Islamic History and Civilization

Islam, the Ancient Near East and Varieties of Godlessness
Islamic History
and Civilization
studies and texts
Editorial Board
Hinrich Biesterfeldt
Sebastian Günther
Honorary Editor
Wadad Kadi
volume 131
The titles published in this series are listed at brill.com/ihc

Patricia Crone, 29 January 2004
photograph by cliff moore/institute for advanced study,
princeton, nj
Islam, the Ancient Near East
and Varieties of Godlessness
Collected Studies in Three Volumes
volume 3
By
Patricia Crone
Edited by
Hanna Siurua
leiden | boston
Cover illustration: Iraq. Kifl. Native Moslem [i.e., Muslim] village with a Jewish shrine to the prophet Ezekiel.
Conical tower over Ezekiel’s tomb. Photographed by the American Colony (Jerusalem), Photo Dept. Date
Created/Published: [1932]. lc-m33-4735 [p&p]. Library of Congress Prints and Photographs Division
Washington, d.c. 20540 usa.
The Library of Congress Cataloging-in-Publication Data is available online at http://catalog.loc.gov

LC record available at http://lccn.loc.gov/2016010221
Want or need Open Access? Brill Open offers you the choice to make your research freely accessible online
in exchange for a publication charge. Review your various options on brill.com/brill-open.
Typeface for the Latin, Greek, and Cyrillic scripts: “Brill”. See and download: brill.com/brill-typeface.
issn 0929-2403
isbn 978-90-04-31927-1 (hardback, volume 131)
isbn 978-90-04-31976-9 (hardback, set, volumes 129, 130 & 131)
isbn 978-90-04-31931-8 (e-book)
Copyright 2016 by Koninklijke Brill nv, Leiden, The Netherlands.

Koninklijke Brill nv incorporates the imprints Brill, Brill Nijhoff, Global Oriental and Hotei Publishing.
All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system,
or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise,
without prior written permission from the publisher.
Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill nv provided
that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive,
Suite 910, Danvers, ma 01923, usa. Fees are subject to change.
This book is printed on acid-free paper and produced in a sustainable manner.

Contents
Editor’s Preface ix
Remarks on Receipt of the 2014 Middle East Medievalists (mem)
Lifetime Achievement Award xi
List of Original Publications and Acknowledgments xvi
1 “Barefoot and Naked”: What Did the Bedouin of the Arab Conquests
Look Like? 1
2 The Ancient Near East and Islam: The Case of Lot-Casting 17
3 Idrīs, Atraḫasīs and al-Khiḍr 44
4 Abū Saʿīd al-Ḥaḍrī and the Punishment of Unbelievers 82
5 The Dahrīs According to al-Jāḥiẓ 96
6 Ungodly Cosmologies 118
7 Post-Colonialism in Tenth-Century Islam 151
8 What Are Prophets For? The Social Utility of Religion in Medieval

Islamic Thought 186
9 Oral Transmission of Subversive Ideas from the Islamic World to

Europe: The Case of the Three Impostors 200
10 How the Field Has Changed in My Lifetime 239
List of Patricia Crone’s Publications 247

Index of Names and Terms 255
Editor’s Preface
The origins of this collection of studies lie in Patricia Crone’s February 2013
visit to Leiden, where she received an honorary doctorate from Leiden Univer-
sity and gave a lecture on how the field of Islamic studies had changed over
her lifetime. Subsequent discussions between her and Petra Sijpesteijn over
the possible publication of that lecture grew into the idea of compiling a col-
lection of her recent, forthcoming and unpublished articles. Professor Crone
herself selected, arranged and in some cases revised the articles to be included
in the collection. Most of the articles are reprinted, but a few are published
for the first time in this collection; these include articles 14 and 15 in volume 1
and articles 3, 8, 9 and 10 (the lecture mentioned above) in the present vol-
ume.
Each volume focuses on a particular theme. The first volume brings together
studies on the community from which Muḥammad emerged and the book that
he brought; the second volume is dedicated to Iranian religious trends both
before and after the arrival of Islam; and this third volume treats Islam in the
historical context of the ancient Near East, with special attention to material-
ists, sceptics and other ‘godless’ people. Each volume includes a bibliography
of Professor Crone’s publications.
All of the articles have been typeset anew, but the page numbers of the orig-
inal publications (wherever available) are indicated in the margin. Where note
numbering has changed in the reprint as a consequence of revisions, the origi-
nal note numbers are given in superscript at the beginning of the affected notes.
I have edited the articles with a very light hand. Errors and misprints have
been corrected, the author’s revisions and additions have been incorporated,
incomplete and previously forthcoming citations have been updated and the
transliteration of Arabic and Persian has been standardised to follow the Arabic
transliteration scheme of the International Journal of Middle East Studies (mod-
ified in the case of elisions). The few editorial interventions beyond these are
bracketed and marked as mine (‘Ed.’). Citation, punctuation and spelling prac-
tices in each article reflect those of the original publication, with only minor,
silent changes.
I would like to thank Sabine Schmidtke, María Mercedes Tuya and Casey
Westerman at the Institute for Advanced Study; Kathy van Vliet, Teddi Dols
and Arthur Westerhof at Brill; Ahmed El Shamsy, Itamar Francez, M. Şükrü
Hanioğlu, Masoud Jafari Jazi, Martin Mulsow, Bilal Orfali, Petra Sijpesteijn and
Frank Stewart for help with queries; Mariam Sheibani for research assistance;
Dana E. Lee for her editorial work; and especially Michael Cook, Professor
x editor’s preface
Crone’s literary executor, who oversaw the finalising of the volumes once Pro-
fessor Crone was no longer able to fill that role herself.
Hanna Siurua
Chicago, January 2016
Remarks on Receipt of the 2014 Middle East iii
Medievalists (mem) Lifetime Achievement Award
When I discussed with Matthew [Ed.: Gordon, then president of mem] what I
should talk about, he said he’d like to hear some manner of reflection on my
work, career, books, students, and the state of the field, or some combination
of these things. Well, I doubt that I shall be able to talk about all these things,
but let me start by telling you a story.
One summer towards the end of my time at school, one of my sisters and I
went to the theatre festival at Avignon, and there for the first time in my life,
I met a live Muslim, a Moroccan. I had decided to study the Muslim world
without ever knowingly having set eyes on an Arab or Persian or heard Arabic
or Persian spoken. There weren’t any of them in Denmark back then: it was
Gilgamesh who had seduced me. I discovered him in my teens and wanted to
be an ancient Near Eastern archaeologist, but for a variety of reasons I became
an Islamicist instead. Anyway, I met this Moroccan in Avignon, and he told
me the story of the Battle of Siffin: the Syrians were losing and responded by
hoisting Qurans on their lances, the battle stopped, and so Ali lost. It never
occurred to me to believe it; I smiled politely and thought to myself, “when
I get to university I’ll hear a different story.” I got to Copenhagen University,
but no Islamic history was taught there, only Semitic philology, which I did not
want to do, and history, meaning European history, which I did do and enjoyed,
but which was not where I wanted to stay. Eventually I got myself to England,
and there I was accepted by soas and heard Professor Lewis lecture on early
Islamic history, including the Battle of Siffin. He told the story exactly as my
Moroccan friend had told it. I could not believe it. It struck me as obvious that
the narrative was fiction, | and besides, everyone knows that battle accounts iv
are most unlikely to be reliable, least of all when they are told by the loser. I
thought about it again many years later, in 2003, when one of Saddam Hussain’s
generals, Muhammad Saeed al-Sahhaf, also known as comical (not chemical)
Ali, persistently asserted that the Iraqis had defeated the Americans and put
them to flight, so that there weren’t any American troops in Iraq any more.
At the very least one would have expected Lewis to say something about the
problematic nature of battle narratives, and was this really true? But no: it was
a truth universally acknowledged that, during the Battle of Siffin, the Syrians
hoisted Qurans on their lances and thereby stopped the battle, depriving the
Iraqis of their victory.
I think this is the biggest academic shock I’ve ever suffered, but I didn’t say
xii remarks on receipt of the 2014 mem lifetime achievement award
anything. I never did, I was too shy. And then I encountered John Wansbrough.
He read Arabic texts with us undergraduates, clearly thinking we were a hope-
less lot, but he was the first person I met at soas who doubted the Siffin story.
As it turned out, he doubted just about everything in the tradition. I was fas-
cinated by him. I wanted to know how he thought we should go about writing
about early Islamic history, so I continued reading texts with him as a graduate,
but I never got an answer. Once, when we were reading Tabari’s account of Ibn
al-Ashʿath’s revolt in the mid-Umayyad period, Wansbrough asked: “What year
are we in?” I thought he simply meant “what year has Tabari put this in?,” but
when I replied “year 82,” or whatever, he acidly retorted, “I see you have the con-
fidence of your supervisor,” meaning Bernard Lewis, my supervisor, whom he
deeply disliked. I think his question was meant to be understood as, “Is all this
really something that happened in year 82 (or whenever) or is it stereotyped
battle scenes interspersed with poetry that could be put in any heroic account
in need of amplification?” I don’t know, for he did not explain. He never did. He
was an imam samit.
From all this you can see two things. First, it was not exposure to Wansbrough
that made me a sceptic or radical or whatever else they like to call me. I was
a sceptic already in Avignon, years before I came to England, without being
aware of it. In my own understanding I was just thinking commonsense. And
secondly, Islamic history was not studied at an advanced level. I don’t know
how the Battle of Siffin is taught these days, but I cannot imagine it is done
with the credulity of those days and, at least in England, Lewis must take part
of the credit for this, for he was very keen for Islamicists to become historians.
After I’d finished my thesis, Michael Cook and I finished Hagarism (1977),
which I assume you have heard about and don’t propose to talk about; and
next, in between some articles, I wrote Slaves on Horses (1980), which was the
first third of my thesis, drastically rewritten. Then it was Roman, Provincial
and Islamic Law (1987), which was a drastically rewritten version of my thesis
part two and which I loved researching because the literature on the Greek,
Roman and provincial side was so superb. The legal learning possessed by these
late nineteenth- and early twentieth-century German and Italian scholars was
incredible, and on top of that they were wonderfully intelligent and lucid. Then
came the First World War and now it is all gone. Apparently it isn’t even done
v to admire them any more. A perfectly friendly | reviewer of my book on law
cautioned his readers that I was an admirer of these scholars, as if it were self-
evident that they were bad people. I don’t see why.
In any case, Meccan Trade came out in the same year. It was delayed by a
report so negative that I withdrew it and sent it to Princeton University Press.
The author of the negative report said that I should have my head examined,
remarks on receipt of the 2014 mem lifetime achievement award xiii
that nothing I’d written would win general acceptance and that I’d never get a
job in America. This last was particularly hilarious since it had never occurred
to me to apply for one there. Serjeant was also outraged by Meccan Trade. He
wrote a furious review in which he accused me of all sorts of misdeeds. But
today the book is perceived as being about the location of Mecca, to which I
devote a page. I’ve even heard somebody introduce me as a speaker and list
Meccan Trade among my books with the comment that it is about the location
of Mecca, to which I had to say sorry, no, actually Meccan Trade is about Meccan
trade.
After Meccan Trade, or at the same time (both this and other books took a
long time to reach print), I published God’s Caliph with Martin Hinds. It was a
short book, but Calder nonetheless thought it was long-winded: I admit I found
that hard to take seriously. It was as usual: the reviewers found fault with this,
that and the other, and you let it pass. The one thing I really disliked about God’s
Caliph was the massive number of misprints, which Martin Hinds was no better
at spotting than I was.
It must have been after God’s Caliph had gone to press that I wrote Pre-
Industrial Societies, which I hugely enjoyed doing because I had to read about
all kinds of places that I didn’t know much about, and also because I wrote
without footnotes. It saves you masses of time. pis, as I called it (pronouncing
it Piss), was barely reviewed and took a while to gather attention, and it too was
riddled with misprints, but the misprints should now have been eliminated and
a fresh print-run with a new cover is on its way.
The next book I wrote was The Book of Strangers: Medieval Arabic Graffiti on
the Theme of Nostalgia (1999), which was completely new to me when I started
translating it. I inherited it from Martin Hinds and was captivated by it, but
had trouble with the poetry in it. However, Shmuel Moreh came to Cambridge
shortly after I’d started, and he was well versed in Arabic poetry, so I asked
him if he’d help me, and he would. So we translated it together and I took
responsibility for the rest.
That book almost generated another Siffin story. The author is traditionally
identified as Abu ʾl-Faraj al-Isfahani, but he himself says that he was in his youth
in 356/967, which makes him considerably younger than Abu ʾl-Faraj.1 Yaqut,
who said he did not know how to resolve the problem, noticed this already.
There is only one way to resolve it: the author is not Abu ʾl-Faraj. The book
doesn’t have much in common with Abu ʾl-Faraj’s works either. But a specialist
in Abu ʾl-Faraj insisted that it was him and came up with the explanation, also
1 Abu ʾl-Faraj allegedly died in 356/967 [Ed.: noted by Antoine Borrut for mem].
xiv remarks on receipt of the 2014 mem lifetime achievement award
tried by older scholars, that Abu ʾl-Faraj was senile when he wrote the book, so
that he had forgotten when he was young. Honestly, the things that Islamicists
will say!
The next book was also a joint project and also connected with Martin Hinds
vi and the so-called “Hinds-Xerox” which Martin | had received from Amr Khalifa
Ennami and which Michael Cook used for his section on the Murjiʾa in his Early
Muslim Dogma. Martin Hinds was working on the last section of the manuscript
when he died. I could have finished that last section, but it seemed a bad idea
to translate yet another fragment. What should be done was a translation of
the whole epistle. But I couldn’t do that on my own—there were parts of the
manuscript that I simply could not decipher. So I asked my former colleague in
Oxford, Fritz Zimmermann, if he would participate, and thank God, he would.
So we started by writing a translation each and then amalgamating them, with
long pauses over passages that seemed impossible. Fritz had some great brain
waves, and somehow we managed to get a complete typescript together. Then
there was all the rest, where the fun for me lay in comparing Salim and the Ibadi
epistles that I had been able to buy in Oman. The Epistle of Salim ibn Dhakwan
was published in Oxford in 2001. Very few people are interested in the Ibadis so
it has not exactly been a bestseller, but I learned an extraordinary amount from
writing it.
After that, I wrote Medieval Islamic Political Thought, which the Americans
called God’s Rule, though it is disagreeably close to God’s Caliph and not par-
ticularly apt in my view. That book started as exam questions in Cambridge.
Carole Hillenbrand was our external examiner, and when she saw the ques-
tions, she asked me if I wanted to write a volume on political thought for her
Edinburgh series. I liked the idea, envisaging the book as much smaller than it
actually became. I also thought I could do it fast because I thought I knew the
field inside out, but that was only true of some of the subjects I wrote about. I
had to do a lot of work on the Ismailis, for example, because I did not know the
sources well enough. I was also acutely aware of having inadequate knowledge
of the last century before the Mongol invasions and don’t think I managed to get
that right. I suppose I was running out of patience. I wasn’t under any pressure,
for I had refused a contract. I usually did until I was close to the end.
My book on political thought was the first book of mine that was uniformly
well received. All the others had a controversial element to them that the
reviewers didn’t like, if only for my refusal to accept that Abu ʾl-Faraj al-Isfahani
had forgotten when he was young. Mercifully, there were also reviewers who
found that a ridiculous argument. Not long afterwards they gave me the Levi
della Vida medal and I also received several honorary doctorates. Altogether, it
was clear that I was no longer an enfant terrible.
remarks on receipt of the 2014 mem lifetime achievement award xv
My latest, and probably also last, book is The Nativist Prophets of Early
Islamic Iran: Rural Revolt and Local Zoroastrianism (2012), which had its roots
in my teaching in Oxford and which was very exciting to write because it was
about villagers, whom we rarely see in the sources, and because their form of
Zoroastrianism was quite different from that of the Pahlavi books. That book
was also well received; it was awarded no less than four book prizes, for its
contribution to Islamic studies, to Iranian studies, to Central Asian studies, and
to historical studies in general.
If I had not fallen ill, I would have started a book on the Dahris, Godless
people on whom I have written some articles, and who are certainly worth a
book. But I don’t think I have enough time.
Patricia Crone
Princeton, November 2014
List of Original Publications and Acknowledgments
We gratefully acknowledge permission to reprint articles that originally ap-

peared in the following publications:
‘Remarks by the Recipient of the 2014 mem Lifetime Achievement Award Writ-
ten for the Annual Meeting of Middle East Medievalists and Read in Absentia by
Matthew S. Gordon (November 22, 2014, Washington, d.c.)’, al-ʿUṣūr al-Wusṭā
23 (2015), iii–vi. Reprinted with permission.
1. ‘“Barefoot and Naked”: What Did the Bedouin of the Arab Conquests Look
Like?’, Muqarnas 25 (‘Frontiers of Islamic Art and Architecture: Essays in Cel-
ebration of Oleg Grabar’s Eightieth Birthday’, ed. Gülru Necipoğlu and Julia
Bailey), 2008, 1–10. © Brill.
2. ‘The Ancient Near East and Islam: The Case of Lot-Casting’ (with Adam Sil-
verstein), Journal of Semitic Studies 55, no. 2 (2010), 423–450. Reprinted by per-
mission of Oxford University Press on behalf of the University of Manchester.
3. ‘Idrīs, Atraḫasīs and al-Khiḍr’. Previously unpublished.
4. ‘Abū Saʿīd al-Ḥaḍrī and the Punishment of Unbelievers’, Jerusalem Studies in

Arabic and Islam 31 (2006), 92–106. Reprinted with permission.
5. ‘The Dahrīs According to al-Jāḥiẓ’, Mélanges de l’Université Saint-Joseph 63

(2010–2011), 63–82. Reprinted with permission.
6. ‘Excursus ii: Ungodly Cosmologies’, in Sabine Schmidtke (ed.), Oxford Hand-

book of Islamic Theology, Oxford: Oxford University Press, 2016. Reprinted by
permission of Oxford University Press.
7. ‘Post-Colonialism in Tenth-Century Islam’, Der Islam 83, no. 1 (2006), 2–38.

Reprinted with permission.
8. ‘What Are Prophets For? The Social Utility of Religion in Medieval Islamic
Thought’. Previously unpublished.
list of original publications and acknowledgments xvii
9. ‘Oral Transmission of Subversive Ideas from the Islamic World to Europe: The
Case of the Three Impostors’. Previously unpublished.
10. ‘How the Field Has Changed in My Lifetime’. Previously unpublished.

chapter 1
“Barefoot and Naked”: What Did the Bedouin of the 1
Arab Conquests Look Like?*
The Syriac churchman Bar Penkaye, who wrote about 690, held the Arab in-
vaders to have been “naked men riding without armor or shield.”1 In the same
vein Michael the Syrian (d. 1199) reports that a certain Hiran sent by the last
Sasanid emperor to spy on the Arabs told his employer that the invaders were “a
barefoot people, naked and weak, but very brave.”2 A Muslim text dating from,
perhaps, the later eighth century similarly insists that the invaders were “bare-
foot and naked, without equipment, strength, weapons, or provisions.”3 In all
three texts the word “naked” seems to be used in the sense of poorly equipped
and lacking body armor rather than devoid of clothes, and all three depict the
Arabs as poorly equipped in order to highlight the extraordinary, God-assisted
nature of the Arab conquests. “I have a sharp arrowhead that penetrates iron,
but it is no use against the naked,” as Rustam says in the Shāhnāma, in his pre-
monition of the fall of the Sasanids.4 But precisely what did the Arab invaders
wear? It would be the first question to spring to Oleg Grabar’s mind. Under
normal circumstances it would be the last to spring to mine, for as Oleg is
fond of telling his colleagues, historians tend to ignore the concrete physical
* I should like to thank Michael Macdonald for invaluable help with images, inscriptions, and
bibliographical references alike. Insofar as this article has any merit, it is really due to him.
(The same most definitely does not apply to the shortcomings.) I am also grateful to Mika
Natiff for teaching me to navigate the Index of Christian Art, to Michael Cook for reading and
commenting on the paper, and to Julia Bailey for spotting visual clues that I had overlooked.
1 Bar Penkaye in A. Mingana (ed. and tr.), Sources syriaques (Leipzig, n.d. [1907?]), 141; trans. in
S.P. Brock, “North Mesopotamia in the Late Seventh Century,” Jerusalem Studies in Arabic and
Islam 9 (1987): 58.
2 Michael the Syrian, Chronique, ed. and tr. J.-B. Chabot, 4 vols. (Paris, 1899–1910), 4:417, 2:421.
3 D. Sourdel, “Un pamphlet musulman anonyme d’ époque ʿabbāside contre les chrétiens,”
Revue des études islamiques 34 (1966): 33 (text), 26 (trans.). For a reconstruction of the
text from which the fragment comes see J.-M. Gaudeul, “The Correspondence between Leo
and ʿUmar,” Islamochristiana 10 (1984): 109–157, with the passage in question on 155. The
transmitter is Ismāʿīl b. ʿAyyāsh.
4 Firdawsī, Shāhnāma, ed. E.E. Bertels, 9 vols. (Moscow, 1960–1971), 9:1.119 (drawn to my atten-
tion by Masoud Jafari).
© koninklijke brill nv, leiden, 2016 | doi: 10.1163/9789004319318_002

2 chapter 1
manifestation of things; in particular, they do not think of the way things looked
and so miss an important dimension of the past. I have always pleaded guilty to
that charge. Having benefited from Oleg’s lively company and warm heart for
over ten years, however, I shall now try to make amends, if only with a trifling
offering: how should we tell a filmmaker who wanted to screen the story of
the Arab conquests to depict the conquerors? More precisely, how should we
tell him to depict the desert Arabs who participated in the conquests? For the
bedouin will not have been dressed in the same way as the settled Arabs, and I
should like to keep things simple.
Most of us would probably reply that the hypothetical filmmaker should
depict the bedouin warriors as men in kaffiyehs and flowing robes, along the
lines familiar from Lawrence of Arabia and countless Hollywood films; but as
far as the bedouin of pre-Islamic Arabia are concerned, it would seem that
we are wrong. Though “naked” may be a little hyperbolic, both literary and
iconographic evidence suggests that it is not far from the truth.
To start with the literary evidence, Ammianus Marcellinus, commander of
the eastern armies about 350 ad, tells us that the Arabs of the Syrian desert
were “warriors of equal rank, half nude, clad in dyed cloaks as far as the loins.”5
The word he uses for their cloaks is sagulum, a short, military tunic, and one
wonders how literally one should take him: were they wearing Roman army
issue, passed down from relatives and friends who had served in the Roman
army, or alternatively stolen from unlucky soldiers? (“When bedouin raiders in
the desert encountered someone from the settled areas, it was their custom to
accost him with the command, Ishlaḥ yā walad, ‘Strip, boy!’ meaning that they
intended to rob him of his clothing,” as Jabbur says of the Syrian bedouin many
centuries later.)6 Ammianus does not tell us what, if anything, the warriors
wore on their heads, but of another Arab, this time one in Roman service at
Adrianople, he says that he was long-haired and naked except for a loincloth.7
In the same vein Malka, a fourth-century Syrian who was captured by bedouin
5 Ammianus Marcellinus, xiv, 4, 3; quoted in J.B. Segal, “Arabs in Syriac Literature before the Rise
of Islam,” Jerusalem Studies in Arabic and Islam 4 (1984): 102; also discussed in J. Matthews, The
Roman Empire of Ammianus (London, 1989), 344, 347–348.
6 Jibrāʾīl Sulaymān Jabbūr, The Bedouins and the Desert: Aspects of Nomadic Life in the Arab East,
trans. L.I. Conrad (Albany, 1995; Arabic original 1988), 1n, with vivid illustrations on 2–3. For
other examples of robbers commanding people to strip naked, see Jacob of Saroug in Khalīl
Alwān, Quatre homélies métriques sur la création (Louvain, 1989), 43; A. Christensen, Contes
persans en langue populaire (Copenhagen, 1918), nos. 9, 33, 42.
7 Matthews, Roman Empire of Ammianus, 348, with reference to Ammianus, xxxi, 16, 6.
“barefoot and naked” 3
figure 1 Ivory carving, right arm of the Chair of Maximianus. Museo Arcivescovile, Ravenna.
photo: alinari/art resource, ny
between Aleppo and Edessa and whose adventures were recorded by Jerome,
describes how the Ishmaelites descended upon his party of about seventy
travelers “with their long hair flying from under their headbands.” He did not
think of them as wearing turbans or kaffiyehs, then, or as shielding their heads
from the sun by any kind of head cover at all. Like Ammianus, he says that
they wore cloaks over their “half-naked bodies,” but he adds that they wore
broad military boots (caligae).8 Again one wonders if they were wearing Roman
army issue. They transported Malka into the desert and set him to work as a
shepherd, and there he “learned to go naked,” he says, presumably meaning
that he learned | to cover himself with a mere skin: this seems to have been all 2
that slaves wore in pre-Islamic Arabia.9 One would infer that he had handed
over his clothes to his captors.
We now turn to the iconographic evidence, looking at it region by region.
Syria
To start in Syria, there is a representation of semi-naked bedouin in an ivory

carving from a chair made in the first half of the sixth century in Antioch
8 Jerome, “Vita Malchi Monachi Captivi,” paragraphs 4–5, in J.-P. Migne, Patrologiae Cursus
Completus, Series Latina, 221 vols. (Paris, 1844–1864), 23: cols. 57–58, trans. in Segal, “Arabs
in Syriac Literature,” 103; cf. I. Shahid, Byzantium and the Arabs in the Fourth Century
(Washington, dc, 1984), 284 ff.; Matthews, Roman Empire of Ammianus, 348.
9 G. Jacob, Altarabisches Beduinenleben (Berlin, 1897), 44 (with reference to ʿAntara’s Muʿal-
laqa).
4 chapter 1
or (under Syro-Palestinian influence) Alexandria (fig. 1).10 It depicts Joseph’s

brothers selling Joseph to two Saracens: the brothers are represented by the
three figures on the left, Joseph stands in the middle, and two Saracens appear
with two camels behind them to the right. The Saracens, who are armed with
a bow and a spear respectively, have long, apparently plaited hair and wear
nothing on their heads or their upper torsos, merely loose garments wrapped
around their waists, which reach as far as their ankles but expose one of their
legs as they walk. The brothers are also scantily clad, but in more military-
looking outfits, and it is they rather than the Saracens who are wearing boots.
The Saracens are shod in sandals. There is of course no guarantee that the
carving is based on observation rather than artistic convention, but one point
is clear: it was not as heavily clad figures in the style of Lawrence of Arabia that
bedouin were envisaged in sixth-century Syria.
Another ivory carving on the same chair shows the Saracens selling Joseph
to Potiphar (fig. 2). Here Joseph is seen twice, first on a camel (on the left)
and next between Potiphar and one of the Saracens, to whom she is handing
money. Potiphar is wearing classical-looking robes. The Saracens’ robes also
appear more flowing than in the first panel, but here as there their lower body
wraps are split in the middle, exposing their legs, and their arms are bare. In
fact, their entire upper torsos could be bare, though it is hard to tell. The short
tunic that Joseph is wearing clearly includes a drape over one shoulder, and the
adult Saracens could have a similar item on their shoulders.11 Maybe the artist
dressed his characters in classical clothes in order to conjure up a bygone age.
In any case, he depicted the Saracens with the same long, apparently plaited
hair as in the first panel, and he gave them sandals, too, but not any kind of
headgear. One would take it to have been long hair of this kind that Malka saw
flowing under headbands.
Yet another sixth-century carving, also a Syrian or Syro-Egyptian work, de-
picts two brothers selling Joseph to a Saracen.12 Joseph and his brothers are
wearing short tunics similar to those in which rural people are depicted on the
mosaic floors of sixth-century churches in Madaba.13 The Saracen is wearing
10 See O.M. Dalton, East Christian Art: A Survey of the Monuments (Oxford, 1925), 172, 205ff.;
idem, Byzantine Art and Archaeology (New York, 1961; orig. publ. 1911), 203ff.
11 Cf. Dalton, Byzantine Art and Archaeology, 206.
12 Berlin, Staatliche Museen, inv. no. 566; cf. Dalton, Byzantine Art and Archaeology, 208;
W.F. Volbach, Elfenbeinarbeiten der Spätantike und des frühen Mittelalters (Mainz, 1952),
80–81; pl. 54, no. 172.
13 M. Piccirillo and E. Alliata, Mount Nebo: New Archaeological Excavations 1967–1997 (Jeru-
figure 2 Ivory carving, right arm of the Chair of Maximianus. Museo Arcivescovile, Ravenna.
photo: alinari/art resource, ny
a mantle that leaves the left part of his chest exposed, but what he is wearing
underneath | is not clear. All four are barefoot and bareheaded. 3
Finally, we have the depiction a man armed with a bow, sword, and whip,
leading a camel (fig. 3); this appears on the mosaic floor of the church of
the monastery of Kayanos at ʿUyun Musa, at the eastern top of the Dead Sea,
dated by Piccirillo to the second half of the sixth century.14 In Piccirillo’s words,
the man “is half naked, wearing a long loincloth reaching beneath his knees
with a cloak thrown over his left shoulder that covers his forearm.” Piccirillo
salem, 1998), 333 (Church of the Deacon Thomas, whole floor); 337 (Stephanos spearing a
lion, wearing “a sleeveless orbiculated tunic … tied to the right shoulder” that seems to be
identical with that of the brother on the left); 338–339 (donkey driver, soldier defending
himself against a bear); 343 (date); 345, 347 (Church of Saints Lot and Procopius, whole
floor).
14 M. Piccirillo, Madaba, le chiese e i mosaici, 207–208; Piccirillo and Alliata, Mount Nebo,
356–358, with a better photo (fig. 224).
6 chapter 1
figure 3 Mosaic from the church of Kaianos at ʿUyun Musa, Mount Nebo
photo courtesy of michele piccirillo
suggests that he was an auxiliary soldier and deems the representation to fit
the “exaggeratedly dramatic” literary accounts of Arab soldiers given by authors
such as Ammianus Marcellinus and Malka in Jerome.15 Whether the Arab was
an auxiliary soldier or not, however, the representation actually seems to be
quite different. The most dramatic feature of the mosaic is the Arab’s bulging
chest. Neither Ammianus nor Jerome says anything about chests, but both
highlight the long, flowing hair of the Arabs; though damage to the mosaic
makes it impossible to say what, if anything, the soldier is wearing on his head,
it is at least clear that he does not have hair (or a kaffiyeh) coming down to his
shoulders. The clothes involved are quite different, too. Ammianus’ Arabs were
wearing short military tunics, Jerome’s were dressed in cloaks and boots, but
the soldier in the mosaic is wearing a waist wrap and shawl along with sandals.
This could well be based on observation, for the waist wrap and shawl (īzār and
ridāʾ) are the two chief items of male clothing in pre-Islamic poetry.16 The main
feature that the three representations have in common is the skimpiness of the
outfits described. Pitched against a horsemen encased in iron, Arabs such as
4 these would indeed have come across as naked. |
In sharp contrast to these representations, an image on a piece of Coptic
tapestry dating from between the sixth and eighth centuries and said to show
15 Piccirillo, in both Madaba, le chiese e i mosaici and Mount Nebo, and with reference to
Ammianus and Jerome in Madaba, le chiese e i mosaici, 225 n. 10.
16 Jacob, Altarabisches Beduinenleben, 44.
Joseph and an Ishmaelite merchant on a camel depicts both Joseph and the
Ishmaelite as thoroughly wrapped up.17 But the alleged camel may well be a
horse,18 and the alleged Ishmaelite seems to be wearing trousers. So this can be
left out of consideration.
South Arabia
If the inhabitants of the Roman empire envisaged the Saracens as wearing

nothing on their heads and not much on their bodies, how were they seen by
the Arabs themselves? We may start in the south.
Here the first image to capture one’s attention is a crude relief on an alabaster
incense burner from Shabwa in the Hadramawt, probably dating from around
the third century ad (fig. 4). It depicts a man riding on an unsaddled camel,
positioned in front of the hump; he holds a short sword or a camel stick or
some such implement in his right hand and the reins in his left, and a water
skin or shield is attached by a strap to the rear of the hump. He is stark naked,
and, apart from the reins, the camel is as naked as he is.
The text gives the name of the person commemorated, presumably identical
with the person represented, as Adhlal ibn Wahabʾil but does not otherwise
tells us anything about him.19 Macdonald wonders whether the incense burner
is a funerary object rather than a dedicatory one (as suggested in the catalogue
of the exhibition in which it was most recently displayed20), for the inscription
does not mention any deity, only a name and a patronym, and the vast majority
of funerary stelae in both North and South Arabia only give the deceased’s
name and patronym. If the object is funerary, the relief might in Macdonald’s
opinion represent the naked soul of the deceased riding his camel on the Day
17 A. Kakovkine, “Le tissu copte des viie–viiie siècles du Musée Metropolitan,” Göttinger
Miszellen 129 (1992): 53–59. It was formerly classified as showing the flight into Egypt.
18 Presumably it was classified as a camel on the basis of its peculiar head (which mostly
looks like that of a dog) and the similarity of its hooves and tail to those of the camel at
Dura Europos (cf. the reference given below, n. 24). But it has no hump, and its legs and
harness are those of a horse.
19 St. J. Simpson, ed., Queen of Sheba: Treasures from Ancient Yemen (London, 2002), 97–98,
no. 110; also in W. Seipel (ed.), Jemen: Kunst und Archäologie im Land der Königin von
Sabaʾ (Vienna, 1998), 86 and 88, no. 20, both without comments on the absence of clothes;
Répertoire d’ épigraphie sémitique, 8 vols. (Paris, 1900–1968), 7, no. 4690.
20 Simpson, Queen of Sheba, 97–98, no. 110.
8 chapter 1
figure 4 Relief from an alabaster incense burner. British Museum,

ane 125682.
reproduced with the permission of the trustees
of the british museum
of Judgment.21 But as Macdonald himself stresses, this is highly conjectural.

Besides, did the pagans of South Arabia believe in the resurrection? There is
nothing to suggest that the deceased was a Jew or a Christian. And the people
depicted on other funerary reliefs are fully clothed. On the whole, it seems more
likely that a bedouin of the Hadrami plateau is being depicted here, for there
are plenty of naked Arabs in the rock reliefs, as will be seen. Why such a man
should figure on a Shabwan incense burner is another question.
A fully clothed camel rider appears on a funerary relief, also of alabaster,
dated to roughly the first | to the third century ad, with an incription identifying 5
the deceased as Mushayqar Hamayat ibn Yashuf (fig. 5).22 He too is holding a
short spear or camel stick in his right hand and the reins in his left, and he is
sitting on a fine camel saddle of a type also attested on a bronze figurine of
a camel thought to be from Yemen.23 Unlike the wild bedouin on the incense
burner, this camel rider was presumably a soldier in the local army, dressed
in conformity with the sense of propriety of the settled people. Of decently
dressed camel-riders, presumably soldiers in the local armies, we also have
an example in a relief from Dura Europos that shows such a rider seated on
a saddled camel, armed with a long lance, and wearing a tunic and mantle.24
But he is bareheaded, and maybe the South Arabian was too: Calvet and Robin
interpret his apparent head cover as a hair style.25
In another funerary relief, a Sabaean alabaster of the second or third century
ad, the lower panel shows a horseman with the north Arabian name of ʿIjl
ibn Saʿdallat touching a camel with his spear, the act by which a camel raider
appropriates a camel. The upper panel shows the deceased sitting at a table
with his wife and child in attendance, or perhaps the deceased at a banquet,
and both the stool and the table indicate that we are in a settled environment,
as also suggested by the fact that the nisba of the deceased was Qryn: he may
have come from Qaryat al-Faw or from Wadi ʾl-Qura.26 He was not a bedouin
21 Michael Macdonald, personal communication with reference to a discussion at the recent

congress “Rencontres sabéennes 10,” in St. Petersburg.
22 Y. Calvet and C. Robin, Arabie heureuse, Arabie déserte: Les antiquités arabiques du Musée
du Louvre (Paris, 1997), 109–110, no. 20, where both the image and the text are reproduced
along with a transliteration, translation, discussion, and bibliography.
23 Reproduced in Simpson, Queen of Sheba, 99, no. 113.
24 A. Perkins, The Art of Dura-Europos (Oxford, 1973), fig. 40.
25 Cf. the reference given above, n. 22 (“Il porte une coiffure arrondée avec une sorte de
pendant à l’ arrière”).
26 Louvre, ao 1029: see Calvet and Robin, Arabie heureuse, 107–108, no. 18 (image, text,
transliteration, translation, discussion, and bibliography); A. Caubert, Aux sources du
10 chapter 1
figure 5 Alabaster funerary relief. Louvre, ao 1128.

photo courtesy of michael macdonald
monde arabe: L’Arabie avant l’ Islam, collections du Musée du Louvre (Paris, 1990), 28 and
39, no. 3 (where the upper panel is interpreted as a banquet scene). For the meaning of
the gesture with the spear see M.C.A. Macdonald, “Camel Hunting or Camel Raiding?,”
Arabian Archaeology and Epigraphy 1, no. 1 (1990): 24–28, with a reproduction of the stela
on 26.
raiding camels, then, but rather a sedentary Arab engaged in what one would
assume to be camel catching staged as a sport.27 All the figures are fully clothed,
the deceased in a long robe and the other two in shorter garments, and the
deceased seems to be wearing some kind of head cover, though his putative
wife and children are clearly bareheaded. The deceased’s headgear, if it is not
simply hair, looks like some sort of stiff bonnet, certainly not like a turban.
South Arabian reliefs, which usually show people bareheaded, do not in fact
seem to depict any turbans at all.
Moving slightly north to Qaryat al-Faw, which flourished from roughly the
second century bc to roughly the fifth century ad, we find a bronze statue of
a man wearing nothing but a loincloth, but he is kneeling reverently, presum-
ably in prayer, and his outfit is more likely to be a form of iḥrām than bedouin
dress.28 Also at Qaryat al-Faw we find two drawings on plaster walls of horse-
men hunting or raiding camels. One horseman could be naked, but the other is
wearing something like a tunic or at least a skirt. Whether they have headgear
is impossible to tell.29
The Desert
That leaves us with the countless rock drawings left by the inhabitants of the
desert themselves. The most striking image among these is a drawing of a horse-
| man hunting an oryx with a short spear (fig. 6). He is wearing a waist wrap 6
similar to that of the Arab soldier in the sixth-century mosaic; the thickened
lines across his shoulders could be taken to suggest that he is also wearing a
ridāʾ, and he has bushy or kinky hair that, although quite long, sticks straight
out from his head, in a style that is quite common in Safaitic drawings.30 Unless
27 This seems at least as likely as that the deceased should be shown as engaged in camel-
raiding, perhaps out of a desire to claim links with a real or nomadic past, as suggested
by Macdonald, “Camel Hunting or Camel Raiding?,” 25–26; idem, “Hunting, Fighting, and
Raiding: The Horse in Pre-Islamic Arabia,” in Furusiyya, ed. D. Alexander, 2 vols. (Riyad,
1996), 1:76. Either is compatible with the conjecture that he was a caravaneer (Calvet and
Robin, Arabie heureuse, 108).
28 A.R. al-Ansary, Qaryat al-Fau: A Portrait of a Pre-Islamic Civilisation in Saudi Arabia (Lon-
don, 1982), 109, no. 3.
29 Ansary, Qaryat al-Fau, 130–133 (where the rider called Salim b. Kaʿb seems to be hunting
rather than raiding, given that the camel appears to have been speared or shot with an
arrow).
30 G.M.H. King, The Basalt Desert Rescue Survey: Safaitic Inscriptions (forthcoming; my
thanks to Dr. King for allowing me to reproduce the image).
12 chapter 1
figure 6 Rock drawing depicting a Safaitic horseman

photo courtesy of g.m.h. king
we take his hair actually to be some sort of hat, he is not wearing anything on
his head. Other drawings do depict headgear, sometimes very elaborate, but
apparently in the form of plumes, which are hardly intended here.31 The author
of the Safaitic inscription on the same stone claims to have made the drawing,
which is thus roughly datable to the period from the first century bc to the
fourth century ad. By then, it would seem, the pre-Islamic “uniform” of īzār
and ridāʾ was in place, but without the turban or other headgear by which it is
usually taken to have been complemented.
By the standards of the rock drawings, this horseman is well dressed, for
most drawings depict males as either naked or wearing skimpy clothes “mainly
meant to cover the private parts,” as Nayeem puts it.32 But these drawings are
difficult to date, and though some are Safaitic,33 many of them are likely to be
much older than the period under consideration here.
31 Cf. Macdonald, “Hunting, Fighting, and Raiding,” 76, 77 fig. 5b, where the upper tier of the
headdress looks like giant feathers.
32 M.A. Nayeem, The Rock Art of Arabia (Hyderabad, 2000), 337. For some striking examples
of naked people see Macdonald, “Hunting, Fighting, and Raiding,” 72, nos. 3, 1d, 1g, 1h.
Unfortunately, these drawings are known only from hand copies, and there is no way of
telling how accurately they represent the originals.
33 M.C.A. Macdonald, “Reflections on the Linguistic Map of Pre-Islamic Arabia,” Arabian
Archaeology and Epigraphy 11 (2000): 45.
figure 7 Rock drawing

There is an example of what the makers of rock art wore in a Thamudic drawing
from the Tabuk region of northern Arabia, which depicts a horseman and two
men in a chariot—a driver and an archer (fig. 7).34 The horseman, who is riding
in front of the chariot, appears to be every bit as naked as the camel on the
Sabaean stela, though one should perhaps envisage him as wearing a loincloth.
He also seems to have long, flowing (rather than bushy) hair. The driver could
be naked, at least as far as his upper torso is concerned (the lower part of
his body is hidden from view), but maybe the draftsman simply refrained
from trying to depict his clothes. He could be bareheaded, but his head is
pointed, perhaps to suggest the conical helmet worn by Assyrian soldiers.35
The footsoldier who is pursuing the chariot and shooting arrows at it, however
(fig. 8), is dressed in a long waist wrap, with a slit at the side or the front to allow
34 Cf. Macdonald, “Hunting, Fighting, and Raiding,” 74, 76ff., with the complete composition
on 224–225.
35 Macdonald, “Hunting, Fighting, and Raiding,” 78.
14 chapter 1
figure 8 Rock drawing, detail

freedom of movement, along the lines of those depicted on the ivory panel of
Saracens buying Joseph from his brothers (see fig. 1). He too seems to have long
hair.
This drawing is likely to be very old. The chariot points to ancient Near
Eastern times, perhaps the seventh to fourth century bc,36 and the footsoldier
has a long, pointed thong between his legs, a feature also found on images of
36 Macdonald, “Hunting, Fighting, and Raiding,” 78; idem, “Wheels in a Land of Camels:
Another Look at the Chariot in Arabia,” Arabian Archaeology and Epigraphy 20 (2009):
156–184.
figure 9 W. Boutcher, detail of an Assyrian relief, Room l of the North Palace at Nineveh.
british museum or. dr. 28. (drawing reproduced with the
permission of the trustees of the british museum)
Arabs on Assyrian reliefs (although precisely what it is meant to represent is

unknown). Indeed, one wonders if the occupants of the chariot should not
actually be identified as Assyrians (or perhaps Babylonians) pursuing one Arab
while being shot at by another.37
The age of the drawing notwithstanding, the clothing and hairstyle of the
Arab archer are not drastically different from those examined above, suggesting
that the desert Arabs dressed in much the same way for over a millennium
before the rise of Islam. In a drawing by W. Boutcher of a detail from the
Assyrian reliefs showing the campaign of Ashurbanipal (688–627 bc) against
the Arabs, the Arabs, with plaited hair, are shown dismounted from their
camels and dressed in wraparounds, each with an opening to allow freedom
of movement (fig. 9). Their wraparounds are not flowing like those of the
37 The main objection to this proposition is that the man on horseback is identified in the
inscription above him as ḥrb, taken by Macdonald to mean enemy warrior on the basis of
modern bedouin dialect. But this is clearly conjectural, and the word may not even have
been correctly deciphered (cf. Macdonald, “Wheels in a Land of Camels,” 175, no. 9).
16 chapter 1
7, 8 Saracens who purchase Joseph from his brothers (fig. 1), | | and their hair
looks shorter and a good deal neater, too, but given that there are more than a
thousand years between the images, the continuity is nonetheless striking. To a
somewhat lesser degree, the same holds true when one compares the Assyrian
representations with the Safaitic rock drawings and the Madaba mosaic.
In sum, what did the bedouin participants in the conquests wear? The
answer seems to be generally not very much at all: either bits and pieces of
what their settled neighbors—whether the latter were Byzantines, Arabians,
or (one assumes) Iranians—wore, or a wraparound and a ridāʾ covering part of
their upper torso, and perhaps even sandals, but rarely, insofar as one can tell,
anything on their heads. It is the absence of headgear that is the most surpris-
ing. Whatever the variations, all the desert dwellers seem to have looked a good
deal more like their ancestors of Assyrian times than like Musil’s Rwala.38 As far
as desert clothing is concerned, Arabia on the eve of Islam seems still to have
been rooted in the ancient Near East.
When and why did the desert Arabs start covering themselves up? I cannot
claim to know. My guess would be that they started doing so in the centuries
after the rise of Islam, and in consequence of the rise of Islam, for Islam drew
the bedouin closer together to the settled people, giving them shared religious
and other norms. Wrapping up was what the people who mattered did, and so
the bedouin came to do so too (at least when they could afford it). According to
Ibn al-Kalbi (d. 819 or later), the Tanukh who met the caliph al-Mahdi (d. 785)
in Qinnasrin were wearing turbans. They were trying to look their best on this
occasion.39 A Byzantine miniature of ca. 976–1025 depicting Simeon Stylites
venerated by Arabs shows Simeon in a hooded monk’s habit and the three
Arabs wearing turbans, now apparently as a matter of course.40 But I had better
leave this question for another birthday.
38 Cf. A. Musil, The Manners and Customs of the Rwala Bedouins (New York, 1928).
39 Shahid, Byzantium and the Arabs in the Fourth Century, 431.
40 Rome, Bibliotheca Vaticana, gr. 1613. The ninth-century miniature of Joseph’s brothers
selling Joseph to a Saracen is uninformative, since no attempt seems to have been made to
distinguish the Saracen from the other figures: all are wearing the same long cloaks and all
are bareheaded (cf. A. Grabar, Les miniatures du Grégoire de Nazianze [Paris, 1943], pl. lxi).
chapter 2
The Ancient Near East and Islam: The Case of 423
Lot-Casting
With Adam Silverstein
In 1993 classical archaeologists made an exciting discovery at Petra. This city,

once the capital of the Nabataean kingdom, thereafter a major town in the
Roman province of Arabia, had long been assumed to have been destroyed in
an earthquake of 551 ce, but this proved to be wrong, and in the church of St
Mary the archaeologists found a cache of papyri. Completely carbonized by
the fire which had destroyed the church in the early seventh century, these
papyri could nonetheless be read by means of sophisticated modern tech-
niques, and an edition is in progress.1 They contained the private archive of
a major family of the city, covering the years from at least 537 to 593 ce. The
papyri are in Greek but reflect a community whose native language appears
to have been Arabic, and among the papyri is a record of a division of an
inheritance between three brothers. The | estate, which consisted of land and 424
buildings, was divided into three equal shares and awarded to the sons by a
procedure which the editors, with reference to a comparable papyrus from
Nessana, take to have been lot-casting.2 The Nessana papyrus, written in 562
ce, also records the division of an estate, here among four sons. The prop-
erty, which consisted of buildings, farmland and personal articles, was divided
into four shares of roughly equal value and awarded to the sons by lot in
the presence of friends and relatives. Here, as at Petra, the parties concluded
1 For all this, see L. Koenen, R.W. Daniel and T. Gagas, ‘Petra in the Sixth Century: the Evidence
of the Carbonized Papyri’, in G. Markoe (ed.), Petra Rediscovered (New York 2003), 250–261;
J. Frösén, A. Arjava and M. Lehtinen (eds), The Petra Papyri, 1 (Amman 2002). Our thanks to
Glen Bowersock for referring us to this literature.
2 Cf. Koenen, Daniel and Gagas, ‘Petra in the Sixth Century’, 251. The papyrus (Inv. 10, P. Petra
Khaled and Suha Shoman) is still unpublished. There is no explicit mention of lots in the
draft edition and translation that Crone has seen, courtesy of her colleague Glen Bowersock,
but the parallels with the Nessana papyrus are certainly striking. [Ed.: The papyrus has now
been published as L. Koenen, J. Kaimio, M. Kaimio and R.W. Daniel (eds), The Petra Papyri, 2
(Amman 2013).]

18 chapter 2
the proceedings by swearing by the Trinity and the Emperor’s health that they
would abide by the division.3
The interest of this discovery to historians of the Near East lies in the fact
that the procedure used for the division of the property in these two papyri
is endorsed in Islamic law. It is also extremely ancient and raises the question
how far, and in what way, the traditions of the ancient Near East lived on to
contribute to Islamic culture. In what follows we briefly survey the attestations
of lot-casting as an official practice from ancient Near Eastern to Islamic times
and discuss what we see as its significance.
Assigning Land, Booty, and Other Property by Lot
In the ancient Near East (by which, for the purposes of this article, we mean
the ancient Fertile Crescent), lot-casting was much used in the division of
inheritances. The standard way of distributing an inheritance in Assyrian and
Babylonian Mesopotamia was to divide the property into parcels and then to
assign the parcels by lot to the heirs (with variations when the eldest son was
privileged).4 The gods themselves are said to have divided the world by this pro-
cedure. ‘They took the box (of lots) …, cast the lots; the gods made the division’:
Anu acquired the sky, Enlil the earth and Enki the bolt which bars the sea.5 This
425 is | famously one of the ancient Near Eastern myths that passed into Greek cul-
ture: Zeus, Poseidon and Hades divide the world among themselves by lot in
the Iliad, and here as in the Akkadian myth, the three gods are brothers.6
The custom is well attested in the Bible, too.7 God Himself distributed the
desolate land of Edom to wild animals by lot (Isa. 34:17), and He also instructed
3 C.J. Kraemer, Excavations at Nessana, iii (Non-Literary Papyri) (Princeton 1958), no. 21. Com-
pare nos. 16, 31, where lots are not mentioned.
4 A History of Ancient Near Eastern Law, ed. R. Westbrook (Leiden 2003), 1, 57f. (general), 395f.
(Old Babylonian), 542 f. (middle Assyrian); 2, 939 (Neo-Babylonian).
5 Atrahasis in B.R. Foster, Before the Muses: an Anthology of Akkadian Literature3 (Bethesda,
Md., 2005), 229; also in S. Dalley (tr.), Myths from Mesopotamia, revised ed. (Oxford 2000), 9.
6 Cf. W. Burkert, The Orientalizing Revolution (Cambridge, Mass., and London 1992), 90f.; id.,
Babylon, Memphis, Persepolis: Eastern Contexts of Greek Culture (Cambridge, Mass., and Lon-
don 2004), 36. For the subsequent history of this myth, see A. Silverstein, ‘From Atrahasīs to
Afrīdūn: on the Transmission of an Ancient Near Eastern Motif to Iran’, Jerusalem Studies in
Arabic and Islam 39 (2012), 95–108.
7 Cf. Th. Gataker, On the Nature and Use of Lots2 (London 1627), modernized and updated by
C. Boyle (Exeter 2008), ch. 4, § 10, an extremely learned work still worth consulting despite its
age; J. Lindblom, ‘Lot-Casting in the Old Testament’, Vetus Testamentum 12 (1962), 164–178.
the ancient near east and islam: the case of lot-casting 19
Moses to divide the Promised Land by lot when it had been conquered;8 Joshua
duly did so.9 Micah seems to have envisaged conquest as the result of divine
or angelic lot-casting: he prophesied that Israel would have nobody in God’s
assembly to cast lots for land for it (Mic. 2:5). Ezekiel added that the land would
be divided up anew by means of arrows in the messianic age (Ezek. 45:1; 47:22).
Land and captives taken by the Babylonians and Assyrians were apparently
divided up in the same way: the Babylonians entered Israel’s gate and ‘cast lots
for Jerusalem’ (Obad. 1:11); but God would punish the nations for having divided
up his land and cast lots for his people (Joel 3:3). When the Assyrians conquered
Thebes in Egypt in 663 bce, ‘lots were cast for her nobles’ (Nah. 3:10). The Bible
does not refer to inherited land being divided by this method.
The idea of allocating new land by lots reappears in Jewish Hellenistic works.
In Jubilees, composed by a Palestinian Jew in the second century bce and later
translated from Hebrew into Greek and Syriac, Noah divides the earth by lot
between his three sons, Shem, Ham and Japheth; Canaan, the son of Ham,
nonetheless settled in Shem’s portion.10 In Maccabees, Antiochus iv (175–163
bce) is described as sending a Syrian commander with orders to wipe out the
residents | of Judaea and Jerusalem and to ‘settle aliens in all their territory, 426
and distribute their land by lot’ (1Macc. 3:36).11 Thereafter, leaving aside mere
retelling of the Biblical passages, the theme of lot-casting for land and/or its
inhabitants seems to disappear from the indigenous sources for a long time.
Lot-casting must be a universal institution, and not just as a private or
ad hoc method of decision making: both land and fortune are things that
one is ‘allotted’ in a great many languages. In Greek, too, a piece of land was
known as a lot (klēros), reflecting the fact that lots were used to distribute land
when colonies were set up in order to ensure that every group received an
equal share. Moveable booty was distributed in the same way,12 but whether
8 Num. 26:52 ff.; 33:50 ff. (at 54); 34:13; cf. also Josh. 21:4 ff.; 1Chron. 6:54ff., where priests and
Levites are given certain cities to dwell in by lot.
9 Josh. 18:3 ff., 10; 19:51; cf. Josephus, Antiquities, book 5, ch. 1, pars. 22, 24, 26.
10 Jubilees 8:11 ff., 10:30 (tr. O.S. Wintermute in J.H. Charlesworth [ed.], The Old Testament
Pseudepigrapha [New York 1983–1985], ii; cf. also his introduction). The detail that the
division was effected by lots seems to have been lost in the later Greek, Latin and Syriac
translations, but it was apparently known to the Muslims, cf. Silverstein, ‘From Atrahasīs
to Afrīdūn’.
11 Settling foreigners on land confiscated from the local population was an Assyrian practice
later adopted by the Achaemenids and Macedonians alike, but this passage could be
inspired by Obadiah on foreigners casting lots for Jerusalem.
12 Cf. G. Wissova, Pauly’s Real-Encyclopädie der classischen Altertumswissenschaft (Stuttgart
20 chapter 2
inherited land was also divided in this way is uncertain.13 The practice is not
attested at Athens14 nor, it would seem, anywhere else in Greek antiquity,
except in a speech once attributed to Dio of Prusa (in Anatolia, d. c. 120), now
held to be by Favorinus (d. mid-second century), a native of Arles: here we
are told that ‘brothers also divide their patrimony that way’.15 Wherever the
orator may have encountered the practice, it certainly sounds similar to that
attested in Petra and Nessana, but it is hard to say more on the basis of a single
passage.
The Romans, who took over from the Greeks, also used lots for the distribu-
tion of land, both at home and in connection with the foundation of colonies.16
Moveable booty, too, was (or might be) distributed by lot.17 But the evidence
427 relating to conquered land and | booty peters out in the third century, and the
Romans do not seem to have used this method in connection with inherited
property either, except in three specific circumstances. First, in actions for the
division of an inheritance or common property, or for the regulation of bound-
aries, it was difficult to decide who was the plaintiff and who the defendant, but
the person who appealed to the law was generally considered plaintiff; to this
Ulpian (d. 223) adds that if the parties appealed at the same time, the matter
was usually decided by lot.18 Secondly, in 428 a law was passed which entitled
the curia (city council) to claim one fourth of the estate left by a member of the
council to an outsider: the estate was to be divided into four parts, of which the
1894–1980, hereafter Pauly-Wissova), s.v. ‘Losung’, col. 1463 (Ehrenberg); D. Asheri, Dis-
tribuzioni di terra nell’antica Grecia (Turin 1966), 13 (drawn to our attention by D. Rous-
sel).
13 Ehrenberg categorically denies it, against earlier authors (cf. Pauly-Wissova, s.v. ‘Losung’,
col. 1478b).
14 Cf. A.R.W. Harrison, The Law of Athens: the Family and Property (Oxford 1968), ch. 5 (where
the possibility is not even discussed).
15 Dio Chrysostom (attrib.), Oratio, 64, 25, where ‘that way’ refers to ‘by lot’ (klērōtas).
Adduced by Gataker, Nature and Use of Lots, ch. 4, §12 (p. 102 of the original work, where
the references are given, misprinted as 46.25); cf. The Oxford Classical Dictionary3, ed.
S. Hornblower and A. Spawforth (Oxford 1996), s.v. ‘Favorinus’. We are much indebted to
Glen Bowersock and Christopher Jones for help with this passage.
16 Pauly-Wissova, s.v. ‘Losung’, col. 1493; D.J. Gargola, Lands, Laws, and Gods (Chapel Hill, n.c.,
1995), 95 ff. For examples, see Dionysius of Helicarnassus, Roman Antiquities, ii, 16; ii, 35;
v, 60; x, 32.
17 Cf. the story of the third-century emperor Probus in Historia Augusta, Life of Probus, 8 (ed.
and tr. D. Magie [London and Cambridge, Mass., 1932], iii, 351).
18 Justinian, Digest, book 5, tit. 1, 13 f. (ed. and tr. T. Mommsen, P. Krueger and A. Watson
[Philadelphia 1985], i, 167).
curia would take one by lot.19 Thirdly, in 531 Justinian ruled that when several
persons had been given the option, by bequest, to pick an item such as a slave
and disagreement arose, they could cast lots: the winner would pick the item
and pay the others the value of their share.20 Division of the estate among the
heirs by lot as the normal procedure in intestate succession does not seem to
be attested.
In line with this, it is mostly as a literary theme that lot drawing for land is
attested in the Near Eastern literature (Jewish and Christian) from the second
century onwards, with no sense of a live practice behind it. The gods cast
lots again, this time for the nations of the earth, in the Pseudo-Clementine
Recognitions, a Jewish Christian work of the mid-fourth century: Simon Magus,
representing heresy, here argues that there are many gods, and that it was to
one of the lower gods that the Jews were assigned (a gnosticizing paraphrase
of Deut. 32:8f.).21 In the same vein, Pirqe de Rabbi Eliezer, a Jewish work of
(perhaps) the mid-eighth century, tells us that when seventy angels descended
in order to confuse the nations building the Tower of Babel, they cast lots
among the nations and Israel fell to God (who is not, of course, a lower God
here).22 The nations are also divided up by lot in the Acts of Thomas, but
now among the apostles rather | than the gods: India fell to Thomas.23 Egypt, 428
Ethiopia, Nubia and the Pentapolis fell to St Mark by lot (qurʿa), as a later
Christian adds.24 The story of the father who divides the earth between his
three sons by lot may have gone into the Persian tradition, though it is only
in Ibn al-Kalbī (d. 204/819 or later) that we see it: according to him, the ancient
king Farīdūn divided his realm (consisting of the entire world) among his three
sons by writing the names of the regions on arrows and telling each son to
choose an arrow.25 There does not seem to be any attestation of this method
of allocating inheritance shares in Persian law or practice, however.
19 Justinian, Codex, 10, 35, 2; cf. 10, 35, 1; A.H.M. Jones, The Later Roman Empire, 284–602
(Oxford 1964), 2, 747 f.
20 Justinian, Codex, 6, 43, 3, 1; cf. id., Institutes, ii, xx, 23.
21 Clement of Alexandria (attrib.), Recognitions, ii, 39 (tr. B.P. Pratten, M. Dods and T. Smith,
The Writings of Tatian and Thophilus and the Clementine Recognitions [Ante-Nicene Chris-
tian Library, iii, Edinburgh 1867], 218 f.).
22 Pirqe de Rabbi Eliezer, tr. G. Friedlander (New York 1971), 176f.
23 Acts of Thomas, 1 (tr. A.F.J. Klijn [Leiden 2003], 17).
24 Severus b. al-Muqaffaʿ, ‘History of the Patriarchs of the Coptic Church of Alexandria’,
ed. and tr. B. Evetts, in R. Graffin and F. Nau (eds), Patrologia Orientalis, i (Paris 1907),
105.
25 Al-Ṭabarī, Taʾrīkh al-rusul waʾl-mulūk, ed. M.J. de Goeje et al. (Leiden 1879–1901), i, 226f.
(Ibn al-Kalbī), with further details in Silverstein, ‘From Atrahasīs to Afrīdūn’. It is not clear
22 chapter 2
At this point one is tempted to conclude that the ancient practice of casting
lots for land, whether conquered or inherited, had disappeared, except for
some special cases where Roman law applied. But it had not. The rabbis discuss
it, apparently as a live institution, with reference to two or three brothers
dividing an inheritance among themselves in material from second-century
Sephhoris (Tiberias) in Palestine onwards;26 and it now proves to have been
practised by Christians in Roman Arabia, too, at Petra and Nessana.
Apparently, it was also alive in the Prophet’s Arabia, at least in connection
with conquered land and booty. We are told that when the Prophet conquered
Khaybar (in the year 7/628), he set aside God’s fifth by lot (using arrows); the
rest of the conquered land was divided into eighteen portions and subdivided,
according to one tradition, into a hundred plots of roughly the same productive
capacity which he distributed to his followers by lot.27 Of the booty from the
429 campaign against B. Qurayẓa we are told that it was | divided into 3072 shares,
consisting partly of land and partly of moveable booty, of which a fifth was
assigned to God and the rest to the Muslims by lot.28 The Muslims also cast
lots for the captives taken at Badr.29 ʿUthmān (644–656) instructed Muʿāwiya
to single out God’s fifth of the booty by writing ‘God’ on one of the five arrows
used for their allocation.30 When ʿAlī’s followers wanted to divide the captives
from the Battle of the Camel among themselves, in 36/656, ʿAlī dissuaded them
by first telling them to bring the lots and next, when they brought the arrows, by
asking them who might get his (spiritual) mother ʿĀʾisha in his lot.31 On another
occasion he used the lots to divide non-Muslim booty.32 Of the Kufan ʿAbīda b.
whether the story should be taken to reflect Persian appropriation of the theme, either
directly from Mesopotamian sources or via para-Biblical literature such as Jubilees, or
simply Ibn al-Kalbī’s own familiarity with the theme.
26 Babylonian Talmud, Baba Bathra, 106a. It is not found in the Jerusalem Talmud.
27 Al-Wāqidī, Kitāb al-maghāzī, ed. M. Jones (London 1966), ii, 680, 692; al-Māwardī, Adab
al-qāḍī, ed. Y.H. al-Sirḥān (Baghdad 1971), ii, 196f., no. 2715 (citing Wāqidī); al-Shāfiʿī in
al-Bayhaqī, Aḥkām al-Qurʾān, ed. M.Z. al-Kawtharī (Cairo 1951), 163; cf. also Ibn Saʿd, al-
Ṭabaqāt, ed. E. Sachau et al. (Leiden 1904–1940), ii/1, 78, 82f.; ed. Beirut 1957–1960, ii, 107,
113 f. (without explicit mention of lots); ei2, s.v. ‘Khaybar’, col. 1141a.
28 Wāqidī, Maghāzī, ii, 522; cited in Māwardī, Adab al-qāḍī, ii, 196, no. 2714.
29 Wāqidī, Maghāzī, i, 100, 107, 139.
30 Al-Sarakhsī, Sharḥ kitāb al-siyar al-kabīr li-Muḥammad b. al-Ḥasan al-Shaybānī, ed. Ṣ.-
D. al-Munajjid (Cairo 1957–1960), iii, 889.
31 Ibn Qutayba (attrib.), al-Imāma waʾl-siyāsa (Cairo 1969), i, 78.
32 Al-Nuwayrī, al-Bidāya waʾl-nihāya (Cairo 1975), xx, 219, where he divides the booty from
Iṣfahān, even including a loaf, into seven portions (one for each of the sevenths into which
Kufa was divided at the time) and distributes them by lot.
Qays (d. 70s/690s) we are told that he would cast lots to assign the leftover from
the division of moveable booty, such as a dirham, saying that this was how it
had been done in past campaigns, but this was more controversial: the point
of the report is that he was persuaded to stop, on the grounds that it was more
equitable to use the dirham to buy something that could be distributed (by lot
or otherwise).33
All these reports are prescriptive and hardly to be taken at face value as his-
torical reports. Taken as literature, however, they certainly suggest that Muslims
who came out of Arabia took the use of lots for the division of conquered land
and booty for granted. This is corroborated by the fact that the standard word
for a share of the booty was sahm (literally ‘arrow’).
As regards inherited land, a Prophetic tradition reports that two men who
had a dispute over inherited property submitted their case to the Prophet
without having anything to prove their respective claims: he told them to cast
lots and take whatever was assigned to them by this method.34 The two men are
not identified as brothers, | however, and the issue is their dispute in a situation 430
without proof rather than the normal procedure in intestate succession. We
are also told that when Abān b. ʿUthmān was governor of Medina in the reign
of ʿAbd al-Malik (685–705), a man manumitted the six slaves who were his
only property on his deathbed; and since bequests were not allowed to exceed
a third of the property, Abān drew lots and manumitted the two slaves who
had the lucky draw.35 The Prophet is said to have used the same solution when
two earlier Medinese manumitted six slaves who were their only property, but
this is presumably a simple reworking of the Umayyad report (though it was
of course the Prophetic precedent which became canonical).36 Here too the
procedure diverges from that attested at Nessana and Petra, for the lots are not
33 Ibn Saʿd, Ṭabaqāt, ed. Sachau, vi, 62f.; ed. Beirut, vi, 93. He was ʿarīf (paymaster) for his
tribal group.
34 Abū Dāwūd, Sunan (Cairo 1982), ii, 295 (K. al-qaḍā, bāb fī qaḍāʾ al-qāḍī idhā akhṭaʾa);
cited in Ibn Qayyim al-Jawziyya, al-Ṭuruq al-ḥukmiyya fī siyāsat al-sharʿiyya, ed. N.A. al-
Ḥamad (Mecca 1428), ii, 743 (in a useful list of Prophetic traditions on qurʿa), where further
references are given. For an Imāmī Shīʿite version, see al-Majlisī, Biḥār al-anwār (Tehran
1357–1392), civ, 324. Our thanks to Aron Zysow for help in connection with this tradition.
35 Al-Shāfiʿī, ‘K. al-qurʿa’, in his Umm (Beirut 1993), viii, 5; cf. J. Schacht, The Origins of
Muhammadan Jurisprudence (Oxford 1956), 201 f. For further references, see Bayhaqī,
Aḥkām al-Qurʾān, 162n.
36 Ibn Abī Shayba, al-Muṣannaf, ed. M.A. al-Nadwī (Bombay 1979–1983), xiv, 158, nos. 17934f.;
Shāfiʿī, ‘K. al-qurʿa’, Umm, viii, 5 (where one manumitter is a woman, the other an Anṣārī
male); further references in Bayhaqī, Aḥkām al-Qurʾān, 162n.
24 chapter 2
being used to allocate equal shares, but rather to pick out two winners. Though
it seems unlikely that the inhabitants of Petra and Nessana should have been
the only Arabs to use lots as the normal procedure for the division of inherited
land, the practice does not seem to be attested in the material on the rise of
Islam. We do however find it in classical Islamic law: here we are told that once
the property had been divided into parcels representing the smallest fractions
to be distributed, the heirs could draw lots among themselves for the parcels;
if the estate consisted of different types of property, such as houses and land,
the different types had to be divided up separately; they could not be bundled
together as was done at Nessana.37
The Near East is not the only region in which lots have been used for
431 the partition of inherited land. It crops up in Europe, too. Thomas | Aquinas
(d. 1274) knew of it,38 and English common law endorsed it for the parti-
tion of land held in coparcenary from medieval down to modern times.39 The
solution is likely to have commended itself wherever property had to be dis-
tributed among equally entitled claimants, and it could in principle turn up
anywhere in unrelated forms. The Near Eastern forms come across as related
in that all they treat lot-casting as a standard way of dividing land and other
property, not simply as a last resort or special solution, as in Roman or com-
mon law. The same may well have been true among many other peoples in
ancient times, however, especially in connection with conquered land, and
the Near Eastern forms are not related etymologically: the usual term for a
37 Māwardī, Adab al-qāḍī, ii, 194f., nos. 2709ff., cf. also 204, nos. 2746ff.; al-Nawawī, Min-
hāj al-ṭālibīn, ed. and tr. L.W.C. van den Berg (Batavia 1882–1884), iii, 395ff.; Ibn Rushd,
Bidāyat al-mujtahid, ed. M.S. al-Muḥaysin and Sh.M. Ismāʿīl (Cairo 1970–1974), ii, 298ff.;
tr. I.A. Khan Nyazee and M. Abdul Rauf (Reading 1996), ii, 319ff. (both with further dis-
cussion); al-Marghīnānī, al-Hidāya (Cairo n.d.), iv, 46; tr. C. Hamilton, 2nd ed. (Lahore
1957), iv, 571 (K. al-qisma); al-Mawsūʿa al-fiqhiyya, xxxiii (Kuwait 1995), 139 (drawn to
our attention by A. Zysow); A. ʿAbd al-ʿAzīz, Fiqh al-kitāb waʾl-sunna (Nablus 1999), iv,
2305.
38 38 He describes it as a method used for the division of inheritances in cases of disagree-
ment, without giving further details (Thomas Aquinas, Commentary on the Ephesians, tr.
M.L. Lamb [Albany 1966], book 1, lecture 4, ad Eph. 1:11).
39 39 An estate held in coparcenary was taken by several heirs as if they were a single person,
for example when the deceased only left daughters (the principle being that there could
only be one heir, normally the eldest son, who would take everything in the absence of
a will). The use of lots for the partition of such estates is first described by Thomas de
Littleton (d. 1481), cited in Gataker, Nature and Use of Lots, ch. 4, §12 (p. 104 of the original
work); it is endorsed in Great Britain, Courts, The Legal Guide, 1 (London 1839), 324f., but
is now obsolete.
lot in the sense of the object used in the procedure is pūr(u) in Assyrian,
isqu in Babylonian, goral in Hebrew (where it also stands for the share allot-
ted), and qurʿa in Arabic, with sahm (‘arrow’) as the normal word for the lot
awarded. But though they may have originated separately in pre-historic times,
by the time we have literary evidence, the Near Eastern institutions stand
apart from those of the neighbouring lands in that they still treat lot-casting
as the standard mode of division, even in connection with inheritance law, and
even, after the coming of Islam, when the heirs were awarded highly unequal
shares. It is with reference to this feature that we treat them as so many mem-
bers of a single family, visible in the cuneiform, Jewish, Greek papyrological,
and Muslim records at different times and places thanks to a combination of
local conditions and the haphazard manner in which the evidence has sur-
vived.
One interesting point here is that if it had not been for the chance preser-
vation of the two Greek papyri, one might have taken lot-casting for the distri-
bution of land in early Islamic society and classical law to represent a case of
Jewish Fortleben in Islam; for until the papyri were discovered, it was only in
rabbinic texts that the practice seemed to be alive in connection with inheri-
tance shares, and the rabbis would of course have had much to say about the
Biblical use of lot | drawing in connection with conquered property, had they 432
been asked. But as the papyri show, the inference would have been false. Lot-
casting for the allocation of inherited property had remained a live practice
in Roman Arabia, too, and also, as the accounts of the Prophet’s procedures
suggest, in connection with conquered land and booty elsewhere in Arabia.
What the striking similarity between Jewish and Islamic law reflects is not,
in this particular case, Jewish Fortleben in Islam, but rather the shared roots
of Jewish and Islamic culture in the ancient Near Eastern tradition. We seem
to have here a case comparable to that of circumcision, practised by both the
Jews and the Arabs (eventually Muslims), not by the one borrowing from the
other, but rather by both retaining an ancient custom which had once been
widespread in the Near East (notably in Egypt). In the case of circumcision,
the Biblical record played a role in endowing the old Arabian practice with a
new religious meaning. There is no suggestion that it did so in the case of lot-
casting.
It is because the Arabs were apt to preserve ancient practices also recorded
in the Jewish scripture that Old Testament scholars (Wellhausen prominent
among them) used to study Arabia with such interest, with special attention
to the bedouin because the ancient Israelites had been pastoralists. It is the
townsmen of Arabia that we see at work at Nessana and Petra, but the bedouin
continued to furnish parallels into modern times: Musil reports that in what
26 chapter 2
he called Arabia Petraea (former Roman Arabia) agricultural land belonging

to the whole tribe would be divided into fields of equal size every year and
distributed among the families or tribal groups by lot.40 The continuity with
ancient Near Eastern practice in Arabia should presumably be related to the
forbidding nature of the peninsula. Difficult to conquer and colonize, it was
the only region of the Near East to escape a millennium of Greek, Roman or
Persian domination, though parts of it (including Petra and Nessana) fell under
foreign rule for periods ranging from centuries to decades. We have to stress,
however, that the Jews and the Arabs may not have been the only inhabitants
of the Near East to use lot-casting for the division of inherited property in late
antiquity, for on the Jewish side it is in rabbinic literature that it is attested, not
in the Bible. This suggests that what the rabbis discussed was a practice they
shared with their neighbours, or in other words that in this particular case the
433 rabbinic literature should not be seen as evidence | for the Jews alone, but also
for the larger Aramaic culture of which they formed part.
Choosing People by Lot
It was not only in connection with the distribution of land and its inhabitants
that lot-casting was used in the ancient Near East; people were selected for
a wide variety of functions by that method, too. The Assyrians used sortition
to choose the annual occupant of the ‘office of the year eponym’, a dignitary
who had the privilege of having a calendar year named after him.41 The king
himself never seems to have been chosen by lot in ancient Mesopotamia,42 nor
do priests. But the Bible tells us that Saul was chosen as king by lot,43 and the
Samaritan Chronicle has it that the first Samaritan king was chosen by the same
method.44 In Pseudo-Philo (c. 50–150) the Israelites also choose Kenaz as their
leader against the Philistines by lot, directed by an angel, and repeatedly try
the same method to find a successor to Phinehas without success.45 By Roman
40 A. Musil, Arabia Petraea (Vienna 1907–1908), 3, 294.

41 W.W. Hallo, ‘The First Purim’, The Biblical Archaeologist 46 (1983), 19f.
42 M.T. Larsen, ‘The City and Its King’, in Le palais et la royauté, ed. P. Garelli (Paris 1971), 298f.
(against Oppenheim).
43 1 Sam. 10:19–21. But God’s answer in v. 22 must have been given by a seer or prophet, cf.
J. Lindblom, ‘Lot-Casting in the Old Testament’, 165n.
44 J. Macdonald, The Samaritan Chronicle, no. ii (Berlin 1969), 99.
45 Latin text (originally Hebrew) and English translation in H. Jacobson, A Commentary on
Pseudo-Philo’s Liber Antiquitatum Biblicarum (Leiden 1996), 25:1f.; 49:1.
times succession to the high priesthood of the Jews had come to be decided in
the same way, with explicit reference to ancient practice.46
In Biblical times, lots were also used to single out the groups and individuals
who were to serve as temple musicians and gate keepers in ancient Israel
(1 Chron. 24:5ff., 25:8ff., 26:13f.), and to allocate rotating responsibilities such
as serving as priests and providing wood offerings to the temple (Neh. 10:35).47
Zachariah was a priest chosen by lot to officiate at a particular time (Luke
1:8 f.),48 and Peter found a replacement for the apostle Judas by selecting two
men and then casting lots (Acts 1:23–26), a procedure which was to be imitated | 434
by later Christians in the Near East and the West alike.49 Indeed, the word
‘clergy’ is derived from klēros, ‘lot’, the clergy being people allocated to God.50
Lot-casting may also have been used to assist the decision who should be
admitted as new members of the community at Qumran, but this is disputed.51
Again, the Greeks and the Romans had similar practices. In Greece lot-
casting was used for the selection of magistrates, especially in democracies,
where it was of fundamental importance as an egalitarian device.52 The Ro-
mans would distribute functions among magistrates already chosen by sorti-
46 Josephus, Wars, book 4, ch. 3, pars. 7 f.

47 Cf. Josephus’ amplifications, Antiquities, book 7, ch. 14, par. 7.
48 Compare Protoevangelium of James, 24:4, in W. Schneemelcher and R. McL. Wilson (eds),
New Testament Apocrypha, 1 (Cambridge 1991), 437, where Zachariah in his turn is replaced
by Simeon by lot.
49 For thirteenth-century nuns choosing an apostle (as patron saint) by lot, see G.G. Coulton
(tr.), Life in the Middle Ages, 1 (Cambridge 1928), 69 f. Thomas Aquinas held that lot-casting
could not be used for ecclesiastical office after the arrival of the Holy Spirit (Commentary
on the Ephesians, book i, lecture 4), but the Mennonites of today choose priests by lot
(personal communication from Christopher Melchert). In the Middle East known to
T. Fahd, monks would decide by lot which novices should receive the habit (ei2 [Leiden
1956–2004], s.v. ‘ḳurʿa’).
50 This too is discussed in Pauly-Wissova, s.v. ‘Losung’, cols. 1466f. (and indeed by Aquinas,
loc. cit.).
51 Lots figure prominently in the Dead Sea scrolls, but almost exclusively in a metaphorical
sense (Y. Licht, ‘The Term Goral in the Writings of the Judean Desert Cult’, Beth Miqra 1
[1956], 90–99 [Hebrew]). For the question of its use in admissions, see W.A. Beardslee,
‘The Casting of Lots at Qumran and in the Book of Acts’, Novum Testamentum 4 (1960),
245–252; S.J. Pfann, ‘The Essene Yearly Renewal Ceremony and the Baptism of Repentance’,
in D. Parry and E. Ulrich (eds), The Provo International Conference on the Dead Sea Scrolls
(Leiden 1999), 337–352; P.S. Alexander, ‘Predestination and Free Will in the Theology of
the Dead Sea Scrolls’, in J.M.G. Barday and S.J. Gathercole (eds), Divine and Human Agency
in Paul and His Cultural Environment (London 2007), 27–49.
52 Oxford Classical Dictionary, s.v. ‘sortition’; Pauly-Wissova, s.v. ‘Losung’, cols. 1475ff.
28 chapter 2
tion. Consuls and praetors, for example, would cast lots among themselves to
determine the assignment of campaigns and provinces (‘What if the casting of
lots had allocated you Africans or Spaniards or Gauls to rule over?’, as Cicero
asked his brother, then governor of Asia);53 lots were also used to determine
voting order and other sequences, to choose officials for special tasks, and in
diverse other connections, including (at least on one occasion) that of select-
ing recruits.54 We even hear of bandits who reputedly used lots to decide which
435 members of the gang should labour or serve the | others,55 but whether this can
count as an example of official use is another question.
On the Greek and Roman side, the official use of lots for the allocation
of office and functions seems to have petered out by late antique times, and
the evidence is thin on the Near Eastern side as well. Rabbinic literature does
admittedly abound in discussions of temple duties and other Old Testament
institutions, but it is all academic. Choosing priests, monks and other ecclesi-
astical personnel by lot is more likely to have continued among the Christians,
thanks to the precedent set by Peter’s choice of Matthew by this method. It
is reflected in the Protoevangelium of James, where Mary is chosen by lot for
the privilege of weaving a particular item,56 but the only attestation relating
to real life that we know of is modern.57 This undoubtedly reflects our igno-
rance of the vast mass of relevant Syriac literature. Once again there is some
ambivalent evidence on the Iranian side:58 in the account of Ardā Virāz’ jour-
ney to heaven and hell, Ardā Virāz is chosen for the journey by three lances
(nēzag) which are thrown at him. But this procedure was in the nature of an
ordeal rather than lot-casting, for the lances were meant to confirm or deny
the suitability of a man already chosen; there were no other candidates.59
53 Cicero, Ad Quintum fratrem, 1, 9, 27.

54 Pauly-Wissova, s.v. ‘Losung’, cols. 1494 ff.; Gargola, Land, Laws, and Gods, 95; R.J.A. Talbert,
The Senate of Imperial Rome (Princeton 1984), 61, 139, 144, 207f., 347–353 (drawn to our
attention by Nathan Rosenstein); N. Rosenstein, ‘Sorting Out the Lot in Republican Rome’,
American Journal of Philology 116 (1995), 43–75, with the recruits at 44, n. 7.
55 B. Shaw, ‘The Bandit’, in A. Giardina (ed.), The Romans, tr. L.G. Cochrane (Chicago and
London 1993), 330 (with ref. to Apuleius, Metamorphoses, 4.8).
56 Protoevangelium of James, 10:2, in Schneemelcher and Wilson, New Testament Apocrypha,
1, 430.
57 Cf. Fahd, above, note 49.
58 The Persians are envisaged as casting lots to fix the day on which the Jews were to be killed
in the Book of Esther (3:7). The institution credited to them here is Akkadian, but whether
it can be inferred that the Persians had adopted it is unclear.
59 Cf. P. Gignoux, ‘Une ordalie par les lances en Iran’, Revue de l’Histoire des Religions 200
(1983), 155–161. The procedure is construed as lot-casting in S. Shaked, ‘Quests and
One would be inclined to conclude that the once prevalent practice of choosing
people for high office and other functions by lot had died out.
Again, however, the practice must have survived in Arabia. Unfortunately,
there does not seem to be any documentary evidence for this. Three pre-
Islamic inscriptions, one from al-Lāt’s temple at Palmyra and two from Yemen,
do refer to lot-casting, but they probably refer to divination.60 We are told,
however, that the pre-Islamic Quraysh would choose men to lead them in war
by lot and accept the candi|date even if he was a minor or very old;61 and 436
the terms qarīʿ and maqrūʿ (chosen by lot) were used in the sense of chief,
leader and person chosen.62 In line with this we later hear of lot-casting for
the selection of caliphs. The Christian astrologer Theophilus of Edessa, active
under the caliph al-Mahdī (d. 169/785), tells us that when Yazīd i died, the
future Marwān i (64/684–685) proposed to solve the succession dispute which
ensued by drawing lots; this was apparently agreed, but when Marwān’s name
came up, his rival al-Ḍaḥḥāk b. Qays refused to accept the result, so the two
of them fought it out at Marj Rāhiṭ.63 Al-Jāḥiẓ also knew of lot-casting in
connection with the choice of caliphs, though he did not think it was necessary:
in his view, the rightful claimant would always be known without the need
for formal procedures, just as everyone knew who was the most generous
man or the best horseman among Qays in the Jāhiliyya without discussion
of their merits or shūrā or casting lots (al-iqrāʿ waʾl-musāhama).64 Lot-casting
was endorsed by some jurists for situations in which two candidates for the
caliphate were equally qualified, or when two of them had come to be elected
by some mishap, but others disagreed.65 ‘In our opinion, lots are required by
Visionary Journeys in Sasanian Iran’, in J. Assmann and G.G. Stroumsa (eds), Transforma-
tions of the Inner Self in Ancient Religions (Leiden 1999), 73.
60 R.G. Hoyland, Arabia and the Arabs (London 2001), 156.
61 Ibn al-Jawzī, al-Muntaẓam (Beirut 1992–1993), ii, 217f., apparently from Ibn al-Kalbī.
62 Ibn Manẓūr, Lisān al-ʿarab (Beirut 1955–1956); Murtaḍā al-Zabīdī, Tāj al-ʿarūs, ed. ʿA. Shīrī
(Beirut 1994), both s.v. ‘qrʿ’.
63 Theophilus as reconstituted by R.G. Hoyland, Seeing Islam as Others Saw It (Princeton
1997), 647, cf. 400 ff.
64 Al-Jāḥiẓ, al-ʿUthmāniyya, ed. ʿA.-S.M. Hārūn (Cairo 1955), 266. In Ibn Ṭāwūs, Fatḥ al-abwāb
bayna dhawī ʾl-albāb wa-bayna rabb al-arbāb fī ʾl-istikhārāt, ed. Kh. al-Khaffāf (Beirut 1989),
267 ff. (chs 20–21), musāhama consists of drawing lots from paper with names written
on them, whereas a qurʿa is an object such as a pebble or a rosary bead, but it was not
necessarily so in Jāḥiẓ’ time. (Our thanks to Etan Kohlberg for drawing Ibn Ṭāwūs’ work to
our attention.)
65 Abū Yaʿlā, al-Aḥkām al-sulṭāniyya, ed. M.Kh. al-Fiqī, second printing, Cairo 1966, 25 (where
lot-casting is prescribed in the first situation and is one out of two acceptable views in the
30 chapter 2
the law to spare people’s feelings, not to establish rights’ (li-taṭyīb al-qulūb dūna
ithbāt al-ḥuqūq), as al-Nasafī (d. 508/1114) observed with reference to the second
situation, meaning that it could only be used for the random distribution of
things to which people had a lawful claim, not to pick out winners.66 No caliph
437 actually seems to have been chosen by this method, but | much later we hear
of an Ottoman grand vizier who was chosen by lot (drawn from pieces of paper
with the names of candidates written on them).67 This was in 1204/1789f., at the
beginning of the reign of Selim iii, and its relevance to our present concerns is
uncertain.
There seems to have been a tradition in Arabia of choosing people for other
functions by lot as well. The Prophet is said to have decided which wife should
accompany him on his travels by lot-casting;68 the Medinese are said to have
used lots to determine who should have the privilege of hosting the Prophet;69
ʿAlī is credited with using lots to settle a case in Yemen in which three men
denied paternity of a child that any one of them could have fathered.70 ʿUmar ii
is said to have included the wives and children of the soldiers in the dīwān and
cast lots to decide who should receive a hundred and who forty dirhams, i.e.
from the income of the immoveable booty which was paid out as stipends.71
All these examples refer to men in official positions, but hardly to lot-casting as
a regular, public institution (though all decisions recorded for the Prophet were
to assume that character). We do, however, encounter lot-casting as a regular
institution in connection with mobilisation.
second); al-Māwardī, al-Aḥkām al-sulṭāniyya, ed. M.J. al-Ḥadīthī (Baghdad 2001), 60.1, 62.–
6; tr. W.H. Wahba (Reading 1996), 6, 8, on unnamed jurists (without verdict on the first
situation, but with arguments against lot-casting in the second).
66 Abū ʾl-Muʿīn al-Nasafī, Tabṣirāt al-adilla, ii, 826f., against al-Qalānisī and al-Kaʿbī. His
position is Ḥanafī, cf. below.
67 Ahmed Cevdet Pasha, Tarih-i Cevdet (Dersaadet 1309), v, 18 (on Ruscuklu Hasan Pasha).
We owe this reference to Şükrü Hanioğlu.
68 Ṭabarī, Taʾrīkh, i, 1519. Compare Babylonian Talmud, Shabbath, 149b, on how Nebuchad-
nezzar would cast lots to decide which of his recently acquired (male) captives to have
sexual relations with.
69 Ibn Saʿd, Ṭabaqāt, ed. Sachau, iii/1, 288; ed. Beirut, 396 (s.v. ‘ʿUthmān b. Maẓʿūn’).
70 He imposed two thirds of the blood-money (for the child) on the man picked out as the
father, presumably on the reasoning that he had caused the other two men to lose a third
of a child each. The Prophet found this solution uproariously funny (Ibn Ḥanbal, Musnad
[Cairo 1313], iv, 373; Wakīʿ, Akhbār al-quḍāḥ, ed. ʿA.-ʿA.M. al-Marāghī [Cairo 1947–1950], i,
91). For a variant involving two men and a slave girl, see al-Majlisī, Biḥār, xl, 244f., cf. also
civ, 63.
71 Ṭabarī, Taʾrīkh, ii, 1367.
When ʿUthmān permitted Muʿāwiya to conduct campaigns by sea, he stipu-

lated that Muʿāwiya was not to select the men himself or cast lots among them
(lā tantakhib al-nās wa-lā tuqriʿ baynahum), but rather to let them decide them-
selves whether to go.72 Sortition was apparently among the methods normally
used in the army to decide who was to go on duty. Of a Syrian soldier who
went on annual summer campaigns in the Byzantine empire in the reign of
Muʿāwiya we are told that he had a bad dream predicting that he would be the
killer of an eminent Medinese and thereby doom himself to Hell; | when people 438
were chosen by lot for Yazīd i’s campaign against Medina (ḍuriba qurʿat baʿth
al-Madīna) in 63/682f., this man had the misfortune to be selected ( fa-aṣābatnī
al-qurʿa).73 In these two examples it is the authorities who use lots, but there
are also stories in which it is the soldiers themselves who do so, some set in
the Prophet’s time. A Medinese desirous of martyrdom told the Prophet that
he had missed the battle of Badr because he drew lots with his son to decide
which one of them should go and it was his son’s lot that had come out (kharaja
sahmuhu).74 Qurʿa was used to select eighty men from a group of volunteers in
connection with another expedition.75 In these stories enlistment is envisaged
as voluntary, but only one man can go because one has to stay behind to look
after the family, or only eighty men are needed of the many who have volun-
teered. In another ḥadīth, Abū Hurayra invokes the example of a man who goes
on campaign with some people and whose lot does not come out (lam yakhruj
sahmuhu) because he has not said ‘amen’;76 here the volunteers are already on
campaign and the question is who should go on a particular expedition in the
course of it. We also hear of men in the mid-Umayyad period who would cast
lots among themselves when they were called up to decide who should actually
go; those who won would stay at home in return for payment of a sum known as
jaʿāʾil.77 Here the assumption seems to be that a particular tribal group would
be told to supply a specified number of men and that the men could decide
72 Ibid., i, 2824.
73 Ibn Qutayba, al-Imāma waʾl-siyāsa, i, 215 f.
74 Wāqidī, Maghāzī, i, 212, on Saʿd b. Khaythama; cited in Majlisī, Biḥār, xx, 125.
75 Ibid., xxi, 77 (on ghuzāt al-silsila).
76 Al-Haythamī, Majmaʿ al-zawāʾid (Beirut 1982), ii, 113 (K. al-ṣalāh, bāb al-taʾmīn).
77 M. Bonner, ‘Jaʿāʾil and Holy War in Early Islam’, Der Islam 68 (1991), 47f., with reference
to T. Nöldeke, Delectus Veterum Carminum Arabicorum (Wiesbaden 1933), 77, and other
sources where the poet is said to have been called up by Muʿāwiya’s governor of Kufa (but
the campaigns in Khwārizm only started in the governorship of Qutayba); Ṭabarī, Taʾrīkh,
ii, 1029, without the poem, where the expedition is despatched by ʿAbd al-Malik. Exactly
how the procedure worked is not clear.
32 chapter 2
for themselves whom to send: they all wanted to stay at home rather than to
be martyred. The Ottomans provide a much later parallel for the use of lots in
connection with military service, too. Al-Majlisī records that when ʿUmar Pāshā
(1764–1776), Mamluk governor of Iraq on behalf of the Ottomans, arrived, he
‘imposed harsh lot-casting on them (ishtadda ʿalayhim al-qurʿa)’ and took sol-
439 diers from villages and the amṣār, high and low, learned | and ignorant, and
ʿAlids and others alike.78 When Muḥammad ʿAlī (1805–1848) introduced con-
scription in Egypt, qurʿa was apparently also meant to be used;79 the Ottoman
conscription system of 1848 was actually known as Qurʿa niẓamnamesi (regula-
tion on the drawing of lots);80 and lots were also used to draft soldiers in Egypt
under Khedive Ismail (1863–1879).81
We abstain from the attempt to account for the Ottoman examples. The
point of interest to us is that in the period with which we can claim some
familiarity (from the rise of Islam to the Mongols), references to the use of lots
in an official context are clustered in the first century, where the Prophet, the
Rāshidūn and the Umayyads form a continuum, to fall off rapidly thereafter,
except in connection with legal procedure. No doubt more will turn up, but it
seems reasonable to infer that the official use of lot-casting for the selection of
persons was a practice rooted in Arabia.
The Qurʾān and the Law
Lot-casting figures in the Qurʾān, but only as a literary theme, not as a live
practice or an object of legislation. Two passages are relevant. The first is
q. 3:44, concerned with Mary. Much of what the Qurʾān has to say about her
78 al-Majlisī, Biḥār, liii, 331.

79 Kh. Fahmy, ‘The Nation and its Deserters: Conscription in Mehmed Ali’s Egypt’, in E.J. Zür-
cher (ed.), Arming the State: Military Conscription in the Middle East and Central Asia
1775–1925 (London and New York 1999), 67, citing Sir John Bowring’s report of 1840 on how
men would be seized without any order, arrangement, inscription ‘or lot-drawing’.
80 E.J. Zürcher, ‘The Ottoman Conscription System in Theory and Practice, 1844–1918’, in
Zürcher, Arming the State, 82f., with a description of the system. Prof. Şükrü Hanioğlu,
to whom we are much indebted for help on Ottoman questions, tells us that the draw of
lots for conscription was called qurʿa-i sherʿiyye in the vernacular, military service being
a religious duty. According to Fahd, qurʿaya girmek came to mean ‘reaching the age of
military service’ (ei2, s.v. ‘ḳurʿa’). See also Granquist in Lindblom, ‘Lot-Casting in the Old
Testament’, 169n, where the system is slightly different from that described by Zürcher.
81 J.P. Dunn, Khedive Ismail’s Army (New York 2005), 43.
life reflects the Protoevangelium of James, a work written in Greek some time
after 150, widely read in the Christian Near East, and translated into Syriac in
the sixth century. In this work we read that Mary grew up in the temple and
that the priests decided to marry her off when she was twelve years old, lest
she pollute the temple by having periods (this passage is strikingly reminiscent
of the story of the ‘Mouse-Maiden’ in the Pañcatantra/Kalīla | wa-Dimna). The 440
priests assemble the widowers of the people and tell them to bring a rod, and
when a dove flies out of Joseph’s rod, they assign Mary to him.82 In other
words, it is a miracle that singles out Joseph as her husband, not lots. But
lots appear in other stories in the Protoevangelium, and on a later occasion
it even mentions that Joseph himself had won his bride by lot.83 The Qurʾān,
on the other hand, briefly declares that ‘you (sg.) were not there when they
threw their rods (to determine) which of them should take care of Mary’ (idh
yulqūna aqlāmahum ayyuhum yakfulu Maryama, 3:44), seemingly referring to
the version with the miracle (and presenting the contest as over kafāla, care,
rather than marriage).84 But the exegetes generally understood the rods as ‘the
arrows with which the lot-drawers (al-mustahimūn) from among the sons of
Israel cast lots (istahama) for the guardianship of Mary’, as al-Ṭabarī puts it.85
The second passage is in the story of how Jonah came to be thrown overboard
from the ship on which he was travelling. In the Bible, Jonah is identified by
lots as the sinner on whose account the storm is sent (Jon. 1:7). In the Qurʾān
there is no reference to the storm, the ship is simply overloaded, so lots are
cast to determine who should be jettisoned; but Jonah is a guilty party here
too, and this does seem to be what the lots indicate: he has run away (abaqa)
and behaved shamefully (wa-huwa mulīm), and when he cast lots, his plea was
rebutted ( fa-sāhama fa-kāna min al-mudḥaḍīn) (37:140–142).
82 Protoevangelium of James, 8:2–9:1, in Schneemelcher and Wilson, New Testament Apoc-

rypha, 1, 429 f.; in the Indian story it is her father’s house that should not be polluted (cf.
the six versions of the passage, including the old Syriac, in F. de Blois, Burzōy’s Voyage to
India and the Origin of the Book of Kalīlah wa Dimnah [London 1990], 7ff.).
83 Protoevangelium of James, 19:1, in Schneemelcher and Wilson, New Testament Apocrypha,
1, 434 (Tischendorf’s version).
84 The Protoevangelium thinks of Mary as a perpetual virgin and accounts for Jesus’ brothers
by casting Joseph as an old widower with children by his first marriage when he wins
Mary. In the Qurʾān, the old man who wins her is Zakariyyā (cf. 3:37), the father of John
the Baptist, and her husband has completely disappeared, an interesting development
which must tell us something about the religious milieu reflected in the Qurʾān.
85 Ṭabarī, Tafsīr, ad 3:44; similarly Fakhr al-Dīn al-Rāzī, Tafsīr, ad loc., and Ibn al-ʿArabī, below,
note 87.
34 chapter 2
The fact that lot-casting is mentioned in the Qurʾān in connection with

venerable figures meant that the procedure had excellent legitimation. It also
generated some stories in which Muḥammad’s kafāla, like Mary’s, is decided
441 by lots.86 But since it was only in accounts of | earlier religious communities
that the procedure is mentioned, it did not shape Islamic law on this topic.
As Ibn al-ʿArabī observed, in the earlier sharīʿa, lot-casting had been sanc-
tioned for general use, whereas it was only used in specific cases in Islamic
law, and not in connection with kafāla; nor would using lots to throw a man
overboard be acceptable under Islamic rules, as both he and others pointed
out.87
As Ibn al-ʿArabī noted, lot-casting did, however, remain acceptable in Islam
in other contexts. First, as mentioned already, the jurists accepted that inheri-
tances (and other joint property) could be allocated by lot. They seem to have
done so without any controversy, and the method is still prescribed for the par-
tition of joint property in the Ottoman Majalla.88 It is a remarkable example of
continuity from the ancient to the modern Near East, if only at a fairly low level
of juristic interest.
That booty could be allocated among equals by lot seems also to have been
widely accepted, at least as long as it was only a method of allocation of the
appropriate shares rather than the assignation of things left over. The imam
was charged with concern for the feelings of his subjects (murāʿāt qulūb al-
raʿiyya) and avoidance of preferential treatment, as al-Sarakhsī explained; for
this reason division of the booty was done by qurʿa, both in connection with
the fifth set aside for the imam and for the distribution of the remaining
four fifths. The four fifths would also be assigned to the pay-masters (ʿurafāʾ)
by lot, and each ʿarīf would divide the portion assigned to him among the
men of whom he was in charge by qurʿa, too, he said (using terminology
from the Umayyad period). He adduced the Prophet’s choice of a wife to
accompany him on his travels by lot as the paradigmatic case in that the
Prophet had used lots to spare their feelings (taṭyīban li-qulūbihinna).89 In
connection with partition, the Mālikī Ibn Rushd also tells us that the jurists
86 Al-Balādhurī, Ansāb al-ashrāf, ed. M. Ḥamīdallāh (Cairo 1959), 85.

87 Ibn al-ʿArabī, Aḥkām al-Qurʾān, ed. ʿA.M. al-Bijāwī (Cairo 1378/1958), iv, 1610f.; al-Qurṭubī,
al-Jāmiʿ li-aḥkām al-Qurʾān (Cairo 1967), xv, 126; and, before both of them, al-Jaṣṣāṣ, Aḥkām
al-Qurʾān (Beirut 1994), iii, 496 f.; all ad 37:141.
88 Al-Majalla (Mecelle-yi ahkām-i ʿadliyye), book x, articles 1151, 1156; cf. also 1180 (available in
English at www.iium.edu.my/deed/lawbase/al_majalle).
89 Sarakhsī, Siyar, iii, 889 f. On his handling of the Prophet’s precedent in connection with
wives, see also below, note 98.
accepted lot-casting taṭyīban li-nufūs al-mutaqāsimīn.90 It was on the same

principle that the Shāfiʿites and others accepted that one could choose prayer
leaders, naqībs | and other persons by sortition when the candidates were 442
equally entitled:91 the contenders had to be mustawīna fī ʾl-ḥujja, as al-Shāfiʿī
said.92
There were situations in which some jurists, above all the Ḥanafīs, deemed
lot-casting to amount to gambling (qimār), however. If a man manumitted
slaves worth more than a third of his property in death, sickness or by will,
the Shāfiʿīs, Ḥanbalīs, Mālikīs and Imāmīs would draw lots and manumit how-
ever many could be accommodated within the third in accordance with the
Prophetic ḥadīth, but the Ḥanafīs held that all the slaves should be set free and
obliged to work until they had paid off the value of the unmanumitted parts.93
Similarly, when two men claimed ownership of some property and adduced
equally valid proof, the Shāfiʿīs, Ḥanbalīs and Imāmīs accepted (among vari-
ous other solutions) that one could cast lots and give the disputed property
to the winner, directly or by having him take the oath which settled the mat-
ter; there were ḥadīths in which the Prophet and ʿAlī did so. But the Ḥanafīs
(and Mālikīs) would divide the property, arguing that the ḥadīths dated from
the period before the prohibition of gambling.94 There were also traditions in
90 Ibn Rushd, Bidāya, ii, 299.2; tr. Nyazee, i, 320 (translated ‘for the satisfaction of the persons
participating in the partition’).
91 Māwardī, Aḥkām, 273 (niqāba), 278 (leadership of prayer), 532.ult. (order on the military
roll), 589.ult. (retaliation); tr. Wahba, 109, 113, 224, 254; Nawawī, Minhāj, iii, 99f., 102
(ḥaḍāna), 119f. (retaliation), 379 (admission to the court room). The Mālikīs and Ḥanbalīs
also accept lot-casting in such situations (Mawsūʿa, xvii, 138ff., 148f.), and the Imāmī
Shīʿites list many more; see Ḥusayn al-Karīmī al-Qummī, Qāʿidat al-qurʿa (Qum 1420
[1999]), 20 f.; Muḥammad Jawād Ashʿarī, Barrasī-yi ḥujjiyat-i qurʿa (Qum 1382 [2003]),
106 ff., 120.
92 Shāfiʿī, ‘K. al-qurʿa’, Umm, viii, 3; Bayhaqī, Aḥkām al-Qurʾān, 158.
93 ʿAbd al-Wahhāb b. ʿAlī al-Baghdādī, al-Ishrāf ʿalā masāʾil al-khilāf, ed. Ḥ. Ṭāhir (Beirut
1999), ii, 990 (no. 2005); al-Ṭūsī, al-Nihāya (Beirut 1970), 105ff.; Ashʿarī, Barrasī, 109. Some
Mālikīs rejected qurʿa if the slaves had been freed in death sickness (Ibn Rushd, Bidāya,
ii, 405 f. [K. al-ʿitq]; tr. Nyazee, ii, 450 f.). Compare the case of a man who divorces one
of his four wives and marries a fifth in death sickness without it being known which
of the four he had in mind: Yaḥyā b. Aktham (eventually classified as a Ḥanafī) would
let all five inherit and observe the ʿidda, the Ḥanbalīs and some Imāmīs would cast
lots for the one who had been divorced (Ibn Qayyim al-Jawziyya, Ṭuruq, ii, 744, 789;
Karīmī, Qāʿida, 21; Ashʿarī, Barrasī, 111). Cf. Mawsūʿa, xvii, for the Mālikī and Shāfiʿī solu-
tions.
94 Al-Sarakhsī, al-Mabsūṭ (Beirut 2001), xvii, 49 f.; al-Khaṣṣāf, Adab al-qāḍī, 391, no. 452;
36 chapter 2
which the Prophet cast lots to decide who should swear first (in the situation
443 in which two parties raise claims against | each other and both have to swear),
but the Ḥanafīs held that the judge should decide in most such situations.95
The reasoning is clearly that lots could not be used in situations in which all
claimants were entitled, but only some could be satisfied in full, or only one
person was entitled, but nobody knew who that person was: picking out the
lucky winners by lots amounted to gambling with their legal rights. Al-Shāfiʿī
also had reservations about lot-casting in the latter case, but Ḥanbalīs endorsed
it in both.96 Those who claimed that qurʿa amounted to gambling and had been
abrogated were ignorant, foul, or positively evil people, Ibn Ḥanbal said; they
had the temerity to label a Prophetic decision qimār.97 Polemicists who credit
Abū Ḥanīfa with the statement al-qurʿa qimār typically cast him as rejecting
the use of lots altogether. The Imāmīs are among them.98 According to them,
sortition was acceptable in all matters unknown (kullu majhūl fa-fīhi ʾl-qurʿa),
a principle they defend to this day.99
Attitudes to Lots
In the Old and New Testaments, too, all forms of lot-casting are consistently
envisaged as an appeal to the divine: God could see differences hidden to the
Baghdādī, Ishrāf, ii, 983 (no. 1993); Nawawī, Minhāj, iii, 440ff.; Mawsūʿa, xxxiii, 142f.; Ṭūsī,
Nihāya, 343 f.; Karīmī, Qāʿida, 105 ff.; Ashʿarī, Barrasī, 108; cf. F. Rosenthal, Gambling in Islam
(Leiden 1975).
95 Cf. Mawsūʿa, xxxiii, 147 f.
96 Cf. Ibn Taymiyya, Ṣiḥḥat uṣūl madhhab ahl al-Madīna (Beirut n.d. [1980?]), 85f. Our thanks
to Aron Zysow for drawing this work to our attention.
97 Ibn Qayyim al-Jawziyya, Ṭuruq, 742, 744 f., 747 f.
98 Thus Karīmī, Qāʿida, 18. He later notes that the Mawsūʿa shows Abū Ḥanīfa to have
accepted qurʿa in general, only to cite a barrage of stories in which Abu Ḥanīfa rejects the
Prophet’s precedent, including the latter’s use of qurʿa for choosing a wife to accompany
him on a journey (pp. 101 f.). Since the Prophet’s use of lots in connection with wives is a
situation in which the procedure was used to pick a winner, Abū Ḥanīfa may well in fact
have disliked this ḥadīth, but according to Sarakhsī (above, note 89), none of the wives
had any legal right to accompany him (whereas the slaves did have a legal right to such
freedom as the estate allowed by virtue of the bequest).
99 Ṭūsī, Nihāya, 345 f.; Majlisī, Biḥār, x, 203; xiv, 325; Ibn Ṭāwūs, Fatḥ al-abwāb, 272 (citing
Ṭūsī); Ashʿarī, Barrasī, 106; Muḥammad Ḥusayn Faḍl Allāh, al-Qurʿa waʾl-istikhāra (Beirut
1417/1997), 24 ff., against Abū Ḥanīfa, Ibn Abī Laylā and Ibn Shubruma at 27, 29; Karīmī,
Qāʿida, 34 f.
human eye; there are passages in which the outcome of lot-casting is explic-
itly equated with His will (1Sam. 10:24; Prov. 16:33; Acts 1:23–26). The Greeks
may once have thought in similar terms, though it has been argued that they
never did so | in connection with divisory lot-casting.100 Divisory lot-casting is 444
an expression coined by Thomas Aquinas for the use of lots to determine who
should have or do what, as opposed to consultative and divinatory lot-casting,
used to decide what to do and to obtain information about the future respec-
tively.101 From ancient times to late antiquity the Greeks seem to have envisaged
lot-casting of the divisory kind as a matter of chance, and the same is true of
the Romans.102 It was a matter of fortuna, as Justinian called it in his legisla-
tion.103 Their attitude affected their Hellenised Near Eastern subjects. Josephus,
for example, famously tells how the rebels at Masada chose ten men by lot to kill
the rest of them, and thereafter each other,104 and how he himself had used lots
to decide who, of his small band about to be captured by the Romans, should
kill whom first (he surrendered as one of the last to survive). He too seems to
think of the outcome as a matter of luck. He does put it to the reader that his
own survival could have been due to God’s providence rather than to chance,
but it sounds like mere self-justification.105
The Sunnī jurists generally seem to have thought of divisory lot-casting
(qurʿa) in much the same sober vein as their Greek and Hellenised predeces-
sors. Their attitudes must of course have varied in place and time and we cannot
claim to have studied them in any detail, but unlike Aquinas who (invoking
Augustine) identified all sortition as ‘a questioning concerning realities whose
occurrence depends on the divine will’, they convey little impression of seeing
the divisory form as an appeal to God. They make no attempt to distinguish it
100 Cf. N.D. Fustel de Coulanges, The Ancient City (New York n.d.; originally published Paris
1864), 182f. (book iii, ch. x); Pauly-Wissova, s.v. ‘Losung’, cols. 1461ff., mostly disagreeing
with Fustel de Coulanges and claiming that the Greeks distinguished between the lot as a
divine oracle and as a tool of equality from the start.
101 Thomas Aquinas, Commentary on the Ephesians, book 1, lecture 4, citing Proverbs 18:18
(‘The lots put an end to dispute’) in justification of the first. He put lot-casting for the
selection of people in the consultative rather than the divisory category. For other classi-
fications, see Gataker, Use and Nature of Lots, ch. 3.
102 Rosenstein, ‘Sorting Out the Lot’, esp. 51.
103 Cf. Justinian, above, note 20; also Favorinus (Ps.-Dio), above, note 15. Fortuna had once
been a goddess, but only in the sense that everything beyond human control could be
seen as divine.
104 Josephus, Wars, book 7, ch. 9, par. 1; cf. Y. Yadin, Masada: Herod’s Fortress and the Zealots’
Last Stand (London 1966), 201.
105 Josephus, Wars, book 3, ch. 8, par. 7.
38 chapter 2
from, or relate it to, consultation (istikhāra) or divination (istiqsām, kihāna),

445 apparently taking it for granted that they were | different; and those who clas-
sified qurʿa as gambling in some situations evidently thought of it as a matter
of chance. Their opponents did sometimes counter this by presenting it as an
appeal to the divine: one ḥadīth displays the Prophet as casting lots in a situa-
tion in which there would be winners and losers with the prayer, ‘O God, give
judgement among your servants with truth’, and Ibn Ḥanbal is credited with
the statement that ‘the lot hits the truth’ (al-qurʿa ṭuṣību ʾl-ḥaqq).106 But such
statements are rare in the Sunnī material we have seen.
Even on a superficial reading, the Shīʿites come across as different. Using lots
was indeed a way of delegating matters to God in their view,107 and particularly
effective if it was done by the imam: his qurʿa never went wrong, being in the
nature of waḥy, they said.108 The seventh/thirteenth-century Shīʿite scholar Ibn
Ṭāwūs did think of lot-casting as a form of consultative divination, istikhāra;109
and qurʿa and istikhāra are also treated together in booklets by contemporary
Shīʿites, including Faḍl Allāh, who repeats that lot-casting is a way of delegating
problems to God. He mentions unidentified persons who hold that only the
imam can do lot-casting, on the grounds that only he knows the special prayer
to be said in connection with it (an argument perhaps designed to eliminate
the whole institution), but he rejects it on the grounds that no special prayer
is needed. The method is only to be used when there is no other solution, he
says, and its purpose is simply to solve a problem, not the unveiling (kashf ) of
anything; but God does not cheat, as he also says.110
By way of contemporary comment, it may be worth noting that there has
been much interest in divisory lots as a political device in both England and
America in recent years.111 Most Westerners probably think of the procedure as
446 archaic, not so much because they see | it as a form of gambling or divination as
106 Saʿīd b. al-Musayyab’s ḥadīth in Sarakhsī, Mabsūṭ, xvii, 49 (with takhrīj); Ibn Ḥanbal in Ibn
Qayyim al-Jawziyya, Ṭuruq, ii, 745.
107 Thus several traditions in Majlisī, Biḥār, xci, 234; civ, 325.
108 Majlisī, Biḥār, ii, 177; xxvi, 32; xl, 245, 328, 363; liii, 331, 332, etc.
109 Ibn Ṭāwūs, Fatḥ al-abwāb, 267 ff.
110 Faḍl Allāh, Qurʿa, 26, 30, 33, 49, 62f., 65. For the question whether lot-casting is the
prerogative of the imam (as claimed in some traditions), see Ashʿarī, Barrasī, 56ff. For
lot-casting, istikhāra and istiqsām in another booklet, see Ḥusaynī, Qāʿida, 123ff.
111 E. Callenbach and M. Phillips, A Citizen Legislature (Berkeley 1985); K. Sutherland, The
Party Is Over: Blueprint for a Very English Revolution (Exeter 2004), revised as A People’s
Parliament (Exeter 2008); B. Goodwin, Justice by Lottery (Exeter 2005). Our thanks to
Anthony Barnett for these references.
because they think they can do better than random chance. (In fact, this seems
to have been the Ḥanafī attitude, too, but since the Prophet had endorsed
sortition, it was only via his prohibition of gambling that they could reject
it.)112 Even today, however, Westerners usually accept the principle of random
selection in connection with juries, which are still chosen by (computerised)
lot-casting, and it is precisely this principle that is attracting attention as a
way of introducing direct representation and popular control to counter what
nowadays goes under the name of the ‘democratic deficit’. As a democratic
device, random selection is what one book on the subject calls ‘the Athenian
option’,113 heartily disliked by a philosopher such as Ibn Rushd because it took
no account of virtue;114 but as an antidote to partiality and special interests in
general it was wholeheartedly endorsed in the Islamic legal tradition. Ancient
though the practice is, it may still be in for new roles, and not just in the West.115
The Return of the Near East
Here, however, our interest is not in modern politics, but rather in the relation-
ship between ancient Near Eastern and Islamic culture. The question has not
been much studied, but it has received some attention of late,116 deservedly in
our view, because it amounts to asking how far we can reconstruct the cultural
and religious history of the Near East as a single, continuous narrative rather
than as dis|jointed parts studied under the rubrics of Biblical, Greek, Roman, 447
ancient Iranian and Islamic history. Between them, the ancient and the Islamic
112 The explanation offered by Rosenthal, Gambling, 33, does not fit the contexts in which
qurʿa was identified as gambling.
113 Cf. A. Barnett and P. Carty, The Athenian Option: Radical Reform for the House of Lords
(Exeter 2008); cf. also O. Dowling, The Political Potential of Sortition (Exeter 2008), which
examines lot-casting as a political device in both Athens and the Western tradition.
114 Cf. P. Crone, Medieval Islamic Political Thought (Edinburgh 2004), 280 and note 111 thereto.
115 Curiously, a ballot or election is actually iqtirāʿ in modern Arabic (see H. Wehr, A Dictio-
nary of Modern Literary Arabic [Wiesbaden 1966], s.v.). Other words may be more com-
mon (notably intikhāb), but iqtirāʿ was used in the Iraqi election in 2005, see http://www
.newsday.com/news/nationworld/nation/ny-oiraqelectiongallery,0,322603.photo gallery
?index=7 (photo 2) [Ed.: This url is now defunct.].
116 Cf. S. Dalley (ed.), The Legacy of Mesopotamia (Oxford 1998); M. Levy-Rubin, ‘On the
Roots and Authenticity of Conquest Agreements in the Seventh Century’, Jerusalem
Studies in Arabic and Islam 34 (2008); and the melammu Project (www.aakkl.helsinki.fi/
melammu).
40 chapter 2
periods cover most of the history of the region, but not all of it: there is a thou-
sand years in between the two, and this is where the problem arises.
The thousand years in question are those in which the Near East was under
colonial rule, first under the Achaemenids, next under Alexander and his
successors, and thereafter under the Greeks and the Romans in its western
part, under the Parthians and the Sasanians in Iraq. As the foreigners moved
in with their own cultural traditions, the high culture of the Near East was
unseated and increasingly reduced to a local tradition of limited interest to
those who mattered. The ancient Near Eastern tradition did not die, of course.
It changed when it ceased to be written in cuneiform languages and was
expressed instead in Aramaic, but as Aramaic culture it lived on. Unfortunately,
very little of it has come down to us. We do have Jewish writings in Aramaic,
and from the third century ce onwards also Christian ones, but the pagans
who formed the vast majority in the region for most of the period have not
left us much. By and large, we are forced to study the Near East through the
eyes of its conquerors, who remained outsiders to the region in the sense that
they continued to be orientated towards their own cultural centres even after
having made themselves thoroughly at home in the land. As ill luck would have
it, the bulk of the Persian tradition is also lost, so that for practical purposes we
only have one pair of foreign eyes, those of the Greeks and the Romans. Some of
those who wrote in Greek were Near Easterners by origin, and some of them did
try to make their native tradition available in Greek, adapted to Greek tastes.
But the bulk of these writings is also lost, and most of the Near Easterners who
wrote in Greek had assimilated the hegemonic culture so thoroughly that they
sound no different from people of other origin writing in that language. The
Jews are again the main exception.
From the third century ce onwards, however, all this begins to change.
In 211 all members of the Roman empire were granted Roman citizenship
(some minor exceptions apart), with the result that all now had to live by
Roman law. Since people could not change their ways overnight whatever the
degree of Roman control, inevitably this meant that much of what they actu-
ally practised was a mixture of Roman and native law. Often called ‘provincial
law’, such native law surfaced in both the eastern and the western parts of
448 the empire, and some of it came to be officially endorsed as Roman | law.117
What this means for us is that the indigenous tradition begins to be visi-
ble in the hegemonic culture. The two Greek papyri from Petra and Nessana
117 See J. Mélèze Modrzejewski, ‘Diritto romano e diritti locali’, in A. Schiavone et al. (eds),
Storia di Roma, iii/2 (Turin 1993), 985–1009.
are perfect examples: the lot-casting by which the shares were allocated was a
provincial practice, not a procedure specified in Roman law.
Christianity made for even greater change. It originated as a Near Eastern
religion carried by speakers of Aramaic, initially Jews, thereafter Jews and
gentiles. A socially inclusive movement in which Greeks and non-Greeks, elite
and masses, were brought together in a manner hitherto unknown in the
Mediterranean, it gradually converted the entire empire to Near Eastern, if
increasingly Hellenised, modes of thought, and in the Near East itself it allowed
for a more extensive resurfacing of Aramaic culture as the Christians of Syria
and Mesopotamia took to writing in Syriac (i.e. the Aramaic dialect current at
Edessa). The establishment of a new capital in Constantinople also contributed
to the ‘Orientalisation’ of the Roman empire, to use the term adopted by those
who see the process from the Greek or Roman point of view. From our point of
view, ‘Orientalisation’ is simply a way of saying that it becomes possible to see
continuities outside the sphere of law as well.
The return of the Near East continued after the Arab conquest, for if Chris-
tianity was a kind of homecoming for the Near Eastern provincials, this was
even truer of Islam. The Arabs were Near Easterners who definitively unseated
the Greeks from their hegemonic role in the region. By then, of course, Greek
culture had served as the high culture of the Near East for close to a thousand
years, so that there was no way of shedding it: it had gone into the blood-
stream of the local culture. But living by Greek culture under the hegemony
of Greeks, who continued to see themselves as its ultimate arbiters even in
its Christian form, was quite different from continuing Greek cultural ways on
one’s own terms, with or without awareness of their Greek origin. Initially, of
course, the Arabs were much like the Greeks in that they saw themselves as
arbiters of Islamic culture, and they too were prejudiced against Aramaeans.
But their hegemonic position was shortlived. As converts to Islam, the Ara-
maeans assumed the legacy, and eventually also the ethnicity, of the Arab con-
querors and became their own cultural masters. When we speak of the Arabs
today, it is largely the former Aramaeans (and Copts) that we have in mind.
Consequently, a great deal of Islamic culture is Aramaic culture, | brought into 449
Islam in the form in which it had developed under Greek and Persian rule, to
develop in new directions thereafter.
This is the overall framework in which the connections between ancient
Near Eastern and Islamic culture have to be pursued. Lot-casting as an official
procedure provides us with a striking example of such a connection, with a
typically uneven distribution of documentation: well attested in the cuneiform
record, its only attestation in Aramaic seems to be in Jewish works. This is
presumably due to the loss of the pagan Aramaic tradition rather than the
42 chapter 2
disappearance of the practice, though it would help if it turned up in Syriac

too. As it is, however, we do have it in Greek, and as good luck would have it,
the Greek evidence comes from Petra and Nessana, which puts us in the rare
situation of having conclusive evidence for pre-Islamic Arabia. Thereafter the
evidence is plentiful, but only for the time of the Prophet, the Rāshidūn and the
Umayyads: as the Arab conquest society wanes, so do the attestations. We do
find discussion of the practice in Islamic law, but incidental references to the
practice in real life seem to disappear until its curious reappearance under the
Ottomans. Even the Jews eventually cease to mention it. Partition by lot-casting
is still discussed in the Gaonic literature, dating from c. 700–1050; but there is
no reference to it in the Kitāb al-mawārīth of Saʿadiya Gaon (d. 942), for all that
it covers inheritance issues in detail, nor do we know of any in the Cairo Geniza.
In short, the overall impression one gets is that what came out of Arabia was in
this case an institution that no longer meshed with the way things were done in
the rest of the Near East. It came and it went, leaving behind only some traces.
One may contrast this with another institution of ancient Near Eastern
origin in Islamic law, the clause requiring a freed slave to remain with his or her
master until the latter’s death, i.e. as a servant. Known as paramonē (‘remaining
by’), it was also found as a labour contract for free people. Originating as a
contract of adoption designed to provide for the manumitter in old age, it was
transmitted from the Near East to Greece at an early stage, and after Alexander’s
conquest of the Near East the indigenous and the Greek forms of the institution
interacted, to breed an amazing range of variations. The Romans accepted the
validity of such contracts when they were made by non-Roman subjects under
their own law, but not as part of Roman law. Inevitably, however, it came to be
practised under Roman law after the universal grant of citizenship, and though
the guardians of Roman law resisted this development, they may eventually
have capitulated. With or without official recognition, the paramonē remained
450 a prominent | part of provincial practice in the Near East. It was also known in
Arabia, where free slaves seem often to have been adopted very much as they
had been in the ancient Near East. We encounter the paramonē as a free labour
contract at Nessana and as an archaic requirement of staying with the master in
the Ḥijāz and elsewhere. After the conquests it is reflected in a wide variety of
forms in a large number of ḥadīths attributed to early jurists and the Rāshidūn,
and it formed the raw material of what the Muslims were to systematise as
kitāba and tadbīr.118
118 Cf. P. Crone, Roman, Provincial and Islamic Law (Cambridge 1987), ch. 5, and the literature
cited there.
If the contract had not been so important outside Arabia, it would presum-
ably have had much the same history as lot-casting: it would have come and
gone, leaving behind some traces. But far from receding into obscurity, it gen-
erated massive discussion and two new formal institutions. Manumission was
of course of much greater practical importance in daily life than lot-casting, so
the examples are not entirely comparable. For all that, it is hard not to suspect
that the key transmitters of originally ancient Near Eastern culture will prove to
be the inhabitants of the Fertile Crescent, now assisted by the Arabian tradition
and now without it, but not usually the Arabians on their own.
chapter 3
Idrīs, Atraḫasīs and al-Khiḍr*
Idrīs
Idrīs is a mysterious figure mentioned twice in the Qurʾān, in both cases in

somewhat unilluminating terms. One passage says of him and two others,
Ismāʿīl and Dhū ʾl-Kifl, that they were among the patient (min al-ṣābirīn) and
that ‘We admitted them to Our mercy, for they were among the righteous ones
(min al-ṣāliḥīn)’ (21:85f.). The other passage exhorts the believers to ‘remember
in the book Idrīs’, once again mentioning him after Ismāʿīl and now identifying
him as a righteous man and prophet (ṣiddīqan nabiyyan) whom God had ‘raised
to a lofty place’ (wa-rafaʿnāhu makānan ʿaliyyan) (19:56f.); the continuation,
perhaps added later, includes him among the prophets who were of the seed
(dhurriyya) of Adam, Noah, Abraham and Israel (19:58).1 It is clear, then, that
Idrīs was a prophet, a biblical figure and someone who had been raised up in
some sense, but it hardly suggests that he was Enoch, who lived before Noah,
Abraham, Israel/Jacob and Ishmael, though he was of course a descendant
of Adam. Nonetheless, it is usually as Enoch (Akhnūkh, Ḥanūkh), the great-
grandfather of Noah, that the exegetes identify him. Of Enoch Genesis twice
tells us that he ‘walked with God (hāʾelōhîm)’ (Gen. 5:22, 24), adding on the
second occasion that ‘then he was no more, because God (ʾelōhîm) took him’.
The Septuagint takes the first statement to mean that Enoch was pleasing to
God and the second that he was moved to heaven without dying, and this
became the standard interpretation (though there was also a tradition that he
died).2 As the Epistle to the Hebrews explains, quoting from the Septuagint,
* I should like to thank Tzvi Abusch, Michael Cook and Adam Silverstein for helpful comments
on earlier drafts of this article, and Tommaso Tesei for a memorable conversation about the
subject.
1 For 19:58 as an interpolation (by the Messenger himself?), see A. Neuwirth, ‘Imagining Mary—
Disputing Jesus’, in B. Jokisch, U. Rebstock and L.I. Conrad (eds.), Fremde, Feinde und Kurioses,
Berlin and New York 2009, pp. 383–416, at 389 f.
2 Thus Jubilees, 7:39, as against 4:23 (tr. O.S. Wintermute in J.H. Charlesworth (ed.), Old Tes-
tament Pseudepigrapha, New York 1983–1985, ii, pp. 71, 63). Targum Onqelos unambigu-
ously declares God to have made Enoch die, though a variant denies it (the translation of
J.W. Etheridge, London 1862, reflects the variant); and both Neofiti and Ps.-Jonathan say that
Enoch was taken away, using a verb that can also mean to die (ʾtngd). For all that, both seem to

idrīs, atraḫasīs and al-khiḍr 45
‘by faith Enoch was taken so that he did not experience death and he was not
found because God had taken him. For it was attested before he was taken
that he had pleased God’.3 The exegetes often take q. 19:57 to refer to Enoch’s
translation: the lofty place to which God had raised him was the fourth or
sixth heaven; he had been moved there without having died, like Jesus, and he
was immortal; Muḥammad met him in the fourth heaven during his heavenly
journey.4
Some exegetes took the Qurʾānic statement that God had raised Idrīs to
a lofty place to mean that He had raised him in terms of rank and status
rather than location.5 This was the view of al-Ḥasan al-Baṣrī (d. 110/728), al-
Jubbāʾī (303/915f.) and Abū Muslim al-Iṣfahānī (d. 322/934), for example.6 But
according to al-Ḥasan al-Baṣrī, the earliest exegete from whom this view is
transmitted, it was in paradise ( fī ʾl-janna) that Idrīs’ rank had been raised.7 In
other words, al-Ḥasan probably shared the belief that Idrīs had been translated
to heaven (or conceivably to some inaccessible place on earth). In principle,
he could have believed that Idrīs died here on earth to be first resurrected and
next moved to paradise after the fashion of Jesus according to the Christians;
but the Qurʾānic Jesus is not said to have been resurrected before ascending
to heaven: either he died a normal death on earth and will be resurrected
along with the rest of mankind (19:33, cf. 19:15) or else he was taken to heaven
as soon as he died (cf. 3:55; 5:117).8 Muslim martyrs are killed here on earth
have his translation in mind; Ps.-Jonathan even adds that Enoch ascended to the firmament
to be known as Metatron (see M. Maher (tr.), Targum Pseudo-Jonathan: Genesis, Collegeville,
mn, 1992, p. 37, n. 8; M. McNamara (tr.), Targum Neofiti 1: Genesis, Collegeville, mn, 1992, p. 70,
n. 11; L.R. Ubigli, ‘La fortuna di Enoc nel giudaismo antico: valenze e problemi’, Annali di Sto-
ria dell’Esegesi 1, 1984, pp. 153–163, at 156f.; J.C. VanderKam, Enoch: a Man for All Generations,
Columbia, sc, 1995, pp. 165 ff.).
3 Hebrews 11:5; similarly Sirach 44:16; 49:14; Philo, De mutatione nominum, 34, 38 (with allegor-
ical interpretation); Josephus, Antiquities, i, 4:85 (cf. ix, 2:28).
4 Al-Ṭabarī, Jāmiʿ al-bayān ʿan tafsīr al-Qurʾān, Beirut 1988, juzʾ xvi, pp. 96f., citing al-Mujāhid,
al-Ḍaḥḥāk and others; al-Ṭabrisī, Majmaʿ al-bayān fī tafsīr al-Qurʾān, Beirut 1995, vi, p. 430,
citing the same and other authorities; al-Suyūṭī, al-Durr al-manthūr, Beirut 1983, v, p. 519, all
ad 19:57. Further references in Encyclopaedia of Islam2, Leiden 1960–2009 (hereafter ei2), s.v.
ʿIdrīs (Vajda); cf. also Encyclopaedia of the Qurʾān, Leiden 2001–2006 (hereafter eq), s.v. ʿIdrīs’
(Erder).
5 Cf. al-Māturīdī, Taʾwīlāt al-Qurʾān, ed. B. Topaloğlu and others, Istanbul 2005–2011, ix, pp. 147f.
(ad 19:57), where this view is preferred.
6 Ṭabrisī, Majmaʿ, vi, p. 430.
7 Māturīdī, Taʾwīlāt, ix, p. 147, ad 19:57.
8 Differently N. Robinson in eq, s.v. ‘Jesus’.
46 chapter 3
and go straight to heaven without being resurrected first (evidently because

this would have required them to be seen alive again before their transfer to
heaven); but although this view is attested already in the Qurʾān (2:154; 3:169),
the martyrs never seem to figure in the discussion of either Jesus or Enoch.
There were certainly scholars who denied Idrīs’ immortality, Ibn Isḥāq and
Muqātil b. Sulaymān (both d. 150/767) among them. Ibn Isḥāq envisaged Idrīs
as dying here on earth, but he says nothing about him being raised to heaven,
with or without being resurrected first.9 But Muqātil and others accepted that
Idrīs was alive when he went to heaven, for in their view it was in the fourth
heaven, or on the way to the fifth, that the angel of death had taken him (an idea
which generated some gripping stories).10 In short, there was almost complete
agreement among the early exegetes that Idrīs was Enoch and that he had gone
to heaven/paradise without dying first.
Needless to say, the agreement was not complete. Ibn Isḥāq (and others?)
apart, there were also exegetes who identified Idrīs with Elijah, another prophet
who had been translated to heaven (2Kings 2:5). Ibn Masʿūd, for example, held
that Idrīs’ name was Elijah and read Idrīs for Elijah in two of the three passages
in the Qurʾān in which Elijah (Ilyās) is mentioned.11 (What he said about the
third does not seem to be recorded.) Elijah in his turn was often identified with
al-Khiḍr/al-Khaḍir, yet another mysterious figure characterised by immortality,
and apparently Idrīs was sometimes held to be al-Khiḍr as well.12 Since Enoch
was believed to be a scribe in heaven no less than on the earth,13 he also came to
be identified with the heavenly scribe of the Egyptians, Thoth, under the latter’s
Greek name of Hermes; and though the identification of the two is likely to
predate the rise of Islam, it seems to be only in Islamic sources that it is attested:
9 Ṭabarī, Taʾrīkh al-rusul waʾl-mulūk, ed. M.J. de Goeje and others, Leiden 1879–1901, ser. i,
p. 176.
10 Muqātil b. Sulaymān, Tafsīr, ed. ʿA.M. Shiḥāta, Beirut 2002, ii, p. 631; Māturīdī, Taʾwīlāt,
ix, p. 147; cf. the neat survey of al-Māwardī, Tafsīr, ed. Kh.M. Khiḍr, Kuwait 1982, ii, p. 529;
Ṭabarī, Jāmiʿ, juzʾ xvi, pp. 96 f. For the gripping stories, see al-Qummī, Tafsīr, Beirut 1991, ii,
pp. 25 f. (here combined with a version of the fallen angels theme); Suyūṭī, Durr, v, pp. 518ff.
(where one story also works in the fallen angels).
11 Ṭabarī, Jāmiʿ, juzʾ xxiii, p. 96, ad 37:130; cf. juzʾ vii, p. 261, ad 6:85; Suyūṭī, Durr, vii, p. 117,
ad 37:123, where both Ibn Masʿūd and Qatāda identify Idrīs and Elijah. Further references
in Encyclopaedia of Islam3, Leiden 2007– (hereafter ei3), s.v. ‘Elijah’ (Rippin).
12 Cf. ei2, s.v. ‘al-Khaḍir (al-Khiḍr)’ (Wensinck); and eq, s.v. ‘Khaḍir/Khiḍr’ (Renard). The
occasional identification of Idrīs and al-Khiḍr is reported by Vajda in ei2, s.v. ‘Idrīs’.
13 Cf. D.E. Orton, The Understanding Scribe: Matthew and the Apocalyptic Ideal, Sheffield
1989, pp. 77 ff.; A.A. Orlov, The Enoch-Metatron Tradition, Tübingen 2005, pp. 50ff.
the earliest author to mention it appears to be al-Jāḥiẓ (d. 255/869).14 It reflects

the thinking of astrologers, alchemists and others to whom Hermes counted as
an authority, not that of theologians or religious scholars. To the latter, Idrīs was
practically always a biblical figure, and almost always Enoch.
Modern scholars sometimes dispute that their predecessors are right. To
Alexander, for example, it seems abundantly clear that the Qurʾān is not, in
fact, referring to Enoch, given that the names Idrīs and Enoch have nothing in
common, that no plausible link between the two figures has been proposed,
and that the statement ‘We raised him to a lofty place’ need not refer to an
ascent to heaven.15 All Alexander’s objections are right, or rather were, for the
first two fall away if the argument presented here is correct; and as we have
seen, even the exegetes who held God to have elevated Idrīs in terms of honour
and status rather than by translation to heaven identified Idrīs as Enoch. A
more serious objection might be that the Qurʾān never identifies Idrīs as an
antediluvian figure, but on the contrary lists him after much later figures in
sura 21:85, and after Ismāʿīl in both that passage and 19:56f. But disregard of
the chronological order is quite common in the Qurʾān: 6:84–86, for example,
enumerates David, Solomon, Job, Joseph, Moses, Aaron, Zakariyah, John, Jesus,
Ishmael, Elisha, Jonah and Lot in that order. Another objection might be that
the Qurʾān associates Idrīs with endurance or patience (ṣabr), as opposed to
repentance, the characteristic with which Enoch had come to be associated.16
But Idrīs does at least share his righteousness with Enoch,17 and Reeves deems
the Qurʾānic passage on his elevation to a lofty place to be arrestingly close
to that used in Jubilees and 1Enoch of Enoch’s transfer to Eden/a lofty place.18
Above all, the exegetes could hardly have been so united in their opinion at
so early a stage if they had first encountered Idrīs as a Qurʾānic figure open
14 Jāḥiẓ, K. al-tarbīʿ waʾl-tadwīr, ed. C. Pellat, Damascus 1955, par. 40: ‘Tell me about Hermes,
is he Idrīs?’.
15 Cf. P.S. Alexander, ‘Jewish Tradition in Early Islam: the Case of Enoch/Idrīs’, in G.R. Hawt-
ing, J.A. Mojaddedi and A. Samely (eds.), Studies in Islamic and Middle Eastern Texts and
Traditions in Memory of Norman Calder, Oxford 2000, pp. 11–29, at 23f.
16 Cf. Sirach 44:16; cf. also Philo, De Abrahamo, 17 f.
17 E.g. Sirach 44:17; 1 Clement (The Letter of the Romans to the Corinthians, ed. and tr. in The
Apostolic Fathers, ed. J.B. Lightfoot, ed. and rev. M.W. Holmes, Grand Rapids, mi, 1992), 9:3
(dikaios); 1 Enoch (Book of Parables), passim (ṣādeq), discussed in Orlov, Enoch-Metatron
Tradition, pp. 77 f.
18 J.C. Reeves, ‘Some Explorations of the Intertwining of Bible and Qurʾān’, in J.C. Reeves (ed.),
Bible and Qurʾān: Essays in Scriptural Intertextuality, Atlanta 2003, pp. 43–60, at 47, with
reference to 1 Enoch, 87:3; Jubilees, 4:23 (in Charlesworth, Old Testament Pseudepigrapha,
ii, pp. 62 f.).
48 chapter 3
to identification as any biblical prophet; on the contrary, one would in that

case have expected some to dispute the identification on the grounds that
the Qurʾānic description of Idrīs as ‘of Abraham and Israel’s seed’ does not fit
the antediluvian Enoch. But no early exegete makes this point; all apparently
took the statement to mean no more than that Idrīs formed part of the line
of biblical prophets and patriarchs from Adam to Jesus of which Abraham
and Jacob/Israel were also members. They never reached any comparable
agreement on the mysterious Dhū ʾl-Kifl, with whom Idrīs is mentioned in
q. 21:85,19 and the early exegetical discussion of this verse is dominated by the
problem of who he could be, not by Idrīs, whose identity with Enoch seems to
be taken for granted. The commentators did not recognise Dhū ʾl-Kifl, but Idrīs
they knew, presumably on the basis of the same tradition as the Qurʾān.
What then was this tradition? Our only clue is the name Idrīs under which
Enoch was held to be known. The etymology of this name is an old problem.
Albright thought that it was derived from Poimandrēs, the name of a Hermetic
treatise and, according to Albright, also of Thoth/Hermes himself;20 Gil, playing
by philological rules all his own, derived Idrīs from the very name Hermes (Hīr-
mīs, Hirmis);21 and Erder connected Idrīs with dōresh ha-torah, the ‘expounder
of the law’ mentioned in the Qumran literature, noting that both the latter and
Idrīs were identified with Hermes.22 It is hard to take the first two suggestions
seriously, and the third falls on the fact that dōresh could neither develop into
Idrīs nor be translated as such. Casanova derived Idrīs from Esdras, the Greek
form of Ezra, via scribal mistakes, but this does not fit the long vowel in Idrīs,
nor did he explain why it was with Enoch rather than Ezra that the exegetes
identified him.23 Philologically speaking, by far the best suggestion was made
back in 1903 by Nöldeke, who observed that Idrīs could easily be derived from
the Syriac forms of Greek Andreas (ʾndrys and the like). But Nöldeke failed to
explain why Andrew, one of the twelve apostles, should have been singled out
19 Cf. ei3, s.v. ‘Dhū l-Kifl’ (Rippin).

20 W.F. Albright, review of P. Boylan, Thot, the Hermes of Egypt, in Journal of the Palestine
Oriental Society 2, 1922, pp. 197f.; id., ‘Islam and the Religions of the Ancient Orient’, Journal
of the American Oriental Society 60, 1940, pp. 283–301, p. 287n.
21 M. Gil, ‘The Creed of Abū ʿĀmir’, Israel Oriental Studies 12, 1982, pp. 9–57, at 35.
22 Y. Erder, ‘The Origin of the Name Idrīs in the Qurʾān: a Study of the Influence of Qumran
Literature on Early Islam’, Journal of Near Eastern Studies 49, 1990, pp. 339–350; succinctly
also in eq, s.v. ‘Idrīs’ (Erder).
23 P. Casanova, ‘Idrîs et ʿOuzaïr’, Journal Asiatique 205, 1924, pp. 356–360. Thus also C.C. Tor-
rey, The Jewish Foundation of Islam, New York 1933, p. 72; followed by Alexander, ‘Jewish
Tradition in Early Islam’, p. 23.
for special attention as a prophet in the Qurʾān or why the exegetes should
have understood this apostle as Enoch.24 Hartmann improved on Nöldeke’s
suggestion by proposing that the relevant bearer of the name Andreas was not
the apostle, but rather the cook who inadvertently acquires immortality in the
Alexander Romance.25 Here at last the name is associated with immortality, but
Andreas the cook is a shifty character who uses his possession of the water of
immortality to seduce Alexander’s daughter and who is punished by transfor-
mation into a sea demon.26 He could not have gone into the Qurʾān as a prophet
and righteous Israelite, nor could he have blended with Enoch outside it. Hart-
mann’s thesis is also open to improvement, however. The suggestion offered
here is that a still unattested Aramaic name lies behind both Greek Andreas and
Arabic Idrīs and that the name in question was that of Atraḫasīs, the ancient
Mesopotamian king who survived the flood and acquired immortality. That
Atraḫasīs is the ultimate source of the name Idrīs has in fact been suggested
before, by the Aramaicist Montgomery, but he presented his case much too
briefly and superficicially to carry conviction, or even to be mentioned there-
after.27 In fact, since we do not know how ‘Atraḫasīs’ was rendered in Aramaic,
there is no way of proving that it is in fact his name that lives on in that of Idrīs;
but as this article will try to persuade the reader, it can at least be shown to be
plausible.
Atraḫasīs
Atraḫasīs (‘exceedingly wise’) is the epithet of several ancient Mesopotamian

mythological figures, including the Mesopotamian flood survivor.28 He appears
under that name in the Akkadian epic known as Atraḫasīs. Here we are told
that humans were created to do all the hard labour of feeding, clothing and
24 Th. Nöldeke, ‘Idrīs’, Zeitschrift für Assyriologie 17, 1903, pp. 83f.
25 R. Hartmann, ‘Zur Erklärung von Sūre 18, 59 ff.’, Zeitschrift für Assyriologie 24, pp. 307–315,
at 314 f. For the Alexander tradition in Arabic, see F. Doulfikar-Aerts, Alexander Magnus
Arabicus, Leuven 2010.
26 Pseudo-Callisthenes, ed. C. Müller, Paris 1846, book ii, chs. 39, 41; The Greek Alexander
Romance, tr. R. Stoneman, London 1991.
27 J.A. Montgomery, ‘Some Hebrew Etymologies’, Jewish Quarterly Review, ns, 25, 1935,
pp. 261–269, at 261. The only scholar to mention his idea seems to be A. Jeffery, The Foreign
Vocabulary of the Qurʾān, Baroda 1938, p. 52.
28 Reallexicon der Assyriologie, ed. E. Ebeling, B. Meissner and others, Berlin and Leipzig 1931–
2008, i, s.v. ‘Atraḫasîs(a)’ (Jensen).
50 chapter 3
housing the gods that the gods (or rather some of them) had formerly done
themselves. This worked well except that humans grew so numerous that their
noise became intolerable. Enlil, the chief of the gods, could not sleep. He
tried to reduce their number with plagues and droughts, but eventually he
and other gods decided to send a flood. On the advice of Enki, a somewhat
mischievous deity, Atraḫasīs managed to escape by building a special boat,
which also carried his family and animals of all kinds. In short, Atraḫasīs is here
the Mesopotamian Noah, not Enoch.29
In the Sumerian deluge story the flood survivor is called Ziusudra (later Zisu-
dra),30 the name under which he also appears in one version of the Sumerian
king list31 and the Babyloniaca of Berossos (fl. c. 290 bc), the Babylonian priest
who made his ancestral heritage available in Greek. Berossos’ work is lost, but
excerpts survive,32 and Zi(u)sudra’s name is here transliterated as Xisouthros,33
Sisithros,34 and perhaps also Sisythes.35 Ziusudra differs from Noah in that the
29 W.G. Lambert and A.R. Millard (eds. and trs.), Atra-ḫasîs, the Babylonian Story of the Flood,
Oxford 1969 (repr. Winona Lake, in, 1999); also in B.R. Foster (tr.), Before the Muses: an
Anthology of Akkadian Literature, 3rd ed., Bethesda, md, 2005, pp. 229ff.
30 M. Civil (ed. and tr.), ‘The Sumerian Flood Story’, lines 254–260, in Lambert and Millard,
Atra-ḫasîs, p. 145; previously translated by S.N. Kramer in J.B. Pritchard (ed.), Ancient
Near Eastern Texts relating to the Old Testament, Princeton, nj, 1955, p. 44 (with a slightly
different line numbering). For the forms of the name, see A.R. George (ed. and tr.), The
Babylonian Gilgamesh Epic, Oxford 2003, i, p. 154n.
31 Cf. S. Langdon, ‘The Chaldean Kings before the Flood’, Journal of the Royal Asiatic Society
1923, pp. 251–259; T. Jacobsen (ed. and tr.), The Sumerian King List, Chicago 1939, p. 76,
n. 34 (on wb 62); cf. also id., ‘The Eridu Genesis’, Journal of Biblical Literature 100, 1981,
pp. 513–529, at 520 (on ct 46.5). Ziusudra was probably also mentioned in a portion of
the California tablet (ucbc 9–1819), now too damaged to be read (J.J. Finkelstein, ‘The
Antediluvian Kings: a University of California Tablet’, Journal of Cuneiform Studies 17, 1963,
pp. 39–51).
32 They are collected in G.P. Verbrugghe and J.M. Wickersham, Berossos and Manetho, Intro-
duced and Translated, Ann Arbor, mi, 2001, pp. 43 ff.
33 Verbrugghe and Wickersham, Berossos and Manetho, pp. 47–50. It is the form used by
Apollodorus, Alexander Polyhistor and Abydenos as preserved in Eusebius’ Chronicle (in
Armenian), pp. 4–6 (also in Cyril of Alexandria, Contre Julien, ed. and tr. P. Burguière and
P. Évieux, i, Paris 1985, book i, 6–8) and by Syncellus, Ecloga Chronographica, pp. 53–56.
34 This form appears in Eusebius’ version of Abydenos in his Praeparatio Evangelica, book ix,
12, 2; but cf. above, note 33.
35 The De Dea Syria attributed to Lucian refers to a myth about the flood of Deukalion
ton Skythea, ‘Deukalion, the Scythian’ (par. 12: Deukalion is the Greek Noah). This was
emended to ton Sisythea by Buttmann in 1828, and the emendation has been so widely
accepted that its conjectural status is often forgotten. Though ‘the Scythian’ is problem-
gods reward him for his role in the preservation of mankind by granting him
eternal life. They do not do this by taking him up to heaven, but rather by
sending down life and eternal breath ‘like [that of] a god’ to him and mak-
ing him dwell in ‘the land of the crossing, the land of Dilmun, the place where
the sun rises’.36 It is in some such mysterious place that Gilgamesh seeks him
out in the Gilgamesh epic. Here the hero, who has been wandering far and
wide in search of immortality, builds a boat on which he reaches the sea of
death which only Shamash can traverse, but which he nonetheless succeeds in
crossing. He reaches the immortalised human, who is here called Ūta-napishti
(or Utnapishtim) the Faraway, except for two passages in which he appears as
Atraḫasīs.37 (Ūta-napishti is an Akkadian interpretation of Ziusudra meaning
‘he [or I] found life’.)38 Ūta-napishti, alias Atraḫasīs, imparts a mystery of the
gods’ to Gilgamesh, meaning knowledge normally beyond the reach of human
beings, by telling him how he was divinely instructed to build the boat on which
he preserved the seed of all living things and how the gods made him and his
wife immortal and made them dwell ‘far away, at the mouth of the rivers’. Again
it is clear that to acquire immortality is to become like a god, but also that it did
not amount to deification: Ūta-napishti remains a human being.39 He tells Gil-
gamesh that if he too is to become immortal, he must start by staying awake for
six days and seven nights, which naturally Gilgamesh fails to do. By way of con-
solation Ūta-napishti shares another ‘mystery of the gods’ with him, namely the
existence of a plant that will rejuvenate him.40 Gilgamesh dives into the deep
and brings up the plant, but while he is bathing in a well a snake eats it and
sloughs its skin. Gilgamesh must grow old and die like everyone else.41
atic, it is preserved in J.L. Lightfoot (ed. and tr.), Lucian on the Syrian Goddess, Oxford 2003,
p. 252, with arguments against the emendation at pp. 342f. It is worth nothing, however,
that the form Sisythes makes sense to George, Gilgamesh Epic, i, p. 154n, as a reflection of
the form Zisuddu. It would imply that Lucian (if he is indeed the author of the Dea Syria)
was familiar with Ziusudra independently of Berossos.
36 Tr. Kramer in Pritchard, Ancient Near Eastern Texts, p. 44; cf. Civil, ‘Sumerian Flood Story’,
in Lambert and Millard, Atra-ḫasîs, p. 145.
37 Gilgamesh epic (standard Babylonian version), tablet xi, lines 49, 197. My transliteration
is George’s.
38 Thus George, Gilgamesh Epic, i, p. 153.
39 Cf. Gilgamesh epic, tablet xi, lines 1–7, 204.
40 It is not clear whether the plant will confer eternal youth, and thus immortality, or
rejuvenation repeatable on further ingestion, and thus immortality again, or just one
rejuvenation. But the narrator probably did not have give thought to these distinctions.
41 Gilgamesh epic, tablet xi, lines 281 ff. For the snake eating the drug against old age in Greek
myth, see W. Burkert, The Orientalizing Revolution, Cambridge, ma, 1992, p. 123.
52 chapter 3
In short, Atraḫasīs (aka Ziusudra, Ūta-napishti) is not just the Mesopota-

mian Noah, but also a figure who was granted immortality like Enoch. This is
what could have enabled him to blend with Enoch. In fact, he may even have
contributed to Enoch’s genesis:42 this is suggested above all by the fact that both
Enoch and Noah are said to have ‘walked with God’ (Gen. 6:9), an expression
which is not used of any other antediluvian figure.43 More commonly, however,
Enoch is seen as a reflection of another Mesopotamian figure, the antediluvian
king Enmeduranki.44 The Sumerian king list shares with Genesis the feature
of narrating the early history of mankind as a succession of enormously long-
lived worthies followed by a flood, and there are ten figures in some versions of
the king list, as also in Genesis. In several versions of the king list the seventh
is Enmeduranki, and in Genesis the seventh is Enoch. Of Enmeduranki we are
told that he ‘sat in the presence of Shamash and Adad’, and that Shamash and
Adad brought him into their assembly, seated him on a throne of gold and
taught him the art of divination.45 We also hear of a sage in Enmeduranki’s
time by the name of Utuabzu, ‘who ascended to heaven’ for initiation into
heavenly secrets.46 One way or the other, then, the seventh generation was
associated with ascent to heaven. But the Sumerian king list and the biblical
genealogies also differ in significant ways, and Enoch does not quite match
his alleged Mesopotamian counterpart.47 Enmeduranki’s presence in heaven
42 For the view that he is lurking in the background, cf. J.J. Collins, The Apocalyptic Imagi-
nation: an Introduction to Jewish Apocalyptic Literature, 2nd ed., Grand Rapids, mi, 1998,
p. 46.
43 Noted by H.S. Kvanvig, Roots of Apocalyptic: the Mesopotamian Background of the Enoch
Figure and of the Son of Man, Neukirchen-Vluyn 1988, pp. 93, 230. For the view that Enoch
could have been the flood hero in the hypothetical flood story in the j stratum, see
E.G. Kraeling, ‘The Earliest Hebrew Flood Story’, Journal of Biblical Literature 66, 1947,
pp. 279–293, at 291 f.
44 Cf. J.C. VanderKam, Enoch and the Growth of an Apocalyptic Tradition, Washington, dc,
1984, pp. 33 ff., with discussion of the earlier literature in ch. 1; Kvanvig, Roots, pp. 185ff.,
230 ff.; Orlov, Enoch-Metatron Tradition, ch. 1.
45 Untitled text dating, probably, from the time of Nebuchadnezzar i (c. 1126–1103bc) edited
and translated in W.G. Lambert, ‘Enmeduranki and Related Matters’, Journal of Cuneiform
Studies 21, 1967, pp. 126–138, at 130, 132.
46 R. Borger, ‘Die Beschwörungsserie Bīt Mēseri und die Himmelfahrt Henochs’, Journal of
Near Eastern Studies 33, 1974, pp. 183–196 (esp. 192 f.).
47 For objections to the assumption of Mesopotamian influence, see C. Westermann, Gen-
esis, Neukirchen-Vluyn 1971, pp. 474ff., 484 ff. (countered in Kvanvig, Roots, pp. 224ff.);
G.F. Hasel, ‘The Genealogies of Gen 5 and 11 and Their Alleged Babylonian Background’,
Andrews University Seminary Studies 16, 1978, pp. 361–374; T.C. Hartman, ‘Some Thoughts
was temporary and the same was probably true of Utuabzu’s, whereas Enoch’s
translation to heaven was permanent. Of course we should not envisage the
priests behind Genesis 5 as working directly with Mesopotamian writings,
as opposed to using motifs and themes of Mesopotamian origin that they
would put together as they saw fit,48 but the motif of ascent to heaven for
a visit is rather different from that of permanent translation, and it of Ūta-
napishti/Atraḫasīs that we hear that the gods ‘took’ him.49 As will be seen, there
were Jewish readers of the Hellenistic period who took the Genesis passage
on Enoch to refer to both temporary ascent à la Enmeduranki and permanent
translation to a remote place on earth à la Atraḫasīs; but by their time there
was a Jewish diaspora to the east of the Euphrates, and it was probably to this
diaspora that they owed their views.
However we are to envisage the origin of the Enoch figure, it is with the
formation of the Jewish diaspora in Mesopotamia that the development of
interest to us begins. In the eighth century bc the northern tribes of Israel
were deported to Assyria, and in the sixth century bc their southern counter-
parts followed them to Babylonia. Enoch and Atraḫasīs, two possibly related
and certainly very similar figures, now came to coexist in Mesopotamia, and
inevitably they interacted there. Initially it may only have been in the minds of
the Jewish captives that they did so, for it was the Jews who had to find ways
of harmonising their native tradition with the more prestigious culture of their
imperial overlords, whereas their overlords could ignore Enoch. As will be seen,
however, the interaction seems eventually to have affected the Mesopotamians
themselves, and it certainly came to do so when they converted to Christianity,
for now it was they who had to find ways of harmonising their own tradition
with that of the Bible. In short, from the exilic period onwards the stage was set
for Enoch and Atraḫasīs to merge.
Unfortunately we cannot follow the interaction between them directly. The
last datable cuneiform tablet was written in 75ad,50 but by then the main
on the Sumerian Kinglist and Gen 5 and 11b’, Journal of Biblical Literature 91, 1972, pp. 25–
32; J.R. Davila, ‘The Flood Hero as King and Priest’, Journal of Near Eastern Studies 54, 1995,
pp. 199–214.
48 Cf. George, Gilgamesh Epic, i, pp. 56 f. Davila, ‘Flood Hero as King and Priest’, p. 211, argues
against direct adaptation of the Mesopotamian lists, but whether direct use has actually
been advocated is not clear to me.
49 Gilgamesh epic, tablet xi, line 206. The verbal root in Akkadian is leqû, that of the Hebrew
text lāqah, cf. K. Luke, ‘The Patriarch Enoch’, Indian Theological Studies 23, 1986, pp. 125–
153, at 133.
50 M. Geller, ‘The Last Wedge’, Zeitschrift für Assyriologie 87, 1997, pp. 43–95, at 46.
54 chapter 3
literary language of Mesopotamia had been Aramaic for some five centuries,
and the bulk of the Aramaic tradition is lost. Of the mythological literature
of the pagan Aramaeans nothing survives except for occasional reflections in
Greek, Jewish and Syriac works. (It undoubtedly left plenty of marks on Persian
literature, but as ill luck would have it, the pre-Islamic Persian tradition is also
largely lost.) These reflections do not tell us anything about Enoch or Atraḫasīs,
but they do allow us to connect Atraḫasīs, Andreas the cook and al-Khiḍr.
We also have Jewish sources, however, and we can observe the post-biblical
transformation of Enoch in the Enoch literature, above all 1 Enoch and Jubilees.
Since the bulk of this literature originated on the Greek side of the Euphrates,
it does not show us how Enoch and Atraḫasīs interacted in Mesopotamia, but
it does allow us to see that Atraḫasīs exercised a magnetic pull on Enoch even
in Palestine. One paragraph in Berossos’ Babyloniaca, moreover, suggests that
Enoch’s pull on Atraḫasīs was no less significant in Mesopotamia. I shall deal
with the material relating to Enoch first.
Enoch and Atraḫasīs
The texts that make up 1 (Ethiopic) Enoch are attributed to Enoch and were
composed in Aramaic at different times between the third or second century
bc and the turn of the era, though they survive in full only in Ethiopic.51 The
texts are apocalypses, or in other words revelations about the past, present and
future of the world, and like most apocalypses they envisage the future as a
violent end to the world in which all sinners are horribly punished and the
righteous rewarded. The story of the flood is repeatedly told and alluded to as
a prototype of the final punishment ahead. The book expands on an enigmatic
passage in Genesis 6:1–2 according to which sons of God mated with daughters
of men and sired giants by them. The sons of God are understood as angels of
the kind called ‘watchers’, and their giant offspring have such trouble satisfying
their enormous appetites that they end up eating people, and even each other,
and drinking their blood. God responds to these developments first by sending
angels to bind the watchers and to induce the giants to destroy themselves
in internecine wars, and next by unleashing the flood to cleanse the earth.52
51 It is used here in the translation of G.W.E. Nickelsburg and J.C. VanderKam, 1Enoch,
Minneapolis 2004. They date its earliest part to the late fourth century bc, perhaps a mere
slip (p. vii).
52 For the Mesopotamian antecedents of this, see S. Bhayro, ‘Noah’s Library: Sources for
1 Enoch 6–11’, Journal for the Study of the Pseudepigrapha 15, 2006, pp. 163–177.
All this is told without reference to Enoch. The latter is introduced as a scribe
whom God sends to the watchers with a decree ordering their punishment
and whom the watchers send back again to God with a petition for forgiveness
(which is rejected); and from this point onwards the texts are about Enoch and
his visions.
Jubilees, on the other hand, is a work composed in Hebrew, perhaps around
125bc,53 and it too is extant only in Ethiopic. It is a retelling of Genesis and the
beginning of Exodus presented as the full revelation received by Moses at Sinai,
and it too covers both Enoch and the flood, if much more briefly than 1 Enoch.
In these works Enoch is reshaped along the lines of Atraḫasīs in three main
ways. First, he has come to be linked with the flood. He is not connected with
it in Genesis, nor would one expect him to be, given that he represents the
seventh rather than the tenth generation after Adam. (There is no attempt
to link Enmeduranki with the flood on the Mesopotamian side.) In 1 Enoch,
however, Enoch is associated with the flood partly by participation in the story
of the watchers who cause it to be unleashed and partly by repeatedly receiving
advance warning of it. An angel tells him that God will open all the chambers
of the waters above the heavens and all the fountains beneath the earth, and
that all dwellers on the earth will be obliterated.54 He sees in a vision that
the earth will sink into the abyss and be utterly destroyed, and he reacts by
imploring God together with his son Methuselah that a remnant of mankind
might remain upon the earth.55 In another passage he predicts ‘the first end’
(as opposed to the day of judgement), here fully aware of the fact that mankind
will be saved.56 There will be a flood and a great destruction, he predicts in yet
another passage, and here he tells his son, who has been sent by his grandson to
consult him about the infant Noah, that ‘this child that was born to you will be
left on the earth, and his three children will be saved along with him’.57 There
is also a passage in which it is Noah who sees that the destruction is near: he
reacts by setting off to speak with his great-grandfather Enoch about it, and the
latter explains all the secrets to him.58 One way or the other, the immortalised
human and the flood survivor are now closely linked.
53 For the date around 125 (as opposed to the vaguer c. 170–100bc), see D. Mendels, The
Land of Israel as a Political Concept in Hasmonean Literature, Tübingen 1987, pp. 57–
88.
54 1 Enoch (Book of Parables), 54:7–9.
55 1 Enoch (Dream Visions), 83:3–9.
56 1 Enoch (Apocalypse of the Weeks), 93:4.
57 1 Enoch (The Birth of Noah), 106:15 f.
58 1 Enoch (Book of Parables), 65:1–68:1.
56 chapter 3
Secondly, Enoch has become a visionary. Again, there is no hint of this

in Genesis, but it is now by revelation that he knows about the flood, and
many other hidden things as well, including astronomy and other secrets of
the cosmos. In some passages he is envisaged as receiving his supernatural
knowledge during a visit to heaven in the style of Enmeduranki. ‘The vision
of heaven was shown to me, and from the words of watchers [i.e. the ones
who stayed in heaven] and holy ones I have learned everything, and in the
heavenly tablets I read everything’, as he says before predicting the ‘the first
end’.59 At other times he receives visions in his sleep. ‘Dreams came upon me,
and visions fell upon me’, he explains with reference to an occasion on which
a voice, apparently God’s, commanded him to go and reprimand the wayward
watchers.60 He had two visions before he married, he informs us, one of the
destruction of the earth and the other of the entire history of mankind from
Adam to the day of judgement, and we are duly given an account of both.61
Since his ascent to heaven is usually cast as a dream vision, the two modes
of revelation are mostly identical.62 In Jubilees, too, Enoch is a visionary who
foresees the entire future of mankind down to the day of judgement in his
sleep.63 He is not a diviner like Enmeduranki. He does not interpret signs or
read omens, but rather communicates directly with God and/or the angels or
receives revelations in dreams, and this is a feature he shares with Atraḫasīs.
Of the latter we are told that he ‘would speak [with his god], and his god
[would speak] with him’;64 and when Enki was under oath not to subvert Enlil’s
plans by speaking to Atraḫasīs, the mischievous deity would communicate with
Atraḫasīs in dreams.65 When Atraḫasīs asked Enki to explain the meaning of a
dream he had received, Enki circumvented his vow by speaking to a reed wall
with Atraḫasīs listening on the other side: this was how Atraḫasīs was warned
of the flood.66 ‘I knew of the counsel of the great gods, I knew of their oath,
though they would not reveal it to me. He repeated their word to the wall’,
59 1 Enoch (Apocalypse of the Weeks), 93:2–4; also Book of Parables, 54:7–9.

60 1 Enoch (Book of Watchers), 13:8.
61 1 Enoch (Dream Visions), chs. 83–90.
62 1 Enoch (Book of Watchers), 13:8; 14:1, 8 ff.
63 Jubilees, 4:19 (in Charlesworth, Old Testament Pseudepigrapha, ii, p. 62).
64 Atraḫasīs, tablet i, lines 366 f., in Lambert and Millard, Atra-ḫasîs, p. 67; in Foster, Before
the Muses, p. 239.
65 Atraḫasīs, tablet ii, col. 3, lines 7–10, in Lambert and Millard, Atra-ḫasîs, p. 77; in Foster,
Before the Muses, p. 243.
66 Atraḫasīs, tablet iii, col. 1, lines 1 ff., in Lambert and Millard, Atra-ḫasîs, p. 89; in Foster,
Before the Muses, p. 247.
as the Ugaritic version has Atraḫasīs declare.67 The theme is present already
in the Sumerian flood story.68 In the Gilgamesh epic Enki (now known by his
Akkadian name Ea) denies that it was he who told Atraḫasīs about the flood: ‘I
did not myself disclose the secret of the great gods; I let Atraḫasīs see a dream,
and so he heard the gods’ secret’.69 Xisouthros is also warned of the flood in a
dream in Berossos.70
Thirdly, the permanent abode to which God took Enoch is sometimes envis-
aged as a distant place on earth rather than heaven. As we have seen, the Bible
was normally taken to say that Enoch was pleasing to God and that God took
him up to heaven on a permanent basis. The Book of Parables, one of the
youngest parts of 1Enoch, adheres to this solution, except that it has Enoch
go on a temporary trip to heaven first. His first ascent was perhaps made in
a dream, though the explanation that ‘a whirlwind snatched me up from the
face of the earth’ suggests that he ascended physically.71 In the second ascent
he was raised ‘on the chariots of the wind’, which sounds much like the whirl-
wind, and though we are twice told that his spirit was taken away and ascended
to heaven, he had his body with him too, for when he saw God and the angels
in heaven, his flesh melted and his spirit was transformed, i.e. he became an
angel; soon thereafter he is addressed as the Son of Man, apparently his heav-
enly double with whom he has now merged.72 But things are less clear in the
Book of Watchers, an older part of 1Enoch. Here the editorial comment with
which Enoch is introduced says that he had been taken (by God) before the
descent of the wayward watchers and that nobody knew where he was because
he was with the (virtuous) watchers and holy ones, clearly in heaven. The bibli-
cal statement that Enoch ‘walked with hāʾelōhîm’ has been taken to mean that
67 rs 22.421 in Lambert and Millard, Atra-ḫasîs, p. 133; in Foster, Before the Muses, p. 255
(whose translation I have reproduced).
68 Civil, ‘Sumerian Flood Story’, lines 148–157, in Lambert and Millard, Atra-ḫasîs, p. 143; in
Pritchard, Ancient Near Eastern Texts, p. 44; in Jacobsen, ‘Eridu Genesis’, pp. 522f.
69 Gilgamesh epic, tablet xi, lines 196–197.
70 Berossos in Syncellus in Verbrugghe and Wickersham, Berossos and Manetho, p. 49; also
tr. in Lambert and Millard, Atra-ḫasîs, pp. 135 f.
71 1 Enoch (Parables), 39:3; cf. 52:1.
72 1 Enoch, 70:2; 71:1, 5, 11, 14; cf. Orlov, Enoch-Metatron Tradition, pp. 167f. and the literature
cited there. One wonders if these passages were overlooked by P.S. Alexander, ‘From
Son of Adam to Second God: Transformations of the Biblical Enoch’, in M.E. Stone and
Th.A. Bergren (eds.), Biblical Figures outside the Bible, Harrisburg, pa, 1998, pp. 87–122,
at 102f.: according to him, the Slavonic Enoch (2 Enoch) marks a radical departure from
the earlier Enoch literature when it unequivocally claims that Enoch ascended bodily to
heaven.
58 chapter 3
he walked with the angels,73 and the statement that ‘he was no more, because
God took him’ is taken to explain how he had come to walk with them: God had
moved him.74 The fact that nobody knew where he was shows that he is envis-
aged as having ascended physically, not just in a dream, so one would assume
God to have moved him to heaven on a permanent basis. But not long there-
after we see him ascend to heaven again, though we have not heard anything
about his descent, and this time he ascends as an earthling who can only do
so in a dream. Clearly, the editorial comment has been inserted without much
attention to coherence. Jubilees, a later work, has tidied things up. Here too the
biblical statement that Enoch ‘walked with hāʾelōhîm’ is taken to mean that he
walked with angels, again for a long time, but not on a permanent basis: he did
so for six jubilees of years. The statement that ‘he was no more, because God
took him’ still refers to his permanent removal, but it is not to heaven that God
removes him: rather, God places him in the garden of Eden, explicitly identified
as a place on earth. There he still was, writing condemnation and judgement of
the world.75
The idea that Enoch was removed to a remote place on earth is not limited
to Jubilees. A text on the birth of Noah in 1Enoch tells us that Noah’s father,
Lamekh, feared that Noah had been sired by an angel and did not believe his
wife’s protestations that the child was his own. For this reason Lamekh asked
his father Methuselah to go and see Enoch about it, and Methuselah came
to Enoch ‘at the ends of the earth’.76 The Book of Parables, the very part of
1Enoch in which Enoch sees his flesh melt in heaven, likewise tells us in what is
probably an older stratum that when Noah had a vision of the destruction of the
earth and set off to speak with Enoch about it, it was at ‘the ends of the earth’
that he found him.77 It also has Noah mention ‘the garden where the chosen
and righteous dwell, where my great-grandfather was taken up, the seventh
from Adam’.78 In the Genesis Apocryphon from Qumran, Methuselah goes off
to find Enoch in ‘Parwain’,79 an exotic, far-off country from which the gold of
the temple came.80 In the Book of Giants, of which fragments were found at
73 Cf. VanderKam, Man for All Generations, p. 32.

74 Cf. VanderKam, Man for All Generations, p. 43.
75 Jubilees, 4:21 ff. (in Charlesworth, Old Testament Pseudepigrapha, ii, p. 62f.).
76 1 Enoch, 106:8.
77 1 Enoch, 65:1.
78 1 Enoch, 60:8.
79 1q20 Genesis Apocryphon, ii, line 23, in G. Vermes (tr.), The Complete Dead Sea Scrolls in
English, 4th ed., Harmondsworth 1997, p. 450.
80 2 Chr. 3:6 (Parwaim); Kvanvig, Roots, p. 89.
Qumran, the giants send one of their own, Mahaway, to Enoch so that he can
interpret a dream for them, and Mahaway finds Enoch past the wastelands, on
the other side of a great desert,81 apparently meaning in the garden of Eden.82
According to Jubilees, the garden of Eden was one of the four places of the
Lord on earth, and it was because of Enoch that Eden was spared inundation
during the flood. Here Enoch is as close as he can get to being the flood survivor,
keeping dry in Eden rather than in the ark. It is presumably on the basis of
Jubilees, which was available in Syriac and left some marks on Islamic literature
too, that al-Ḥasan al-Baṣrī envisaged Idrīs as being in the garden (al-janna), i.e.
paradise, when God raised his rank, though whether he located the garden in
heaven or on earth one cannot tell.83
In short, Enoch became more like Atraḫasīs. As Kvanvig observes, Noah did
too: he also figures as a visionary who foresees the flood in 1 Enoch, and some-
times it is hard to tell whether it is Enoch or Noah that the book is speaking
about.84 The flood survivor and the immortalised human are flowing together,
exactly as one would expect. That Mesopotamian rather than Greek culture
was the engine behind these developments is nicely illustrated by the fact that
although the man-eating, blood-sucking giants undoubtedly typify the Hel-
lenistic rulers under whose control the Jews had fallen, it is the Mesopotamian
Gilgamesh, not the Greek Hercules, who figures among them in the fragments
of the Aramaic Book of Giants (omitted from 1Enoch). It was presumably the
Jews of Babylonia who first depicted Gilgamesh in this negative light, with ref-
erence to the rulers they had to bear with there.85
81 qg5, 5 f., with identification of the speaker in qg4a, 21–23, in J.C. Reeves, Jewish Lore in
Manichaean Cosmogony, Cincinnati 1992, pp. 63f.; corresponding to 4q530, col. 2, 21–23;
col. 3, 5 f., in L.T. Stuckenbruck, The Book of Giants from Qumran: Texts, Translation, and
Commentary, Tübingen 1997, pp. 126, 130.
82 Cf. the material in Reeves, Jewish Lore, p. 104.
83 For echoes of Jubilees in Qudāma b. Jaʿfar and in the Persian tradition, see A. Silverstein,
‘From Atraḫasīs to Afrīdūn: on the Transmission of an Ancient Near Eastern Motif to
Islamic Iran’, Jerusalem Studies in Arabic and Islam 39, 2012, pp. 1–14, at 5, 8ff.
84 Cf. 1 Enoch 60:23, where the speaker refers to an event in the life of Enoch, implying that the
speaker is somebody other than Enoch (presumably Noah). Nickelsburg and VanderKam,
however, emend Enoch to Noah, thus retaining Enoch as the speaker.
85 For the polemical nature of the appearance of Gilgamesh and other figures from the Gil-
gamesh epic in the Book of Giants, see Reeves, Jewish Lore, p. 126. According to D.R. Jack-
son, ‘Demonising Gilgameš’, in J. Azize and N. Weeks (eds.), Gilgameš and the World of
Assyria, Leuven 2007, pp. 107–114, at 113, the author(s) chose Gilgamesh rather than a Greek
figure in order to hide his significance from their opponents, while M. Goff, ‘Gilgameš the
Giant: the Qumran Book of Giants’ Appropriation of Gilgameš Motifs’, Dead Sea Discover-
60 chapter 3
A passage from the lost work of Berossos suggests that by the third cen-
tury bc the interaction between Jewish and Babylonian models had affected
not only the Jewish understanding of Enoch, but also the Babylonian under-
standing of Atraḫasīs. The passage concerns the grant of immortality to the
flood survivor, which Berossos narrates in wording quite different from that
of the two earlier works known to us, the Sumerian deluge story and the Gil-
gamesh epic. According to the Sumerian account, when the flood was over,
Ziusudra sacrificed and prostrated himself to An and Enlil, who responded
favourably: ‘Life like [that of] a god they give him, breath eternal like [that of]
a god they bring down for him. Then Ziusudra, the king, the preserver of the
name of vegetation [and] of the seed of mankind, in the land of crossing, the
land of Dilmun, the place where the sun rises, they caused to dwell’.86 In the
Gilgamesh epic Ūta-napishti sacrifices while still in the boat; we then hear of
a dispute between the gods about Ea’s role in Ūta-napishti’s survival and the
questionable merits of Enlil’s use of so drastic a remedy as the flood; Enlil then
enters the boat and touches the foreheads of Ūta-napishti and his wife, who
are kneeling before him, and declares that ‘In the past Ūta-napishti was (one of)
mankind, but now he and his wife shall be like us gods! Ūta-napishti shall dwell
far away, at the mouth of the rivers’. Ūta-napishti reports, ‘They took me and
settled me far away, at the mouth of the rivers’.87 Berossos’ account is initially
similar. Xisouthros disembarks together with his wife and daughter, prostrates
himself and sacrifices to the gods. But the continuation says that ‘after this he
disappeared together with those who had left the ship with him. Those who
remained on the ship and had not gone out with Xisouthros … searched for
him and called out for him by name all about. But Xisouthros from then on was
seen no more, and then the sound of a voice that came from the air gave the
instruction that … Xisouthros, because of the great honour he had shown the
gods, had gone to the dwelling place of the gods’.88
There are several noteworthy changes in Berossos’ account. First, the gods
who are present in person in the two earlier Mesopotamian accounts are here
ies 16, 2009, pp. 221–253, sees more ‘creative appropriation’ than polemics here; and I. Fröh-
lig, ‘Enmeduranki and Gilgamesh: Mesopotamian Figures in Aramaic Enoch Traditions’, in
E.F. Mason and others (eds.), A Teacher for All Generations: Essays in Honor of James C. Van-
derKam, Leiden 2012, ii, pp. 637–653, at 652 f., denies that Gilgamesh is envisaged as a giant.
86 Sumerian flood story, final lines, in Pritchard, Ancient Near Eastern Texts, p. 44; slightly
differently (and less powerfully) in Lambert and Millard, Atra-ḫasîs, p. 145; Jacobsen,
‘Eridu Genesis’, p. 525.
87 Gilgamesh epic, tablet xi, lines 157 ff., 199–206.
88 Syncellus in Verbrugghe and Wickersham, Berossos and Manetho, p. 50.
replaced by a disembodied voice: the concept of the divine had drastically

changed.89 Secondly, it is not clear that the immortalised Atraḫasīs is being
moved to a remote place on earth: he goes to the dwelling of the gods, prob-
ably meaning heaven rather than Dilmun.90 Finally, Berossos tells the story of
Xisouthros’ reward from the point of view of those left behind: those on the
boat could not find him, ‘Xisouthros from then on was seen no more’. The word-
ing here is strikingly similar to the biblical ‘he was no more’ and 1 Enoch’s more
expansive ‘none of the sons of men knew where he had been taken or what
had happened to him’.91 This has been noticed before,92 but the assumption
has been that Berossos may show us the source of the biblical formulation,
though he wrote around 290bc. It is possible, of course, that Berossos here pre-
serves an ancient Akkadian formulation that passed into Genesis even though
it happens not to have come down to us. It could even be argued that Berossos
is giving us the missing Mesopotamian source for Enoch’s permanent trans-
lation to heaven, assuming his ‘dwelling place of the gods’ to be ancient too.
But given that Berossos freely departs from the tradition to accommodate a
new concept of the divine, it seems more likely that the other unprecedented
elements are also new, or in other words that they reflect exposure to Enoch.
Berossos would not, of course, have read the Bible or any other Jewish writ-
ings, but Jews would have retold the story of Enoch along lines that fused
Enoch and Atraḫasīs, and these versions could easily have passed to Baby-
lonian priests and scribes, especially in the Persian and Hellenistic periods,
when the Babylonians lost their hegemonic status and the position of the Jews
improved.
89 Differently E.G. Kraeling, ‘Xisouthros, Deucalion and the Flood Traditions’, Journal of the
American Oriental Society 67, 1947, pp. 177–183, at 178, 179, according to whom Berossos is
covering up the polytheism of the original narrative out of consideration for enlightened
Greek taste. Why the polytheist Greeks should be more enlightened than the polytheist
Babylonians, or indeed why polytheism should be unenlightened, is not explained.
90 Luke, ‘The Patriarch Enoch’, pp. 132, 135, takes Dilmun to be the abode of the gods in the
Gilgamesh epic with reference to Gilgamesh’s question to Ūta-napishti: ‘(Tell me) how
you joined the assembly of all the gods in your quest for life’ (Gilgamesh epic, tablet xi,
line 7; cf. the translation by Speiser in Pritchard, Ancient Near Eastern Texts, p. 93). But
George translates, ‘How was it you attended the gods’ assembly, and found life?’, which
suggests a temporary meeting with the gods (presumably when he came out of the boat),
not permanent residence in their midst. There is no suggestion in the Gilgamesh epic that
Ūta-napishti was surrounded by gods.
91 1 Enoch, 12:1.
92 C. Westermann, Genesis 1–11, a Continental Commentary, Minneapolis 1994 (German orig-
inal 1974); Kvanvig, Roots, pp. 226, 228.
62 chapter 3
In short, in the Hellenistic period there were Jews who cast Enoch as a figure
connected with the flood, a visionary who received communications from the
divine, and a recipient of immortality who was removed to a remote place
on earth. All three features assimilated Enoch to Atraḫasīs. Conversely, there
were Babylonians who thought of Atraḫasīs as a figure who had disappeared,
apparently by being taken to heaven, when he was granted immortality, a
feature which assimilated him to Enoch (or alternatively reveals him as one of
the sources of this figure). Either way, Enoch and Atraḫasīs were now difficult
to tell apart. The learned will hardly have gone so far as to identify them, but
it is no wonder if Atraḫasīs came to be regarded as simply another name for
Enoch at a popular level in Babylonia.
The Name
At this point the reader may be ready with two objections. First, how could
the Mesopotamian flood hero have blended with Enoch under the name of
Atraḫasīs rather than Ziusudra or Ūta-napishti when, with the exception of two
passages in the Gilgamesh epic, it is only under the name of Ziusudra (and
variants) or Ūta-napishti that the flood hero is associated with immortality
in the Akkadian literature known to date? The most obvious response is that
although the immortality theme is absent from the Atraḫasīs epic as it has
come down to us, it must in fact have been present there too. We do not have a
complete version. There is a lacuna of 34 lines at the end of tablet iii, containing
the final part of the epic. Here Enlil, after first being enraged by Atraḫasīs’
survival, institutes new measures of population control that will not wipe out
mankind, and this is where one would expect to hear that he also granted
immortality to Atraḫasīs and his wife and moved them to a distant place. ‘The
apotheosis of the flood hero could have been contained in the damaged ending
of Atra-ḫasîs’, as Lambert and Millard remark.93 That this was actually the case
is further suggested by the fact that the flood survivor is granted immortality
both in the earlier Sumerian flood story and in the later Akkadian Gilgamesh
epic: how could these themes have been absent from the Atraḫasīs epic in
between? It is a version of this very epic that is being retold in the Gilgamesh
epic.94 In short, the flood hero was probably granted immortality under all
three names under which he appears in the tradition. This does not, of course,
93 Lambert and Millard, Atra-ḫasîs, p. 137.

94 Lambert and Millard, Atra-ḫasîs, p. 11.
explain why the name Atraḫasīs was preferred over the other two. The reason
could be that it stressed the great wisdom of the hero. But at all events, there
is nothing particularly problematic about the use of this rather than the other
two names.
It has been suggested that Enoch also came to be known as Ūta-napishti in
circles which surface in Manichaeism. Mani’s Book of Giants mentioned a fig-
ure called At(a)nabīsh (ʾtnbysh), a name which Reeves tentatively explained as
derived from Ūta-napishti. In Reeves’ view the book downgraded Ūta-napishti
and other figures from the Gilgamesh epic to the status of iniquitous giants.95
Huggins provisionally accepts the derivation of At(a)nabīsh from Ūta-napishti,
but he denies the downgrading. He sees a parallel between a passage in the
Qumran Book of Giants and a line in Mani’s Book of Giants (both known only
from fragments) which would identify At(a)nabīsh as Enoch.96 If so, Enoch
appears both under his own name and as At(a)nabīsh in the Manichaean book.
Pace Stuckenbruck, this is hardly a problem, given that Ūta-napishti himself
appears both under his own name and that of Atraḫasīs in the Gilgamesh
epic;97 but there simply is not enough information in the fragment to clinch
the reality of the parallel, and both the form of his name and another two frag-
ments suggest that At(a)nabīsh was indeed a giant.98
95 J.C. Reeves, ‘Utnapishtim in the Book of Giants?’, Journal of Biblical Literature 112, 1993,
pp. 110–115; id., Jewish Lore, pp. 126, 159, n. 373 (using the form Atambīsh).
96 R.V. Huggins, ‘Noah and the Giants: a Response to John C. Reeves’, Journal of Biblical
Literature 114, 1995, pp. 103–110. In the Qumran Book of Giants the giant Mahaway is sent to
ask Enoch for the interpretation of a dream. In the Manichaean Book of Giants ‘Māhawai
went to Atambīsh (and) related everything’ (Reeves, ‘Utnapishtim in the Book of Giants?’,
p. 114).
97 Cf. Stuckenbruck, Book of Giants, p. 73n; id., ‘Giant Mythology and Demonology: from the
Ancient Near East to the Dead Sea Scrolls’, in A. Lange, H. Lichtenberger and K.F.D. Röm-
held (eds.), Die Dämonen: die Dämonologie der israelitisch-jüdischen und frühchristlichen
Literatur im Kontext ihrer Umwelt, Tübingen 2003, pp. 318–338, at 334.
98 Similarly Stuckenbruck, Book of Giants, p. 73n; id., ‘Giant Mythology’, pp. 333ff. There
are two figures presumed to come from the Gilgamesh epic in the Qumran Book of
Giants, Gilgames(h) and Ḥobabis(h), both written now with a šin and now with a samek;
there are also two in the fragments of the Manichaean book, Ḥobabīsh (thus written in
Manichaean Middle Persian) and At(a)nabīsh. The name Ḥobabis(h) is generally held to
be derived from Ḫumbaba, the monstrous guardian of the cedar forest, and the -ish ending,
which has generated much learned speculation, was presumably just stuck on to make
the names rhyme (for more learned explanations, see Stuckenbruck, ‘Giant Mythology’,
pp. 327f.). The fact that At(a)nabīsh fits the rhyming pattern strengthens the case for his
identification as Ūta-napishti, and also for his status as a giant rather than as Enoch.
64 chapter 3
The second objection perhaps present in the reader’s mind is philological.

In order for Atraḫasīs to turn into Idrīs the velar fricative ḫ would have to dis-
appear, but how could it? Needless to say, we lack the material with which to
follow a gradual transformation of the name, but the development postulated
does at least have to be plausible, and ḫ (or kh in the common Islamicist translit-
eration) is not a sound that is easily elided. Nonetheless, its disappearance is
not a problem. Aramaic did once distinguish between the pharyngeal ḥ and
the velar fricative ḫ, though it used the same letter to express them in writ-
ing; but by about 200bc the velar fricative ḫ had turned into the pharyngeal
ḥ.99 In the subsequent development the pharyngeal ḥ was weakened in several
Aramaic dialects (as also the ʿayn), and it completely disappeared in Babylo-
nian Aramaic as known from the Talmud and Mandaic: ḥ was reduced to h or
eliminated.100 Transmission through Pahlavi also reduced ḥ to h.101 In Babylo-
nian Aramaic and Pahlavi, then, Atraḫasīs would have become something like
Atra(ha)sīs. Transformation of the t into d and contraction would have done the
rest. A form such as *Addarasīs, easily shortened to Idrīs, could have turned into
*Andarasīs by dissimilation of gemination in Pahlavi (well attested in connec-
tion with other Aramaic words) and thus yield the Andreas of the Alexander
Romance.102 Perhaps even an ungeminated form such as Adrasīs could yield
Idrīs and Andreas alike. If not, we evidently need to explain how *Addarasīs
got to be geminated in the first place, but though I do not have an answer, it is
hardly an insuperable problem.
99 E. Lipinski, Semitic Languages: Outline of a Comparative Grammar, Leuven 1997, p. 146;

K. Beyer, Die aramäischen Texte vom Toten Meer, Göttingen 1984, i, p. 102 (my thanks to
Kevin van Bladel for drawing the second work to my attention).
100 S. Weninger and others (eds.), The Semitic Languages, Berlin and Boston 2011, pp. 612,
624 f., 633f. (Jewish Palestinian, Samaritan and Christian Palestinian Aramaic), 662 (Jew-
ish Babylonian Aramaic), 674 (Mandaic); M. Morgenstern, Studies in Jewish Babylonian
Aramaic, Winona Lake, in, 2011, pp. 73 ff. (Jewish Babylonian Aramaic).
101 In fact the development seems usually to be explained with reference to the influence of
Iranian languages in the east, of Greek in the west. The mystery is how Syriac escaped.
102 For such dissimilation, compare the transformation of Syriac guddā into Middle Persian
gond, shabbta into shamba, and Manichaean Aramaic zaddīq into Middle Persian zindīk
(all adduced by F. de Blois in ei2, s.v. ‘Zindīḳ’).
The Reflections of Gilgamesh’s Search for Immortality
We may now turn to the reflections of pagan Aramaic mythology in the lit-
erature of the neighbours that take us to al-Khiḍr. All are reflections of Gil-
gamesh, an enormously popular figure who lived on under both his own name
and those of others; indeed, thanks to the conservatism of magic his name
appears in an amazingly faithful form even in a work attributed to al-Suyūṭī
(d. 911/1505), who reproduces an incantation of Solomon that includes Gil-
gamesh (Jiljamīsh) among the spiritual beings. As Reeves observes, this reflects
the use of Gilgamesh’s name in incantations, a practice well attested in Akka-
dian times.103 Outside the domain of magic Gilgamesh may appear twice in a
list of ancient kings in Theodore Bar Koni (fl. late 8th century ad), but no infor-
mation is offered about these figures.104 The last author to mention him by his
own name (Gilgamos) with some information about him is the Greek Aelian
(d. c. 235), but most of what he says about him was originally told about oth-
ers.105
103 Reeves, Jewish Lore, pp. 120 f. and 159, n. 370; cf. George, Gilgamesh Epic, i, pp. 112ff., 130ff.;
T. Abusch, ‘Ishtar’s Proposal and Gilgamesh’s Refusal: an Interpretation of “the Gilgamesh
Epic”, Tablet 6, Lines 1–79’, History of Religions 26, 1986, pp. 143–187, at 150f. and the
literature cited there; M. Schwartz, ‘Qumran, Turfan, Arabic Magic, and Noah’s Name’, in
R. Gyselen (ed.), Charmes et sortilèges, magie et magiciens, Bures-sur-Yvette 2002, pp. 231–
238.
104 Theodore Bar Koni, Livre des scolies (recension de Séert), ed. A. Scher, Liber Scholiorum
(csco 55, 69/Syr. 19, 26), Paris 1910, 1912; tr. R. Hespel and R. Draguet (csco 431–432/Syr.
187–188), Louvain 1981–1982, mimrā ii, par. 120 (Gamigos and Ganmagos).
105 Aelian, De natura animalium, xii, 21. Gilgamos, son of the daughter of the king of Babylon,
was hurled from a tower by the king who had been warned that the son of his daughter
would oust him; saved by an eagle, he was brought up by a gardener and eventually
became king of Babylon. Not much of this fits Gilgamesh (cf. George, Gilgamesh Epic, i,
pp. 61, 106 ff.). For the gardener, compare Sargon of Akkad (3rd millennium bc), whose
father or foster-father is said to have been a gardener, cf. B. Lewis, The Sargon Legend,
Cambridge, ma, 1980; S. Dalley and A.T. Reyes, ‘Mesopotamian Contact and Influence
in the Greek World: 2. Persia, Alexander, and Rome’, in S. Dalley (ed.), The Legacy of
Mesopotamia, Oxford 1998, pp. 107–124, at 119. By Aelian’s time the motifs had also been
transferred to the Achaemenids: Achaemenes was supposedly nursed by an eagle, as
Aelian himself mentions, while Cyrus was supposedly brought up by a Median cowherd
for the same reason that Gilgamesh was brought up by a gardener, cf. Herodotus, Histories,
i, 107 ff. Cf. also W.F.M. Henkelman, ‘Beware of Dim Cooks and Cunning Snakes: Gilgameš,
Alexander, and the Loss of Immortality’, in R. Rollinger and others (eds.), Interkulturalität
in der alten Welt, Wiesbaden 2010, pp. 323–358, at 323f. (my thanks to Tommaso Tesei for
drawing this splendid study to my attention); id., ‘The Birth of Gilgameš (Ael. na xii.21):
66 chapter 3
In the material reflecting Gilgamesh’s search for immortality, by contrast, it

is the names of other people that have been affixed to a story originally told
about him. The earliest reflection is in a narrative found in some recensions
of the Greek Alexander Romance.106 The role of Gilgamesh is here played
by Alexander the Great.107 Alexander goes off with his troops to the land of
perpetual darkness with the intention of reaching the Land of the Blessed.
Unlike Gilgamesh he does not find an immortal human resident on the other
side: there is no Atraḫasīs/Ūta-napishti in this version. (A wise old man does
appear, but he is an ordinary human being and a member of Alexander’s own
camp.) On the way, however, Alexander and his troops come to a spring with
twinkling water, and when Alexander’s cook Andreas (whose name is not given
in β) goes to wash a dried fish in it, the fish comes alive and swims away.108
The cook does not tell anyone about this, but drinks of the water himself
and fills a silver vessel with it, which he later gives to Alexander’s daughter in
order to seduce her. Alexander, who reaches the Land of the Blessed without
being able to enter it, only hears the story of the fish on his way back and never
gets to drink of the life-giving water. He reacts by angrily turning Andreas into a
daimōn of the sea, while his own daughter becomes a daimōn of the desert.109
a Case Study in Literary Receptivity’, in R. Rollinger and B. Truschnegg (eds.), Altertum

und Mittelmeersraum: die antike Welt diesseis und jenseits der Levante. Festschrift für Peter
W. Haider, Stuttgart 2006, pp. 807–856.
106 For the recensions in question (β, the expanded version of β contained in a Leiden
manuscript l, and the versions of β known as λ), see Henkelman, ‘Dim Cooks’, pp. 325–
328.
107 This is denied by R. Stoneman, ‘Oriental Motifs in the Alexander Romance’, Antichthon
25, 1992, pp. 95–113, at 99, to whom the similarity between the Gilgamesh epic and the
Alexander Romance is slight and superficial. His views are ably countered by Henkelman,
‘Dim Cooks’, pp. 342 f.
108 The cook is nameless in the regular manuscripts of β. The name Andreas first appears in l
(the Leiden ms containing an expanded version of β) at ii, 41 in a passage explaining the
name of the Adriatic as derived from his, and also once in the margin (ii, 39); two later
versions have the name in the text in both passages, cf. Henkelman, ‘Dim Cooks’, n. 68.
The folk etymology of the name of the Adriatic is presumably based on knowledge that
the cook who became a sea demon was called Andreas in some circles. The narrator can
hardly have invented Andreas’ name to explain that of the Adriatic, which is incidental to
the story.
109 Pseudo-Callisthenes, book ii, 39 ff.; tr. Stoneman, Greek Alexander Romance, pp. 119–122,
cf. 8 ff., 28 f., for the dates; cf. T. Tesei, ‘Survival and Christianization of the Gilgamesh
Quest for Immortality in the Tale of Alexander and the Fountain of Life’, Rivista degli Studi
Orientali, ns, 83, 2011, pp. 417–440.
Andreas the cook is playing the role of the snake in the Gilgamesh epic: it is
he who robs Alexander of his immortality (the dried fish is just a passive ben-
eficiary of the cook’s action). The substitution of a human being for the snake,
as also the transformation of this human into a maritime daimōn, reflects the
presence in the narrator’s mind of a Greek mythological figure, Glaukos, who
achieved immortality as a sea god or sea monster after eating grass brought
up from the sea.110 The intrusion of this figure meant that there came to be a
second immortalised human in the story originally told of Gilgamesh and Ūta-
napishti (Alexander’s daughter, in principle the third, is treated in too perfunc-
tory a manner to count), but the story is only designed to have one, and this may
be why Atraḫasīs/Ūta-napishti has disappeared from the version in the Alexan-
der Romance. The immortal human who remained in the story, however, seems
to have inherited Atraḫasīs’ name, in an Aramaic version that sounded some-
what like Andreas to a Greek ear.
The story of the cook who washes the fish in the spring of life is not found
in the Syriac version of the Alexander Romance, but it appears in the Syriac
Alexander Poem (or Song, or metrical Homily) which is attributed to Jacob
of Sarug (d. 521) but was actually composed between 628 and 636.111 In this
work Alexander does meet a wise old man after traversing the land of darkness.
Alexander tells him he has come to find the spring of life, and the wise old man
advises him to let his cook test the diverse springs in the area by washing a
salted fish in them; if the fish comes alive, he has found it. When the cook comes
to the spring of life, the fish swims away and the cook jumps into the water to
catch it, without success. He then tells Alexander about it, but Alexander does
not succeed in bathing in the spring, apparently because he cannot find it in
the darkness. The wise old man consoles him, and thereafter the story shifts to
questions asked by Alexander and the wise man’s answers. The momentous
fact that the cook has become immortal by jumping into the water is left
unmentioned. The focus is on the old man and the wisdom he imparts to
110 Pauly’s Real-Encyclopädie der classischen Altertumswissenschaft, ed. G. Wissova, Stuttgart

1894–1963, vii/1, s.v. ‘Glaukos’ (no. 8).
111 Thus G.J. Reinink’s introduction to his translation of the work, Das syrische Alexanderlied:
die drei Rezensionen, Louvain 1983, p. 12; also tr. E.A.W. Budge, The History of Alexander the
Great, Cambridge 1889, repr. 2003, pp. 163–200, at 172ff. (lines 170ff.). Cf. also Henkelman,
‘Dim Cooks’, pp. 328 f.; and the further references in Tesei, ‘Survival and Christianization’,
p. 419n. The otherwise interesting study by S. Dalley, ‘Gilgamesh in the Arabian Nights’,
Journal of the Royal Asiatic Society, 3rd series, 1, 1991, pp. 1–17, at 9, summarises this story
in a mongrel form in which al-Khiḍr is Alexander’s servant, supposedly on the basis of the
version in Budge, History, p. 168.
68 chapter 3
Alexander with his answers. That the old man himself is immortal is also left
unmentioned, and neither he nor the cook is given a name.
The fish episode also went into the Babylonian Talmud, where Alexander
once more travels through the land of darkness, but here both the cook and
the wise old man have disappeared. It is Alexander himself who washes the
fish that comes alive, and we are told that according to some he responded
by washing his face in the water: the significance of this is left unspecified.
Others said that Alexander responded by tracing the water to its source at the
entrance to the garden of Eden, where he clamoured to be let in on the grounds
that he was a king, unsuccessfully of course.112 Here the garden of Eden to
which Enoch was moved reappears as the Jewish version of the Land of the
Blessed.113
There is also a reflection of Gilgamesh’s search for immortality in an obscure
account of the origins of Zoroastrianism in the Syriac Cave of Treasures. Here
we are told that Nimrod was the first to worship fire and that he went to Yoq-
dora in Nod, where he found Yonṭon (or Maniton), son of Noah, by the sea
of Aṭras (or Ukaras or the like). Nimrod bathed in that sea and then went
and prostrated before Yonṭon, saying he had come for his sake. Yonṭon taught
Nimrod wisdom and the writing of the revelations (or just the revelations)
and told him not to come any more; and when Nimrod came up from the
east, he astounded people with his wisdom.114 The identification of Nimrod
as the first to worship fire and/or as Zoroaster is a late antique common-
place, but the rest is distinctly unusual. Nimrod is playing the role of Gil-
gamesh while Yonṭon plays Atraḫasīs/Ūta-napishti. The latter lives in the land
of Nod, located to the east of Eden according to Genesis 4:16, and it is duly
112 Babylonian Talmud, Tamid 32b. The spring of life also originates in paradise in ʿUmāra’s
Alexander story in I. Friedlaender, I. Friedlaender, Die Chadirlegende und der Alexander-
roman, Berlin 1913, pp. 135, 309.20.
113 For other features shared by 1 Enoch and the Gilgamesh epic, see Tesei, ‘Survival and
Christianization’, p. 425 and the literature cited there.
114 The passage is translated S.M. Ri, Commentaire de la Caverne des Trésors, Louvain 2000,
pp. 341 f., on the basis of ch. 27.6–12 of his edition and translation of the text (La Caverne
des Trésors, Louvain 1987). For further comments on the passage, including variants, see
his Commentaire, esp. pp. 79–81, 319 ff., 327 ff.; 355. The variant versions of the names in the
Syriac manuscripts are listed at p. 341n. The passage is also cited with a partial translation
in M. Lidzbarski, ‘Wer ist Chadhir?’, Zeitschrift für Assyriologie 7, 1892, p. 115; and with a
full translation by R.H. Gottheil, ‘References to Zoroaster in Syriac and Arabic Literature’,
Classical Studies in Honour of Henry Drisler, New York and London 1894, pp. 25f., with
reference to C. Bezold (ed. and tr.), Die Schatzhöhle, Leipzig 1833, 1888, p. 230 = 136f.
from the east that Nimrod returns.115 As the cook jumps into the water in
the Alexander poem and Alexander washes his face in the life-giving water
in the Talmud, so Nimrod bathes in the Sea of Aṭras, but in all three cases
the significance of the act is left unidentified; and although Nimrod worships
Yonṭon, we are given to understand that this was for his wisdom, not for his
immortality (or quasi-divinity), which is not mentioned. There is no sign of the
ancient names either, unless we take Aṭras to be another version of the name
Atraḫasīs.
Finally, the fish episode appears in the Qurʾān (18:60–64). Here the role of
Gilgamesh is played by Moses, who vows not to give up until he reaches the
confluence of the two seas. When he and his servant ( fatā, lit. young man)
get there, they ‘forget’ the fish, which swims away. Later Moses is hungry and
asks his servant for food; the servant, who is clearly his cook, replies that he
(not they) forgot the fish, thanks to Satan, and that the fish has swum away.
Moses realises that this water is what they are seeking and they retrace their
steps, with what degree of success we are not told. Instead, the text shifts to
an account of an enigmatic superior being, identified only as a servant (ʿabd,
lit. slave) of God, who imparts wisdom to the hero. The nature of the wisdom
relates to theodicy: the anonymous servant of God justifies God’s seemingly
unjust ways by engaging in seemingly evil acts. This is quite different from the
wisdom imparted by the old man to the hero in the Alexander Poem,116 to which
115 For a different explanation of Nod, see Ri, Commentaire, p. 322; but cf. also 350f. Ri does
not seem to be aware of the longer roots of this passage in the Gilgamesh epic, and this the
main reason why his understanding of Nod and other aspects of the passage differs from
mine.
116 It is a version of the folktale motif ‘God’s justice vindicated’ (type 759 in the Aarne-
Thompson motif index), and many hold the Qurʾānic story to be based on a midrash
concerning Rabbi Joshua b. Levi and Elijah. This theory was apparently first proposed by
Zunz, but it was endorsed by Geiger and so came to be accepted by luminaries such as
Friedlaender and Wensinck among many others. As Jellinek and others pointed out long
ago, however, and as Wheeler has stressed again more recently, the rabbinic story is not
attested until the eleventh century; it was originally written in Arabic, and it is more likely
to be dependent on the Qurʾān than the other way round (cf. H. Schwarzbaum, ‘Some
Theodicy Legends’, in his Jewish Folklore between East and West, ed. E. Yassif, Beersheva
1989, pp. 75–125; B.M. Wheeler, ‘The Jewish Origins of Qurʾān 18:65–82? Re-examining
Arent Jan Wensinck’s Theory’, Journal of the American Oriental Society 118, 1998, pp. 153–
171). The pre-Islamic version closest to the Qurʾān that has been found to date is in John
Moschus’Leimon, where Moses’ role is taken by a monk and the superior being is an angel
(R. Paret, ‘Un parallèle byzantin à Coran xviii, 58–81’, Revue des Études Byzantines 26, 1968,
pp. 137–159).
70 chapter 3
the Qurʾānic passage is otherwise closely related;117 but here as there, neither
the servant of Moses nor the servant of God is given a name.
All in all, then, the only name of interest yielded by all these accounts
is Andreas. That apart, the most striking feature of the stories is the virtual
disappearance of the immortality theme, presumably due to Christianisation.
The only human to become immortal is Andreas (if we discount Alexander’s
daughter); nothing is said about the acquisition of immortality by the nameless
cooks in later versions of the story. The wise old man, where he appears, is not
said to be immortal either. It is still to find the spring of life that Alexander
seeks him out in the Syriac Alexander Poem, but it is from his wisdom that he
benefits, and wisdom is also what the enigmatic sage imparts to Nimrod and
Moses. The association of the waters of life with wisdom is found already in
the Bible, both Jewish and Christian, and thereafter in the Odes of Solomon and
Gnostic literature.118 It is also the association we find in the Qurʾān.
Al-Khiḍr
The exegetes read the immortality back into the Qurʾānic story by identifying
the enigmatic servant of God in sura 18:65–82 as al-Khiḍr, an immortal fig-
ure first encountered in the commentaries on this passage. He is introduced
as a character familiar to the reader, without any sign of disagreement over
the identification until we reach the rationalising theologians (mutakallims).
117 It is identified as the direct source in Th. Nöldeke, Beiträge zur Geschichte des Alexanderro-
mans, Vienna 1890, p. 32, and again in Friedlaender, Chadirlegende, p. 61. But this is unlikely
if it dates from the 630s, as proposed by Reinink, Syrische Alexanderlied, p. 12; id., ‘Alexan-
der the Great in 7th-Century “Apocalyptic” Texts’, Byzantinorossika 2, 2003, pp. 150–178, at
165. The shared features are unduly minimised by B.M. Wheeler, Moses in the Qurʾān and
Islamic Exegesis, London 2002, pp. 11–19.
118 Tesei, ‘Survival and Christianization’, pp. 428 f. As he notes, the living waters were also
associated with baptism and resurrection, and the substitution of a fish (a symbol of
Christ) for the snake certainly resonates with Christian concepts. But though the editor
of recension β was a Christian who did his best to eliminate the most pagan features of
the Alexander Romance (Tesei, op. cit., p. 432), it is difficult to see the fish as a symbol
of Christ here, or even in the Alexander Poem attributed to Jacob of Sarug. There is no
special interest in or sympathy for the fish in either version, the emphasis is on its revival
at Alexander’s expense, and it does not stand for us even in the version attributed to
Jacob of Sarug (contrast the Infancy Gospel of Thomas and the Apocryphal Acts of Peter,
where Christ and Peter revive a salted fish). But the resonance with Christianity may have
mattered to Christian readers even if it did not fit the story line.
Al-Jubbāʾī (d. 303/915f.), for example, objected that al-Khiḍr was sent as a
prophet after Moses and so could not be the servant of God that Moses encoun-
tered (he is probably identifying al-Khiḍr with Elijah); and Fakhr al-Dīn al-Rāzī
(d. 606/1209) adds that if the servant of God was al-Khiḍr, then al-Khiḍr must
have been a more important person in the Torah than Moses, who plays the role
of pupil here, and this he deems to be impossible.119 But we can leave these
developments aside. The earliest material is narrative rather than analytical
and takes the form of a story narrated in different versions with different isnāds
that all go back to Ibn ʿAbbās. According to this story, God rebuked Moses for
declaring himself to have greater knowledge than anyone else on earth and told
him that He had a servant who knew more than he did. When Moses asked
how he could find this servant, God replied that he would have reached his
destination when a salted fish came alive in the water. Moses and his servant
(identified as Joshua) duly set off, the fish came alive, but the servant forgot to
tell Moses; he remembered when Moses became hungry and asked for food, so
they retraced their steps and found al-Khiḍr, the man of superior knowledge
that Moses had set out to locate.120 Like the earlier narrators, these exegetes
saw the hero as searching for wisdom rather than immortality; but unlike them,
they knew the dispenser of wisdom to be immortal.
Who then was this al-Khiḍr? In the long run there were to be many answers
to this question, for al-Khiḍr was a popular figure, and a massive amount of
material accumulated around him.121 The bulk of it is irrelevant to us, however,
because it is not tied to the story of Moses and the waters of life in sura 18.
In the non-exegetical tradition the predominant image of al-Khiḍr is that of a
wanderer who turns up in unexpected places to offer his help.122 This was an
idea was of great appeal to both the popular and the Sufi imagination, and it is
still current today,123 but there is no mention of it in the early interpretations
of the Qurʾānic passage. In fact, though the early exegetes took familiarity with
119 Fakhr al-Dīn al-Rāzī, al-Tafsīr al-kabīr, Tehran 1413, xxi, p. 149, ad 18:65.
120 See the exegetes ad loc., e.g. Ṭabarī, Jāmiʿ, xv, pp. 277–279, 281, 282; also id., Taʾrīkh, ser. i,
pp. 417ff.; al-Kisāʾī, Qiṣaṣ al-anbiyāʾ, ed. I. Eisenberg, Leiden 1922, pp. 230ff.; tr. W.M. Thack-
ston, Tales of the Prophets, Chicago 1997, pp. 247 f.; Friedlaender, Chadirlegende, pp. 75ff.
For the ḥadīth collections, see ei2, s.v. ‘al-Khaḍir (al-Khiḍr)’ (Wensinck), bibliography.
121 There is a helpful survey of all this material in ei2, s.v. ‘al-Khaḍir (al-Khiḍr)’ (Wensinck).
122 For this feature see K. Vollers, ‘Chidher’, Archiv für Religionswissenschaft 12, 1909, pp. 235ff.,
with the proverb asyaru min al-Khiḍr, ‘more of a traveller than al-Khiḍr’, recorded by al-
Maydānī (d. 518/1124).
123 Cf. P. Franke, Begegnung mit Khidr: Quellenstudien zum Imaginären im traditionellen Islam,
Beirut and Stuttgart 2000.
72 chapter 3
al-Khiḍr for granted, they were not sure who he was. Some said that he was an
angel sent by God to Moses and others that he was a human being who had
lived a long time ago, such as a figure connected with Alexander, or someone
mentioned in the Bible, or he was a Babylonian or a Persian rather than an
Israelite. The idea of al-Khiḍr as an angel fits John Moschus’ version of the
theodicy motif. It is admittedly also an angel who justifies God’s ways (to a
monk rather than to Moses) in John Moschus’ version of the theodicy motif,
but this solution is rare in the Islamic tradition: all we are told is that God sent
an angel to teach Moses,124 or that al-Khiḍr was transformed into an angel, not
in heaven after the fashion of Enoch, but here on earth.125
As regards the explanations of al-Khiḍr as a historical figure, the exegetical
attempt to connect al-Khiḍr with Alexander reflects recognition of the fact that
the Qurʾān was retelling a story familiar from the Alexander Romance; but it
was hampered by the fact that there was no immortal sage in this version of
the story. Ibn Isḥāq tells us (on the authority of Ibn ʿAbbās, needless to say)
that Moses’ servant drank of the water of life and so became immortal, and
that since he had no right to drink this water, the learned man (i.e. the servant
of God or al-Khiḍr) punished him by sending him out to sea, where he would
remain until the day of judgement.126 This is a remarkably faithful version of
the cook Andreas who turned into a sea daimōn, and it is explicitly told in
response to a question about Moses’ cook rather than the servant of God. But it
obviously could not explain how the servant of God had become immortal.127
According to other scholars, al-Khiḍr was a commander in charge of Dhū ʾl-
Qarnayn’s vanguard who reached the river of life and drank of it, with the
result that he became immortal and remained alive to this day. He drank of
it inadvertently, or without having set out to do so, or because he and Dhū ʾl-
Qarnayn had been searching for it, and he found it when a salted fish came
124 For attestations, see Māwardī, Tafsīr, ii, p. 495; Friedlaender, Chadirlegende, p. 274. The
idea that al-Khiḍr was an angel did not find many takers, but it was taken up by Mawdūdī
(Franke, Begegnung mit Khidr, pp. 366 ff.).
125 Thus ʿUmāra (fl. 2nd/8th century) in Friedlaender, Chadirlegende, pp. 135f., 145, 146f.;
Arabic text pp. 309, 313 f., 314 f.
126 Ṭabarī, Taʾrīkh, ser. i, p. 428; Friedlaender, Chadirlegende, pp. 105f.
127 Friedlaender nonetheless thinks that al-Khiḍr’s origins are to be sought in the wayward
cook (Chadhirlegende, p. 108). But the two are properly distinguished even in the much
later story of Bulūqiyā in the Arabian Nights. Here the cook/servant is not just a demon
but king of the entire demon world, and we are told that he would never grow old or die
because he had drunk from the fount of immortality guarded by the sage al-Khiḍr (Dalley,
‘Gilgamesh in the Arabian Nights’, p. 5, on the basis of Mardrus’ version).
alive, but in any case his behaviour was morally impeccable.128 Here the servant
has been upgraded to the status of upright sage, suggesting that the exegetes
did not know of Alexander stories in which the sage was still present: the cook
was the only figure they had to work with. It was not easy, and there was also
a problem of chronology in that Moses lived long before Dhū ʾl-Qarnayn in
the sense of Alexander the Great. Some responded by asserting that the Moses
who was associated with Dhū ʾl-Qarnayn was not the Moses who had led the
Israelites out of Egypt, an idea against which Ibn ʿAbbās is said to have protested
vigorously.129 Accordingly, al-Ṭabarī places the Dhū ʾl-Qarnayn connected with
al-Khiḍr in the time of Abraham and calls him ‘Dhū ʾl-Qarnayn the Elder’,
perhaps meaning Nimrod or perhaps just creating a doublet of Alexander the
Great sufficiently old for things to fit.130 This was the best one could do with
the Alexander material.
No wonder, then, that others tried to find al-Khiḍr in the biblical tradition.
He really ought to be mentioned there, given his exalted status as somebody
more knowledgeable than Moses, but who was he? Muqātil and ʿUmāra iden-
tified him as Elisha (al-Yasaʿ).131 For a figure connected with Moses this was an
odd choice, perhaps suggested to them by a comparison of the two verses of the
Qurʾān that mention Dhū ʾl-Kifl. One says of Ismāʿīl, Elisha and Dhū ʾl-Kifl that
all of them were among the good (38:48; cf. 6:86), and another says of Ismāʿīl,
Idrīs and Dhū ʾl-Kifl that all of them were among the patient and the righteous
(21:85f.). This could obviously be taken to imply that Elisha was identical with
Idrīs, and the latter in his turn was easily identified with al-Khiḍr. According
to Ibn Isḥāq citing Wahb b. Munabbih, however, al-Khiḍr was a prophet sent
to the Israelites in the days of Josiah, namely the Aaronid called Jeremiah, son
of Hilkiah.132 Jeremiah is also the biblical equivalent of al-Khiḍr in a passage
128 Ṭabarī, Taʾrīkh, ser. i, p. 414; Ibn Bābawayh, ʿUmāra, al-Thaʿlabī and Ibn Hishām citing Wahb
b. Munabbih, and in Friedlaender, Chadhirlegende, pp. 125ff., 143ff., 169f., 199f.
129 E.g. Ṭabarī, Taʾrīkh, ser. i, p. 424, cf. 417, 419f.
130 Ṭabarī, Taʾrīkh, ser. i, pp. 414, 416. Lidzbarski, ‘Wer ist Chadhir?’, p. 115n.
131 Muqātil, Tafsīr, ii, p. 594; ʿUmāra in Friedlaender, Chadirlegende, p. 137, with the Arabic
text at p. 310. The identification is maintained in what follows, and Elisha/al-Khiḍr is Dhū
ʾl-Qarnayn’s cousin and wazīr.
132 Ṭabarī, Taʾrīkh, ser. i, pp. 415 f., 657f., 661; Friedlaender, Chadirlegende, pp. 269f. The associ-
ation of Josiah and al-Khiḍr is preserved even in al-Thaʿālibī’s version of the Bulūqiyā story,
though Jeremiah himself has fallen by the wayside here (S. Dalley, ‘The Tale of Bulūqiyā
and the Alexander Romance in Jewish and Sufi Mystical Circles’, in J.C. Reeves (ed.), Trac-
ing the Threads: Studies in Jewish Pseudepigrapha, Atlanta 1994, pp. 239–269, at 248; more
briefly also ead., ‘Gilgamesh in the Arabian Nights’, pp. 6f.).
74 chapter 3
in al-Jāḥiẓ’ Tarbīʿ in which a number of identifications are paraded as open to

doubt.133 This too is an odd choice for a figure associated with Moses. Maybe it
is rooted in Matthew 16:13f., where Jesus asks his disciples, ‘Who do people say
the Son of Man is?’, and the disciples reply, ‘Some say John the Baptist, others say
Elijah, and still others Jeremiah or one of the prophets’. All these figures are seen
as alive in some sense and capable of coming back as the ‘Son of Man’. John the
Baptist and Elijah were often identified (Mark 9:12 f.; similarly Matthew 11:13 f.),
and it is still John the Baptist who is Elijah in al-Jāḥiẓ’ passage; only Jeremiah,
then, was free for candidacy as al-Khiḍr. The underlying assumption would be
that al-Khiḍr was the Son of Man, a heavenly being identified with Enoch in the
Parables of Enoch134 and with Jesus in the Gospels.
A fair number of other biblical figures were to be suggested,135 and the
winner in the long run proved to be Elijah. Like Idrīs, he possessed the requisite
immortality, but he too was an odd choice for a figure in the time of Moses, for
he was not credited with pre-existence. The reasons he won out do not seem to
have anything to do with the exegesis of sura 18, however.136
Al-Ṭabarī did not like the identification of al-Khiḍr as Elijah, which he does
not even mention, though he cites a tradition that rules it out,137 and he explic-
itly argues against the theory that al-Khiḍr was Jeremiah. He held al-Khiḍr to
be a much earlier figure, as indeed he would have to be if he was the instructor
of Moses. The only biblical figure before Elijah to achieve immortality was the
antediluvian Enoch/Idrīs, but al-Ṭabarī does not propose him, perhaps because
he envisages Enoch as Idrīs ensconced in heaven or perhaps because he did
not think that Enoch could have been known under two names in Arabic.
Instead, he cites a nameless scholar or scholars who claimed that al-Khiḍr was
the offspring of a Babylonian who lived in the reign of the mythical Persian
king Farīdūn, corresponding to the time of Abraham, and who emigrated to
133 Al-Jāḥiẓ, al-Tarbīʿ waʾl-tadwīr, ed. C. Pellat, Damascus 1955, §40, asks whether Jeremiah
(Armiyā) is al-Khiḍr, reserving Elijah for John the Baptist.
134 1 Enoch, 71:14ff.
135 Friedlaender, Chadirlegende, pp. 258 ff., 268 ff., 272ff., on Melchizedek (Malkān), Job and
others.
136 The same is true when we are told that some held al-Khiḍr’s mother to be a daughter
of Pharaoh, or that he was a pure Arab, or that he descended from Cain (Ibn Ḥajar, al-
Iṣāba fī tamyīz al-ṣaḥāba, Cairo 1328, i, p. 429; all ten suggestions are reproduced in Vollers,
‘Chidher’, p. 258).
137 Ṭabarī, Taʾrīkh, ser. i, p. 415: al-Khiḍr was a Persian, Elijah an Israelite, and they used to
meet every year. For other traditions to the same effect, see U. Rubin, Between Bible and
Qurʾān: the Children of Israel and the Islamic Self-Image, Princeton, nj, 1999, p. 42.
Palestine together with Abraham.138 This made al-Khiḍr a monotheist without

requiring him to be mentioned by name in the Bible. It also made him come
from Babel, whether as an Aramaean, an Israelite or a Persian settled there. Al-
Ṭabarī held him to be a Persian, apparently in the third sense (min wuld al-furs,
as he says shortly thereafter).139 There were also some who held that his father
was a Persian and his mother a Byzantine, or the other way round, but this looks
like mere embroidery.140
What is so interesting about al-Ṭabarī’s suggestions is that he seems to
have believed al-Khiḍr to belong in Mesopotamia, but lacked a framework of
Mesopotamian history in terms of which to position him. He and his likes
wrote at a time when ancient history was either Persian or biblical. Practically
nothing was known of the civilisations behind the tablets strewn all over Iraq.
Pagan Aramaic culture was almost extinct, and whereas ancient Mesopotamia
is nowadays seen as the background to the Bible, in al-Ṭabarī’s time it conjured
up Hellenised magic, astrology, alchemy and other esoteric wisdom of the
type that could be credited to Hermes and envisaged as written on tablets in
antediluvian times. Respectable religious scholars could not attach al-Khiḍr
to that tradition. Yet their sense that he belonged in Babylonia was right.
As Lenormant, Guyard and Lidzbarski recognised long ago, al-Khiḍr in his
role as the instructor of Moses is a late descendant of Atraḫasīs/Ūta-napishti,
the exceedingly wise and immortal flood survivor who is the instructor of
Gilgamesh, Alexander, Nimrod and Moses in Akkadian and Syriac literature
and the Qurʾān.141 One takes it that stories about a famous hero’s search for
immortality had been told by story-tellers in Aramaic and Arabic, and that the
immortal human in these versions bore a name that was Arabised as al-Khiḍr.
138 Ṭabarī, Taʾrīkh, ser. i, pp. 414 f.; mentioned by Ibn Ḥajar, Iṣāba, i, p. 429. Some identified
him as Abraham’s nephew or simply as Lot (cf. Friedlaender, Chadirlegende, p. 273).
139 Ṭabarī, Taʾrīkh, ser. i, p. 415.
140 Ibn Ḥajar, Iṣāba, i, pp. 429 f.
141 Lidzbarski, ‘Wer ist Chadhir?’ (cf. also id., ‘Zu den arabischen Alexandergeschichten’,
Zeitschrift für Assyriologie 8, 1893, pp. 263–312); S. Guyard, ‘Bulletin critique de la religion
assyro-babylonienne’, Revue de l’ Histoire des Religions 1, 1880, pp. 327–345, at 344f., with
the observation that Lenormant had noted the parallel before him (he does not say
where); cf. also Henkelman, ‘Dim Cooks’, pp. 334 ff.
76 chapter 3
Ancient Mesopotamia
The reason that European scholars of the late nineteenth and early twenti-
eth centuries could do better than the early exegetes is that they could read
Akkadian. In 1857 the language was declared to have been deciphered; in 1872
George Smith announced the existence of a Babylonian flood story; and by 1880
Guyard had connected the Babylonian flood survivor with al-Khiḍr, if only as a
hunch.142 The documentation soon followed. The link between the Qurʾān and
the Alexander material was established by Nöldeke in 1890, that between the
Alexander Romance and the Gilgamesh epic by Meissner in 1892 and 1894, and
it was also in 1892 that Lidzbarski documented the link between the Babylonian
flood survivor and al-Khiḍr.143 Al-Khiḍr was the object of intense Orientalist
discussion, with some scholars tracing his roots to Glaukos and others accept-
ing his descent from Atraḫasīs. Friedlaender made as good a case for al-Khiḍr’s
Greek origins as could be made,144 but though his book is a most impressive
piece of scholarship that can still be consulted with profit, it is the ‘Babyloni-
anist’ thesis that carries conviction today.145
What, then, is the name al-Khiḍr? It is often explained as meaning ‘Mr Ever-
green’, the eternally young man,146 but as Lidzbarski noted, this is unlikely
to be right, for no early source associates his name with either eternal youth
or immortality. One early explanation is that he was called green because he
sat on white fur that gave off a green sheen; another is that he was so called
because of his shining beauty, or because he wore green clothes, or because
everything turned green around him, or under him.147 Only the fourth explana-
142 Cf Guyard, ‘Bulletin critique’, pp. 344 f.

143 Nöldeke, Beiträge; B. Meissner, ‘De servitute babylonico-assyriaca’, dissertation, University
of Berlin, defended in 1892; id., Alexander und Gilgamos, Leipzig 1894 (neither seen);
Lidzbarski, ‘Wer ist Chadhir?’. Lidzbarski refers to Meissner’s thesis at p. 109n.
144 K. Dyroff, ‘Wer ist Chadir?’, Zeitschrift für Assyriologie 7, 1892, pp. 319–327; Friedlaender,
Chadirlegende, pp. 113ff., 241 ff., with arguments against his roots in the Gilgamesh epic at
pp. 37 f.
145 Lidzbarski’s thesis was accepted already by K. Vollers, ‘Chidher’, Archiv für Religionswis-
senschaft 12, 1909, pp. 234–284, at 274, though he cites Lidzbarski only to disagree with
him (pp. 281 f.). It is Lidzbarski’s thesis that is immortalised by Wensinck’s entry ‘al-Khaḍir
(al-Khiḍr)’ in ei2. For fair criticism of Friedlaender’s thesis, see Henkelman, ‘Dim Cooks’,
pp. 336 ff.
146 E.g. Vollers, ‘Chidher’, p. 235; ei2, s.v. ‘al-Khaḍir (al-Khiḍr)’; eq, s.v. ‘Khaḍir/Khiḍr’.
147 Cf. Friedlaender, Chadirlegende, pp. 110 ff. The Abū ʾl-Fatḥ he cites at p. 112n as explain-
ing the name with reference to al-Khiḍr’s immortality flourished in the tenth/sixteenth
century.
tion fits ‘Mr Evergreen’, and then only just, for it is not he who is evergreen, but
rather the vegetation that becomes green (again) thanks to him.148 Clermont-
Ganneau, writing in 1877, held the name al-Khaḍir to be a simple translation of
Glaukos, the Greek mythological figure who became an immortal sea daimōn;
Dyroff independently reached the same conclusion in 1892, and Friedlaen-
der agreed.149 But even granting that the colour designations may correspond
(which Lidzbarski disputed) and that al-Khiḍr has a maritime side to him, this
is extremely unlikely, for Glaukos is not actually mentioned in any version of the
heroic quest for immortality: he was merely present in the narrator’s mind as
the latter reshaped his material.150 And more importantly, Glaukos fused with
Andreas, the wayward cook, not with the immortal sage who lived on as al-
Khiḍr, the instructor of Moses.
It may well be by accident that the name of the immortal sage acquired
a form that happened to mean green. Lidzbarski derived al-Khiḍr from ‘Cha-
sisadra’, an inversion of Atraḫasīs’ name assumed at the time to lie behind
Berossos’ Xisouthros: Arabs doing their best to reduce foreign words to three
radicals could only end up with al-Khaḍir, he claimed, carried away by youth-
ful exuberance (he was twenty-four at the time).151 In fact, as we now know,
Berossos’ Xisouthros reflects the Sumerian Zisudra and the form Khasīsadra is
a chimaera (retained in the second edition of the Encyclopaedia of Islam!).152
Given that al-Khiḍr may have been a multifaceted figure already in the sec-
ond/eighth century, we cannot be sure that his name originated in the context
of stories descended from Gilgamesh’s search for immortality. If it did, it would
have to be derived from Zisudra. This has in effect been proposed,153 but it
requires the sibilant z to turn into the velar fricative ḫ, which sounds impossi-
ble. It is noteworthy, though, that Berossos transliterated Zisudra as Xisouthros,
with a xī rather than a zēta. Did he hear the initial letter as a palatalised velar
fricative? I have not seen a discussion of Berossos’ transliteration and would
prefer to leave the question for the experts in Sumerian and Semitic languages
to decide.
148 For this aspect of him, see Franke, Begegnung mit Khidr, pp. 80ff.
149 Cf. the references given above, note 144.
150 Both Dyroff (‘Wer ist Chadir?’, p. 327) and Friedlaender (Chadirlegende, pp. 116, 242) held
that there must have been versions in which the cook was called Glaukos. For al-Khiḍr as
a maritime figure, see Friedlaender, op. cit., pp. 116 ff.
151 Lidzbarski, ‘Wer ist Chadhir?’, pp. 109 f.
152 The explanation is that Wensinck’s entry ‘al-Khaḍir (al-Khiḍr)’ is a reprint from the first
edition of the ei, published in 1913–1936.
153 Guyard, ‘Bulletin critique’, pp. 344 f.
78 chapter 3
Conclusion
As noted already, the hypothesis presented in this paper is not amenable to

proof. It is a fact that Enoch had come to be identified with somebody known as
Idrīs by the time the exegetes were active, and there is no doubt that Atraḫasīs
and Enoch were similar, perhaps related, figures who grew even more alike in
the course of time. But that they actually fused at a popular level cannot be
demonstrated in the present state of the evidence, and the same is true of the
claim that the name Atraḫasīs lives on in that of Idrīs. Both propositions have
a fair degree of plausibility, however.
Whether the speaker of the Qurʾān himself had Enoch in mind when he
spoke of Idrīs is a good deal more uncertain. Obviously, if the name ‘Idrīs’ is
derived from ‘Atraḫasīs’, the answer has to be yes, regardless of whether God
had raised him to heaven or to an exalted position; but it cannot be ruled out
that the reference is to another figure. That Atraḫasīs (alias Zisudra and Ūta-
napishti) lived on, without any sign of fusion with Enoch, as the figure known
to the exegetes as al-Khiḍr is not in doubt. It is one out of several cases in which
the Muslim material preserves features of Akkadian origin not found in the
Alexander Romance.154 But it is impossible to tell whether the speaker of the
Qurʾān had the same figure in mind as did the exegetes. All that can be said
is that the exegetes are drawing on an ultimately Mesopotamian tradition on
which the Qurʾān itself may be drawing as well. This is also what they are doing
when they explain the angels Hārūt and Mārūt (2:102) as angels who came
down to earth in Enoch’s time and sinned, though these angels are not easy
to recognize in their Qurʾānic version. Modern scholars probably would not
have been able to identify them on the basis of the Qurʾān alone, though in this
particular case the Qurʾān and the exegetes are certainly drawing on the same
tradition. The ease with which the exegetes identified figures from the Enoch
literature contrasts strongly with their handling of other passages, where they
plainly do not know whom or what the Qurʾān is talking about and so resort to
guessing (as they do in connection with Dhū ʾl-Kifl, for example).
Why there should be such a high degree of continuity between the Qurʾān
and exegesis in connection with the Enoch material is hard to say, but it may
have something to do with the fact that the material originated in Iraq, where
most of the early exegetes were active. For it was evidently in Iraq, not Ethiopia,
154 Cf. Tesei, ‘Survival and Christianization’, pp. 418, n. 2; 426, n. 27 (citing Wensinck, ‘al-Khaḍir
(al-Khiḍr)’, and D. Bodi, ‘Les mille et une nuits et l’ épopée de Gilgamesh’, in A. Chraïbi (ed.),
Les mille et une nuits en partage, Paris 2004, pp. 407f., on Bulūqiyā).
that Enoch acquired the Babylonian name under which he appears in the
Qurʾān, if the thesis advanced here is accepted; and it was also in Iraq rather
than Ethiopia that the fallen watchers were reduced to two and endowed with
the Zoroastrian names of Haurvatāt and Ameretāt, to pass into the Qurʾān
as Hārūt and Mārūt.155 The Slavonic Enoch book (2 Enoch) must have some
connection with Iraq as well, since there are Zoroastrian features in its views
on animals and time;156 and the Hebrew Enoch book (3 Enoch, alias Sefer
Hekhalot) is assumed to have reached its final shape in Iraq in the sixth or
seventh century. In short, the Enoch literature was well known in Iraq and
probably more familiar to the exegetes active there than Qurʾānic material of
other provenance.
It is noteworthy that the Enoch literature continued to be read on the
Sasanian side of the Euphrates, for on the Greek side the Jews and Christians
had ceased to regard it as authoritative in the course of the third and fourth
centuries. Both had come to dislike the story of angels mating with humans
and now interpreted the biblical ‘sons of God’ as humans of elevated status.157
The rabbis were also wary of the idea of Enoch’s translation to heaven, which
they associated with heretics.158 They rarely mention Enoch, and they take a
poor view of him when they do. In a famous passage in Genesis Rabba one rabbi
interprets the biblical statement that ‘he was not’ to mean that Enoch was not
inscribed in the scroll of the righteous; another passage declares that Enoch was
sometimes righteous, sometimes wicked and that God took him in a righteous
phase (to save him from further sins); or what the Bible means when it says
155 See P.J. de Ménasce, ‘Une légende indo-iranienne dans l’angélologie judéo-musulmane: à
propos de Hārūt et Mārūt’, Études Asiatiques 1, 1947, pp. 10–18; P. Crone, ‘The Book of Watch-
ers in the Qurʾān’, in H. Ben-Shammai, S. Shaked and S. Stroumsa (eds.), Exchange and
Transmission across Cultural Boundaries: Philosophy, Mysticism and Science in the Mediter-
ranean, Jerusalem 2013 [Ed.: reprinted in P. Crone, The Qurʾānic Pagans and Related Mat-
ters, vol. 1 of Collected Studies in Three Volumes, ed. H. Siurua, Leiden 2016, art. 7], pp. 16–51.
156 S. Pines, ‘Eschatology and the Concept of Time in the Slavonic Book of Enoch’, in R.J. Zwi
Werblowsky and C.J. Bleeker (eds.), Types of Redemption, Leiden 1970, pp. 72–87; cf. F.I. An-
derson’s introduction to his translation of 2 Enoch in Charlesworth, Old Testament Pseud-
epigrapha, i, esp. p. 95.
157 Judges according to the rabbis, sons of Seth as opposed to descendants of Cain according to
the Christians; see for example B.J. Bamberger, Fallen Angels, Philadelphia 1952, pp. 78ff.,
91, 149 ff.; A.Y. Reed, Fallen Angels and the History of Judaism and Christianity: the Recep-
tion of Enochic Literature, Cambridge 2005. Further literature is cited in Crone, ‘Book of
Watchers’, nn. 11–20.
158 Genesis Rabba, 25:1: heretics asked R. Abbahu why they did not find any mention of Enoch’s
death (in Genesis).
80 chapter 3
that God took him is simply that he died, as we are also told.159 Around 600 ad
the circles viewed with suspicion by the rabbis surface in 3 Enoch, alias Sefer
Hekhalot, in which Enoch is the angel Metatron and the ‘lesser yhwh’, second
only to God himself.160
The Christians did not turn against Enoch as a person. They continued to
mention him in connection with the two eschatological witnesses of Revelation
11 (where John predicts that at the end of times, between the sixth and seventh
trumpets, two witnesses will come forth to give testimony, to be killed by the
beast of the abyss, revived after three and a half days, and then translated to
heaven). The witnesses are unnamed, but they were usually held to be Enoch
and Elijah, the two biblical figures who had not died.161 Other Christian works,
however, presented Enoch as living in paradise right now: thus for example the
much read Apocalypse of Paul, composed in Greek in probably the mid-third
century and translated into Syriac, Coptic and many other languages thereafter
(like Muḥammad, Paul met Enoch in heaven).162 That Enoch was translated
is also affirmed, for example, by Epiphanius (d. 403),163 Ephraem of Amida
(patriarch of Antioch under Justinian), Theodosius of Alexandria (d. 566) and
Timothy of Antioch (sixth/seventh century).164 Byzantine historians continued
to quote from the Enoch book as well, though not without warning their readers
159 Genesis Rabba, 25:1; VanderKam, Man for All Generations, pp. 161ff.; Alexander, ‘Jewish
Tradition in Early Islam’ (above, note 15), p. 17; M. Himmelfarb, ‘A Report on Enoch in
Rabbinic Literature’, in P.J. Achtemeier (ed.), Society of Biblical Literature 1978 Seminar
Papers, i, Missoula, mt, 1978, pp. 259–269.
160 Tr. P.S. Alexander in Charlesworth, Old Testament Pseudepigrapha, i, pp. 223–315. He
ascends to heaven and turns into Metatron, the great scribe, in Targum Ps.-Jonathan, too,
but not in the other targums (VanderKam, Man for All Generations, pp. 165–168; Orlov,
Enoch-Metatron Tradition).
161 VanderKam, Man for All Generations, pp. 180 ff.; id. and W. Adler, The Jewish Apocalyptic
Heritage in Early Christianity, Assen, mn, 1996, pp. 89ff.; cf. the History of Joseph the
Carpenter in J.K. Elliott (tr.), The Apocryphal New Testament, Oxford 1993, p. 115, pars.
31–32 (4th–5th century); Oecumenius (6th century?), Commentary on the Apocalypse, tr.
J.N. Suggit, Washington, dc, 2006, ch. 6, 4 (p. 102); Andrew of Caesarea (early 7th century),
Commentary on the Apocalypse, tr. E.S. Constantinou, Washington, dc, 2011, ch. 30, ad 11:3–
4 (pp. 131 f.); W. Bousset, The Antichrist Legend, Atlanta 1999, pp. 203ff.
162 In Elliott, Apocryphal New Testament, p. 628.
163 Epiphanius, Panarion, tr. F. Williams, Leiden 1987–1994, ii, p. 622 (heresy 79, 2:4).
164 Cf. D. Krausmüller, ‘Timothy of Antioch: Byzantine Concepts of the Resurrection, Part 2’,
Gouden Hoorn 5, no. 2 (1997–1998), http://goudenhoorn.com/2011/11/28/timothy-of-
antioch-byzantine-concepts-of-the-resurrection-part-2/ (unpaginated), at note markers
71, 85 (Timothy himself), 87 ff. (Ephraem of Amida) and 114 (Theodosius).
of corruptions ‘by Jews and heretics’. According to Jacob of Edessa, however, the
Enoch book had been unjustly anathematised. It was a genuine antediluvian
work in his view, and the only reason Athanasius (d. 373) had proscribed it was
that heretics in his time had incorporated the work into their library of secret
books.165
There is no sign in either the Qurʾān or the early exegetical tradition of the
rabbinic denigration of Enoch or of the Christian view of him as an eschatolog-
ical witness; but here as in the Christian tradition, Enoch is a prophet,166 and
the Qurʾānic association of Idrīs with ṣabr (endurance, patience), for which
there is no precedent in either the Bible or the Enoch literature, is perhaps
also rooted in the Christian tradition.167 Hārūt and Mārūt are still angels in the
Qurʾān, however, not human beings of elevated status, as both the Jews and the
Christians had come to affirm; so if we assume the Qurʾānic material on Enoch
and these two angels to have been transmitted by the same circles (which is
not certain), the circles in question would seem to be Iraqis who had parted
ways with mainstream Christianity by the third or fourth century, to develop
along lines of their own. This fits the Manichaeans, who certainly liked Enoch
and read books ascribed to him, but the Qurʾānic material is not likely to go
back to them.168 In fact, the circles in question were not necessarily sectarian
at all, as opposed to simply poorly policed by the rabbis, churchmen or Zoroas-
trian priests. It may be that just as the Arab conquerors inadvertently turned
the social map of the Near East upside down,169 so they inadvertently elevated
marginal traditions to high cultural status.
165 W. Adler, ‘Jacob of Edessa and the Jewish Pseudepigrapha in Syriac Chronography’, in
Reeves, Tracing the Threads, p. 145.
166 Cf. Reed, Fallen Angels, pp. 152 f.
167 Cf. the Apocalypse of Paul in Elliott, Apocryphal New Testament, p. 644; in E.A.W. Budge, tr.,
Miscellaneous Coptic Texts, London 1915, p. 1076, where Enoch declares that ‘the sufferings
which a man endures for the sake of God God will not afflict him with when he leaves the
world’.
168 My reasons for doubting that there is thought of Manichaean origin in the Qurʾān are
presented in P. Crone, ‘Jewish Christianity and the Qurʾān (Part Two)’, Journal of Near
Eastern Studies 75, 2016 [Ed.: reprinted in P. Crone, The Qurʾānic Pagans and Related
Matters, vol. 1 of Collected Studies in Three Volumes, ed. H. Siurua, Leiden 2016, art. 10],
section no. 10.
169 Cf. P. Crone, The Nativist Prophets of Early Islamic Iran: Rural Revolt and Local Zoroastrian-
ism, New York 2012, p. 17.
chapter 4
92 Abū Saʿīd al-Ḥaḍrī and the Punishment of

Unbelievers*
In his al-Imtāʿ waʾl-muʾānasa, the littérateur al-Tawḥīdī (d. 414/1023) tells of a

theologian called Abū Saʿīd al-Ḥaḍrī/Ḥuṣrī/Ḥaṣīrī/Ḥuḍarī/Ḥaḍramī who held
that God would admit all human beings to Paradise. This passage, which is of
considerable interest for the intellectual climate of ninth-century Baghdad, has
been translated and discussed by Van Ess, who has also assembled the little we
know about the theologian:11 his ism was al-Ḥasan b. ʿAlī and he was a Basran
Shīʿī of the mid to late ninth century; originally he was a Muʿtazilī of the Ṣūfī
variety; later, according to Ibn al-Nadīm, his mind became unhinged and he
struck out of his own (khulliṭa wa-abdaʿa).2 The passage, which must refer to
Abū Saʿīd’s unorthodox phase, goes as follows:3
Abū Saʿīd al-Ḥaḍramī, one of the clever theologians (ḥudhdhāq al-muta-

kallimīn) in Baghdad and the one who openly professed belief in the
equipollence of proofs (wa-huwa ʾlladhī taẓāhara biʾl-qawl bi-takāfuʾ al-
adilla),4 said that if God is just, generous, munificent, omniscient, kind
and merciful, He will admit all human beings ( jamīʿ khalqihi) to Paradise,
for all in their different ways endeavour to seek His pleasure and avoid
His anger as far as their knowledge and intelligence allows. They only fail
to follow His command because they have been deceived. Falsehood has
been decked out as truth for them. They are like a man carrying a gift to
93 a king who | was stopped on the way by people engaging in deception,
trickery and theft. They set up a man and called him by the name of the
king he was travelling to, so he handed the gift to them. If the king he
* I should like to thank Fritz Zimmermann for reading an early draft of this article and Michael
Cook for reading the final version.
1 Van Ess, tg, vol. 4, pp. 91–93, 333, with the translation at vol. 5, p. 344; see also Monnot,
Penseurs musulmans et religions iraniennes, pp. 61–63.
2 Van Ess’ translation reads khallaṭa, “spread confusion”, which is also possible, but cf. tg, vol. 2,
p. 4, note 1. My thanks to F. Zimmermann for the reading adopted here.
3 Tawḥīdī, Imtāʿ, vol. 3, pp. 192 f.
4 Not everyone who believed in takāfuʾ was willing to say so openly (cf. Tawḥīdī, Muqābasāt,
nos. 35, 54, pp. 159, 227).

abū saʿīd al-ḥaḍrī and the punishment of unbelievers 83
meant to see was generous, he would excuse him and have mercy on him
and treat him with extra generosity and kindness when he learnt of what
had happened to him. That would be more proper to him (awlā bihi) than
getting angry and punishing him.
In brief, God as normally conceived was mean: He punished people who failed
to worship Him even though He knew full well that they were innocent victims
of deception. Abū Saʿīd does not seem to think that God really is mean or that
therefore there must be a higher God above Him, but rather that God really is
generous and merciful and that therefore He cannot engage in the behaviour
imputed to Him. His statement is one out of many arguments mounted by
ninth-century theologians from a dualist background against the punitive God
of the Judaic tradition. But who are the tricksters and precisely whom are they
deceiving?
According to Van Ess, the tricksters are theologians and their victims are
sinners: Abū Saʿīd’s message is that God will admit all human beings in the
sense of all Muslims to Paradise, even sinners, because all would worship the
true God if only they followed reason, but they are misled by the theologians,
who offer lies and enrich themselves at the expense of simple folk. Van Ess
does wonder whether Abū Saʿīd meant to include unbelievers along with the
sinners, but he leaves it uncertain. He also observes that one could read the
tricksters as false prophets, noting that Abū Saʿīd’s parable would in that case
give us something approaching the “three impostors” thesis, i.e., the idea that
Moses, Jesus and Muhammad were tricksters who used religion to accumulate
worldly power.5 He does not accept this reading, however. In what follows I
shall argue that the victims are indeed unbelievers, but that the tricksters are
neither theologians nor false prophets; rather, they are demons. I shall conclude
with some further thoughts on Abū Saʿīd’s views.
Who Were the Tricksters?
The tricksters are unlikely to be theologians because Abū Saʿīd’s parable is

plainly about false religions, not theological fabrications. The traveller is made
to honour a man impersonating the king, i.e. a false deity; he | is misled into wor- 94
shipping somebody other than God, not simply into having some wrong ideas
about him. One could hardly blame the mutakallimūn for the existence of false
5 Van Ess, tg, vol. 4, pp. 93, 333 (small print).

84 chapter 4
religions. It does not even come easily to see them as corrupting the beliefs of
simple folk, for they were normally accused of doing the very opposite, namely
making religion so abstruse that simple folk could not understand them. They
were guilty of takfīr al-ʿawāmm, holding ordinary people to be unbelievers for
taking their religion on trust even in respect of fundamentals. Abū Saʿīd may
have written against takfīr al-ʿawāmm, for he is credited with a book denying
the superiority of theologians over the common people ( fī taswiyat aṣḥāb al-
kalām biʾl-ʿawāmm).6 That he saw them as corrupting the common people is
not implied. If the choice is between understanding the tricksters as theolo-
gians or as false prophets, it surely comes much more naturally to see them as
prophets. As Van Ess notes, some later readers may actually have understood
them as such,7 perhaps even ʿAbd al-Jabbār (d. 415/1025), who repeatedly men-
tions Abū Saʿīd along with Abū ʿĪsā al-Warrāq, Ibn al-Rāwandī and their likes as
Shīʿites guilty of slandering God and the prophets.8 But since ʿAbd al-Jabbār on
one occasion includes Hishām b. al-Ḥakam in the list of slanderers, he is prob-
ably indulging in polemical exaggeration.9 Van Ess is in any case right that the
tricksters are unlikely to be false prophets, for they are envisaged as operating
as a team rather than following one another. Unlike both pseudo-prophets and
theologians, moreover, they enrich themselves at God’s expense, not at that of
the traveller.
By Abū Saʿīd’s time, however, there was a long tradition in the eastern
Mediterranean of comparing God with a human king in order to illustrate His
relationship with other celestial beings, usually angels, but in the case of the
Christians also demons. This tradition was shared by monotheists of both the
pagan and the Biblical type, and it is above all in polemics between them that it
is attested. I shall now give a brief aperçu of how the different groups used the
imagery to show that Abū Saʿīd’s parable continues the usage of the Christians.
Late antique pagans liked to defend their polytheist heritage by casting God
as a king who ruled with the assistance of largely autonomous governors after
the fashion of such monarchs as the Persian emperor. Zeus had appointed the
lesser gods to the various regions of the world and they were like his governors
and satraps, Aelius Aristides (d. 181 or later) said.10 One God was king of all and
95 many gods ruled together | with him, according to Maximus of Tyre (d. 185).11
6 Van Ess, tg, vol. 5, p. 344 (from ʿAbd al-Jabbār, Tathbīt, p. 51.–6).
7 Van Ess, tg, p. 333 (small print), in the context of the three impostors thesis.
8 ʿAbd al-Jabbār, Tathbīt, pp. 51, 129.
9 ʿAbd al-Jabbār, Tathbīt, p. 232, supra.
10 Orations, xliii, 18.
11 Maximus of Tyre, Orations, xxxix, 5.
God had allotted different parts of the earth to different overseers, Celsus
(c. 180) and Julian (d. 361) agreed in polemics against the Christians.12 Celsus
added that one should pay due reverence to all beings who had been allotted
control by God over earthly things: for just as the satrap or sub-governor of the
Persian or the Roman emperor and other officials, including lesser ones, could
do one much damage if they were slighted, so it went without saying that all
God’s underlings could cause much harm if they were insulted.13 Ambrosiaster,
writing in fourth-century North Africa, tells us that if one asked a pagan how he
could worship a whole lot of gods, he would reply that they were like dignitaries
interceding with the sovereign on his behalf.14 God delegated matters to such
dignitaries because it would be unseemly for him to attend to the details of
petty administration, just as it was below the dignity of a human king such as
Xerxes to do so, according to the first-century Pseudo-Aristotelian De Mundo,
where the comparison between God and the Persian emperor is developed at
length.15 In the same vein a fragment attributed to the Zoroastrian Mazdak
(d. 530s) depicts God as seated on his throne as Khusraw sits on his in the lower
world; in front of God and Khusraw alike are four powers, who rule through
seven powers, and so on.16
The pagans never seem to envisage God as a king in connection with mali-
cious powers. They did see a link between such powers and false religious
claims: thus Celsus entertained the possibility that Jesus and other wonder-
workers were “wicked men possessed by an evil demon”,17 while the mushrikūn
immortalized in the Qurʾān asked themselves whether there was a spirit ( jinna)
in the man who claimed to have been sent to them, when they did not simply
dismiss him as mad (majnūn).18 But the demons are not cast as usurpers of the
prerogatives of the supreme God in these examples, nor is there any suggestion
that they took possession of their victims with a deliberate intention to mislead
mankind.
The combination of God as king and demons as usurpers also seems to
be missing on the Gnostic side, though evil powers actively seeking to | trick 96
12 Celsus in Origen, C. Celsum, v, 25; Julian, Against the Galilaeans, p. 402 (= Cyril, Pro
Christiana Religione, 290e).
13 In Origen, C. Celsum, viii, 33, 35; cf. also vii, 68.
14 Cumont, “Polémique”, pp. 426 f. Compare Celsus in Origen, C. Celsum, viii, 2. The mushri-
kūn say much the same in Qurʾān 39:3, but without the governmental imagery.
15 Aristotle (attrib.), De Mundo, ch. 6, pp. 398a–b.
16 Shahrastānī, Milal, p. 193.
17 Origen, C. Celsum, i, 68.
18 Qurʾān 15:6; 34:8; 37:36; 44:14; cf. 26:27, where Pharaoh dismisses Moses as majnūn.
86 chapter 4
people into worshipping false deities are extremely common here. In fact, it is
typically the evil powers that are cast as rulers (prince of darkness, archons, and
so on), not the hidden God, who was apparently too pure and too transcendent
to be conceived in terms relating to government.
The governmental image reappears when we turn to the Jews, however.
According to Philo (died ca. 50 ce), it would be most unwise to give the same
tribute to the creatures as to their maker, just as it would be most unwise to
give subordinate satraps the honour due to the great king.19 A famous rabbinic
vignette conveys much the same message by depicting a king as sitting in
a chariot together with a governor: when the subjects mistakenly greet the
governor as lord, the king pushes the governor out of the chariot. The rabbis
mention this in illustration of God’s response when the angels mistook Adam
for a divine being: God pushed Adam out of the chariot by putting him to sleep,
thereby demonstrating that he was a mere mortal.20 Humans were all too prone
to casting Adam or a principal angel such as Metatron as God’s vice-regent and
magnifying his position to the point where it rivalled God’s. A famous story tells
of a third-century rabbi who made a mystic ascent to heaven, where he mistook
the angel Metatron in all his glory for God. On this occasion, too, God pushed
the governor out of his chariot, this time by having Metatron whipped and the
rabbi excommunicated.21 In all three examples, the lesser beings are legitimate
subordinates of God, however, and though humans sometimes overdo their
worship of them, there is no suggestion that the subordinates are trying to
mislead them.
The Jews were also familiar with malicious celestial powers, and like the
pagans they would invoke them in explanation of false religious claims. In
the Gospels, for example, they sometimes react to Jesus by dismissing him as
possessed: “he has a demon and is out of his mind”, as many of them said with
reference to his presumptuous statements (John 10:19); “you have a demon”,
they insisted when he denied it (John 8:48f., 52); “he has Beelzebub, and by
the ruler of the demons he casts out demons”, the scribes said (Mark 3:22). In
fact, fallen angels and demons had played a major role in the explanation of
evil among Jews in the Greek and Roman periods, and in the Book of Watchers
(part of the Book of Enoch), perhaps dating from the third century bce, it is
demons who are responsible for the existence of idolatry: the fallen angels
here generate evil spirits which lead people into error by inducing them to
19 Philo, De Decalogo, p. 61.

20 Genesis Rabba, viii, 10; for a translation and further references, see Schäfer, Rivalität
zwischen Engeln und Menschen, p. 82.
21 Deutsch, Guardians of the Gate, chs. 3–4.
offer sacrifices to these spirits themselves in the mistaken belief | that they are 97
gods.22 The idea that the gods venerated by the pagans were actually demons
is also encountered in the Septuagint.23 But the Jews did not to my knowledge
cast the demons who led mankind astray as usurpers of the prerogatives of the
true king; and in any case the rabbis played down the idea of demonic powers
as it rose to prominence in Gnosticism and Christianity.24
It is among the Christians that we find the right combination of God as
king and demons as usurpers of His prerogatives. According to the Christians,
the analogy between divine and human kingship did not serve to vindicate
polytheism, as the pagans claimed; rather, it refuted it, for monarchy was the
best constitution: polyarchy meant anarchy, so that if there were many gods,
all things would go to pieces.25 (This argument also appears in the Qurʾān.)26
A pagan philosopher, perhaps Porphyry (d. c. 305 ce), retorted that a monarch
is unique in being a ruler, not in being a human: on the contrary, one would
not call him a king at all if he did not rule over other human beings, only
over beasts; it followed that God would not be king at all if he did not rule
over other gods, only over humans.27 To this and other pagan arguments the
Christians responded, much like Philo, that if a servant of the king allowed
himself to be called Caesar, both he and those who had called him by that name
would perish.28 It was quite wrong to claim that God’s underlings would harm
those who slighted them by refusing to call them gods, Origen (d. 254 or 255 ce)
explained in refutation of Celsus, for the angels were true satraps, subordinate
governors and officers of God. If demons had the ability to hurt people, it was
22 1 Enoch 19 (in Charlesworth, ed., Old Testament Pseudepigrapha, vol. 1, p. 23).

23 Psalms 96:5 declares that “the gods of the nations are idols”. The Septuagint (95:5) rendered
this as “the gods of the nations are demons”.
24 Cf. Bamberger, Fallen Angels, ch. 16.
25 Lactantius, Divine Institutes, i, 3; Eusebius, Laus Constantini, iii.6 (trans. Drake, In Praise of
Constantine, p. 87); Gregory of Nazianzus, Discours, no. 29, 2; Gregory of Nyssa, Poemata,
in Dvornik, Early Christian and Byzantine Political Philosophy, vol. 2, p. 689 (mpg, vol. 37,
p. 414); cf. also the tenth-century Moses Bar Kepha, Hexaemeronkommentar, ch. 3, 9 (trans.
Schlimme, p. 101).
26 Qurʾān 21:22; cf. also 17:42 (here with echoes of the old combat myth, cf. Forsyth, Satan and
the Combat Myth).
27 Macarius Magnes, Apocriticus, iv, 20, 2 (English trans. Hoffmann, Porphyry’s Against the
Christians, pp. 83f.). For a thorough discussion of this work and the philosopher it refutes,
see Goulet’s edition and French translation, vol. i.
28 Ambrosiaster and Pseudo-Maximus of Tyre in Cumont, “Polémique”, p. 427. The philoso-
pher in Macarius Magnes disagrees again (Apocriticus, iv, 23, 3; trans. Hoffmann, Por-
phyry’s Against the Christians, p. 88).
88 chapter 4
precisely because they had not received any appointment from God, but were
evil powers who would cause suffering to those who submitted to them. Origen
98 implies that even Christians were known to submit themselves “to the demon |
of the locality”; a real Christian, however, meaning one who submitted himself
to God alone and His Logos, would be safe from such powers, for the angel of
the Lord would be with him.29
Here the demons seem to be envisaged as local power-holders of an ille-
gitimate kind, such as barbarian usurpers, warlords, or robbers; and though
Origen does not say so, the Christians held such usurpers to be trying actively
to lead people astray. The demonic offspring of the fallen angels had enslaved
the human race, among other things by teaching people how to offer sacrifices,
incense and libations to them, as Justin Martyr (d. 160s ce) said, developing
the theme from the Book of Watchers.30 That demons were the forces behind
paganism became the standard Christian view: evil spirits lurked behind the
idols, coming out in all their hideousness when the idols were cut down (as
they were to do in Muhammad’s Arabia too; early Muslims also held that it
was demons [al-jinn] who made infidels worship idols and ascribe partners
to God).31 According to Eusebius (d. 340ce), “spirits and demons, also called
principalities, powers, world-rulers, spiritual hosts of wickedness”, hate God so
much that “they wish themselves to be proclaimed gods and steal away for
themselves the honours intended for God, and attempt to entice the simple
by divinations and oracles as lures and baits”.32 Here the imagery is very close
indeed to Abū Saʿīd’s, though it is only implicitly that God is cast as king.
The imagery reappears in a work by the Christian mutakallim Theodore
Abū Qurra (d. ca. 825ce) on how to identify the true religion. Like Abū Saʿīd
al-Ḥaḍrī, he uses a parable: a king had a son who went away on a journey
and fell ill; the king sent a messenger with a prescription that would cure
him, but the king’s enemies heard of this and sent their own messengers with
harmful prescriptions, hoping to harm the king and his son; their plot was
foiled by a wise physician accompanying the son: he told him to scrutinize
all the messages to determine which was the right one, and only one proved
to be true.33 The king was God, the son was Adam/mankind, and the wise
29 Origen, C. Celsum, viii, 36, cf. also 33.

30 Second Apology, 5 (trans. Barnard, pp. 76f.). Cf. 1 Enoch 19 (above, note 22); Reed, Fallen
Angels, p. 164.
31 Thus, for example, al-Ḥasan al-Baṣrī in Fakhr al-Dīn al-Rāzī, Tafsīr, vol. 13, p. 115, ad 6:100.
32 Eusebius, Praeparatio Evangelica, vii, 16.9 f.
33 Griffith, “Faith and Reason in Christian Kalām”, pp. 34f., citing Abū Qurra, Traité de
l’ existence du créateur et de la vraie religion, viii.
physician was reason, Abū Qurra explains, adding that God’s enemies were
the demons (al-shayāṭīn).34 In both his and Abū Saʿīd’s parable, the demons
interfere with communications between the king and his subjects, in the one
by | sending false messages to the travellers, in the other by falsely giving them 99
to understand that they have arrived at their destination. In both, the object
of the exercise is to divert royal prerogatives to the illegal operators, stealing
honours intended for God as Eusebius puts it. In short, it comes naturally to
read the evil-doers in both as demons.
By the ninth century, idolatry was no longer a problem. What troubled
people now was the existence of rival scriptural religions, and Abū Qurra’s
demons no longer operated as they did in Eusebius’ time: instead of seducing
people into worshipping idols they now sent messengers in imitation of the
true God. Their behaviour is shaped by the rise of Islam, in other words; the
paradigmatic bearer of a false message is clearly Muhammad, whom Abū
Qurra characterizes as a false prophet possessed by a demon elsewhere as
well.35 It cannot be said that the adaptation of the old imagery to the new
conditions is entirely felicitous, however, for the demonic explanation only
works in connection with false religions, not when we add superseded ones.
Abū Qurra inadvertently suggests that even Moses was an impostor, given that
all the messengers other than the one true one are sent by the enemies of God.
In Abū Saʿīd’s parable the demons even operate in the old style by setting up
one of themselves as a rival god, which is hardly a good characterization of any
of the religions with which Islam was in competition; and again no distinction
is made between false religions and superseded ones, so that Moses and Jesus
are implicitly put on a par with figures such as Mani or Zoroaster. Maybe both
authors were using the old imagery in an offhand manner, or maybe they would
have explained that demons worked through many kinds of people: pseudo-
prophets in some cases, rabbis and priests or the obstinate infidels themselves
in others. But both parables would have worked better if the demons had been
envisaged as working in the same way in all cases.
When we meet the demonic explanation again, it is precisely in connection
with the view that all prophets were victims of demonic trickery. According
to Abū Bakr al-Rāzī (d. 313/925), “the souls of evildoers who have turned into
demons show themselves in the form of angels, who come to people and com-
mand them to go and tell people that an angel has appeared to them and told
them that God has given them prophethood … with the result that discord
34 Abū Qurra, Traité, viii, 33 (p. 217).

35 Meyendorff, “Byzantine Views of Islam”, p. 120.
90 chapter 4
appears among people”.36 Here the demons are imitating Gabriel, the paradig-
matic prophet being Muhammad yet again. Al-Rāzī hardly meant the explana-
tion literally; rather, he was using mythical language for didactic purposes to
100 show how | his view of the prophets, above all the Prophet, fitted in with the
historical record and to bring out that he took them to believe in their own
mission even though he did not believe in it himself. Demons were no longer
routinely invoked in explanation of evil by his time, however; they sound curi-
ously out of date even here. Once they had been discarded, the explanation of
false prophets had to be that they were cynical manipulators rather than inno-
cent victims of deception, for now they were acting on their own, yet every bit as
evil as before. In effect, the removal of the demons simply secularised the expla-
nation: the pseudo-prophets turned into demons stripped of their supernatural
status. It was in this guise, smacking of conspiracy theory, that the concept of
the three impostors was exported to Europe.37
Abū Saʿīd’s Position
Even if Abū Saʿīd’s tricksters had been false prophets, his parable would not
have been an early version of the “three impostors” thesis, for like Abū Qurra’s,
it is based on the assumption that there was a true religion, centered on wor-
ship of the real king. At least one revealed religion is right, and one assumes it
to be Abū Saʿīd’s own. In keeping with this, his parable is not in fact concerned
with the question how far people can reach God by rational means, unaided
by prophets (or for that matter theologians), but rather with the importance
of their intentions: all humans do their best to please God in their very differ-
ent ways (ʿalā ikhtilāfihim), he says, presumably meaning that all try to please
Him even though they belong to different religious communities. The issue he
is addressing is whether God is being fair to those of them who are in the wrong
communities. Since Abū Saʿīd openly professed belief in the doctrine of takāfuʾ
al-adilla, the sceptical view that an argument in favour of a particular propo-
sition could always be matched by another of equal weight to the contrary,38
he plainly cannot have regarded reason as a better guide to truth than prophet-
hood.
36 Nāṣir-i Khusraw, Jāmiʿ al-ḥikmatayn, in Kraus, Rasāʾil, p. 177; also translated in Stroumsa,
Freethinkers, p. 106.
37 Cf. Van Ess, tg, vol. 4, p. 333, for literature.
38 Cf. Hankinson, The Sceptics, p. 27; Van Ess, Erkenntnislehre, pp. 221ff.
On the contrary, his problem must have arisen from the very fact that reason
did not offer any guidance here. In Abū Qurra’s parable the wise physician
shows the prince how to tell the difference between healing and harmful
prescriptions: one could tell a true revelation from a false one by rational
means. This is precisely what Abū Saʿīd denied with his doctrine of takāfuʾ.
What his parable is saying is surely that it would be | unfair of God to punish 101
those who have been duped by demons, for all would follow the truth, if
only they knew what it was. All have the best of intentions, all are trying to
please Him to the best of their ability; it is precisely their reason which is
deficient. How were people to guard themselves against tricksters if they did
not even know when they were being deluded? Their sharpened intellects
notwithstanding, theologians were not in fact in a better position than anyone
else, for their attempts to establish criteria of judgement came to grief on the
equipollence of proofs. One would assume this to be what Abū Saʿīd said in his
book Fī taswiyat aṣḥāb al-kalām biʾl-ʿawāmm. It would certainly do something
to explain why ʿAbd al-Jabbār found it deeply offensive.39
If one could not tell a true religion from a false one, what was Abū Saʿīd’s own
faith? Al-Tawḥīdī has a wonderful vignette of a sceptic who decides to stay in
the religion he has grown up in on the grounds that if one does not know where
the truth is, one may as well stay where one is.40 This was also a well-known
reaction of sceptics in antiquity: entertaining a rational distrust of reason, they
practised suspension of judgement and so were apt to cope with the problem
of what to do and think by following tradition.41 One would assume Abū Saʿīd
to have reacted similarly, for there is no suggestion that he abandoned Islam.42
In fact, he seems to have remained not only a Muslim, but also a Muʿtazilī.
That he remained a Muʿtazilī is suggested by a comparison of his presenta-
tion of the problem of God’s justice with that of his contemporary, the Zoroas-
trian Martān Farrūkh. What the latter disliked about the Muslim conception of
God was not just that He punished unbelievers, but also that He punished peo-
ple for evil that He Himself had created and made them follow. Martān Farrūkh
could not see how such a God could possibly be called just, merciful or wise. If
God was just and wise, as he believed Him to be, He could not be omnipotent:
evil had to have autonomous existence. Martān Farrūkh mentions omniscience
39 Cf. his Tathbīt, 51.–6.

40 Imtāʿ, vol. 3, pp. 193 f.; cf. Van Ess, “Skepticism”, pp. 6 f.
41 Burnyeat, “Can the Sceptic Live his Scepticism?”, p. 33; Schofield, “Cicero for and against
Divination”, pp. 55 f. (my formulation reflects his).
42 There was a Manichaean Abū Saʿīd about the same time, but he does not seem to be
identical with ours (Van Ess, tg, vol. 4, p. 92).
92 chapter 4
as a problem too, but it is the incompatibility of omnipotence and justice that

he stresses time and again.43 By contrast, Abū Saʿīd makes no reference to God’s
omnipotence at all, only to His omniscience. That God should punish people
for evil that He Himself had created and made His servants prone to follow did
not apparently trouble him: he does not ask how the tricksters came into being
102 or why God allows them to mislead people. | He may of course have done so
in other works of his, but apparently he had found an answer, for it would not
otherwise have made sense for him to worry about the subsidiary problem of
God’s punishment of infidels on its own. This is explicable on the assumption
that he was still a Muʿtazilī: God did indeed have power over all things, but this
did not make Him unfair, for He had given people free will; what was unfair
was only that He should punish them even when their free will could not help
them, i.e., when rational choice was rendered impossible by takāfuʾ al-adilla.
According to Ibn Ḥazm, believers in the equipollence of proofs fell into three
groups. The first took the doctrine to mean that one could neither prove nor
disprove the existence of God or anything that followed from it. The second
accepted the existence of God and affirmed that the truth was available in
some belief system or other, but held takāfuʾ to rule out certainty as regards
the rest. The third accepted that the true religion was Islam and so limited
the applicability of takāfuʾ to inner-Islamic divisions (an anti-sectarian use of
scepticism of the type first attempted by the Murjiʾīs).44 Abū Saʿīd appears to
belong in the second group, for although his parable is based on the assumption
that Islam is the true religion, he does not limit his takāfuʾ to inner-Islamic
divisions. This was what made him heterodox: only one religion was true, but
given the inability of human reason to tell which one it was, he held that one
could be saved in all of them. No doubt he said this too in his book Fī taswiyat
aṣḥāb al-kalām biʾl-ʿawāmm, and it was probably this feature that struck Ibn
al-Nadīm as bidʿa:45 it overstressed the importance of good intentions.
Though Abū Saʿīd was far less radical than Abū ʿĪsā al-Warrāq and Ibn al-
Rāwandī, with whom ʿAbd al-Jabbār associates him and with whom he seems
to have been personally acquainted too,46 he shared with them the feature
of being a Muʿtazilī mutakallim with strong Shīʿī sympathies who fell into
zandaqa, which in a ninth-century context meant heresy of a vaguely Iranian
43 Martān Farrūkh, Škand-Gumānīk Vičār, ch. 11.

44 Ibn Ḥazm, Faṣl, vol. 5, pp. 118 f. For the Murjiʾīs and scepticism, see Cook, Early Muslim
Dogma, ch. 7.
45 Cf. above, note 2.
46 Ashʿarī knew them to have participated in a debate together in Baghdad (Van Ess, tg, vol. 6,
p. 364 [no. 43]).
and thoroughly rationalist kind. All three were troubled by the behaviour of
an all-powerful God who declared Himself to be just and merciful, but who
nonetheless inflicted eternal pain on tiny beings that He had made Himself;
and all three allowed reason to sit in judgement of the revelation, though they
did not all go so far as to reject it altogether. Unlike his associates, however,
Abū Saʿīd does not seem to have impressed other mutakallims. The view that
God | would not punish infidels, or anyone, or that at least He would not do so 103
for ever, is aired at some length in al-Qirqisānī (10th century) and Fakhr al-Dīn
al-Rāzī (d. 606/1209).47 Neither mentions any names, but both reproduce the
objections of Abū ʿĪsā and Ibn al-Rāwandī to divine punishment, namely that
it would amount to inflicting harm of no benefit to either God or the victims,
which was morally repugnant (qabīḥ),48 and that God knew in advance that the
infidel would not believe: since He created human beings for beneficial rather
than harmful purposes, He could not have given them obligations that He knew
He would have to punish them for eternally, as the argument continues in
the formulation of later mutakallims.49 In Fakhr al-Dīn al-Rāzī, the opponents
of divine punishment add that God is the creator of the impulses that led to
sin, for the stupidity and foolishness that cause people to disobey God are
not something they have chosen themselves, but rather something built into
their natures (al-aḥwāl al-gharīziyya), so that it would be morally repugnant
for God to punish them for it. (“Should the forms be ugly, whose fault is
it?”, as ʿUmar Khayyām asked.)50 And even if one accepted that He would in
fact punish them, why should He do so for ever?51 The Qurʾān did of course
threaten unbelievers with eternal punishment, but God was not obliged to
carry out His threats, and even the sternest human master who inflicted that
kind of torments on his slaves would eventually be moved to forgive them.
God’s words to the unbelievers in Qurʾān 2:7, “theirs is a mighty punishment”
(lahum ʿadhābun ʿaẓīm), simply meant that they deserved such punishment,
47 Qirqisānī, Anwār, iii, 9 (vol. 2, pp. 246 ff.); Fakhr al-Dīn al-Rāzī, Tafsīr, vol. 2, pp. 54ff. (ad
Qurʾān 2:7), vol. 27, pp. 74 f. (ad Qurʾān 40:56–60).
48 Cf. Abū ʿĪsā al-Warrāq in Tawḥīdī, Imtāʿ, vol. 3, p. 192 (trans. Van Ess, tg, vol. 6, p. 432); Ibn
al-Rāwandī in Ibn al-Jawzī, Muntaẓam, vol. 6, pp. 101 (sub anno 298).
49 The early formulation was simpler and ruder, cf. Abū ʿĪsā al-Warrāq, loc. cit.; Ibn al-
Rāwandī in Khayyāṭ, Intiṣār, p. 12.5. It is also simpler in Qirqisānī, who attributes it to the
Manichaeans (Anwār, iii, 9, 9 f.).
50 Dāya, God’s Bondsmen, trans. Algar, p. 54.
51 Compare Ibn al-Rāwandī on how a God who condemns people who disobey or do not
believe in Him to eternal, everlasting hellfire is stupid and ignorant of the right measure
of punishment (lā ʿālim bi-maqādīr al-ʿiqāb) (Khayyāṭ, Intiṣār, p. 12.6).
94 chapter 4
but His magnanimity would necessarily make Him forgive them. (To all these
arguments Fakhr al-Dīn al-Rāzī laconically replies that God is above human
reasoning.) One wonders why Abū Saʿīd’s argument has been left out. It does
come across as rather Ṣūfī in its concern with the human heart where the other
mutakallims focus on the nature of God. Maybe even those who agreed with
him found him to overstress the importance of good intentions.
104 Bibliography
ʿAbd al-Jabbār. Tathbīt dalāʾil al-nubuwwa. ʿA.-K. ʿUthmān, ed. Beirut, 1966.
Abū Qurra. Traité de l’existence du créateur et de la vraie religion. I. Dick, ed. Jounieh
and Rome, 1982.
Aelius Aristides. Orations. In C.A. Behr, trans., The Complete Works. Leiden, 1981–1986.
Aristotle (attrib.). De Mundo. J. Tricot, trans. Paris, 1949.
Bamberger, B.J. Fallen Angels. Philadelphia, 1952.
Burnyeat, M.F. “Can the Sceptic Live his Scepticism?”. In M. Schofield, M. Burnyeat and
J. Barnes, eds., Doubts and Dogmatism. Oxford, 1980.
Celsus, see Origen.
Charlesworth, J.H., ed. The Old Testament Pseudepigrapha. New York, 1983–1985.
Cook, M. Early Muslim Dogma. Cambridge, 1981.
Cumont, F. “La polémique de l’Ambrosiaster contre les paiens.” Revue d’Histoire et de
Littérature Religieuse 8 (1903): 417–440.
Dāya, Najm al-Dīn Rāzī. The Path of God’s Bondsmen. H. Algar, trans. New York, 1982.
Deutsch, N. Guardians of the Gate. Angelic Vice Regency in Late Antiquity. Leiden, 1999.
Dvornik, F. Early Christian and Byzantine Political Philosophy. Washington, dc, 1966.
Ess, J. van. Die Erkenntnislehre des ʿAḍudaddīn al-Īcī. Wiesbaden, 1966.
. “Skepticism in Islamic Religious Thought”. Al-Abḥāth 21 (1968): 1–18.
. tg = Theologie und Gesellschaft im 2. und 3. Jahrhundert Hidschra. Berlin and
New York, 1991–1997.
105 Eusebius. Laus Constantini. H.A. Drake, trans., In Praise of Constantine. Berkeley, 1976.
. Praeparatio Evangelica, book vii. G. Schroeder and É. des Places, eds. and
trans. Paris, 1975.
Fakhr al-Dīn al-Rāzī. Al-Tafsīr al-kabīr. Tehran, 1413.
Forsyth, N. The Old Enemy: Satan and the Combat Myth. Princeton, 1987.
Gregory of Nazianzus. Discours. P. Gallay, ed. and trans. (nos. 27–31). Paris, 1978.
Griffith, S.H. “Faith and Reason in Christian Kalām: Theodore Abū Qurrah on Discern-
ing the True Religion”. In S.Kh. Samir and J.S. Nielsen, eds., Christian Arabic Apolo-
getics during the Abbasid Period. Leiden, 1994.
Hankinson, R.J. The Sceptics. London and New York, 1995.
Ibn Ḥazm. Al-Faṣl fī ʾl-milal waʾl-ahwāʾ waʾl-niḥal. Cairo, 1317.

Ibn al-Nadīm. Kitāb al-fihrist. R. Tajaddud, ed. Tehran, 1971.
Ibn al-Jawzī. Al-Muntaẓam, vol. 6. Hyderabad, 1357.
Julian. Against the Galilaeans, in his Works. W.C. Wright, ed. and trans. Loeb Classical
Library, vol. 3. Cambridge, Mass., and London, 1923.
Justin. The First and Second Apologies. L.W. Barnard, trans. New York, 1997.
Al-Khayyāṭ. Kitāb al-intiṣār. A.N. Nader, ed. and trans. Beirut, 1957.
Kraus, P., ed. Rasāʾil falsafiyya li-Abī Bakr … al-Rāzī, vol. 1 (no more published). Cairo,
1939.
Lactantius. The Divine Institutes, books i–vii. M.F. MacDonald, trans. Washington, d.c.,
1964.
Macarius Magnes. Apocriticus. R. Goulet, ed. and trans., Le Monogénès. Paris, 2003.
R. Hoffmann, partial trans., Porphyry’s Against the Christians. Amherst, 1994.
Martān Farrūkh. Škand-Gumānīk Vičār: La solution décisive des doutes. J.P. de Menasce,
ed. and trans. Fribourg, 1945.
Maximus of Tyre. The Philosophical Orations. M.B. Trapp, trans. Oxford, 1997.
Meyendorff, J. “Byzantine Views of Islam,” Dumbarton Oaks Papers 18 (1964): 113–132. 106
Monnot, G. Penseurs musulmans et religions iraniennes: ʿAbd al-Jabbār et ses devanciers.
Paris and Beirut, 1974.
Moses Bar Kepha. Der Hexaemeronkommentar. L. Schlimme, trans. Wiesbaden, 1977.
Nāṣir-i Khusraw, see Kraus.
Origen. Contra Celsum. H. Chadwick, trans. Cambridge, 1953.
Philo. De Decalogo. In F.H. Colson, ed. and trans., Philo. Loeb Classical Library, vol. 7.
Cambridge, Mass., and London, 1950.
Porphyry, see Macarius Magnes.
Al-Qirqisānī, Yaʿqūb. Kitāb al-anwār waʾl-marāqib: Code of Karaite Law. L. Nemoy, ed.
New York, 1939.
Reed, A.Y. Fallen Angels and the History of Judaism and Christianity: the Reception of
Enochic Literature. Cambridge, 2005.
Schofield, M. “Cicero for and against Divination.” Journal of Roman Studies 76 (1988):
47–63.
Schäfer, P. Rivalität zwischen Engeln und Menschen. Berlin and New York, 1975.
Al-Shahrastānī. Al-Milal waʾl-nihal. W. Cureton, ed. London, 1842–1846. D. Gimaret and
G. Monnot, trans., Livre des religions et des sectes. Paris, 1986.
Stroumsa, S. Freethinkers of Medieval Islam. Leiden, 1999.
Al-Tawḥīdī. Al-Imtāʿ waʾl-muʾānasa. A. Amīn and A. al-Zayn, eds. Cairo, 1939–1944.
. Al-Muqābasāt. M.T. Ḥusayn, ed. Baghdad, 1970.
chapter 5
63 The Dahrīs According to al-Jāḥiẓ*
In the third/ninth-century Islamic world we encounter people of whom it is

said that they denied the existence of God, angels, prophets, spirits, the resur-
rection, post-mortem reward and punishment, and the afterlife altogether. In
effect, they rejected the entire metaphysical realm as either false or beyond the
limits of human reasoning, on the understanding that there was no point in try-
ing to know about anything unless it was accessible to human reasoning. It was
this understanding which made them radical even when or if they were willing
to consider the possibility of a reality beyond us: they did not accept revelation
as an alternative source of knowledge. They were empiricists in the sense that
they held all genuine knowledge to be based on sense impressions in conjunc-
tion with reasoning. The sources call them Dahrīs, eternalists, aṣḥāb al-hayūlā,
adherents of prime matter, and aṣḥāb al-ṭabāʾiʿ, adherents of the four “natures”,
i.e. the four elementary qualities (heat, cold, moisture, and dryness) of which
they held the world to be composed. It is not in the third/ninth century alone
that we hear of them: there are intimations that they existed earlier and the
polemics against them continue down to at least the sixth/eleventh century.
But it seems to have been in the third/ninth century that they attracted most
attention.1
The Dahrīs sound so weirdly out of place in the early Islamic world that
modern Islamicists often have trouble believing that they really existed, unless
they have studied them themselves.2 No Dahrī writings survive, most of the
evidence is polemical, and with some minor exceptions no individual Dahrīs
64 are known by | name, so it comes naturally to suspect that all there is to them
* I should like to thank the participants in a graduate seminar on Dahrism I taught at Princeton
University in 2006 for assisting my attempt to understand the texts we read, Everett Rowson
for the generosity with which he shares his expertise, and Michael Cook and Emma Gannagé
for commenting on a draft of this article.
1 See I. Goldziher and A.-M. Goichon (1965), “Dahriyya”, Encyclopaedia of Islam, 2nd edition,
Brill, Leiden, vol. ii, pp. 95a sq.; M. Shaki and D. Gimaret (1993), “Dahrī”, Encyclopaedia Iranica,
http://www.iranicaonline.org; J. Van Ess (1991–1997), Theologie und Gesellschaft im 2. und
3. Jahrhundert Hidschra, eine Geschichte des religiösen Denkens im frühen Islam, 6 vols., De
Gruyter, Berlin/New York (hereafter tg), esp. vol. iv, pp. 451sqq.
2 I have never encountered any doubts about their reality in the literature on them, but
suspicion of polemical invention is a common response to oral presentations of their views.

the dahrīs according to al-jāḥiẓ 97
is heresiographical stereotyping and construction of the “other”. This makes

the testimony of al-Jāḥiẓ (d. 255/869) particularly important. He is one out
of many Muʿtazilites who wrote on the Dahrīs in the third/ninth century, but
the others wrote refutations and heresiographical accounts of the Dahrīs, and
their works are lost except for extracts in later sources.3 Al-Jāḥiẓ, by contrast,
wrote as a littérateur, most of his work is extant, and though his attitude to the
Dahrīs is also polemical, he gives us a vivid picture of them as a live presence.
In what follows I go through the information he provides, restraining myself
from the temptation to adduce material from other ninth-century sources, so
that the reader will have a clean picture of the Dahrīs as perceived by a single,
contemporary author.
Overall Portrait
Most of al-Jāḥiẓ’s references to Dahrīs are found in his book on animals, and
the single most informative passage comes in the last volume of that work.4 It
is long and convoluted, and it starts with a relative clause of which the first part
goes on for so long that it can be read either as incomplete or as completed in a
way suggesting that the author (or copyist) had himself lost his sense of where
he was. I have read it as incomplete and inserted some words that seem to be
missing; the alternative is to remove two that would be superfluous, and the
reader can construe the sentence either way, as I have underlined the words
that introduce the relative cause and those that could be taken to initiate its
3 A Kitāb al-Radd ʿalā ʾl-dahriyya is listed for al-Aṣamm (d. 200/816 or the year after), Bishr
b. al-Muʿtamir (d. 210/825), and al-Naẓzām (d. before 232/847) (Muḥammad b. Isḥāq Ibn al-
Nadīm [1971], Kitāb al-Fihrist, ed. R. Tajaddud, Maktabat al-Asadī, Tehran, pp. 206, 13; 214, 15;
tg, vol. v, p. 285, no. 48). Of these, all we have are the samples of al-Naẓzām’s polemics against
Dahrī cosmology preserved in al-Jāḥiẓ’s animal book (cf. the references below, notes 29sq.).
The polemics against the Dahrīs by Muḥammad b. Shabīb (d. 230/840), presumably from his
Kitāb al-Tawḥīd, survive in Muḥammad b. Muḥammad al-Māturīdī (d. 833/944) (1970), Kitāb
al-Tawḥīd, ed. F. Kholeif, Dar al-Mashreq, Beirut, pp. 141 sqq.; cf. tg, vol. iv, pp. 124sqq. on Ibn
Shabīb. The section on the Dahrīs from the heresiography of the third/ninth-century Abū ʿĪsā
al-Warrāq survives in Maḥmūd b. Muḥammad Ibn al-Malāḥimī (d. 536/1141) (1990), Kitāb al-
Muʿtamad fī uṣūl al-dīn, ed. M. McDermott and W. Madelung, al-Hoda, London, pp. 548sqq.;
cf. S.M. Stern (1960), “Abū ʿĪsā al-Warrāq”, ei2, vol. i, p. 130.
4 Al-Jāḥiẓ (1938), Kitāb al-Ḥayawān, ed. ʿA.S.M. Hārūn, 7 vols., Maktabat Muṣṭafā al-Bābī al-
Ḥalabī, Cairo, vol. vii, pp. 12 sqq.; also discussed in H. Daiber (1999), “Rebellion gegen Gott.
Formen atheistischen Denkens im frühen Islam”, in F. Niewöhner and O. Pluta (eds.), Atheis-
mus im Mittelalter und in der Renaissance, Harrassowitz, Wiesbaden, pp. 23–44, pp. 24sq.
98 chapter 5
completion. Al-Jāḥiẓ has just said that nobody who prays towards the qibla will
65 disagree with what he has said, and that this holds | true even of the mulḥids
who believe in the resurrection and revealed religion/law (al-sharāʾiʿ), so that
the only one who will disagree is the Dahrī:
for the one who denies divinity (al-rubūbiyya), makes the command and
prohibition something absurd, rejects the very possibility of the prophecy
( jawāz al-risāla), holds matter (al-ṭīna) to be eternal, flatly denies ( yajḥa-
du) reward and punishment, does not recognize the prohibited and the
permitted, does not acknowledge that there is any proof in the entire
world of a maker and things made or a creator and things created, and
who considers the heavenly sphere—which does not know itself from
others, which cannot distinguish between that which appears in time
and the eternal, or between the doer of good and of evil, which cannot
increase its movement or decrease its circular motion, and which can-
not follow movement with rest, stand still for one moment, or deviate
from its direction—to be the one5 through which everything is held firm
and destroyed, and which accounts for all things fine or great, includ-
ing these marvelous, wise arrangements, perfect forms of governance,
the wonderful composition and wise construction in accordance with a
known computation and familiar order exhibiting the subtlest ways of
wisdom and perfect workmanship [such a Dahrī cannot accept what we
say], but such a Dahrī has no right to object to our book, even if it goes
against his views and calls to the opposite of what he believes. For the
Dahrī does not think there is any revealed religion (dīn) or creed (niḥla)
or religious law (sharīʿa) or religious system (milla) on earth. He does not
think the permitted has any sanctity (ḥurma) or know what it is, nor does
he think that the forbidden has any limit or know what it is. He does not
expect any punishment for evil-doing, nor does he hope for any reward for
doing good. What is right in his view and true in his judgment is that he
and undiscriminating quadrupeds (al-bahīma) are the same and that he
and predatory animals (al-sabuʿ) are the same. Moral wrong (al-qabīḥ) in
his opinion is simply that which goes against his inclination, moral good
(al-ḥasan) is merely what conforms with his inclination: things turn on
(madār al-amr) failure and success, pleasure and pain, and what is right
lies simply in that which confers benefit, even killing a thousand upright
5 Daiber’s translation is clearly wrong here (compare the editor’s helpful gloss at K. al-Ḥayawān,
ed. Hārūn, vol. vii, p. 13, n. 2).
men for the sake of a bad dirham. This Dahrī does not fear that he will be
punished and chastised, temporarily or for ever, if he stops criticizing the
scriptures6 or the imams, nor does he hope for any reward in this world
or the next if he finds fault with them and displays hostility to them.
The Dahrī is here depicted as an atheist in the sense of someone who denies
the existence of a God outside nature, or God in any sense at all. In so far as
the Dahrī operates with anything that could be called a deity, it is the celestial
sphere, which he sees as ruling the universe, and which he may have credited
with intelligence, though al-Jāḥiẓ does not say so; he even seems to reject that
possibility by having the Dahrī deny divinity (al-rubūbiyya) outright. To the
Dahrī, the cosmos is ruled by itself, not | regulated by a being outside it, and 66
it has not been created by such a being either: there is not in his view any
evidence for creation anywhere in the universe. Matter has always existed, and
by implication always will, though it is only the first of these points that al-Jāḥiẓ
singles out for attention. Since the Dahrī does not believe in a personal God,
he also denies that there can be any such thing as a divine message, meaning
one carried by a prophet, and accordingly he also rejects the possibility of
“command and the prohibition”, here as elsewhere in al-Jāḥiẓ meaning divine
law.7 What God has forbidden and allowed means nothing to the Dahrī: it has
no inviolability and sets no limits in his view. Since there is no God, there is not
any religion on earth either in his view, or in other words, he does not think that
any of the many religions found on earth is true; and since it is only from the
revelation that we know about rewards and punishments after death, he does
not believe in them either. He is described as an outright denier, not a sceptic or
agnostic. Elsewhere, al-Jāḥiẓ cites his teacher al-Naẓẓām as observing that he
had engaged in disputation with two kinds of mulḥids, the denier (al-jāḥid) and
the doubter (al-shākk), and that he had found the latter to be better at kalām
than the former.8 But the term Dahrī is not used there, and al-Jāḥiẓ himself
always seems to think of a Dahrī as a jāḥid.
What the Dahrī does believe, in al-Jāḥiẓ’s presentation, seems to be that the
combination of eternal matter and the motion of the celestial sphere suffices
to explain everything in the world around us. Al-Jāḥiẓ highlights the absurdity
of this belief by recourse to an old argument against the divinity of the planets
6 Al-kutub, clearly not al-Jāḥiz’s own animal book, as Daiber says (“Rebellion”, p. 25).
7 See for example al-Jāḥiẓ, “Al-Maʿāsh waʾl-maʿād”, in ʿA.-S.M. Hārūn (ed.) (1965–1979), Rasāʾil
al-Jāḥiẓ, 4 vols., Maktabat al-Khānjī, Cairo, vol. i, pp. 100, 1; 104, 2; id., “Maqālat al-Zaydiyya
waʾl-Rāfiḍa”, in ibid., vol. iv, p. 320, 1.
8 Al-Jāḥiẓ, K. al-Ḥayawān, ed. Hārūn, vol. vi, p. 35.
100 chapter 5
and stars. The very regularity of the motion of the heavenly bodies which had
constituted proof of their divinity to the Greeks proved to the Christians that
they were ruled by a higher power, and this was how al-Jāḥiẓ saw it, too: how
could the heavenly bodies, which did not have the ability to vary their own
movements, be the regulators of everything?9 That the Dahrī should claim to
find no evidence for a creator or maker anywhere in the world also strikes al-
Jāḥiẓ as absurd in view of the wonders of nature and the exquisitely intricate
67 ways of things, clearly meant as a reference | to the wonderful things he has
described in his animal book. His response to the Dahrī, in other words, is
recourse to the argument from design. He is envisaged as having developed this
argument at greater length in a book against deniers of God and providence
which is falsely attributed to him.10
Though the Dahrī rejects divine law, he operates with a concept of morality
and distinguishes between good and bad, al-ḥasan waʾl-qabīḥ, literally the
beautiful and the ugly (or the nice and the nasty), the standard terms for
good and bad as perceived by the human intellect, as opposed to al-ḥalāl
waʾl-ḥarām, the forbidden and the allowed, or in other words the good and
the bad as defined by divine legislation. The Dahrī thinks that humans are
capable of defining good and bad themselves, with reference to concepts such
as benefit or utility: to al-Jāḥiẓ, this boils down to setting moral standards to
suit your own convenience. He takes it for granted that the Dahrī will set the
standard with exclusive reference to his own personal advantage, so that he
could in principle approve of killing a thousand good people for a bad coin.
The possibility that the Dahrī thought of right and wrong in terms of collective
welfare is not considered. Like so many believers, al-Jāḥiẓ cannot help thinking
that an atheist must be a deeply immoral and selfish person: his moral rules
are not set by an external authority higher and wiser than himself; and he does
not expect to be either rewarded or punished for anything he does after death,
9 This argument had been disseminated in Iraq by Christians, cf. J.T. Walker (2004), “Against
the Eternity of the Stars: Disputation and Christian Philosophy in Late Sasanian Mesopo-
tamia”, in G. Gnoli and A. Panaino (eds.), La Persia e Bisanzio (Atti dei Convegni Lincei
2001), Accademia Nazionale dei Lincei, Rome, pp. 518–535, where the Christian ʿAbdishoʿ
uses it against the Zoroastrian Qardagh (who converts); id. (2006), The Legend of Mar
Qardagh: Narrative and Christian Heroism in Late Antique Iraq (Transformation of the
Classical Heritage, 40), University of California Press, Berkeley/Los Angeles/London,
pp. 29; 190 sqq.
10 Cf. al-Jāḥiẓ (attrib.) (1928), Kitāb al-Dalāʾil waʾl-iʿtibār ʿalā ʾl-khalq waʾl-tadbīr, ed. M.R. al-
Tabbākh, al-Maṭbaʿa al-ʿIlmiyya, Aleppo; M.A.S. Abdel Haleem (tr.) (1995), Chance or
Creation? God’s Design in the Universe, Garnet, Reading.
so what motives could he possibly have for behaving unselfishly? The Dahrīs
familiar to al-Jāḥiẓ seem to have argued that humans have it in them to manage
their lives, including the determination of right and wrong, on the basis of their
innate intelligence much as animals do; and al-Jāḥiẓ is on shaky grounds here,
for his book is full of praise for the wonderful governance that one can see
in nature, and he sometimes adduces animals as examples of the way things
work in human societies too. If the wonderful general governance of the world
suffices to make animals flourish, why must humans have prophets, revealed
law, or beliefs in Paradise and Hell in addition? It was a good question, later
taken up by Abū Bakr al-Rāzī as an argument against the idea of prophecy.11 The
Sincere Brethren, too, adduced the animals in illustration of natural as opposed
to prophetic religion.12 But al-Jāḥiẓ wriggles out of the question by simply
appealing to human self-esteem: the Dahrī downgrades us to undiscriminating
quadrupeds (al-bahīma) | and predatory animals (al-sabuʿ), he says. In his 68
epistle on the cultivation of virtue he credits animals with the same self-seeking
drives as human beings and casts the divine law as the antidote in the human
case, again without telling the reader why animals could manage without it, or
even whether they could:13 there were people in his time, in fact pupils of his
own teacher al-Naẓẓām, who held that animals did have prophets and religious
laws just as humans did,14 an idea that al-Jāḥiẓ derided.15 This makes his own
refusal to explain the difference all the more surprising.
The last point that al-Jāḥiẓ makes in this passage is that the Dahrī is given
to criticizing the scriptures and finding fault with both them and the imams,
presumably including the prophets. This is a sign of the Dahrī’s perversity, for
he does not expect to gain any reward for it in the next world, nor does he think
that he would be punished for it after his death if he stopped.
11 Cf. Abū Ḥātim al-Rāzī (1977), Aʿlām al-nubuwwa, ed. Ṣ. al-Ṣāwī, Muʾassasa-yi Pizhūhishī-i
Ḥikmat wa Falsafa-yi Īrān, Tehran, p. 183, 2 sq. (and, implicitly, 3, 11; 181, 7; 274, 2).
12 Rasāʾil Ikhwān al-Ṣafā wa-khullān al-wafāʾ, 4 vols., Dār Bayrūt, Beirut 1957, vol. ii, pp. 203–
377, esp. pp. 324–329; L.E. Goodman (tr.) (1978), The Case of the Animals versus Man before
the King of the Jinn: A Tenth-Century Ecological Fable of the Pure Brethren of Basra, Twayne,
Boston, esp. pp. 156–165.
13 Al-Jāḥiẓ, “Al-Maʿāsh waʿl-maʿād”, Rasāʾil, ed. Hārūn, vol. i, p. 102, 12sq.
14 The best known is Aḥmad b. Khābiṭ/Ḥāʾiṭ, cf. Muṭahhar b. Ṭāhir al-Maqdisī (1899–1919),
Kitāb al-Badʾ waʾl-taʾrīkh, ed. C. Huart, 6 vols., Ernest Leroux, Paris, vol. iii, pp. 8sq.; ʿAlī b.
Aḥmad Ibn Ḥazm (1317–1321h.), Kitāb al-Faṣl fī ʾl-milal waʾl-ahwāʾ waʾl-niḥal, 5 vols., Cairo,
vol. i, pp. 78 sqq.; tg, vol. iii, pp. 430 sqq.
15 Al-Jāḥiẓ, K. al-Ḥayawān, ed. Hārūn, vol. v, p. 424; tr. tg, vol. vi, p. 214, on Aḥmad b. Khābiṭ
and prophets to the bees.
102 chapter 5
Rule-Bound Universe versus Divine Intervention
Elsewhere in his animal book al-Jāḥiẓ gives us concrete examples of Dahrī

criticism of the scriptures. In one passage it is merely implicit. He tells us that
some Dahrīs flatly denied the existence of metamorphosis (maskh), the process
whereby God had turned some humans into monkeys and pigs (q. 5:60; cf. 2:65;
7:166); other Dahrīs accepted its existence, but explained it as the outcome of
environmental damage and its effect on the elementary qualities for which no
supernatural intervention was required.16 In another passage the Dahrīs attack
the Qurʾān directly, finding fault with the story of Solomon and the Queen
of Sheba. They object that on the one hand, this story presented Solomon as
having asked for, and apparently been granted, kingship of a kind never granted
to anyone else (q. 38:35), that is to say power over not just humans, but also
spirits ( jinn) and the winds, and knowledge of the language of the birds. But
on the other hand, the story claims that Solomon needed the hudhud bird to
tell him about the existence of the Queen of Sheba, though she was hardly all
that far away. Al-Jāḥiẓ quotes the Dahrīs as saying,
69 “Our kings today, who have less power than Solomon, are not unaware
of the rulers of the Khazars, the Rūm, the Turks, or the Nubians, so how
could Solomon be unaware of this queen, when their lands were so close
and also contiguous, without any seas or rugged land in the way?” This
and the like, they said, “is evidence of the corrupt nature of your historical
tradition” (dalīl ʿalā fasād akhbārikum).17
Al-Jāḥiẓ replies by granting that if it were the case that God abstained from
intervention in the world and governance of its affairs, letting them run in
their normal way, then the Dahrīs would be right; but sometimes God diverted
people’s minds (ṣarafa awhāmahum), so that for example Jacob and Joseph
did not recognize each other in Egypt even though both were prophets, and
so that the Israelites wandered in the desert for forty years without finding
their way to their destination, which would not normally have been so difficult
either. His argument is that “they were diverted from the chance to learn the
truth by divine providence”, because it was not yet right for them to learn it, as
Lactantius had put it some six hundred years earlier.18 Al-Jāḥiẓ adduces several
16 Al-Jāḥiẓ, K. al-Ḥayawān, ed. Hārūn, vol. iv, pp. 70sqq. (with the ṭabāʾiʿ at p. 73, 4); cf.
M. Cook (1999), “Ibn Qutayba and the Monkeys”, Studia Islamica 89, pp. 43–74 (p. 60).
17 Al-Jāḥiẓ, K. al-Ḥayawān, ed. Hārūn, vol. iv, pp. 85 sq.
18 Ibid., vol. iv, pp. 86 sqq; Lactantius, Divine Institutes, iv, 2:5, tr. by A. Bowen and P. Garnsey,
Liverpool Univ. Press, Liverpool, 2003, p. 227; cf. tg, vol. iii, pp. 411sq.
other examples of ṣarfa, including the jinn who keep trying to eavesdrop on
conversions in heaven, apparently never learning better; but he seems to be
aware that this is an argument that only believers would accept, for he adds that
the Dahrī cannot expect the same (sort of reasoning) “from people who accept
worship and messengers as from the pure Dahrī (al-dahrī al-ṣirf ), who does not
acknowledge anything other than what he sees himself (mā awjadahu ʾl-ʿiyān)
and that which works in the same way as seeing things for oneself (mā yajrī
majrā ʾl-ʿiyān)”.19 Here as in the reference to the corrupt nature of the historical
tradition we are being told something about Dahrī epistemology: a Dahrī is
someone who deems information transmitted from others to be unacceptable
if it does not conform to reason and to whom evidence consists in what he
sees for himself and what is of the same nature as that (which is not further
explained). Al-Jāḥiẓ continues:
The Dahrī knows [that we believe]20 we have a lord who has brought the
bodies into existence (ikhtaraʿa al-ajsām) and that He is alive, but not
through life, knowing, but not through knowledge, that He is a thing, but
cannot be divided, that He has no length, breadth, or depth, and that the
prophets can revive the dead, all of which the Dahrī holds to be impossible
(mustankar).21
Once again, the Dahrī, or rather the “pure Dahrī”, is identified as somebody who 70
does not believe in God, not even God as defined by the Muʿtazilites. So he holds
the bodies of which the world is made up to exist on their own and denies that
prophets (or anyone, presumably) can bring people back to life.
Al-Jāḥiẓ makes several of the same points in another discussion of the jinn
who try to eavesdrop on discussions in heaven.22 The Qurʾān says that the jinn
in question had balls of fire thrown at them when they tried to do so (q. 72:8sq.;
cf. 15:17sq.; 37:7sq.). “Some people”, later identified as Dahrīs, claimed that it was
absurd to suppose that creatures endowed with superior intelligence should go
on trying: they would have learnt from the Qurʾān (which they had heard) that
God always does as He threatens; and that apart, they would have learnt from
their long experience (ṭūl al-tajriba), from plain seeing for themselves (al-ʿiyān
19 Al-Jāḥiẓ, K. al-Ḥayawān, ed. Hārūn, vol. iv, pp. 89 sq.

20 Inserted by the editor, though it seems superfluous.
21 Al-Jāḥiẓ, K. al-Ḥayawān, ed. Hārūn, vol. iv, p. 90.
22 Ibid., vol. vi, pp. 265 sqq.; partial tr. in C. Pellat (1967), The Life and Works of Jāḥiẓ, transla-
tions of selected texts, tr. from the French by D.M. Hawke, Routledge, London, pp. 176sqq.
104 chapter 5
al-ẓāhir), and from some telling each other about it (ikhbār baʿiḍihim li-baʿḍ).23
That the jinn should have learnt from the Qurʾān is an argument based on the
opponents’ premises. The rest tells us what counted as legitimate sources of
knowledge to the Dahrī: experience, seeing for oneself, and information from
others (empeiria, autopsia, and historia in the terminology of Greek empiricist
doctors).24
Al-Jāḥiẓ once more seems to accept that the Dahrīs are right in terms of
the normal rules of things, for he responds by invoking the ṣarfa theory again,
once more adducing the Israelites in the desert and other examples, including
Solomon and the Queen of Sheba, and explaining that God diverts the minds of
people so as to expose them to trials (al-miḥna), for it is when people are tested
that obedience and disobedience become manifest. Once more he is aware that
his explanation will not be acceptable to Dahrīs, for he mentions that there
are other examples “which go against the Dahrī method (mimmā yukhālafu
fīhi ṭarīq al-dahriyya)”, and explains that “the Dahrī does not acknowledge any-
thing other than sense impressions and regularities (al-maḥsūsāt waʾl-ʿādāt),
in contrast with this doctrine (of ṣarfa)”.25 Again, the Dahrīs are identified as
empiricists. Earlier we were told that they only believed in what they saw for
themselves or what was of the same nature, or in experience, seeing for them-
selves, and information from others (when it accorded with reason); here, the
71 basis on which one accumulates experience and acquires the | ability to reason
about it is implicitly defined as sense impressions, including the observation of
regularities.
Al-Jāḥiẓ adds that the Dahrī cannot use the ṣarfa argument
as long as he persists in not believing in monotheism (al-tawḥīd) and con-

tinues not to recognise anything but the heavenly sphere (al-falak) and its
doings, and continues to hold the dispatch of messengers (irsāl al-rusul)
to be impossible, and to believe that the command and prohibition, as
also the reward and punishment, are other than what we say, and that
God cannot order by way of testing things, only by way of uniform/irre-
versible decree (lā yajūzu an yaʾmara min jihat al-ikhtibār illā min jihat
al-ḥazm/jazm).26
23 Al-Jāḥiẓ, K. al-Ḥayawān, ed. Hārūn, vol. vi, pp. 4 sq.

24 Cf. R.J. Hankinson (1995), The Sceptics, Routledge, London, ch. 13, esp. pp. 227sq.
25 Al-Jāḥiẓ, K. al-Ḥayawān, ed. Hārūn, vol. vi, p. 269.
26 Ibid., vol. vi, pp. 269 sq. The text has al-ḥazm; al-jazm was suggested to me by Joseph
Witztum.
In other words, the regularities one observed did not admit of exceptions:
God could not break His own laws in the Dahrī view. Once again the Dahrī
replaces God with the celestial sphere, rejecting the reality of prophets, divine
law and otherworldly retribution, but here he actually speaks of God, possibly
because al-Jāḥiẓ has made him do so, but more probably because he would do
so in actual fact, if only for purposes of the argument. The Dahrīs also argue
on the basis of their opponents’ premises in the discussion of the jinn, where
they refer to the jinn as creatures endowed with superior intelligence even
though they did not believe in jinn themselves, or for that matter in devils,
angels, veridical dreams (al-ruʾyā), or charms, as al-Jāḥiẓ tells us elsewhere.27
But the Dahrī was not a monotheist to al-Jāḥiẓ: he did not believe in al-tawḥīd.
Elsewhere, al-Jāḥiẓ casually refers to “the difference between the madhāhib
al-dahriyya and madhāhib al-muwaḥḥidīn (the doctrines of the Dahrīs and
the doctrines of the monotheists)”.28 A Dahrī failed to count as a monotheist
because he had no God, not because he had many: a muwaḥḥid is here the
opposite of an atheist, not of a mushrik.
On the question of Dahrī cosmology al-Jāḥiẓ says more in a passage in which
he is quoting from his teacher al-Naẓzām. According to the latter, Dahrīs did
not all have the same beliefs. “Some of them say that this world of ours is
made of four principles (lit. pillars, arkān), heat, cold, dryness, and moisture,
and that other things are outcomes, combination, and generation (natāʾij wa-
tarkīb wa-tawlīd)”. Others also claimed that the world is made of four principles,
but identified them as “earth, air, water, and fire”, i.e. the elements rather
than the elementary qualities. The first group cast the elementary qualities as
bodies (ajsāman), the latter cast the elements | as substances ( jawāhir) and the 72
elementary qualities as accidents.29 They, apparently all of them, gave priority
to the sense of touch (by which the four elements and elementary qualities
could be perceived) and held all smells, colours, and sounds to be composed
of those four. Al-Jāḥiẓ devotes many pages to his teacher’s refutations of their
physics, but we may leave them aside here, except for his observation that some
people held there to be a fifth pillar in the form of spirit (rūḥ).30
27 Ibid., vol. ii, p. 139.

28 Ibid., vol. i, p. 217, 8 sq.
29 Ibid., vol. v, p. 40.
30 Ibid., vol. v, p. 47, 5.
106 chapter 5
Other Issues
So far, we have only considered al-Jāḥiẓ’s book on animals, but he discusses

Dahrīs in other works as well. In his ʿUthmāniyya he has a passage in which
he accuses the Shīʿites of making the imamate unduly complicated: they made
it more difficult than problems such as assessing the probity of transmitters
(taʿdīl and tajwīr), distinguishing between things done on the basis of innate
nature and those stemming from free choice (al-faṣl bayna ʾl-ṭibāʿ waʾl-ikhtiyār),
the arguments for and against anthropomorphism (al-tashbīh), and the relative
status of information handed down by the tradition versus rational arguments
(majīʾ al-akhbār wa-ḥujaj al-ʿuqūl). He held this to be excessive because
we have never seen anyone turning godless (alḥada) or dualist (tazan-

daqa) because of an error in the doctrine of the imamate or disagreement
over it, whereas those who have apostatised as dualists and Dahrīs (man
irtadda zindīqan aw dahriyyan) over these questions are uncountable.31
Two of the questions over which people would turn dualist or Dahrī are theo-
logical (anthropomorphism, the determination of our acts), and two are epis-
temological (the reliability of transmitters and the value of transmitted infor-
mation versus rational arguments). To start with the former, the Dahrīs must
have been among those who stripped God of His anthropomorphic features, in
their case by reducing Him to mere nature in the form of the heavenly sphere or
to nothing at all; and they must have held our acts to be determined by nature.
On this second point al-Jāḥiẓ offers some corroborating evidence. He tells us
that al-Naẓẓām had a brother-in-law called Abū ʾl-ʿAbbās who “believed in the
stars (kāna yadīnu biʾl-nujūm) and did not believe anything to happen except in
accordance with nature (ṭibāʿ)”.32 One takes it that he was a Dahrī, though al-
73 Jāḥiẓ politely avoids branding a member of his teacher’s family as | such; and the
ṭibāʿ which Abū ʾl-ʿAbbās saw as determining events was presumably the partic-
ular mixture of the four elementary qualities in things, including ourselves, in
conjunction with the rotation of the heavenly sphere. Elsewhere al-Jāḥiẓ men-
tions the importance of distinguishing between the science of the natures (ʿilm
al-ṭabāʾiʿ) and free will (al-ikhtiyār), implying that if one did not, belief in the
four natures would lead to determinism.33
31 Al-Jāḥiẓ (1955), Kitāb al-ʿUthmāniyya, ed. ʿA.-S.M. Hārūn, Dār al-Kitāb al-ʿArabī, Cairo,
pp. 270 sq.
32 Al-Jāḥiẓ, K. al-Ḥayawān, ed. Hārūn, vol. i, p. 148, 6 sqq.
33 Ibid., vol. i, p. 218, 5.
As regards the reliability of transmitters and the relative role of tradition

(akhbār) and rational arguments (ḥujaj al-ʿuqūl), we have seen that Dahrīs
took the implausible features of the story of Solomon and Sheba as evidence
of the corrupt nature of the historical tradition ( fasād akhbārikum). In line
with this, they seem to have impugned the veracity of the Muslim transmitters,
possibly taking up the science of taʿdīl and tajwīr in order to demonstrate its
uselessness. They themselves clearly judged reports from others on the basis of
their contents, rejecting them if they failed to conform to reason. They operated
on the tacit assumption that divine intelligence would not be radically different
from that of humans, so that information coming from a divine source would
not clash with our own sense of true and false, or right and wrong, and they
picked out what to them were absurd stories in the Qurʾān in order to prove that
the book could not be of divine origin. To traditionalist believers, by contrast,
there was no guarantee that information coming from a divine intelligence
vastly superior to ours would conform to our limited reasoning. Since humans
had no access to the metaphysical realm apart from the revelation that God
had made available to them, they could not evaluate the authenticity of the
information coming from that realm (in the form of Qurʾān and Ḥadīth) on the
basis of its contents, only on the basis of the chains of transmission through
which it had been passed down: if the material had been properly transmitted,
the authoritative nature of the contents was guaranteed. Al-Jāḥiẓ occupies a
position midway between the two, for as a Muʿtazilite he shared the Dahrī
view that there was continuity between divine and human reason, without
wanting to jettison either the Qurʾān or the tradition whenever they seemed
to say something unreasonable: this is why he likes the ṣarfa doctrine, which
offers a rational explanation for seemingly irrational claims in the Qurʾān.
Dahrism on the Ground 74
Al-Jāḥiẓ clearly thinks of Dahrīs and Zindīqs as closely related: it was as one or
the other that people had been brought to apostasy by the difficult questions,
and he links the two elsewhere as well.34 The fact that he thinks of them as
apostates shows they are people within the Muslim community who have come
to subscribe to unacceptable ideas, not unbelievers from outside it. This is also
34 Cf. ibid., vol. iv, pp. 432–434, where he tells the Christians that they are neither Zindīqs
nor Dahrīs, Muslims nor Jews; “Ḥujaj al-nubuwwa”, in Rasāʾil, ed. Hārūn, vol. iii, p. 281,
where he notes that no hypocrite, Zindīq, or Dahrī can relate that Muḥammad ever held
back or fled from a battle; and the references given below, nn. 37, 45, 48.
108 chapter 5
clear from the Dahrī familiarity with the Qurʾān, and from the general manner
in which al-Jāḥiẓ refers to them: it is within his own community that they
are dissenters. He intimates that they were numerous: uncountable numbers
have apostatised as Zindīqs and Dahrīs, as he says. But he is clearly speaking
hyperbolically. In his Tarbīʿ he tells us that they had never constituted a polity.
“How come that we have never seen a nation of Dahrīs when we know that
it is not possible for a Dahrī to claim prophethood?” “How come that no king
has ever become a Dahrī?”35 This seems to be meant as a teasing question: the
obvious explanation is precisely that no Dahrī could claim prophethood, for
nations were assumed to be formed on the basis of revealed laws; this is also
why kings had no use for Dahrī doctrine. But if his reader had replied along
these lines, al-Jāḥiẓ would probably have come up with a counter-example.
In the next passage, in which he lets the reader try the obvious answer to the
question why there had never been a nation of Manichaeans, namely that they
did not allow fighting, he responds by adducing the Byzantines, whose religion
did not endorse fighting either. Here he might have replied that actually there
had been a nation of Dahrīs, for elsewhere he tells us that the ancient Greeks
had been Dahrīs.36 In any case, the crucial observation is the one that follows:
“How come that we only find the doctrine of the Dahriyya among individuals,
people here and there, and the occasional man ( fī ʾl-khāṣṣ waʾl-shādh waʾl-rajul
al-nādir)?” In other words, we should not envisage the Dahrīs as a sect or a
school. Dahrism was an individual opinion, no doubt more commonly found
in some circles than in others, like atheism today, but not a doctrine that could
serve as the basis of community life. In so far as the Dahrīs had any collective
existence, it will not have been as Dahrīs, but rather as devotees of sciences and
75 professions in which their opinions were widely encountered. It is | clear from
what al-Jāḥiẓ has told us that the Dahrīs he knew were aṣḥāb al-ṭabāʾiʿ, people
concerned with the four elementary qualities. Three sciences in particular are
known to have been associated with the four elementary qualities, namely
astrology, medicine, and alchemy. It seems to have been particularly in circles
engaged in the study of the first two that Dahrism was common in al-Jāḥiẓ’s
time. At least we do not hear of any alchemists among them.
Al-Jāḥiẓ has already told us that the Dahrīs assigned God’s role as governor
of the universe to the heavenly sphere, suggesting that they were astrologers.
In his refutation of Christianity he adds further evidence that they were often
35 Al-Jāḥiẓ (1955), Kitāb al-Tarbīʿ waʾl-tadwīr, ed. C. Pellat, Damascus, §137 (drawn to my
attention by Kevin van Bladel). Pellat suggests emending the passage to say that a Dahrī
could claim prophethood, but this makes no sense.
36 Cf. below, n. 54.
envisaged as having studied astronomy/astrology and medicine, and also that

they were associated with philosophy. He claims that the Jews did not approve
of philosophical enquiry (al-naẓar fī ʾl-falsafa) or rationalising theology (al-
kalām fī ʾl-dīn) on the grounds that such enquiries engendered all kinds of
specious questions (shubha), that the only true knowledge is in the Torah and
the books of the prophets, and that belief in medicine and the astrologers was
among the things that caused people to become Zindīqs and Dahrīs (wa-anna
ʾl-īmān biʾl-ṭibb wa-taṣdīq al-munajjimīn min asbāb al-zandaqa waʾl-khurūj ilā
ʾl-dahriyya).37 Medicine and astrology, then, had the same effect as difficult
theological and epistemological questions.
It seems to be medicine and astrology that count as al-naẓar fī ʾl-falsafa as
opposed to al-kalām fī ʾl-dīn in this passage. Elsewhere al-Jāḥiẓ speaks of kalām
al-falsafa as opposed to kalām al-dīn,38 here as there contrasting philosophy
as pursued by mutakallims (dialectitians) with theology as practised by them.
Kalām was dialectics, a method of debating by questions and answers which
was practised in disputations. Like so many other things in the Near East, it
had a long history in the region, but all we need to note here is that disputation
was the prime vehicle of intellectual enquiry in the Near East in al-Jāḥiẓ’s time,
both within religious communities and between them, and among friends no
less than opponents. Eventually, kalām came to mean theological enquiry as
pursued in books written in the style inherited from disputations, but this is
not what it meant to al-Jāḥiẓ. To him, theological enquiry was just one branch
of kalām; the other was philosophy, and it is clear from another passage in
his book on animals that philosophy in this context meant enquiry into the
physical universe, above all the elements or the elementary qualities of which
it was widely assumed to be composed. A good mutakallim had to master the
whole field, he said; he would not count as a leader
unless he becomes equally good at the kalām of religion and the kalām 76
of philosophy. The (true) scholar, in our opinion, is the one who com-
bines the two, and the person who has got things right is the one who
harmonises verification of monotheism (taḥqīq al-tawḥīd) with recogni-
tion of the essential characters (ḥaqāʾiq) of the actions of the natures/ele-
mentary qualities (al-ṭabāʾiʿ). He who claims that there can be no true
monotheism without rejection of the essential characters of the elemen-
37 Al-Jāḥiẓ, “Al-Radd ʿalā ʾl-naṣārā”, in Rasāʾil, ed. Hārūn, vol. iii, p. 314 (drawn to my attention
by Krisztina Szilagyi).
38 Cf. below, n. 39.
110 chapter 5
tary qualities (anna ʾl-tawḥīd lā yaṣiḥḥu illā bi-ibṭāl ḥaqāʾiq al-ṭabāʾiʿ) has
carried over into monotheism his own weakness at kalām; and likewise,
if he claims that there can be no true elementary qualities when they
are linked with monotheism (anna ʾl-ṭabāʾiʿ lā taṣiḥḥu idhā qarantahā biʾl-
tawḥīd): whoever says [that] has carried his own weakness at kalām into
the elementary qualities. The godless person (mulḥid) will only despair
of you when your respect for monotheism does not cause you to belittle
the truth about the elementary qualities … By my life, there is some dif-
ficulty (shidda) in their combination. I implore God that I will not tear
down a pillar from my own doctrine every time my spear touches a gate
of kalām that is difficult of entry! There is no benefit in anybody who is
like that.39
There are two fields of kalām and there is tension between them. Some claim
that one cannot be a true monotheist without rejecting everything said about
the elementary qualities (ṭabāʾiʿ), evidently because the elementary qualities
are what mulḥids will discuss; and conversely there are people, whom one takes
to be the mulḥids, who say that one cannot be a good mutakallim without
rejecting monotheism, i.e. belief in God. Al-Jāḥiẓ thinks that this is a mistake.
In his view it is only by getting into the field and mastering it that one can
make the mulḥid despair, undoubtedly because the mulḥid does not want the
monotheists to colonise his science and take it over, hitching it to their world
view at the cost of his own. Al-Jāḥiẓ is aware that taking over the field is
a dangerous enterprise: he prays that he will not tear down any of his own
doctrines whenever he comes to a difficult subject of kalām. But what he wants
to do is precisely to make natural science compatible with monotheism and
to expropriate it for the believers. The mulḥid whom al-Jāḥiẓ sees himself as
confronting is presumably a Dahrī and/or Zindīq.
Here then we have a first-hand admission that getting into the science of
the four natures was difficult for a believer, coupled with an assurance that it
could be done. It is not surprising if some Jews held that it was best to stay away
from medicine and astronomy, or that the jurist Abū Yūsuf held, or was reputed
to have said, that “he who seeks the religion by means of kalām has become a
Zindīq (man ṭalaba ʾl-dīn biʾl-kalām tazandaqa)”.40
39 Al-Jāḥiẓ, K. al-Ḥayawān, ed. Hārūn, vol. ii, pp. 134 sqq.

40 Cited in ʿAbd Allāh b. Muslim Ibn Qutayba (1966), Taʾwīl mukhtalif al-ḥadīth, ed. M.Z. al-
Najjār, Maktabat al-Kulliyyāt al-Azhariyya, Cairo, p. 61, 6; cf. also Kull al-kalām siwā ʾl-
Qurʾān zandaqa, attributed to an anonymous scholar from Shāsh in Abū Bakr Aḥmad b.
Among the tricky questions that al-Jāḥiẓ knew Dahrīs to ask was that “of 77
the anvil and the hammer, and the egg and the chicken”, both clearly designed
to prove that the world must always have existed: you cannot have an anvil
without a hammer, but you cannot make the hammer without an anvil; you
cannot have an egg without a chicken or a chicken without an egg.41 It was such
questions that caused people to tear down the pillars of their own doctrine
when they came to difficult gates in kalām. But al-Jāḥiẓ practised what he
preached: among his lost books there is one on the actions of the elementary
qualities (afʿāl al-ṭabāʾiʿ).42
He was not the only one to be keen on science. All the Muʿtazilites were
busy getting into physics at the time, writing books on atoms, bodies, natures,
and more besides. They were appropriating the entire domain of ancient sci-
ence as it had been transmitted to them in Iraq. At the same time, they were
busy writing refutations of mulḥids, Zindīqs, and the Dahriyya.43 It is precisely
because the Muʿtazilites were the scientific pioneers of the Muslims that they
were the ones to confront the Dahrīs and so our chief sources of informa-
tion about them. It is for the same reason that Muʿtazilites had a constant
tendency to go off the rails, tearing down one pillar after another of their
good monotheist beliefs as they got into the dangerous domain in which the
Dahrīs specialised. Abū Saʿīd al-Ḥaḍrī, al-Ḥaddād, Abū ʿĪsā al-Warrāq, and Ibn
al-Rāwandī are the best known examples.44 But all that takes us away from al-
Jāḥiẓ.
Al-Jāḥiẓ tells us more about the Dahrīs in his book in defence of prophet-
hood, devoted to the criteria by which genuine reports from the past (akhbār)
can be distinguished from false ones. Here he admits that many people can
agree on an error: the Jews, Christians, Zoroastrians, Zindīqs, Dahrīs, and Bud-
dhists (aṣḥāb al-bidada) all deny that the Prophet had wrought miracles and
brought a revelation.45 He insists that he is not writing his book because the
criticisms of the godless (ṭaʿn al-mulḥidīn) were having any effect whatever
ʿAlī al-Khaṭīb al-Baghdādī (1971), Sharaf aṣḥāb al-ḥadīth, ed. M.S. Khaṭīboghlu, Dār Iḥyāʾ
al-Sunna al-Nabawiyya, Ankara, p. 79, no. 170.
41 Al-Jāḥiẓ, K. al-Tarbīʿ, ed. Pellat, § 46.
42 tg, vol. vi, p. 314, no. 14.
43 Cf. above, n. 3. For refutations of unspecified mulḥids, zindīqs, and aṣḥāb al-ṭābāʾiʿ written
about the same time, see Ibn al-Nadīm, Fihrist, ed. Tajaddud, pp. 204, 3, 11; 205, 8; 206, 13sq.;
214, 15; 215, 2, 4, 7, 9.
44 Cf. tg, vol. iv, pp. 89 sqq., 289 sqq.
45 Al-Jāḥiẓ, “Ḥujaj al-nubuwwa”, in Rasāʾil, ed. Hārūn, vol. iii, p. 250. Cf. also the reference to
some who yatabaddadu and some who yatadahharu at vol. iii, p. 246, 9.
112 chapter 5
on the community.46 For all that, he observes that if the pious ancestors who
collected the Qurʾān had also collected the signs, miracles, and proofs of the
prophet,
78 then it would not have been possible today for a denying Zindīq, a stub-
born Dahrī, a libertine dandy, a misled person of feeble intelligence/edu-
cation, or a duped young man (lā zindīq jāḥid wa-lā dahrī muʿānid wa-lā
mutaẓarrif 47 mājin wa-lā ḍaʿīf makhdūʿ wa-lā ḥadath maghrūr) to deny
the reality and truth of these events […]. Nor would the godless person
(mulḥid) have found an opportunity to win over the stupid person or
deceive the young48 […]. If we didn’t have so many people of weak intel-
ligence/education (ḍuʿafāʾ) and so many intruders (dukhalāʾ) who speak
our language and seek the help of our intellects against our stupid and
foolish ones, then we would not take it upon ourselves to lay bare what is
already clear.
The ancestors had omitted this task because it had not been necessary in their
time, and what had caused the “ignorant, young, foolish, and reprobate people”
to appear now was that they would “apply to their intellects more subtleties of
kalām than they can master before having learnt the bulk of it”, with the result
that they “stray from the truth to the right and to the left”.49
Here we have another indication of the social circles in which Dahrism, as
also Zandaqa, flourished: the smart set. The questions over which people were
in danger of apostatising appealed to the young and clever who liked to see
themselves as sophisticated (ẓarīf ) and who would adopt a nonchalant atti-
tude to conventions, indulge in mujūn (playful inversion of norms), and gen-
erally madden their elders with their inappropriate behaviour. The Dahrīs and
the Zindīqs are depicted as interlopers: they are non-Arabs using our language,
that is to say people who have been brought into the community by the con-
quests and whose baleful influence is now all too widely felt. They seek the
help of “our intellects” (ʿuqūlinā) against the foolish, presumably meaning that
46 Ibid., vol. iii, pp. 224 sqq.

47 Thus the edition of Ḥ. al-Sandūbī (1933), Rasāʾil al-Jāḥiẓ, al-Maktaba al-Tijāriyya al-Kubrā,
Cairo, p. 119, 7 (Hārūn has mutaṭarrif ).
48 Reading ghabī for ghanī with al-Sandūbī, and yastamīluhu for yastamlihu with Hārūn.
49 Al-Jāḥiẓ, “Ḥujaj al-nubuwwa”, in Rasāʾil, ed. Hārūn, vol. iii, pp. 226sq. (ed. al-Sandūbī,
p. 119); for other translations, see C. Pellat (1953), Le Milieu baṣrien et la formation de Ǧāḥiẓ,
Librairie d’ Amerique et d’ Orient Adrien-Maisonneuve, Paris, p. 84; id., Life and Works,
p. 40.
they seek to mobilise our rationality against our faith, succeeding among the
young.50 They are numerous and dangerous in that they seduce the young, and
their means of seduction is kalām, which clever young people think they master
without having any proper knowledge of it. “One misfortune is that every Mus-
lim thinks that he is a mutakallim and that nobody is more entitled to argue
with the mulḥids than anyone else”, as al-Jāḥiẓ remarks elsewhere, implicitly
admitting that the encounters took place as much because the Muslims sought
them out as because the godless were conspiring to undermine the faith of the
believers.51
Kalām in this material is not simply a defensive tool, as later Muslims were 79
often to see it,52 but on the contrary the all too enticing instrument of the very
people who had to be combated, certainly the only means by which they could
be combated, but also a lure and a snare, even in the eyes of someone like al-
Jāḥiẓ, to whom kalām was the queen of the sciences. He depicts the half-studied
people who thought they were masters of the craft as one of its banes in his
epistle extolling the virtues of kalām as well.53 As a professional, he wanted to
keep control of his craft. But there can be no doubt that all those who were
hostile to kalām, whether Jews or Muslims, had good reason to be worried by
it; they were not simply being obscurantist.
Al-Jāḥiẓ also gives us some evidence on the cultural origins of Dahrism. “We
all know that the intelligence of the ancient Greeks (al-yūnāniyya) was greater
than suggested by their belief in Dahrism (al-diyāna biʾl-dahriyya) and their
attentive worship of the signs of the zodiac and the stars”, he casually remarks,
noting that the intelligence of the Indians is likewise greater than suggested
by their obedience to al-budd and worship of al-bidada, presumably meaning
the Buddha and Buddha-idols.54 Dahrism was pagan Greek thought to him; he
did not associate it with India. In line with this, the Christians of al-Jāḥiẓ’s time
claimed that the Muslim philosophers were made on the model of the Chris-
tians (iqtadaw ʿalā mithālihim).55 Christians did in fact speak of their theolo-
gians as philosophers, and they were prominent in medicine and astrology too,
50 Pellat has “s’aident de nos spéculations” (Le Milieu, p. 84), and “taking advantage our
debates” (Life and Works, p. 40).
51 Al-Jāḥiẓ, “Al-Radd ʿalā ʾl-naṣārā”, in Rasāʾil, ed. Hārūn, vol. iii, p. 320.
52 Thus for example Ibn Khaldūn (n.d.), Muqaddima, Beirut, p. 507; F. Rosenthal (tr.) (1958),
The Muqaddimah: An Introduction to History, 3 vols., Princeton Univ. Press, Princeton,
vol. iii, p. 34.
53 Al-Jāḥiẓ, “Ṣināʿat al-kalām”, in Rasāʾil, ed. Hārūn, vol. iv, p. 246.
54 Al-Jāḥiẓ, K. al-Ḥayawān, ed. Hārūn, vol. v, p. 327, 5.
55 Id., “Al-Radd ʿalā ʾl-naṣārā”, in Rasāʾil, ed. Hārūn, vol. iii, p. 315, 10.
114 chapter 5
as al-Jāḥiẓ himself noted with regret, stressing that the sciences which made
them so prestigious were in fact taken over from the ancient Greeks.56 The reli-
gion of the Christians resembled Zandaqa and Dahrism in some respects, he
says, and the Christians were the source of all perplexity, being more strongly
affected by Zandaqa, confusion, and perplexity than anyone else.57 Indeed, it
was thanks to their mutakallims, doctors, and astrologers that the Manichaean,
Marcionite, Bardesanite, and other books had fallen into the hands of the ele-
gant set, the frivolous and foolish young men whose desire to put on airs he
bewails again. But here the complaint is entirely about Zandaqa without refer-
ence to Dahrism.58
80 Conclusion
Al-Jāḥiẓ’s information may be summarised as follows. Dahrism was a convic-

tion based on an empiricist epistemology to the effect that all genuine knowl-
edge must be based on sense impressions, especially things one has seen for
oneself, that is personal experience, and that information from others must
be judged on the basis of reason, which in its turn must respect the rule-
bound nature of the universe. Nothing in our experience or reason enables
us to affirm the existence of God, angels, spirits, prophets, divine revelation,
prophetic revival of the dead, or other divine violations of the natural order,
nor does it allow for any belief in an afterlife. It does enable us to affirm that
matter has always existed, that it has not been created, and that it is not reg-
ulated by a deity standing outside it, but on the contrary regulates itself. The
regulatory mechanism is the heavenly sphere or (though this is only hinted at
in al-Jāḥiẓ) a fifth entity called spirit. The irreducible components of the world
are the four elementary qualities (heat, cold, dryness, and moisture) envisaged
as bodies, or the four elements (fire, earth, air, and water) envisaged as sub-
stances, with the elementary qualities as accidents. Everything in the world is
made of combinations of these four or, in the case of those who admitted the
existence of spirit, these five.
Adherents of such views did not form a sect or a school. Dahrism was an
individual opinion associated above all with the study of medicine and astrol-
ogy, but like Zandaqa, with which it is often concatenated, it also appealed
56 Ibid., pp. 313 sq.

57 Ibid., p. 315.
58 Ibid., pp. 320 sq.
to those who saw themselves as sophisticated. It was pursued in disputations

through the medium of kalām, both that of the type called kalām al-falsafa
(medicine, astrology/astronomy, cosmology) and that called kalām al-dīn (the-
ology, including epistemology). It was dangerous because it clashed with the
monotheist world view and appealed to the young. In terms of cultural origin,
it was seen as Greek, not Indian, and associated, again like Zandaqa, with the
sciences of the Christians.
Dahrism (again like Zandaqa) was combated by the Muʿtazilites, but the two
sides should not be envisaged as hostile camps in social terms. Like so many
adherents of opposing views at the time, they moved in the same circles and
were often personal friends, or even related by marriage. Al-Jāḥiẓ himself writes
about them in a calm tone, though he disagrees sharply with their views and
likes to think of them as immoral foreigners corrupting the good old Muslim
ways.
The information in al-Jāḥiẓ tallies well with that given in other ninth-century
sources, notably Ibn Qutayba, Muḥammad b. Shabīb, and Abū ʿĪsā, though this
has not been demonstrated here. Needless to say, much information about the
Dahriyya in the Islamic tradition is problematic and some of it is certainly use-
less, but the automatic assumption that hostile accounts of minorities credited
with views | completely unlike those of the majority must be dismissed as mere 81
constructions of the “other” is no better than the automatic belief that every-
thing they say must be true. Not only were the Dahrīs real, they clearly played a
major role in the formation of Muʿtazilite doctrine, some of which was formu-
lated in opposition to them, and which occasionally came dangerously close to
their views. In short, they were people of considerable importance.
Bibliography
Sources
Abdel Haleem, M.A.S. (tr.) (1995), Chance or Creation? God’s Design in the Universe,
attributed to al-Jāḥiẓ, Garnet, Reading.
Goodman, L.E. (tr.) (1978), The Case of the Animals versus Man before the King of the Jinn:
A Tenth-Century Ecological Fable of the Pure Brethren of Basra, Twayne, Boston.
Ibn Ḥazm, ʿAlī b. Aḥmad (1317–1321h.), Kitāb al-Faṣl fī ʾl-milal waʾl-ahwāʾ waʾl-niḥal, 5
vols., Cairo.
Ibn al-Malāḥimī, Maḥmūd b. Muḥammad (1990), Kitāb al-Muʿtamad fī uṣūl al-dīn, ed.
M. McDermott and W. Madelung, al-Hoda, London.
Ibn al-Nadīm, Muḥammad b. Isḥāq (1971), Kitāb al-Fihrist, ed. R. Tajaddud, Maktabat
al-Asadī, Tehran.
116 chapter 5
Ibn Qutayba, ʿAbd Allāh b. Muslim (1966), Taʾwīl mukhtalif al-ḥadīth, ed. M.Z. al-Najjār,
Maktabat al-Kulliyyāt al-Azhariyya, Cairo.
Ibn Khaldūn, see also Rosenthal.
Ibn Khaldūn (n.d.), Muqaddima, Beirut.
Ikhwān al-Ṣafā (1957), Rasāʾil Ikhwān al-Ṣafā wa-khullān al-wafāʾ, 4 vols., Dār Bayrūt,
Beirut.
Al-Jāḥiẓ, see also Hārūn, Pellat, al-Sandūbī.
Al-Jāḥiẓ (attrib.) (1928), Kitāb al-Dalāʾil waʾl-iʿtibār ʿalā ʾl-khalq waʾl-tadbīr, ed. M.R. al-
Ṭabbākh, al-Maṭbaʿa al-ʿIlmiyya, Aleppo.
Al-Jāḥiẓ (1938), Kitāb al-Ḥayawān, ed. ʿA.-S.M. Hārūn, 7 vols., Maktabat Muṣṭafā al-Bābī
al-Ḥalabī, Cairo.
Al-Jāḥiẓ (1955), Kitāb al-ʿUthmāniyya, ed. ʿA.-S.M. Hārūn, Dār al-Kitāb al-ʿArabī, Cairo.
Al-Jāḥiẓ (1955), Kitāb al-Tarbīʿ waʾl-tadwīr, ed. C. Pellat, Damascus.
Al-Khaṭīb al-Baghdādī, Abū Bakr Ahmad b. ʿAlī (1971), Sharaf aṣḥāb al-ḥadīth, ed.
M.S. Khaṭīboğlu, Dār Iḥyāʾ al-Sunna al-Nabawiyya, Ankara.
Lactantius (2003), Divine Institutes, tr. A. Bowen and P. Garnsey, Liverpool Univ. Press,
Liverpool.
Al-Maqdisī, Muṭahhar b. Ṭāhir (1899–1919), Kitāb al-Badʾ waʾl-taʾrīkh, ed. C. Huart, 6
vols., Ernest Leroux, Paris.
Al-Māturīdī, Muḥammad b. Muḥammad (1970), Kitāb al-Tawḥīd, ed. F. Kholeif, Dar al-
Mashreq, Beirut.
Pellat, C. (1967), The Life and Works of Jāḥiẓ, translations of selected texts, tr. from the
French by D.M. Hawke, Routledge, London.
82 Al-Rāzī, Abū Ḥātim (1977), Aʿlām al-nubuwwa, ed. Ṣ. al-Ṣāwī, Muʾassasa-yi Pizhūhishī-yi
Ḥikmat wa Falsafa-yi Īrān, Tehran.
Rosenthal, F. (tr.) (1958), The Muqaddimah: An Introduction to History, 3 vols., Princeton
Univ. Press, Princeton.
Al-Sandūbī, Ḥ. (1933), Rasāʾil al-Jāḥiẓ, al-Maktaba al-Tijāriyya al-Kubrā, Cairo.
Studies
Cook, M. (1999), “Ibn Qutayba and the Monkeys”, Studia Islamica 89, pp. 43–74.
Daiber, H. (1999), “Rebellion gegen Gott. Formen atheistischen Denkens im frühen
Islam”, in F. Niewöhner and O. Pluta (eds.), Atheismus im Mittelalter und in der
Renaissance, Harrassowitz, Wiesbaden, pp. 23–44.
Goldziher, I., and Goichon, A.-M. (1965), “Dahriyya”, Encyclopaedia of Islam, 2nd edition,
Brill, Leiden, vol. ii, pp. 95–97.
Hankinson, R.J. (1995), The Sceptics, Routledge, London.
Pellat, C. (1953), Le milieu baṣrien et la formation de Ǧāḥiẓ, Librairie d’Amérique et
d’Orient Adrien-Maisonneuve, Paris.
Shaki, M., and Gimaret, D. (1993), “Dahrī”, Encyclopaedia Iranica, http://www
.iranicaonline.org.
Stern, S.M. (1960), “Abū ʿĪsā al-Warrāq”, Encyclopaedia of Islam, 2nd edition, Brill, Lei-
den, vol. i, p. 130.
Van Ess, J. (1991–1997), Theologie und Gesellschaft im 2. und 3. Jahrhundert Hidschra, eine
Geschichte des religiösen Denkens im frühen Islam, 6 vols., De Gruyter, Berlin/New
York.
Walker, J.T. (2004), “Against the Eternity of the Stars: Disputation and Christian Philos-
ophy in Late Sasanian Mesopotamia”, in G. Gnoli and A. Panaino (eds.), La Persia
e Bisanzio (Atti dei Convegni Lincei 2001), Accademia Nazionale dei Lincei, Rome,
pp. 518–535.
Walker, J.T. (2006), The Legend of Mar Qardagh: Narrative and Christian Heroism in Late
Antique Iraq (Transformation of the Classical Heritage, 40), University of California
Press, Berkeley/Los Angeles/London.
chapter 6
Ungodly Cosmologies1
The reader may wonder both what the title means and why a subject of this
nature should be included in a volume on Islamic theology. The answer is that
a number of cosmologies of late antique origin which left little or no room for
God in the creation and management of the world played a major role in the
development of Muslim kalām, a field normally translated as (dialectical) the-
ology. In fact, kalām covered much the same range of topics as Greek physics, if
in a very different way: the principles (in the sense of the ultimate constituents
of the universe), the origin and end of the material world, the nature of man,
God and his relationship with us. To Greek philososphers, physics was a key
to the nature of the gods; to Muslim theologians, it was God who was a key to
physics. This was a well-known source of tension between reason as the sole
basis of the search for the truth and reason as the handmaid of revelation.
Al-Jāḥiẓ (d. 255/869), who distinguished between kalām al-falsafa, dialectical
philosophy (covering natural science), and kalām al-dīn, dialectical theology
(covering God and his relationship with us), readily admitted that philosophy
was dangerous, but nonetheless insisted that a good practitioner of kalām had
to master both fields (Crone 2010–2011: 75f.).
When the curtain opens on Muslim kalām in the mid-second/eighth cen-
tury, the field of kalām al-falsafa was dominated by thinkers whom Muslims
called Zindīqs and Dahrīs and bracketed as mulḥids, a term sometimes trans-
lated as ‘atheists’ but better rendered as ‘godless’ or ‘ungodly people’. All mulḥids
denied that God had created the world from nothing, and some denied his
creation, government, and ultimate judgement of the world altogether along
with any form of afterlife. The Muslims had to develop their own cosmology
to counter the ungodly systems, and they did so by assimilating and gradu-
ally transforming those of their rivals. The ungodly cosmologies thus show us a
bridge between late antique and Islamic thought.
Cosmology had acquired great religious importance in late antiquity, for
Zoroastrians, Gnostics, and Platonists (Christian, pagan, and other) had all
come to share the convictions that the key to our troubled human condition
was to be found in primordial events leading to the creation of this world, rather
than in early human history. All offered detailed accounts of these events,
1 I am indebted to Michael Cook for reading and commenting on a draft of this article.

ungodly cosmologies 119
and most drew on Greek philosophy for their formulation. Thinkers such as
Basilides (fl. 120–140), Valentinus (d. c. 160), Marcion (d. c. 160), Bardesanes/Bar
Dayṣān (d. 222), and Origen (d. c. 254), who had a huge impact on Near East-
ern thought on both sides of the Euphrates, all drew their main philosophical
inspiration from Middle Platonism and Stoicism. So too did the immensely
influential physician Galen (d. c. 200). The Platonic-Stoic legacy is still dis-
cernible in the thought of the Zindīqs and Dahrīs, and in kalām influenced
by them, along with occasional input from the rival Sceptical and Epicurean
schools and intriguing suggestions of a strong interest in the Presocratics. Also
discernible, however, is the magnetic pull exercised from perhaps the sixth
century onwards by Aristotle’s Categories, treated as a guide to ontology, not
just to logic. But by the fourth/tenth century the irresistible force was Neo-
platonism, carried by Ismailis and philosophers ( falāsifa) of a new type who
owed their ideas to Arabic translations of Plato, Aristotle, and the Neoplatonist
commentators. Henceforth it was the emanatory scheme of the Neoplatonists
that dominated cosmological debates; the old-style mulḥids no longer played
a major role in them, though they still attracted attention, especially for their
denial of the creator and of the afterlife (Dhanani 1994: 4 f., 182–187; Encyclopae-
dia of Islam2, s.v. ‘Dahriyya’; Encyclopaedia Iranica, s.v. ‘Dahrī’; Encyclopaedia of
Islam3, s.v. ‘Dahrīs’).
i The Actors
The mulḥids had complicated backgrounds. Some were Marcionites, Bardesan-

ites, or Manichaeans by origin, that is to say they came from Christian com-
munities of a type proscribed by the victorious Christian churches. (Even the
Manichaeans counted themselves as Christians.) But by early Islamic times
the Marcionites and Bardesanites had become so heavily Iranianized that they
were barely recognizable as Christians, and the Muslims classified all three
sects as dualist, deeming them ineligible for protected status. The communities
nonetheless survived, but many of their members appear to have been forced to
convert, or to have found it prudent to do so. It was nominal converts from these
three religions and others attracted to their beliefs who were called Zindīqs.
The term is derived from the Aramaic ṣaddīq by which the Manichaean ‘elect’
were known,2 and the Muslims sometimes used it of real Manichaeans too. Just
2 Cf. Encyclopaedia of Islam2, s.v. ‘zindīḳ’ (de Blois), decisively eliminating the derivation of the
word from zand.
120 chapter 6
as the Zindīqs were not true Muslims, however, so they were not true adherents
of the religions they had left behind. A Zindīq in the period c. 750–900 was usu-
ally a man who had lost faith in any positive religion, or even in any God.
The Dahrīs mostly seem to have their intellectual roots in the older belief
systems dismissed by Christians as ‘pagan’. When the emperor Justinian (r.
527–565) set out to eradicate paganism from the Roman empire, he took the
precaution of also persecuting those pagans who had ‘decided to espouse in
word the name of Christians’ (Procopius, Anecdota, 11: 32), and it was prob-
ably as nominal Christians that most of them survived. Those persecuted by
the Sasanians, who imposed Zoroastrianism as understood in Pārs (Ar. Fārs)
on their Iranian and occasionally also non-Iranian subjects, seem likewise to
have included pagans in the sense of people who were not Zoroastrians, Jews,
or Christians,3 but mostly they were bearers of local, non-Persian forms of
Zoroastrianism (cf. Crone 2012a: chs. 15–16). The Baga Nask, an Avestan book
preserved only in a Pahlavi summary, tells of ‘apostates’ ( yašarmogān) who
had been defeated and kept their apostasy concealed, reluctantly calling them-
selves Zoroastrian priests and teaching the good religion despite their heretical
inclinations (Dēnkard, book ix, 52: 3). These ‘apostates’ would hardly have been
forced to officiate on behalf of official Zoroastrianism if they had not been
priests of what the Sasanians took to be deviant forms of their own faith.
Whatever their origin, Dahrīs shared with Zindīqs the feature of having lost
belief in their ancestral religion without having acquired belief in another. A
dillusioned attitude is attested even among pagans who had not been forced
into any religious community. In the Jewish-Christian Pseudo-Clementines,
probably composed in Antioch or Edessa c. 300–360, one of the heroes is a
well-born pagan who believes in astrology and denies the existence of both
God and providence on the grounds that everything is governed by chance
and fate, meaning the conjunctions under which one happens to have been
born, and who resists conversion because he simply cannot believe that souls
are immortal and subject to punishment for sins. Nemesius of Emesa (c. 390)
also mentions deniers of providence and the afterlife (Nemesius, Nature, 213f.,
217). So too does Theodoret of Cyrrhus (d. c. 460), but now they were nominal
Christians to whom it was still physics that provided a key to God rather than
the other way round: it is by appeals to nature and the ancient Greeks that
Theodoret tries to persuade them (Theodoret, Providence, 9: 23f.). Saint Simeon
3 Cf. Theodore Bar Koni, Liber, mimrā i, 29 f.; Moses Bar Kepha, Hexaemeronkommentar, i.13.1–
15; Muqammiṣ, ʿIshrūn, 7: 6, where they are ṣābiʿa, clearly in the sense of pagans, not Sabians
of Harran; compare Yaʿqūbī, Tārīkh, 1: 166, 179 (Greek, Roman, and Iranian kings as Sabians);
Balīnūs, Sirr, 1: 2.3.6, p. 35.
the Younger (d. 592) found Antioch to be teeming with impious mockers whose
errors included denial of the resurrection, astrological beliefs to the effect that
natural disasters and human misbehaviour were caused by the position of the
stars, ‘automatism’ (presumably meaning the view that the world had arisen on
its own), and the claim, here characterized as Manichaean, that the creation
was due to fate or chance (van den Ven 1962: §§157, 161). On the Sasanian side
there is evidence for denial of the resurrection already in the third century.
The first attestations could concern belief in reincarnation, widespread in the
Jibāl and elsewhere, but by the sixth century the denial is coupled with loss
of faith in God/the gods, the creation, and afterlife of any kind. When the
famous physician Burzoē, active under Khusraw i (r. 531–570), lost faith in his
ancestral religion, he tried not to ‘deny the awakening and resurrection, reward
and punishment’. A Pahlavi advice work informs us that man becomes wicked
on account of five things, one of which is lack of belief ‘in the (imperishableness
of) the soul’, i.e denial of afterlife of any kind; and several other works stress that
one should be free of doubts concerning the existence of the gods, paradise,
hell, and the resurrection (Crone 2012a: 373ff.). Burzoē remained an unhappy
sceptic who held the truth to be beyond us, but others turned into assertive
materialists, that is to say Dahrīs.
In short, the mulḥids had their roots in proscribed communities whose
members had been directly or indirectly forced into Christianity or Persian
Zoroastrianism, and thereafter into Islam. Dahrīs were insincere Muslims who
professed Islam out of fear of the sword, as al-Qummī remarks (Tafsīr, 2: 270, ad
q. 45: 24).4 There can hardly be much doubt that the massive use of coercion
on behalf of God in late antiquity and early Islam had played a role in eroding
their faith in anything except their own reason, but other factors were also at
work. One was the sheer diversity of rival religions. When religions compete in
a free market situation, as in modern America, the competition can apparently
increase religiosity (Stark and Finke 2000, and other works by the same authors;
Kraus 1934: 15ff.), but it certainly did not do so in the past, when religion was
not a freely purchased commodity and when the competition between rival
forms was often felt to undermine the truth of all of them. In the sixth century
the sheer diversity of beliefs troubled Burzoē and Paul the Persian; by the tenth
century it troubled Muslims too (Crone 2006: 21 f.). The only way to evaluate
the competing claims was by use of reason.
One way in which reason came to sit in judgement over religious claims was
by disputation, a competitive sport of enormous popularity on both sides of
4 For the Dahrīs as interlopers, see also Jāḥiẓ, Ḥujaj, 118.

122 chapter 6
the Euphrates both before and after the rise of Islam (Lim 1995; Cook 1980;
Cook 2007). The rules required the disputers to base their arguments on shared
premisses, meaning that appeals to scripture and tradition were only allowed
in disputation with co-religionists, and even then it was reason which had
to sit in judgement over the different interpretations. Debaters thus learnt
to translate their beliefs into claims that could stand on their own and be
defended by Aristotelian logic. The Categories was the disputer’s Bible. Already
the third-century Apelles, a deviant Marcionite, had used dialectical syllogisms
to discredit the Pentateuch, and the Manichaeans soon learned to set aside
their extravagant mythology to become fearsome disputers (Grant 1993: ch. 6;
Lim 1995: ch. 3). There is no trace of mythology in the debate staged by Justinian
at Constantinople between a (chained) Manichaean and a certain Paul the
Persian representing the Christian side,5 nor is there in the cosmologies of
Manichaean, Marcionite, Bardesanite, and Zoroastrian origin that the Zindīqs
and Dahrīs fielded in disputation with the Muslims. Inevitably, many disputers
came to regard reason rather than scripture and tradition as the ultimate
authority at all times, not just for purposes of disputation. Al-Jāḥiẓ complains
that young men would foolishly rush into disputations with mulḥids, convinced
of their own dialectical skills, only to be seduced by them, and roundly declares
that ‘countless’ people had apostatized as Zindīqs and Dahrīs over complicated
questions of kalām (Crone 2010–2011: 72). It was in their relentless refusal of
claims based on scripture and tradition that both the godlessness and the
seductiveness of the Zindīqs and Dahrīs lay.
Zindīqs and Dahrīs are first mentioned in the 120s/740s and receive partic-
ular attention in the third/ninth century, though they continue to be attested
down to the Mongol invasions. They formed loose clusters of individuals, not
sects. Dahrīs seem mostly to have been doctors, astrologers, and others inter-
ested in the workings of nature; Zindīqs were predominantly secretaries, cour-
tiers, poets, and other members of the elegant set. How far similar convictions
flourished among uneducated urbanites and villagers is unknown.6 In learned
gatherings Zindīqs and Dahrīs would pick out inconsistencies in the Qurʾān and
ḥadīth, scoff at accounts of claims running counter to normal experience, and
sometimes mock Islamic ritual. But they lived like everyone else, observing the
normal rules of propriety and formalities of the law (Masʿūdī, Murūj, 5: 84 [3,
§1846]; Ṭabarī, Tārīkh, 3: 422f.; Van Ess 1991–1997: ii. 17; al-Rāzī, Tafsīr, 23: 18, ad
5 Photinus, Disputationes. On the several persons called ‘Paul the Persian’, see Gutas 1982:
239 n.
6 For a suggestion that the ʿāmmī might be a Dahrī, see Maqdisī, Badʾ, 1: 121.2; cf. also Mai-
monides on the multitudes (below, n. 73 and the text thereto).
q 22: 17f.), and relations between them and Muʿtazilite mutakallims appear to
have been friendly. Al-Naẓẓām (d. c. 220–230/835–45), who wrote against both
Dahrīs and mulḥids, had a brother-in-law who attributed everything to natural
causes and the stars (Jāḥiẓ, Ḥayawān, 1: 148). Zindīqs were particularly close
to the Shīʿites. Shīʿite sources abhor them and invariably depict the imams as
refuting them in Medina (Vajda 1938: esp. 222f.; Chokr 1993: esp. 109, 111–113), but
it is clear from the doctrines of the Shīʿite mutakallim Hishām b. al-Ḥakam (d.
c. 179/795) that the interaction was in Iraq and involved Muslim appropriation
and reshaping of the rival doctrines, not just refutation of them.
Dahrīs seem rarely to have been persecuted,7 but Zindīqs came in for a
purge under the caliph al-Mahdī (r. 775–785), to whom a Zindīq seems to have
been anything from a genuine Manichaean to an irreverent courtier. There
is no mention of Dahrīs in this connection, perhaps because the two terms
were sometimes used synonymously, but more probably because the Zindīqs
flourished at the court, where they sometimes inclined to Manichaeism in a
religious sense and where the poets would shamelessly jockey for position by
denouncing their rivals as Zindīqs. Mutakallims, by contrast, would close ranks
against outsiders (Jāḥiẓ, Ḥayawān, 4: 450; 6: 37). Al-Mahdī is reported to have
ordered the mutakallims to write refutations of the mulḥids (Masʿūdī, Murūj,
8: 293 [5, §3447]; Yaʿqūbī, Mushākala, 24), and whatever he may have meant
by that term (if he used it), the mutakallims did not limit their refutations to
Zindīqs. Books against dualists, Manichaeans, Dahrīs, and mulḥids in general
were composed by theologians active under and after al-Mahdī. But only their
titles survive, and we have no statements by the Zindīqs or Dahrīs themselves.
We do, however, have works presenting cosmologies closely related to theirs in
the Book of Treasures by the Christian doctor Job of Edessa (writing c. 817), the
Sirr al-khalīqa attributed to Apollonius of Tyana (Balīnūs, Balīnas) (c. 205/820?),
and the mostly fourth/tenth-century alchemical corpus attributed to the Shīʿite
Jābir (heavily Neoplatonized). We hear of books by Zindīqs, including a Kitāb
al-shukūk by a Zindīq espousing Sceptical views, but not of books by Dahrīs
(Ibn al-Nadīm, Fihrist, 204, 401; trans. Dodge, i. 387; ii. 804).8 Whether they
wrote or not, all mulḥids aired their views in disputations, the main vehicle of
religious and philosophical discussion at the time.
7 For an exception, see Rashīd b. al-Zubayr, Dhakhāʾir, 140.

8 Cf. Van Ess 1991–1997: ii. 17 and n. 20. This Zindīq, Ṣāliḥ b. ʿAbd al-Quddūs, is also credited with
dogmatist views.
124 chapter 6
ii Epistemology
(a) Scepticism
The mulḥids included both doubters and deniers (Jāḥiẓ, Ḥayawān, 6: 35 f.).
Some doubters were people suffering from religious uncertainty and loss of
faith, like Burzoē, but those who fielded doubts in disputations were Sceptics
in the technical sense of adherents of an epistemology to the effect that we can
never know the true nature of things. Such Sceptics were known as shākkūn,
juhhāl, mutajāhilūn, ḥisbāniyya, muʿānida, lā adriyya, and the like, and also, for
reasons that remain obscure, as Sūfisṭāʾiyya, ‘sophists’ (Van Ess 1966: index s.v.
‘Skepsis’; Van Ess 1968).
Scepticism is attested both as dogmatic assertion of our inability to know
and as suspension of judgement. Al-Jāḥiẓ mentions a Sceptic who held that
one could only know things by preponderance (biʾl-aghlab). This was the posi-
tion of Academic Sceptics, and Galen had expounded both their views and
those of their Pyrrhonic rivals in his De optimo docendi; perhaps al-Jāḥiẓ’s Scep-
tic had found inspiration in this work (Jāḥiẓ, Ḥayawān, 6: 37; Floridi 2002:
17). More commonly, however, it is Pyrrhonic Scepticism with its suspension
of judgement that is reflected in the sources. Pyrrhonic Scepticism had gone
into empiricist medicine (Hankinson 1995: ch. 13), and also into disputation
practice. As Gregory of Nazianzus (d. 389) remarked, Pyrrho, Sextus, and the
practice of ‘arguing to opposites’ had infected the churches (Floridi 2002: 12);
the sixth-century disputer Uranius is reported by Agathias to have been a Scep-
tic in Sextus’ tradition, and Manichaean missionaries would apparently field
Sceptical arguments in order to undermine the beliefs of potential proselytes
and convert them (Agathias, Histories, 2: 29.1, 7; Pedersen 2004: 207).
According to Sceptical mulḥids, all claims about reality had to be based on
sense impressions, preferably or exclusively autopsy (ʿiyān, what one had seen
for oneself) (Jāḥiẓ, Ḥayawān, 4: 449; Ḥujaj, 247; Muqammiṣ, ʿIshrūn, 14: 1; Ibn
Qutayba, Taʾwīl, 133; trans. 149 [§170]). Bashshār b. Burd (d. 163/783), a poet
variously classified as a Zindīq, Dahrī, and mutaḥayyir (somebody perplexed
or sceptical),9 is said to have believed only in what he had seen for himself
and what was similar to it (mā ʿāyantuhu aw ʿāyantu mithlahu) (Abū ʾl-Faraj,
Aghānī, 3: 227). The meaning of ‘similar’ is unclear. Perhaps he was referring to
the principle of ‘transition to the similar’ current in empiricist medicine (if you
had personal experience of a disease affecting the upper arm, you could apply it
9 Ibn Durayd, Ishtiqāq, 299; Abū ʾl-Faraj, Aghānī, 3: 147 (mutaḥayyir mukhallaṭ); Chokr 1993:
285.
to the upper leg);10 but he could also have meant unanimous transmission from
others. In any case, as this and other passages show, Scepticism was based on
empiricist premisses.
The premisses were meant for rejection, however, for even sense impres-
sions were unreliable, the Sceptics said. They would trot out the better-known
tropes of their Greek predecessors (honey tastes bitter to a jaundiced patient;
buildings appear small at a distance; poles appear bent under water, and so
on); and as in antiquity their exasperated opponents would react by wanting to
slap or beat them in order to demonstrate the reality of the sense impressions
they were dismissing (Van Ess 1966: 172f.; Van Ess 1968: 1 f.; Māturīdī, Tawḥīd,
153.18). As Sextus said, this rested on lack of familiarity with Sceptical doctrine:
Sceptics did not reject the sense impressions that induced assent involuntarily,
but merely refused to dogmatize about the reality behind them; they granted
that honey appeared to be sweet, but whether it was sweet in essence only
a dogmatist would claim to know (Sextus Empiricus, Outlines, 1.13.19f.). This
was the position of the Sūfisṭāʾīs too. Unlike their Greek predecessors, how-
ever, they are often presented as doubting the very existence of such a truth
or essence (ḥaqīqa), not just its knowability (this could reflect Buddhist influ-
ence, cf. Crone 2012b: 31f.).
A Sceptic who asserted that we cannot know the truth laid himself open to
the charge of self-contradiction, since his assertion was a truth-claim. The pru-
dent Sceptic would suspend judgement. Though both positions are reflected
in the arguments against Sceptics in the Muslim material, there is no term
for suspension of judgement there: the prudent Sceptic merely says, ‘I don’t
know’ (e.g. Baghdādī, Uṣūl, 319). Two terms for it turn up among the believ-
ers, however. One is irjāʾ, coined around 100/720 by Murjiʾites on the basis of
q 9: 107. The Murjiʾites subscribed to the Sceptical claim that one could only
judge things on the basis of autopsy and unanimous information from others;
since neither was available in the case of the caliph ʿUthmān (killed in 35/656),
one had to suspend judgement on the divisive question whether he had been
rightly guided or a sinner (Cook 1981: chs. 5, 7). The scope of their scepticism
was narrow and the term irjāʾ remained tied to their doctrines. The other term
is wuqūf or tawaqquf. Al-Jāḥiẓ, for example, observes that the common peo-
ple are less prone to doubt than members of the elite because they do not
‘hold back’ ( yatawaqqafūna), but rashly declare things to be true or false (Jāḥiẓ,
Ḥayawān, 6: 36f.). The term appears in later texts too, but it is less prominent
10 Hankinson 1995: 229. Ḥunayn was later to translate ‘transition to the similar’ as al-intiqāl
min al-shayʾ ilā nāẓirihi (Strohmaier 1981: 188).
126 chapter 6
than takāfuʾ al-adilla, the expression for the equal weight (isostheneia) of com-
peting proofs that made suspension of judgement necessary. We first hear of
belief in the equipollence of proofs in the mid-third/ninth century; a century
later the philosopher Abū Sulaymān al-Manṭiqī (d. c. 375/985) depicted it as
a characteristic of mutakallims in general, including their leading men, saying
that he would give their names if he did not prefer to leave them alive (Tawḥīdī,
Muqābasāt, 227 [no. 54]).11 The proofs that were so often found to be of equal
weight, and thus to cancel each other out, were those tried and tested in dispu-
tations about kalām al-dīn. Some adherents of takāfuʾ al-adilla would suspend
judgement on inner-Islamic disagreements alone, but others found it impossi-
ble to affirm anything apart from the existence of the creator; and still others
would suspend judgement even on him (Ibn Ḥazm, Faṣl, 5: 119f.).12 There were
also Sceptics who declared all religious tenets to be sound, the truth being rel-
ative to those who asserted it (Baghdādī, Uṣūl, 319.10; Ibn al-Jawzī, Talbīs, 41,
citing Nawbakhtī); the judge al-ʿAnbarī (d. 168/784) upheld this principle in
inner-Islamic disagreements (Goldziher 1920: 178f.). Scepticism affected Chris-
tians and Jews no less than Muslims (Jāḥiẓ, Radd, 315; Saadia, Amānāt, 13, 65 ff.;
trans. 17, 78ff.), and it had its uses for believers too. The tropes against the reli-
ability of sense impressions were apparently adduced in support of Ashʿarite
atomism (Macdonald 1927: 336; Van Ess 1966: 178), and all arguments against
the ability of humans to reach the truth could be used in a fideist vein.
(b) Dogmatism
Most mulḥids were dogmatists. They agreed with the Sceptics that all claims
about the realities of things had to be based on sense impressions, preferably
or only on autopsy,13 but unlike the Sceptics they deemed sense impressions to
be reliable and admitted a modest amount of inference from them. One could
make deductions (istidlāl) from perceptions to the reality of things, provided
that they were perceptions of regularities (al-ʿādāt) (Jāḥiẓ, Ḥayawān, 6: 269).
Anything regularly observed in large or common objects could be postulated
for small or rare ones too, since quantity did not affect their epistemological
status (ḥukm qalīl al-shayʾ ka-ḥukm kathīrihi). The nature of invisible or absent
things could similarly be observed from those observed (mā ghāba ʿanhum
mithl alladhī shūhida), but only as long as they were of the same type: ‘they
assign everything to its likes (ashkāl) and oblige it to follow the rules of the
11 Cf. Van Ess 1966: 221 ff.; Van Ess 1991–1997: index, s.v. ‘takāfuʾ al-adilla’.
12 Typically, he does not name any Muslims, only two Jewish doctors.
13 E.g. Jāḥiẓ, Ḥayawān, 4: 89 f., 449.4; 6: 269.5; Ibn Qutayba, Taʾwīl, 133; trans. 149 (§170);
Māturīdī, Tawḥīd, 111.–2; Saadia, Amānāt, 63; trans. 75; Ibn al-Jawzī, Talbīs, 41.
genus ( jins)’ (Abū ʿĪsā al-Warrāq in Ibn al-Malāḥimī, Muʿtamad, 550 f./597 f.).
They would reject all postulates about the invisible world (al-ghāʾib ʿanhum)
that ran counter to what they themselves could observe (al-ḥāḍir ʿindahum);
they applied ‘criteria for corporeal things to spiritual entities’, as Ibn Qutayba
said in defence of ḥadīth that the mulḥids deemed ridiculous (Ibn Qutayba,
Taʾwīl, 127.1; trans. 142f. [§164f.]). Information from others (akhbār, samʿ) they
admitted only if it conformed to these rules. Accordingly, they rejected the
Qurʾānic account of sinners who were transformed into monkeys and pigs, or
accepted it only in a naturalist interpretation. They scoffed at the Qurʾānic story
of the jinn who tried to listen in to conversations in heaven only to have balls
of fire thrown at them (q 72: 8f.; cf. 15: 17f.; 37: 7 f.), objecting that creatures
supposedly endowed with superior intelligence would have learnt better from
the Qurʾān (which they had supposedly heard), from their long experience,
from plain seeing for themselves, and from information passed around among
themselves. They also found fault with the Qurʾānic story of Solomon and
the Queen of Sheba, deeming it to be ‘evidence of the corrupt nature of your
historical tradition’ (dalīl ʿalā fasād akhbārikum) (Jāḥiẓ, Ḥayawān, 4: 70ff., 85 f.;
6: 265ff.; cf. Cook 1999: 60). That the jinn should have learnt from the Qurʾān
is an argument based on the opponents’ premisses; the rest tells us what
counted as legitimate sources of knowledge to the mulḥid: experience, seeing
for oneself, and information from others (empeiria, autopsia, and historia in the
terminology of Greek empiricist doctors) (Hankinson 1995: 227 f.).
Both al-Aṣamm (d. c. 200/815) and al-Naẓẓām were empiricists in some
respects (Ashʿarī, Maqālāt, 331.7; 335.13; Van Ess 1991–1997: ii. 399; iii. 334 f.).
For the rest the believers refuted the mulḥids on the latter’s own premisses
by means of the argument from design: one could see with one’s own eyes
that the world had been created by a wise and provident maker; it simply was
not credible that so intricate and well-designed a construction should have
come about on its own (Jāḥiẓ, Ḥayawān, 7: 12f.; Eutychius, Burhān, § 4).14 These
points are developed at length in a work falsely attributed to al-Jāḥiẓ and in the
Imāmī Shīʿite works Kitāb al-Tawḥīd and Kitāb al-Iḥlīlija (Jāḥiẓ, Dalāʾil; Chokr
1993: 97ff.).
14 Other arguments include the need for someone to hold the conflicting ‘natures’ (cf. below)
together.
128 chapter 6
iii Cosmology
All the godless people denied creation ex nihilo. Some believed God to have
created the world out of pre-existing material, others held it to have originated
on its own, and still others held that it had always existed. We may start with
the Zindīqs.
(a) Zindīqs
Zindīqs believed the pre-eternal principles to be two, light and darkness, and
explained the world as the outcome of their mixture. Those who retained
belief in God typically held the highest God to have sent a figure, variously
identified as Jesus, the holy spirit, or the apostle of light, to impose order on
the chaos resulting from the mixture; the Marcionites diverged by crediting
this task to the devil. Other Zindīqs explained the formation of the world in
terms of natural processes that are not further identified. Both the creation-
ists and the automatists often saw the mixture as having come about by acci-
dent.15
The synthesis of Middle Platonism and Stoicism was attractive to dualists
because the Platonists shared their negative view of matter, sometimes deem-
ing it positively evil (Dunderberg 2008: 125f.), while the Stoics also explained
the world as a mixture of two pre-eternal principles, one active, that is God/
logos/pneuma, and the other passive, that is matter or ‘unqualified substance’.
The concept of a divine logos (reason, word) or pneuma (spirit) that shapes
and regulates pre-existing matter, now as a demiurge sent by the highest God
and now as an impersonal principle, appears in several Platonizing and Gnos-
ticizing systems in late antiquity, including that of Bardesanes. The latter is said
also to have shared the Stoic view that everything which exists is a body (Syr-
iac gushmā, Arabic jism) (Furlani 1937: 350), even a line or a sound (Ephrem,
Prose Refutations, 2: 20, 29f.; trans. ix, xiii; cf. Ramelli 2009: 19). This implies
that he also held that bodies could completely interpenetrate and blend with
one another without losing their separate substance, a doctrine developed
by the Stoics to explain how pneuma could be present throughout matter;16
15 Cf. Encyclopaedia Iranica, s.v. ‘Bardesanes’; Encyclopaedia of Islam3, s.v. ‘Dayṣanīs’; Crone
2012a: ch. 10. The beginning was bi-ihmāl lā ṣanʿa fīhi wa-lā taqdīr wa-lā ṣāniʿ wa-lā mu-
dabbir, as Ibn Abī ʾl-ʿAwjāʾ says in Jaʿfar al-Ṣādiq (attrib.), Tawḥīd, 9.
16 Cf. Long and Sedley 1987: no. 48: the soul pervaded the whole body while preserving its
own substance in mixture with it, as did fire and glowing iron, and a drop of wine in the
ocean (contrary to what Aristotle said). Long and Sedley adopt ‘blending’ for complete
instead, however, Bardesanes is reported to have been an atomist. According to

Ephrem, he held that the pure elements (light, air, water, and fire), suspended
in the vacuum between God and darkness (inert matter), were composed of
atoms (perdē, seeds) and that the same was true of darkness;17 some Bardesan-
ites held reason (hawnā), power (ḥaylā), and thought (tarʿīthā) likewise to be
composed of atoms (Ephrem, Prose Refutations, 2: 220; trans. civ; Possekel 1999:
119 f.). Both the Stoic concept of interpenetration, based on the premiss that
bodies are infinitely divisible, and the Epicurean concept of atoms, directed
against infinite divisibility, allow two ingredients to blend completely without
losing their identity, a crucial point to those who saw the world as composed
of ultimately separable light and darkness. (The Zoroastrians, to whom the
world was composed out of Ohrmazd’s own substance, saw darkness as mixed
in by juxtaposition.18) And whatever Bardesanes himself may have said, both
doctrines seem to have been current in his and other schools. All things com-
mingled were capable of being separated again, as third-century Sethians of
apparently Mesopotamian origin declared, encouraging their disciples to study
the doctrine of krasis and mixis (Hippolytus, Refutatio, 5: 21.1 f., 4 f.).19 Inter-
penetration is reported under the name of mudākhala in Muslim sources on
the Manichaeans (Ashʿarī, Maqālāt, 327.15),20 and it appears without a name
of its own in the Melkite Christian Eutychius (d. 940) in explanation of the
mixture of the divine and human nature in Christ.21 The idea that all things
are bodies interpenetrating one another went into early Muslim cosmology
interpenetration without destruction of the bodies involved (fire and red-hot iron; a drop
of wine in the ocean), and use ‘fusion’ for the mixture of the type in which the bodies
are destroyed and another generated (as in drugs); but there seems to be no consistent
terminology in the Greek material: the qualification di’ holou/holōn is used in connection
with both blending and fusion, and both are called krasis and mixis too.
17 Ephrem, Refutations, 1: 53 (vacuum); 2: 214 ff.; trans. lv; ii, ciff. (darkness at 215; trans. cii);
Encyclopaedia Iranica, s.v. ‘Bardesanes’; Possekel 1999: 116ff. Ephrem is the only source for
Bardesanite atomism.
18 Cf. de Ménasce 1973: no. 403: light and darkness do not mix absolutely, as proved by fire;
light has merely adjoined smoke.
19 For these Sethians, cf. Crone 2012a: 200 f. Note also the Valentinian idea that Jesus, the
Church, and Wisdom formed a complete blending of bodies (di’ holōn krasis tōn sōmatōn)
in Casey 1934: 17.1.
20 Cf. Ashʿarī, Maqālāt, 349.11 on the Dayṣānīs, where the term is imtizāj.
21 Eutychius, Burhān, nos. 122f., with the soul and body, fire and glowing iron as examples.
The use of Stoic mixture theory in this context goes back to Gregory of Nazianzus (cf.
Stewart 1991: 182, 186).
130 chapter 6
in the physics of Hishām b. al-Ḥakam, al-Aṣamm (at least partially), and al-
Naẓẓām.22 Other mutakallims rejected infinite divisibility and interpenetration
in favour of atomism.
Muslim sources report atomism for some Manichaeans/dualists, including
one al-Nuʿmān al-Thanawī (executed by al-Mahdī), Isḥāq b. Ṭālūt, and Ibn
Akhī Abī Shākir (al-Dayṣānī) (Ibn al-Malāḥimī, Muʿtamad, 566f., 590/611, 631;
ʿAbd al-Jabbār, Mughnī, 5: 20; trans. 173). But more mainstream Christians also
seem to have included atomists, for Epicurus, normally denounced by Chris-
tians as an atheist and hedonist, is praised as one of the great philosophers by
the West-Syrian David Bar Paulos (Brock 1982: 25);23 and the mid-third/ninth-
century Muʿtazilite Ibn Mānūsh, a pupil of al-Naẓẓām of Origenist/Evagrian
background, envisaged humans in pre-existence as atoms (Baghdādī, Farq,
258, trans. Van Ess 1991–1997: vi. 220; cf. Crone 2014). The idea of disembod-
ied humans as atoms was probably due to Plato, who had defined the soul
as ‘uncompounded, indissoluble, and indivisible’, according to Albinus’ hand-
book, or, as Israel of Kashkar (d. 877) put it, as a jawhar wāḥid ghayr munqasam
ajsāman, ‘one substance/an atom, not divisible into bodies’.24 The idea of man
as an atom was also espoused by the Muʿtazilites Muʿammar (d. 215/830) and
Hishām al-Fuwaṭī (d. 220s/840s?), both atomists in cosmological terms as well
(Ashʿarī, Maqālāt, 331.13; ʿAbd al-Jabbār, Mughnī, 11: 311). In short, atomism prob-
ably reached the Muslims from both Christians and dualists.
Muslim mutakallims seem to have accepted the existence of atoms as a
matter of course, reserving their ire for the infinite divisibility of bodies because
there could not in their view be infinity in the created world. Atoms and
accidents were all there was to it in their view. Some third/ninth-century
mutakallims held atoms to have sides, explained as accidents, while others
denied that they had either sides or magnitude (Ashʿarī, Maqālāt, 316.1, 10, cf.
also 8; trans. with comments in Dhanani 1994: 99, nos. 1, 3, cf. also 2). Both
groups seem to have conceived of the atom as an Epicurean minimal part:
several such minimal parts (elachista, minima) made up an atom according
to Epicurus, though it could not in practice be divided. To Epicurus, however,
the minimal parts had magnitude. To the mutakallims, by contrast, magnitude
was either added as accidents which could not in practice be separated from
it, or else it was generated by the combination of several atoms. On their
22 Cf. Van Ess 1991–1997: i. 362, 365 f.; ii. 398 ff.; iii. 335ff.; Van Ess 1967: 250ff. The doctrine of
mudākhala is not mentioned in the exiguous material on Ḍirār.
23 Democritus is also lauded, but he had come to stand for many things.
24 Albinus, Didaskalikos, 59 (cf. Plato, Phaedo, 80b); Israel of Kashkar, Unity, no. 49. The date
of the work is not certain.
own, the minimal parts had lost their dimensions. The first known Muʿtazilite
propounder of the atom without dimensions is Abū ʾl-Hudhayl (d. 226/841),
according to whom bodies had length, breadth, and depth, whereas atoms did
not.25 It has long been suspected that he and others were indebted to dualists
such as Bardesanites or Manichaeans for their atomism (Pretzl 1931: 127ff.;
Dhanani 1994: 4f., 182ff.), and he must be refuting dualists when he denies that
atoms have life, power, or knowledge, the characteristics of light. He also denied
that they possessed colour, taste, or smell, the properties possessed by Bar
Dayṣān’s elements and, presumably, the atoms of which they were composed
(Ashʿarī, Maqālāt, 315.5). But only corporeal atoms are attested for the dualists.
Bar Dayṣān’s elements varied from light to heavy and fine to coarse;26 and
the atoms of al-Nuʿmān al-Thanawī, a Manichaean who disputed with Abū
ʾl-Hudhayl (Van Ess 1991–1997: i. 443), certainly had three dimensions (Ibn
al-Malāḥimī, Muʿtamad, 590/631; ʿAbd al-Jabbār, Mughnī, 5: 20; trans. 173). By
contrast, humans in pre-existence are unlikely to have possessed corporeal
dimensions, since they were with God; and some Christians or dualists do in
fact seem to have envisaged the lightest atoms as mere points, for the sixth-
century Barḥadbeshabbā envisages Epicurus and Democritus as believing in
fine bodies which were ‘incorporeal atoms’ (perdē delā geshūm).27
It was probably from Christians of some kind that atoms passed to the author
of the Sirr al-khalīqa (c. 210/825?). He operates with a prime substance (al-
jawhar al-awwal) which is present in everything (Sirr, 1: 1.1.3, p. 3.9), which
was clearly pre-eternal in the work he was adapting (Sirr, 2: 4.1, pp. 104f.; 2: 5.1,
pp. 109ff.),28 and which must be the source of the atoms (ajzāʾ lā tatajazzaʾu) of
which he says that the world was built and the whole macrocosmos made (Sirr,
2: 18, p. 197.9; 2: 19.1, p. 203.ult.). As to how this happened, all we are told is that
the substance was uniform until the accidents arose in it, whereupon its par-
ticles or atoms (ajzāʾ) diversified (Sirr, 1: 1.1.3, p. 3.10). Mostly the author writes
as one of the aṣḥāb al-ṭabāʾiʿ (discussed in Section iii[b]) to whom ‘everything
is from the four natures, which are heat, cold, moisture, and dryness’ or ‘which
are fire, air, water, and earth’ (Sirr, 1: 1.1.3, p. 3.4; 3: 20, p. 307.5), and the only
atoms that interest him are those of light and subtle things such as fire, the
subtlest of all bodies, composed of heat and atoms, or ‘resting air’, composed
25 Ashʿarī, Maqālāt, 307.10, where Muʿammar and al-Jubbāʾī agree. Abū ʾl-Hudhayl died after
Muʿammar, but at the age of around a hundred.
26 Ephrem, Refutations, 1: 52 f.; trans. livf.; 2: 159; trans. lxxiv; cf. Ehlers 1970: 346f.
27 Barḥadbeshabbā, Cause, 365. He locates them in Alexandria.
28 Cf. Weisser 1980: 174 f.
132 chapter 6
of warmth, moisture, and atoms, or the air between the spheres, which is full
of atoms (Sirr, 2.18, p. 197.9, cf. 2: 17.2, p. 192; 2: 16.3, p. 190.1; 2: 19.1, p. 203.11).
The different types of spiritual beings (rūḥāniyyāt) or angels were created out
of the subtle (particles) of the prime substance (laṭīf al-jawhar al-awwal), more
precisely from the heat of the wind, the light of fire, and the flow of water. Like
the prime substance before the onset of accidents, they were jawhar wāḥid (lit.
‘one substance’), here in the sense of uncompounded, and they were so subtle
that they had no corporeal matter (lā ajrāma lahā) and did not take up space;
‘everything which is not a body with six sides ( jirm musaddas) does not take
up space (makān)’ (Sirr, 2: 15.1, p. 149; 2: 15.3, pp. 153 f.). In short, spiritual beings
formed part of the created, material world, but not that of gross, tangible mat-
ter ( jirm, ajrām). They had spiritual bodies, as one might say. Like everything
else, they must have been made of atoms, but apparently these atoms lacked
dimensions. Abū ʾl-Hudhayl called an atom a jawhar wāḥid and he too dis-
tinguished them from bodies with six sides, meaning top, bottom, front, back,
left, and right, an archaic definition of bodies which appears four times in the
Sirr (Ashʿarī, Maqālāt, 302f.; Sirr, 1: 3.5.2, p. 64; 1: 3.9.4, p. 94; 6: 28.7, p. 510),
but which is replaced by the standard three dimensions in later summaries of
Abū ʾl-Hudhayl’s doctrine.29 The evidence of the Sirr suggests that it was the
desire to identify the atomic structure of intelligibilia below the level of God
himself (angels, humans in pre-existence and in spiritual afterlife, numbers,
and ideal geometric figures) that had generated the concept of incorporeal
atoms.30
It was clearly atoms of Greek rather than Indian origin that the dualists trans-
mitted (Dhanani 1994: 97ff.), though the Muslim recipients are unlikely to have
been aware of their ultimate cultural origin. The Mīzān al-ṣaghīr attributed to
Jābir, which expounds a cosmology related to that of the Sirr, tells us that the
prime substance is dust which becomes visible when the sun shines on it (Haq
1994: 55). According to Lactantius (d. c. 325), who wrote against Epicureans,
Leucippus had compared the atoms to ‘little particles of dust in the sun when
it has introduced its rays and light through a window’.31 This comparison could
29 Thus already Ashʿarī, Maqālāt, 307.11, 314.14; two further examples in Van Ess 1991–1997:
v. 37.
30 Cf. Dhanani 1994: 185, who points to the role of geometry. Sextus Empiricus’ Against the
Mathematicians and the late antique development of Aristotle’s concept of noetic matter
might repay a study from this point of view. Both Epicureans and Pyrrhonic Sceptics
rejected Euclidean geometry (Dhanani 1994: 103). Cf. also Langermann 2009, suggesting
that Galen played a role.
31 Lactantius, De ira Dei, 10: 9. Lactantius quotes him as calling the atoms seeds (semina, 10:
also have reached the Muslims via Platonist Christians and/or dualists, whose
formative period lay in the second and third centuries; back then the Epicurean
school tradition was still alive.
(b) Dahrīs: Aṣḥāb al-ṭabāʾiʿ

Dahrīs were either aṣḥāb al-ṭabāʾiʿ or aṣḥāb al-hayūlā. The former, whom I shall
henceforth call physicists, owed their name to their belief that everything in
this world is composed by four ‘natures’ (Greek physeis, Syriac kyānē, Arabic
ṭabāʾiʿ), that is the four elementary qualities, hot, cold, dry, and wet, which
combined to form the four elements, fire, water, air, and earth. Each element
had two qualities according to Aristotelians (fire was hot and dry), but only
one according to the Stoics (fire was hot). Since the Stoics identified both the
elements and their qualities as bodies, they did not distinguish sharply between
the two, as Plutarch (d. 120), Galen (d. c. 200), and Alexander of Aphrodisias
(fl. c .200) complained (Lammert 1953: 489f.); and assisted by the medical
humour theory, the qualities came to acquire ontological, as opposed to purely
analytical, priority. When late antique authors speak of the elements, they
often mean the qualities,32 and the term ‘natures’ was used of both.33 In Arabic
the ‘natures’ are usually the qualities, but sometimes the elements, otherwise
known as usṭuqussāt, ʿanāṣir, and ummahāt (mothers).34
Some physicists refused to affirm the existence of anything other than the
four elementary qualities, whereas others added a fifth (Abū ʿĪsā in Ibn al-
Malāḥimī, Muʿtamad, 547.13/594.17; Ashʿarī, Maqālāt, 348.5 f.). Just as the di-
verse colours produced by dyers were all mixtures of white, red, black, and
green, so all things in this world were really mixtures of hot, cold, moisture, and
dryness, the former said, using a comparison strikingly similar to that of Empe-
docles, the ultimate author of the four-elements theory (Māturīdī, Tawḥīd, 112,
3), cf. Syriac perdē. For the dust as partless (habāʾ lā juzʾ lahu), see Kraus 1942: 154 n.; Fakhr
al-Dīn al-Rāzī in Pines 1997: 157, on the atomic theories of the ancients (who could be
Greeks or Muslims).
32 The elements are identified as the qualities in, for example, Philastrius, Diversarum, xix:
5 (47, 5 f.), citing the mid second-century Apelles; Athanasius, Contra Gentes, par. 27; Job
of Edessa, Treasures, 1: 1 (p. 78; trans. 5).
33 Cf. Kraus 1942: 45, 165 n. 7; Ephrem, Commentary, 75 and n. 24, ad Gen. 1: 1; Jacob of Sarug,
Sermons, 2: 177, cf. 4: 319 f.; Jacob of Edessa in Teixidor 1997: 125.
34 For the mothers, see Yaʿqūbī, Tārīkh, 1: 170.11; Sirr, 2: 16.2, p. 187.ult.; 3: 20, p. 308.2; mulḥaq 1,
pp. 532 f.; Weisser 1980: 176, citing K. Isṭamāṭīs; Abū Ḥātim al-Rāzī, Iṣlāḥ, 166.15; Māturīdī,
Tawḥīd, 60.17, where they are coupled with ‘fathers’, i.e. the spheres and the stars or the
lords in charge of their motion, cf. Walker 1993: 103 (al-Sijistānī); Madelung 2005: 159.
134 chapter 6
141).35 The fifth nature added by others was often identified as spirit (rūḥ),
which pervaded and regulated everything and was also life: this was presum-
ably another Stoic legacy (Abū ʿĪsā in Ibn al-Malāḥimī, Muʿtamad, 547/594;
Ashʿarī, Maqālāt, 335.4, 11).36 Others held the fifth nature to be a wind differ-
ent from moving air, perhaps related to the breath or breeze (nasīm) that some
held to be life (Baghdādī, Uṣūl, 53.10; cf. Abū ʿĪsā in Ibn al-Malāḥimī, Muʿtamad,
549.9/596.3), or else it was space (al-faḍāʾ), identified as the place of things
(makān al-ashyāʾ) (Abū ʿĪsā in Ibn al-Malāḥimī, Muʿtamad, 549.2/596.10), or
knowledge (Yaʿqūbī, Tārīkh, 1: 170.14, of Greek and Roman Dahrīs). Still others
opted for the heavenly sphere (Maqdisī, Badʾ, 1: 132.–2; Baghdādī, Uṣūl, 320.12),37
which acted on the four qualities and so caused generation and corruption, or
which was the source of the four natures and everything else in the world.38
Al-Māturīdī had heard an astronomer compare the universe to a giant weav-
ing machine, with the heavenly bodies producing the variegated textile that
is life down here (Māturīdī, Tawḥīd, 143). Those who identified the heavenly
spheres as the source of everything else often credited their science to Hermes
and associated figures,39 but devotees of Hermes believed in spiritual realities
and credited themselves with both inner and external senses,40 whereas Dahrīs
had no inner eye (Asadī, Garshāspnāma, 140.11; trans. 2: 31).
The Christian physician and philosopher Job of Edessa (writing c. 817) held
God to have created the ‘simple elements’ (i.e. the qualities) and put them
together as ‘compound elements’, meaning the fire, water, air, and earth of
which everything was composed (Job of Edessa, Treasures, 1: 4; 1: 6). Several
Muslim mutakallims, al-Jāḥiẓ, Thumāma b. Ashras, and al-Māturīdī among
them, also operated with ‘natures’ created by God, without being Dahrīs, as
al-Juwaynī noted (disapproving of their view that the natures had causative
power).41 But the author of the Sirr is a creationist only in the sense that his
35 Cf. Empedocles, fr. 23, on painters who mix pigments to make pictures of everything.
36 Cf. al-Naẓẓām in Jāḥiẓ, Ḥayawān, 5: 47; Baghādī, Uṣūl, 53.12; Daiber 1999: 40.
37 This view is ascribed to Aristotle (e.g. Maqdisī, Badʾ, 2: 9) and to Hermes and Ptolemy
(Israel of Kashkar, Unity, no. 34).
38 Jāḥiẓ, Ḥayawān, 7: 12 f.; Māturīdī, Tawḥīd, 60.16; Maqdisī, Badʾ, 1: 126.12; Asadī, Garshāsp-
nāma, 139; trans. 2: 30; al-Rāzī, Tafsīr, 27: 269 f., ad q. 45: 24. cf. Balīnūs, Sirr, 2: 19.8, p. 212,
where their motion generates the mawālīd; cf. also Saadia, Amānāt, 58; trans. 70.
39 For a (perhaps) ninth-century summary of Hermetic doctrine, see Israel of Kashkar, Unity,
nos. 28–35; cf. also van Bladel 2009.
40 Balīnūs, Sirr, 1: 1.1.1, p. 2 and index s.v. ‘al-ḥawāss al-bāṭina/ẓāhira’.
41 Juwaynī, Shāmil, 237f.; Frank 1974 (where the ṭabāʾiʿ are not properly distinguished from
ṭabʿ); cf. Ashʿarī, Maqālāt, 517.2, where we hear of physicists with views on God’s speech.
God sets the formation of the elements in motion with his creative command;
for the rest the process unfolds on its own. Other Dahrīs agreed that the world
had originated in time, but not that it had a creator: it had been born of the four
eternal ‘uncompounded simples’ (al-afrād al-sawādhij), i.e. the elementary
qualities, which made things grow on their own without intent, wish, or will.42
Still other physicists held the natures to be pre-eternal, but put together by God;
and one Ibn Qays apparently held God to have joined them since pre-eternity,
so that the world was pre-eternal too (Baghdādī, Uṣūl, 70, 320). This aligned
him with the common physicist view that the four or five natures had always
existed in a state of combination or mixture (both mechanical and chemical
terms are used), so that the world as we know it had always been and always
would be.43 The universe had neither beginning nor end, be it in terms of time
or extent (misāḥa), and apparently not in terms of number (kathra) either;44
the several worlds implied were presumably successive rather than concurrent,
and separated by Stoic-type conflagrations, for at least some Dahrīs saw time
as cyclical.45
In agreement with the Stoics the aṣḥāb al-ṭabāʾiʿ identified the four or five
natures as bodies rather than incorporeal characteristics (al-Naẓẓām in Jāḥiẓ,
Ḥayawān, 5: 40; Ashʿarī, Maqālāt, 348.4). Space (al-faḍāʾ), defined as the place
of things (makān al-ashyāʾ), is explicitly said not to have been a body, suggest-
ing that it is the Stoic topos or place, identified as ‘that which is able to be
occupied by what is’ and counted as one of the four incorporeals (Abū ʿĪsā
in Ibn al-Malāḥimī, Muʿtamad, 549.2/596.3; cf. Long and Sedley 1987: nos. 27,
49). According to the pneumatic physicists, the four bodies had always been
in motion, either because movement was natural to them or because the
spirit was moving them, and their movements caused them to come together.
42 Balīnūs, Sirr, 2: 3, p. 103; Yaʿqūbī, Tārīkh, 1: 170.7, of Greek and Roman Dahrīs (sawādhij
is an Arabic plural of the Middle Persian form of Persian sādha, simple); compare Saadia,
Amānāt, 61; trans. 73, where those who hold heaven and earth to have originated by chance
explain the process along the same lines as the Sirr, without God’s creative command to
set the process going.
43 Abū ʿĪsā in Ibn al-Malāḥimī, Muʿtamad, 547.12, 549.18/594.18, 596.19; Māturīdī, Tawḥīd,
143.12. But Saadia, Amanat, 55; trans. 66, and Juwaynī, Shāmil, 239.5, present them as
claiming that the four originally existed in isolation.
44 Abū ʿĪsā in Ibn al-Malāḥimī, Muʿtamad, 549.19, 552.9/596.20, 598.21; Sirr, 1: 3.9.3, p. 93.10.
45 Yaʿqūbī, Tārīkh, 1: 168.6 (inna ʾl-dahr dāʾir), of Greek and Roman Dahrīs; Maimonides,
Guide, 2: 13 (28b); Ibn Kathīr, Tafsīr, 4: 150, ad q. 45: 24 (cycles of 36,000 years); cf. the
cycles in the thought of the communities from which Dahrīs seem often to have been
drawn (Crone 2012a: 209 f., 235 f., 239, 245 f., cf. also 481).
136 chapter 6
This sounds Epicurean, but they interpenetrated in the Stoic style ( yaghullu
baʿḍuhā fī baʿḍin) instead of simply combining. By mixing in different ways
they became sounds, smells, minerals, plants, and so on (Abū ʿĪsā in Ibn al-
Malāḥimī, Muʿtamad, 547.16; 548.4; 551.12/594.21,46 595.9, 598.4). The matter
(mādda) formed by their mixture was composed of particles (ajzāʾ), presum-
ably infinitely divisible, and things were strengthened and weakened by con-
junction with similar and contrasting forms (ashkāl and aḍdād). When a living
being died, the particles dispersed to join the concordant forms closest to it,
and the same particles might accidentally come together to form a living being
of the same kind, or of a different kind, or just a plant, or the particles might
simply be dispersed in water or the earth.47 In short, the physicists allowed for
the possibility of what others called reincarnation, but explained it in materi-
alist terms. If their roots went back to the third century, they could have picked
up this explanation from the Epicurean school tradition (cf. Lucretius, On the
Nature of Things, 3: 845–860). But whether they did so or not, it is not the only
evidence to suggest that they hailed from communities in which belief in rein-
carnation was widespread. In fact, while some members of these communities
were making godless science out of their ancestral beliefs, to be dismissed as
Dahrīs, others were reformulating them as Muslim doctrine, to be dismissed as
Khurramīs and Ghulāt (Crone 2012a: 248f.).
Neither the dualists nor the aṣḥāb al-ṭabāʾiʿ needed a material substratum
to carry their corporeal qualities, for even qualities were bodies, so they did
not accept the Aristotelian concept of prime matter,48 nor the Aristotelian
distinction between substance ( jawhar) and accidents (aʿrāḍ). Some had come
round to accepting one accident, however, namely motion, a key concept in
that it was coterminous with action and change.49 But there were also some
who claimed that there was no such thing as motion or any other accident.50
The Muʿtazilite al-Aṣamm shared this view (Ashʿarī, Maqālāt, 343.12; Baghdādī,
Uṣūl, 7.14; cf. Van Ess 1991–1997: ii. 398f.; v. 194f.). Motion was a body, i.e. the
46 Wrongly yuqillu for yaghullu in the new edition.

47 Abū ʿĪsā in Ibn al-Malāḥimī, Muʿtamad, 548.1, 9/595.6, 13; cf. Sirr, 1: 1.1.3, p. 4.4; Ashʿarī,
Maqālāt, 329.6; Maqdisī, Badʾ, 1: 127.11.
48 It is rejected as nonsense in Job of Edessa, Treasures, 1: 2. Jābir, who does operate with a
substrate, mentions those who do not (Kraus 1942: 169f.).
49 Abū ʿĪsā in Ibn al-Malāḥimī, Muʿtamad, 548.17, 566.13/595.20, 611.8; Ashʿarī, Maqālāt, 348.7,
12; 349.12; Ibn Shabīb in Māturīdī, Tawḥīd, 141.15, 143.21.
50 Ibn al-Malāḥimī, Muʿtamad, 549.15/596; Muqammiṣ, ʿIshrūn, 3:11; Ashʿarī, Maqālāt, 348.11;
349.6, 15; Baghdādī, Uṣūl, 52.16.
body moving, as some put it, which is also what a Stoic would have said.51 As
a certain Plato the Copt from Ḥulwān is reported to have declared, we do not
see motion or any other action, only the person or thing moving or acting.52
The Sirr refutes him as if he were a Sceptic, assimilating him to a different set of
people who denied the reality of change as an illusion, claiming that the created
word was all one and the same, and who seem to have invoked Parmenides
(‘Munīs’).53 It is those who dismiss diversity (ikhtilāf ) as an illusion generated
by the senses who trot out Sceptical tropes in al-Yaʿqūbī’s account of Greek and
Roman Dahrīs (Yaʿqūbī, Tārīkh, 1: 168f.).
Many Dahrīs had succumbed to the advancing tide of Aristotelianism, how-
ever. They defined the elements as substance and the elementary qualities as
accidents (al-Naẓẓām in Jāḥiẓ, Ḥayawān, 5: 40), and postulated a substrate in
the form of prime matter (hayūlā, ṭīna).
(c) Aṣḥāb al-hayūlā

Some people held the world to have been created from nothing while others
held it to be drawn from matter (hylē), Paul the Persian observed (Land 1862–
1875: 4, fo. 56r; trans. 2). Two centuries later the adherents of the latter view
were known as aṣḥāb al-hayūlā and singled out for refutation by al-Naẓẓām
(Van Ess 1991–1997: vi. 1 [no. 3]). Some aṣḥāb al-hayūlā were creationists who
held God to have created the world out of pre-existing matter (Greek hylē) by
means of movement and rest, which caused accidents to arise. The author of
the Sirr, who tacitly operates with prime matter, is an example.54 Al-Maqdisī,
who deemed them guilty of dualism, informs us that they also held that the
creator had always created (a Platonist view rooted in the Timaeus), so they
were eternalists too.55 Judging from the frequency with which the emergence
of the world is described in impersonal terms, other aṣḥāb al-hayūlā were
automatists. Their Platonism notwithstanding, the adherents of prime matter
are mostly envisaged as Aristotelians,56 with some justice in that their hayūlā
51 Ibn al-Malāḥimī, Muʿtamad, 566.–5/611.13 (Manichaean majority); Ashʿarī, Maqālāt, 349.2;

cf. 346.6, on Jahm b. Ṣafwān (on different grounds); Sedley in Algra et al. 1999: 399.
52 Sirr, 1: 2.2.11, p. 28.
53 Sirr, 1: 2.2.10, pp. 26 f.; cf. Rudolph 1995: 133 f.
54 Theodore Bar Koni, Liber, mimrā i, 30; Maqdisī, Badʾ, 1: 92; compare the Sirr, 2: 3ff.,
pp. 103 ff.
55 Maqdisī, Badʾ, 1: 92; Māturīdī, Tawḥīd, 86.13; Pines 1997: 41, 48, on the tenth-century
Īrānshahrī, one of the aṣḥāb al-hayūlā; Goodman 1993: 148; Plato, Timaeus, 29e.
56 Job of Edessa, Treasures, 1: 2; Yaʿqūbī, Tārīkh, 1: 170.14 (aṣḥāb al-jawhar); Māturīdī, Tawḥīd,
147; cf. Bar Koni, Liber, mimrā xi, 9, and Zurqān in Maqdisī, Badʾ, 1: 140, on Aristotle himself.
138 chapter 6
(also called ṭīna) was clearly Aristotle’s protē hylē, a material substrate devoid
of extension, dimensions, or any other properties, endowed with the potential
to be anything. (They do not seem to have known about Simplicius’ and Philo-
ponus’ modifications of Aristotle on this point. Māturīdī, Tawḥīd, 147.5; Sorabji
1988: ch. 2.) Hayūlā was empty of accidents, as the sources will say (Maqdisī,
Badʾ, 1: 47.8; Baghdādī, Uṣūl, 57.5), thinking in terms of substance and accidents
(as in the Categories) rather than matter and form.57 Thanks to its potential-
ity (quwwa), which often seems to be envisaged as a separate entity, accidents
arose in it, and the appearance of accidents transformed the hayūlā into sub-
stance ( jawhar) (Māturīdī, Tawḥīd, 147; cf. also 30.17). Some called prime matter
‘substance’ or ‘simple substance’ or ‘first substance’ ( jawhar basīṭ/awwal) from
the start. The term ʿunṣur also came to be used. Some held every species of
being to have its own prime matter (Baghdādī, Uṣūl, 53.5).
The aṣḥāb al-hayūlā, then, held that matter/substance was pre-eternal (qa-
dīm), but accepted that accidents originated in time (ḥadītha), with or without
divine intervention. They held that the bodies preceded the accidents, as al-
Baghdādī puts it (Uṣūl, 55.8). He held this to distinguish the aṣḥāb al-hayūlā
from other Dahrīs, for most of the Dahrīs who operated with accidents were
eternalists in respect of them too, in three different ways. Some, labelled Aza-
liyya Dahrīs by al-Baghdādī, did agree that the accidents originated in time,
but they added that before every origination there had always been another:
the process had no beginning; the world had always existed as we see it now
with its stars, animals, procreation, and so on.58 Others held that the acci-
dents had always existed in potentiality (biʾl-quwwa). According to them, and
also to (some?) Manichaeans, the accidents or the world or the phenomena (?
maʿānī) were in the prime matter/substance in potentiality and emerged from
there into actuality (ẓaharat biʾl-fiʿl); in support of this they would adduce the
presence of the man in the sperm, of the animal in the sperm or egg, of the
tree in the kernel, and so on.59 This doctrine was also known to the Zaydī al-
Qāsim b. Ibrāhīm (d. 246/860), whose mulḥid opponent adduces the date palm
57 All things are either substance (ousia) or accident, as Job of Edessa remarks (Treasures, 1: 3,
p. 81; trans. 10). The terminology was to be revised in the light of the translations, cf. matter
versus form (ṣūra) and the elementary qualities as kayfiyyāt in Shahrastānī, Nihāya, 163ff.;
Shahrastānī, Milal, 257.ult.; trans. 2: 187.
58 Baghdādī, Uṣūl, 55, 59; Muqammiṣ, ʿIshrūn, 5: 36, 42; Ibn al-Malāḥimī, Muʿtamad, 566.14/
611.9 (of some dualists, apparently Manichaeans); Maqdisī, Badʾ, 1: 123.4.
59 Māturīdī, Tawḥīd, 63.9, cf. 30.16; Muqammiṣ, ʿIshrūn, 5: 8, 10, 14 (claiming to know nobody
adhering to this view, but associating it with Dahrīs and Manichaeans); Guidi, Lotta, 46.9;
trans. 107.
in the pit (Pines 1997: 165f.). Finally, some Dahrīs held that the accidents had
always existed in the bodies, apparently in actuality. Colours, tastes, and smells
were hiding in the earth, water, and fire and became manifest in fruit by trans-
fer (intiqāl) and the conjunction of likes (ashkāl) (Ashʿarī, Maqālāt, 329.4; cf.
Maqdisī, Badʾ, 1: 47, 134.6). The adherents of this view were the aṣḥāb al-kumūn
waʾl-ẓuhūr, ‘those who believed in latency and manifestation’, and al-Baghdādī
may have conflated them with the defenders of the second position (Baghdādī,
Uṣūl, 55.12 [where the second position is omitted]; Maqdisī, Badʾ, 1: 47.4). They
too seem to have adduced the chicken and the egg, the wheat in the grain, and
so on by way of confounding those who believed the world to have a begin-
ning and an end, or perhaps all Dahrīs did so.60 At all events, they said that
when one accident was manifest, its opposite disappeared from view and was
hidden in the body until the roles were reversed, as for example in the case
of motion and rest, and so it would go on forever.61 There was no origination
(ḥudūth).62
Wolfson thought that the Dahrīs were Aristotelians, with reference to their
doctrines of potentiality and kumūn (Wolfson 1976: 504ff.); Horovitz related
these views to the Stoic concept of ‘seminal reasons’ (logoi spermatikoi), accord-
ing to which the creative fire or reason was ‘like a seed’ containing the causes of
all things past, present, and future (Horovitz 1903: 186); and Nyberg thought that
al-Naẓẓām’s kumūn theory (cf. below) must be rooted in the concept of Plato’s
ideas as thought (and thus potentiality) in the mind of God.63 But whatever
philosophical language the Dahrīs may have used, what they, and sometimes
also Zindīqs, really wished to express was a deep-seated Near Eastern convic-
tion, namely that everything is endless recurrence. This is what shaped their
understanding of Greek philosophy, and also what gave them an affinity with
the Presocratics. Whether the chicken or the egg was originated or pre-eternal,
hidden in the body, in Aristotelian potentiality, in Stoic ‘seminal reasons’, or in
the mind of God, the point was that there was nothing new under the sun. The
chicken produced eggs which produced chickens which produced eggs; so it
had always been and so it always would be. Denial of origination and destruc-
60 Jāḥiẓ, Tarbīʿ, no. 46; Kraus 1935: 299 f. (where the doctrine is primarily Manichaean);
Maqdisī, Badʾ, 1: 118 f., 133; 2: 134; Baghdādī, Uṣūl, 319.14; Juwaynī, Shāmil, 224.1; Ibn al-
Malāḥimī, Muʿtamad, 160/152.
61 Muqammiṣ, ʿIshrūn, 5:12; Baghdādī, Uṣūl, 55; Baghdādī, Farq, 139.
62 Yaʿqūbī, Tārīkh, 1: 168.3; Guidi, Lotta, 45.6; trans. 105.
63 Nyberg 1919: 52, adding that al-Naẓẓām linked it with Anaxagoras’ homoiomery theory,
which must be a slip for Anaxagoras’ opposite theory that ‘there is a portion of everything
in everything’.
140 chapter 6
tion coupled with belief in eternal recurrence and pantheism also appears in
the Hermetic corpus (Copenhaver 1992: xii. 15–17). Simon Magus is credited
with the view that fire, the principle of all things, possessed hidden and mani-
fest parts corresponding to the potentiality and actuality of Aristotle, the intel-
ligible and sensible of Plato (Hippolytus, Refutatio, 6.9.5 f., adduced by Wolfson
1976: 510). The Gnostic Basilides, who believed in a ‘not-being God’ (ouk ōn
theos) utterly beyond us, held this deity to have caused a seed to exist in which
all things were contained just as the entire plant is contained in the mustard
seed and the multicoloured peacock and other birds in the egg (Hippolytus,
Refutatio, 7: 21).64 Basilides’ system, or something similar to it, was known to al-
Yaʿqūbī, according to whom one of the Dahrite groups among the pagan Greeks
and Romans believed the origin (aṣl) of things in pre-eternity (al-azaliyya) to
be a seed (ḥabba) which split open, whereupon the world with all the diver-
sity of colours and other sense impressions appeared from it (Yaʿqūbī, Tārīkh, 1:
168.16): here as elsewhere, al-Yaʿqūbī’s ancient Dahrīs are actually late antique
and/or Islamic. Al-Maqdisī also knew them.65
Al-Naẓẓām, who shared the view of everything as interpenetrating bodies,
also held that motion was the only accident and subscribed to the theory of
kumūn: God created everything in one go, hiding future things in the bodies;
and fire was not originated, but hidden in the stone.66 His view that God created
the world all at once aligns him with Origen, but almost all his other views on
physics align him with the Dahrīs. His affinities were with the physicists, as al-
Shahrastānī said.67 The same was true of other early Muʿtazilites.68 The aṣḥāb
al-hayūlā also had an afterlife as falāsifa, represented by Īrānshahrī and Abū
Bakr al-Rāzī (the latter an atomist) (Pines 1997: 41 f., 47, 48).
64 Hippolytus saw him as a follower of Aristotle.

65 Maqdisī, Badʾ, 1: 141.11, on aṣḥāb al-juththa (read aṣḥāb al-ḥabba? For inqalaʿat, read
infalaqat).
66 Cf. Van Ess 1991–1997: iii. 339 ff., 360 ff., 367ff. (where it is noted that he is also credited with
the opposite doctrine that God creates everything new in every moment).
67 Shahrastānī, Milal, 1: 39; trans. 208; cf. Baghdādī, Farq, 113f., 127, 139; Baghdādī, Uṣūl, 48
(with much polemical exaggeration); cf. Van Ess 1991–1997: iii. 307, 332.
68 Shahrastānī, Milal, 1: 44, 52, 53; trans. 228, 257, 260, on Bishr b. al-Muʿtamir and Jāḥiẓ;
Baghdādī, Uṣūl, 36.ult., on al-Aṣamm; cf. also Van Ess 1991–1997: iii. 333.
iv Godless Religion
Dahrīs are often said not to have believed in God,69 and some must indeed
have denied his existence. But others clearly believed in him,70 and in any
case the key issue between Dahrīs and ‘monotheists’ (muwaḥḥidūn) was not
whether God existed or not, but rather what significance he had for humans.
To monotheists he had created the world and administered it, sent prophets
to mankind to make his wishes known, and would eventually call everyone
to account. To ‘pure Dahrīs’ all this was nonsense: whether there was a deity
or not, there was no creator, providential ruler (mudabbir), or lord (rabb) of
the world, nor any angels, spirits, prophets, religious laws, veridical dreams,
or afterlife of any kind.71 The alleged miracles of prophets could be explained
rationally, and demons (shayāṭīn), spirits ( jinn), paradise, and hell had been
invented to deceive people and make them obey.72 Like the Zindīqs, the Dahrīs
saw the world as simply too full of inequality, injustice, illness, violence, hos-
tility, pain, and death to have a creator or providential overseer.73 Some, how-
ever, accepted that the world had a creator (muḥdith), but held that he had
ceased to exist. ‘We see people fall into water without being able to swim, or
into fire, and call upon the provident maker (al-ṣāniʿ al-mudabbir), but he does
not rescue them, so we know the creator is non-existent (maʿdūm)’, uniden-
tified philosophers observed. After completing the world and finding it good
the creator had destroyed himself so as not to add or detract from his hand-
iwork, leaving behind the laws (aḥkām) current among the living beings and
things he had made. Alternatively his particles had dispersed in the world so
that every force in it was of the divine essence. Or a defect (? tawalwul) had
appeared in the essence of the creator so that all his power and light had been
69 E.g. Abū ʾl-Faraj, Aghānī, 13: 280; al-Māturīdī, Taʾwīlāt, 4: 94, ad q. 4: 150; cf. Kulaynī, Kāfī,
1: 76.9, on a Zindīq.
70 Cf. Ibn Qays and his likes (above, note 43 and the text thereto).
71 Jāḥiẓ, Hayawān, 7: 12 ff.; Abū ʿĪsā in Ibn al-Malāḥimī, Muʿtamad, 587.13; Khushaysh in al-
Malaṭī, Tanbīh, 72; Yaʿqūbī, Tārīkh, 1: 168.1; Maqdisī, Badʾ, 1: 119.3. For the ‘pure Dahrī’, see
Jāḥiẓ, Ḥayawān, 4: 90.1. For tadbīr (and siyāsa) as a translation of Syriac purnāsā, rendering
Greek pronoia, see Daiber 1980: 12.
72 Jāḥiẓ, Ḥujaj, 3: 263f. (cf. also 278, 281); Māturīdī, Taʾwīlāt, 17: 400.ult., ad q. 114: 4–6; Maqdisī,
Badʾ, 5: 25; Asadī, Garshāspnāma, 139; trans. 30 (ch. 44); Pretzl 1933: *23; trans. 46.
73 Kaʿbī on Dahrīs in Maqdisī, Badʾ, 1: 116; Ṣāliḥ b. ʿAbd al-Quddūs in Van Ess 1991–1997: ii. 18;
another Zindīq (Ibn al-Muqaffaʿ?) in Guidi, Lotta, 22.23, 24.3; trans. 52, 54; cf. Maimonides,
Guide, 3: 2 (18a) on Abū Bakr al-Rāzī, noting that the multitudes often shared this view. Sex-
tus had also shared it, showing us yet another affinity between Sceptics and Manichaeans
(cf. Hankinson 1995: 238).
142 chapter 6
sucked out of him and into this world; all that remained of him was a cat (!
sinnawr), which would suck the light out of this world again so that eventu-
ally he would be restored; meanwhile he was too weak to attend to his created
beings; their affairs were left unattended with the result that injustice had
spread.74 The sinnawr could be a misreading for something to do with nūr, but
the members of the Hāshimite movement in Khurāsān were accused of wor-
shipping cats, so maybe we should take it as it stands; al-Māturīdī confirms that
there were mulḥids who held God to suffer defects and illnesses (āfāt) (Akhbār,
282; Māturīdī, Taʾwīlāt, 15: 283, ad q. 67: 1). All these explanations accounted
for the orderly design of the world, the key argument against Dahrism, while
also explaining its unjust nature. There was nobody up there to look after us
anymore. The heavens were no longer inhabited, as Zindīqs reportedly said
(Kulaynī, Kāfī, 1: 75 [kharāb laysa fīhā aḥad]; cf. Māturīdī, Taʾwīlāt, 16: 309, ad
q. 75: 36).
Opponents occasionally accused Dahrīs of making the elements or the heav-
enly sphere divine, but rarely of actually worshipping them. Though natural
scientists often had a strong occult side to them, as they do in the Sirr al-
khalīqa and the Jābir corpus, the ‘pure Dahrīs’ and their Zindīq counterparts
come across as reductionists singularly lacking in religious feelings. Their ethics
were rationalist. People were obliged to know and avoid naturally evil things
such as anger, killing, and theft, nothing else, as Bashshār al-Burd said (Ibn al-
Malāḥimī, Muʿtamad, 590/631f.; ʿAbd al-Jabbār, Mughnī, 5: 20; trans. 173); Dahrīs
determined right and wrong (ḥasan, qabīḥ) on the basis of their own fancy, as
al-Jāḥiẓ caricatured them (Jāḥiẓ, Ḥayawān, 7: 13). Like atheists everywhere, they
were often envisaged as utterly immoral and depraved.
v The Persistence of Godlessness
Muʿtazilite and Shīʿite mutakallims who interacted with Zindīqs and Dahrīs
sometimes became unhinged (khulliṭa), as their colleagues said. They include
the third/ninth-century Abū Saʿīd al-Ḥaḍrī/Ḥuṣrī, the fourth/tenth-century
Abū Isḥāq al-Naṣībī,75 and Abū Ḥafṣ al-Ḥaddād (Van Ess 1991–1997: iv. 89–91),
as well as the notorious Ibn al-Rāwandī (d. mid or late fourth/tenth century).76
74 Yaḥyā b. Bishr b. ʿUmayr al-Nihāwandī (writing before 377/987f.) in Ibn al-Jawzī, Talbīs, 46
(ch. against the falāsifa).
75 Tawḥīdī, Imtāʾ, 1: 141; cf. id., Akhlāq al-wazīrayn, 202, 211f., 297.
76 Cf. Encyclopaedia of Islam2, s.v. ‘Ibn al-Rāwandī’; Stroumsa 1999: ch. 2; Van Ess 1991–1997:
iv. 295 ff.
The latter is said to have written a book on the eternity of the world and another
on its evil, but he is more famous for his view that prophets were tricksters
whose alleged miracles were open to rational explanation. This was a theme of
considerable prominence in fourth/tenth- and fifth/eleventh-century theology
and philosophy (another famous exponent was Abū Bakr al-Rāzī); so too was
the denial of the afterlife, but covering these developments would require
another chapter. Dahrī cosmology, on the other hand, went into a phase of
kumūn,77 to make a ẓuhūr in post-Mongol Iran. It was now Sufis who said that
‘there is nobody here except us’, that the world has always existed, that God does
not look after it, that he does not send messengers to it, that there is no afterlife,
and that time is endless recurrence, while Dahrī materialism reappeared in
the Nuqṭawī heresy of Maḥmūd Pasīkhānī (d. 831/1427f.). But the tone was no
longer scoffing, nor was the materialism irreligious. Maḥmūd claimed that the
four elements were all that existed, but what he meant was that God was those
elements, not that he did not exist, and though his explanation of reincarnation
was materialist (humans had no soul), it was merit which determined how one
was reborn.78 Such cosmologies were still heterodox, but they were no longer
ungodly.
References
ʿAbd al-Jabbār (Mughnī). al-Mughnī. Vol. 5. Ed. M.M. al-Khuḍayrī. Cairo: Wizārat al-
Thaqāfa waʾl-Irshād al-Qawmī, 1965 [partially trans. G. Monnot, Penseurs musul-
mans et religions iraniennes. Paris: J. Vrin, 1974]. Vol. 9. Ed. M.ʿA. al-Najjār and ʿA.-
Ḥ. al-Najjār. Cairo: Wizārat al-Thaqāfa waʾl-Irshād al-Qawmī, 1965.
Abū Ḥātim al-Rāzī (Iṣlāḥ). Kitāb al-Iṣlāḥ. Ed. Ḥ. Mīnūchihr and M. Muḥaqqiq. Tehran:
Muʾassasa-yi Muṭālaʿāt-i Islāmī, Dānishgāh-i Tihrān, McGill University, 1998.
Agathias (Histories). The Histories [= De imperio et rebus gestis Iustiniani]. Trans. J.D.
Frendo. Berlin: de Gruyter, 1975.
Abū ʾl-Faraj al-Iṣfahānī (Aghānī). Kitāb al-Aghānī. Cairo: Dār al-Kutub al-Miṣriyya/al-
Hayʾa al-Miṣriyya al-ʿĀmma lil-Kitāb, 1927–1974.
Akhbār al-dawla al-ʿAbbāsiyya wa-fīhi akhbār al-ʿAbbās wa-waladuhu li-muʾallif min al-
qarn al-thālith al-Hijrī. Ed. ʿA.-ʿA. al-Dūrī and ʿA.-J. al-Muṭṭalibī, Beirut: Dār al-Ṭalīʿa
lil-Ṭibāʿa waʾl-Nashr, 1971.
77 The last presentation in which it is alive, as opposed to an object of routinized refutation,

is Asadī’s Garshāspnāma (ch. 44), completed in 458/1065f.
78 See the eighth/fourteenth- or ninth/fifteenth-century heresiography Haftād u sih millat,
nos. 5, 19, 26, 33–35, 71; Crone 2012a: 481 ff.
144 chapter 6
Albinus (Didaskalikos). The Platonic Doctrines of Albinus [= Didaskalikos]. Trans. J. Ree-

dy. Grand Rapids, mi: Phanes Press, 1991.
Algra, K., et al. (eds.) (1999). The Cambridge History of Hellenistic Philosophy. Cambridge:
Cambridge University Press.
Asadī Ṭūsī, Aḥmad b. ʿAlī (Garshāspnāma). Garshāspnāma. Ed. Ḥ. Yaghmāʾī. Tehran:
Būrūkhīm, 1317/1938 [trans. C. Huart and H. Massé, Le Livre de Gerchasp: Poème
persan. 2 vols. Paris: P. Geuthner, 1926].
al-Ashʿarī (Maqālāt). Die dogmatischen Lehren der Anhänger des Islam [= Maqālāt al-
islāmiyyīn wa-ikhtilāf al-muṣallīn]. 2 vols. Ed. H. Ritter. Istanbul: Devlet Matbaasi,
1929–1933.
Athanasius (Contra Gentes). Contra Gentes. Ed. and trans. R.W. Thomson. Oxford: Cla-
rendon Press, 1971.
al-Baghdādī, ʿAbd al-Qāhir (Farq). al-Farq bayna ʾl-firaq. Ed. M. Badr. Cairo: Maṭbaʿat
al-Maʿārif, 1328/1910.
al-Baghdādī, ʿAbd al-Qāhir (Uṣūl). Uṣūl al-dīn. Istanbul: Devlet Matbaasi, 1928.
Balīnūs al-Ḥakīm (attrib.) (Sirr). Buch über das Geheimnis der Schöpfung und die Dar-
stellung der Natur, [oder], Buch der Ursachen = Sirr al-khalīqa wa-ṣanʿat al-ṭabīʿa,
[aw], Kitāb al-ʿilal. Ed. U. Weisser. Aleppo: Maʿhad al-Turāth al-ʿIlmī al-ʿArabī, Jāmiʿat
Ḥalab, 1979.
Barḥadbeshabbā (Cause). Cause de la fondation des écoles. Texte syriaque publié et
traduit par A. Scher. Paris: Firmin-Didot, 1908.
Brock, S. (1982). ‘From Antagonism to Assimilation: Syriac Attitudes to Greek Learning’.
In N. Garsoian et al. (eds.), East of Byzantium: Syria and Armenia in the Formative
Period. Washington, dc: Dumbarton Oaks, Center for Byzantine Studies, Trustees
for Harvard University, 17–34.
Casey, R.P. (ed. and trans.) (1934). The Excerpta ex Theodoto of Clement of Alexandria.
London: Christophers.
Chokr, M. (1993). Zandaqa et zindīqs en Islam au second siècle de l’hégire. Damascus:
Institut français de Damas.
Cook, M. (1980). ‘The Origins of Kalām’. Bulletin of the School of Oriental and African
Studies 43: 32–43.
Cook, M. (1981). Early Muslim Dogma. Cambridge: Cambridge University Press.
Cook, M. (1999). ‘Ibn Qutayba and the Monkeys’. Studia Islamica 89: 43–74.
Cook, M. (2007). ‘Ibn Saʿdī on Truth-Blindness’. Jerusalem Studies in Arabic and Islam
33: 169–178.
Copenhaver, B.P. (1992). Hermetica: The Greek Corpus Hermeticum and the Latin Ascle-
pius in a New English Translation, with Notes and Introduction. Cambridge: Cam-
bridge University Press.
Crone, P. (2006). ‘Post-colonialism in Tenth-Century Islam’. Der Islam 83: 2–38 [Ed.:
included as article 7 in the present volume].
Crone, P. (2010–2011). ‘The Dahrīs According to al-Jāḥiz’. Mélanges de l’Université Saint-

Joseph 63: 63–82 [Ed.: included as article 5 in the present volume].
Crone, P. (2012a). The Nativist Prophets of Early Islamic Iran: Rural Revolt and Local
Zoroastrianism. Cambridge: Cambridge University Press.
Crone, P. (2012b). ‘Al-Jāḥiẓ on aṣḥāb al-jahālāt and the Jahmiyya’. In R. Hansberger et
al. (eds.), Medieval Arabic Thought: Essays in Honour of Fritz Zimmermann. London:
Warburg Institute, 27–40 [Ed.: reprinted in P. Crone, The Iranian Reception of Islam:
The Non-Traditionalist Strands, vol. 2 of Collected Studies in Three Volumes, ed. H. Siu-
rua, Leiden: Brill, 2016, art. 8].
Crone, P. (2014). ‘Pre-existence in Iran: Zoroastrians, Ex-Christian Muʿtazilites, and Jews
on the Human Acquisition of Bodies’. Aram 26: 1–27 [Ed.: reprinted in P. Crone, The
Iranian Reception of Islam: The Non-Traditionalist Strands, vol. 2 of Collected Studies
in Three Volumes, ed. H. Siurua, Leiden: Brill, 2016, art. 13].
Daiber, H. (ed. and trans.) (1980). Aetius Arabus: Vorsokratiker in arabischer Überliefer-
ung. Wiesbaden: Steiner.
Daiber, H. (1999). ‘Rebellion gegen Gott: Formen atheistischen Denkens im frühen
Islam’. In F. Niewöhner and O. Pluta (eds.), Atheismus im Mittelalter und in der
Renaissance. Wiesbaden: Harrassowitz, 23–44.
Dēnkard. The Dinkard: The Original Péhlwi Text, the same transliterated in Zend charac-
ters, translations of the text in the Gujrati and English languages, a commentary and
a glossary of select terms by P.D.B. Sanjana. 19 vols. Bombay: Duftur Ashkara Press,
1874–1928.
Dhanani, A. (1994). The Physical Theory of Kalām: Atoms, Space, and Void in Basrian
Muʿtazilī Cosmology. Leiden: Brill.
Dunderberg, I. (2008). Beyond Gnosticism: Myth, Lifestyle, and Society in the School of
Valentinus. New York: Columbia University Press.
Ehlers, B. (1970). ‘Bardesanes von Edessa, ein syrischer Gnostiker’. Zeitschrift für Kir-
chengeschichte 81: 334–351.
Ephrem (Commentary). ‘Commentary on Genesis’. In E.G. Matthews and J.P. Amar
(trans.), St Ephrem the Syrian: Selected Prose Works. Washington, dc: Catholic Uni-
versity of America Press, 1994, 59–213.
Ephrem (Refutations). Ephraim’s Prose Refutations of Mani, Marcion, and Bardaisan:
Transcribed from the palimpsest bm Add. 14623. 2 vols. Ed. and trans. C.W. Mitchell.
London: Williams and Norgate, 1912 [repr. Piscataway, nj: Gorgias Press, 2008].
Eutychius (Saʿīd b. Biṭrīq) (Burhān). Kitāb al-Burhān. Ed. and trans. P. Cachia and
W.M. Watt. Louvain: Secrétariat du Corpus Scriptorum Christianorum Orientalium,
1960–1961.
Floridi, L. (2002). Sextus Empiricus: The Transmission and Recovery of Pyrrhonism. Ox-
ford: Oxford University Press.
Frank, R.M. (1974). ‘Notes and Remarks on the ṭabāʾiʿ in the Teaching of al-Māturīdī’. In
146 chapter 6
P. Salmon (ed.), Mélanges d’islamologie: volume dédié à la mémoire de Armand Abel

par ses collègues, ses élèves et ses amis. Leiden: Brill, 137–149.
Furlani, G. (1937). ‘Sur le Stoicisme de Bardesane d’Édesse’. Archiv Orientální 9: 347–
352.
Goldziher, I. (1920). Die Richtungen der islamischen Koranauslegung. Leiden: Brill.
Goodman, L.E. (1993). ‘Time in Islam’. In A.N. Balslev and J.N. Mohanty, Religion and
Time. Leiden: Brill, 138–162.
Grant, R.M. (1993). Heresy and Criticism: The Search for Authenticity in Early Christian
Literature. Louisville, ky: Westminster/J. Knox Press.
Guidi, M. (ed. and trans.) (1927). La lotta tra l’Islam e il manicheismo: un libro di Ibn
al-Muqaffaʿ contro il Corano confutato da al-Qāsim b. Ibrāhīm. Rome: R. Accademia
Nazionale dei Lincei.
Gutas, D. (1982). ‘Paul the Persian on the Classification of the Parts of Aristotle’s Philos-
ophy’. Der Islam 59: 231–267.
Haftād u sih millat, yā iʿtiqādāt-i madhāhib, risāla-yi dar farq-i Islām az āthār-i qarn-i
hashtum-i Hijrī. Ed. M.J. Mashkūr. Tehran: Muʾassasa-yi Maṭbūʿātī-yi ʿAṭāʾī, 1341/1962.
Hankinson, R.J. (1995). The Sceptics. London: Routledge.
Haq, S.N. (1994). Names, Natures and Things: The Alchemist Jābir ibn Ḥayyān and His
Kitāb al-Aḥjār (Book of Stones). Dordrecht: Kluwer.
Hippolytus (Refutatio). Refutatio omnium haeresium. Ed. M. Marcovich. Berlin: de Gruy-
ter, 1986 [trans. J.H. MacMahon, The Refutation of all Heresies. Edinburgh: t. &
t. Clark, 1868 (the chapter divisions are those of the edition)].
Horovitz, S. (1903). ‘Über den Einfluss des Stoicismus auf die Entwickelung der Philoso-
phie bei den Arabern’. Zeitschrift der Deutschen Morgenländischen Gesellschaft 57:
177–196.
Ibn Durayd (Ishtiqāq). al-Ishtiqāq. Ed. ʿA.-S.M. Hārūn. Cairo: Maṭbaʿa al-Sunniyya al-
Muḥammadiyya, 1958 [references are to the marginal pagination].
Ibn Ḥazm (Faṣl). al-Faṣl fī ʾl-milal waʾl-ahwāʾ waʾl-niḥal. Cairo: n.p., 1899–1903.
Ibn al-Jawzī (Talbīs). Talbīs Iblīs. Ed. M.M. al-Dimashqī. Cairo: Idārat al-Ṭibāʿa al-Munī-
riyya, 1928.
Ibn Kathīr (Tafsīr). Tafsīr. Cairo, n.d., n.p.
Ibn al-Malāḥimī (Muʿtamad). al-Muʿtamad fī uṣūl al-dīn. Ed. M. McDermott and W. Ma-
delung. London: Al-Hoda, 1991 [revised and enlarged edition by W. Madelung, Teh-
ran: Institute of Philosophy, 2012 (cited in that order, separated by a slash)].
Ibn al-Nadīm (Fihrist). al-Fihrist. Ed. R. Tajaddud. Tehran: Ibn-i Sīnā, 1971 [trans. B.
Dodge, The Fihrist of al-Nadīm: A Tenth-Century Survey of Muslim Culture. 2 vols. New
York: Columbia University Press, 1970].
Ibn Qutayba (Taʾwīl). Taʾwīl mukhtalif al-ḥadīth. Ed. M.Z. al-Najjār. Cairo: Maktabat al-
Kulliyyāt al-Azhariyya, 1966 [trans. G. Lecomte, Le Traité des divergences du hadīt
d’Ibn Qutayba (mort en 276/889). Damascus: Institut français de Damas, 1962].
Ibn al-Zubayr, Aḥmad b. al-Rashīd (Dhakhāʾir). Kitāb al-Dhakhāʾir waʾl-tuḥaf. Kuwait:

Dāʾirat al-Maṭbūʿāt waʾl-Nashr, 1959.
Israel of Kashkar (Unity). A Treatise on the Unity and Trinity of God by Israel of Kashkar
(d. 872): Introduction, edition, and word index by B. Holmberg. Lund: Plus Ultra,
1989.
Jacob of Sarug (Sermons). Quatre homélies métriques sur la creation. Ed. and trans.
K. Alwan. Louvain: Peeters, 1989.
Jaʿfar al-Ṣādiq (attrib.) (Tawḥīd). Kitāb al-Tawḥīd. Ed. M.ʿA.-R. Ḥamza. N.p.: Dār al-
Salām, 1329/1911.
al-Jāḥiẓ (Ḥayawān). Kitāb al-Ḥayawān. Ed. ʿA.-S.M. Hārūn. Cairo: Maktabat Muṣṭafā al-
Bābī al-Ḥalabī, 1938–1958.
al-Jāḥiẓ (Ḥujaj). Ḥujaj al-nubuwwa, in al-Jāḥiẓ, Rasāʾil. Ed. ʿA.-S.M. Hārūn. Cairo: Mak-
tabat al-Khānjī bi-Miṣr, 1979, 3: 222–281.
al-Jāḥiẓ (Radd). al-Radd ʿalā ʾl-naṣārā, in al-Jāḥiẓ, Rasāʾil. Ed. ʿA.-S.M. Hārūn. Cairo:
Maktabat al-Khānjī bi-Miṣr, 1979, 3: 302–351.
al-Jāḥiẓ (Tarbīʿ). Kitāb al-Tarbīʿ waʾl-tadwīr. Ed. C. Pellat. Damascus: Institut français de
Damas, 1955.
al-Jāḥiẓ (attrib.) (Dalāʾil). Kitāb al-Dalāʾil waʾl-iʿtibār ʿalā ʾl-khalq waʾl-tadbīr. Cairo:
Maktabat al-Kulliyyāt al-Azhariyya, 1987 [trans. M.A.S. Abdel Haleem, Chance or
Creation? God’s Design in the Universe, Reading: Garnet, 1995].
Job of Edessa (Treasures). Encyclopædia of Philosophical and Natural Sciences as Taught
in Baghdad about a.d. 817, or, Book of Treasures, by Job of Edessa. Ed. and trans.
A. Mingana. Cambridge: W. Heffer, 1935.
al-Juwaynī (Shāmil). al-Shāmil fī uṣūl al-dīn. Ed. ʿA.S. al-Nashshār et al. Alexandria:
Munshaʾāt al-Maʿārif, 1969.
al-Khayyāṭ (Intiṣār). Kitāb al-Intiṣār waʾl-radd ʿalā Ibn al-Rāwandī al-mulḥid. Ed. A.N.
Nader. Beirut: al-Maṭbaʿa al-Kāthūlīkiyya, 1957.
Kraus, P. (1934). ‘Zu Ibn al-Muqaffaʿ’. Rivista degli Studi Orientali 14: 1–20.
Kraus, P. (ed.) (1935). Mukhtār Rasāʾil Jābir b. Ḥayyān. Cairo: Maktabat al-Khānjī wa-
Maṭbaʿatihā.
Kraus, P. (1942). Jābir b. Ḥayyān. Contribution à l’histoire des idées scientifiques dans
l’Islam. Vol. ii: Jābir et la science grecque. Cairo: Impr. de l’Institut français d’archéo-
logie orientale.
al-Kulaynī (al-Kāfī). Al-Uṣūl min al-Kāfī. 8 vols. Ed. ʿA.A. al-Ghaffārī. Tehran: Dār al-
Kutub al-Islāmiyya, 1377–1381/1957–61.
Lactantius (De ira Dei). De ira dei. Vom Zorne Gottes. Ed. and trans. H. Kraft and
A. Wlosok. Darmstadt: Gentner, 1971.
Lammert, H. (1953). ‘Zur Lehre von Grundeigenschaften bei Nemesios’. Hermes 81: 488–
491.
Land, J.P.N. (1862–1875). Anecdota syriaca. Collegit, edidit, explicavit. 4 vols. Leiden:
Brill.
148 chapter 6
Langermann, Y.T. (2009). ‘Atomism and the Galenic Tradition’. History of Science 47:
277–295.
Lim, E. (1995). Public Disputation, Power and Social Order in Late Antiquity. Berkeley:
University of California Press.
Long, A.A., and D.N. Sedley (1987). The Hellenistic Philosophers. 2 vols. Cambridge:
Cambridge University Press.
Lucretius (De rerum natura). De rerum natura = On the Nature of Things. Ed. and trans.
W.H.D. Rouse, rev. M. Ferguson Smith. Cambridge, ma: Harvard University Press,
1982.
Macdonald, D.B. (1927). ‘Continuous Re-creation and Atomic Time in Muslim Scholas-
tic Theology’. Isis 9: 326–344.
Madelung, W. (2005). ‘An Ismaili Interpretation of Ibn Sīnā’s Qaṣīdat al-Nafs’. In T. Law-
son (ed.), Reason and Inspiration in Islam: Theology, Philosophy and Mysticism in
Muslim Thought. Essays in Honour of Hermann Landolt. London: I.B. Tauris, 157–168.
Maimonides (Guide). The Guide of the Perplexed. Trans. S. Pines. Chicago: University of
Chicago Press, 1963.
al-Malaṭī (Tanbīh). Kitāb al-Tanbīh waʾl-radd ʿalā ahl al-ahwāʾ waʾl-bidaʿ. Ed. S. Dedering.
Istanbul: Maṭbaʿat al-Dawla, 1936.
al-Maqdisī (Badʾ). Kitāb al-Badʾ waʾl-tārīkh = Livre de la création et de l’histoire. Ed. and
trans. C. Huart, Paris 1899–1919 [references are to the Arabic text].
al-Masʿūdī (Murūj). Murūj al-dhahab. 7 vols. Paris, 1861–1917 [ed. C. Pellat, Beirut:
al-Jāmiʿa al-Lubnāniyya, 1966–1979 (cited in that order, by page and paragraph
respectively; the volumes are identical)].
al-Māturīdī (Tawḥīd). Kitāb al-Tawḥīd. Ed. F. Kholeif. Beirut: Dar el-Machreq, 1970.
al-Māturīdī (Taʾwīlāt). Taʾwīlāt al-Qurʾān. Ed. B. Topaloğlu et al. Istanbul: Dār al-Mīzān,
2005–2010.
de Ménasce, J. (trans.) (1973). Le Troisième Livre du Dēnkart. Paris: C. Klincksieck.
Moses Bar Kepha (Hexaemeronkommentar). Der Hexaemeronkommentar des Moses bar
Kepha. Einleitung, Übersetzung und Untersuchungen von L. Schlimme. Wiesbaden:
Harrassowitz, 1977.
al-Muqammiṣ (ʿIshrūn). Dāwūd ibn Marwān al-Muqammiṣ’s Twenty Chapters (ʿIshrūn
Maqāla). Ed. and trans. S. Stroumsa. Leiden: Brill, 1989.
Nemesius of Emesa (Nature). On the Nature of Man. Trans. R.W. Sharpless and P.J. van
der Eijk. Liverpool: Liverpool University Press, 2008.
Nyberg, H.S. (1919). Kleinere Schriften des Ibn al-ʿArabī. Leiden: Brill.
Pedersen, N.A. (2004). Demonstrative Proof in Defence of God: A Study of Titus of Bostra’s
Contra Manichaeos. The Work’s Sources, Aims, and Relation to its Contemporary The-
ology. Leiden: Brill.
Philastrius (Diversarum). Diversarum Hereseon Liber. Ed. F. Marx. Vienna: F. Tempsky,
1898.
Photinus (Disputationes). ‘Disputationes Photini Manichaei cum Paulo Christiano’.

Patrologia Graeca 88 (1864): 529–578.
Pines, S. (1997). Studies in Islamic Atomism. Trans. M. Schwarz. Jerusalem: Magnes Press,
1997.
Possekel, U. (1999). Evidence of Greek Philosophical Concepts in the Writings of Ephrem
the Syrian. Louvain: Peeters.
Pretzl, O. (1931). ‘Die frühislamische Atomenlehre’. Der Islam 19: 117–130.
Pretzl, O. (1933). Die Streitschrift des Ġazālī gegen die Ibāḥīja, im persischen Text hrsg.
und übersetzt. Munich: Verlag der Bayerischen Akademie der Wissenschaften, 1933.
Procopius (Anecdota). Anecdota. Trans. H.B. Dewing. London/Cambridge, ma, 1969.
al-Qummī (Tafsīr). Tafsīr. 2 vols. Ed. Ṭ. al-Mūsawī al-Jazāʾirī. Beirut: Dār al-Surūr, 1991.
Ramelli, I. (2009). Bardaisan of Edessa: A Reassessment of the Evidence and a New
Interpretation. Piscataway, nj: Gorgias Press.
al-Rāzī, Fakhr al-Dīn (Tafsīr). al-Tafsīr al-kabīr [Mafātīḥ al-ghayb]. Tehran: Intishārāt-i
Asāṭīr, 1413/1992–1993.
Rudolph, U. (1995). ‘Kalām im antiken Gewand: Das theologische Konzept des Kitāb Sirr
al-Ḫalīqa’. In A. Fodor (ed.), Proceedings of the 14th Congress of the Union Européenne
des Arabisants et Islamisants. Budapest … 1988. Part 1. Budapest: Eötvös Loránd
University Chair for Arabic Studies & Csoma de Kőrös Society, Section of Islamic
Studies, 123–136.
Saadia Gaon (Amānāt). Kitāb al-Amānāt waʾl-iʿtiqādāt. Ed. S. Landauer. Leiden: Brill,
1880 [trans. S. Rosenblatt, The Book of Beliefs and Opinions. New Haven, ct: Yale
University Press, 1948].
Sextus Empiricus (Outlines). Outlines of Pyrrhonism. Ed. and trans. R.G. Bury. Cam-
bridge, ma: Harvard University Press, 1933.
al-Shahrastānī (Milal). Kitāb al-Milal waʾl-niḥal. Ed. W. Cureton. London: Printed for
the Society for the Publication of Oriental Texts, 1842–1846 [trans. D. Gimaret and
G. Monnot, Livre des religions et des sectes. Paris: Peeters, 1986 (which preserves the
paginations of both Badrān’s and Cureton’s editions)].
al-Shahrastānī (Nihāya). The summa philosophiae of al-Shahrastānī = Nihāyat al-aqdām
fī ʿilm al-kalām. Ed. with a summary translation A. Guillaume. Oxford: Oxford Uni-
versity Press, 1934.
Sorabji, R. (1988). Matter, Space, and Motion: Theories in Antiquity and their Sequel.
Ithaca, ny: Cornell University Press.
Stark, R., and R. Finke (2000). Acts of Faith: Explaining the Human Side of Religion.
Berkeley: University of California Press.
Stewart, C. (1991). ‘Working the Earth of the Heart’: The Messalian Controversy in History,
Texts, and Language to ad 431. Oxford: Oxford University Press.
Strohmaier, G. (1981). ‘Galen in Arabic’. In V. Nutton (ed.), Galen: Problems and Prospects.
London: Wellcome Institute for the History of Medicine, 197–212.
150 chapter 6
Stroumsa, S. (1999). Freethinkers of Medieval Islam: Ibn al-Rāwandī, Abū Bakr al-Rāzī
and their Impact on Islamic Thought. Leiden: Brill.
al-Ṭabarī (Tārīkh). Tārīkh al-rusul waʾl-mulūk. Ed. M.J. de Goeje et al. Leiden: Brill, 1879–
1901.
al-Tawḥīdī (Akhlāq). Akhlāq al-wazīrayn. Ed. M. al-Ṭanjī. Beirut: Dār Ṣādir, 1992.
al-Tawḥīdī (Imtāʾ). Kitāb al-Imtāʾ waʾl-muʾānasa. Ed. A. Amīn and A. al-Zayn. Cairo:
Lajnat al-Taʾlīf waʾl-Tarjama waʾl-Nashr, 1939–1944.
al-Tawḥīdī (Muqābasāt). al-Muqābasāt. Ed. M.T. Ḥusayn. Baghdad: Maktabat al-Irshād,
1970.
Teixidor, J. (1997). ‘Les Textes syriaques de logique de Paul le Perse’. Semitica 47: 117–137.
Theodore Bar Koni (Liber). Liber scholiorum (Seert version). Ed. A. Scher. Paris: E Typo-
grapheo Reipublicae, 1910, 1912 [trans. R. Hespel and R. Draguet. Louvain: Peeters,
1981–1982].
Theodoret (Providentia). On Divine Providence [= De providentia]. Trans. T. Halton. New
York: Newman Press, 1988.
Vajda, G. (1938). ‘Les Zindīqs en pays d’Islam au début de la période abbaside’. Rivista
degli Studi Orientali 17: 173–229.
van Bladel, K. (2009). The Arabic Hermes: From Pagan Sage to Prophet of Science. Oxford:
Oxford University Press.
van den Ven, P. (ed. and trans.) (1962). La Vie ancienne de S. Syméon Stylite le Jeune (521–
592). 2 vols. Brussels: Société des bollandistes.
Van Ess, J. (1966). Die Erkenntnislehre des Aḍudaddīn al-Īcī: Übersetzung und Kommen-
tar des 1. Buches seiner Mawāqif. Wiesbaden: Steiner.
Van Ess, J. (1967). ‘Ḍirār b. ʿAmr und die “Cahmiya”, i’. Der Islam 43: 241–279.
Van Ess, J. (1968). ‘Skepticism in Islamic Religious Thought’. al-Abḥāth 21: 1–18.
Van Ess, J. (1991–1997). Theologie und Gesellschaft im 2. und 3. Jahrhundert Hidschra: Eine
Geschichte des religiösen Denkens im frühen Islam. Berlin: de Gruyter.
Walker, P.E. (1993). Early Philosophical Shiism: The Ismaili Neoplatonism of Abū Yaʿqūb
al-Sijistānī. Cambridge: Cambridge University Press.
Weisser, U. (1980). Das ‘Buch über das Geheimnis der Schöpfung’ von Pseudo-Apollonios
von Tyana. Berlin: de Gruyter.
Wolfson, H.A. (1976). The Philosophy of the Kalam. Cambridge, ma: Harvard University
Press.
al-Yaʿqūbī (Mushākala). Mushākalat al-nās li-zamānihim wa-mā yaghlibu ʿalayhim fī
kull ʿaṣr. Beirut: Dār al-Kitāb al-Jadīd, 1962.
al-Yaʿqūbī (Tārīkh). Tārīkh. 2 vols. Ed. M.T. Houtsma. Leiden: Brill, 1883.
chapter 7
Post-Colonialism in Tenth-Century Islam* 2
Carl-Heinrich Becker, the scholar who is commemorated in these lectures,

wrote about the Arabs as colonisers, comparing them with modern colonial
powers such as the British, and he would probably have been interested in post-
colonialism, too, if he had he lived to see it.1 In a way you could say that he did
live to see it, for the term “post-colonialism” is often taken to refer to the culture
of peoples affected by colonial government from the very moment they were
conquered, not simply from their recovery of independence.2 But it was only
after the collapse of the colonial empires, in the wake of the Second World War,
that the concept of post-colonialism acquired prominence, and Becker died in
1935. Even if he had been familiar with the concept, moreover, the fact that
he saw the parallel between the Arab and the modern European empires does
not necessarily mean that he would have deemed it appropriate to analyse the
result in terms of post-colonialism. The wisdom of applying a concept referring
to a modern experience to the tenth-century Muslim world may well strike
many readers of this paper as questionable, too.
The Two Rāzīs 3
For the moment I shall leave such readers to their scepticism, for I should
like to start by discussing something completely different, namely a public
disputation which took place around 920 or 930 in Rayy, the medieval precursor
* I should like to thank Prof. Lawrence Conrad for inviting me to deliver the Becker lecture. I
am also indebted to audiences in Cambridge, Napoli, Berkeley, Paris and above all Hamburg
for their responses to different versions of that lecture, and to Sarah Savant for most helpful
comments on the penultimate draft. [Ed.: This article is reproduced in the form in which it
originally appeared in Der Islam, with the exception of silent correction of minor typograph-
ical or editorial errors and a few bracketed editorial interventions.]
1 C.H. Becker, “Die Araber als Koloniatoren”, in his Islamstudien, ii (Leipzig, 1932), esp. 2f.
For a more recent invocation of the similarity between the Arab and the European con-
quests, see A. Hannoum, Colonial Histories, Post-Colonial Memories: the Legend of the Kahina
(Portsmouth, 2001), ch. 1, esp. 5, 9. Cf. also below, n. 31.
2 Thus for example B. Ashcroft, q. Griffiths and H. Tiffin, The Empire Writes Back2 (London,
2002), 2.

152 chapter 7
of modern Tehran.3 The two participants in the debate were both called Rāzī.
One was Abū Bakr al-Rāzī (d. 313/925 or 323/935), the famous physician and
philosopher who was known in medieval Europe as Rhazes. The other was Abū
Ḥātim al-Rāzī (d. 322/934), a missionary on behalf of Ismāʿīlism, the radical Shīʿī
movement which had begun some 50 years before the disputation took place.
We know about the disputation because the Ismāʿīlī missionary wrote a book
refuting the philosopher’s claims, both as presented on that occasion4 and as
recorded in a lost book (or books) of his.5
The disputation was about revealed religion—religion in the sense of a mes-
sage sent down by God to mankind through a specially selected human being, a
4 prophet. Was there any such thing? The philosopher de|nied it. More precisely,
he said that there was no such thing as prophets. The idea was not compatible
with divine wisdom and mercy in his view. If God wanted to communicate the
truth to mankind, why should He only tell one single person? Why should He
favour one man over all others?6 It was a well-known source of conflict and war-
fare, he said, stressing the role of religion as a provoker of bloodshed.7 Besides,
3 The debate is said by al-Kirmānī (see the following note) to have taken place in Rayy in the
presence of the amīr Mardāwīj. Since Mardāwīj only occupied Rayy in 318/930, this clashes
with al-Bīrūnī’s information that Abū Bakr al-Rāzī died in 313/925. Maybe al-Rāzī only died
in 323/935, as other authorities say, or maybe the amīr was Aḥmad b. ʿAlī (d. 311/923f.) rather
than Mardāwīj, as suggested by S.M. Stern, “The Early Ismāʿīlī Missionaries in North-West
Persia and in Khurāsān and Transoxania”, in his Studies in Early Ismāʿīlism (Jerusalem and
Leiden, 1983), 202, cf. also 196, 198.
4 Abū Ḥātim al-Rāzī, Aʿlām al-nubuwwa, ed. Ṣ. al-Ṣāwī (Tehran, 1977), 3–28; I refer to the pages
because the absence of the chapter, section, and paragraph numbers from the running heads
makes it difficult to locate passages by means of them. The parts relating to Abū Bakr al-Rāzī
were first edited by P. Kraus in “Raziana ii”, Orientalia 5 (1936), 35–56, 358–378; it was re-edited,
this time including al-Kirmānī’s account of the debate, by P. Kraus, al-Rasāʾil al-falsafiyya li-
Abī Bakr … al-Rāzī, i (no sequel published) (Cairo, 1939). For an English translation of the
first chapter of the Aʿlām, which contains the disputation, see L.E. Goodman, “Rāzī vs Rāzī—
Philosophy in the Majlis”, in H. Lazarus-Yafeh, M.R. Cohen, S. Somekh and S.H. Griffith (ed.),
The Majlis: Interreligious Encounters in Medieval Islam (Wiesbaden, 1999) (based on the text
as given in Kraus, Rasāʾil), 84–107.
5 The title of the book (cf. Aʿlām, 28.1) is not given. Kraus and Pines identify it as Fī ʾl-nubuwwāt,
also known as Naqḍ al-adyān; ei1, s.v. “al-Rāzī”; cf. al-Bīrūnī, Risāla fī fihrist kutub Muḥammad
b. Zakariyyāʾ al-Rāzī, ed. P. Kraus (Paris, 1936), no. 173. For another possibility, see below, n. 14.
6 Abū Ḥātim, Aʿlām, 3 (Goodman, 85). As noted by S. Stroumsa, Freethinkers of Medieval Islam
(Leiden, 1999), 95n, he uses qawm to mean “certain individuals” rather than “some people”
(cf. Abū Ḥātim’s response at 8.9).
7 Abū Ḥātim, Aʿlām, 3 f. (Goodman, 85 f.), 181 ff., 186; Nāṣir-i Khusraw citing Rāzī’s Theology in
Kraus, Rasāʾil, 177; in Stroumsa, Freethinkers, 106.
post-colonialism in tenth-century islam 153
it was not easy for a single man to persuade the rest of mankind that he, and he
alone, possessed the truth. Why should God use so cumbersome a method?8 It
struck the philosopher Rāzī as much more plausible that God in His wisdom
and mercy should have given all humans equal access to the truth, by endow-
ing them with innate knowledge of what was good and bad for them, in respect
of this world and the next alike, just as he had given animals innate knowledge
of what they needed to know.9 All humans should engage in critical investiga-
tion to the best of their ability, for it was only by philosophical study that one
could reach salvation—which he envisaged as release from this world.10 His
own philosophy was certainly a religion, but it was a religion based entirely on
reason. As he saw it, the revealed variety only gave you lies and fairy tales (al-
akādhīb wa-l-khurāfāt).11 All the different revelations claimed to be true with
reference to the same arguments, and they all contradicted one another, and
indeed themselves as well, as he demonstrated with merciless criticism of the
scriptures.12 People only accepted them as true because they took things on
trust from their leaders, from whom they had heard them for so long that these
things had become second nature to them.13 The miracles supposedly per-
formed by the would-be prophets were mere juggleries and sleights of hand, in
so far as people had actually seen them.14 The so-called prophets were people | 5
who caused discord and bloodshed because demons had appeared to them in
the guise of angels and persuaded them that God had chosen them, he said,15
presumably adopting mythological language for didactic purposes, but show-
ing that he saw the prophets as deluded people rather than swindlers. As for the
religious scholars, they were mere “goatbeards”—men who impressed unedu-
cated people with their long beards and white clothes and who transmitted
inconsistent material from past authorities, prohibiting critical investigation,
and branding every opponent as an unbeliever who could be freely killed.16
8 Aʿlām, 181.
9 Ibid., 3 f. (Goodman, 86), 181, 183; cf. also 274.2.
10 Aʿlām, 12 f. (Goodman, 91).
11 Aʿlām, 13.5 (Goodman, 92, on the sharāʾiʿ of the prophets), 32.7 (on the doctrines of
religious scholars).
12 Aʿlām, 69 ff., 171.
13 Ibid., 31 f., 171.
14 Ibid., 192. He wrote a book on this subject (Fī ḥiyal al-mutanabbiyyīn, also known as
Makhāriq al-anbiyāʾ, in Bīrūnī, Fihrist, no. 174), and Abū Ḥātim could be drawing on it
here. Conceivably, he had it as part of the book referred to above, n. 5.
15 Nāṣir-i Khusraw citing Rāzī’s Theology in Kraus, Rasāʾil falsafiyya, 177; Stroumsa, Free-
thinkers, 106.
16 Abū Ḥātim, Aʿlām, 31 f.
154 chapter 7
The Ismāʿīlī Rāzī was horrified by all this. Prophets were real to him, and
he vehemently refutes the philosopher’s assertions. But one soon notices that
there is something peculiar about his view of prophets, too. He sees them
first and foremost as communal leaders. Moses and Jesus were the men best
endowed in their time with the qualities that an imam needs to govern people
in this world and the next, he says; of course this was even truer of Muḥam-
mad, whose power and extensive conquests he vaunts.17 Even those who deny
their prophetic status and their miracles ought to accept that they were men of
superior intelligence and ability, he says, sounding rather like a modern histo-
rian.18 Prophets discipline people and keep them in order with their wondrous
governance (siyāsa ʿajība).19 They are needed because people are equal only in
respect of the nutritional and reproductive needs they share with other ani-
mals, not in respect of the knowledge they require for moral and civilized lives
in this world and salvation in the next.20 Their different endowments in this
regard are plain for everyone to see: this is why some have to act as teach-
ers and leaders to others. Abū Ḥātim clearly sees himself as having refuted
his opponent with this statement, but the philosopher did not of course dis-
6 agree: all he denied was that such teachers had superhuman knowl|edge.21
Like so many heirs to the ancient Near Eastern tradition, however, the Ismāʿīlī
Rāzī found it impossible to think of religion, morality and culture as some-
thing that humans had evolved on their own: even medicine and other sciences
owed their existence to revelation in his view, not to human use of innate gifts,
as the philosopher claimed. Once the human need for teachers and leaders
had been established, the need for prophets thus followed automatically as
he saw it.22 He also argues on the basis of his own premises when he tacitly
17 Ibid., 89.
18 Ibid., 89.17, 90.10. A century later another missionary (al-Muʾayyad) said much the same
in response to Ibn al-Rāwandī’s K. al-zumurrud: even if the deniers of prophethood were
right, they ought to speak of the prophets with respect, given the latter’s ability to govern
people and keep order. See P. Kraus, “Beiträge zur islamischen Ketzergeschichte”, Rivista
degli Studi Orientali 14 (1933), 109; Stroumsa, Freethinkers, 139.
19 Abū Ḥātim, Aʿlām, 8 f. (Goodman, 89 f.).
20 Abū Ḥātim, Aʿlām, 6 f. (Goodman, 88), 183 ff.
21 Abū Bakr al-Rāzī does come across as an “epistemological democrat”, as Goodman puts
it (“Philosophy in the Majlis”, 104), but when he says that humans are equal, he means
that they all have the same generic abilities, not that these abilities are evenly distributed
among them or that humans do not learn from one another. Abū Ḥātim’s presentation
does not allow for subtle distinctions, however.
22 For the Ismāʿīlī Rāzī’s vehement denial that the philosophers have developed the sciences
on their own, see Aʿlām, 273 ff.
assumes leadership to rest on knowledge, so that the political and social hier-
archy of a particular community reflects (or ought to reflect) the distribution
among its members of knowledge originating from above.23 What prophets did
in his view was to establish such a hierarchy. God was very wise to send the
truth to just one man, so that people had to defer to others in order to get
access to it: hierarchy and subordination were what the religious law was all
about.
Religion and Political Organization
What is so striking about the debate is that both the Rāzīs associated prophets
with power and war. Up to a point, of course, this is as might be expected, for
Islam owed its existence to the fact that Muḥammad had established a polity
in Medina. He had brought a law and united the Arabs in obedience to it, and
this had indeed involved warfare, which had continued when the Arabs pro-
ceeded to conquer the world outside Arabia. Tenth-century Muslims generally
assumed their own case to be paradigmatic: all prophets were founders of poli-
ties in their view, or rather this was true of all the prophets who brought laws.
In explanation of this idea they said that human beings were social (madanī)
animals who depended on one another for their many needs, meaning that
they had to live together, but that they were also anti-social animals given to
ruthless competition and fighting, meaning that they would perish if they were
left | alone: they needed a higher authority, a neutral outsider, to set the rules 7
of the game for them, and to enforce them. God in His mercy set the rules for
them in the form of a law; a prophet would transmit the law to human beings
and found a polity in which it could be enforced; and after the death of the
prophet, other rulers would take over the task of maintaining the polity and
ensuring that the law was maintained. What God revealed, in short, was first
and foremost a moral order, shaped as a law, and what the prophet created was
a polity within which people could live together in safety and trust, by adhering
to the shared rules.24
This view of prophets was particularly popular with rationalizing theolo-
gians (mutakallims), philosophers, and Shīʿīs, but practically all educated Mus-
lims knew that revealed religion was first and foremost a blueprint for commu-
23 Cf. Goodman, “Philosophy in the Majlis”, 103.

24 Cf. P. Crone, Medieval Islamic Political Thought (Edinburgh, 2004; American title God’s
Rule: Government and Islam: Six Centuries of Medieval Islamic Political Thought), ch. 17.
156 chapter 7
nal organization and that man would go to rack and ruin without it, in this
world and the next alike. It enabled them to think about the socio-political
functions of religion in very sophisticated terms. What modern sociologists
call the “latent functions” of religion was mostly perfectly manifest to them.
Religion existed for the organization of collective affairs, they said; it created
communities by enjoining obedience to higher powers, it enabled humans to
internalize moral codes and thus to counteract the destructive effects of indi-
vidual desire (hawā), keeping them on the straight and narrow by a combi-
nation of carrot and stick—the promise of Paradise and the threat of Hell. It
stabilized government by legitimating rulers, increasing people’s respect for
them, and so on.25 In short, revealed religion and societal organization were
two sides of the same coin.
The two Rāzīs took this view of prophethood for granted. But they went
further than that, for they thought that the law brought by a prophet was
only about communal order. This was where they took off into heresy. To
the philosopher Rāzī, the so-called revelation was simply politics in disguise:
the so-called prophets claimed that their warfare and (by implication) the
political activities leading to it were ordered by God, but God had nothing to do
with mundane affairs. As he saw it, the truth was elevated above such affairs,
and accessible through the intellect which all humans shared, not through
membership of this or that community, and it was not a prescription for order
in this world at all, but on the contrary something that purified your soul of
8 worldly concerns and | caused you to be released from this world. Genuine
religion was spiritual. Had the philosopher Rāzī lived today, he would have been
a secularist—an adherent of the view that religion is an individual matter and
must be kept out of public affairs.
To the Ismāʿīlī Rāzī, on the other hand, revealed religion was genuine e-
nough, not a mere mask for political interests: organizing people was exactly
what God meant His prophets to do.26 The law they brought just was not the
highest form of religion. There was a spiritual realm above it. For religion had
two levels, a higher and a lower or, as the Ismāʿīlīs preferred to say, an inner
and an external one. It was only the external, overt and literal meaning of the
revelation that concerned communal order. At the level of the literal mean-
ing of the revelation (or law: sharʿ), religion was indeed mundane, and also
changeable: every scriptural prophet brought a new religion/law abrogating
that of his predecessor, reflecting the new circumstances of his time. But the
25 Ibid., 265 f., 285, 393.

26 Abū Ḥātim, Aʿlām, e.g. 108.15.
apparent contradictions between their messages to which this gave rise did
not affect the inner meaning (al-bāṭin) of the revelation, which was eternal,
unchanging, the same for all human beings anywhere. At this level the revela-
tion had nothing to do with communal organization. On the contrary, it was
totally divorced from the particulars in which we live in the here and now,
totally unmired by matter, wholly spiritual, just as the philosopher said. The
philosopher’s mistake lay in his failure to understand that there were two sides
to religion. If the revealed laws were not from God and the ẓāhir were all there
was to them, then he would be right, Abū Ḥātim says, but they were indeed
from God, and there were spiritual meanings behind their literal wording.27
One found these meanings by treating the literal meaning of the revelation
as parables, symbols and allegories pointing to higher things, relying for guid-
ance here not on the prophets, but rather on the imams who followed them.
You could not live properly in this world without the law and its socio-political
prescriptions, but otherwordly salvation lay entirely in the inner spiritual mes-
sage.28
There was a further twist to Ismāʿīlī doctrine. The Ismāʿīlīs were awaiting a 9
mahdī, a messiah. He was due to come any moment, and he would be the last
prophet. Like the earlier prophets, he would abrogate the law of his predecessor,
but unlike them, he would not bring a new one. Mankind would live by the
inner spiritual meaning alone, without all the limitations imposed on us by our
incarceration in gross bodies. The sociopolitical and legal apparatus associated
with the law would wither way. There would be no more organized religion, no
more hierarchy, and also no more division of mankind into different polities.
The inner spiritual meaning would be directly accessible to all of us. Then we
would indeed have equal access to the truth. And then there would be no more
27 Ibid., 114.11; similarly 104.7, 113.12, 115.3.

28 Modern Ismāʿīlīs find it difficult to accept that their distant forebears denied the sav-
ing role of the law, but Abū Ḥātim makes a clear distinction between the ẓāhir, which
people must be forced to accept for reasons of social and political order, and the inner
meanings which they are free to seek for themselves and in which their salvation lies
(mā fīhi najātuhum min al-maʿānī allatī taḥta sharāʾiʿihim al-ẓāhira) (Aʿlām, 111f.; simi-
larly 110.13). Compare also the account of the Ismāʿīlīs in al-Nawbakhtī and Qummī, com-
posed in the 280s/890s, in which the Ismāʿīlīs claim that “the whole of the Book and
the Sunna, which outwardly contain obligations imposed by God on men, are parables
expressing inner meanings: it is these inner meanings which must one act upon in order
to be saved. If one follows the outward meaning, which consists of prohibitions, one per-
ishes” (W. Madelung, “The Account of the Ismāʿīlīs in Firaq al-Shīʿa”, in Stern, Studies,
52).
158 chapter 7
communal divisions and war. Mankind would be united in what would amount
to a return to Adam’s Paradise. But the philosopher was mistaken in thinking
that humans had been made that way.29
In other words, both Rāzīs denied that salvation lay in the revealed law: it
lay in reason according to the one, in the inner allegorical meaning of the law
according to the other. One Rāzī said that prophets could not save you, meaning
that you had to seek the truth yourself; the other said that prophets could not
save you on their own, meaning that you had to turn to the imams, the religious
leaders from the Prophet’s family, for elucidation of the inner meaning. One
Rāzī said that prophets did not actually exist, the other said that they did,
but that the era of prophets was about to come to an end: either way, they
saw the highest truth as lying beyond prophethood. And the two Rāzīs were
not alone. Doubts about the existence of prophethood (and other aspects of
revealed religion) are common in the tenth-century literature, and Ismāʿīlism
was spreading like wildfire.
Post-Colonialism
Why did people have such strange ideas? What was going on? This, at last, is
where I get to the subject announced in the title: post-colonialism.
10 Like the author and the reader of this article, the two Rāzīs were living in a
society dominated by the cultural after-effects of a great imperial expansion.
In their case as in ours, the after-effects owed their character to a combination
of three basic facts. First, the conquerors had passed on their key beliefs and
values to the conquered peoples: just as the elites that took over government
from the French and the British were westernized, so the elites that took over
from the Arabs were islamized, and in both cases these new elites presided
over further westernization/islamization of the people below them. Secondly,
the conquered peoples nonetheless retained their own identity, invariably in
the case of the European expansion, and sometimes in that of the Arabs: just
as the Indians under British rule did not become Englishmen even when they
were fully anglicized, so the Iranians under Arab rule did not become Arabs
even when they were fully islamized (whereas converts in Egypt and the Fertile
Crescent eventually did). Thirdly, the empire broke up without putting an
end to the close relationship between the former rulers and subjects. Just as
29 For a concise account of Ismāʿīlī doctrine, see for example H. Halm, Die Schia (Darmstadt,
1987; tr. J. Watson, Shīʿism, Edinburgh, 1991), ch. 4.
the West and its former colonies could not simply forget about each other
when the Western powers withdrew, so the Arabs and the peoples they had
conquered could not simply revert to the pre-conquest situation when the
caliphate collapsed. In both cases there was a political divorce, but (for very
different reasons) not a cultural one. In both cases the parties continued to live
together, on new terms, with much recrimination and uncertainty and much
effort to find new standards acceptable to both sides. It is this tense relationship
that I like to call post-colonialism. The term seems more commonly to be used
with reference to the culture and outlook of the conquered peoples during and
after their political subjection,30 but the empire evidently affects both sides,
and nobody would talk about post-colonialism today if it were not for the
continuing relationship: the term was coined to articulate a grievance against
the former bearers of empire by people writing in the latter’s language and
sharing their conceptual world. In short, post-colonialism as I see it refers to
a situation in which the conquered peoples have adopted the key beliefs and
values of their conquerors without having been being absorbed by them in
ethnic terms, and also without being able to ignore the former conquerors
when they cease to be ruled by them.
Now let me give you a bird’s-eye view of how the Muslims got themselves 11
into the post-colonial relationship.
The Arabs began their expansion in the 630s and had a major empire a
mere 30 years later. This was a colonial empire of the classic type, with a sep-
arate metropole (Arabia) and periphery (Syria, Egypt, Iraq and Iran). But for
all the well-known similarities between the ports of the British and the gar-
rison cities of the Arabs,31 the Arab empire was terrestrial rather than mar-
itime, so the distinction between metropole and periphery did not remain
sharp for long; and since the metropole was also considerably less well devel-
oped than the conquered lands, it soon lost its politically dominant role. In
41/661 “Muʿāwiya placed his throne in Damascus and refused to go to the
30 Usage varies enormously. Sometimes, colonialism and post-colonialism seem to mean

little more than domination and exploitation of a capitalist type, making them terms
of abuse rather than analytical tools (a fate suffered by all terms of great contemporary
political relevance).
31 Both were located on the edge of the lands they controlled to facilitate easy retreat,
via the desert in the case of the Arabs, via the sea in the case of the British, and both
accommodated a population that had no desire to mix with the natives. The comparison
is so old that I do not know who first came up with it. See also N. AlSayyad, “The Islamic
City as a Colonial Enterprise”, in N. AlSayyad (ed.), Forms of Dominance: on the Architecture
and Urbanism of the Colonial Enterprise (Aldershot, 1992).
160 chapter 7
seat of Muḥammad”, as a Christian observer put it.32 It was a fateful step—

somewhat as if the capital of the British empire had been moved from London
to Cairo.33
The capital remained in Syria down to 132/750, when the Umayyad caliphate
was toppled by revolutionaries from eastern Iran. Contrary to what many
people expected, the revolutionaries did not chase out the Arabs or restore
the Persian empire.34 On the contrary, they enthroned another Arab dynasty.
But they moved the capital to Iraq, where the Persian emperors had also had
their centre, so now it was somewhat as if the capital of the British empire was
being moved to Delhi, where the Mughal emperors had resided (though Iraq
was of course less alien to the Arabs than India to the British). The bureaucrats
recruited in Iraq were all natives, usually from families who had served under
12 the Persians;35 and | the revolutionaries themselves were a mixed bunch of
Arabs and Iranians,36 so the ruling elite was losing its Arab ethnicity. And Spain
seceded in 756, so the empire was also beginning to break up. Most of it was kept
together for another hundred years. But by the 860s it was fast disintegrating.37
There still was an Arab caliph. In fact, there continued to be one all the way
down to 1258, but he was becoming ceremonial. Real power had passed to
others, both at the centre and in the provinces. By the time of the disputation
in Rayy the whole of Iran was ruled by Iranians again.
In short, the ninth century was a period of decolonization, and by the tenth
century the process was complete. It has to be stressed that unlike the British
and the French, or for that matter the Mongols in China, the Arabs were never
forced to withdraw physically. On the contrary, they stayed on for long enough
to arabize the indigenous peoples of Syria, Egypt and Iraq. This is important, for
this was one way in which the relationship between the two parties continued:
not by economic ties, globalization, or immigration by the conquered peoples
to the old metropole, but rather by the conquerors’ bequeathing their identity
32 Maronite chronicler in R.G. Hoyland, Seeing Islam as Others Saw it (Princeton, 1997), 136.
33 No Muslim comment comparable to that of the Maronite chronicler seems to survive, but
the change of capital is clearly one factor behind the conviction that Muʿāwiya’s accession
marked the end of the rightly guided caliphate.
34 For these expectations, see the references in P. Crone, “The ʿAbbāsid Abnāʾ and Sāsānid
Cavalrymen”, Journal of the Royal Asiatic Society, Third Series, 8 (1998), 12, n. 101.
35 The closest to a prosopography is D. Sourdel, Le Vizirat ʿabbāside (Damascus, 1959–1960).
36 S.S. Agha, The Revolution which Toppled the Umayyads: Neither Arab nor ʿAbbāsid (Leiden,
2003), part iii; cf. also Crone, “ʿAbbāsid Abnāʾ”, 11 f.
37 On which, see H. Kennedy, “The Decline and Fall of the First Muslim Empire”, Der Islam
81 (2004) (the first Becker lecture).
to a substantial segment of the conquered population. As rulers the Arabs

lost out, but as colonists they not only stayed on but hugely expanded their
originally tiny ranks.
The other way in which the relationship continued was by the converts
having adopted an Arab prophet and scripture, so that their relationship with
their own ancestral tradition had permanently changed. This was the crucial
factor. They could not get the conquerors out of their cultural system even when
they resumed political control of themselves.38 Their inability to do so would
probably have sufficed to produce a reaction among them even if the Arabs
had reverted to their pre-conquest insignificance when the caliphate broke up.
Perhaps it would have done so even if the Arabs had disappeared altogether, as
the Romans so kindly did | after the collapse of their empire in the West. But 13
disappearance was not on the cards, since the Arabs retained their homeland
intact and stayed on for long enough outside the Arabian peninsula to generate
a substantial population of neo-Arabs. It was the converts who had not (or
not yet) been arabized, and above all the Iranians, who found themselves in
a situation resembling that of the post-colonial world today.
The reader may object that it is absurd to speak about decolonization and
post-colonialism in a situation in which the colonists stayed on, and so in a
sense it is. In fact, there is something inept about the entire modern termi-
nology. A colony properly speaking is a settlement on foreign soil of people
who remain culturally or politically connected with their homeland, like the
Greeks in Anatolia, the Romans in their newly conquered lands, the Arabs in
their garrison cities or the British in Rhodesia.39 Decolonization thus ought
to mean the removal of the foreign settlers. There was no decolonization in
that sense in the Muslim case, except much later, in the Iberian peninsula,
where the Arabs had lived for so long by the time they were expelled that
one can hardly call them colonists anymore. Nor was there any decoloniza-
tion in that sense in British India, since there were hardly any colonists there.
38 Unlike Jesus, who lost his Jewish identity when he was adopted by the gentiles, Muḥam-
mad remained an Arab, just as the Qurʾān remained in Arabic and the sanctuary remained
in Arabia. The fact that the Arabs had arrived as conquerors had given them a control
over their own religion vis-à-vis the non-Arab converts that the Jews who disseminated
the Jesus-movement among the gentiles had not enjoyed over theirs.
39 For a narrower definition, see M.I. Finley, “Colonies—an Attempt at a Typology”, Trans-
actions of the Royal Historical Society 26 (1976), according to whom the settlement is only
a colony if the continuing relationship is political, and then only if it is one of depen-
dence. This eliminates most of what is normally called colonies, including the Greek ones
(stressed at 173 f.).
162 chapter 7
But nowadays a colony has come to mean a foreign dependency, with or with-
out colonisation. A colony is distinguished from a protectorate or a sphere
of influence: the words are about degrees of control, not about settlement,
and the entity they designate is no longer the community planted on for-
eign soil but rather the much larger area it controls as the representative of
an imperial power. So decolonization has come to mean the end of empire,
and post-colonialism is a word for the cultural state of the indigenous peo-
ples affected by this empire or, as I prefer to use the term, for the cultural
relationship between the two parties brought together by an empire. It is in
that sense there was both decolonisation and post-colonialism in the Muslim
case.
The reader may also object that if the modern terminology is inept, there is
no point in using it, and that on the contrary it might be better to apply termi-
nology derived from the Arab caliphate (or some other imperial experience in
14 the past) to our modern situation. So indeed it might. | But the modern termi-
nology has the advantage of being known to everyone and conjuring up a famil-
iar world complete with a sense of the main actors, their ways of interaction, the
feelings they voice and the sheer variety and complexity of the relationships,
all of which tends to get lost when the fullness of experience possessed by the
living is reduced to a couple of pages in a handful of ancient sources. The world
encountered in the Muslim sources is not our own, but it has strong similarities
with ours because in some crucial respects it was shaped by similar develop-
ments, and historians have a habit of focusing on what they recognize best in
the past. In retrospect, it may look as if each generation is rewriting history in
its own image, but what is actually happening is that the past and the present
are allowed to illuminate each other, often in ways that permanently change
our perceptions of the historical events in question even when the next gener-
ation deems the recognition to have been exaggerated or debatable. It is in the
hope of providing such illumination that the comparison of the Arab past and
our own present is offered here.
Shuʿūbiyya
With this apologia let me return to the Arabs. The cultural effects of the devel-
opment sketched above manifested themselves soon enough, in two separate
stages, the Shuʿūbī movement before the break-up of the empire and what we
may call the tenth-century crisis after it.
The Shuʿūbī movement was a literary attack on the Arabs and their heritage
by assimilated natives who were heard with increasing frequency after the rev-
olution of 750.40 The natives in question were mostly Iranian Muslims who had
risen high in the conquerors’ society. Typically, they occupied high bureaucratic
or academic positions in the capital, where, like many articulate descendants of
the victims of colonialism today, they were active participants in what is nowa-
days called the production of hegemonic culture. They always wrote in Arabic,
addressing themselves to the bearers of empire and assimilated natives, never
to natives back in their original homes; and what they wrote often reflected
the prejudice to which their fathers and grandfathers had been exposed under
Arab rule: sheer anger is prominent in their statements, as are horror stories
of the | ways in which the Arabs had maltreated converts to their faith.41 For 15
the Arab conquerors had regarded themselves as ethnically superior, much as
did the Europeans. A native who adopted the culture of the British, including
the scientific and other “progressive” beliefs which the British saw as their dis-
tinguishing feature and in terms of which they explained their own success,
did not thereby become a full member of British society (nor did a native con-
vert to Christianity, whatever his degree of assimilation).42 Rather, he would
be seen as a “westernized Oriental gentleman” (or “wog” for short). Similarly, a
native who adopted the culture of the Arabs, including the monotheistic reli-
gion which the Arabs saw as their distinguishing feature and in terms of which
they explained their own success, did not thereby become a full member of
Arab society. Rather, he became a mawlā, “client”, a legal term which came to
be widely used in the broad sense of “assimilated native”. To be a mawlā was
to be someone who had lost his position in his native society without being
fully accepted into the new one; it was to have one’s career circumscribed and
to endure regular humiliation by people less able and intelligent than one-
self, because of prejudice, not a legitimate hierarchy: this is what had made
it unbearable.43
40 In general, see I. Goldziher, Muhammedanische Studien, i (Halle, 1889), chs. 3–5; S. Ender-
witz, Gesellschaftlicher Rang und ethnische Legitimation (Freiburg, 1979).
41 See for example Rāghib al-Iṣbahānī, Muḥāḍarāt al-udabāʾ (Beirut, 1961), i, 347; Ibn ʿAbd
Rabbih, al-ʿIqd al-farīd, ed. A. Amīn, A. al-Zayn and I. al-Abyārī (Cairo, 1950–1953), iii,
413 f. = B. Lewis (tr.), Islam from the Prophet Muhammad to the Capture of Constantinople
(Oxford, 1987), ii, 204 f.
42 The British expansion was not legitimated in religious terms, and it was only where the
missionaries dominated that conversion to Christianity was seen as the key that unlocked
the door to the conquest society.
43 Cf. P. Crone, “Mawālī and the Prophet’s Family: a Shīʿite View”, in M. Bernards and J. Nawas
(eds.), Patronate and Patronage in Early and Classical Islam (Leiden, 2005), 184–185, where
I first made this point.
164 chapter 7
By the ninth century, however, all this was in the past. All Muslims now
said that prejudiced behaviour was wrong; Arab and non-Arab Muslims were
all the same, or almost the same (the sense that the Arabs were a chosen
people never entirely disappeared), and in terms of careers, non-Arab ethnicity
was not the slightest impediment any more; on the contrary, non-Arabs now
dominated at elite level. Yet Shuʿūbism continued, or indeed intensified. For
what was at stake was not just career prospects, but also self-respect and,
above all, the character of the culture that converts were now sharing with the
conquerors.
Converts to Islam were in the disagreeable position of owing their innermost
16 convictions to people they disliked. The Arabs had dragged them | to paradise in
chains, as a famous saying had it.44 How were they supposed to react? By being
grateful? Yes, many people said, on the grounds that the Arabs had brought
the truth, whatever else they had done. All religious scholars seem to have
taken this view regardless of their ethnic origins. In the caliphal army, too,
allegiance to the Arabs was widely seen as essential even though the soldiers
were more often than not assimilated Iranians: without the original bearers of
the religion, they feared, Islam might drown in the sea of unconverted and/or
unassimilated natives.45 But there were also people who, whatever gratitude
they might feel to God for being Muslims, found it impossible to feel grateful to
the Arabs for having conquered them. Typically, they were Iranians working in
and around the court, as bureaucrats, translators, copyists and other purveyors
of professional knowledge and skills.
What do you do if you owe your beliefs and values to people who have
defeated your ancestors and treated them badly thereafter? If you cannot, or
do not want to, become one of them, the only solution is to dissociate the
beliefs that you want to retain from the carriers that you want to discard. Just as
modern science and other aspects of secular modernity are coming to be seen
not as something specifically Western, but rather as a human development
which simply happens to have played out its most recent phase in the West,
so Islam had to be seen as part of a divine process which simply happened to
have culminated in Arabia. Both interpretations are eminently defensible in
historical terms, yet neither made its appearance before the respective bearers
of empire had lost their monopoly on power: it was the desire to have the belief
44 Al-Bukhārī, al-Jāmiʿ al-ṣaḥīḥ, ed. L. Krehl (Leiden, 1862–1908), ii, 250; al-Haythamī, Majmaʿ
al-zawāʾid, third printing (Beirut, 1982), v, 333; cited in Ibn ʿAbd Rabbih, ʿIqd, iii, 412; tr.
Lewis, ii, 203.
45 Cf. Crone, “ʿAbbāsid Abnāʾ”, 14 f.
system without indebtedness to the bearers of empire that caused people to

rethink history, and it was the new distribution of power that caused the result
to be heard on both sides.
The spirit in which the Shuʿūbīs presented their rethinking was usually
polemical rather than academic. They argued, quite correctly, that all the
prophets before the rise of Islam had been non-Arabs, some minor exceptions
apart, and inferred that it was really the non-Arabs who had discovered the
truth, or most of it (the equivalent claims nowadays mostly refer to science);
they added that various early converts to Islam had been non-Arabs, too, so that
Islam could be said (with some exaggeration) to have been half non-Arab from
the start. Besides, the Shuʿūbīs intimated, the Arabs had shown themselves to
be bad Muslims by their ter|rible treatment of non-Arab converts, whereas the 17
latter had taken Islam to heart: assimilated natives were now better bearers of
the belief system than the conquerors. Moreover, they said, with the partial
exception of the belief system, the non-Arabs owed nothing to the Arabs, for
all the kings before the rise of Islam had been non-Arab, as had all science,
technology, art and literature, with the partial exception of poetry.46 Their tone
was as shrill as that of their modern counterparts: we had civilization while you
Arabs were still eating lizards in the desert, as they put it (while you Westerners
were still swinging in the trees, as their modern counterparts say today). And
just as their modern counterparts talk more about prejudice and colonial atti-
tudes today than they did in the past when they were truly exposed to them,
so the Shuʿūbīs harped on the prejudiced behaviour of the Arabs at the very
time when it had ceased to affect them much. By constantly pouring abuse on
the Arabs while stressing their own contribution to religion, government and
culture before (and indeed after) the rise of Islam, the Shuʿūbīs informed the
world that the Arabs did not deserve a special place in Islam, let alone in the
high culture with which the belief system was associated.
Though the Shuʿūbīs disliked the Arabs, it was not their ambition to destroy
the caliphate, in which they were doing very well, and with one famous excep-
tion, they expressed no hope for the return of the Persian empire.47 Rather, they
took it for granted that all Muslims were now sharing the same political house:
what they were debating was their own status within this house, and the char-
acter of the culture it was to accommodate.48 They did not resent using Arabic
46 Ibn ʿAbd Rabbih, ʿIqd, iii, 404 ff.; partial tr. Lewis, Islam, ii, 201ff.
47 S.M. Stern, “Yaʿqūb the Coppersmith and Persian National Sentiment”, in C.E. Bosworth
(ed.), Iran and Islam (Edinburgh, 1971).
48 Thus H.A.R. Gibb, “The Social Significance of the Shuubiya”, in his Studies on the Civiliza-
tion of Islam, ed. S.J. Shaw and W.R. Polk (Princeton, 1962).
166 chapter 7
as the shared imperial language, either. But they did not want to think of
the beliefs they had internalized as something they owed to conquerors, and
what they wanted to read in Arabic, apart from the Qurʾān, were islamized
versions of their own cultural traditions, not traditions relating to Arabia. It is
no accident that debates over the literary canon were raging at the same time
as the Shuʾūbī controversy, though it is unclear how far the poets pioneering
18 “modern” (muḥdath) poetry were Shuʿūbīs themselves.49 People were | tired of
reading the output of dead tribal males. They wanted poetry, Persian culture,
Greek philosophy, Indian statecraft and anything else available in the Near East.
In short, their outlook could be summarized as “Hey ho, Arab civ. has gotta go”,
except that they denied that there was any such thing as Arab civilization.50
The Tenth-Century Crisis
The “tenth-century crisis” is a shorthand for developments over the next three
centuries, roughly 850–1150, for which no name seems to exist. Fazlur Rahman
spoke of them as a crisis,51 and it peaked in the tenth and early eleventh
centuries: hence the nomenclature adopted here.
There were still Shuʿūbīs in the tenth century, but the intellectual climate
had changed and they no longer held the centre stage, for by now the caliphate
had broken up and the differences between the conquerors and the conquered
peoples had been even further effaced. In political terms, both Muslims and
non-Muslims were now living under secular kings, usually of non-Arab origin:
it was an upstart Iranian ruler who presided over the disputation between the
two Rāzīs at Rayy.52 The new rulers were secular (or profane) in the sense of
“not prescribed by the Sharīʿa”, not in the sense that they kept religion out of the
public sphere; on the contrary, they saw themselves as servants of Islam, or at
least they were supposed to, so Islam retained its political dominance. But the
49 For the question whether Abū Nuwās was actually a Shuʿūbī, see E. Wagner, Abū Nuwās
(Wiesbaden, 1965), 136 ff.
50 The Berkeley students who shouted this slogan in 1968 (in its original version, “hey ho,
Western civ. has gotta go”) were mostly members of the empire-bearing people, however,
or rather of their American successors, whereas Arabs never seem to have been Shuʿūbīs
(Ḍirār b. ʿAmr, sometimes adduced as an exception, is not really one). The post-imperial
bad conscience displayed by Westerners should presumably be related to the weakness of
secularism as an imperial creed.
51 F. Rahman, Prophecy in Islam: Philosophy and Orthodoxy (London, 1958), 63.
52 See above, n. 3.
sacred polity distinguishing Muslims from all others had disappeared, or rather
turned into a purely notional religious community. Moreover, as the Iranians
were returning to power inside the Muslim community, so the Byzantines were
returning outside it, conquering northern Syria and broadcasting wild visions
of reconquering Jerusalem, Egypt, and more besides.53
Culturally, too, the pre-conquest Near East was resurfacing in a recognizable 19
way. We are now in the period that some call the Iranian intermezzo and others
the Renaissance of Islam, with reference to the return of the above-mentioned
Iranian rulers plus Persian culture and the Persian language on the one hand
and that of Greek science and philosophy (without the rulers) on the other. The
debate between the two Rāzīs is symptomatic in that respect, too, for both men
were Iranians and most of what they said had long roots in Near Eastern culture.
In cultural terms, Muslims and dhimmīs, too, were converging, especially at the
level of the elite. Educated Muslims and non-Muslims were now speaking and
writing the same language (if not usually in the same scripts) and participating
in the same high culture. As secretaries, astrologers and doctors, dhimmīs often
moved in courtly circles, enjoyed great wealth, and were hard to distinguish
from the Muslims. At elite level, in other words, the natives had been largely
assimilated now even though they had not all converted.
In short, the Muslims were no longer clearly marked off from their non-
Muslim subjects by ethnicity, culture or worldly success. Of course, Islam was
still politically dominant, but things did not look good on the ground. When
Daylamite mercenaries established a protectorate over the caliph in Baghdad,
adopting the Persian imperial title of shāhanshāh (“King of Kings”) and order-
ing their protegé to treat them with proper honours, it was somewhat as if
a Gurkha mercenary had taken power in London after the dissolution of the
British empire, calling himself Maharaja and telling members of Parliament
to get down on their knees before him.54 Symbolically, the conquerors had
been forced to withdraw. As the poet al-Maʿarrī put it, if al-Manṣūr had risen
from the grave, his reaction would have been to regret having killed Abū Mus-
lim on the grounds that “the sons of Hāshim dwell in the desert, and their
empire has passed to the Daylamites”.55 Moreover, the transfer of power from
53 A.A. Vasiliev, History of the Byzantine Empire (Madison and Milwaukee, 1964), i, 306–311;
G. von Grunebaum, “Eine poetische Polemik zwischen Byzanz und Bagdad im x. Jahrhun-
dert”, in Analecta Orientalia 14 (Studia Arabica i, Rome 1937).
54 Cf. ʿAḍud al-Dawla’s message to the caliph in 370/980 in H. Busse, “The Revival of Persian
Kingship under the Būyids”, in D.S. Richards (ed.), Islamic Civilisation 950–1150 (Oxford,
1973), 62.
55 R.A. Nicholson (ed. and tr.), “The Meditations of al-Maʿarrī”, in his Studies in Islamic Poetry
168 chapter 7
20 the conquerors | to the conquered peoples had involved extreme political frag-
mentation: there no longer was a unitary Muslim state to counter the Byzantine
empire, and the political control of the new rulers was limited. Al-Masʿūdī
shuddered at the thought of invasions by Turks, Allans, Khazars and others
“with the weakness and evanescence of Islam at this time, the victory of the
Byzantines over the Muslims, the ruination of the pilgrimage, the absence of
jihād, the unsafe and dangerous nature of the roads, and what with people
setting themselves up as independent rulers in any locality they inhabit after
the fashion of the ‘party kings’ after the death of Alexander up to the reign of
Ardashīr”.56 Above all, the cultural fusion, though deeply exciting to a modern
scholar, was painful to live through. It is no secret that multi-culturalism and
the incipient fusion of traditions observable in the West today looks to many
as the beginning of the end of Western civilization even though Western sci-
ence and technology, political models, gender roles, clothing, eating patterns,
and many other things are spreading throughout the world (where they are per-
ceived as threats to the prevailing cultures in their turn). In the same way, the
resurfacing of pre-conquest culture in the Near East struck many Muslims as
heralding the end of what they took to be Islam, even though the religion was
constantly recruiting new adherents both within and beyond its political bor-
ders while at the same time Arabic and New Persian, as well as the high culture
associated with them, were spreading among the Jews, Christians, and Zoroas-
trians (generating fears for the survival of their traditions in turn). As far back as
the eighth century there were Arabs who grumpily blamed all ills on non-Arab
Muslims, whom they saw as an unwelcome presence in their society, much as
many people in Britain see Asian immigrants today;57 and Ismāʿīlism was com-
monly identified as a conspiracy by the conquered peoples to subvert Islam
(Cambridge, 1921), 237 = 100 f. (no. 106). Compare the poem put into the mouth of the
Byzantine emperor: “You have accepted the Daylamite as caliph and become slaves of the
Daylamite slaves. Return in ignominy to the land of the Ḥijāz and leave the land of the
Byzantines, noble men” (von Grunebaum, “Poetische Polemik”, verses 38f.).
56 al-Masʿūdī, Murūj al-dhahab, ed. C. Pellat (Beirut, 1966–1979), i, §504.
57 ʿUthmān supposedly predicted that things would go wrong with the coming of prosperity,
the achievement of adulthood by the children of captive women, and both Arabs and
non-Arabs reciting the Qurʾān (al-Ṭabarī, Taʾrīkh al-rusul wa-l-mulūk, ed. M.J. de Goeje
et al. (Leiden, 1879–1901), i, 2803f.); ʿUmar predicted that the Arabs would perish when
the children of Persian women grew up and said that the Israelites had done well until
the muwalladūn abnāʾ al-sabāyā led them astray. See Ibn Abī Shayba, al-Muṣannaf, ed.
M.A. al-Nadwī (Bombay, 1979–1983), xii, no. 12516; Sayf b. ʿUmar al-Tamīmī, Kitāb al-ridda
wa-l-futūḥ, ed. q. Al-Samarrai (Leiden, 1995), 18, no. 21. Abū Ḥanīfa and others said the
same of the Muslims; Abū Zurʿa, Taʾrīkh, ed. Sh. Al-Qawjānī (Damascus, 1980), no. 1339.
from within by | means of a fatal mixture of Zoroastrian, Manichaean and Greek 21

philosophical ideas.58 Then as now, too, the fusion generated the phenomenon
of trahison des clercs: Western intellectuals denouncing their forebears for their
colonial sins and attacking their own cultural tradition; Muslim intellectu-
als embracing Greek and/or Iranian ideas in order, as it seemed, to subvert
their own religion. To ʿAbd al-Jabbār there were enemies of Islam everywhere,
but above all in the Muslim community itself: translators of Greek, Persian
and Indian books into Arabic, physicians, philosophers, crypto-Manichaeans,
Ismāʿīlīs and other Shīʿīs, all came across to him as so many auto-immune dis-
eases.59
The fact was that at elite level all the confessional communities of the Middle
East were coming together in a single cultural world, all of them were finding
that their cherished beliefs were being transformed and relativized by the
encounter. Back in the seventh and eighth centuries, the absolute truth of Islam
had seemed self-evident to its adherents: nothing else in the world was the
source of so much dazzling power and success; God was clearly siding with the
Arabs. Even those who refused to convert found it impossible to resist the pull
of the new religion, borrowing this or that key idea in the hope of defeating its
bearers with their own arms, and imitating them in other ways, too, because the
ways of the powerful are attractive. (These factors are conspicuous in the spread
of Western ideas, too.) But the very power which makes a belief system seem
self-evidently true while its bearers are on top of the world has a way of placing
a question mark over its validity when they lose their dominant position. Just
as the collapse of the European empires has been followed by doubt about the
validity of Western institutions (even as they are spreading), so the collapse
of the Arab caliphate was followed by doubts about the beliefs it left behind
(even as they were spreading, too). Now that Islam had lost its epistemological
privilege, it was no longer self-evident what it had over other systems of belief.
In the Muslim case, 300 years of Islamic dominance had endowed all the
competing systems (Zoroastrianism included) with the same basic structure:
all operated with a single (good) God, saw Him as having com|municated with 22
mankind through prophets, preserved the communication in a scripture, and
authenticated the scripture with reference to a tradition which was deeply
58 Thus, among many others, Ibn al-Nadīm, Kitāb al-fihrist, ed. R. Tajaddud (Tehran, 1971),
239 f.; tr. B. Dodge (New York, 1970), i, 469; al-Ghazālī, Faḍāʾiḥ al-bāṭiniyya (Amman, 1993),
24 (ch. 3, ii).
59 ʿAbd al-Jabbār, Tathbīt dalāʾil al-nubuwwa, ed. ʿA.-K. ʿUthmān (Beirut, 1966), 70ff. (trans-
lators), 51, 128 f. (Abū ʿĪsā, Ibn al-Rāwandī and other mulḥids and zindīqs), 129f. (Baḥrayn
affair), 623 ff. (Rāzī), 626 ff. (physicians), 631 (Kindī), etc.
170 chapter 7
meaningful to insiders but had no probative value to outsiders. Which one

of them was true? The only way of judging between them was by reason, but
reason proved incapable of delivering a verdict, for rational arguments in favour
of one tenet could always be countered by others of equal weight against it,
as the mutakallims soon found out thanks to disputations in which the rival
religions or sects were defended on the basis of rational arguments alone. The
superiority of Islam could not be proved, except to the converted; nor could
that of any other religion, or of any subdivisions within them. To those in search
of proof, the increasingly even positioning of the various systems in the socio-
political hierarchy made all of them look much the same in epistemological
terms as well, generating the feelings of relativism and doubt that so often
appear where rival belief systems compete on an equal footing, and causing
Muslims and dhimmīs alike to go on real or imagined journeys in quest of
wisdom. By the later ninth century, religious scepticism with reference to the
equipollence of proofs (takāfuʾ al-adilla) is well-attested.60 There were also
philosophers who denied that humans could know anything for certain at all,
claiming that all truth was relative so that everything was both true and false at
the same time, or even that life itself was an illusion.61 And then as today, the
suspicion arose that all that the privileged system had ever had over the others
was power, or at the very least that it was badly contaminated by power.
In the Muslim case the problem posed by power owed its formulation to
the fact that back in the days when the natives were Christians, Zoroastrians,
and Gnostics, they had often claimed that Islam was false because it was
spread by the sword (debiting its invincibility to their own sins rather than to
God’s agreement with it). Initially they said it in Greek, Syriac and other local
languages, then they said it in Arabic, and from the tenth century onwards one
finds it in Muslim writings as an embarrassing charge that had to be refuted.62
By then, the charge was also being made by Muslims. There were Shīʿīs, for
example, who used it to discredit the Companions, claiming that they had only
followed Muḥammad for the sake of plunder and power, not because they knew
23 him to be a true | prophet (which ʿAbd al-Jabbār took to mean that the Shīʿīs in
question denied Muḥammad’s prophethood as well);63 and the issue also came
60 J. van Ess, “Skepticism in Islamic Religious Thought”, al-Abḥāth 21 (1968), 7.

61 Van Ess, “Skepticism”, 1 f.; P. Crone, “al-Fārābī’s Imperfect Constitutions”, Mélanges de
l’ Université Saint-Joseph 57 (2004) [Ed.: reprinted in P. Crone, The Iranian Reception of
Islam: The Non-Traditionalist Strands, vol. 2 of Collected Studies in Three Volumes, ed.
H. Siurua (Leiden, 2016), art. 12], notes 79–87.
62 Crone, Medieval Islamic Political Thought, 375 f.
63 ʿAbd al-Jabbār, Tathbīt, 35.
up between the two Rāzīs, for it was above all with warfare that the philosopher
associated prophets, requiring the Ismāʿīlī Rāzī to explain why the fact that
Muḥammad used the sword did not invalidate his message.64 In sum, what
worried people now was not the role of Arabs in the rise and spread of Islam,
but rather that of power.
Accordingly, the tenth and eleventh centuries are dominated by attempts
to dissociate the ultimate truth from the political and military concerns with
which the Prophet had fused it. People were looking for a single absolute truth
which had nothing to do with power, which all humans could accept regardless
of the perspective from which they saw it, and which spoke to them as indi-
viduals rather than members of this or that confessional community. Unlike
the philosopher Rāzī, who simply discarded the confessional boundaries as the
creations of deluded men, most people wished to combine belief in this abso-
lute truth with continued membership of the communities into which they
had been born, remaining loyal to their prophet and the tradition of which he
was seen as the founder. But one way or the other, the universalism that the
Shuʿūbīs had fought for within Islam now had to embrace all human beings. It
was a disturbing development to the religious scholars, whether mutakallims or
traditionalists, Sunnīs or Imāmīs, given that it threatened to reduce the truths
they worked with to parochial formulations of something higher shared by all
mankind. But though they wrote against the new trends, they do not seem to
have had any answers to the questions they posed. It was the philosophers and
the Ismāʿīlīs who embraced the new developments and who knew how to han-
dle them.
The New Leaders
Post-colonialism was not the only factor at work: another was the rise to promi-
nence of educated laymen. Secretaries, administrators, doctors, astrologers,
copyists, and other professionals (and to some extent also poets) all owed their
wealth and status to secular know-how rather than mastery of the religious tra-
dition (though they were usually well schooled in that tradition too). Highly
educated and trained to think on the basis of human rather than revealed infor-
mation, they were often disinclined to | defer to religious scholars, whom they 24
frequently rivalled in terms of wealth and influence as well. They rose to promi-
nence after the revolution in 132/750, when they benefitted from the ʿAbbāsid
64 Abū Ḥātim, Aʿlām, 3 f., 181 ff., 186 ff.

172 chapter 7
expansion of the bureaucracy, and they benefitted again from the political
break-up of the caliphate from the ninth century onwards because the new
rulers usually modelled their courts on that of Baghdad and so felt obliged to
patronize whole bevies of such men. In the ninth century the professionals
tended to be rationalizing theologians (mutakallims) rather than traditional-
ists, and it was also from their ranks that the Shuʿūbīs were recruited; but in
the tenth century they tended to be philosophers. As philosophers, they were
rivals of the mutakallims (and had no time for traditionalists at all), so there
is sometimes an element of anti-clericalism in their thinking, most obviously
in that of Abū Bakr al-Rāzī. This gives them a similarity with the philosophes
of enlightenment in Europe, with whom they have much in common in terms
of their actual ideas as well. But unlike the philosophes, they were also heirs
to an empire that had united different ethnic and religious communities, and
anti-clericalism is less pronounced in their thinking than a desire simply to rise
above the clerics. Jewish, Christian and Muslim members of the professional
elite often had more in common with each other than with their own core-
ligionists: in such circles the idea of single truth above the many had strong
appeal.
Ismāʿīlīs were sometimes secretaries, too, but their first leaders seem to
have been villagers and petty townsmen engaged in local transport, trade or
crafts, in keeping with the humble milieux in which Gnosticism appears to
have flourished in the first centuries of Islam. At least some of them were
literate and wrote books, but they were not truly educated, and they had no
links with the political and cultural establishments. Why such people should
have felt the need to project themselves onto the public scene is hard to say,
though the fact that the agrarian economy seems to have undergone a fair
degree of commercialization in (or by) the tenth century may come into it.65
The Ismāʿīlīs moved closer to elite level in both social and intellectual terms
in the course of the tenth century, when they overlaid their Gnosticism with
Neoplatonist philosophy (especially in Iran) and rose to political power in
Fāṭimid North Africa and Egypt. But their leaders (known as missionaries,
25 though they soon became the | equivalent of bishops) were primarily suppliers
of pastoral care to local communities, and they were less willing and/or able
to transcend their own familiar world than the philosophers. All prophets,
according to Abū Ḥātim, had preached the same inner message which was
65 Cf. A. Mez, Die Renaissance des Islams (Heidelberg, 1922), ch. 24 (still useful); A.L. Udo-
vitch, “International Commerce and Rural Society in Egypt of the 11th Century”, in A.K.
Bowman and E. Rogan (eds.), Agriculture in Egypt from Pharaonic to Modern Times (Lon-
don, 1999).
soon to become the creed for all mankind, but when he set out to explain
this thesis, the inner message he discerned in all the revelations was in effect
Islam.66
The Symptoms
It was among the rationalizing theologians (mutakallims) of the ninth century

that the doubts about the absolute truth of traditional religion began. Its ear-
liest manifestation was scepticism (ḥayra, also translated “perplexity”) about
the truth of any one religion. Asked why he followed his particular religion, one
such sceptic in Sīstān replied that he did not know it to be truer than any other
(he accepted the principle of takāfuʾ al-adilla), that he had simply been brought
into it by his parents, but that long familiarity had made it dear to him: he was
like a traveller in a caravanserai, he said; the manager had showed him into a
room without consulting him, and when it began to rain, the ceiling proved to
be leaking; so he had wondered whether to get himself another room, but then
he saw that the courtyard was muddy and that the ceiling was leaking in the
other rooms too, so he decided to stay where he was.67
Rationalizing theologians were also the first to have doubts about the exis-
tence of prophethood. Such doubts were not in fact the only way in which the
dwindling of confidence in conventional religion displayed itself: loss of faith
in bodily resurrection, or in any kind of afterlife at all, was also prominent, as
was the problem of reconciling a single omnipotent God with the existence
of evil.68 But prophets were at the centre of | the debate because it was in 26
them that the confessional communities originated: remove them and you had
what the European Enlightenment thinkers called natural religion; that is, a
religion in which the relationship between God and the individual was based
directly on human nature, without the intermediary of institutions posited by
66 When other religions differ from Islam, their tenets are declared not to come from the
prophets, but rather from later innovators who corrupted their faith in a bid for power
(Abū Ḥātim, Aʿlām, 160, 171ff.). The once common view that the Ismāʿīlīs were particularly
inclined to supra-confessionalism rests on the K. al-balāgh, a forgery in which the (grossly
distorted) ideas often seem to be rooted in philosophy rather than Ismāʿīlīsm. See Stern,
Studies, ch. 4; cf. Crone, “Fārābī’s Imperfect Constitutions”, note 69, on grades of initiation;
below, notes 93–94, on the prophets as impostors.
67 al-Tawḥīdī, K. al-Imtāʿ wa-l-muʾānasa, ed. A. Amīn and A. al-Zayn (Cairo, 1939–1944), iii,
193 f.; cited in Van Ess, “Skepticism”, 6 f.
68 I hope to deal with this in a longer work on the subject.
174 chapter 7
prophets and maintained by others claiming to have special knowledge about

God, such as priests, imams or religious scholars. Medieval Muslims did not
use the expression natural religion, though they came close at times; rather,
they spoke about rational religion. But what they meant was the same: a reli-
gion which freed the thinking individual from dependence on the institutions
and conventions of the community in his relationship with God, allowing him
instead to approach God directly, as a single soul on his own.
Among the mutakallims the first to attack communal religion were Abū ʿĪsā
al-Warrāq (d. 247/861 or later) and Ibn al-Rāwandī (d. between 240s/860s and
298/912). One or the other, or both, famously declared that either the prophets
said things in conformity with reason, in which case they were superfluous, or
else they said things contrary to reason, in which case they were wrong; and Ibn
al-Rāwandī apparently added that prophets were magicians and tricksters.69
We also hear of a tenth-century mutakallim, Abū Isḥāq al-Naṣībī (fl. around
370/980), who had his doubts about prophetic missions.70 But by his time
the initiative had passed to the philosophers. Thus al-Sarakhsī (d. 286/899) is
credited with a book dismissing the prophets as tricksters;71 the tenth-century
philosopher Abū l-ʿAbbās al-Īrānshahrī is said not to have believed in any
existing religion, only in one which he had devised for himself.72 Abū Bakr
27 al-Rāzī allegedly plagiarized his scientific ideas.73 After al-Rāzī we find | al-
Fārābī (d. 339/950) writing against people who dismissed the prophets (or
“lawgivers”, as he called them) as jugglers and tricksters, al-Rāzī presumably
among them.74 But there were others of the same kind. According to the
Brethren of Purity (wrote 360s/970s?), there were intelligent people who would
engage in philosophy and reject the stories of Adam, Eve, the angels and the like
because they took them literally instead of following their spiritual meaning, so
69 Stroumsa, Freethinkers, ch. 2.

70 Tawḥīdī, Imtāʿ, i, 141 ( yashukku fī l-nubuwwāt kullihā); cf. id., Akhlāq al-wazīrayn, ed. M. al-
Ṭanjī (Beirut, 1991), 202, 211 f., 297.
71 Bīrūnī in F. Rosenthal, Aḥmad b. aṭ-Ṭayyib al-Saraḫsī (New Haven, 1943), 51.
72 Al-Bīrūnī, Taḥqīq mā li-l-Hind, ed. E. Sachau (London, 1887), 4. According to Abū l-Maʿālī,
Bayān al-adyān, ed. H. Raḍī (Tehran, 1342), 67, Īrānshahrī claimed to be a prophet sent
to the ʿajam and wrote a book in Persian which he claimed to have from an angel; i.e. he
is here a nativist prophet rather than a rationalist freethinker. But his scientific views as
recorded by Bīrūnī in a variety of works rule out this interpretation, cf. S. Pines, Studies in
Islamic Atomism (Jerusalem, 1997), 41 f., 48, 54, 65–67.
73 Nāṣir-i Khusraw, Zād al-musāfirīn, in Kraus, Rasāʾil, 255f., 259.
74 Al-Fārabī, K. ārāʾ ahl al-madīna al-fāḍila, ed. and tr. R. Walzer (Oxford, 1985), 17, §6; cf.
Crone, “Fārābī’s Imperfect Constitutions”.
that they would fall into scepticism and doubt, though they might hide it for
fear of the sword, and that sometimes they would reject the prophetic books on
the grounds that reason made revelation unnecessary.75 According to al-ʿĀmirī
(d. 381/996) there were “pretentious people” (mutaẓarrifa), probably in Iran,
who dismissed all religions as conventional institutions designed to facilitate
social life, arguing that they would not have been based on revelation (tawqīf )
rather than reason if there had been any truth to them, and that there would
not have been so many of them either.76 Rāghib al-Iṣfahānī (fl. early 5th/11th
c) also knew of people who rejected positive religion, some of them on the
grounds that there were too many rival forms of it,77 while a friend of Ibn Sīnā
(d. 428/1037) had trouble believing in prophethood, causing Ibn Sīnā to write an
epistle affirming it.78 The poet al-Maʿarrī (d. 449/1058) repeatedly voiced views
strikingly similar to al-Rāzī’s: prophets were tricksters in search of a livelihood,
all positive religion was instituted by humans, he said (or presented others
as claiming); “They all err, Muslims, Christians, Jews and Zoroastrians; two
make humanity’s universal sect: one man intelligent without religion, and one
religious without intellect”, as he put it in what must be his most famous line on
the subject.79 Both al-Māwardī (d. 450/1058) and al-Ghazālī (d. 505/1111) wrote
against the belief that reason made revelation unnecessary, that prophetic
miracles were mere sleights of hand, and that the prophets were liars whose
untruths were meant to deceive the world | according to some, to benefit it 28
according to others;80 and al-Ghazālī reported that loss of faith in prophethood
was widespread.81 The real or pseudonymous ʿUmar Khayyām (d. c. 517/1123)
provides us with yet another example: “Will no one ever tell us truthfully
whence we have come and whither we go?”, as one of the quatrains circulating
under his name exclaims.82 This takes us into the twelfth century but thereafter
the attestations peter out.
75 Rasāʾil Ikhwān al-Ṣafā (Beirut, 1957), iv, 10, 100.

76 al-ʿĀmirī, K. al-iʿlām fī manāqib al-Islām, ed. A.ʿA.-Ḥ. Ghurāb (Cairo, 1967), 101. [Ed.: The
originally published text read, erroneously, “based on reason rather than revelation.”]
77 Rāghib al-Iṣfahānī, al-Iʿtiqādāt, ed. Sh. al-ʿIjlī (Beirut, 1988), 109f.
78 Ibn Sīnā, “Fī ithbāt al-nubuwwāt”, in his Tisʿa rasāʾil, ed. H. ʿĀsī ([Beirut], 1986), 95.
79 Nicholson, “Meditations of al-Maʿarrī”, no. 239, with discussion at pp. 164ff.
80 al-Māwardī, Aʿlām al-nubuwwa, ed. M.M. al-Baghdādī (Beirut, 1987), 50f.; al-Ghazālī, Tahā-
fut al-falāsifa, ed. and tr. M.E. Marmura (Utah, 1997), muqad. 4; id., Fayṣal al-tafriqa, ed.
S. Dunyā (Cairo, 1961), 184; tr. S.A. Jackson (Oxford, 2002), 101; tr. F. Griffel (Zürich, 1998),
67.
81 al-Ghazālī, al-Munqidh min al-ḍalāl, ed. and tr. F. Jabre (Beirut 1959), 46 = 110.
82 A. Dashti, In Search of Omar Khayyam (London, 1971), 117.
176 chapter 7
The Remedies
What kind of truth did reason supply? To al-Maʿarrī and others, the answer
seems to have been, not much of one, in the sense that he and others lived with
uncertainty about the metaphysical realm and based their moral decisions
on rational considerations as best they could. Al-ʿĀmirī’s “pretentious people”
recommended following the injunctions shared by all religions and leaving off
the rest; Rāghib’s sceptics held it best to stop thinking about religious divisions
and to work in fields known to be good for mankind, such as medicine and
agriculture, a solution also recommended (as Rāghib notes) by Burzoē in his
introduction to Kalīla wa-Dimna.83 It is about as far as many people get today.
But to others, reason meant philosophy in the technical sense, and that in its
turn meant a two-tiered concept of religion similar to that adopted by the
Ismāʿīlīs. The upper level was occupied by Aristotelian and/or Neoplatonist
philosophy, which gave you eternal verities for all mankind; the lower level was
occupied by positive religion, which gave you approximations of the highest
truth expressed in mythical and allegorical form for the many who could not
understand philosophy. The revealed religions differed from one community
to the next, but there was no need to be worried by this, for the differences
were required for socio-political functions they served, and the eternal verities
they reflected were the same. Unlike the Ismāʿīlīs, however, the philosophers
29 had no intention of ever abolishing the lower | level. They mostly accepted
it in its Sunnī form, and though they did not usually display enough of an
interest in this level to be associated with a particular legal school (Ibn Rushd
is the great exception), they held communal life to be impossible without the
law. They did not believe in the spiritual perfectibility of man and had no
hopes for a world without religious or political divisions. At the most they
held that individual philosophers could perfect themselves to the point of
dispensing with the Prophet’s injunctions, but this was not something they
would broadcast. The Ismāʿīlīs had higher hopes because they expected the
final unification of mankind to be effected by God, that is they awaited a
new revelation. They were not alone in this; and unlike the philosophers and
others who placed their faith in reason, those who expected Muḥammad’s law
to be abrogated often seem to have expressed themselves in anti-Arab terms
reminiscent of Persian restorationism. Back in the ninth century, for example,
a certain ʿAbdallāh al-ʿĀdī or ʿAbdī had written an astrological work predicting
the coming of a man who would unite all of mankind in a single community
83 Above, nn. 77 f., 79.

and put an end to evil: he would do this by restoring Zoroastrianism and

eliminating the power of the Arabs (mulk al-ʿarab).84 An Ibāḍī by the name
of Yazīd b. Unaysa, perhaps also active about this time, predicted that God
would raise up a non-Arab/Iranian prophet who would bring a new book and
follow the religion of the Sabians mentioned in the Qurʾān, i.e. he would bring
a religion foretold in the scripture which would both fulfil and abrogate that
scripture as a more universal form of its predecessor.85 In a more violent vein
a number of apocalyptic traditions preserved in a tenth-century Imāmī Shīʿī
book predict the coming of a messiah who would conquer the Chinese, the
Turks, the Indians and others, bring a new law, and slaughter the Arabs.86 When
the tenth-century Ismāʿīlīs took political action in the belief that the coming of
this messiah was imminent, they found him in an Iranian captive of whom it
was said that he descended from the kings of Persia and hailed from | Isfahan, 30
a city from which astrologers other than (or perhaps including) ʿAbdallāh al-
ʿĀdī had predicted the rise of a new religion;87 and there was an obvious
Zoroastrian element in some of the outrageous measures with which this
messiah, inaugurated in Baḥrayn in 319/931, tried to show that Muḥammad’s
law had been abrogated.88
At first sight, this anti-Arab streak is surprising, especially in Shīʿism, for
neither Imāmism nor Ismāʿīlism was a movement to restore the Persian empire
or rehabilitate the Iranians at the cost of the Arabs. On the contrary, Ismāʿīlism
was a movement to overcome all such earthly divisions so as to unite mankind
in a single spiritual religion. But Islam was still felt to be too closely tied to the
Arabs to allow for ethnic divisions to be completely transcended within it, just
84 al-Bīrūnī, al-Āthār al-bāqiya ʿan al-qurūn al-khāliya, ed. C.E. Sachau (Leipzig, 1923), 213 =
The Chronology of Ancient Nations, tr. C.E. Sachau (London, 1879), 196f. His nisba is given
as al-ʿAdī.
85 J. van Ess, Theologie und Gesellschaft (Berlin 1991–1997), ii, 614–618, summarizing id.,
“Yazīd b. Unaisa und ʿĪsā al-Iṣfahānī”, in Studi in onore di Francesco Gabrieli (Rome, 1984),
places him in the first/seventh century. Unfortunately, there is no real evidence either
way.
86 Ibn Abī Zaynab al-Nuʿmānī, al-Ghayba (Beirut, 1983), 154f.; cf. S.A. Arjomand, “Islamic
Apocalypticism in the Classic Period”, in The Encyclopaedia of Apocalypticism, ed. B. Mc-
Ginn, ii (New York, 1999), 264.
87 al-Masʿūdī, K. al-tanbīh wa-l-ishrāf, ed. M.J. de Goeje (Leiden, 1894), 391f.; id., Murūj, v,
§ 3600; Bīrūnī, Āthār, 132 = 129; W. Madelung, “The Assumption of the Title Shāhānshāh
by the Būyids and ‘the Reign of the Daylam’ (Dawlat al-Daylam)”, Journal of Near Eastern
Studies 28 (1969), 87n.
88 Bīrūnī, Āthār, 213 = 196f. For an account of the entire episode, see H. Halm, Das Reich des
Mahdi (Munich 1991), 222–236; tr. M. Bonner (Leiden, 1996), 247–264.
178 chapter 7
as its law was felt to be too externalist to allow for the religious unification of
mankind. Complete universalism could only be achieved at the cost of both.
Ismāʿīlī (or more precisely Qarmaṭī) missionaries in Baḥrayn expressed this
by preaching that it was the Arabs who had killed Ḥusayn.89 What they were
articulating was the Shīʿī equivalent of the Christian charge that the Jews had
killed Christ: just as one could not be both a Jew (a Christ-killer) and true Israel
(i.e. a Christian), so one could not be both an Arab (a Ḥusayn-killer) and a true
Muslim (i.e. an Ismāʿīlī Shīʿī). Most Ismāʿīlīs in Baḥrayn were ethnic Arabs, just
as most early Christians were ethnic Jews, but the issue was not ethnicity on
its own. Just as a “Jew” was an ethnic Jew who clung to the old dispensation
instead of following Christ, so an “Arab” was an ethnic Arab who clung to the
externalist features that the Ismāʿīlīs were abolishing instead of following the
Mahdī: all those who adopted the right belief were ipso facto gentiles. In both,
ethnicity rested on a combination of descent and belief.90 By refusing to be
31 Jews, the Christians broke with | the community in which they originated to
form a separate religion of their own.91 By rejecting Arab ethnicity the Qarāmiṭa
did the same.
The fact that the Qarāmiṭa chose a Persian prophet to preside over their
break with old Islam does not mean that they had a particular attachment to
things Persian, but rather that they envisaged their messiah as everything that
Muḥammad was not: the man who abrogated the old community was simply
an inversion of the man who had founded it. His various Persian qualifications
served to identify him as anti-matter to Islam, so to speak, not to mark him out
as the representative of a highly valued political, religious or cultural past. The
Qarāmiṭa would not of course have needed such anti-matter if they had broken
with old Islam gradually rather than in one single radical operation, but unlike
the Christians, they were political no less than religious revolutionaries; the
severance had to be total, public, and enacted with dramatic, preferably deeply
89 Goldziher, Muhammedanische Studien, i, 175. Compare Akhbār al-dawla al-ʿabbāsiyya wa-

fīhi akhbār al-ʿabbās, ed. ʿA. ʿA. al-Dūrī and ʿA.-J. al-Muṭṭalibī (Beirut, 1971), 198.8.
90 Compare the participants in the ʿAbbāsid revolution, who also saw Arab identity as resting
on a combination of descent (or naturalisation) and a belief they rejected: they too saw
themselves as gentiles whether they were Arab by ancestry or not. See P. Crone, “The
Significance of Wooden Weapons in al-Mukhtār’s Revolt and the ʿAbbāsid Revolution”, in
I.R. Netton (ed.), Studies in Honour of Clifford Edmund Bosworth, i (Leiden 2000), 179f.; cf.
also ead., “Mawālī and the Prophet’s Family”, 184 ff.
91 More precisely, that is how they talked, but reality was a good deal more complicated, cf.
A.H. Becker and A.Y. Reed (eds.), The Ways that Never Parted: Jews and Christians in Late
Antiquity and the Early Middle Ages (Tübingen, 2003).
shocking, rituals which brought it home to the participants that the old world
had been destroyed, that they were on the threshold to a new world, and that
they were on their own.
Among the deeply shocking rituals that served this purpose was ceremonial
cursing of the prophets, including the founder of Islam. Unlike the Christians,
the Qarāmiṭa could not retain the founder of the parent religion among their
sacred figures. It had not in fact been easy for the Christians to do so either:
Marcion had rejected Moses as representing the God of law overcome by Chris-
tianity, Gnostics of various kinds had rejected the Old Testament God as down-
right evil, deriding his law as shackles that had to be cast off for the sake of
spiritual perfection. Since the Qarāmiṭa were Gnostics by origin and moreover
revolutionaries, they too saw the law as shackles and their messiah now told
them to cast it off, instituting ritual cursing of the lawgiver prophets, Moses,
Jesus and Muḥammad, or perhaps of all prophets, as mere tricksters in search
of power.92 Das war also des Pudels Kern, his enemies responded. But the thesis
of the three impostors (which the Sunnīs also credit to the Ismāʿīlīs | in other 32
contexts)93 actually reflects the sentiments of Ibn al-Rāwandī, al-Sarakhsī, Abū
Bakr al-Rāzī, and other radical philosophers better than those of the Ismāʿīlīs,
who must have borrowed it from such philosophers, wittingly or unwittingly,94
not because they hated the prophets, but on the contrary because they loved
them too much: they had to vilify and throw dirt at them in order to enable
themselves to part with them for the sake of the new world, and what the radi-
cal philosophers offered was a ready-made language with which to do it. (One
wonders whether it was really via the Ismāʿīlīs rather than the philosophers
92 Ibn Rizām citing Abū Ṭāhir’s physician, cf. Halm, Reich, 231f.; tr. Bonner, 258ff.; Niẓām
al-Mulk, Siyāsatnāme2, ed. H. Darke (Tehran, 1985), 309; tr. H. Darke (London, 1960), 236
(ch. 46, § 36), probably reflecting the same source.
93 K. al-balāgh in al-Baghdādī, al-Farq bayna l-firaq, ed. M. Badr (Cairo, 1910), 278ff.; cf. L. Mas-
signon, “La legende ‘De tribus impostoribus’ et ses origines islamiques”, in his Opera
Minora, ed. Y. Moubarac (Beirut, 1963), i, 83 f. See also Maḥmūd of Ghazna to the caliph in
420/1029 in Ibn al-Jawzī, al-Muntaẓam, viii (Hyderabad, ah 1359), 39, where the Ismāʿīlīs
are said to regard all religions as made up by sages; Ghazālī, Bāṭiniyya, 24 (ch. 3, ii), where
godless philosophers, dualists and sceptics concoct such beliefs for Shīʿī consumption.
Both Abū Ḥātim al-Rāzī and Abū Yaʿqūb al-Sijistānī wrote books affirming their belief in
prophethood in no uncertain terms, but to no avail.
94 Cf. ei1, s.v. “al-Rāzī” (Kraus and Pines), where it is suggested that the Qarāmiṭa studied
Abū Bakr al-Rāzī’s books, on the basis of questionable evidence. It seems more likely
that the Qarāmiṭa had simply picked up this language, which was widely diffused at the
time.
180 chapter 7
themselves that the theme of the three impostors passed to Europe, where it
was to serve as dynamite against established religion in the Enlightenment.)95
As it turned out, the Persian messiah did not prove equal to the task, but
rather lost control of his community, to be killed by his own adherents. The
transition to the new post-prophetic order had failed. The coming of a new reli-
gion continued to be predicted,96 and the Baḥrayn Ismāʿīlīs did eventually suc-
ceed in abolishing the law in circumstances unknown, but by then they were
too peripheral to count. Meanwhile, another branch of Ismāʿīlīs had decided
to postpone the coming of utopia. This second and, as it turned out, much
more important branch consisted of the followers of the Fāṭimids, who estab-
33 lished themselves in North | Africa in 297/909, moving on from there to Egypt in
358/969; and having acquired real power, the Fāṭimids unsurprisingly did their
best to suppress messianic expectations. The prophets were not cursed, but on
the contrary venerated as indispensable for salvation in North Africa and Cairo.
Individual Ismāʿīlīs seem to have thought, much like the philosophers, that they
could rise above the rules laid down by the Prophet, but the era of collective
liberation from externality ceased to be just around the corner. Ismāʿīlīsm thus
lost the ability to conjure up a new world on which its early magnetism had
rested. When it reappeared as a major attraction in the sixth/twelfth century,
it was as a very different creed.97
The Seljuqs
A new era did none the less come, just not as people had imagined it. In 431/1040
the eastern frontier broke, and Turkish tribes poured into Iran, Iraq and Syria.
They reached Baghdad in 447/1055 under the leadership of the Seljuq family.
More Turks were to follow a century later, and still more in the 650s/1250s, when
they came as participants in the Mongol invasions. It was the end of both Arab
and Iranian power: from 1055 down to 1918 practically all rulers in the Muslim
Middle East were Turks.
95 It is first attested in Frederick ii’s Sicily around 1239 and reappears in Lisbon in the
1340s; M. Esposito, “Les hérésies de Thomas Scotus”, Revue d’Histoire Ecclésiastique 33
(1937), 59, 65, 69. See further F. Niewöhner, Veritas sive Varietas (Heidelberg, 1988); S. Berti,
F. Charles-Daubert and R.H. Popkin (eds.), Heterodoxy, Spinozism, and Free Thought in
Early-Eighteenth-Century Europe: Studies on the Traité des Trois Imposteurs (Dordrecht,
1996).
96 Nicholson, “Meditations of Maʿarrī”, no. 263:4 (wa-qīla yajīʾu dīnun ghayru hādhā).
97 Cf. Crone, Medieval Islamic Political Thought, 205–208, 325f.
Here the parallel with our own post-colonial world comes to a drastic end.
Nothing comparable has happened to us and nothing comparable probably
will, given that there are no outsiders left to play the Turks any more. It was
also the beginning of the end of the post-colonial malaise in the Muslim world
itself. After the Turkish invasions, the religious scholars return to the driving
seat, the confessional borders reassert themselves, and by the twelfth century
the evidence for scepticism, relativism, and unbelief begin to peter out along
with that for Shuʿūbism. Exactly how all this happened remains unknown. The
question has traditionally been discussed under the name of “the Sunnī revival”,
an unfortunate label which is now so heavily associated with the religious activ-
ities of the last caliphs of pre-Seljuq Baghdad on the one hand98 and with
conscious policies rather than the inadvertent effects of a barbarian in|vasion 34
on the other,99 that it seems best to do without it. But pursuing the question is
in any case impossible here. I shall confine myself to some comments on the
solution that won the day.
Sufism and al-Ghazālī
Tenth-century Sufism could not be said to provide a two-tiered religion. In its

pietist form of renunciation, asceticism and observance of the law for love of
God rather than fear of Him, it represented an interiorized form of conven-
tional religion rather than an upper-level form shorn of ties with this world.
As a spiritual search for direct experience of God, it left the status of conven-
tional religion undefined. Its attractions were limited, too, or so at least to the
educated elite in Iraq: Sufis were there seen as people who moved in humble
circles, mixed with low life, had questionable morals, knew how to milk peo-
ple for money, and spoke nonsense in grandiloquent and abstract terms.100 A
friend of al-Tanūkhī even tried the equivalent of a Sokal spoof on them.101 One
would not have guessed that it was with them that the future lay.
98 Cf. G. Makdisi, Ibn ʿAqīl et la résurgence de l’ Islam traditionaliste au xie siècle (Damascus,
1963); id., “The Sunnī Revival”, in D.S. Richards (ed.), Islamic Civilisation 950–1150 (Oxford,
1973).
99 Cf. for example G. Leiser’s introduction to his edition and translation of I. Kafesoglu, A
History of the Seljuqs (Carbondale and Edwardsville, 1988), 4.
100 F. Sobieroj, “The Muʿtazila and Sufism”, in F. de Jong and B. Radtke (eds.), Islamic Mysticism
Contested (Leiden, 1999); P. Crone and S. Moreh, The Book of Strangers (Princeton, 2000),
175 f.; Crone, “Fārābī’s Imperfect Constitutions”, notes 61–68.
101 Al-Tanūkhī, Nishwār al-muḥāḍara, ed. ʿA. Shāljī (Beirut, 1971–1972), i, 99; tr. D.S. Margo-
182 chapter 7
As spiritually centred on direct experience of God, Sufism shared with Ismāʿ-

īlism and philosophy the feature of addressing the believer as a naked soul,
shorn of worldly attachments, and of handing the key to salvation to the indi-
vidual rather than his community: one could not be born as a Sufi, a philoso-
pher or an initiate into the bāṭin that constituted the upper level in Ismāʿīlism;
one had to choose one’s own path, to take one’s salvation into one’s own hands.
In all three cases this involved relativization of the law and society into which
one was born, but Sufism was by far the most otherworldly persuasion of the
three: all institutions in this world were deemed to be impediments to the quest
for God; all had to be abandoned in the course of the journey to Him. At best,
35 this reduced | the law and the society based on it to secondary importance;
at worst, it completely drained them of religious significance, or even deprived
them of regulatory force, given the tendency for antinomian behaviour to blend
into immoral or criminal behaviour of the normal type. This has to be borne in
mind when it comes to explaining why there was so much hostility to Sufism
of the type centred on direct experience of God in the early days. The fact
that most Sufis probably lived by the law did not answer the question how
they expected their coreligionists to accommodate a spirituality that placed a
question mark over the value of marriage, homes, gainful employment, wealth,
power, book-learning, cleanliness or even clothes. When al-Ghazālī became a
Sufi, he resigned from his job and abandoned his wife and small children to
save his soul.102 As it happens, the outcome was books that seemed to save the
soul of the Muslims at large, so that in retrospect his behaviour looks noble; but
this may not have been how his family and pupils saw it, and in any case one
could not maintain a society by indiscriminate encouragement of this kind of
behaviour, as al-Ghazālī himself was well aware. It was up to the Sufis, then, to
demonstrate not only that they accepted both levels of religion, but also that
they knew how to fit the two together. Al-Qushayrī (d. 465/1072) made a contri-
bution to this with his Risāla, in which he denounced antinomianism (ibāḥa)
as a corruption of the original movement, but it was al-Ghazālī himself who
answered the question by providing a complete guide to observance of the law
as part of a spiritual life. He too wrote against antinomianism (in Persian, sug-
liouth, The Table-talk of a Mesopotamian Judge (London, 1922), 58f. (like the editors of
Social Texts, the members of the circle accepted it, but the shaykh saw through it).
102 Ghazālī, Munqidh, 38 = 99. That his children were small is clear from the fact that
he had still been unmarried when he arrived in Baghdad four years earlier; D. Kra-
wulsky (tr.), Briefe und Reden des Abū Ḥāmid Muḥammad al-Ġazzālī (Freiburg, 1971),
135.
gestive of where he saw it as prevalent),103 but his key contribution was his
Iḥyāʾ, in which to be a Muslim is to be a Sufi.
It should be noted that al-Ghazālī was fighting on two fronts, for he also
had to argue against those who held the law to be so important that the
whole of Muslim society was vitiated by its failure to live in accordance with
it. Both attitudes led people to reject normal society; both resulted in a view
of Muslim society as standing in the way of salvation. The obverse of ibāḥa
was refusal to handle money, earn a living or live in the Muslim community
in anything but a geographical sense, claiming | that it had no caliph and that 36
the whole umma had merged with the abode of kufr. Al-Ghazālī did his best
to get both groups back into the community, assuring them that it was still
a legitimate version of the community founded by the Prophet, that it still
had a legitimate caliph, and that it was with God’s blessing that power had
passed to the Turks; and he wrote in great detail on precisely what kind of
dealings one could and could not have with rulers without violating the law,
what kind of money one could take from them and what not.104 Throughout,
his aim is to impress on people that the Muslim community was still the saving
vehicle, that it had not been corrupted to the point of disappearing, and that
people should concentrate on getting their social life onto a moral footing
again.
In the fourth/tenth century, all the greatest minds had been trying to tran-
scend the Muslim community, to seek some unification of thinking men above
it. This is what is reversed with al-Ghazālī in the fifth/eleventh. Like his pre-
decessors, he had a strong sense of the difference between the conventional
religion and the natural (God-given) capacity of the human mind to know the
ultimate truth,105 and he seems to have been more of a Neoplatonist philoso-
pher in private than one would guess from his pastoral works.106 But at the
same time he had a genuine sympathy and respect for traditional believers
and common people, and also an intense sense of the importance of keeping
the Muslim community together. Accordingly, he refused to cast positive reli-
gion as mere parables or fairy tales for the masses designed to keep them in
order while the elite pursued the highest truth. He insisted that the Prophet’s
103 O. Pretzl, “Die Streitschrift des Ġazālī gegen die Ibāḥīja”, Sitzungsberichte der Bayerischen
Akademie der Wissenschaften, Phil.-hist. Abt., 1933; cf. also Krawulsky, Briefe, 210ff.
104 Al-Ghazālī, Fātiḥat al-ʿulūm (Damascus, n.d.), 139ff.; id., Iḥyāʾ ʿulūm al-dīn (Cairo, ah 1282),
ii, 110 ff. (K. al-ḥalāl wa-l-ḥarām, ch. 5); Crone, Medieval Islamic Political Thought, 237ff.,
305, 348.
105 H. Landolt, “Ghazālī and ‘Religionswissenschaft’ ”, Asiatische Studien 45 (1991), 19.
106 Cf. Landolt on his Mishkāt al-anwār (in the article in the preceding note).
184 chapter 7
revelation, the law on which Muslim society was based, was meant for all mem-
bers of this society: the revelation was the starting point for the exploration of
higher spirituality, not a substitute for it. Conversely, all members of Muslim
society were free to participate in the pursuit of the highest truth, that is to say
as Sufis: spiritual gifts were randomly distributed, did not require expensive
education, and did not have to be licensed by an imam. But however high the
Sufis soared, they had to respect the confessional boundaries on the ground.
37 In effect, al-Ghazālī was herding his coreligionists | back into the community
and providing them with their lower and higher forms of religion alike within
it.
From Scepticism to Sufism
It was as a person who had experienced the post-colonial malaise in per-

son that al-Ghazālī found Sufism to be the remedy, and set about pairing
it with conventional religion: he had suffered deeply from scepticism in his
youth. Where Sufism came to the rescue was in its epistemology. Like so many
others, al-Ghazālī had reasoned his way to the limits of reason: where was
he to go from there? One option was to live with uncertainty: many clearly
did. But this he found impossible. The only alternative was to postulate that
some humans in the here and now possessed a faculty higher than reason
through which such knowledge could be obtained. This he fully accepted.
The question was what humans? According to the Ismāʿīlīs, the higher fac-
ulty was possessed by the Imām, in whose instruction (taʿlīm) the believer
could find the escape from perplexity: it was taʿlīm, not messianism, that
was the great attraction of Ismāʿīlism when it reappeared as a major chal-
lenge. But to al-Ghazālī, the only bearer of instruction so authoritative was
the Prophet, who was dead and gone. What he accepted instead was that
there were ordinary people in the present who had similar gifts in the form
of dhawq, the lived, intuitive and entirely subjective experience of divine real-
ities by direct vision (mushāhada) and “unveiling” (mukāshafa) that the Sufis
cultivated.107 It was by seeking such subjective experience, or by recognizing
that others had it, that one prevented reason from running wild in scepti-
cism and kept it working instead for the belief system that one knew to be
true.
107 Cf. Ghazālī, Munqidh, 15 = 67f., on the four groups in which the truth had to be found, each
representing an epistemological route to the ultimate truth.
Many were to opt for the same way out. It was in Sufism that ʿUmar Khayyām
tried to find certainty, inspired by al-Ghazālī,108 but apparently with consider-
ably less success: he was deemed to have remained in perplexity on the basis
of his quatrains.109 It was on “unveiling” and direct experience (al-kashf wa-l-
dhawq) that Yaḥyā al-Suhrawardī, a former Peripa|tetic, based his philosophy of 38
illumination. One had to start by observing the spiritual realities, he said, then
build up to the divine sciences: whoever did it differently would remain a prey
to doubt.110 Reason produced a thousand explanations but ultimately it just
produced doubt, ʿAṭṭār agreed: knowledge of God was better reached through
the heart and the soul.111 In short, as a two-level religion, Sufism was increas-
ingly to supplant and absorb the systems developed by the philosophers and
the Ismāʿīlīs.
108 He reproduces Ghazālī’s four groups in his Risālat al-wujūd; see the translation in S.H. Nasr,
Islamic Cosmological Doctrines, second ed. (Albany, 1993), 20, and the preceding note.
109 Najm al-Dīn Rāzī Dāya, The Path of God’s Bondsmen, tr. H. Algar (New York, 1982), 54 and
n. 10.
110 Al-Suhrawardī, Ḥikmat al-ishrāq, ed. and tr. J. Walbridge and H. Ziai (Provo, 1999), xvii, 4.
111 H. Ritter, Das Meer der Seele (Leiden, 1978), 79 f.
chapter 8
What Are Prophets For? The Social Utility of

Religion in Medieval Islamic Thought*
Why do humans have religion? Many years ago I was surprised to discover that
there were people in both antiquity and the Islamic world who thought they
knew why. That is what I shall talk about here, or rather I shall talk about the
Muslim case, with occasional reminders of the Greek precedent.
So let me start with al-Jāḥiẓ, a famous litterateur and theologian who died
in 868. He tells us that human beings need a God-given law in order to sur-
vive. He notes that there is a big difference between what he calls “original
nature” (al-ṭabʿ al-awwal) and acquired habit which, as he says, becomes sec-
ond nature (ṭabʿan thāniyan).1 As regards our original nature, he explains that
God has given all living beings a strong desire to secure benefits for them-
selves and to avoid harm. That is built into all of them, humans and animals
alike, he says,2 but he only discusses the case of humans. It is in the nature
of the self to crave wealth and ease, power, influence, high status and so on,
and if God left people alone to follow their own natural habits,3 the result
would be disastrous, for there would be nothing but rivalry. There would be
no mutual affection or kindness (al-tabārr), and without that, there could
be no society: people would stop reproducing, and mankind would die out.4
But God knew that mankind would not be able to have any social life with-
out discipline (taʾdīb), so he issued commands and prohibitions—meaning a
revealed law. He also knew that his commands and prohibitions would not
* A version of this essay was presented at the University of California at Davis on March 31,
2011.
1 Jāḥiẓ, ‘Al-Maʿād waʾl-maʿāsh’, in his Rasāʾil, ed. in ʿA.-S.M. Hārūn, Rasāʾil al-Jāḥiẓ, Cairo 1964–
1979, i, 91–133, at 97. Habit was identified as second nature by Hippokrates (M. Ullmann,
Islamic Medicine, Edinburgh 1978, 57), but the saying was better known from Aris-
totle.
2 Jāḥiẓ, ‘Al-Maʿād’, i, 102: hādhā fīhim ṭabʿun murakkab wa-jibilla mafṭūra, lā khilāf bayna ʾl-khalq
fīhi, mawjūdun fī ʾl-ins waʾl-ḥayawān.
3 Law tarakahum wa-aṣl al-ṭabīʿa (ibid., i, 103; cf. 104: law tarakahum waʾl-ṭibāʿ al-awwal wa-jarū
ʿalā sunan al-fiṭra wa-ʿādat al-shīma). He has a remarkably large vocabulary for “nature”, but
it is always the nature of things, never nature in the sense of the cosmos.
4 Ibid., i, 103.

what are prophets for? 187
have any effect without reward and punishment, so he instituted hell to restrain
people from following their own desires and paradise as a compensation for
all the many things they have to renounce in this world in order to obey
him.5 In short, Jāḥiẓ is saying that God made civic life possible by giving
people laws to suppress their anti-social tendencies and by instituting par-
adise and hell as the carrot and the stick to ensure that His law would be
obeyed.
So here we have a ninth-century author wondering what a revealed law is
for. In effect, he is asking why human beings need religion, or more precisely
religion of the type variously called positive or conventional, for revealed law
(sharʿ) and positive religion (dīn) were practically synonymous concepts in
medieval Islam. What’s more, Jāḥiẓ formulates his answer in terms of functional
sociology: a religious law has certain social functions that enable human groups
to survive; it serves to curb human selfishness; it makes people sacrifice their
own individual interests for the sake of the common good. In effect, that is
also the explanation that the sociologist Durkheim offered in 1912. The reason
that Jāḥiẓ could think like a sociologist is that he shared two fundamental
presuppositions with his contemporaries.
The first is that prophets are lawgivers, not spiritual figures. (Prophets are
not actually mentioned in the epistle, but they are presupposed, as they are
the intermediaries though whom God’s law is transmitted to mankind.) Their
role is to get people together in a single community and subject them to the
same law, so that they can escape from moral, social and political anarchy. Reli-
gion means unity and order. It brings people together in the same vehicle of
salvation and makes them obey rules that enable them to travel together in
peace and quiet to their destination in this world and the next. This is mod-
elled on Muḥammad, who united the Arabs in a polity. It also fits Moses, who
organised his people for the exodus from Egypt. But it does not fit Jesus, and
modern Westerners do not usually think of religion as a synonym for law and
order either. To them, religion is first and foremost an individual relationship
with God, a source of spiritual sustenance, direction and support, and its social
functions are what the sociologist Merton called latent functions, that is to
say side-effects that people do not notice, though they may be exceedingly
important in practice. But these functions were not latent at all to Muslims
of al-Jāḥiẓ’s time, for to them, religion was first and foremost about commu-
nity formation. As a tenth-century work tells us, no religion was ever instituted
for the benefit of the individual, or as another says, religion is collective obe-
5 Ibid., i, 104 f.
188 chapter 8
dience to a single authority.6 All the so-called pillars of Islam were (and are)
collective acts: the daily ritual prayers, the weekly congregational service, the
annual fast, the pilgrimage once in a lifetime, and the charity that should be
practised at all times. These were all public, external acts that people performed
together or at least at the same time, in obedience to the ruler of the polity,
God.7
In other words, as medieval Muslims saw it, revealed religion was first and
foremost a civic religion—a religion that regulates your life as a member of a
polity and marks it out from others. God stood for the community. In worship-
ping God a society is worshipping itself, as Durkheim famously declared. He
should know. He came from a strictly observant Jewish family: God and the
community were two sides of the same coin in that tradition too.
So that was one of al-Jāḥiẓ’s presuppositions: revealed religion is civic reli-
gion. Al-Jāḥiẓ’s second presupposition was that human nature is highly anti-
social. Left on their own, people would engage in ruthless competition, nobody
would defer to anyone else, and they would destroy one another. God made
them that way: they need their strong drives in order to survive. But God also
made them incapable of living alone. They need to come together and take up
different occupations to satisfy their many needs, as many observed, including
al-Jāḥiẓ himself in another work.8 So humans are both social and deeply anti-
social animals, and they could not resolve that contradiction on their own. If
God had left them alone in what Westerners call a state of nature, they would
have perished. In short, al-Jāḥiẓ tells us that revealed religion exists because it
is eminently useful, indeed indispensable for social life. Brought by prophets,
it made civilisation possible.9
The Muslim view of things has often been compared with Hobbes’ contract
theory. Hobbes famously said that humans in a state of nature would fight:
their lives would be nasty, brutish and short. A contract with a sovereign solved
the problem. In Hobbes’ view, the sovereign was a human king. In the Muslim
vision, he was God.
6 P. Crone, Medieval Islamic Political Thought (American title God’s Rule), Edinburgh and New
York 2004, 393, citing al-ʿĀmirī and Rasāʾil Ikhwān al-Ṣafā.
7 “Wherever there is a general need, there the obligation is to God”, as Ibn Taymiyya put it (cited
by Crone, Medieval Islamic Political Thought, 394).
8 Cf. Crone, Medieval Islamic Political Thought, 260, 341f., citing both al-Jāḥiẓ and many other
authors.
9 The Muslims also inherited Aristotle’s contrary view that humans are social/political animals
by nature, and they often combined them. But even those who stressed the social nature of
mankind tended to agree that without prophets there would be no law or government.
In fact, the two contract theories probably share a remote ancestor in Dem-
ocritus. According to Democritus, followed by Epicurus, humans originated
without language, love, altruism, group solidarity and so on; they had painfully
and gradually worked their way to civilisation, but it all rested on convention
and institutions, not on their inbuilt nature, and it could collapse at any time.
As a third-century Epicurean said:
Those who have established the laws, customs, kings and magistrates in
cities have placed our life in the greatest security and peace and done
away with trouble. But if one did away with all that, we would live the
life of beasts, and everyone would eat anyone if they could.10
That’s exactly what al-Jāḥiẓ is saying, except that he did not think that humans
had done it on their own: without divine intervention, it would never have
happened.
Or take Critias, Plato’s uncle. According to him, human beings originally
lived like beasts, ruled by force, without any reward for good men or punish-
ment for the wicked. That’s the Democritan view we have just met. According
to Critias, humans eventually established laws so that they would be ruled
by justice; but people would still commit crimes when they were alone, so a
wise man hit on the idea of fear of the gods: he told people about immor-
tal divinities who see and hear everything we do, even when we are alone.
In short, Critias held that an ancient lawgiver had invented the gods to curb
people’s anti-social behaviour.11 Critias’ gods punished people by means of nat-
ural phenomena such as thunderbolts, not requital in the hereafter, but that
was soon added. Polybius, for example, said that the common people were
fickle, full of lawless desires and violent passions, so the only way to keep
them in check was by religion; if all men could be philosophers, it might not
be necessary, but they couldn’t, so the ancients were very wise when they
introduced beliefs about the gods and punishments in Hades.12 This argument
10 Plutarch, ‘Reply to Colotes’, 30, 1124d.

11 Sextus Empiricus, Against the Physicists, i, 54. Compare Aristotle, Metaphysics Lambda,
1074b, on how some myths had been added “with a view to the persuasion of the multitude
and to its legal and utilitarian expediency”.
12 Polybius, 6.56 (cf. P.A. Brunt, ‘Laus imperii’, in Imperialism in the Ancient World, ed.
P.D.A. Garnsey and C.R. Whittaker, Cambridge 1978, 166). According to Polybius, it was
scrupulous fear of the gods that kept the Roman commonwealth together. By this he meant
that fear of the gods had an extraordinary effect on Roman behaviour, not that the gods
were rewarding the Romans for their observance; but nor is he lending support to the view
190 chapter 8
was extremely widespread, in both Greek and Latin, and it lived on in new
forms in Islam, though one can only speculate as to how it had been transmit-
ted.13
In al-Jāḥiẓ’s rendition the argument has changed in two important ways.
First, the ancient sages and lawgivers have turned into prophets: they no longer
invent myths about the gods and Hades, but rather bring messages from God
about paradise and hell. What they say is true. Secondly, we all need to be
restrained by laws, we all need reward and punishment, not just the ignorant
masses. That’s also true of the Democritan tradition, but not usually otherwise.
In al-Jāḥiẓ, however, monotheism has done away with the sharp distinction
between a philosophically trained elite and common people;14 and it is this
monotheist reworking that transforms the old argument into good sociology,
for now it has become an explanation of how a whole society works—not just
an elitist argument about the management of the masses.
So the lawgivers now bring true messages from God, but for all that al-
Jāḥiẓ has not the slightest compunction about explaining religion in utilitarian
terms. To him, it merely goes to demonstrate God’s providence: everything God
does for us is for the best; God gave us competitive natures and God gave us
the religion to keep our competitive natures under control. One encounters
this view elsewhere as well. The theologian al-Māwardī (d. 1058), for exam-
ple, tells us that revealed religion keeps people in order by getting to dominate
people’s inner lives, so that they feel ruled by it even when they are alone.15
That’s what Critias said, but al-Māwardī doesn’t doubt that the revelation is
that “all elite Romans were complete sceptics who were in a conspiracy to deceive other
sections of the population” (pace J.A. North, Roman Religion, Oxford 2000, 30, cf. 77). What
impresses him is the cohesion that fear of the gods induces and the sacrifices that every-
body will make for the sake of the common good, not the manipulation of the masses.
In fact, it is not only the masses that the Roman gods affect: it is everybody, including
unphilosophical members of the elite. All had to be virtuous, only the means differed.
Polybius wishes that the unphilosophical common people would fear religion as much
back home in Greece, where those who do not cultivate philosophy have no virtue at
all.
13 You can follow it down to the third century ad, then it disappears. The Christians derided
it as an example of how the nasty pagans had deliberately lied to the masses so as to
exploit them. The Christians did not think of their own religion in utilitarian terms. How
the Muslims came to do so is unknown.
14 God is so infinitely greater than human beings that the differences among us cease to
matter: we all turn into the same tiny, fallible specks. And God’s revelation is so infinitely
above anything that human reason can work out that philosophy is neither here nor there.
15 Al-Māwardī, Adab al-dunyā waʾl-dīn, ed. M. al-Saqqā, Cairo 1973, 136 (ed. Beirut 1987, 133).
true. By contrast, when Durkheim discovered the social functions of religion

he felt that he had unmasked positive religion as a purely human creation.
But then metaphysical truth and social utility had grown up in different com-
partments in the West and were not compatible. They still aren’t, as one can
see in the interminable debates whether this or that is really religious or just
political.
But it has to be admitted that al-Jāḥiẓ’s argument is not entirely watertight.
What about animals and infidels? al-Jāḥiẓ explicitly says that God has endowed
animals and humans with the same inbuilt nature. So why do animals have
social lives without religion of any kind? Why have they not gone extinct for
lack of cooperation? You might have expected him to consider that problem.
He wrote a whole book about animals, frequently adduces animal parallels
to human features and had a colleague who said that animals did have reli-
gion. According to this colleague, Ibn Khābiṭ, all animals received prophets
from their own species and were punished and rewarded by reincarnation in
animal or, eventually, human form in accordance with their deeds until they
reached salvation. But al-Jāḥiẓ just ridicules him.16 That didn’t make the prob-
lem go away. Less than a century later you have the famous Iranian philosopher
al-Rāzī—Rhazes to the Latin Christians. He did not think that animals had
prophets, but he argued that since animals managed without them, humans
could do so too: they had no genuine need of prophets; wittingly or unwit-
tingly, the prophets were actually frauds.17 This view takes us from one extreme
to the other. There was a third variant, represented by a tenth-century Shi-
ite group. They accepted both that animals don’t have prophets and that we
humans do, but they hankered for the day when we humans wouldn’t need
positive religion any more. They wanted natural religion, like the animals as
they imagined them. They told a fable in which the animals praise God and do
16 J. van Ess, Theologie und Gesellschaft im 2. und 3. Jahrhundert Hidschra, Berlin and New
York 1991–1997, iii, 430 ff.; cf. P. Crone, ‘Al-Jāḥiẓ on aṣḥāb al-jahālāt and the Jahmiyya’,
in Medieval Arabic Thought: Essays in Honour of Fritz Zimmermann, ed. R. Hansberger,
M. Afifi al-Akiti and C. Burnett, London and Turin 2012 [Ed.: reprinted in P. Crone, The
Iranian Reception of Islam: The Non-Traditionalist Strands, vol. 2 of Collected Studies in
Three Volumes, ed. H. Siurua, Leiden 2016, art. 8], 27–40, at 34ff., here in connection with
the claim that everything, even animals and stones, was rational, though he does mention
Ibn Khābiṭ for his views on animal prophets too.
17 For his views in brief, see P. Crone, ‘Post-colonialism in tenth-century Islam’, Der Islam 83,
2006, 2–38 [Ed.: included as article 7 in the present volume], at 4–6 (with references);
for a longer treatment, see S. Stroumsa, Freethinkers of Medieval Islam, Leiden 1999,
ch. 2.
192 chapter 8
the right things of their own accord, without law, scriptures, mosques, religious
scholars, prayers, fasting or any of the paraphernalia we need to achieve some
kind of decency.18 These Shiites were suffering from the discontent of civilisa-
tion. But al-Jāḥiẓ doesn’t seem to have devoted any thought to the contrasting
cases of animals and humans.
What about infidels, then? Many infidels managed to live social lives with-
out a revealed law or belief in reward and punishment after death. Al-Jāḥiẓ did
actually take some note of that, for in another epistle he has a Turkish chief in
Central Asia and a Muslim general compare the relative merits of manmade
and revealed law. Here he simply has the general say that you Turks have man-
made law, law based on reason; we Muslims prefer a revealed law. He doesn’t
say there could be no community without it. So he is being inconsistent. But the
Turks were tribesmen, a bit like the Arabs before the rise of Islam: good fighters,
but not civilised. al-Jāḥiẓ did not know of any civilised people who lived without
a religious law. And he certainly didn’t know of whole societies without belief
in reward and punishment after death. So the counter-examples didn’t weigh
on his mind. It was not until the fourteenth century that it was pointed out by
two quite different thinkers, Ibn Taymiyya and Ibn Khaldūn, that it was per-
fectly possible to form a polity without a revealed law, and that many peoples
had done so. Eventually, this was to become all too well known, for the peo-
ples in question included the Europeans, and it was when they rose to world
dominance that the idea of a purely manmade law and political order acquired
major importance.
So far, so good, but not everybody held that positive religion was both true
and useful. Some said that it was neither true nor useful—that was the old
Epicurean view—and still others said it was useful all right, but not true—the
same view as in Critias.
Neither True nor Useful
Of those who dismissed positive religion as neither true nor useful, the earliest
were the so-called Dahrīs, who seem to have existed in Iraq well before the con-
quests, but who were at their height in al-Jāḥiẓ’s time, the ninth century. There
was a bewildering variety of them.19 All denied the creation ex nihilo; many
18 Rasāʾil Ikhwān al-Ṣafā, Beirut 1957, epistle 22; ed. and tr. L.E. Goodman and R. McGregor,
Oxford 2009.
19 See P. Crone, ‘Ungodly cosmologies’, in The Oxford Handbook of Islamic Theology,
denied that there was a creator at all or at least explained the creation without
recourse to such a figure; the most extreme of them dismissed the entire meta-
physical realm along with the creator: there were no angels, demons, prophets,
revelation, holy law or scriptures, and no afterlife of any kind either. All this
is fascinating, but what did they say about why we have religion? We don’t
know, though there are suggestions that some of them regarded the prophets as
tricksters: they credited them with knowledge of the astrological and medical
sciences.20
They are quoted as speaking about positive religion in the same dismissive
tone as Richard Dawkins or Daniel Dennett, and they may not have given any
more thought to the question why something so stupid (in their view) should
have won close to universal acceptance than the latter do. There was also an
interesting set of people about whom, unfortunately, we only have a couple of
lines. They were creationists all right, but they didn’t think the universe had
a ruler any more. “We see people fall into water without being able to swim,
or into fire, and call upon the provident maker (al-ṣāniʿ al-mudabbir), but he
does not rescue them, so we know the creator is non-existent (maʿdūm)”. One
group explained that after completing the world and finding it good the creator
had destroyed himself so as not to ruin his handiwork, leaving behind the laws
(aḥkām) current among the created things and living beings. Another group
held that, rather, a tawalwul had appeared in the essence of the creator and it
had sucked all his power and light out of him and into itself: that tawalwul was
the world, and all that remained of the creator was a cat (sinnawr), which would
suck the light out of this world again so that eventually he would be restored;
meanwhile he was too weak to attend to his created beings; their affairs were
left unattended with the result that injustice had spread.21 A third group agreed
that all the divine power had gone into the world, but envisaged the process as
a dispersal of particles rather than light. Or he had run out of energy in some
other way and was now too feeble to do anything. I don’t know what tawalwul
means, but it is clearly some kind of defect, and I suspect it is a medical term:
the world was some kind of parasitic growth on the creator and had reduced
ed. S. Schmidtke, Oxford 2016 [Ed.: included as article 6 in the present volume]. P. Crone,
‘The Dahrīs according to al-Jāḥiẓ’, Mélanges de l’ Université Saint-Joseph 63, 2010–2011, 63–
82 [Ed.: article 5 in the present volume], does not cover all of them.
20 Cf. al-Jāḥiẓ, ‘Ḥujaj al-nubuwwa’, in his Rasāʾil, ed. ʿA.-S.M. Hārūn, iii, 263f., on why Muḥam-
mad cannot be dismissed as an astrologer; Māturīdī, Taʾwīlāt, xvii, ed. A. Vanlioğlu and
B. Topaloğlu, Istanbul 2010, 400.ult. (ad q. 114:4–6).
21 Yaḥyā b. Bashīr b. ʿUmayr al-Nihāwandī (wr. before 377/987f.) in Ibn al-Jawzī, Talbīs Iblīs,
ed. M.M. al-Dimashqī, Cairo 1928, 46 (ch. against the falāsifa).
194 chapter 8
him to such a feeble state that he might as well not exist. So why did these
people think that people accepted positive religion? We don’t know.
There were also people who believed in God all right, just not in prophets.
We know several of them by name, but we only have details about two of them.
The first is Ibn al-Rāwandī, who flourished in Baghdad a bit after al-Jāḥiẓ. He
began as a theologian of the same theological school as al-Jāḥiẓ, and there is
general agreement that he was brilliant. He was too clever for his own good,
they said. I suspect he had also had too many disputations with the Dahrīs,
for a fair number of those who argued with them ended up by going off the
rails one way or the other. In any case, at some point Ibn al-Rāwandī started
writing highly offensive books attacking the prophets, especially Muḥammad,
and the Quran. The extant fragments have the scoffing tone of the Dahrīs. Then
he wrote more books refuting all his outrageous works, though he died before
he had finished the task. Maybe he was a sceptic trying to prove that for every
argument in favour of something there was another against it; in short, you
could not know anything. We don’t know.
In any case, in his outrageous books he said that there was no need for rev-
elation, for either it was in conformity with reason, in which case it was super-
fluous, or else it was contrary to reason, in which case it was false. So why did
people believe in prophets? Because they were duped. The so-called prophets
were frauds who used trickery and sorcery to produce their alleged miracles.
They knew about the powers of magnets, for example; their predictions were
of the kind that any half-decent astrologer could come up with; some of the
alleged miracles could be dismissed because of the small number of witnesses,
and others simply did not make sense: for example, if angels assisted the believ-
ers in the battle at Badr, where the believers won, where were they at Uḥud,
where the believers lost?
What motivated the men who claimed to be prophets? Ibn al-Rāwandī does
not say, but others did: the so-called prophets were after power, money or both.
That was a very old explanation of other people’s false prophets. What was
unusual about Ibn al-Rāwandī is that he applied it to all the prophets, including
his own. He and his likes denied the whole category. There was not and could
not be any such thing as a genuine prophet.
The other person we have some details about you know already: the Iranian
philosopher and medical doctor al-Rāzī. Unlike the brash, offensive Ibn al-
Rāwandī, he was by all accounts a very affable and likable man, and he was not
a scoffer. But he didn’t like revealed religion, and in his case we know why. He
didn’t want civic religion to come between himself and God. True religion in his
view spoke directly to the individual, it was above communal divisions, it was
universally true for all men, and it had nothing to do with mundane polities,
law or war. He said that you reached God through philosophy, through your own
reason, and that all humans had the same ability to reach him. If God wanted to
reveal himself to mankind, why should do so to just one man? It struck al-Rāzī
as an absurd idea. God must have revealed himself to everyone, by implant-
ing the ability to reach him—by means of reason—in every human being, just
has he had implanted knowledge of what animals need to know in every ani-
mal. It followed that prophets were impostors. Their miracles were mere magic
and sleights of hand, their books contradicted each other and themselves. The
would-be prophets had been seduced by evil spirits who appeared to them in
the form of angels: in other words, in Rāzī’s view they were deluded, and later
generations only believed them because they had been reared on such beliefs
since childhood. We don’t know how he explained why the prophets’ contem-
poraries had followed them, but others said that they, too, were after power and
money.
It was a widespread view at the time, in part because there were so many
competing politico-religious leaders, all with their own religious messages and
preachers who would wheedle money out of people with their stirring sermons.
Al-Maʿarrī, an eleventh-century Syrian who also held that true religion rested
on reason, gives us this example:
For his own sordid ends

The pulpit he ascends,
And though he disbelieves in resurrection,
Makes all his hearers quail
Whilst he unfolds a tale
Of Last Day scenes that stun the recollection
And here is his most famous verse:
They all err—Muslims, Christians, Jews and Magians [i.e. Zoroastrians].

Two make humanity’s universal sect:
One man intelligent without religion, and one religious without
intellect.22
You hear more about such cynical views of prophetic religion from the famous
theologian al-Ghazālī (d. 1111). He wrote, among other things, against those
22 Al-Maʿarrī, ed. and tr. in R.A. Nicholson, ‘The meditations of Muʿarrī’, in his Studies in
Islamic Poetry, Cambridge 1921, nos. 128, 239.
196 chapter 8
who said that hell was invented to scare people and that everything said about
paradise was just blandishment, to make people behave.23 So we are back
with Critias and Polybius, except that we don’t know whether al-Ghazālī’s
opponents held religion to be useful.
In short, a fair number of people held that the entire religious institution was
a giant fraud, created by scheming tricksters who were after power and money.
Useful but Not True
Finally, we have those who held that religion was useful, but not true. Most or
all of them seem to have been Iranians. Round about 900 we hear this about an
obscure sect in eastern Iran:
They claim that it is impossible (muḥāl) that God should send a mes-
senger to mankind from among themselves; rather, Muḥammad was a
sage (ḥakīm) who copied this book about [sic; from?] the remains of the
ancients to be of use for people’s lives/livelihoods.24
In other words, Muḥammad made up a book to provide the shared norms

which enable people to have peaceful dealings with each other—get married,
inherit, engage in commercial transactions and so forth. Muḥammad was a
wise man who had instituted the law, just like the ancient lawgivers that Critias
and Polybius talked about. Apparently, these sectarians were Sufis of some
kind, so they probably saw true religion as spirituality.
In a related vein the philosopher al-ʿĀmirī (d. 381/996) tells us that there were
people, clearly in Iran, who dismissed all religions as consisting of nothing but
legal rules from which everyone picked what enabled them to provide for their
material needs. If these religions were true, they said, they would not resort to
revelation (tawqīf ).25 And in 1066, the year in which the Normans conquered
England, an Iranian called Asadī wrote an epic which mentions the Dahrīs. He
said that there were two kinds of them, and he credited one kind with a softer
view of revealed religion than they had had in the past. Their view was that
23 Al-Ghazālī, Kimiyā-yi saʿādat, ed. Ḥ. Khadīvjam, Tehran 1380, i, 113; cf. id., ‘Die Streitschrift
des Ġazālī gegen die Ibāhīja’, ed. and tr. O. Pretzl, Sitzungsberichte der Bayerischen Akade-
mie der Wissenschaften, Phil.-hist. Abt., 7, 1933, 23 = 46.
24 Abū Muṭīʿ Makḥūl al-Nasafī, ‘Kitāb al-radd ʿalā ʾl-bidaʿ’, ed. M. Bernand, Annales Islam-
ologiques 16, 1980, 111.
25 Al-ʿĀmirī, K. al-Iʿlām bi-manāqib al-islām, ed. A.ʿA.-Ḥ. Ghurāb, Cairo 1967, 101.
every now and again a wise man appears and shows another religion and
road. “I am sent by God”, he says, “from the creator: all the things he says
I bring to you”. He puts in front of people a hell and a paradise, so that
everyone may think about his work.26
There’s no scoffing here. The prophet is a wise man who deceives people for
altruistic reasons. The same view reappears in a Persian heresiography written
in the fourteenth or fifteenth century, probably in Tabriz. Here some people say
that
the prophets and messengers were intelligent and learned men … and
philosophers (ḥakīmān) who … used wisdom out of mercy for people.
They made a law and rules (for use) among people, called it Sharīʿa and
said it was God’s decisions, and they formulated wisdom, saying it was
God’s speech, to make it more effective.
Again, the prophets are actually human lawmakers, but they credit their law
to God to lend authority to it; their motives are entirely altruistic. How do we
know that they were not sent by God? Because
the people of the earth are much too puny for a messenger to come
to them from heaven. As has been proved in the science of astronomy,
the body of the sun and the width of its disc are 7 times 7,000 by 7,000
parasangs. … As has been said:
The earth from the view of this coloured glass ceiling (the sky)
is like a poppy seed on the surface of the sea.27
It is astronomy which has bred unbelief again. Astronomy shows you that we
humans are nothing on the scale of the universe. These people did not doubt
the existence of God, but they saw him as far, far removed from us and our
affairs. Religious law was just a human institution for the regulation of worldly
affairs; but they clearly regarded it as useful.
26 Asadī, Garshāspnāma, ed. Ḥ. Yaghmaʿī, Tehran 1938, 139; tr. H. Massé, Paris 1951, ii, 40.
27 Anon., Haftād u sih millat, ed. M.J. Mashkour, 2nd printing, Tehran 1962, sect no. 35. The
opening lines (in Arabic) are taken from Abū Muṭīʿ Makḥūl al-Nasafī, cited above, note 24.
198 chapter 8
Taqiyya
How did people who held such negative views of positive religion manage
to coexist with ordinary believers? Well, in the same way that they had in
antiquity, by participating in the public religious cult and keeping discreet
about their real beliefs—practicing taqiyya, as the Muslims called it. Even the
Epicureans, who believed that the gods had done nothing whatever to this
world, would participate in the public cult. As Plutarch said:
Out of fear of public opinion he goes through a mummery of prayers and

obediences that he has no use for and pronounces words that run counter
to his philosophy.
Dahrīs, al-Rāzī and their likes also practiced taqiyya. Civic religion was the
price of citizenship; you had to conform in public: if you didn’t, you would be
persecuted. You were not allowed to rock the boat that everybody was sailing
in, or you would be a traitor to your own people. But as an individual you
could pursue any truth you liked, more or less, as long as you were discreet.
You could discuss your views with likeminded individuals, in private scholarly
gatherings, to some extent even in books, because in the good old days the
masses were illiterate. There was no confession, no inquisition, no prying into
your innermost heart; what you concealed in your innermost conscience was
between you and God. In short, you could have your private convictions as long
as you behaved as if you believed in the established religion.
If you really could not keep quiet in public, you should air your views in
poetry, as al-Maʿarrī did, and/or use ambiguity. By al-Maʿarrī’s time you could
also become a Sufi holy fool and say the most outrageous things about God,
or even to him: the fool could get away with it because he had stepped out
of normality. In effect he was playing the role of court jester: he’d tell God all
the nasty truths about the way he treated mankind that ordinary people would
not dare to say. Some of Ibn al-Rāwandī’s outrageous statements lived on as the
sayings of holy fools, but usually in an affectionate tone quite different from Ibn
al-Rāwandī’s own. Some people, though, were extremely rude. The Bektashis
were among them. You could find everything among them, from the highest
mysticism to the purest atheism, with the whole range of beliefs in between.28
Alas, all this has completely changed. Practising taqiyya or expressing your-
self in poetry will not protect you any more, nor is it accepted that you have a
28 Cf. J.K. Birge, The Bektashi Order of Dervishes, London and Hartford, Conn., 1937, 87.
private space in your interior where you are alone with God and where other
humans cannot interfere. That is one respect in which modernity has made the
Muslim world a less agreeable place than it was before.
chapter 9
Oral Transmission of Subversive Ideas from the

Islamic World to Europe: The Case of the Three
Impostors*
In 1239 the good Christians of Europe were shocked by a bulletin from Pope
Gregory ix announcing that the most powerful man in Latin Christendom,
the emperor Frederick ii, had made a terrible claim, namely that Jesus, Moses
and Muḥammad were tricksters, or in other words that all the religions he
knew were false.1 This is the first mention in Europe of what European histori-
ans call “the three impostors thesis” (one allegedly earlier attestation notwith-
standing),2 but it was not the last. It crops up elsewhere in the Mediterranean
thereafter, and by the sixteenth century it was everywhere. Rumours of an
actual book called De Tribus Impostoribus generated intense interest in free-
thinking circles and much speculation about its authorship, without anyone
succeeding in finding a copy.3 But at the end of the 1680s a Latin work of
* This article was originally drafted in connection with a series of workshops on the transmis-
sion of radical ideas from the Islamic world to Europe at the Institute for Advanced Study.
I should like to thank Jonathan Israel for co-organizing the first of them together with Mar-
tin Mulsow and myself, and Martin Mulsow for co-organizing the second with me as well.
The third unfortunately had to be cancelled and no publication was produced, except in the
form of individual articles, of which this is one. I must also thank several of the participants
in the workshops, not least Thomas Gruber, who introduced me to the three impostors in
Europe; Robert Lerner, who provided important information on the same subject; and Kevin
van Bladel, who opened my eyes to Near Eastern beliefs in pre-Adamites. I am also indebted
to Michael Cook and Stefania Pastore for reading earlier drafts.
1 The classic account is by M. Esposito, ‘Una manifestazione d’incredulità religiosa nel medio-
evo: il detto dei “Tre Impostori” e la sua trasmissione da Federico ii a Pomponazzi’, Archivio
Storico Italiano, serie vii, 16, 1931, 3–48.
2 Simon of Tournai (1190s?) was accused of having voiced the impostor thesis by Thomas
Cantimpré, who wrote between 1256 and 1263 (cf. Esposito, ‘Manifestazione’, 33ff.). But since
the charge was made after Gregory’s publication of his bull against Frederick, when Simon
had been dead for many years, it is not usually taken seriously.
3 The suspected authors included Boccaccio, Pomponazzi, Pietro Aretino, Guillaume Postel,
Campanella and many others, cf. G. Ernst, ‘Campanella e il De Tribus Impostoribus’, Nouvelles
de la Republique des Lettres 2, 1986, 143–170. See also the highly informative entry ‘Trattato dei
tre impostori’ in the Italian Wikipedia.

oral transmission of subversive ideas 201
this title materialised,4 and it was soon followed by a French treatise enti-
tled Traité des Trois Imposteurs or L’Esprit de Spinoza, which counts as “one of
the most significant irreligious clandestine writings available in the Enlighten-
ment”.5
It has long been suspected that the idea, which worked so powerfully on
the European imagination, originated in the Islamic world. Medieval authors
sometimes attributed it to Averroes,6 if only because “Averroism” was the stan-
dard rubric to which heresies suspected of Islamic origins were assigned, and
early modern authors also thought it might be a Muslim theme.7 In fact, Aver-
roes had nothing to do with it, nor was there a book on the subject before Euro-
pean freethinkers took it upon themselves to produce one; but it was indeed
Muslims who had developed the subversive idea. In what follows I briefly sur-
vey the history of this idea up to the time of its transmission to the West
and examine the channels of transmission, arguing that more than one was
involved.8
4 The date (late 1680s) and the author (Joachim Müller, professor of law at Hamburg) were
established by W. Schröder in his introduction to his edition of De impostoris religionum (De
tribus impostoribus), Stuttgart 1999.
5 For all this, see S. Berti, F. Charles-Daubert and R.H. Popkin (eds.), Heterodoxy, Spinozism, and
Free Thought in Early Eighteenth-Century Europe: Studies on the Traité des Trois Imposteurs,
Dordrecht and Boston 1996. The French work circulated in manuscript form from probably
around 1678 onwards and was clandestinely printed for the first time in 1718 in The Hague by
the Huguenot Spinozist Charles Levier (I believe I owe this information to Jonathan Israel).
By then, the impostor idea was also encountered in many other works.
6 M. Esposito, ‘Manifestazione’, 29, 31, citing the De Erroribus Philosophorum of Aegidius Ro-
manus (Giles of Rome, d. 1316), written between 1260 and 1274 (but without explicit mention
of the impostor theme), and Benvenuto da Imola (d. 1388), Commento latino sulla Divina
Commedia di Dante Alighieri, 1855, 138.
7 Cf. S. Åkerman in Berti, Chales-Daubert and Popkin, Heterodoxy, 403.
8 For literature on the question, see L. Massignon, ‘La légende “de tribus impostoribus” et ses
origines islamiques’, in his Opera Minora, ed. Y. Moubarac, i, Beirut 1963, 82–85; F. Niewöhner,
Veritas sive Varietas: Lessings Toleranzparabel und das Buch Von den drei Betrügern, Heidel-
berg 1988; D. de Smet, ‘La théorie des trois imposteurs et ses prétendues origins islamiques’,
in C. Cannuyer and J. Grand’Henry (eds.), Incroyance et dissidences religieuses dans les civili-
sations orientales, Bussels and Louvain-la-Neuve 2007, 81–93 (drawn to my attention by S. Tra-
boulsi). His title notwithstanding, de Smet does not dispute the Muslim origin of the idea. See
also F. Gunny, ‘Le traité des trois imposteurs et ses origins arabes’, Dix-huitième Siècle 28, 1996,
169–174, dealing with a motif attested in the French treatise, but not in earlier reports (drawn
to my attention by G. Paganini).
202 chapter 9
Antiquity
The ultimate roots of the impostor idea lie in classical antiquity. A prophet in
the ancient Greek world was a soothsayer or oracle, a person inspired by the
divine who had the ability to predict the future, heal and work other wonders—
in short, what the pre-Islamic Arabs called a kāhin. The Greeks, including Hel-
lenised non-Greeks, often suspected such prophets of being swindlers who
faked their apparent contact with the divine and had no genuine religious
knowledge.9 When, for example, a Syrian slave by the name of Eunus raised
a major slave revolt in Sicily (135–131bc), working miracles and making pre-
dictions, he was assumed to be a charlatan who had “deceived many” with his
magic.10 Eunus was unusual in that he used his divine inspiration to establish
himself as a political leader. Most prophets in the Greek Mediterranean served
private needs and acquired political importance only when rulers consulted
them on the outcome of the acts they were planning, as Croesus did at Delphi.
Accordingly, they were usually associated with money grubbing rather than
political ambitions. There is a memorable portrait of the prophet as a swindler
who milks the superstitious masses for money by the Syrian Lucian of Samosata
(d. after 180) in his satirical account of the prophet Alexander, a contemporary
of his.11 Another Syrian, the Cynic Oenomaus of Gadara (d. c. 120) composed a
withering critique of oracular practice combining satire, ridicule and invective
under the title Goētōn phōra, “Exposé of Charlatans” or “Detection of Impos-
tors”.12 The oracles, he said, did not proceed from a daimon or god, but were
rather “frauds and tricks of human impostors cunningly contrived to deceive
the multitude”.13
The relevance of this to the present theme begins when Greek-speaking
people unsympathetic to the Jews began to dismiss Moses as a swindler of
this type. Moses was not, of course, a diviner or soothsayer, but rather a man
whose contact with the divine had resulted in the liberation of his followers
from bondage and the revelation of a divine law—in effect the role to which
Eunus had aspired in Sicily. In Greek terms, Moses was a lawgiver (nomothetēs)
9 For early examples, see W.H.C. Guthrie, The Sophists (A History of Greek Philosophy, iii),
Cambridge 1971, 246 f.
10 Diodorus Siculus, Bibliotheca Historica, xxxiv, 2, 5–9, possibly from Posidonius.
11 Lucian of Samosata, Alexander, the False Prophet.
12 The fragments are preserved in Eusebius, Praeparatio Evangelica, starting at v, 18, 6. For
a commentary (drawn to my attention by Yannis Papadoyannakis), see J. Hammerstaedt,
Die Orakelkritik des Kynikers Oenomaus, Frankfurt am Main 1988.
13 Eusebius, Praeparatio, v, 21, 6.
and the founder of a colony. But it was as prophētēs that Hebrew nabiʾ had
been translated into Greek, and Moses had also worked miracles, so he fell
into the category of oracular soothsayer too. Josephus (1st century ce) reports
that a number of Greeks “have maligned our lawgiver Moses as a magician
(goēs) and impostor (apateōn)”;14 and there were probably Hellenised Jews
who played around with the same idea, for Philo (d. c. 50 ce) envisages the
Jews themselves as maligning Moses as a trickster during their sojourn in the
desert.15 Jesus too could be seen as a nomothetēs (in a less political sense) as
well as an oracular soothsayer; and Lucian, who refers to him as the “first law-
giver” of the Christians, implicitly places the “crucified sophist” in the same
company as the pseudo-prophet Alexander.16 The second-century Celsus (wr.
c. 180) dismissed both Moses and Jesus as magicians in his famous attack
on Christianity, claiming that both of them had learnt magic in Egypt, and
he tells the story of Jesus’ life along lines known from the Jewish polemi-
cal work Toledoth Ieshu, in which Jesus is also a magician.17 Dismissing Jesus
as a magician became a standard Jewish18 and pagan ploy;19 and the pagans
would also dismiss other Christian figures such as Paul and Peter as dissemi-
nators of deceit who owed their successes to sorcery or other trickery.20 The
14 Josephus, Against Apion, ii, 145.

15 Philo, Hypothetica, 356 (in Eusebius, Praeparatio, viii, 6, 2).
16 Lucian, The Death of Peregrinus, 5, in S. Benko, ‘Pagan Criticism of Christianity’, in H. Tem-
porini and W. Haase (eds.), Aufstieg und Niedergang der römischen Welt (hereafter anrw),
ii, xxiii.2, 1095.
17 Cf. Benko, ‘Pagan Criticism of Christianity’, 1102; cf. S. Krauss, Das Leben Jesu nach jüdischen
Quellen, Berlin 1902.
18 Cf. P. Schäfer, Jesus in the Talmud, Princeton 2007, 64, 102ff., 137. For the earliest suspicions,
see Mark 3:22. In the Acts of Pilate (Gospel of Nicodemus), i, 1; ii, 1 (in W. Schneemelcher
(ed.), New Testament Apocrypha, i, Louisville, Ky., 1991, 506f.), the Jews are envisaged as
calling Jesus a sorcerer on the basis of this passage. The Jews of third/ninth-century Iraq,
on the other hand, would sometimes laugh at Jesus’ miracles and sometimes get angry
and dismiss him as a magician (ṣāḥib ruqan wa-nīranjāt), according to al-Jāḥiẓ, ‘al-Radd
ʿalā ʾl-naṣārā’, in his Rasāʾil, ed. ʿA.-S.M. Hārūn, Cairo 1964–1979, iii, 325.ult.
19 P. Courcelle, ‘Anti-Christian Arguments and Christian Platonism: from Arnobius to St.
Ambrose’, in A. Momigliano (ed.), The Conflict between Paganism and Christianity in the
Fourth Century, Oxford 1963, 153.
20 Origen, Contra Celsum, passim (Celsus on the Christians in general); Jerome, Tract on
Psalm 81 (some said Paul did it all for money; Porphyry said that they (the disciples?)
worked miracles with magic because they were poor); Julian, Against the Christians, ed.
and tr. W.C. Wright, Cambridge, Mass., and London 1923, 340f. (100 a: Paul surpassed all
magicians and charlatans); A. Meredith, ‘Porphyry and Julian against the Christians’, in
anrw, xxiii/2, 1120 f. (Hierocles on Paul, Peter and others).
204 chapter 9
target in these attacks was not prophecy as such, but rather the authoritative
status of the particular figures in question.
The Islamic World
Casting other people’s prophets as mere magicians was a convenient way of

protecting one’s own religious institutions, and the practice continued in the
non-Greek Near East. Thus the Qurʾānic pagans dismissed Muḥammad as a
magician (though they also found fault with his miracle-making abilities),21
and they seem to have suspected him of using religion for political ends.22
When claimants to prophethood appeared thereafter, the Muslims themselves
dismissed them as magicians;23 and when holy men and Sufis emerged serving
much the same needs as the oracular prophets of the past, many regarded
them too as tricksters who cleverly milked people for money.24 But it is clear
that something had changed. For one thing, the hegemonic polity of the Near
East now owed its existence to a prophet of the Mosaic type, that is a lawgiver
who had brought a law and founded a polity, and whose status was vindicated
by miracles and oracular predictions. When the Greek philosophical tradition
was translated into Arabic, the Muslims duly called Muḥammad a wāḍiʿ al-
nāmūs/al-sharīʿa, a translation of “lawgiver” (nomothetēs). That the founder of a
religion was a political figure now came to be taken for granted. Indeed, thanks
to Muḥammad’s paradigmatic status, Muslims commonly assumed that all
polities were based on a revealed law, at least in so far as they were monotheist
(a term with much the same connotations as “civilised”). Left on their own,
it was argued, people would pursue their own selfish interests and engage
in constant rivalry and strife: God had to give them rules in order for social
life to be possible, and he told them about Paradise and Hell in order to
21 q. 6:7; 34:43; 37:15; 38:4f.; 43:30; 46:7; 74:24. Moses and Jesus are envisaged as having been
rejected in the same way; see for example 5:110; 28:36; 40:24; 61:6.
22 “Have you come to turn us away from what we found our fathers following, so that you two
may become great in the land?” (10:78), as the Egyptians say to Moses and Aaron, typifying
the pagans that Muḥammad was up against.
23 Thus for example al-Muqannaʿ (cf. Encyclopaedia Iranica, s.v. ‘Moqannaʿ’).
24 Cf. al-Tanūkhī, Nishwār al-muḥāḍara, ed. ʿA. al-Shāljī, Beirut 1971–1972, i, 165ff. (no. 84, al-
Ḥallāj); ii, 324 ff. (no. 170, street astrologer), 351 ff. (no. 187, holy man), 359 (no. 190, Sufi
preacher); iii, 119 (no. 75, Junayd), 120 (no. 77, Sufi); tr. D.S. Margoliouth, The Table-Talk of
a Mesopotamian Judge, part i, London 1922, 86 ff., 277ff., 289–292, 294; parts ii and viii,
Hyderabad n.d., 180 f.
make them obey.25 Seen as indispensable to civilised life, prophets came to

occupy a much more central role in high-cultural thought than they had in
the Greek Near East even after the victory of Christianity. Dismissing such
figures as impostors was far more subversive than it had been in the Greek
world.
For another thing, the Muslims operated with an abstract concept of proph-
ethood (nubuwwa). Nubuwwa is usually translated as prophecy, and maybe this
is what it means in the Qurʾān.26 In classical Arabic, however, it stands for
the status occupied by a prophet and the mission with which he is charged,
not for the revelations (let alone predictions) that he utters; and what came
to be discussed in Arabic was the very idea that there could be such a thing
as a prophet, or in other words the proposition that God communicated with
humans by means of revelation. The Muslims developed a new literary genre
called “the proofs of prophethood” (dalāʾil al-nubuwwa), devoted to the defence
of Muḥammad’s prophetic status.27 Most works of this genre concern them-
selves with his case alone, but some begin by vindicating the concept of proph-
ethood as such,28 for already by the second/eighth century, and probably be-
fore, there were some who rejected it. The target in these attacks was not just
the authoritative status of the particular figures in question, but the very con-
cept of a prophet.
The Deniers of Prophethood, c. 750–900

(a) Dahrīs and Zindīqs
That the entire idea of prophets was false was first maintained by thinkers
within the Muslim community who went by the name of Dahrīs. They are
attested from the mid-second/eighth century onwards, with the best evidence
falling in the third/ninth, and they are presented as likeminded individuals
who occupied themselves with medicine and astronomy/astrology, above all
in Iraq.29 According to ʿAbd al-Jabbār (d. 415/1025), a Sunnī theologian with
25 Cf. P. Crone, Medieval Islamic Political Thought (American title God’s Rule), Edinburgh
2004, 261 ff.
26 q. 3:79; 6:89; 29:27; 45:16; 57:26.
27 Cf. S. Stroumsa, ‘The Signs of Prophecy: the Emergence and Early Development of a Theme
in Arabic Theological Literature’, Harvard Theological Review 78, 1985, 101–114.
28 The best known example is al-Māwardī (d. 450/1058), Aʿlām al-nubuwwa, ed. M.M.-A. al-
Baghdādī, Beirut 1987, ch. 4 (pp. 49 ff.).
29 Cf. Encyclopaedia of Islam, 2nd ed., Leiden 1960–2004 (hereafter ei2), s.v. ‘Dahriyya’ (Goldz-
iher and Goichon); Encyclopaedia Iranica, ed. E. Yarshater, London and Boston 1982–,
s.v. ‘Dahrī’ (Shaki and Gimaret); P. Crone, ‘The Dahrīs according to al-Jāḥiẓ’, Mélanges de
206 chapter 9
a hyperbolic bent, most doctors rejected the idea of prophethood, deemed

Muslims and other believers in revealed religion to be ignorant, and denied
both God (al-rubūbiyya) and the resurrection.30 Doctors and astrologers are
singled out for their Dahrī views down to at least the seventh/thirteenth cen-
tury.31 Dahrīs are described as empiricists who held that all knowledge must
be based on sense impressions and some limited forms of reasoning to the
exclusion of revelation. In terms of physics, they were materialists who held
the universe to be eternal, rule-bound, and explicable in terms of the mixture
of the four elementary qualities (hot, cold, dry, wet) or the four elements, or
in terms of accidents arising in prime matter. They came in many varieties.
Some assigned God a role in the creation, but most denied that the world had
a creator, let alone a providential ruler or judge, and some denied the very exis-
tence of God, along with that of prophets, angels, spirits, revealed scriptures
and veridical dreams. All were adamant that there was no such thing as life
after death. Dahrīs are strongly associated with other radicals called Zindīqs,
who had slightly different views and who were drawn from different religious
communities too. Whereas Dahrīs seem to have been drawn from Zoroastrian
and pagan groups, Zindīqs were Manichaeans, Bardesanites and Marcionites by
origin, and whereas Dahrīs were typically doctors and astrologers, Zindīqs were
mostly bureaucrats, courtiers and poets. All, however, had lost their ancestral
faith.32
l’ Université Saint-Joseph 63, 2010–2011, 63–82 [Ed.: included as article 5 in the present vol-
ume]. For more on the scientific views involved, see P. Crone, ‘Ungodly Cosmologies’, in
S. Schmidtke (ed.), The Oxford Handbook of Islamic Theology, Oxford 2016 [Ed.: article 6 in
the present volume].
30 ʿAbd al-Jabbār, Tathbīt dalāʾil al-nubuwwa, ed. A.-K. ʿUthmān, Beirut 1966, i, 62.
31 Al-Masʿūdī, Murūj al-dhahab, ed. C. Pellat, Beirut 1966–1979, iii, §1846; al-Jāḥiẓ, Kitāb
al-ḥayawān, vi, 269 f.; vii, 12 f.; cf. al-Jāḥiẓ, ‘al-Radd ʿalā ʾl-naṣārā’, in his Rasāʾil, iii, 313f.,
320 f.; al-Maʿarrī, Luzūmiyyāt, ed. Ḥ. ʿAbd al-Majīd and others, Cairo 1992–1994, iii,
no. 1074; al-Ghazālī, Kimiyā-yi saʿādat, ed. Ḥ. Khadīvjam, Tehran 1380, i, 57, 65; Gaon
Shmuel ben Eli (d. 1195), Treatise on the Resurrection of the Dead, Hebrew tr., quoted in
S. Stroumsa, ‘Twelfth-Century Concepts of Soul and Body: the Maimonidean Controversy
in Baghdad’, in A.I. Baumgarten, J. Assmann and G.G. Stroumsa (eds.), Self, Soul and Body
in Religious Experience, Leiden 1998, 317; Maimonides in S. Stroumsa, ‘“Ravings”: Mai-
monides’ Concept of Pseudo-Science’, Aleph 1, 2001, 146. Note also the claim that Abū
Maʿshar studied astronomy until he ‘turned godless’ (ḥattā alḥada, Tanūkhī, Nishwār, iv,
66).
32 Cf. M. Chokr, Zandaqa et zindīqs en Islam au second siècle de l’hégire, Damascus 1993;
Crone, ‘Ungodly Cosmologies’; cf. also ead., The Nativist Prophets of Early Islamic Iran,
Cambridge 2012, index, s.v. ‘Dahrīs’.
Both Dahrīs and Zindīqs aired their views in disputations, the dominant
vehicle of intellectual pursuit at the time, and we know about them only
from their opponents, above all the Muʿtazilite theologians who shared their
interest in cosmology while trying to refute their views on metaphysics. Their
opponents’ responses centred on the evidence for design in nature, from which
one could infer the existence of a creator God and providence, not on the
existence of prophethood, since it went without saying that without belief in
God one could not believe in prophethood either. Prophethood must have been
discussed as well, however, for we know that Dahrīs and Zindīqs had rational
explanations for the alleged miracles of the prophets: they credited them with
knowledge of the astrological and medical sciences.33
(b) “Brahmans”
There were also ninth-century thinkers who believed in God while rejecting the
idea of prophets. In the terminology of the Enlightenment they were deists. The
theologians called them “Brahmans” and presented them as good monotheists
who denied the existence of revelation on the grounds that one could reason
one’s way to God and proper behaviour alike.34 There are also reports in which
the Barāhima accept one prophet, Adam, or two, Adam and Abraham. No
genuine knowledge of the Brahmans of India is reflected in the information
about them, and there can be little doubt that the label was used as a cover
for views which originated in the Islamic world itself. Of one thinker, Ibn al-
Rāwandī (d. mid-ninth century or later) we are explicitly told that he used the
Brahmans as his mouthpieces.35
Ibn al-Rāwandī was one of a fair number of the Muʿtazilites who were suf-
ficiently affected by the arguments of their Dahrī and Zindīq opponents for
their own faith to be shaken.36 He seems to have become a Skeptic, in the
technical sense of a believer in the principle of the equipollence of proofs (isos-
theneia/takāfuʾ al-adilla), according to which every argument in favour of one
view could always be balanced by another against it.37 In the Greek world, this
33 Cf. al-Jāḥiẓ, ‘Ḥujaj al-nubuwwa’, in his Rasāʾil, iii, 263f., on why Muḥammad cannot be
dismissed as an astrologer; al-Māturīdī, Taʾwīlāt al-Qurʾān, xvii, ed. A. Vanlioğlu and
B. Topaloğlu, Istanbul 2010, 400.ult. (ad 114:4–6).
34 Cf. Encyclopaedia of Islam, 3rd ed., Leiden 2008– (hereafter ei3), s.v. ‘Barāhima’ (Crone).
35 Cf. S. Stroumsa, Freethinkers of Medieval Islam, Leiden 1999, 48.
36 Cf. ei2, s.v. ‘Ibn al-Rāwandī’ (Kraus and Vajda); Stroumsa, Freethinkers, ch. 2; J. van Ess,
Theologie und Gesellschaft im 2. und 3. Jahrhundert Hidschra, Berlin and New York, 1991–
1997 (hereafter tg), iv, 295 ff.
37 Cf. R.J. Hankinson, The Sceptics, London and New York 1995, 27. On the principle in
208 chapter 9
principle had applied to all knowledge claims about things which are not open
to immediate perceptual inspection, not specifically to metaphysics; but in the
Muslim world it was always in connection with claims rooted in revelation that
the principle was used.38 How far Ibn al-Rāwandī’s Skepticism went we do not
know, but like Protagoras of old and other Skeptics, he would write for and
against the same position. Unlike his predecessors, however, he wrote his attack
and defence in separate books. In his Zumurrud (“Emerald”), attributed to the
Brahmans, he argued that there was no need for prophets, since their message
would be either in conformity with reason, in which case it was superfluous, or
contrary to it, in which case it had to be rejected. All prophets were imposters
and their miracles were mere trickery (makhāriq) and sorcery:39 for example,
they knew about the powers of magnets;40 their predictions were of the kind
that any half-decent astrologer could come up with;41 some of the alleged mira-
cles could be dismissed because of the small number of witnesses,42 and others
simply did not make sense—for example, if angels assisted the believers in the
battle at Badr, where the believers won, where were they at Uḥud, where the
believers lost?43 He also wrote a book in criticism of Muḥammad called al-Farīd
(“The Unique”) and yet another book, entitled al-Dāmigh (“The Brainbasher”),
in which he discussed the inconsistencies in the Qurʾān which proved it not to
Islamic thought, see J. van Ess, Die Erkenntnislehre des ʿAḍudaddīn al-Īcī, Wiesbaden 1966,
221 ff., with Ibn al-Rāwandī at p. 223; cf. also index, s.v. ‘Skepsis’; id., ‘Skepticism in Islamic
Religious Thought’, al-Abḥāth 21, 1968, 1–18, with Ibn al-Rāwandī at p. 7.
38 Cf. Ibn Ḥazm, al-Faṣl fī ʾl-milal waʾl-ahwāʾ waʾl-niḥal, Cairo 1317, v, 119f. For a relatively mild
case, see P. Crone, ‘Abū Saʿīd al-Ḥaḍrī and the Punishment of Unbelievers’, Jerusalem Stud-
ies in Arabic and Islam 31, 2006, 92–106 [Ed.: included as article 4 in the present volume].
39 Stroumsa, Freethinkers, 76–86; cf. the texts in tg, vi, 457ff.
40 Ibn al-Jawzī, al-Muntaẓam, Hyderabad 1357–1359, vi, 100f. (year 298); also in H. Ritter (ed.
and tr.), ‘Philologika. vi: Ibn al-Ǵauzīs Bericht über Ibn ar-Rēwendī’, Der Islam 19, 1930, 4
= 12; cf. al-Māturīdī, Kitāb al-tawḥīd, ed. F. Kholeif, Beirut 1970, 186, where the argument is
credited to Abū ʿĪsā al-Warrāq; tg, vi, 474 f., with discussion.
41 Ibn al-Jawzī, Muntaẓam, vi, 101; Ritter, ‘Philologika’, 4 = 12 (regarding a prediction by
Muḥammad); cf. al-Warrāq in Māturīdī, Tawḥīd, 195.17 (corrupt); ʿAbd al-Jabbār, Tathbīt, ii,
413.11; tr. Stroumsa, Freethinkers, 63 and note 104 (regarding Moses’ and Jesus’ prediction
of Muḥammad).
42 Al-Muʾayyad in P. Kraus, ‘Beiträge zur islamischen Ketzergeschichte’, Rivista degli Studi
Orientali 14, 1933, 101 = 113 (no. 7). Even many people could agree on a falsehood, such as
that Jesus had been crucified (Muʾayyad in Kraus, ‘Ketzergeschichte’, 104 = 115 (no. 12); cf.
al-Warrāq in tg, vi, 479–481).
43 Cf. Kraus, ‘Ketzergeschichte’, 105 f. = 115 f. (no. 13); attrib. al-Warrāq in Māturīdī, Tawḥīd, 199;
tr. tg, vi, 477 f.
be divine,44 picking out verses also used by Zindīqs and Dahrīs.45 He then wrote
a book against the Zumurrud, as well as a book in proof of prophethood, and a
Naqḍ al-Dāmigh (“Refutation of the Brainbasher”), which he did not complete,
presumably because he died.46
That Ibn al-Rāwandī was a Skeptic, or even a heretic, has recently been
denied with reference to the fact that the tenth-century theologian al-Māturīdī
(d. 333/944) quotes him as defending Islam against the very views he had
voiced in the Zumurrud. In al-Māturīdī’s work he does not acknowledge the
subversive views as his own, but rather attributes them to Abū ʿĪsā al-Warrāq,
a slightly earlier Muʿtazilite of dubious repute with whom he had been person-
ally associated.47 To Van Ess, this shows that Ibn al-Rāwandī was the victim of
a black legend.48 Van Ess does not dispute that Ibn al-Rāwandī wrote books
containing outrageous views, but he thinks that he always credited these views
to their real authors (such as Abū ʿĪsā al-Warrāq) and presented them sim-
ply to demonstrate the impossibility of proving anything: his aim was to cast
doubt. Yet he was not a Skeptic: according to Van Ess, the fact that he eventu-
ally wrote refutations of these views shows that he knew where the solution
lay.
It is difficult to agree. For one thing, the black legend hardly amounts to more
than a normal reaction: a man who devoted whole books to the presentation of
outrageous propositions in a strikingly impudent tone without refuting them
44 As Van Ess (Erkenntnislehre, 223) notes, the title of this book is curiously reminiscent
of Protagoras’ Kataballontes, “The Knocker-down”. There is another Skeptical work of a
similar title in the Tattvopaplavasimha, “The lion destroying all principles”, written c. 800
(drawn to my attention by Michael Cook; cf. B.-A. Scharfstein, Comparative History of
World Philosophy, Albany 1998, 252 f.).
45 Compare the questions in Ibn Ḥanbal (d. 241/855), al-Radd ʿalā ʾl-Zanādiqa waʾl-Jahmiyya,
Cairo 1393, 8 ff., and Ibn al-Rāwandī’s Kitāb al-dāmigh in Ibn al-Jawzī, Muntaẓam, vi, 99ff.;
H. Ritter, ‘Philologika’, 2 ff. Unfortunately, it is not clear that the Radd is actually by Ibn
Ḥanbal, so some of the objections attributed to the Zindīqs could in principle be Ibn al-
Rāwandī’s (e.g. Radd, 8; Muntaẓam, vi, 103; Ritter, ‘Philologika’, 7 = 15, on 4:56).
46 Ibn al-Nadīm, al-Fihrist, ed. R. Tajaddud, Tehran 1971, 216f.; cf. tg, vi, 434ff., nos. 20, 34, 36,
38, 40, 42.
47 On him, see ei2, s.v. (Stern); Van Ess, tg, iv, 289–294; Stroumsa, Freethinkers, 40ff. The
idea, present in some sources, that Abū ʿĪsā had written or contributed to the Zumurrud,
is presumably a result of this ploy, cf. Stroumsa, Freethinkers, 75 (with a different expla-
nation). Ibn al-Nadīm does not credit Abū ʿĪsā with any works against prophethood, nor
does anyone say that he used the Brahmans as his cover.
48 J. van Ess, ‘Ibn ar-Rēwandī, or the Making of an Image’, al-Abḥāth 27, 1979, 5–26; shorter
version under the title ‘Al-Fārābī and Ibn al-Rēwandī’, Hamdard Islamicus 3, 1980, 3–15.
210 chapter 9
there and then was bound to be understood as an adherent of the propositions

in question, as Ibn al-Rāwandī must have known very well. For another thing,
if his heart was not in the outrageous arguments, what reason do we have to
believe that it was in the counter-arguments? He is simply switching sides in al-
Māturīdī, casting doubt in the opposite direction, as one would expect him to
do in his anti-works.49 Of course he may have repented (as some said he did),50
but that presupposes that he meant what he said when he was being outra-
geous. And if he was not a Skeptic, why was it so important to him to cast doubt?
Whatever his motives, Ibn al-Rāwandī was not the only theologian to air
subversive views of this kind at the time. Of another lapsed Muʿtazilite, Abū
Ḥafṣ al-Ḥaddād, we are told that he wrote a book on the equipollence of proofs
(takāfuʾ al-adilla) and held the Prophet (Muḥammad) to have used tricks to
convert his followers.51 Maybe he too was simply trying to demonstrate that
refuting the idea of prophethood was just as easy as defending it. Of yet another
theologian, Abū Isḥāq al-Naṣībī (d. c. 370/980), it was said that he had doubts
about all prophetic missions, but whether he suspected trickery we are not
told.52 Nor are we told how any of them explained the motives of the alleged
impostors, but Ibn al-Rāwandī’s tone certainly suggests that he saw them as self-
seeking. We also hear of unnamed deniers of the prophets who would dismiss
demons and jinn as invented by the would-be prophets in order to scare people
into following them.53
49 Van Ess’ interpretation originally rested on the assumption that al-Māturīdī was quoting
from the Zumurrud (an assumption shared by Stroumsa in her argument against Van Ess):
he inferred that all the outrageous views in that book had been willfully understood as
Ibn al-Rāwandī’s own, though the latter had clearly identified them as Abū ʿĪsā’s. That
Māturīdī was quoting from one of Ibn al-Rāwandī’s works refuting his own position was
first suggested by Madelung (cf. his review in Zeitschrift der Deutschen Morgenländischen
Gesellschaft 124, 1974, 150, proposing the Ithbāt al-rusul in which he affirmed the reality
of prophets). Van Ess himself has now suggested that al-Māturīdī could be quoting the
anti-Zumurrud, yet he has not changed his position (tg, iv, 343).
50 Ibn al-Nadīm, Fihrist, 216.19. This, incidentally, shows that he had his defenders in Iraq
as well. Van Ess’ idea that Ibn al-Rāwandī was maligned in Iraq and remembered as a
good theologian in Khurāsān also suffers from the fact that the quotation from al-Balkhī’s
Maḥāsin Khurāsān in Ibn al-Nadīm, Fihrist, 216.14, must include the statement that Ibn
al-Rāwandī turned heretical: it is not formulated in the phrase that Ibn al-Nadīm himself
uses of Muʿtazilites who went astray (cf. Fihrist, 215.–2; 216.5, 7).
51 For him, see Ibn al-Nadīm, Fihrist, 216; tg, iv, 89 f.
52 Al-Tawḥīdī, Kitāb al-imtāʿ waʾl-muʾānasa, ed. A. Amīn and A. al-Zayn, Cairo 1939–1944, i,
141; cf. id., Akhlāq al-wazīrayn, ed. M. al-Ṭanjī, Beirut 1992, 202, 211f., 297.
53 Māturīdī, Taʾwīlāt, xvii, 400 f. (ad q. 114:4–6).
There were others, however, who held the prophets, or at least Muḥammad,
to have practised deceit in an altruistic vein. According to the theologian
Abū Muṭīʿ al-Nasafī (d. 318/930), a certain sect, probably in Iran, regarded
Muḥammad as a wise man who had composed the Qurʾān himself in order to
make it easier for people to pursue their livelihoods, namely by giving them
shared norms and thus allowing them to live together in peace.54 This was
a new version of the idea, widespread in pagan antiquity, that the ancient
lawgivers had invented the gods and/or the punishments of afterlife in order
to curb people’s anti-social behaviour;55 and it was generally accepted in the
medieval Islamic world that religion had this effect.56 The Greeks and the
Romans had generally seen religion (or in other words positive religion as
opposed to philosophy) as either true and necessary for political order, or
else as false but still necessary.57 The vast majority of Muslims adhered to the
first view, but as al-Nasafī shows us, there were also some who adhered to
the second. In the classical world, only the Epicureans held positive religion
to be both false and unnecessary: this is the view of which Ibn al-Rāwandī’s
“Brahmans” give us a new version.
The Tenth and Eleventh Centuries

By the tenth century it was philosophers rather than theologians who were the
pioneers in thoughts about prophethood, usually as defenders of the concept,
but sometimes as its critics. Al-Sarakhsī (d. 286/890), a pupil of al-Kindī, wrote
a book called Takshīf asrār al-mumawwihīn, “Revelation of the Secrets of Trick-
sters”, in which he ridiculed the prophets.58 Of another philosopher, the famous
medical doctor Abū Bakr al-Rāzī (d. 313/925 or later), we know a little more.59
54 M. Bernand (ed.), ‘Le Kitāb al-radd ʿalā ʾl-bidaʿ d’ Abū Muṭīʿ Makḥūl al-Nasafī’, Annales
Islamologiques 16, 1980, 111 (109w). Their reasoning was that God would not have sent a
fellow-human to mankind. Compare Ibn Dāʿī, Tabsirat al-ʿawāmm, ed. I. ʿAbbās, Tehran
1313, 65, where a judge who died in 463/1071 credits Ibn Karrām, perhaps by way of parody,
with the question why God had not sent an angel rather than a human prophet, so that
everyone would have believed (cf. A. Zysow, ‘Two Unrecognized Karrāmī Texts’, Journal of
the American Oriental Society 108, 1988, 582 f. and note 44 on the judge).
55 Some of the best known passages are conveniently assembled in T.R. Glover, The Conflict
of Religions in the Early Roman Empire, 11th ed., London 1927, 3ff.
56 Cf. Crone, Medieval Islamic Political Thought, 187 f., 261ff., 265f.
57 Cf. A. Wardman, Religion and Statecraft among the Romans, Baltimore 1982, 53f., 56.
58 J. Fück (ed.), ‘Sechs Ergänzungen zu Sachaus Ausgabe von al-Bīrūnīs “Chronologie orien-
talischer Völker” ’, in his Documenta Islamica Inedita, Berlin 1952, 78; F. Rosenthal, Aḥmad
b. aṭ-Ṭayyib al-Saraḫsī, New Haven 1943, 51.
59 On him, see ei1, s.v. ‘al-Rāzī, Abū Bakr’ (Kraus and Pines); also ei2, s.v. (Goodman);
212 chapter 9
Al-Rāzī rejected the idea of prophethood on the grounds that all humans had
the same ability to reach God. It struck him as absurd that God should choose
to reveal himself to mankind by informing just one person. In his view, God
must have implanted the requisite knowledge in all human beings, just as he
had given all animals the knowledge they needed to flourish. It followed that
prophets were impostors. They had been seduced by evil spirits who appeared
to them in the form of angels, and their miracles were mere magic and sleights
of hand.60
Al-Rāzī engaged in scriptural criticism with a view to showing that the
prophets contradicted one another, and sometimes themselves as well.61 He
seems to have known more about contradictions in the Old Testament than in
the New, presumably because he was indebted to Marcionite and Manichaean
criticism of the Old Testament (devoted to proving that the Old Testament
deity was not the highest God). Many critics of monotheism of the Biblical
type drew on Marcionite and Manichaean arguments, not just al-Rāzī. A ninth-
century Zoroastrian also made heavy use them in his polemics against Judaism,
as earlier Zoroastrians may have done already in Sasanian times;62 and a ninth-
century Jew, Ḥīwī of Balkh (wr. c. 870ce), drew on them for his critique of his
Judaism too.63 Ḥīwī rejected prophetic miracles, explaining them rationally,64
and found fault with his own scripture in the tradition of the scoffers with
whom Philo had contended in Alexandria many centuries earlier.65 Zindīqs
came from a Marcionite and Manichaean (and Bardesanite) background, and
L.E. Goodman, ‘Muḥammad ibn Zakariyyāʾ al-Rāzī’, in S.H. Nasr and O. Leaman (eds.),
History of Islamic Philosophy, i, London and New York 1996, ch. 13; M.M. Bar Asher, ‘Abū
Bakr al-Rāzī’, in F. Niewöhner (ed.), Klassiker der Religionsphilosophie: von Platon bis
Kierkegaard, Munich 1995.
60 Cf. Stroumsa, Freethinkers, ch. 3; P. Crone, ‘Post-colonalism in Tenth-Century Islam’, Der
Islam 83, 2006 [Ed.: included as article 7 in the present volume], 3–5.
61 Abū Ḥātim al-Rāzī, Aʿlām al-nubuwwa, ed. Ṣ. al-Ṣāwī, Tehran 1977, 69ff.
62 See P.J. de Menasce (ed. and tr.), Škand-Gumānīk Vičār: une apologétique mazdéenne du ixe
siècle, Fribourg en Suisse 1945, ch. 13, and the introduction, pp. 179f.
63 On him, see I. Davidson, Saadia’s Polemic against Ḥiwi al-Balkhi, New York 1915; J. Rosen-
thal, ‘Ḥiwi al-Balkhi’, Jewish Quarterly Review 38, 1947–1948, 317–342, 419–430 (with his
date at 319, n. 15); 39, 1948, 79–94. The pagan, Marcionite and Manichaean antecedents
of his questions are fully documented in Rosenthal’s ‘Ḥiwi’. The view, voiced from time to
time, that Ḥiwi was himself a Marcionite or Manichaean is both unnecessary and unper-
suasive.
64 Cf. Rosenthal, ‘Ḥiwi’, 334 f.
65 Cf. Philo, Questions and Answers in Genesis, iii, 43; De mutatione nominum, 61; De confu-
sione linguarum, 2; De ebrietate, 65 ff.; Quis rerum divinarum heres, 81; De somnis, 93f.
it was presumably from disputations with them that Ḥīwī,66 Ibn al-Rāwandī67
and al-Rāzī had learned how to pick holes in a scripture.68
Unlike Ibn al-Rāwandī, al-Rāzī was not a Skeptic (nor was he a scoffer).
Philosophy to him was an avenue to God, and what he attacked was not religion
as such, only positive or conventional (waḍʿī) religion, that is to say religion
embodied in a set of institutions that come between the individual and God
and that are credited with the right to lay down what others should believe and
do to reach salvation. Al-Rāzī did not want the panoply of religious scholars
and theologians (the “goatbeards”, as he called them) to dictate to him. What
he wanted was natural (Arabic ʿaqlī, rational) religion, that is the truth about
God reached by the individual himself on the basis of his own inner resources,
which had been implanted in him by nature/God and which were shared by
all human beings; such religion, it was widely thought, would be the same for
all mankind, unlike conventional religion, which sanctified one confessional
community against another and divided mankind instead of uniting it. It was
this desire for natural/rational religion and the corresponding hostility to the
conventional institutions erected by all confessional communities that made
Abū Bakr al-Rāzī and his likes freethinkers. (The word stands for a specific type
of religious radical, not for any kind of them.) In early modern Europe al-Rāzī
would have formed part of the radical Enlightenment; in later modern Europe,
he would have been a secularist, in the proper sense of someone who wishes
religion to be a private matter for the individual to decide on his own, not in
the debased sense of an anti-religious person in which the word is often used
today.
Al-Rāzī lived in a period that some have duly dubbed the “Muslim Enlighten-
ment” (c. 300–500/900–1100). It was a period of political fragmentation in which
Islam was competing on almost equal terms with other religions and in which
educated laymen of Muslim, Christian and Jewish background mixed freely at
the courts, where they were enjoying unusual prominence at the expense of the
religious scholars.69 Al-Rāzī was not the only Muslim to think in terms of a uni-
versal, rational truth versus the conventional religion represented by diverse
66 Cf. G. Vajda, ‘Judeo-Arabica’, Revue des Études Juives 99, 1935, 81–91, comparing Ḥīwī’s
questions with those asked by Zindīqs and rightly suggesting that he was a “radical
freethinker” like Ibn al-Rāwandī rather than a Manichaean, a Christian Gnostic or the like,
as proposed by earlier authors.
67 Cf. above, note 45.
68 Compare Abū Bakr al-Rāzī in Rāzī, Aʿlām, 69 f., on the burnt offering, and Ḥīwī in Rosen-
thal, ‘Ḥiwi’, 332.
69 For the wider context, see Crone, ‘Post-colonialism’, 18ff.
214 chapter 9
confessional communities. There were many ways of coping with the diversity,
however. One could postulate that all the prophets had really preached the
same truth and envisage all confessional communities other than one’s own
as the outcome of some kind of corruption: only one form of positive religion
was true (though they might all be useful). This was probably the most com-
mon view. Or one could see all the confessional communities as true in a more
relative sense by casting them as local, time-bound and socially determined
reflections of the universal truth. This was how most philosophers coped. Phi-
losophy, they said, conveyed the absolute and universal truth to the few who
could understand it wherever and whenever they were; revelation conveyed the
same insights in a more metaphorical form accessible to the masses, adjusted
to both their intellectual level and the particular time in which their communi-
ties flourished. This was also what the Ismaili Shīʿites said, with the difference
that in their view the need for such time-bound metaphorical forms would dis-
appear in the great spiritual resurrection with which they expected their Mahdī
(messiah) to bring the world as we know it to an end. Finally, one could dismiss
all revealed religion as devoid of truth value, but nonetheless indispensable for
political order: this view continued to have adherents in (apparently) Iran, now
among philosophically inclined people who held all revealed religions to be
mere legal institutions used by the nations for the maintenance of their liveli-
hoods.70 But like Ibn al-Rāwandī, al-Rāzī took the even more extreme view
once associated with the Epicureans, denying not only that positive religion
was true, but also that it was useful. He held that all human beings, not just
the elite, were capable of living without conventional religion here and now,
not just when the messiah came: all could be saved through philosophy in his
view. Accordingly, the prophets had to be impostors. Others only followed them
because they had grown up with all these “superstitions” (khurāfāt), which
had been dinned into them since childhood by the religious scholars (tacitly
accused of imposture too).71
Al-Rāzī’s views seem to have influenced the poet Abū ʾl-ʿAlāʾ al-Maʿarrī
(d. 449/1058), whose verses abound in statements directed against positive reli-
gion: “Awake, awake, you dupes! All these religions (diyānāt) are mere trickery
(makr) on the part of the ancients who wished to secure worldly goods for
themselves”. “In every nation falsehoods are taken as religion”.72 “They all err—
Muslims, Christians, Jews and Magians; two make humanity’s universal sect:
70 Al-ʿĀmirī, K. al-iʿlām bi-manāqib al-islām, ed. A.-ʿA.-Ḥ. Ghurāb, Cairo 1967, 101.
71 For the references, see Crone, ‘Post-colonialism’, 5 f.
72 R.A. Nicholson (ed. and tr.), ‘The Meditations of Maʿarrī’, in his Studies in Islamic Poetry,
Cambridge 1921, nos. 249, 252 (with more poetic translations).
one man intelligent without religion, and one religious without intellect”.73 But
unlike al-Rāzī’s, al-Maʿarrī’s tone is sarcastic, sceptic and deeply pessimistic.
In principle, the thought of all these men was politically explosive. Positive
religion meant law, political obedience and social order; and the adherents
of the impostor theory would now openly explain the prophets as motivated
by a desire for political leadership, not simply for money, though the desire
for wealth continued to be well represented too. To the philosopher al-Fārābī
(d. 339/950), for example, those who dismissed revelation as downright false
(which he did not) would cast its recipient as a mere “swindler seeking ruler-
ship and other things”.74 In practice, however, only the Ismailis were politi-
cally active. Whether they were Dahrīs or Zindīqs, theologians or philosophers,
recluses or courtiers, the freethinkers convey no impression of wishing the
political house that Muḥammad had built to come tumbling down, or even to
purge it of its theologians and religious scholars, though they certainly resented
the latter’s encouragement of what they saw as uncritical attitudes and their
tendency to brand anyone who disagreed with them as an infidel who could be
lawfully killed.75 The freethinkers come across first and foremost as educated
laymen who wanted to make sense of the world for themselves, without regard
for the custodians of the established order. They disliked the pairing of the high-
est truth with mundane social and political arrangements, and some of them,
al-Rāzī included, were also offended by the concomitant linkage of the highest
truth with war and bloodshed.76 But it is not clear that they thought it could be
changed.
To their opponents, freethinkers often came across as intellectual snobs. No
doubt they often were, for then as now, it was chic to flirt with radical positions.
Dahrism appealed to the smart set, as al-Jāḥiẓ (d. 255/869) informs us;77 it
was those who wanted to look clever who would take up positions against
73 Nicholson, ‘Meditations’, no. 239 (his translation).

74 Al-Fārābī, Mabādiʾ ārāʾ ahl al-madīna al-fāḍila, ed. and tr. R. Walzer, Al-Farabi on the Perfect
State, Oxford 1985, ch. 17, 6 (283 = 284). Compare ch. 18, 12, 13 (304= 305): among the
ancients [whose pernicious views have followers today] there were people who would
dismiss the ideas of a provident deity, prayer, abstinence, and reward and punishment in
the afterlife as mere tricks and ruses used by those who lacked the military strength to
take the good things of life by force.
75 Abū Bakr al-Rāzī in Abū Ḥātim al-Rāzī, Aʿlām al-nubuwwa, 30f.
76 Cf. Crone, ‘Post-colonialism’, 4 f., 22 f.
77 Jāḥiẓ, ‘Ḥujaj al-nubuwwa’, in his Rasāʾil, iii, 226f. (quoted in C. Pellat, Le milieu baṣrien et
la formation de Ǧāḥiẓ, Paris 1953, 84); id., ‘Radd ʿalā ʾl-naṣārā’, in his Rasāʾil, iii, 320f.; id.,
‘Ṣināʿat al-kalām’, in his Rasāʾil, iv, 246; Crone, ‘The Dahrīs according to al-Jāḥiẓ’.
216 chapter 9
established religion, as many later authors observed, both Muslims78 and Jews
among them.79 Precisely how deeply committed people were to the radical
views they played with is often unclear, not least because it was unwise to reveal
it. But some were certainly battling with loss of faith. According to Ibn Sīnā
(d. 428/1037), Satan has followers who secretly whisper to the innermost hearts
of man “that there is no resurrection, no requital for good and bad acts, and no
being existing eternally on its own, reigning eternally over the kingdom (wa-lā
qayyūm ʿalā ʾl-malakūt)”.80 He knew what he was talking about, for he had a
friend who lost belief in the reality of prophethood and wrote a philosophical
letter to persuade him of its truth.81
By medieval standards, the freethinkers did however enjoy considerable
freedom to air their views, as long as they did so in private discussions, in
the salons of the elite, and at the courts of tolerant rulers. One could debate
radical propositions as if for the sake of argument alone. One could also voice
them as part of mujūn, risqué statements or behaviour which bordered on the
blasphemous, the scurrilous or the pornographic, and which were an accepted
part of high culture as long as one expressed oneself with literary elegance and
wit, and/or in poetry, and coupled one’s daring with a good sense of precisely
where to stop.82 As regards literary expression, the fifth/eleventh-century ʿAbd
al-Jabbār claims that many of the godless people (mulḥida, here probably
Dahrīs), Zindīqs, and errant Muʿtazilites who had written against the prophets
back in the days when the caliphate was strong had composed their books in
secret, without even their wives and children knowing about them, and shown
them only to individuals engaged in similar practices, though the books had
eventually acquired such diffusion that one could now buy them in the Muslim
78 Cf. al-ʿĀmirī (above, note 70); al-Ghazālī, Tahāfut al-falāsifa (The Incoherence of the Philoso-
phers), ed. and tr. M.E. Marmura, Provo, Utah, 1997, 1ff. (first muqaddima, 2–5).
79 They would flaunt their erroneous views and look down on the followers of truth, as
Saʿadya Gaon (d. 942) observes with reference to such people among his own coreligionists
in his Kitāb al-amānāt waʾl-iʿtiqādāt, ed. S. Landauer, Leiden 1880, 4; tr. S. Rosenblatt, The
Book of Beliefs and Opinions, New Haven 1948, 7; they would ridicule the midrashim, casting
themselves as cultivated men, physicians and philosophers who were wiser than the sages
(Maimonides (d. 1204) in Stroumsa, ‘ “Ravings”: Maimonides’ Concept of Pseudo-Science’,
146).
80 Ibn Sīnā, Ḥayy b. Yaqẓān, ed. A. Amīn, 1947, 51; tr. A.-M. Goichon, Paris 1959, 174f.
81 Ibn Sīnā, ‘Fī ithbāt al-nubuwwāt’, in his Tisʿa rasāʾil, Cairo 1989, risāla no. 6.
82 See now Z. Szombathy, Mujūn: Libertinism in Medieval Muslim Society and Literature,
E.J.W. Gibb Memorial Trust 2013; more briefly, see ei2, s.v. ‘Mudjūn’ (Pellat); also P. Crone
and S. Moreh (trs.), The Book of Strangers, Princeton 2000, 174f., 178f., for brief characteri-
sations and some examples.
markets and everybody was talking about them.83 This may be broadly true: we
should perhaps envisage the medieval Islamic world as having a clandestine
literature. But one could say many things in published books as well as long as
one hid behind dead or otherwise absent dissenters and took care to counter
them with some appropriate objections. (Oddly, we do not hear of anonymous
publications.) In public, the opponents of conventional religion would conform
to prevailing norms and practise taqiyya (precautionary dissimulation), as Abū
Bakr al-Rāzī openly admitted.84
Very few freethinkers of the early Muslim world seem to have been penalised
for their views. Some Zindīqs were executed in a purge of people broadly classi-
fied as Manichaeans in the later second/eighth century, but the subject is highly
obscure.85 Al-Sarakhsī was also executed, but not for his religious views. He was
a polished courtier who would say daring things as witticisms, and though at
least one of his fellow-courtiers professed not to be amused,86 nobody knew
why he suddenly fell from grace;87 it was only later that his impiety seemed
to be the obvious explanation.88 Abū ʿAlī al-Jubbāʾī (d. 303/915f.), a Muʿtazilite
theologian, claims that both Abū ʿĪsā al-Warrāq and Ibn al-Rāwandī were pur-
sued by the authorities and that Abū ʿĪsā died in jail while Ibn al-Rāwandī was
forced to flee to a Jewish home, where he composed the Dāmigh and soon
after died.89 But it sounds like wishful thinking; it was certainly by wishful mis-
reading of al-Jubbāʾī that the Ḥanbalite scholar Ibn ʿAqīl (d. 513/1119) declared
Ibn al-Rāwandī to have been crucified.90 That Abū Bakr al-Rāzī had died in his
bed after a distinguished career as a doctor at diverse courts in Iran was never
denied, but then his thought barely seems to have reached the religious schol-
ars.
83 ʿAbd al-Jabbār, Tathbīt, i, 129.

84 He tells us that he had been blamed for not living like Socrates, an ascetic who did not
consort with kings or “practise taqiyya vis-à-vis the masses or the ruler” (P. Kraus (ed. and
tr.), ‘La conduite du philosophe: Traité d’ éthique d’ Abū Muḥammad b. Zakariyyā al-Rāzī’,
Orientalia 4, 1935, 309 = 322; also tr. C. Butterworth, ‘The Book of the Philosophic Life’,
Interpretation 20, 1993, 227).
85 Cf. Chokr, Zandaqa et zindīqs.
86 “Your unbelief will never be considered nice and witty”, as Ibn al-Munajjim (d. 300/912)
said in a poem dismissing al-Sarakhsī’s religious observance as mere hypocrisy (F. Rosen-
thal, Aḥmad b. aṭ-Ṭayyib al-Saraḫsī, New Haven 1943, 32).
87 Many different explanations were offered, cf. Rosenthal, Saraḫsī, 32, 35f.
88 Cf. Ibn al-Jawzī (d. 597/1200) and al-Ṣafadī (d. 764/1363) in Rosenthal, Saraḫsī, 29, 31.
Rosenthal tends to agree with them (p. 34).
89 Ibn al-Jawzī, Muntaẓam, vi, 102.3; in Ritter, ‘Philologika’, 5 = 13.
90 Ibn al-Jawzī, Muntaẓam, vi, 105.4; in Ritter, ‘Philologika’, 9 = 17 (he read ṣuliba for ṭuliba).
218 chapter 9
The Twelfth to Fourteenth Centuries

Al-Ghazālī (d. 505/1111), a towering Sunnī theologian, devoted considerable
attention to the enfeeblement of people’s belief in the very idea (aṣl) and real-
ity (ḥaqīqa) of prophethood, which he had found to be widespread.91 In his
view, it was people who mouthed philosophical views, thinking themselves
ever so clever, who would dismiss the revealed law as manmade and explain
its provisions as “embellished tricks” (ḥiyal muzakhrafa).92 Some said that Hell
was invented to scare people and that everything said about Paradise was just
blandishment to make them behave.93 Some of them were Dahrīs, of whom we
are now explicitly told that they held the prophets to be tricksters.94 Accord-
ing to the Persian epic poet Asadī (d. 465/1072f.), there were also Dahrīs who
regarded the prophets as learned ( farzāna), men who had established new reli-
gions in what they appear to have regarded as a benevolent vein even though
their claim to have been sent by God was false:95 it is curious that the adher-
ents of the theory of benevolent deceit always seem to be Iranians. It was
known to the Andalusian Ibn Ḥazm, too, however: the laws of Islam were either
given by God or else posited (mawḍūʿa) by agreement among the most vir-
tuous sages (afāḍil ḥukamāʾ) for the governance of people and restraint from
mutual oppression and vile things, as he says in polemics against unidenti-
fied philosophers who took the second view.96 To al-Ghazālī it did not mat-
ter whether the Lawgiver was held to have aimed at deception (talbīs) or the
welfare of the world (maṣlaḥat al-dunyā): either way, such views were incom-
patible with membership of the Muslim community.97 Those who claimed to
believe in prophecy but equated the prescriptions of the revealed law with
philosophical wisdom did not believe in genuine prophecy either in his view,98
91 Al-Ghazālī, al-Munqidh min al-ḍalāl, ed. and tr. F. Jabre, Beirut 1959, 46; tr. W.M. Watt, The
Faith and Practice of al-Ghazálí, Edinburgh 1953, 76.
92 Ghazālī, Tahāfut, 1 ff. (first muqaddima 2–5).
93 Ghazālī, Kimiyā-yi saʿādat, i, 113.
94 Al-Ghazālī, Fayṣal al-tafriqa bayna ʾl-islām waʾl-zandaqa, ed. S. Dunyā, Cairo 1961, 194; tr.
F. Griffel, Über Rechtgläubigkeit und religiöse Toleranz, Zürich 1998, 75; tr. S.A. Jackson, On
the Boundaries of Theological Tolerance in Islam, Oxford 2002, 111, without use of the word
Dahrī (they deny the creator, deem the world always to have existed, deny prophethood
and afterlife, and hold death to be pure nothing).
95 Asadī Ṭūṣī, Garshāspnāma, ch. 44, ed. Ḥ. Yaghmāʾī, Tehran 1317, 139; tr. H. Massé, Paris
1926–1951, ii, 30.
96 Ibn Ḥazm, Faṣl, i, 95.
97 Ghazālī, Tafriqa, 184; tr. Jackson, 101; tr. Griffel, 67.
98 Ghazālī, Munqidh, 50 f.; tr. Watt, 84.
and he resented the ostentatious piety that such men would display in pub-
lic. If one asked them why they would join the prayer when they did not hold
prophethood to be genuine (ṣaḥīḥa), they would affirm that they did believe
prophethood to be true and accepted the revealed law as genuine, or they
would come up with explanations such as that praying was good exercise, or
that it was local custom, or that they wanted to keep their wealth and chil-
dren.99
Al-Ghazālī’s efforts notwithstanding, the prophets continued to have their
critics. Abū Shāma (d. 665/1266 f.) mentions the death, in 656/1258, of a Zindīq
(here in the general sense of heretic) by the name of Shihāb al-Naqqāsh,
who would speak in the manner of the philosophers (ḥukamāʾ) and deny the
prophetic missions; this man lived in the Nūriyya Madrasa in Damascus, where
a number of Zindīqs of his kind would gather around him.100 Abū Shāma also
records the death, in 657/1259, of another real or alleged Zindīq by the name
of al-Fakhr (i.e. Fakhr al-Dīn) b. al-Badīʿ al-Bandahī, who had occupied himself
with philosophy and ancient sciences; he lived in the “madrasas of the jurists”
and would corrupt the creed of the young men who studied with him there
by openly belittling the prophets.101 How he belittled them we are not told.
Some works on logic by this man are extant, without furnishing evidence of
heretical views, and he may simply have cast the prophets as philosophers.102
But it is at least clear from all this that philosophers and religious scholars
were no longer distinct social groups, as they had been in the old days: they
now lived and worked in the same institutions. Another philosopher, the Jew
Ibn Kammūna (d. 683/1284f.), still found it necessary to refute the arguments
of those who denied the prophetic missions (al-nubuwwāt);103 and by then
philosophical ideas about the prophets had penetrated Sufi circles, too. Ibn al-
Jawzī (d. 597/1200) knew of Sufis who did not believe in God and other Sufis
99 Ghazālī, Munqidh, 47 f.; tr. Watt, 78 f. There is no mention of keeping their lives.
100 Abū Shāma, Tarājim al-rijāl al-qarnayn al-sādis waʾl-sābiʿ, ed. M.Z. al-Kawtharī, Cairo
1366/1947, 200.–2. I owe this reference to Denis McAuley.
101 Abū Shāma, Tarājim, 202.10. This man’s father claimed that he (the father?) had been a
pupil of Fakhr al-Dīn al-Rāzī.
102 He is listed as the author of a commentary on al-Khūnajī’s Kashf al-asrār in Ḥājjī Khalīfa,
Kashf al-ẓunūn, Istanbul 1941–1947, ii, 1486. I owe my knowledge of his identity, his works,
and their apparent orthodoxy to Khaled El-Rouayheb, who is editing al-Khūnajī’s work.
103 Ibn Kammūna, Tanqīh al-abḥāth lil-milal al-thalāth, ed. M. Perlmann, 18ff.; tr. M. Perl-
mann, Ibn Kammūna’s Examination of the Three Faiths, Berkeley, Los Angeles and London
1967 and 1971, 33 ff. Niewöhner, Veritas, 227–231, saw this work as a response to the three
impostors idea as supposedly formulated by the Ismailis three centuries earlier (cf. below).
220 chapter 9
who denied prophethood.104 The mystic Ibn ʿArabī (d. 638/1240) encountered a
philosopher who denied prophethood and miracles in 586/1190f., possibly in al-
Andalus;105 and Shams-i Tabrīzī, beloved of the mystic poet Rūmī (d. 672/1273),
mentions philosophers who rejected the probative value of prophetic mira-
cles, claiming that proof had to rest on the intellect and that the prophets had
been deceived by the angels: this was why they had been orientated towards
this world, busying themselves with people and taking wives; they had been
“ambushed in the road by love for position and prophethood”. One of the
philosophers would say things of this kind with a wink.106 But Rūmī himself
was familiar with people who would cast the prophets as ordinary humans and
compare their miracles with magic: such people, he said, were hypocrites who
would join the ritual prayer “for quarrelling’s sake, not for supplication”.107 A
Persian heresiography probably composed in Tabrīz in the eighth/fourteenth
or ninth/fifteenth century, which takes issue with numerous radical Sufi ideas,
includes among its targets the claim that the prophets were intelligent men
who used their wisdom in a benevolent vein to make rules for mankind, cred-
iting them to God to make them sound impressive. The adherents of this view,
an old one in Iran by now, held that humans were much too puny for a mes-
sage to come to them from heaven: the earth was a mere poppy seed in rela-
tion to the sun, they said, specifying the size of the sun as calculated by the
astronomers.108
In sum, there is ample evidence in the Islamic Near East for the view of
the prophets as impostors from the third/ninth century to beyond the time by
which the theme had appeared in Latin Christendom. It was aired in several
books and was known in a wide variety of different formulations, some more
radical than others, and was combated from Syria to Iran.
104 Ibn al-Jawzī, Talbīs Iblīs, ed. M.M. al-Dimashqī, Cairo 1928, 352; cf. Kraus, ‘Ketzerge-
schichte’, 348.
105 Ibn ʿArabī, al-Futūḥāt al-makkiyya, Būlāq n.d., ii, 490.5 (ed. of bāb 185). I owe this reference
to Denis McAuley.
106 W.C. Chittick (tr.), Me & Rumi: the Autobiography of Shams-i Tabrīzī, Louisville, Ky., 2004,
26 f., cf. also 62. The man who said it with a wink was Shihāb Hariwa (sic, Harawī?),
perhaps a student of Fakhr al-Dīn al-Rāzī (d. 607/1209), a major theologian well versed
in philosophy who was active in Iran.
107 Rūmī, Mathnawī, ed. and tr. R.A. Nicholson, London 1925–1940, verses 263ff.
108 M.J. Mashkūr (ed.), Haftād u sih millat, Tehran 1341, 45f. (This work was drawn to my
attention by Masoud Jafari Jazi.)
The Focus on Three and the Ismailis
What is missing in the material reviewed so far is a focus on three prophets.

Ibn al-Rāwandī dismissed Abraham along with Moses, Jesus and Muḥammad
in his Zumurrud;109 and Abū Bakr al-Rāzī discussed Zoroaster and Mani along
with the three.110 Naturally, Moses, Jesus and Muḥammad were the three most
relevant prophets to the Muslims, and we do sometimes find them enumerated
together on their own,111 for example in the poetry of al-Maʿarrī:112
The astrologer of the peoples is like a blind man

who has scrolls with him that he reads by touch.
He has been labouring for a long time, and how much he has struggled
with lines that their writer has effaced.
Moses preached, then Jesus stood up,
and Muḥammad came with the five prayers.
It is said that a religion other than this one will come
while people are perishing between tomorrow and yesterday.
Who assures me that the religion will become fresh again
and quench the thirst of the one who has engaged in devotional
exercises
after going without water for a long time?
In other words, the astrologer has long been trying to figure out the truth on
the basis of the stars, but is doing no better than a blind man trying to read
by touch;113 while he has been doing this, Moses, Jesus and Muḥammad have
appeared and now a new religion is predicted (the Ismailis were among those
who held their own belief system to be the religion in question), but will it be
any better than its predecessors? The answer is clearly no.
For prose formulations of the view that the impostors were three, however,
we need to turn to the Ismailis, adduced by Massignon in his note on the impos-
109 Al-Khayyāṭ, al-Intiṣār, ed. A.N. Nader, Beirut 1957, 12.8.

110 In Abū Ḥātim al-Rāzī, Aʿlām, 70.
111 E.g. Abū Ḥātim al-Rāzī, Aʿlām, 73.12, 91.15.
112 Maʿarrī, Meditations, ed. and tr. Nicholson, no. 263 (I have replaced Nicholson’s beautiful
translation with a more literal version); cf. no. 252 on the Furqān (i.e. Qurʾān), the Torah
and the Gospel. See also the examples in the Ismaili Abū Ḥātim al-Rāzī, Aʿlām, 73.12, 91.15.
113 This could be taken to imply that some kind of Braille had been devised in al-Maʿarrī’s
time, but more probably it simply means that just as the blind cannot read books by touch,
so astrologers cannot read the stars by sight.
222 chapter 9
tor theme many years ago. The Ismailis differed from the men considered so
far in that they were not hostile to the prophets, but on the contrary devoted to
them.114 They did, however, form part of a wider phenomenon labelled “Bāṭin-
ism”, roughly translatable as a preference for religion as spirituality rather than
law. How far a single attitude to the law prevailed among them before the
establishment of the Fāṭimid caliphate is unclear, but all agreed that however
indispensable it might be in our current, imperfect state, the law would be abol-
ished when the Mahdī (messiah) came: he would preside over the political and
spiritual regeneration of the world that they called the resurrection (qiyāma);
positive religion would wither away, and unmediated access to the truth would
prevail as it had done (according to some) in the time of Adam.115 Since one
cannot show that a religious law has been abrogated without acting contrary
to its precepts, the Ismailis were in principle committed to a great act of ritual
violation of the external aspect (ẓāhir) of Islam.
In practice, most of them ducked out of it. The movement split in 286/899,
in the course of preparations for the coming of the Mahdī, and one branch
(the one in which all modern Ismailis have their roots) proceeded to estab-
lish the Fāṭimid caliphate in North Africa and Egypt, where it affirmed its
allegiance to the law and postponed the spiritual resurrection to a distant
future. Another branch, usually known as Qarmaṭī, remained committed to the
abolition of the external institutions of Islam, however. In 310/922f. the Qar-
maṭīs in Baḥrayn began to launch regular attacks on Iraq, hoping to unseat
the caliph; in 317/929 they attacked Mecca, slaughtered pilgrims, and carried
away the black stone of the Kaʿba, reputedly desecrating it further back home,
in order to demonstrate that Islam as everyone knew it was finished; and
in 319/931 they proclaimed an Iranian captive of theirs to be the Mahdī and
proceeded to engage in a number of outrageous acts under his direction. In
the course of all this they are said to have declared the true religion to have
come, namely the religion of Adam, and to have cursed the prophets as impos-
tors.116 This is not impossible. It is certainly hard to see how they could have
parted with their beloved prophets without persuading themselves that they
hated them. But the account is both sensationalist and polemical, and exactly
what the Qarmaṭīs said we shall never know. Their acts deeply shocked other
114 For an account of their beliefs, aims and history in this period, see H. Halm, Das Reich des
Mahdi, Munich 1991; tr. M. Bonner, The Empire of the Mahdi, Leiden 1996.
115 Whether there had or had not been (religious) law in the time of Adam was hotly debated
by two fourth/tenth-century Ismailis, al-Nasafī and Abū Ḥātim al-Rāzī (see W. Madelung,
‘Das Imamat in der frühen ismailitischen Lehre’, Der Islam 37, 1961, 102ff.).
116 For all this, see Halm, Reich, 225 ff./Empire, 250 ff.
Muslims, however, and gave the Ismailis a scandalous reputation that has made
them reluctant to discuss the episode to this day.
Their enemies responded by casting Ismailism as a conspiracy by the con-
quered peoples who lacked the military strength to recover their lands and
who therefore planned to destroy Islam from within, namely by seducing Mus-
lims into a doctrine which, though disguised as Shīʿism, would eventually be
revealed to them as pure atheism. This theory was set out in a pamphlet known
as “The Book of the Highest Initiation” (Kitāb al-balāgh) or “The Book of Pol-
icy” (Kitāb al-siyāsa), which survives only in quotations. Supposedly an Ismaili
work, it is in fact a forgery not unlike the Protocols of the Elders of Zion in that
it purports to be a record of the planning of the cynical masterminds believed
to be behind the movement.117 Formulated as instructions by the leader of the
movement to the missionaries, it informs the reader that the highest law of
the prophets was to deceive this perverted world, that the missionaries had
to familiarise themselves with their impostures and contradictions, of which
some illustrations relating to Jesus, Moses and Muḥammad are given, and
that the missionaries also had to learn juggling and conjuring tricks so that
they could secure the world and everything in it for themselves.118 Thanks to
this pamphlet, all good Muslims “knew” that the Ismailis were really enemies
of the prophets, however many works in proof of prophethood they might
compose.119 After his conquest of Rayy (the precursor of modern Tehran) in
420/1029, for example, the Sunnī ruler of eastern Iran, Maḥmūd, reported to the
caliph that he had uprooted heretics there, including Bāṭiniyya (i.e. Ismailis)
who did not believe in God or his angels, or (revealed) books, messengers or
the last day, but rather regarded all religion as trickery by the philosophers
(makhāriq al-ḥukamāʾ); Maḥmūd or his secretaries had probably read all this in
the “Book of Highest Initiation”.120 Al-Ghazālī, who must actually have known
better, also credits the Ismailis with the idea that all the prophets were impos-
117 Cf. W. Madelung, ‘Fatimiden und Baḥrainqarmaṭen’, Der Islam 34, 1959, 69ff.; S.M. Stern,
‘The “Book of the Highest Initiation” and Other Anti-Ismāʿīlī Travesties’, in his Studies in
Early Ismailism, Jerusalem and Leiden 1983, 56–83. (The comparison with the Protocols
is also made by de Smet, ‘La théorie’, 89.) For Ismailism as a conspiracy of the conquered
peoples, see for example Stern, ‘Abū ʾl-Qāsim al-Bustī and His Refutation of Ismāʿīlism’, in
the same work, 305 f.; al-Ghazālī, Faḍāʾiḥ al-Bāṭiniyya, ch. ii.2 (ed. Amman 1993, pp. 13f.).
118 Stern, ‘Book of Highest Initiation’, 66 ff.
119 Abū Ḥātim al-Rāzī (d. 322/934) wrote one against Abū Bakr al-Rāzī (thereby preserving
the latter’s views for us, cf. above, note 61). Abū Yaʿqūb al-Sijistānī (d. after 361/971) wrote
another, entitled Ithbāt al-nubuwwāt (ed. ʿA. Tāmir, Beirut 1966).
120 Ibn al-Jawzī, Muntaẓam, viii, 39.
224 chapter 9
tors and depicts their missionaries as deceivers spreading false ideas in order
to gain wealth and power, probably drawing on the same work.121 A fourteenth-
century author familiar with the pamphlet similarly assures us that the Ismailis
denied the prophetic missions and miracles and claimed that the Prophet
wrote the Qurʾān himself.122
Massignon, who wrote at a time when the history of Ismailism was still
highly obscure, took the forgery to be a genuine Ismaili work and quoted the
snippet to do with the imposture of Moses, Jesus and Muḥammad in his famous
note on the Islamic origin of the three impostors theme.123 He also adduced
a passage from the Siyāsatnāma of Niẓām al-Mulk (d. 485/1092) along with an
anti-Ismaili passage from al-Maʿarrī. According to Niẓām al-Mulk, the man who
presided over the abduction of the black stone and the abolition of externalist
Islam in Baḥrayn, Abū Ṭāhir, wrote to the first Fāṭimid caliph, informing him
that “three persons have ruined mankind, a shepherd [Moses] and a doctor
[Jesus] and a camel-driver [Muḥammad], and the camel-driver was more of
a conjurer and juggler than the others”.124 A different version of Abū Ṭāhir’s
statement is found in an eighth/fourteenth-century Arabic source, where Abū
Ṭāhir says that “it was a shepherd, a physician and a camel-driver that led this
nation astray (mā aḍalla hādhihi ʾl-umma illā rāʾin wa-ṭabīb wa-jammāl)”.125
Thus is undoubtedly also based on the “Book of Highest Initiation”, which Abū
Ṭāhir had studied according to Niẓām al-Mulk.126
Whatever the Qarmaṭī leaders may or may not have said when they abol-
ished exoteric Islam in Baḥrayn, all we have is a Sunnī formulation of what
the Sunnīs believed them to have said. It could have been in this formulation
121 Ghazālī, Faḍāʾiḥ al-Bāṭiniyya, ch. ii.3 (ed. Amman 1993, pp. 15ff.).
122 Muḥammad b. al-Ḥasan al-Daylamī al-Yamānī (wrote 707/1307), Qawāʿid ʿaqāʾid al-bāṭiniy-
ya, ed. M.Z. al-Kawtharī, Cairo 1950, 90; cf. also Abū ʿUthmān b. ʿAbdallāh b. al-Ḥasan
al-Ḥanafī al-ʿIrāqī (6th/12th century?), al-Firaq al-muftariqa bayna ʾl-zaygh waʾl-zandaqa,
ed. Ankara 1961, 100, where they dismiss Muḥammad as an impostor who deceived the
rude Arabs thanks to their ignorance of philosophy and astronomy, without reference to
the other prophets (my thanks to Masoud Jafari Jazi for drawing this work to my attention).
The Assassins themselves are never credited with dismissing the prophets as impostors,
nor is al-Ḥākim, to whom the legend of the Old Man of the Mountain was transferred (pace
de Smet, ‘La théorie’, 92, who claims so with reference to F. Daftary, The Assassin Legends,
London 1994, 118–120).
123 Massignon, ‘La légende’, 83 f.
124 Niẓām al-Mulk, Siyāsatnāma, ed. H. Darke, Siyar al-mulūk, 2nd ed., Tehran 1985 = The Book
of Government or Rules for Kings, tr. H. Darke, London 1960, ch. 46, §36 (p. 309 = 236).
125 Yamānī, Qawāʿid, 90.
126 Niẓām al-Mulk, Siyāsatnāma, ch. 46, § 32 (p. 306 = 234).
that the idea of the prophets as impostors reached Frederick ii’s court. It does
not have to be, however. Poetry such as al-Maʿarrī’s could have had the same
effect.
Europe: Frederick ii
In his bull of 1239, directed to all of Latin Christianity, Pope Gregory ix charged
the emperor Frederick ii with saying that “the whole world had been deceived
by three deceivers (barattatoribus), to use his words, namely Christ Jesus, Moses
and Muhammad”.127 The three impostors are enumerated in the order of impor-
tance to Christians, but no motive is imputed to them, nor are we told where
Frederick had picked up the idea.128 A fuller account is found in Vita Gregorii ix,
and this work, composed in 1240, is of particular importance in that it was writ-
ten for papal in-house use and so cannot be dismissed as a mere propaganda
tool.129 Here we are told that Frederick owed his bad ideas to conversation with
Greeks and Arabs, who,
mendaciously affirming that all things relative to government derive from

the stars, instilled in him the pagan error that a man rejected by God
appears to himself to be a God in human form; and he publicly affirmed
that Moses, Christ and Muhammad were three tricksters (truffatores) who
had come to deceive people, that Moses, after having been saved from the
water, nourished himself with the bread of others, that Muhammad was a
camel-driver of servile birth, that both of them by means of their cunning
completed their lives supported by public favour; Christ was actually the
son of an artisan and an impoverished woman who, deceived by a false
belief, was justly recompensed by condemnation to the torments of the
cross; he then accuses him with various arguments of not being God,
affirming that the union of the creator and the created is impossible.
127 Esposito, ‘Manifestazione’ (above, note 1), 6.

128 Pace de Smet, ‘La théorie’, 91, neither the bull nor the Cronica S. Petri Erfordensis Moderna
says that he had picked it up from the Assassins, though the bull implies and the Cronica
says that he had obtained hired assassins from them.
129 Vita Gregorii ix in Le liber censuum de l’ église romaine (written 1240), ed. P. Fabre, ii, Paris
1905, 32 f.; cf. P. Fabre, ‘Les vies de papes dans les manuscrits du Liber Censuum,’ Mélanges
d’ Archéologie et d’ Histoire 6, 1886, 147–161, at 154, 155n; P. Montaubin, ‘Bastard Nepotism’,
in Frances Andrews (ed.), Pope, Church, and City, Leiden 2004, 129–176, at 154. I owe these
references to Robert Lerner.
226 chapter 9
Here it is from Greek and Arab astrologers that Frederick learns that one
can believe oneself to be divine even while being rejected by God. They are not
explicitly identified as the source of the impostor thesis, but the text can cer-
tainly be read to imply it. Moses, Christ and Muḥammad, now in chronological
order, are declared to be tricksters; Moses, a mere foundling, “nourished him-
self with the bread of others”, while Muḥammad was “a camel-driver of servile
birth” who used his cunning to complete his life “supported by public favour”.
Christ too had humble origins, but he suffered the punishment that the other
two avoided. It is notable that here as in the statement imputed to the Qarmaṭī
Abū Ṭāhir, Muḥammad is identified as a camel-driver, but the professions of
Moses and Jesus are missing, so this is perhaps less important than it seems. As
regards imposture for the purpose of living off public funds, this may be what
al-Maʿarrī is speaking of when he says:
Some parties declared that your God did not send Jesus and Moses (as
prophets)
to mankind,
but they only provided a means of livelihood (maʾkala) for their
followers
and made a net/a law/a deceit to catch them all (wa-ṣayyarū li-jamīʿi
ʾl-nāsi nāmūsan).130
But the reference could be to altruistic deceit: by providing a law, however

fraudulent, Jesus and Moses enabled their followers to live and make transac-
tions together. Muḥammad is not mentioned for obvious reasons, but nobody
will have been fooled.
As many have surmised, Frederick ii’s heresy must have originated in the
Islamic world. It is of course perfectly possible that the idea of religious leaders
as tricksters in search of worldly wealth and power has suggested itself inde-
pendently several times in history, but what we have here is not a case of the
wheel being invented twice. In the first place, Moses, Jesus and Muḥammad
were profoundly different figures to the Christians of Europe, who could hardly
have cast them as tricksters of the same type on their own. By contrast, the
Muslims venerated all three as prophets and so would naturally reject all three
as embodiments of the same error if they turned against them.131 In the sec-
130 Maʿarrī, Meditations, tr. Nicholson, no. 248.

131 Noted by D. Weltecke, “Der Narr Spricht: Es ist kein Gott”. Atheismus, Unglauben und
Glaubenszweifel vom 12. Jahrhundert bis zur Neuzeit, Frankfurt am Main 2010, 143.
ond place, Christ is the figure of central concern to Frederick ii, yet he does not
really fit the impostor pattern. Moses and Muḥammad successfully deceive oth-
ers into granting them positions of wealth and honour, but Jesus does not seem
to deceive anyone apart from himself: what he is punished for is his belief in his
own divinity, though it is not clear that it got him anywhere. Finally, the theme
appears suddenly on the Latin Christian side and remains rare for centuries,
whereas it has a continuous history from antiquity onwards on the Islamic side,
where we find it with a profusion of variations.
If we accept that the theme is of Islamic origin, by what channels was it trans-
mitted to Europe? First, was it to Frederick ii’s court that it was transmitted?
Some scholars deny it, if not on good grounds.132 An alternative hypothesis
would be that it was the Pope himself who had picked up the three impos-
tors idea from Muslim informants and fathered it on Frederick ii. This is not so
ridiculous a thesis as it may sound, for the papal curia spearheaded the same
type of culture as Frederick’s court: Michael Scotus had been patronised by the
popes Honorius iii and Gregory ix before passing into the service of Freder-
ick ii, for example.133 In addition, there was much traffic between the Roman
curia and the Holy Land. But Gregory ix claims to be quoting Frederick ii’s own
words in his bull (“the whole world had been deceived by three deceivers, to use
his words”), and his word for deceivers (barattatoribus) is an unusual one, which
he would hardly have imputed to Frederick if it had not figured in the reports he
had received about him.134 As we have seen, the account in the Life of Gregory ix
also makes it difficult to dismiss the charge as a mere propaganda ploy. More
probably, the idea of Moses, Jesus and Muḥammad as impostors was brought
to Frederick ii’s court by people to whom it was a well-known view, which they
did not necessarily share, but which they would air along with other explana-
tions of prophethood in discussions of precisely what the prophets had been:
philosophers who had achieved such perfection that they had come to be in
receipt of revelation (from the First Intellect)? Philosophers who had not in
fact received any revelation, but who had claimed to have their message from
132 Cf. D. Abulafia, Frederick ii: a Medieval Emperor, London 1988, 254, claiming that the
charge was “a stock accusation against disbelievers in the west well before he was born”.
In fact, only one possibly earlier case is known, that of Simon of Tournai (above, note 2),
and it is normally rejected in favour of Frederick as the first case.
133 Cf. A. Paravicini Bagliani, ‘Federico ii e la corte dei papi: scambi culturali e scientifici’,
in his Medicina e scienze della natura alla corte dei papi nel Duecento, Spoleto 1991, 53–84;
S.J. Williams, ‘The Early Circulation of the Pseudo-Aristotelian Secret of Secrets in the West:
the Papal and Imperial Courts’, Micrologus 2, 1994, 132, 140, 142.
134 I owe this point to Thomas Gruber.
228 chapter 9
God for the sake of the good of mankind? Or just men (philosophers or other-
wise) who had claimed to receive revelations in order to accumulate worldly
power and wealth? Discussions of this kind are likely to have been conducted
in philosophical circles all over the Muslim world, spiced with quotations of
poetry by al-Maʿarrī and his likes; and since the statement about the shepherd,
the physician and the camel-driver was witty, and probably also well known,135
it may have formed part of such discussions as well, as a succinct formulation
of the most extreme view.
Such discussions are likely also to have been conducted at Frederick’s court
in southern Italy (he did not return to Sicily after his youth). Among his cour-
tiers were an otherwise unknown astronomer sent to him by al-Kāmil, the ruler
of Egypt; a Christian doctor by the name of Theodore of Antioch, who had stud-
ied philosophy and science at Antioch, Mosul and Baghdad and who perplexed
several Dominicans with philosophical arguments that they were unable to
refute during Frederick ii’s siege of Brescia in 1238;136 the astronomer/astrologer
Michael Scotus, a Scot who had worked in Toledo, where he learned Arabic and
translated al-Biṭrūjī’s astronomical work Kitāb fī ʾl-hayʾa into Latin; and Jacob
Anatoly, an in-law of the famous Ibn Tibbon family of translators in Provence
(refugees from the Iberian peninsula), who worked as a translator of Aristotle
and Averroes from Arabic to Hebrew at Frederick’s court.137 These are precisely
the sort of men who would have felt free to discuss the nature of revelation.
Theodore of Antioch, for example, will have thought of the founders of the great
religions, including his own, as lawgiver prophets of the same type; and since
from a Christian point of view, Muḥammad fell into the category of impostors
motivated by a desire for worldly power, it raised the question how once could
be sure that the same was not true of the other founders, meaning Jesus and
Moses (since Zoroaster and Mani were not relevant in Italy). Theodore may
have been genuinely worried by that question or he may just have liked to per-
plex other people with it. It was a nicely radical view for an intellectual to play
with. The presence of just one scandalised observer from another part of Latin
Christendom, where discussions of this risqué kind were not part of the high
135 According to de Smet, ‘La théorie’, 90, it is cited by innumerable Sunnī and Zaydī authors
up to the Mamluk period. But he does not give any examples apart from Niẓām al-Mulk.
136 B.Z. Kedar and E. Kohlberg, ‘The Intercultural Career of Theodore of Antioch’, Mediter-
ranean Historical Review 10, 1995, 165 ff., 171; C. Burnett, ‘Master Theodore, Frederick ii’s
Philosopher’, in Federico ii e le nuove culture, Spoleto 1995, 225–285 (reprinted in his Ara-
bic into Latin in the Middle Ages, Farnham 2009, no. ix), 225f., 228, 255f.
137 Abulafia, Frederick ii, ch. 8; cf. T. Hockey and others (eds.), The Biographical Encyclopedia
of Astronomers, New York 2007, s.v. ‘al-Biṭrūjī’; Williams, ‘Early Circulation’, 138.
culture, would have been all that was required for Gregory ix to receive the hor-
rendous news that Frederick ii held the whole world to have been deceived by
three impostors, namely Jesus, Moses and Muḥammad.138 A Franciscan writ-
ing in 1261 claims that the person who heard Frederick utter this blasphemy
was Heinrich Raspe, the landgrave of Thuringia who was elected Holy Roman
emperor with papal backing when Frederick ii was excommunicated in 1246.
But the landgrave, contemptuously known as the Pfaffenkönig, had too strong
an interest in supporting the pope against Frederick for this to carry much
weight.139
There is of course no way to prove precisely how it happened. Maybe the idea
had been brought to Italy in some other way. The main point is that we need
not assume that Frederick ii actually meant what he said, if indeed it was he
who said it, or that anyone else at this court was convinced of it. They may have
been or they may not, but the presence of the idea at Frederick’s court does not
depend on it. However it happened, the transmission must have been oral, for
no Arabic book translated into Latin, whether at Frederick’s court or elsewhere,
contained the idea. It arrived by virtue of people from two different sides of the
civilisational fence talking to each other, as they did in Sicily, southern Italy, the
Iberian peninsula and elsewhere in the Mediterranean.
Later Attestations
After Frederick ii the impostor thesis disappears from sight for a hundred
years, then it turns in the Iberian peninsula. Here Thomas Scotus (no relation
of Michael Scotus) declared in the 1340s that “There were three impostors in
the world, sc. Moses who deceived the Jews, Christ who deceived the Chris-
tians and Muḥammad who deceived the Saracens …”140 This Thomas Scotus
has been plausibly identified as Thomas of Braunceston, a Franciscan necro-
mancer, alchemist and heretic who had been patronised by Pope John xxii
and enrolled, on the latter’s order, as a Dominican at Carcassonne in 1328. In
1333 “Thomas the Englishman”, probably the same person, was appointed lec-
138 Compare the Andalusian scandalised by disputations in Baghdad in which Muslims and
non-Muslims, even Dahrīs and Zindīqs, would debate on an even footing (M. Cook, ‘Ibn
Saʿdī on Truth-Blindness’, Jerusalem Studies in Arabic and Islam 33, 2007, 169–178).
139 Cronica S. Petri Erfordensis Moderna in mpl 30, 398, cited in Niewöhner, Veritas sive
Varietas, 149.
140 M. Esposito, ‘Les hérésies de Thomas Scotus d’ après le “Collyrium Fidei” d’Alvare Pélage’,
Revue d’ Histoire Ecclésiastique 33, 1937, 59.
230 chapter 9
turer in natural philosophy for the convent at Rieux at the foot of the Pyrenees;
and between 1341 and 1344 Alvarus Pelagius (d. 1352), bishop of Silves in Por-
tugal, tells us of Thomas Scotus, an apostate of both the Franciscan and the
Dominican orders whose heresies were known “in various parts of Spain and
elsewhere”.141 Apart from dismissing the founders of Judaism, Christianity and
Islam as liars, this Thomas denied the virgin birth and the divinity of Christ (as
did Frederick ii), as well as Christ’s miracles (dismissed as magic), the angels,
the afterlife and papal power. In positive terms he affirmed the eternity of the
world, the superiority of philosophy over positive religion (Aristotle was better
than Christ, a bad man hanged for his sins) and, most strikingly, the existence of
human beings before Adam.142 Thomas’ impostor thesis seems to be identical
with Frederick ii’s, but the pre-Adamites are new. The bishop of Silves linked
Thomas’ belief in pre-Adamites with his Aristotelian affirmation of the eter-
nity of the world, and there can of course be little doubt that Thomas, a natural
philosopher, was an Aristotelian. But a great many Aristotelians believed the
world to be eternal without affirming that there were humans before Adam.
Like the three impostors thesis, it was a view at home in the Islamic world
which is sporadically reported in late medieval Europe and shoots to great pop-
ularity in the sixteenth and seventeenth centuries.143 By then it was well known
that pre-Adamites were an idea of Persian, Arabic or Jewish origin.144
141 P. Nold, ‘Thomas of Braunceston o.m./o.p.’, in T. Prügl and M. Schlosser (eds.), Kirchenbild
und Spiritualität: Festschrift für Ulrich Horst op, Paderborn 2007, 179–195. I owe my knowl-
edge of this study to Robert Lerner.
142 Latin text in Esposito, ‘Hérésies de Thomas Scotus’, 59–62; summary English tr. in Nold,
‘Thomas of Braunceston’, 192–195.
143 M. Mulsow, ‘Pre-Adamites and Astrology of History between the Middle East and Europe:
Longue-Durée-Transfer or Entanglement?’, unpublished paper (2013), partly published
as ‘Vor Adam. Ideengeschichte jenseits der Eurozentrik’, Zeitschrift für Ideengeschichte 9,
2015, 47–66; R.H. Popkin, ‘The Pre-Adamite Theory in the Renaissance’, in E.P. Mahoney
(ed.), Philosophy and Humanism: Renaissance Essays in Honor of Paul Oskar Kristeller,
New York 1976, 59 f., cf. 61 (Thomas Nash declared in 1592 and 1593 that “I hear say there
be mathematicians abroad that will prove men before Adam”); A. Hamilton, The Family
of Love, Cambridge 1981, 118, on the Surrey sectarians, confession of 1561; Paul Kocher,
Christopher Marlowe, Chapel Hill 1946, 34, 43 f.; cf. the Diggers, Ranters and others in
P.C. Almond, Adam and Eve in Seventeenth-Century Thought, Cambridge 1999, 51; W. Poole,
‘Seventeenth-Century Preadamism, and an Anonymous English Preadamist’, Seventeenth
Century 19, 2004, 2, 7 f.; and Isaac de La Peyrère, Prae-Adamitae, published in 1655 (English
tr. Men before Adam, 1656; he had aired the ideas from the 1640s onwards).
144 Popkin, ‘Pre-Adamite Theory’, 53, where this is explained to La Peyrère by the Biblical
scholar Father Richard Simon.
Like the three impostors thesis, the idea of pre-Adamites may be rooted in
antiquity. According to Photius (d. c. 893), Clement of Alexandria (d. c. 215)
talked marvels about transmigrations of souls and about “many worlds having
existed before Adam”.145 If this is correct, Clement, a Christian moulded by
Platonism, was presumably trying to accommodate the Stoic-Platonic doctrine
of many successive worlds in a Christian scheme in which the world containing
Adam was the last. But it is only from Photius that we learn this, and Photius
was not known in fourteenth-century Europe. Augustine (d. 430), who was very
well known, was also aware of people who believed in many successive worlds,
but he infers that they must believe in spontaneous generation of humans out
of the elements, not that they must believe in humans before Adam.146 The
idea of successive worlds is also surprisingly well attested in rabbinic sources.
Here we are told that God created worlds and destroyed them before creating
this one, or that there were 974 generations before the creation of this world, all
destroyed because of their wickedness, or that the 974 generations wanted to
be created, but were not, though they were distributed as evil ones in every
generation, and the like.147 That there were humans before Adam is never
explicitly stated, however.
In the Islamic world we hear both of many Adams and of humans before
Adam from the mid-ninth century onwards. Some heretics said that God had
created seven Adams, each of whom would preside over an era lasting 50,000
years.148 Others merely insisted that Adam himself had ancestors.149 The idea
145 Photius, Bibl., 109.

146 Augustine, City of God, xii, 11.
147 Bereshit Rabba iii, 7 (ad Gen. 1:5); ix, 2 (ad Gen. 1:31), xxviii, 4 (ad Gen. 6:7); Babylonian
Talmud, Hagiga, 13b–14a; cf. E.E. Urbach, The Sages, tr. I. Abrahams, Jerusalem 1975, ch. 9,
210 f., citing these and other sources. (I owe my knowledge of the rabbinic material to
Reimund Leicht and Oded Zinger.)
148 Cf. P. Crone, The Nativist Prophets of Early Islamic Iran: Rural Revolt and Local Zoroastrian-
ism, Cambridge 2012, 209 f., citing al-Nāshiʾ (attrib.), K. uṣūl al-niḥal, ed. J. van Ess, Frühe
muʿtazilitische Häresiographie, Beirut and Wiesbaden 1971, §58, probably composed in the
first half of the ninth century; each era has a different population, without any carry-over
from one to the other. Compare Pirqe de Rabbi Eliezer (c. 800?), tr. G. Friedlander, New York
1970, 141 (on the Sabbath): God created seven aeons, six “for the going in and coming out”
and the seventh entirely Sabbath. Similar ideas appear in the Ḥurūfī Maḥramnāma, com-
posed in the late fourteenth/early fifteenth century (cf. O. Mir-Kasimov, ‘Notes sur deux
textes ḥurūfī’, Studia Iranica 35, 2006, 219f.), among the modern Yazīdīs (cf. P. Kreyenbroek,
Yezidism, London 1995, 37: the Christians only know history from the last Adam onwards);
and in an impeccably Imāmī Shīʿite village in the Zagros mountains in the 1970s (R. Loef-
fler, Islam in Practice: Religious Beliefs in a Persian Village, Albany 1988, 37, 39).
149 Cf. Khushaysh b. Aṣram (d. 253/867) in al-Malaṭī, K. al-tanbīh waʾl-radd ʿalā ahl al-ahwā
232 chapter 9
of several Adams inaugurating successive eras was widely accepted by the

Ismailis,150 and it also turns up in Sufism: the Andalusian mystic Ibn ʿArabī
(d. 638/1240) recollected a saying of the Prophet to the effect that “God has
created a hundred thousand Adams”.151 The idea of humans before Adam was
current in historical astrology, the study of the conjunctions determining the
rise and fall of kings, dynasties, prophets, religions and other major events on
earth and the predictions which can be made on that basis. The main Muslim
authority on this subject was Abū Maʿshar (d. 272/886), whose Book of Reli-
gion and Dynasties was translated into Latin in the mid-twelfth century; but
contrary to expectation, it does not mention pre-Adamites. Back in Iraq, how-
ever, an astrologer who flourished around 900 held that before Adam there
had been “many nations, created beings, monuments, habitations, civilisations,
religions, (forms of) kingship and kings”, all quite different from ours, and that
“Hermes lived a long time before Adam”. He also wrote a book about con-
junctions and predicted the coming of a man who would restore Zoroastrian
sovereignty, unite the whole world and do away with the rule of the Arabs and
others.152 The self-proclaimed Chaldaean Ibn Waḥshiyya (fl. c. 320/930) simi-
larly held that a sage called Dawanay lived before Adam, and that Adam was
called the Father of Mankind only because of his contributions to science.153
waʾl-bidaʿ, ed. S. Dedering, Istanbul 1936, 72 (“They do not acknowledge Adam [as the first],
but say that Adam also had ancestors”); similarly an old trader in Loeffler, Islam in Practice,
37, 39.
150 Cf. P.E. Walker, Early Philosophical Shiism, Cambridge 1993, 112, on Abū Yaʿqūb al-Sijistānī
(d. c. 975); al-Baghdādī, al-Farq bayna ʾl-firaq, ed. M. Badr, Cairo 1910, 280; al-Ḥusayn b. ʿAlī
(Yemeni Ismaili, d. 667/1268) in B. Lewis, ‘An Ismaili Interpretation of the “Fall of Adam”’,
Bulletin of the School of Oriental and African Studies 9, 1937–1939, 694, 697, cf. 697n on
the modern Ismailis; H. Corbin, Cyclical Time and Ismaili Gnosis, London 1983, 42ff., 78ff.
Kevin van Bladel, who has drafted a provisional article on pre-Adamism in the Islamic
world, has many more Ismaili references.
151 Ibn ʿArabī, al-Futūḥāt al-makkiyya, Dār Ṣādir reprint, Beirut 1968, iii, 549, line 13.
152 Al-Maqdisī, Kitāb al-badʾ waʾl-taʾrīkh, ed. and tr. C. Huart, Paris 1899–1919, ii, 97f., 147f.;
cf. Baghdādī, Farq, 271 (where he is a Bāṭinī, i.e. Ismaili); al-Bīrūnī, al-Āthār al-bāqiya
ʿan al-qurūn al-khāliya, ed. C.E. Sachau, Leipzig 1878 (repr. 1923), 213; tr. C.E. Sachau, The
Chronology of Ancient Nations, London 1879 (repr. 1984), with the same pagination in the
margin. Most of the material on this man (whose name appears in different forms) was
presented by Kevin van Bladel at the first workshop on the transmission of radical ideas
from the Islamic world to Europe.
153 Ibn Waḥshiyya in J. Hämeen-Anttila, The Last Pagans of Iraq: Ibn Waḥshiyya and His
Nabatean Agriculture, Leiden 2006, 298, text 44. The works in this and the next note were
also covered by van Bladel.
Ibn Waḥshiyya’s views on pre-Adamites were reported by Judah ha-Levi (fl.

c. 1130) in al-Andalus and by Maimonides in his Guide, written in Egypt, and
so were known to Jews well before the Guide was translated into Latin (at Fred-
erick ii’s court).154 It was presumably from Jewish and/or Muslim astrologers
in the Iberian peninsula that Thomas Scotus had picked up the idea.
Then there is silence for another hundred years.155 In 1459, however, the
impostor theme turns up again, this time among the heresies of Zaninus, a
canon of Solcia in Lombardy.156 Zaninus repented of a fair number of errors,
including that God had created a world other than this one and that in its time
many other men and women had existed, so that Adam was not the first man;
that Jesus Christ, Moses and Muḥammad had ruled the world by the pleasure
of their wills; that Jesus Christ suffered and died by the law of the stars, not to
redeem the human race; that Christian law would come to an end through the
succession of another law, just as the law of Moses has been terminated by the
law of Christ; and that the world would be destroyed naturally rather than by
divine fiat, by the heat of the sun consuming the humidity of the land and air
and setting the elements on fire. Zaninus had clearly been studying historical
astrology, and he too combined the impostor thesis with belief in pre-Adamites.
In Zaninus’ case there cannot be much doubt that we have to do with
independent transmission of the impostor theme, for he envisages all three
founders of religion as rulers: they ruled the world by the pleasure of their wills,
he says, i.e. as they saw fit, not on the basis of divine instructions. The political
dimension of the impostor theory had surfaced in Europe well before Zaninus,
for Matthew Paris (whose chronicle stops in 1258) credits Frederick ii with the
view that three conjurers (prestigiatores) cleverly seduced the world in order
to dominate it (rather than to live off public funds, as the Life of Gregory ix has
it).157 How Matthew Paris had picked up this idea is unknown, but it can hardly
be a Christian development, given that Jesus had not ruled anything at all. Cast-
ing him as a wielder of political power made sense because of the dominant role
that the church had acquired, but it violated the historical record, and it does
so in a particularly drastic form in Zaninus’ formulation, since Jesus had asked
for his death to be taken away from him. Further, the concept of a succession
154 Judah ha-Levi, al-Kuzari, i, 61; Maimonides, Guide, iii, ch. 29, tr. C. Rabin, Indianapolis
1995, 177, 179; tr. S. Pines, Chicago 1963, ii, 515 f.
155 According to R.W. Southern, Western Views of Islam, Cambridge, Mass., 1962, 75, the
impostor idea turned up again in Aragon in the 1380s, but his only reference is to Esposito
on Thomas Scotus and I do not know what he has in mind.
156 Esposito, ‘Manifestazione’, 41 ff.
157 Esposito, ‘Manifestazione’, 8.
234 chapter 9
of religious laws, each one abrogating its predecessor, is Islamic. It is particu-

larly well developed in Ismailism, but historical astrologers were into the same
game. Once again, the intermediaries could be Jews, for Zaninus also held that
Jesus was illegitimate, presumably because he had been exposed to the Tole-
doth Ieshu, and the cyclical concept of time had gone into the Kabbalah by then.
In addition, however, Zaninus subscribed to two views which are strikingly
reminiscent of Muslim heresy, namely that “wantonness outside of matrimony
is not a sin, except by the prohibition of positive laws”—only ecclesiastical pro-
hibition stopped people from following the opinion of Epicurus—and that the
taking of other people’s property is not a mortal sin even when it is against
the will of the owner.158 The idea that women and property were free for all,
once associated with the sectarians of western Iran known as Khurramīs, was
current in Sufi circles, including those condemned by the heresiographer writ-
ing (probably) in fourteenth-century Tabrīz.159 They were well known in Latin
Europe too, where they were associated with the so-called Free Spirits. But
the Free Spirits had no interest in astrology, science or pre-Adamites. Given
that Zaninus’ heresies form a coherent cluster of ideas, all well attested in the
Islamic world, the chances are that all of them had travelled from Tabrīz to
Europe, either via the Balkans, carried by Bektashis and other Sufis, or else via
Constantinople, the route by which Maraghan astronomy made its way from
Azerbaijan to Italy and, in ways still not precisely known, to Copernicus.160
The political dimension of the impostor theory is absent from the report of
1468 on the members of the Roman academy who dismissed Moses, Christ and
Muḥammad as deceivers and seducers.161 In some of these reports, the founders
of religion are at least identified as lawgivers,162 but it is not until the sixteenth
century that positive religion is routinely cast as “but a device of policy”, as
Marlowe (d. 1593) reputedly described it when he dismissed Jesus and Moses
as deceivers and the Bible as idle stories.163
158 Cf. Esposite, ‘Manifestazione’, 42 f.

159 Cf. M. Mashkūr (ed.), Haftād u sih millat, nos. 27, 37, 76; Crone, Nativist Prophets, 261ff.,
440, 448 ff., 482; cf. also 137, 257, 261 ff.
160 Cf. G. Saliba, Islamic Science and the Making of the European Renaissance, Cambridge,
Mass., and London 2007, ch. 6, esp. 194 f.
161 E. Garin, History of Italian Philosophy, ed. and tr. G.A. Pinton, Amsterdam and New York
2008, 199.
162 Thus the versions attributed to Averroes, the report of 1468, and Matteo da Acquasparta
(d. c. 1302) on Frederick ii (Esposito, ‘Manifestazione’, 15).
163 J. Hotson, The Death of Christopher Marlowe, London and Cambridge, Mass., 1925, 11f. (“a
device of pollicie”).
By then we also encounter a related idea, namely that religion, whether true
or false, was a useful institution in that it allowed us to live together. In fact, we
encounter this notion already in the Policraticus of John of Salisbury (d. 1180),
who tells us that King Numa had civilised the barbarous Romans by means of
(false) religious institutions.164 In John’s time this was a radically new idea: to
Augustine and his many readers, Numa was a cynical manipulator exemplify-
ing the deceit that pagan rulers would practise in order to consolidate their own
power.165 But as John had learnt from Pope Adrian iv (d. 1195), one should con-
sider “the utility of all” instead of focusing on the harshness used by the church
or secular princes, such as for example when they extracted money from all and
sundry.166 Are Islamic ideas lurking behind this as well? It is certainly striking
that John speaks of “external worship” (cultus exteriores) for what the Ismailis
called the ẓāhir, meaning external, manifest or plain religion (public worship to
John), as opposed to the inner, esoteric (bāṭin) meaning pursued by initiates.167
However this may be, there were soon Christians who denied that the law had
anything but utilitarian value. Thus it was said of Pope Boniface viii (r. 1294–
1304) that already back in his days as a cardinal he had considered religious
laws, including those of Muḥammad and of Christianity, to have been “invented
by men in order to take people away from evil by means of the fear of punish-
ment”. There was no eternal punishment in his opinion: “thus, in religious laws
the truth is nothing but the condition for men to live together civilly and qui-
etly (civiliter et quiete) because of fear of spiritual punishment”.168 Thereafter
164 John of Salisbury, Policraticus, tr. J. Dickinson, New York 1927, v, 3 (68f.).
165 Thus Augustine, City of God, iii, 4, 9; iv, 31, 32; vii, 34; viii, 5, on Numa and Varro. John’s
most important source seems to be the Epitome of Florus (d. c. 130), based on Livy, still
incompletely known in John’s time. (John did not use Plutarch’s Numa, for all that he freely
invokes Plutarch’s name.) Florus, a pagan, is favourable to Numa, but lacks the idea that
the divine support was a sham: to him the “immortal gods” are real. John is the first to have
the view that Numa’s institutions were good even though they were what Augustine saw
as devilish inventions.
166 John had naively assumed that the Church and the Pope should not take bribes, an idea to
which Pope Adrian responded by laughing and telling him the story of the ancient Roman
Menenius Agrippa, who taught his soldiers that all parts of the human body contributed
to the body’s welfare, even the stomach, which seemingly did nothing: if it was starved,
the whole body would die (Policraticus, vi, 24).
167 John of Salisbury, Policraticus, iii, 68.
168 In P. Dinzelbacher, Unglaube im “Zeitalter des Glaubens”, 2nd ed., Badenweiler 2009, 23;
Ruggero di Simone in Boniface viii en procés, ed. J. Coste, Rome 1995, 504. Many other
witnesses made similar statements, also reproduced in Coste’s book (of which I owe my
knowledge to Gianluca Briguglia).
236 chapter 9
the idea of religion as socially useful surfaces in different forms in the works
of Albert the Great (d. 1280),169 the Paduan judge Geremia da Montagnone
(d. 1321),170 Marsilio of Padua (d. 1342)171 and later figures; and by the sixteenth
century we hear of Spanish and Italian Christians who held that religion was “a
human invention for living well (al ben vivere)”172 and that religion existed “so
that we may live in peace (para que viviéramos en paz, ut viverimus in pace)”.173
In a related vein the miller Menocchio (Domenico Scandella) held the func-
tion of the eucharist to be to control people, in a civilising sense: it had been
instituted so that men would not be like beasts.174 But was religion really useful
for everyone or just for kings and/or churchmen? Augustine’s view that pagan
169 Albert, Commentary on De Anima of Aristotle, 407b19ff., cited in M. Silk, ‘Numa Pompilius
and the Idea of Civil Religion in the West’, Journal of the American Academy of Religion 72,
2004, 875: Albert held that it was because Pythagoras wanted “to make citizens cultivate
piety and justice” that he made up the story that the souls of bad citizens would depart
from one body into another of worse condition, e.g. into the body of a lion or an ass. Albert
is developing a point made by Averroes in his Long Commentary (Sharḥ) on Aristotle,
De Anima, Latin translation probably by Michael Scotus (d. 1231?), ed. F. Stuart Crawford,
Cambridge 1953, 74; tr. R.C. Taylor and T.-A. Druart, New Haven 2009, 67 (book i, 53):
Averroes briefly says that Pythagoras spoke of reincarnation in order to correct the souls
of the citizens.
170 He knew the idea from classical sources such as Cicero’s De natura deorum, i, 118 (“some
have said that the whole opinion about immortal gods was made up by wise men for the
sake of the commonwealth”), which he cites without agreeing with it (see A. Brett (tr.),
Marsilius of Padua: The Defender of the Peace, Cambridge 2005, 29n).
171 Brett, Marsilius, i, 5, 11–12 (pp. 28 f.).
172 Thus Girolamo Busale (d. 1541, probably of Marrano origin) in M. Firpo, Tra alumbrados
e “spirituali”: Studi su Juan Valdés e il valdesianesimo nella crisi religiosa nel ’500 italiano,
Florence 1990, 94 (drawn to my attention by Stefania Pastore). According to a Venetian
Inquisitorial trial in 1553 (asv, Sant’Uffizio, Processi, b. 159, f. ii, f. 113r, made available to me
by Stefania Pastore), a student of law by the name of Giulio Basalú passed from believing
“only in that which tallied in the one and the other law, i.e. the Hebrew and the Christian”,
to the conviction that religion was no more than an “invention by humans for living well”.
Under the influence of a Spanish refugee in Italy, he came to hold that “Christ was purely
human, but abundantly full of holy spirit”, that the soul was mortal, that no religion was
true, whether Christian, Jewish or other, and that “Christ was a good man who taught how
to live well (Christo era stato homo da bene che haveva insegnato el ben viver)”. He also
held concubinage to be no sin, and laughed at everything. See further L. Addante, Eretici
e libertini nel Cinquecento italiano, Rome 2010, 25–30 (drawn to my attention by Stefania
Pastore).
173 S. Pastore, Una herejía Española: Conversos, alumbrados e Inquisición, Madrid 2010, 218, on
Juan de Castillo, 1537.
174 C. Ginzburg, The Cheese and the Worm, Baltimore 1980, 11.
religion was invented for the enslavement of the masses appealed to Boccaccio
(d. 1375) and was to play a major role in the radical Enlightenment as a thinly
disguised attack on the Christian church, seen as manipulating and defrauding
the common people, while a list of “articles in which modern heretics err” dis-
missed Easter observance, confession and penance as devices permitting the
church to collect money.175 Hell had been invented by priests in order to cheat
people for the sake of money; it was an illusion created by the authorities so that
they could rule as they liked.176 The concept of religion as socially useful thus
follows the same pattern as the impostor theme and the idea of pre-Adamites:
well attested in many forms in the Islamic world from the third/ninth century
onwards, it appears like a bolt out of the blue in the Latin West in the twelfth
century, a bit earlier than the other themes, and surfaces from time to time
thereafter until it takes off in the sixteenth century.
Conclusion
The three impostors illustrate a process that still has not received much atten-
tion, namely transmission from (and to) the Islamic world by word of mouth
rather than by books. Where people live next to each other, they talk to each
other, learn from each other and adjust to one another’s positions, whether for
purposes of living in peace or on the contrary to fight. We know a great deal
about the relations between the Islamic world and Christian Europe in terms
of war, political negotiations, polemics and translations, all of which left plenty
of paper trails; but the same does not apply to oral contacts because they were
not usually recorded, and on top of that they often took place at social levels
that did not count from a high cultural point of view. Of such oral exchanges
there must have been plenty, since Muslims and Christians were living cheek
by jowl in the Mediterranean, with plenty of Jewish neighbours too. In fact,
oral transmission must have played a major role even in connection with the
translation of Arabic texts into Latin, for the collectors had to talk to (Muslim,
Jewish or Christian) bearers of Islamic culture in order to acquire manuscripts,
and the translators must have looked for informants to tell them how the texts
175 W. Wakefield, ‘Some Unorthodox Ideas of the Thirteenth Century’, Medievalia et Human-
istica, ns, 4, 1973, 30.
176 Thus Christiern Pedersen (d. 1554) on “mad people” in Dinzelbacher, Unglaube, 70; the
miller Pellegrino Baroni, nicknamed “Pighino” (1570), in Ginzburg, The Cheese and the
Worm, 118; Costatino Saccardino in 1622 (in G. Schwerhoff, ‘Die alltäglische Auferstehung
des Fleisches’, Historische Anthropologie 12, 2004, 309–337, at 332).
238 chapter 9
were to be understood. In fact, translations were often cooperative enterprises,

and the middlemen were often Jews, who were more likely than others to mas-
ter more than one high cultural language. But most exchanges will have taken
place without the parties being aware of it, causing ideas to travel in imper-
ceptible ripples from one community to the other and to display themselves
in subtle adjustments to traditional ways of reading well-known material, for
example in the new evaluation that John of Salisbury puts on King Numa or
Alvarus Pelagius’ accommodation of belief in pre-Adamites under the rubric
of Aristotelianism. The impostor theme has unusual visibility in the sources
because it was so scandalous, but it should be treated as symptomatic of a much
broader process that we need to learn how to track; for we cannot otherwise
know precisely what it meant for the development of Christian Europe that it
had the Islamic world rather than some other civilisation as its neighbour.
chapter 10
How the Field Has Changed in My Lifetime*
Let me start by telling you that there is one monumental change that I am
not going to talk about even though it indisputably occurred in my lifetime,
and that is the technological revolution that has given us computers, the web,
databases and more besides. It is not that these changes have not affected me,
far from it; but I don’t feel they are really part of my scholarly history because
they haven’t shaped me. I use these gadgets, but only up to a point, and I still
think as if I lived in a world of typewriters. So it is for the next generation to
assess the effect that electronics have had on our field.
Back in the 1960s, when I started studying Islamic history, the field was still
dominated by the work of the great Orientalists who had created the field.
Most of them were Western Europeans working primarily in Germany, France,
Britain and Holland. They usually came to Islamic studies from Biblical studies,
but there were also many whose academic interest had been aroused by their
work as colonial administrators. They all had Greek and Latin from school,
they normally combined Arabic either with Persian and Turkish or with all the
Semitic languages, and they often worked in many fields relating to Near East,
not just Islamics. They were pretty impressive people. They edited the main
texts and wrote the first scholarly accounts, started source-critical studies, and
looked all set to raise the study of Islam to the level achieved in Biblical studies
when the First World War broke out, soon followed by the second, and so
everything changed.
The key characteristic of Orientalism was a sharp distinction between the
subject and object of study. The Orientalists—the subject—were studying an
alien world—the object—in order to explain it to a Western audience. They
had no intention of converting the people they were studying, or of polemi-
cising against them, or even of interacting with them in any way, except when
they were colonial administrators or missionaries. Of course there were excep-
tions such as Goldziher, who studied at al-Azhar, or Edward Browne, to whom
the study of Persian was intimately (if briefly) linked with interaction with Ira-
nians.1 But even when the Orientalists actually interacted with Muslims, their
* A version of this essay was presented as a lecture in Leiden on 9 February 2013 at a colloquium
that marked the launch of Ibn Ḥazm of Cordoba: The Life and Works of a Controversial Thinker,
edited by C. Adang, M. Fierro and S. Schmidtke (Leiden 2013).
1 Another exception is Bernard Lewis, who travelled extensively in the Muslim world through-

240 chapter 10
approach was distinctive in that their aim was to explain them in terms intel-
ligible to Westerners. In short, the Orientalists were doing much the same as
Bīrūnī (d. c. 1050) had done in his India book. He wanted to explain Indian reli-
gions to a Muslim audience. He wasn’t out to convert his object of study either,
nor was he writing polemics against them; on the contrary, he complains that
all earlier treatments of the subject were biased and partisan. He did collect
information from them, and he was able to do so because his patron took him
to India to assist his colonial expansion. But he was not interested in how the
Indians would respond to his portrait of them—he didn’t expect them to read
it.
I shall have to leave Bīrūnī aside. The point I am trying to make is that the Ori-
entalists studied Muslims as if they were distant stars, translating things Islamic
or Indian into categories and patterns that their own people could understand,
without regard for what the distant stars would make of it. Of course their writ-
ings were Eurocentric, partly because they were writing for Europeans using
European concepts and categories, and partly because they assumed their own
civilization to be superior. People usually do, especially when they are on top of
the world. But the objects of study did not remain distant stars. Muslims were
increasingly being schooled in Western languages and academic methods, and
works written for Westerners began to look offensive when the audience came
to include Easterners, who didn’t like reading about themselves in translation
and who disliked having to learn foreign categories in order to understand their
own traditions. After ww2 they ceased to be colonial subjects whose opinion
could be dismissed; they became independent and many of them moved to
the West, where they acquired a voice in the universities and other elite insti-
tutions. So now you have the great drama whereby the subjects and the objects
of enquiry begin to merge and have to work out new rules of intellectual coexis-
tence. The landmark here is Edward Said’s attack on Orientalism, which came
out in 1978.
I grew up in that rather remote outpost called Denmark, and I had the Ori-
entalist conception of Islamic studies. I originally wanted to study the ancient
Near East, and I decided to do Islamic history instead without ever having
set eyes on a Muslim or heard any Middle Eastern language spoken. I didn’t
think I would ever even get to the Middle East. To me, studying Sumerians
and studying Muslims were much the same. Of course that changed when I
out his long career. But he moved in diplomatic, governmental and royal circles, so that his
experience was quite different from that of younger specialists in Islamic studies (see B. Lewis
(with B.E. Churchill), Notes on a Century: Reflections of a Middle Eastern Historian, New York
2012).
how the field has changed in my lifetime 241
went to England, but even there the Orientalist conception only seemed to
be frayed at the edges. That’s fifty years ago, and things have totally changed
since then, in our field as in Indian and Chinese studies. There is no Western
study of the Orient anymore. The key distinction is no longer between West
and East, but rather between Islamic (or Hindu or Confucian or whatever)
history as done in the universities and as done in seminaries or madrasas or
other traditional institutions of learning. This development has boosted our
numbers, opened up new libraries in the Middle East, and led to the publi-
cation of a vast array of new sources. Islamic history is now a much bigger
field with much better source material than it was in the fifties and sixties.
It has also had some drawbacks, such as the intrusion of politicized history-
writing, identity politics and victim culture, but overall it has been a good
thing.
There are two further ways in which the end of colonialism has affected
the field. One is what you might call the rise of post-colonial bad conscience,
which became particularly pronounced after the publication of Edward Said’s
Orientalism. It is still with us and still rampant, and it requires you to denigrate
Western civilization for its colonial and other sins whenever you can, whereas
you may not say anything that could be construed as criticism of Islam. I could
give many examples, but no doubt many others could too, and although some
people seem to be unaware of the degree to which they are engaging in double-
think, I prefer to say no more about it.
The second way is the degree to which intellectual developments in the
Muslim world now affect Islamic studies in the West, which I think is not
generally noticed. There is a good example in approaches to the Quran. When I
started my studies, the general Islamicist view was that all interpretations of
the Quran had to come out of the exegetical literature. All Islamicists were
unwittingly subscribing to the rule that the tradition sat in judgement of the
Quran, not vice versa: al-sunna qāḍiya ʿalā ʾl-qurʾān. But that’s no longer the
case, and the change started in the Islamic world, in Egypt in the 1950s, in
Pakistan a bit later. Muslims were rebelling against the tradition because they
wanted to adapt to modern ways, and they wanted the Quran to validate their
views. So they started doing tafsīr al-qurʾān biʾl-qurʾān, and the results were
startlingly different. In the west the first person to study the Quran on the basis
of the Quran alone seems to have been Angelika Neuwirth, in her first book
published in 1981. But now it is commonplace, and it has contributed to making
Quranic studies a very live field. Everything has to be rethought. I shouldn’t
think Neuwirth was aware of following Muslim trends. I have found myself
saying things which reflected modernist Muslim interpretations and which I
myself was not aware of at the time. One thing I’d like to know is how these
242 chapter 10
things spread. In the old days you called it Zeitgeist, now I think we have to
come up with something better, such as some suggestions as to the mechanics,
but I don’t think anyone has worked on it.
To give you another example, when I started studying Islamic history, Arab
nationalism was reigning supreme. And back then, Islamic history was studied
in isolation from other fields, and with an overwhelming stress on the Arabs.
Islam was seen as the fruit of a marriage between God and Arabia, and as hav-
ing developed thereafter in accordance with its own internal needs, shaped by
the Quranic spirit, and so on. Nobody paid any attention to the cultural tradi-
tions of the conquered peoples. Basically the Arab conquerors were assumed to
have brought Islam as we know it to the Middle East, where it erased everything
that went before it, except for a few so-called “foreign elements” that somehow
slipped through. To suggest anything else was to detract from “the originality
of the Prophet” or “the originality of Islamic civilization”. Islamic history began
in Arabia and ended with Arab nationalism, and it did so because Arab nation-
alism set the tone, yet we were not aware of being shaped by it.
Of course, other factors came into it too. In the 1950s and ’60s it became fash-
ionable to adopt a functionalist approach and to deride an interest in origins.
There was much impatience with “diffusionism”. This was a trend that affected
all the humanities; I think it started in anthropology, and it is presumably also
connected with the rise of the former colonies to independence. In any case,
at a time when the whole world was being transformed by Western influence,
academic orthodoxy required you to deny that there was any such thing as
“influence” at all. All social and cultural transformations were allegedly due
to “inner, organic” developments; foreign elements were only borrowed when
people positively needed them and wanted them for their own internal rea-
sons. That put a nice gloss on what was going on in the post-colonial world. It
also did have some salutary effects, but like so many trends it became tyranni-
cal.
There are still people who think that history-writing is about giving prizes
for “originality” and “creativity”, as if that was what people in the past were
striving for, but Islamic studies have now been enormously affected by the
developments inaugurated by Peter Brown. He put late antiquity on the map
and made it so prestigious that now it is the height of fashion to connect Islam
with developments in the Middle East outside Arabia and to stress the degree
to which Islam originated as a late antique religion. In addition, globalization
has caused people in just about every field to stress the porous nature of
borders and the numerous ways in which neighbouring civilisations affected
(and affect) each other. You still aren’t allowed to speak of “influence”, with the
result that people sometimes resort to silly euphemisms to avoid it. It is true
that it isn’t a good word, but there are times when you need it, if only as a
shorthand. But be that as it may, there can be no doubt that Islamic studies
have come out of their isolation.
ok, so much for the aftermath of colonialism. Now for the changes internal
to the West itself. The first big change to note here has been the rise and fall
of the Soviet Union and, along with it, the rise and fall of socio-economic
history, in Islamic studies as elsewhere. The interest in socio-economic history
among Islamicists began in the late nineteenth century, but the big change
came with the Russian revolution, and by the 1950s Islamicists were deeply
into economic organization, social classes, the rise of capitalism and so on.
Socio-economic history continued to reign supreme down to the eighties. My
esteemed colleague, Michael Cook, started as an economic historian; it was his
job description until he defected to America and he has published in that field
as well. Though he has left it, there are still people in it, I am glad to say, for
it is important, but it no longer plays a dominating role. It was replaced by
questions to do with race, identity and gender. For some reason, the American
preoccupation with race and slavery didn’t have much of an impact on Islamic
studies—the only one who took up the subject was Bernard Lewis. But identity
was a different matter. For years after my move to the Institute for Advanced
Study, the applications were dominated by questions of identity, and to a lesser
extent gender. Gender is still going strong, but the favourite subjects are now
porous boundaries along with agency (ascribed to everything, even inanimate
objects).
The third big change has been the rise of Islamic studies in America together
with the enormous expansion of the universities, which is still going on. Back
in the ’50s and ’60s the only Americans who mattered in Islamic studies were
immigrants from Europe, such as Rosenthal, Schacht or von Grunebaum. The
field was dominated by Western Europe—especially England, France, Ger-
many and Holland. Now America dominates the field. Lots of Islamicists are
being produced there; and though many Europeans continue to be imported,
there can be no doubt that America is setting the tone.
This has had both good and bad consequences. Among the good conse-
quences is the sheer increase in our numbers. There just weren’t enough of
us before to get things moving—now there are. But the expansion of higher
education has also resulted in a huge bottom of semi-educated people who are
barely literate and whose entrance has introduced Gresham’s law to Islamicist
scholarship: the bad is driving out the good.2 It is not sheer numbers alone that
2 As Lewis observes in his Notes on a Century, 193.

244 chapter 10
are at work here, though. All over the world, especially in the West, there has
been a great wrench away from the tradition, including the traditional rules of
how to write, spell and construe sentences and arguments, which were delib-
erately withheld from the next generation by teachers who were young in the
’60s, so that the next generation reached maturity without the ability to express
themselves and went on to teach their impoverished language skills to the next
generation in their turn. Sheer numbers in combination with poor schooling
and lack of interest in the tradition have also resulted in an increased number
of people who only have one foreign language (in Islamic studies, usually Ara-
bic), which condemns you to mediocrity.
Connected with this expansion is the rise of the publish or perish syn-
drome, which is something that did not affect me at all as a young scholar and
which now dominates the lives of young scholars everywhere, partly thanks to
increased competition and partly thanks to the victory of the business concep-
tion of academic output. It results in a lot of publications that are premature,
hurried, second-rate and often much too long for what they have to say. It
also means that certain types of enquiry simply don’t get attempted any more
because they would take too long. Or they get attempted only as part of big col-
laborative enterprises such as the ones that are funded by the European Union.
All in all you could say that back in the ’50s one thought of great scholars as
geniuses toiling away in the attic—it was a Romantic conception of the scholar
as the lone pursuer of the truth. But now it is the factory mode of production
that prevails and the ideal great scholar now is not a genius in the attic, but
rather an entrepreneur and broker. It has been accompanied by a huge increase
in bureaucratic chores. All this is bad.
Another effect of the rise of American scholarship has been the rise of theory.
Like most trends it has been both a good and a bad thing. The worst side effect
has been the rise of pretentiousness. A lot of scholarship gets written in stilted
Latinate jargon, almost beyond comprehension, and often very solipsistic. The
only world that exists is the author’s mind. It has also led to a sad contempt
for philological skills, including those required to produce editions, which is a
serious problem, for many of the standard editions we rely on are not proper
editions, just printings of a particular manuscript, often without indication
of variants, with lots of corruptions and so on. We desperately need better
editions, but a graduate who decides to make one will not get a job. What
people want is “originality”. Every young person is trained to think that he is
going to turn the field upside down. We have a situation in which people think
that a healthy field is one in which there is nothing but paradigm shifts. It’s the
academic version of the doctrine of permanent revolution, and it is not doing
anyone any good. But not all is lost. What could sound more philological than
papyrology? Yet Petra Sijpesteijn has managed to make the topic sexy, and to
integrate this formerly rather marginal field into mainstream historical studies.
Finally, there is a most positive effect. Americans take it for granted that
people doing Islamic history belong in a history department, that those doing
Arabic or Persian or Urdu literature belong in a literature department, and
that those doing religion belong in a religion department. It may sound self-
evident, but it is not at all how things were done back in England. There all
Islamicists were put in faculties of Oriental studies, with a few in the faculty of
divinity. The history faculty in Oxford, for example, did not recognize any non-
European history as history, and believe it or not, that was a common attitude.
It contributed to the isolation of our field. But over here Islamicists have been
exposed to the ways of other historians, or literary scholars, or specialists in
religion, and learned to ask the kind of questions that others ask, and it seems
to me that the study of Islamic history, at least in the period that I tend work
on, has become a lot more sophisticated.
There is no doubt that the Orientalists were amazing scholars. Somebody
like Nöldeke puts us all to shame. But it has to be admitted that he was not a
good historian, and that the same tended to be true of the other great Oriental-
ists, though of course there were exceptions (e.g. Wellhausen). Most Oriental-
ists practised what you might call the scissors and paste approach to history:
first you separated fact from fiction—there was nothing in between: things
were either true or fabricated, and if they were fabricated they were useless—
and then you pasted the facts together, one piece here, another there, until
eventually you had used them up, leaving you with a picture that had no depth,
no perspective, no sense of real people interacting in a real world.3 That’s how
Islamic history tended to be written when I started. You still see it today. But
most Islamicists these days will treat information as just a tiny fragment of a
lost world, an accidental survivor which is of value not only for what it explic-
itly says, but also, sometimes only, for what it presupposes, for the kind of beliefs
and institutions it takes for granted, and for what it is trying to say even if every
word is invented. A single potsherd can suffice to give you an idea of the whole
vase. So we no longer insist on gluing all our sherds together. Rather, we will
mount them the modern way, one here, one there, with a bit of conjectural
metal to hold them together, so that what you see is a real, three-dimensional
vase even though the actual fragments are tiny.
That has been a huge gain. The crude pictures with which the pioneers began
are being discarded; there is a lot less “essentialism”, a lot more sensitivity to
3 Nöldeke’s Sketches from Eastern History is a good example.

246 chapter 10
changes over time and place and greater awareness of the different interests of
competing groups, including women and children. There is also a much better
understanding of how societies actually worked in the past, how social control
was maintained, how politics were negotiated, how propaganda was shaped to
dress it up, and so forth. We owe a lot to the social sciences here, but also of
course to literary studies and deconstruction.
All in all, despite the loss of respect for philology, the excessive respect for
originality, the tyranny of the factory model of academic work, the publish or
perish syndrome, and other negative factors, the developments of the last fifty
years strike me as largely positive.
List of Patricia Crone’s Publications
Books
Hagarism: The Making of the Islamic World. Cambridge: Cambridge University Press,
1977 (with Michael Cook). Paperback edition, 1980. Unauthorized Arabic translation
by Nabīl Fayyāḍ, al-Hājariyyūn, Damascus: n.p., 2003.
Slaves on Horses: The Evolution of the Islamic Polity. Cambridge: Cambridge University
Press, 1980. Paperback edition, 2003.
God’s Caliph: Religious Authority in the First Centuries of Islam. Cambridge: Cambridge
University Press, 1986 (with Martin Hinds). Paperback edition, 2003.
Roman, Provincial and Islamic Law: The Origins of the Islamic Patronate. Cambridge:
Cambridge University Press, 1987. Paperback edition, 2002.
Meccan Trade and the Rise of Islam. Princeton, nj: Princeton University Press, 1987.
Reprint, Piscataway, nj: Gorgias Press, 2004. Arabic translation by Āmāl Muḥam-
mad al-Rawabī, Tijārat Makka wa-ẓuhūr al-islām, Cairo: al-Majlis al-Aʿlā lil-Thaqāfa,
2005.
Pre-Industrial Societies. Oxford: Blackwell, 1989. Swedish translation by Birger Hedén
and Stefan Sandelin, Förindustriella samhällen, Lund: Studentlitteratur, 1991. Ger-
man translation by Marianne Menzel, Die vorindustrielle Gesellschaft: Eine Struk-
turanalyse, Munich: Deutscher Taschenbuch-Verlag, 1992. Second edition, Oxford:
Oneworld, 2003. Revised edition, 2015.
The Book of Strangers: Medieval Arabic Graffiti on the Theme of Nostalgia. Princeton,
nj: Markus Wiener, 1999 (translated with Shmuel Moreh). Danish translation by
Sune Haugbølle, De fremmedes bog: Arabisk nostalgisk graffiti fra middelalderen,
Copenhagen: Vandkunsten, 2004.
The Epistle of Sālim ibn Dhakwān. Oxford: Oxford University Press, 2001 (with Fritz
Zimmermann).
Medieval Islamic Political Thought. Edinburgh: Edinburgh University Press, 2004. Pub-
lished in America under the title God’s Rule: Government and Islam; Six Centuries
of Medieval Islamic Political Thought. New York: Columbia University Press, 2004.
Paperback edition, 2005. Turkish translation by Hakan Köni, Ortaçağ İslam dünya-
sında siyasi düşünce, Istanbul: Kapı, 2007. Persian translation by Masʿūd Jaʿfarī Jazī,
Tārīkh-i andīsha-yi siyāsī dar islām, Tehran: Sukhan, 2011; chapter 13 reprinted as an
article in Bukhārā 14, no. 80 (2011).
From Kavād to al-Ghazālī: Religion, Law and Political Thought in the Near East, c. 600–
1100. Aldershot: Ashgate, 2005 (Variorum reprint of 12 articles).
From Arabian Tribes to Islamic Empire: Army, State and Society in the Near East, c. 600–
850. Aldershot: Ashgate, 2008 (Variorum reprint of 12 articles).
248 list of patricia crone’s publications
The Nativist Prophets of Early Islamic Iran: Rural Revolt and Local Zoroastrianism. Cam-
bridge: Cambridge University Press, 2012. Paperback edition, 2013.
Muqannaʿ wa sapīdjāmagān. Tehran: Māhī, 2013 (with Masʿūd Jaʿfarī Jazī; four articles
translated into Persian).
Collected Studies in Three Volumes (reprinted, revised and previously unpublished arti-
cles). Vol. 1: The Qurʾānic Pagans and Related Matters. Vol. 2: The Iranian Reception
of Islam: The Non-Traditionalist Strands. Vol. 3: Islam, the Ancient Near East and Vari-
eties of Godlessness. Edited by Hanna Siurua. Leiden: Brill, 2016.
Edited Volumes
Studies in Early Islamic History, by Martin Hinds, ed. Jere Bacharach, Lawrence I. Con-
rad and Patricia Crone. Princeton, nj: Darwin Press, 1996.
The Greek Strand in Islamic Political Thought, ed. Emma Gannagé, Patricia Crone,
Maroun Aouad, Dimitri Gutas, and Eckart Schütrumpf. Mélanges de l’Université
Saint-Joseph 57. Beirut: Université Saint-Joseph, 2004.
Princeton Encyclopedia of Islamic Political Thought, ed. Gerhard Böwering with Patricia
Crone, Wadad Kadi, Devin J. Stewart and Mahan Mirza. Princeton, nj: Princeton
University Press, 2013.
Articles
‘Islam, Judeo-Christianity and Byzantine Iconoclasm’. Jerusalem Studies in Arabic and

Islam 2 (1980): 59–96 (= From Kavād to al-Ghazālī, no. iii).
‘Jāhilī and Jewish Law: The Qasāma’. Jerusalem Studies in Arabic and Islam 4 (1984):
153–201 (= From Kavād to al-Ghazālī, no. iv).
‘The Tribe and the State’. In John A. Hall (ed.), States in History, Oxford: Blackwell, 1986,
48–77. Revised version in John A. Hall (ed.), The State: Critical Concepts, London:
Blackwell, 1994, vol. 1, 446–476 (= From Arabian Tribes to Islamic Empire, no. i).
‘Max Weber, das islamische Recht und die Entstehung des Kapitalismus’. In Wolfgang
Schluchter (ed.), Max Webers Sicht des Islams: Interpretation und Kritik, Frankfurt
am Main: Suhrkamp, 1987, 294–333. Revised English version, ‘Weber, Islamic Law,
and the Rise of Capitalism’, in Toby E. Huff and Wolfgang Schluchter (eds.), Max
Weber and Islam, New Brunswick, nj: Transaction Books, 1999, 247–272 (= From
Kavād to al-Ghazālī, no. vi).
‘Did al-Ghazālī Write a Mirror for Princes?’ Jerusalem Studies in Arabic and Islam 10
(1987), 167–191 (= From Kavād to al-Ghazālī, no. xii).
‘On the Meaning of the ʿAbbasid Call to al-Riḍā’. In C.E. Bosworth, Charles Issawi, Roger
list of patricia crone’s publications 249
Savory and A.L. Udovich (eds.), The Islamic World from Classical to Modern Times:
Essays in Honor of Bernard Lewis, Princeton, nj: Darwin Press, 1989, 95–111 (= From
Arabian Tribes to Islamic Empire, no. vii).
‘Kavād’s Heresy and Mazdak’s Revolt’. Iran 29 (1991), 21–42 (= From Kavād to al-Ghazālī,
no. i; = Collected Studies in Three Volumes, vol. 2, The Iranian Reception of Islam: The
Non-Traditionalist Strands, art. 1).
‘Serjeant and Meccan Trade’. Arabica 39, no. 2 (1992), 216–240.
‘Tribes and States in the Middle East’ (review article). Journal of the Royal Asiatic Society
3, no. 3 (1993), 353–376 (= From Arabian Tribes to Islamic Empire, no. ii).
‘“Even an Ethiopian Slave”: The Transformation of a Sunnī Tradition’. Bulletin of the
School of Oriental and African Studies 57, no. 1 (1994), 59–67 (= From Kavād to al-
Ghazālī, no. viii).
‘Were the Qays and Yemen of the Umayyad Period Political Parties?’ Der Islam 71, no. 1
(1994), 1–57 (= From Arabian Tribes to Islamic Empire, no. iv).
‘Zoroastrian Communism’. Comparative Studies in Society and History 36, no. 3 (1994),
447–462 (= From Kavād to al-Ghazālī, no. ii; = Collected Studies in Three Volumes,
vol. 2, The Iranian Reception of Islam: The Non-Traditionalist Strands, art. 2).
‘The First-Century Concept of Hiǧra’. Arabica 41, no. 3 (1994), 352–387 (= From Arabian
Tribes to Islamic Empire, no. iii).
‘Two Legal Problems Bearing on the Early History of the Qurʾān’. Jerusalem Studies in
Arabic and Islam 18 (1994), 1–37 (= From Kavād to al-Ghazālī, no. v).
‘The Rise of Islam in the World’. In Francis Robinson (ed.), The Cambridge Illustrated
History of the Islamic World, Cambridge: Cambridge University Press, 1996, 2–31.
‘A Note on Muqātil b. Ḥayyān and Muqātil b. Sulaymān’. Der Islam 74, no. 2 (1997), 238–
249 (= From Arabian Tribes to Islamic Empire, no. v).
‘The ʿAbbāsid Abnāʾ and Sāsānid Cavalrymen’. Journal of the Royal Asiatic Society 8, no. 1
(1998), 1–19 (= From Arabian Tribes to Islamic Empire, no. viii).
‘A Statement by the Najdiyya Khārijites on the Dispensability of the Imamate’. Studia
Islamica 88 (1998), 55–76 (= From Kavād to al-Ghazālī, no. ix).
‘The Early Islamic World’. In Kurt A. Raaflaub and Nathan S. Rosenstein (eds.), War
and Society in the Ancient and Medieval Worlds: Asia, the Mediterranean, Europe, and
Mesoamerica, Cambridge, ma: Harvard University Press and Washington, dc: Center
for Hellenic Studies, 1999, 309–332 (= From Arabian Tribes to Islamic Empire, no. ix).
‘The Significance of Wooden Weapons in the Revolt of al-Mukhtār and the ʿAbbāsid
Revolution’. In Ian Richard Netton (ed.), Studies in Honour of Clifford Edmund
Bosworth, vol. 1: Hunter of the East: Arabic and Semitic Studies, Leiden: Brill, 2000,
174–187 (= From Arabian Tribes to Islamic Empire, no. vi).
‘Ninth-Century Muslim Anarchists’. Past and Present, no. 167 (2000), 3–28 (= From
Kavād to al-Ghazālī, no. x).
‘The Khārijites and the Caliphal Title’. In Gerald R. Hawting, Jawid A. Mojaddedi and
Alexander Samely (eds.), Studies in Islamic and Middle Eastern Texts and Traditions
in Memory of Norman Calder, Oxford: Oxford University Press, 2000, 85–91 (= From
Kavād to al-Ghazālī, no. xi).
‘Shūrā as an Elective Institution’. Quaderni di Studi Arabi 19 (2001), 3–39 (= From Kavād
to al-Ghazālī, no. vii).
‘A New Source on Ismailism at the Samanid Court’ (with Luke Treadwell). In Chase
F. Robinson (ed.), Texts, Documents and Artefacts: Islamic Studies in Honour of D.S.
Richards, Leiden: Brill, 2003, 37–67 (= Collected Studies in Three Volumes, vol. 2, The
Iranian Reception of Islam: The Non-Traditionalist Strands, art. 10).
‘What Was al-Fārābī’s “Imamic” Constitution?’ Arabica 50 (2003), 306–321 (= Collected
Studies in Three Volumes, vol. 2, The Iranian Reception of Islam: The Non-Traditionalist
Strands, art. 11).
‘The Pay of Client Soldiers in the Umayyad Period’. Der Islam 80, no. 2 (2003), 284–300
(= From Arabian Tribes to Islamic Empire, no. x).
‘Al-Fārābī’s Imperfect Constitutions’. Mélanges de l’Université Saint-Joseph 57 (2004),
191–228 (= Collected Studies in Three Volumes, vol. 2, The Iranian Reception of Islam:
The Non-Traditionalist Strands, art. 12).
‘Mawālī and the Prophet’s Family: An Early Shīʿite View’. In Monique Bernards and John
Nawas (eds.), Patronate and Patronage in Early and Classical Islam, Leiden: Brill,
2005, 167–194 (= From Arabian Tribes to Islamic Empire, no. xi).
‘How Did the Quranic Pagans Make a Living?’ Bulletin of the School of Oriental and
African Studies 68, no. 3 (2005), 387–399 (= Collected Studies in Three Volumes, vol. 1,
The Qurʾānic Pagans and Related Matters, art. 1). Danish translation in Tidskrift for
Islamforskning 1 (2006), art. 2 (online).
‘Post-Colonialism in Tenth-Century Islam’. Der Islam 83, no. 1 (2006), 2–38 (= Collected
Studies in Three Volumes, vol. 3, Islam, the Ancient Near East and Varieties of Godless-
ness, art. 7).
‘Imperial Trauma: The Case of the Arabs’. Common Knowledge 12, no. 1 (2006), 107–116
(= From Arabian Tribes to Islamic Empire, no. xii).
‘Abū Saʿīd al-Ḥaḍrī and the Punishment of Unbelievers’. Jerusalem Studies in Arabic and
Islam 31 (2006), 92–106 (= Collected Studies in Three Volumes, vol. 3, Islam, the Ancient
Near East and Varieties of Godlessness, art. 4).
‘Quraysh and the Roman Army: Making Sense of the Meccan Leather Trade’. Bulletin of
the School of Oriental and African Studies 70, no. 1 (2007), 63–88 (= Collected Studies
in Three Volumes, vol. 1, The Qurʾānic Pagans and Related Matters, art. 2).
‘“Barefoot and Naked”: What Did the Bedouin of the Arab Conquests Look Like?’
Muqarnas 25 (2008), 1–10 (= Collected Studies in Three Volumes, vol. 3, Islam, the
Ancient Near East and Varieties of Godlessness, art. 1).
‘No Compulsion in Religion: q. 2:256 in Medieval and Modern Interpretation’. In
Mohammad Ali Amir-Moezzi, Meir M. Bar-Asher and Simon Hopkins (eds.), Le
Shīʿisme imāmite quarante ans après, Turnhout: Brepols, 2009, 131–178 (= Collected
Studies in Three Volumes, vol. 1, The Qurʾānic Pagans and Related Matters, art. 13).
‘The Muqannaʿ Narrative in the Tārīkhnāma’ (with Masoud Jafari Jazi). Part i (Introduc-
tion, Edition and Translation), Bulletin of the School of Oriental and African Studies
73, no. 2 (2010), 157–177. Part ii (Commentary and Analysis), Bulletin of the School of
Oriental and African Studies 73, no. 3 (2010), 381–413. (= Muqannaʿ wa sapīdjāmagān,
arts. 1 and 2 [in Persian]; = Collected Studies in Three Volumes, vol. 2, The Iranian
Reception of Islam: The Non-Traditionalist Strands, arts. 6 and 7.)
‘The Ancient Near East and Islam: The Case of Lot-Casting’ (with Adam Silverstein).
Journal of Semitic Studies 55, no. 2 (2010), 423–450 (= Collected Studies in Three
Volumes, vol. 3, Islam, the Ancient Near East and Varieties of Godlessness, art. 2).
‘The Religion of the Qurʾānic Pagans: God and the Lesser Deities’. Arabica 57, no. 1–2
(2010), 151–200 (= Collected Studies in Three Volumes, vol. 1, The Qurʾānic Pagans and
Related Matters, art. 3).
‘The Dahrīs According to al-Jāḥiẓ’. Mélanges de l’Université Saint-Joseph 63 (2010–2011),
63–82 (= Collected Studies in Three Volumes, vol. 3, Islam, the Ancient Near East and
Varieties of Godlessness, art. 5).
‘Abū Tammām on the Mubayyiḍa’. In Omar Alí-de-Unzaga (ed.), Fortresses of the Intel-
lect: Ismaili and Other Islamic Studies in Honour of Farhad Daftary, London: I.B. Tau-
ris and Institute of Ismaili Studies, 2011, 167–187 (= Muqannaʿ wa sapīdjāmagān, art. 4
[in Persian]; = Collected Studies in Three Volumes, vol. 2, The Iranian Reception of
Islam: The Non-Traditionalist Strands, art. 5).
‘Angels versus Humans as Messengers of God: The View of the Qurʾānic Pagans’. In
Philippa Townsend and Moulie Vidas (eds.), Revelation, Literature, and Community
in Late Antiquity, Tübingen: Mohr Siebeck, 2011, 315–336 (= Collected Studies in Three
Volumes, vol. 1, The Qurʾānic Pagans and Related Matters, art. 4).
‘Al-Jāḥiẓ on Aṣḥāb al-Jahālāt and the Jahmiyya’. In Rotraud Hansberger, M. Afifi al-
Akiti and Charles Burnett (eds.), Medieval Arabic Thought: Essays in Honour of Fritz
Zimmermann, London: Warburg Institute and Turin: Nino Aragno Editore, 2012, 27–
39 (= Collected Studies in Three Volumes, vol. 2, The Iranian Reception of Islam: The
Non-Traditionalist Strands, art. 8). Persian translation in Bukhārā 18, no. 108 (2015),
64–82.
‘Buddhism as Ancient Iranian Paganism’. In Teresa Bernheimer and Adam Silverstein
(eds.), Late Antiquity: Eastern Perspectives, n.p.: E.J.W. Gibb Memorial Trust, 2012,
25–41 (= Collected Studies in Three Volumes, vol. 2, The Iranian Reception of Islam:
The Non-Traditionalist Strands, art. 9).
‘The Quranic Mushrikūn and the Resurrection’. Part i, Bulletin of the School of Oriental
and African Studies 75, no. 3 (2012), 445–472. Part ii, Bulletin of the School of Oriental
and African Studies 76, no. 1 (2013), 1–20. (= Collected Studies in Three Volumes, vol. 1,
The Qurʾānic Pagans and Related Matters, arts. 5 and 6.)
‘The Book of Watchers in the Qurʾān’. In Haggai Ben-Shammai, Shaul Shaked and Sarah
Stroumsa (eds.), Exchange and Transmission across Cultural Boundaries: Philosophy,
Mysticism and Science in the Mediterranean, Jerusalem: The Israel Academy of Sci-
ences and Humanities, 2013, 16–51 (= Collected Studies in Three Volumes, vol. 1, The
Qurʾānic Pagans and Related Matters, art. 7).
‘Pre-Existence in Iran: Zoroastrians, Ex-Christian Muʿtazilites, and Jews on the Human
Acquisition of Bodies’. Aram 26, no. 1 & 2 (2014), 1–20 (= Collected Studies in Three
Volumes, vol. 2, The Iranian Reception of Islam: The Non-Traditionalist Strands, art. 13).
‘Traditional Political Thought’. In Gerhard Böwering (ed.), Islamic Political Thought: An
Introduction, Princeton, nj: Princeton University Press, 2015, 238–251.
‘Problems in Sura 53’. Bulletin of the School of Oriental and African Studies 78, no. 1 (2015),
15–23 (= Collected Studies in Three Volumes, vol. 1, The Qurʾānic Pagans and Related
Matters, art. 12).
‘Jewish Christianity and the Qurʾān’. Part One, Journal of Near Eastern Studies 74, no. 2
(2015), 225–253. Part Two, Journal of Near Eastern Studies 75, no. 1 (2016): 1–21. (=
Collected Studies in Three Volumes, vol. 1, The Qurʾānic Pagans and Related Matters,
arts. 9 and 10.)
‘Excursus ii: Ungodly Cosmologies’. In Sabine Schmidtke (ed.), Oxford Handbook of
Islamic Theology, Oxford: Oxford University Press, 2016 (= Collected Studies in Three
Volumes, vol. 3, Islam, the Ancient Near East and Varieties of Godlessness, art. 6).
Commentaries on q 37:6–11, q 43:81–83, q 52 and q 72. In Mehdi Azaiez, Gabriel
S. Reynolds, Tommaso Tesei and Hamza M. Zafer (eds.), The Qurʾan Seminar Com-
mentary: A Collaborative Study of 50 Qurʾanic Passages, Berlin: Walter de Gruyter,
2016.
‘“Nothing but Time Destroys Us”: The Deniers of Resurrection in the Qurʾan.’ Journal of
the International Qurʾanic Studies Association 1, no. 2 (2016).
‘Tribes without Saints’. In Collected Studies in Three Volumes, vol. 1, The Qurʾānic Pagans
and Related Matters, art. 15.
‘Idrīs, Atraḫasīs and al-Khiḍr’. In Collected Studies in Three Volumes, vol. 3, Islam, the
Ancient Near East and Varieties of Godlessness, art. 3.
‘What Are Prophets For? The Social Utility of Religion in Medieval Islamic Thought’. In
Collected Studies in Three Volumes, vol. 3, Islam, the Ancient Near East and Varieties
of Godlessness, art. 8.
‘Oral Transmission of Subversive Ideas from the Islamic World to Europe: The Case of
the Three Impostors’. In Collected Studies in Three Volumes, vol. 3, Islam, the Ancient
Near East and Varieties of Godlessness, art. 9.
‘Pagan Arabs as God-Fearers’. In Carol Bakhos and Michael Cook (eds.), Islam and Its
Past: Jāhiliyya and Late Antiquity in Early Muslim Sources, Oxford: Oxford University
Press, forthcoming (= Collected Studies in Three Volumes, vol. 1, The Qurʾānic Pagans
and Related Matters, art. 11).
Encyclopaedia Entries
A Companion to Samaritan Studies, ed. Alan D. Crown, Reinhard Pummer and Abraham
Tal, Tübingen: Mohr Siebeck, 1993: ‘Athinganoi’.
Encyclopædia Iranica, online edition (2011): ‘Ḵorramis’, http://www.iranicaonline.org/
articles/korramis (= Collected Studies in Three Volumes, vol. 2, The Iranian Re-
ception of Islam: The Non-Traditionalist Strands, art. 3); ‘Moqannaʿ’, http://www
.iranicaonline.org/articles/moqanna (= Muqannaʿ wa sapīdjāmagān, art. 3 [in Per-
sian]; = Collected Studies in Three Volumes, vol. 2, The Iranian Reception of Islam: The
Non-Traditionalist Strands, art. 4).
Encyclopaedia of Islam, second edition: ‘Khālid b. al-Walīd’, ‘Khiṭṭa’, ‘Masāmiʿa’, ‘Maʿūna’,
‘Mawlā’, ‘al-Muhallab b. Abī Ṣufra’, ‘Muhallabids’, ‘Sulaymān b. Kathīr’, ‘ʿUthmāniyya’,
‘Yazīd b. Abī Muslim’.
Encyclopaedia of Islam, third edition: ‘Anarchism’, ‘ʿArīf’, ‘Atheism (Pre-Modern)’,
‘Bābak’, ‘Barāhima’, ‘Dahrīs’, ‘Dayṣanīs’.
Encyclopaedia of the Qurʾān: ‘War’ (= Collected Studies in Three Volumes, vol. 1, The
Qurʾānic Pagans and Related Matters, art. 8).
Princeton Encyclopedia of Islamic Political Thought, ed. Gerhard Böwering with Patricia
Crone, Wadad Kadi, Devin J. Stewart and Mahan Mirza, Princeton, nj: Princeton
University Press, 2013: ‘Clients’, ‘Philosophy’, ‘Quraysh’, ‘Sunna’, ‘Traditional Political
Thought’.
Other Writings
‘Vom Studium vorindustrieller Gesellschaften’. Börsenblatt (1992), 78–80.

‘The Rise of the Muslim Sects’ [in Chinese]. In Chung-tung yen-chiu tao-lun, Taipei:
Sino-Arabian Cultural and Economic Association, 1993.
‘Til Paradis i lænker: Jihad i historisk perspektiv’. Kritik (Copenhagen) 36, no. 12/162
(2003), 37–43.
‘What Do We Actually Know about Mohammed?’ Open Democracy (online publica-
tion), 31 August 2006, available at https://www.opendemocracy.net/faith-europe
_islam/mohammed_3866.jsp.
‘“Jihad”: Idea and History’. Open Democracy (online publication), 30 April 2007, avail-
able at https://www.opendemocracy.net/faith-europe_islam/jihad_4579.jsp. Re-
printed (without the final section) in Cosmopolis 1 (2015), 99–104 [83–88]. Danish
translation in Weekendavisen, 1–22 June 2007.
‘Islam and Religious Freedom’. Keynote speech at the 30th Deutscher Orientalistentag,
Freiburg im Breisgau, 24 September 2007, published at http://orient.ruf.uni-freiburg
.de/dotpub/crone.pdf [no longer online] (= Collected Studies in Three Volumes, vol. 1,
The Qurʾānic Pagans and Related Matters, art. 14). Also published (without the
beginning) under the title ‘No Pressure, Then: Freedom of Religion in Islam’ at
Open Democracy (online publication), 7 November 2009, available at https://www
.opendemocracy.net/patricia-crone/no-compulsion-in-religion.
‘Remarks by the Recipient of the 2014 mem Lifetime Achievement Award Written for
the Annual Meeting of Middle East Medievalists and Read in Absentia by Matthew
S. Gordon (November 22, 2014, Washington, d.c.)’. Al-ʿUṣūr al-Wusṭā 23 (2015), iii–vi
(= Collected Studies in Three Volumes, vol. 3, Islam, the Ancient Near East and Varieties
of Godlessness, xi–xv).
‘Zandagīnāma-yi khūdniwisht’. Bukhārā 18, no. 108 (2015), 37–63.
‘Safarnāma-yi Tirmiz’. Bukhārā 18, no. 108 (2015), 83–86.
‘How the Field Has Changed in My Lifetime’. In Collected Studies in Three Volumes, vol. 3,
Islam, the Ancient Near East and Varieties of Godlessness, Leiden: Brill, 2016, art. 10.
Festschrift
Behnam Sadeghi, Asad Q. Ahmed, Adam Silverstein and Robert Hoyland (eds.), Islamic
Cultures, Islamic Contexts: Essays in Honor of Professor Patricia Crone. Leiden: Brill,
2015.
Index of Names and Terms
Note: The definite article (al- or ʾl-) is ignored in alphabetisation.
ʿAbd al-Jabbār 84, 91, 169, 170, 205–206 desert-dwelling, appearance of 1–16
ʿAbdallāh al-ʿĀdī/ʿAbdī 176–177 loss of political dominance by 159–160,
Abū ʾl-Faraj al-Isfahānī xiii–xiv 166–167, 169
Abū ʾl-Hudhayl 131, 132 lot-casting among 22–24, 25–26, 29–32,
Abū ʿĪsā al-Warrāq 84, 92–93, 111, 174, 209, 38, 42
217 prejudice against non-Arabs of 163–164,
Abū Qurra, Theodore 88–89, 90, 91 165, 168
Abū Saʿīd al-Ḥaḍrī/Ḥaḍramī/Ḥuṣrī 82–83, preserving Jewish customs 25
90–93, 111, 142 role in Near Eastern history 41, 43, 81
Abū Ṭāhir 224 settled, sartorial norms of 2, 9–11, 16
accidents Aramaeans
elementary qualities as 105, 114, 137 as Arabs 41
motion as sole instance of 136, 140 language and culture of 40, 54, 65, 75
in prime matter/substance 131–132, 137– Aristotle 119, 138, 139, 188n9, 230
139, 206 Aṣamm, al- 127, 130, 136
sides and magnitude of atoms as 130 astrology/astronomy
Adam 86, 231–232 among Dahrīs 108–109, 114, 122, 197, 205–
humans before 230–233 206
Albright, W.F. 48 as esoteric knowledge 56, 75
Alexander, Philip S. 47 historical 232, 233, 234
Alexander Poem (Syriac) 67–68, 68–70 as philosophy/kalām 109, 115
Alexander Romance 49, 66–67, 72 used by prophets 193, 194, 207, 208,
Alexander the Great 66–68, 72, 73 224n122
ʿAlī, lot-casting by 22, 30, 35 See also kalām: as science/philosophy
Andreas the cook atheism. See Dahrīs; mulḥids; Zindīqs
in Alexander stories 66–68, 70, 72 atoms 129, 130–132
as Atraḫasīs 64, 67 Atraḫasīs
as Glaukos 67, 77 as Andreas 67
as Idrīs/Enoch 49 as Enoch 52–54, 55, 56, 59, 62–64, 78
angels as flood survivor 49–52, 56–57, 62
demonic offspring of 54, 88 Idrīs as 49, 78
demons posing as 89, 153, 195, 212 as immortal 49, 51–52, 60–61, 62, 67
fallen 78, 79, 86–87 al-Khiḍr as 75, 78
as subordinates of God 86, 87 as receiving revelations 56–57
animals, religion among 101, 191–192 as taken by the gods 53, 61
anthropomorphism 106 as Ūta-napishti 51–52, 53, 60, 62–63, 66–
anti-Arabism 163–166, 177–178 67, 78
antinomianism Yonṭon as 68
among Ismāʿīlīs 157, 177–180, 222 as Zi(u)sudra/Xisouthros 50–52, 57, 60–
among Sufis 182–183 61, 62–63, 78
apostasy 106, 107–108, 112, 122 Augustine 231
Arabs
converts’ resentments against 163–166, Babylonians, Jewish influence on view of
177–178 Atraḫasīs of 53–54, 60–62
256 index of names and terms
Bandahī, Fakhr al-Dīn b. al-Badīʿ al- 219 reasons for becoming 106, 108, 109, 112,
Bardesanes (Bar Dayṣān) 119, 128–129, 131 121, 122
Bashshār al-Burd 124, 142 role of reason for 96, 103, 107, 115
Bāṭinism 222, 223. See also Ismāʿīlīs and science 108–109, 110, 114–115
Battle of Siffin xi social backgrounds of 112, 115, 122, 205–
Becker, Carl-Heinrich 151 206, 215
Berossos 50, 60–61, 77 See also mulḥids; prime matter (hayūlā,
Bīrūnī, al- 240 ṭīna); ṭabāʾiʿ (elementary qualities);
“Brahmans” 207–208, 211 Zindīqs
Brown, Peter 242 Democritus 189, 190
Buddhism 111, 113 demons
Burzoē 121, 124, 176 deceiving prophets 89–90, 153, 195, 212
as false prophets 89
Casanova, Paul 48 possessing/misleading people 85, 86–87,
Christians 88, 89, 91
Iranianised 119 usurping divine prerogatives 87–89
lot-casting among 17–18, 22, 28 design, argument from 100, 127, 142, 207
role in Near Eastern history 41 dhimmīs 167, 170
and science 113–114 Dhū ʾl-Kifl 44, 48, 73, 78
views on Enoch 79, 80–81 Dhū ʾl-Qarnayn 72–73
view of God and demons 87–89 disputation 109, 121–122, 123, 124
circumcision 25 dualism 119, 128, 131, 132–133, 136. See also
Clement of Alexandria 231 Manichaeans; Zindīqs
clothing, as marker of propriety 9, 16 Durkheim, Émile 187, 188, 190
consultative divination (istikhāra), as lot-
casting 38 Elijah
Cook, Michael 243 Idrīs as 46
cosmology in late antiquity 118–119 John the Baptist as 74
Critias 189–190, 196 al-Khiḍr as 46, 71
Elisha, Idrīs/al-Khiḍr as 73
Dahrīs empiricism 96, 103–104, 114, 124–127
absence of community of 108–109, 114, Enki/Ea 18, 50, 56–57, 60
122 Enlil 18, 50, 56, 60, 62
as apostates 106, 107–108, 112, 122 Enmeduranki 52–53
as aṣḥāb al-hayūlā 96, 138–139 Enoch
as aṣḥāb al-ṭabāʾiʿ 96, 108, 133–137 association with flood of 55, 59
belief in a creator of 141–142, 192–194, 206 as At(a)nabīsh 63
criticising scripture 99, 101–102, 103–104, Atraḫasīs as 52–54, 55, 56, 59, 62–64, 78
107 books of 54–55, 79–81, 86
denying God and prophethood 96, 98– Enmeduranki as 52–53
99, 103, 104–105, 114, 118, 193, 206–207 as Hermes/Thoth 46–47
empiricism of 96, 103–105, 114, 206 Idrīs as 44, 46, 47–48, 78
eternalism of 96, 98, 99, 111, 114, 139–140 as Metatron 80
Greek influence on 113, 115, 136 taken away by God 44–45, 47, 53, 57–59,
heavenly sphere for 98, 99–100, 104–105, 61, 79–80
106, 114, 134, 142 as Ūta-napishti 63
historical reality of 96–97, 115 as visionary 56
morality of 98–99, 100–101, 142 equipollence of proofs (takāfuʾ al-adilla) 82,
as (former) pagans 120, 121 90–91, 92, 126, 170, 173, 207–208, 210
index of names and terms 257
Epicureanism Gregory ix (pope) 200, 225, 227, 229

atoms in 129–131, 132–133, 136
on human nature 189 Ḥaddād, Abū Ḥafṣ al- 111, 142, 210
influence on Dahrīs and Zindīqs 119 Hartmann, Richard 49
on positive religion 192, 198, 211, 214 Hārūt and Mārūt 78, 79, 81
Erder, Yoram 48 Ḥasan al-Baṣrī, al- 45, 59
Ess, Josef van 82–83, 84, 209, 210n49 heaven, ascent to
eternalists. See Dahrīs by Atraḫasīs 61
by Elijah 46
Fārābī, al- 174, 215 by Enmeduranki 52–53
fire by Enoch 44–45, 53, 56, 57–58
as elementary substance 105, 114, 132 by Idrīs 45–46, 47
nature of 133, 140 by martyrs 45–46
worship of 68 by Utuabzu 52–53
fish, in spring of life story 66–68, 69, 70n118, heavenly sphere 98, 99–100, 104–105, 106,
71, 72–73 114, 134, 142
Frederick ii 200, 225–229, 230, 233 Hermes 46–47, 48, 75, 134
Free Spirits 234 Ḥīwī of Balkh 212–213
free will 106 Huggins, Ronald V. 63
Friedlaender, Israel 72n127, 76
Ibn Khābiṭ 191
Ghazālī, al- 175, 182–184, 195–196, 218–219, Ibn Mānūsh 130
223–224 Ibn al-Rāwandī
giants on God 92–93
in watcher story 54 on prophets 84, 143, 174, 179, 194, 208–
Manichaean Book of 63 209, 210
Qumran Book of 58–59, 63 reception of 198, 217
Gil, Moshe 48 scepticism of 207–211
Gilgamesh as unhinged Muʿtazilite 92–93, 111, 142
Alexander the Great as 66–68 Ibn Rushd (Averroes) 34, 39, 176, 201
in Book of Giants 59 Ibn Waḥshiyya 232
in magic incantations 65 Idrīs
Moses as 69 as Andreas/Andrew 48–49
Nimrod as 68–69 as Andreas the cook 49
search for immortality of 51, 66–68 as Atraḫasīs 49, 78
Glaukos 67, 76, 77 elevated by God 44, 45–46, 47, 59
God as Elijah 46
as active principle 128 as Elisha 73
denial of (see under Dahrīs; Zindīqs) as Enoch 44, 46, 47–48, 78
fairness of 83, 90–92, 93–94 as Ezra/Esdras 48
likened to human king 82–83, 84–89, 188 as Hermes 48
punishing infidels 83, 90–92, 94 as al-Khiḍr 46
gods, lesser 84–86, 87. See also demons in the Qurʾān 44, 47, 81
Grabar, Oleg 1–2 Ikhwān al-Ṣafā 101, 174–175, 191–192
Greeks 40, 41 ilḥād. See mulḥids
lot-casting among 19–20, 27–28, 37 imams (Ismāʿīlī) 157, 158, 184
as source of Dahrism 113–114, 115 immortality, search for 66–68, 70, 75
as source of “prophets as impostors” thesis intentions, as determining salvation 90–91,
225–226 92, 94
interpenetration of bodies 128–130, 140 Khiḍr, al-

Iranians. See Persians as Andreas the cook 72n127
Īrānshahrī, Abū ʾl-Abbās al- 140, 174 as Atraḫasīs 75, 76, 78
irjāʾ (suspension of judgement) 125 as angel 72
Islamic empire as Babylonian 74–76
post-colonialism in 158–162, 181 as Dhū ʾl-Qarnayn’s commander 72–73
tenth-century cultural fusion in 166–171, as Elijah 46, 71, 74
183, 213–214 as Elisha 73
Ismāʿīlīs 152 as Enoch 74
backgrounds of 172 as Glaukos 76, 77
denounced as subversive 168–169, 223– as helpful wanderer 71
224 as Idrīs 46, 73, 74
imams of 157, 158, 184 as immortal 46, 70–72
on levels of religion 156–158, 182, 214, 222, as Jeremiah 73–74
235 name of 76–77
messiah (Mahdī) of 157–158, 177, 178–180, as servant of God 71, 72, 73
214, 222 as servant of Moses 70–71
on prophets 154–155, 156, 172–173, 179– as Son of Man 74
180, 222–224 Khurramīs 234
relationship to mainstream Islam of 177– kumūn (latency), doctrine of 139, 140
180, 222–223
Lewis, Bernard xi, xii, 243
Jāḥiẓ, al- Lidzbarski, Mark 76, 77
on Dahrīs 98–106, 107–108, 111–112, 113, lot-casting
114–115 connecting ancient and Islamic culture
on kalām 109–110, 113, 118 18, 41
on religion 100–101, 186–189, 190–192 as contemporary political device 38–39
Jesus 45, 128, 154, 230, 233–234 to divide inheritances 17–18, 20–22, 23–
as impostor 83, 179, 200, 203, 221, 223– 25, 34
227, 229, 233–234 as gambling 35–36, 38–39
as possessed by a demon 85, 86, 89 by gods 18–19, 21
Jews in Islamic law 34–36, 37–38
Christian polemics against 178 as Islamic practice 18, 22–24, 25, 29–32,
disapproval of sciences of 109 38, 42
lot-casting among 19–22, 25–27, 28, 37, in the Qurʿān 32–34
42 to select people 26–32, 34–35, 39
scriptural critics among 212–213 to share land and booty 18–20, 22–23,
views on Enoch of 53–54, 59–60, 61–62, 24–26, 34
79–80 as standard practice in the Near East 18,
view of God and demons of 86–87 24–25
jinn, eavesdropping in heaven 103–104, See also under ʿAlī; Arabs; Christians;
127 Greeks; Jews; Muḥammad; Ottomans;
John of Salisbury 235, 238 Persians; Romans; Shīʿites
Jubilees 55, 59
Maʿarrī, al- 175–176, 195, 198, 214–215, 221,
kalām 225, 226
dangers of 110–111, 112–113, 115, 122 Macdonald, Michael 7–9
as science/philosophy 109–110, 115, 118 Mahdī (Ismāʿīlī) 157–158, 177, 178–180, 214,
as theology 109, 115, 118, 126 222
Maḥmūd Pasīkhānī 143 “Muslim Enlightenment” 213–214. See also

Malka 2–3 Islamic empire: tenth-century cultural
Manichaeans fusion in
cosmology of 122, 129–130, 131, 138 Muʿtazilites
on Enoch 63, 81 and Dahrism 97, 107, 111, 115, 123, 140, 142,
and Zandaqa 114, 119, 122, 123, 169, 206, 207
212, 217 God’s justice for 91–93
Marcion 119, 179 interest in science of 111
Marcionites and Zandaqa 114, 119, 122, 128,
206, 212 Naṣībī, Abū Iṣḥāq al- 142, 174, 210
Martān Farrūkh 91–92 nationalism, Arab 242
martyrs, transfer to heaven of 45–46 natural religion 101, 173–174, 183, 191–192,
Mary, chosen/assigned by lot 28, 32–33 213
Massignon, Louis 221–222, 224 natures. See ṭabāʾiʿ (elementary qualities)
Māwardī 175, 190–191 Naẓẓām, al- 99, 101, 105, 123, 127, 139, 140
mawlā, status of 163 Dahrī brother-in-law of 106, 123
medicine 108–109, 110, 115, 124–125, 154, 176 Neoplatonism 119, 172, 176
metamorphosis (maskh) 102 Neuwirth, Angelika 241
Metatron 80, 86 Nimrod 68–69, 73
Michael Scotus 227, 228 Noah 52, 55, 58, 59
monotheism Nöldeke, Theodor 48–49, 76, 245
vs. Dahrism 105–105, 141
vs. elementary qualities 109–110 Orientalism 239–241, 245
Montgomery, James A. 49 Ottomans, lot-casting among 30, 32, 34, 42
Moses
as false prophet 83, 89, 179, 200, 202–203, pagans, Near Eastern 84–85, 87, 120, 203–
221, 223–227, 229, 233–234 204
as Gilgamesh 69–70 paramonē 42–43
servant of 69–71, 72–73 (see also Khiḍr, Persians
al-) arabization of 160–161, 163–166
Muḥammad lot-casting among 21, 28
as deceived by demons 89–90 Petra, papyrus finds in 17
as impostor 83, 179, 200, 204, 210, 221, philosophers 140, 172, 174, 211, 219–220
223–229, 233–234 as intellectual elite 176, 189–190, 214
lot-casting by 22, 23, 30, 34, 35–36, 38 prophets as 197, 219, 227–228
as lawgiving sage 196, 211 as source of three impostors thesis 179–
as leader 154, 155, 170–171, 187, 204 180
See also prophets; three impostors thesis vs. theologians (mutakallims) 118, 172, 211,
mulḥids 213
backgrounds of 119, 121 philosophy 109, 118–119, 153, 176, 182,
beliefs of 98, 99, 118 195, 213–214. See also kalām: as
converting others 112, 113 science/philosophy
empiricism of 124–125, 126–127 Platonism 118–119, 128, 133, 137, 231
and kalām 110, 113 Polybius 189, 196
reasons for becoming 106 post-colonialism 151, 158–162, 181, 241–243
as sceptics 99, 124–126 prime matter (hayūlā, ṭīna) 136, 137–138
See also Dahrīs; Zindīqs proponents of 137–140
Murjiʾites 92, 125 prophets
Musil, Alois 25–26 among animals 101, 191
in ancient Greek world 202–203 and power 155–156, 169–171, 204–205

as bearers of divine law 155–157, 187, 190, socio-political functions of 155–156, 176,
202–205, 218, 228, 233–234 186–192, 196–197, 204–205, 211, 234–237
as bearers of manmade law 156, 196–197, true vs. false 83, 88–89, 90–92, 153, 156,
211, 218, 220, 226 170, 194–195, 196, 200, 214
ceremonial cursing of 179 as unnecessary 99–101, 174–175, 191–192,
concept of 152–153, 158, 173–175, 187, 194, 194, 207–208, 211, 214
205 Roman empire, “Orientalisation” of 40–41
deceived by demons/angels 89–90, 153, Romans 40–41
195, 212, 220 lot-casting among 20–21, 24, 25, 27–28,
deniers of (see Dahrīs; mulḥids; Zindīqs) 37
vs. imams 157, 158
as impostors 83, 84, 174–175, 179, 194, 200, Said, Edward 240, 241
202–204, 208, 218, 223–229, 233 (see also Sarakhsī, al- 174, 179, 211, 217
three impostors thesis) ṣarfa, doctrine of 102–104, 107
as leaders 154–156, 170–171, 202, 204–205, scepticism 91, 124–126, 170, 173, 181, 207–209.
233 See also equipollence of proofs (takāfuʾ
non-Arabs among 165, 177 al-adilla)
as philosophers 227–228 science 108–111, 113–114, 118. See also kalām;
Protoevangelium of James 28, 32–33 ṭabāʾiʿ (elementary qualities)
“provincial law” 40–41, 42 Shīʿites
criticising Companions 170
Qarāmiṭa 178–179, 222, 224. See also lot-casting among 38
Ismāʿīlism and Zindīqs 123
qurʿa. See lot-casting Shuʿūbī movement 162–166, 181
Qurʾān, interpretation of 241 Sijpesteijn, Petra 245
Sincere Brethren 101, 174–175, 191–192
Raspe, Heinrich 229 Sirr al-khalīqa 131–132, 134–135, 137, 142
Rāzī, Abū Bakr al- slaves, conditional manumission of 42–43
on prophets 89–90, 101, 152–153, 156, 158, Solomon 102
191, 195, 212, 214 spirit (rūḥ) as fifth nature 105, 133–134
on true religion 153, 194–195, 213–214 Stoicism 119, 128–129, 133–134, 135–136, 137,
Rāzī, Abū Ḥātim al- 152, 154–155, 156–158, 139, 231
171, 172–173 Sufism
Rāzī, Fakhr al-Dīn al- 93–94 Dahrī views in 143
reason epistemology of 184–185
as criterion of knowledge 96, 103, 107, 114, philosophical ideas about prophets in
118, 121–122, 174, 194, 208 219–220
limits of 90–91, 92, 170, 176, 184–185 role of law in 181–184, 196
as philosophy 153, 176, 195 Sūfisṭāʾiyya 124, 125
vs. revelation 93, 114, 118, 122, 175, 207 “Sunnī revival” 181
Reeves, John C. 47, 63 suspension of judgement (irjāʾ, wuqūf ) 125–
reincarnation 136, 143 126
religion
expression of doubts on 112, 122, 198–199, ṭabāʾiʿ (elementary qualities) 102, 105, 106,
216–217, 228 114, 133–136
higher vs. lower levels of 156–158, 176, God as creator of 134–135
183–184, 214, 235 proponents of 96, 108, 131, 133–137
natural 101, 173–174, 183, 191–192, 213 science of 109–111
takāfuʾ al-adilla 82, 90–91, 92, 126, 170, 173, al-Khiḍr as 75, 78
207–208, 210 Yonṭon as 68
taqiyya 198, 217 Utuabzu 52–53
“tenth-century crisis” 166–171
theodicy motif 69, 72 Wansbrough, John xii
Theodore of Antioch 228 watchers, story of 54–55, 56, 57, 78, 79, 86, 88
theologians (mutakallims) water of life 49, 66–68, 69–70, 72
doubting religion 171, 173–174 wuqūf (suspension of judgement) 125
and philosophy/physics 109–110, 118, 126,
130, 170 Yazīd b. Unaysa 177
vs. philosophers 118, 172, 211, 213 Yonṭon (son of Noah) 68–69
refuting mulḥids 113, 123
as tricksters 83–84 Zaninus 233–234
See also kalām; Muʿtazilites Zindīqs
Thomas Scotus 229–230, 233 backgrounds of 119, 122, 206
three impostors thesis Christians as 114, 115, 119
in Islamic Near East 83, 224, 226–227 cosmology of 128–133, 139
among Ismāʿīlīs 221–224 denying God and prophethood 111–112,
in medieval Europe 90, 200–201, 225– 118, 120, 141–142, 206–207, 219
230, 233–234 persecution of 123, 217
Muslims as source of 201, 225–227 rationalism of 92–93, 142
among philosophers 179–180 reasons for becoming 106, 107, 109, 110,
transmission 114–115, 122
as guarantee of content 107 and science 109, 118
oral, of ideas 237–239 Shīʿite view of 123
Turks 180–181, 192 Zi(u)sudra/Xisouthros 50–52, 57, 60–61, 62,
77, 78
ʿUmar Khayyām 93, 175, 185 Zoroastrianism
Ūta-napishti Islamic influence on 168, 169
as Atraḫasīs 51–52, 53, 60, 62, 66–67, non-Persian forms of 120
78 as source for Muslim heresies 122, 169,
as Enoch 63 177, 206

Islamic History and Civilization

Uploaded by

Copyright:

Available Formats

Islamic History and Civilization

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

Islamic History and Civilization

Uploaded by

Copyright:

Available Formats

Islam, the Ancient Near East and Varieties of Godlessness

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

The Library of Congress Cataloging-in-Publication Data is available online at http://catalog.loc.gov

Copyright 2016 by Koninklijke Brill nv, Leiden, The Netherlands.

This book is printed on acid-free paper and produced in a sustainable manner.

2 The Ancient Near East and Islam: The Case of Lot-Casting 17

3 Idrīs, Atraḫasīs and al-Khiḍr 44

4 Abū Saʿīd al-Ḥaḍrī and the Punishment of Unbelievers 82

5 The Dahrīs According to al-Jāḥiẓ 96

6 Ungodly Cosmologies 118

7 Post-Colonialism in Tenth-Century Islam 151

8 What Are Prophets For? The Social Utility of Religion in Medieval

9 Oral Transmission of Subversive Ideas from the Islamic World to

10 How the Field Has Changed in My Lifetime 239

List of Patricia Crone’s Publications 247

Medievalists (mem) Lifetime Achievement Award

We gratefully acknowledge permission to reprint articles that originally ap-

3. ‘Idrīs, Atraḫasīs and al-Khiḍr’. Previously unpublished.

4. ‘Abū Saʿīd al-Ḥaḍrī and the Punishment of Unbelievers’, Jerusalem Studies in

5. ‘The Dahrīs According to al-Jāḥiẓ’, Mélanges de l’Université Saint-Joseph 63

6. ‘Excursus ii: Ungodly Cosmologies’, in Sabine Schmidtke (ed.), Oxford Hand-

7. ‘Post-Colonialism in Tenth-Century Islam’, Der Islam 83, no. 1 (2006), 2–38.

10. ‘How the Field Has Changed in My Lifetime’. Previously unpublished.

“Barefoot and Naked”: What Did the Bedouin of the 1

Arab Conquests Look Like?*

© koninklijke brill nv, leiden, 2016 | doi: 10.1163/9789004319318_002

To start in Syria, there is a representation of semi-naked bedouin in an ivory

or (under Syro-Palestinian influence) Alexandria (fig. 1).10 It depicts Joseph’s

If the inhabitants of the Roman empire envisaged the Saracens as wearing

figure 4 Relief from an alabaster incense burner. British Museum,

of Judgment.21 But as Macdonald himself stresses, this is highly conjectural.

21 Michael Macdonald, personal communication with reference to a discussion at the recent

figure 5 Alabaster funerary relief. Louvre, ao 1128.

figure 6 Rock drawing depicting a Safaitic horseman

figure 7 Rock drawing

figure 8 Rock drawing, detail

Arabs on Assyrian reliefs (although precisely what it is meant to represent is

The Ancient Near East and Islam: The Case of 423

With Adam Silverstein

In 1993 classical archaeologists made an exciting discovery at Petra. This city,

© koninklijke brill nv, leiden, 2016 | doi: 10.1163/9789004319318_003

Assigning Land, Booty, and Other Property by Lot

he called Arabia Petraea (former Roman Arabia) agricultural land belonging

Choosing People by Lot

40 A. Musil, Arabia Petraea (Vienna 1907–1908), 3, 294.

46 Josephus, Wars, book 4, ch. 3, pars. 7 f.

53 Cicero, Ad Quintum fratrem, 1, 9, 27.

When ʿUthmān permitted Muʿāwiya to conduct campaigns by sea, he stipu-

The Qurʾān and the Law

78 al-Majlisī, Biḥār, liii, 331.

82 Protoevangelium of James, 8:2–9:1, in Schneemelcher and Wilson, New Testament Apoc-

The fact that lot-casting is mentioned in the Qurʾān in connection with

86 Al-Balādhurī, Ansāb al-ashrāf, ed. M. Ḥamīdallāh (Cairo 1959), 85.

accepted lot-casting taṭyīban li-nufūs al-mutaqāsimīn.90 It was on the same

from, or relate it to, consultation (istikhāra) or divination (istiqsām, kihāna),

The Return of the Near East

disappearance of the practice, though it would help if it turned up in Syriac

Idrīs, Atraḫasīs and al-Khiḍr*

Idrīs is a mysterious figure mentioned twice in the Qurʾān, in both cases in