Professional Documents
Culture Documents
The Flowering of Muslim Theology Van Ess
The Flowering of Muslim Theology Van Ess
,::J ( '\
THE FLOWERING OF
MUSLIM THEOLOGY
Josef van Ess
Translated by Jane Marie Todd
HARVARD UNIVERSITY PRESS
Cambridge, Massachusetts, and London, England
2006
'"'
(
Copyright 2006 by the President and Fellows
of Harvard College
ALL RIGHTS RESERVED
Printed in the United States of America
Originally published as Prtfmices de Ia theologie musulmane,
Editions Albin Michel S.A., 2002
This book was published with the support of the
French Ministry of Culture-National Book Center I
Cer ouvrage a ere publie avec !'assistance du Ministere
charge de la culture-Centre National du Livre.
Library of Congress Cataloging-in-Publication Data
Ess, Josef van.
[Premices de la theologie musulmane. English]
The flowering of Muslim theology I Josef van Ess;
translated by Jane Marie Todd.
p.cm.
"This book began as a series of four lectures given by Josef van Ess
at the Institut du monde arabe, Paris, in 1998."
Includes bibliographical references and index.
ISBN 0-67 4-02208-4
1. Islam-Theology. 2. Islam-Doctrines. I. Title.
BP166.E7713 2006
299.2-dc22 2005052792
[ ~ Q A S !JBRARY]
I
2
3
4
CONTENTS
Note to the English-Language
Edition
Introduction
Theology in Its Own Eyes:
Division and Heresy in Islam
Theology and the Koran:
The Mi'raj and the Debate on
Anthropomorphism
Theology and Science:
Mu'tazilite Atomism
Theology and Human Reality:
Historical Images and
Political Ideas
Vll
I
9
Ill
45
79
II7
CONTENTS
5 Theology and Its Principles:
Hermeneutics and Epistemology 153
Notes
193
Index 209
vi
NOTE TO THE ENGLISH-LANGUAGE
EDITION
THIS BOOK BEGAN as a series of four lectures given
by Josef van Ess at the Institut du monde arabe, Paris,
in 1998. These lectures constitute the first four chap-
ters; the author added a fifth chapter and an intro-
duction for publication as the book Premices de la
thiologie musulmane (2002). In a way, this book is an
introduction in French to van Ess's masterwork, The-
ologie und Gesellschaft im 2. und 3 ]ahrhundert Hidschra
(Theology and society in the second and third centu-
ries AH).
THE FLOWERING OF
MUSLIM THEOLOGY
INTRODUCTION
IN THE MASS MEDIA, Islam is the "Other," repellent
and strange. The notion commonly associated with it
is the Sharia, Islamic law based on the Koran, which
would seem to be incompatible with the rules of en-
lightened reason. That view contrasts sharply with
what Taha Husayn wrote in The Future of Culture in
Egypt (1938): "Everything seems to indicate that there
is nothing to distinguish a European mind from the
Eastern mind."
1
And he added: "I am certjin that
there is no difference in essence or in nature between
us and the Europeans."
2
Mustafa 'Abd al-R.aziq re-
peated that idea in 1945, in the official speech he gave
to inaugurate his new duties as sheikh al-Azhar, that
is, as the head of a university that virtually embodies
the spirit of the Sharia: "I see no real reason to set
Islam and the West in opposition to each other."
3
At the time, these two Egyptian intellectuals could
count on the assent of some Europeans. Johann
Wolfgang von Goethe had declared: "Orient unci
I .
INTRODUCTION
Okzident sind nicht mehr zu trennen" (East and
West can no longer be separated). Like later Roman-
tic writers, though, Goethe was thinking about po-
etry, especially Iranian poetry; by contrast, Taha
Husayn and Mustafa 'Abd al-Raziq were thinking of
medieval philosophy and what emerged from it. In
the Germany of the Romantics, Ibn Rushd (Averroes)
had been lauded as the representative of Islamic en-
lightenment and considered a "rationalist" who had
transmitted the legacy of enlightened Islam to a Eu-
rope still in limbo.
4
The film Destiny) by the Egyptian
filmmaker Youssef Chahine, also disseminates that
image. For Chahine, however, Islam has two faces, a
notion that has become quite familiar to us.
It is not enough to speak of philosophy, or merely
of fundamentalism and the law. A place must also be
assigned to theology, 'ilm al-kalam) as it was called,
the "science of dialectical speech" practiced by those
who knew how to "hold conversations" about their
religion with those defending other interpretations,
even if their interlocutors were not Muslim.5 The
word suggests that the "dialecticians" were engaged
in apologetics. That is only partly true, however; the-
ology would soon make other claims. The role it en-
visaged for itself was to provide an authentic explana-
tion of the world. Hence it was naturally tal\:en to be a
2
INTRODUCTION
"philosophy" at a time when the true falsafa) that of
al-Kindi and his circle, of al-Farabi and others, on up
to Ibn Rushd-the only one that deserves to be called
a philosophy in our modern view-had not yet made
its appearance.
At that time, as in the eighteenth century in
France, "philosophers" were simply intellectuals, and
among them a group of mutakallimun (those who
practiced kalam) had the very highest status, along-
side physicians and "men of science" -that is, astrolo-
gers and alchemists.
6
In his Ketaba de slmata (Book
of treasures), Job of Edessa complained about the
"new philosophers" who were gravitating around al-
Nazzam, "vainly seeking the glory of the world."
7
For
Job of Edessa, a Christian physician and defender of
the Greek legacy, such men were mere grandstand-
.,
ers, undeserving of the reputation they enjoyed. For
Muslims, however, they performed a function similar
to that of the Church Fathers. They were not as im-
mersed in the ideas of Plato and Aristotle as their
Christian counterparts, but they had the same ad-
vantage Origen and Clement of Alexandria had had
some centuries earlier: they still had historical op-
tions available to them, options that would become
more limited later on. As a result, they enjoyed a free-
dom of thought that later generations could only
3
INTRODUCTION
dream of For that reason, the kaldm phenomenon
reached its zenith very early; its most creative period
did not occur after it had come of age, but well be-
fore, at a time when signs of tedium and paralysis
had not yet appeared.
That fascinating world has now collapsed, sub-
merged by the waves of puritanism that prefigured
modern fundamentalism. In Saudi Arabia, kaldm is
excluded from the university curriculum-as is phi-
losophy. We must not forget, however, that theology
in the sense described here has always been confined
to certain regions and certain moments in history,
particularly Iraq during the early centuries of the
Abbasid dynasty, and then Iran, where it flourished
again until the Mongol empire. We are interested
here in the beginnings, the time of open options
when kaldm was still deeply rooted in Muslim society.
It held a rank equivalent to that of jurisprudence.
The first Iraqi jurist, Abu Hanifa, who made a lasting
mark on Islamic law, left behind only his theological
treatises (Letter to 'Uthmdn al-Battf and the dialogues
collected by his students, Fiqh al-absat and Kitdb al-
'alim wa'l-muta'allim). His legal teachings seem to
have been transmitted only through the oral tradi-
tion. The two disciplines were still rivals, and it was
not at all obvious that jurisprudence would prevail.
4
INTRODUCTION
Both benefited from their close relationship to their
environment. Theologians were not concerned only
with God and eternal truths; far from it. They also
dealt with believers' everyday problems and personal
worries. The theologians would begin to lose ground
only after they grew hungry for power and allied
themselves with the Abbasid court. The shock pro-
duced by the mihna, the inquisition launched under
al-Ma'mun, was immeasurable. Those accused of be-
ing the instigators-namely, the Mu'tazilite theolo-
gians-were not directly responsible for it. The initia-
tive for that inquisition had been taken, rather, by
the caliph. But the persecution campaign cost the
Mu'tazilites the sympathy of all its victims, members
both of the lower classes and of the middle class.
Before religious thinkers congregated in
41
the cap-
ital, local orthodoxies of a sort had been created in
cities such as al-Kufa, Basra, and Damascus, and they
varied according to which option was chosen. This
state of affairs is especially noteworthy when it comes
to political theories; the positions differed from one
city to the next. At a particular time, the following
ideologies were "true" wherever they had arisen:
Qadarism, Murji'ism, and the different forms of
Shiism. Later, however, they were all considered here-
sies. Such was the influence exerted by the capital. In
5
INTRODUCTION
Baghdad, the imported "orthodoxies" were amal-
gamated and underwent a process of attrition. The
court acted as a catalyst; scholars in the "provinces"
who wanted to pursue a career in the capital were
forced to abandon local particularities or to refrain
from openly propagating them. The geographer Ibn
al-Faqih wrote, "The good thing about Baghdad is
that the government does not have to fear that one
school will prevail over another as has happened in
al-Kufa. There, the 'Alids, alongside the Shiites, often
manage to rule the population. In Baghdad, all of
them-Shiites, Mu'tazilites, and Kharijites-coexist;
each camp holds the others in check and prevents it
from asserting its dominance."
8
Mu'tazilism, though
originally an import from the city of Basra, was able
to take advantage of the situation, creating the first
orthodoxy in the Muslim world that extended be-
yond one locality. In broadening its theoretical base
and establishing a new balance between the various
views, it managed to avoid becoming diluted. The ra-
tionalism in which it took such pride was its most ef-
fective tool.
The aim of this book is to clarifY a few aspects of
that development. Its subject is historical, nothing
more. Nevertheless, the Muslim worldview has con-
served many traces of it. Mu'tazilism was replaced by
6
INTRODUCTION
other orthodoxies, but the ideas it developed had a
subversive influence on later movements. Although
it was threatened in Baghdad in the late third cen-
tury AH (ninth century cE), it managed to prevail in
Khwarazm in about AH 400 (ww cE), thanks to its
missionary efforts. In AH 803 (I400-J40I CE), during
the Syrian campaign ofTimur Lenk (Tamerlane), the
world conqueror was accompanied by a Hanafite ju-
rist who was a Mu'tazilite. Ibn Khaldun met him in
Damascus during the famous audience that Timur
Lenk granted him. In the end, Mu'tazilite orthodoxy
also became a local orthodoxy. Theological problems
may be eternal, but they are not static. The responses
that befit a given situation at a given time quickly
become rigid stereotypes. But sometimes they also
prove to be viable alternatives to those offered in our
own time. Therein lies their importance, for-Muslims
and for us as well. Modern thought is heir to a rich
past, and recalling views and decisions that were once
taken seriously will always prove useful.
7
1
THEOLOGY IN ITS OWN EYES
Division and Heresy in Islam
HERESY OCCUPIES as vast a field in Islam as in
Christianity. Is it really legitimate to confine it to the
classical age? During the mihna) or inquisition, the
marriage of Ibn Malaj, a traditionalist who rejected
the idea of the created Koran (khalq al-Qur)dn)) was
voided by the courts.l There is a striking parallel be-
tween that action and the present-day case of Nasr
fl
Abu Zayd in Egypt, despite the twelve centuries that
separate them. In our own time, the Egyptian Islam-
ist group Jama'at al-Takfir wa'l-Hijra inspires fear;
but even in the ninth century CE, the Mu'tazilites had
earned a reputation for their use of the takftr; that is,
for the anathematizing of their adversaries. Abu Hay-
yan al-Tawhidi deplored the practice. Abu Hashim,
he said, had called his own father, al-Jubba'I, an in-
fidel, and vice versa; Abu Hashim's sister, he contin-
ued, had anathematized both her father and her
9
THE FLOWERING OF MUSLIM THEOLOGY
brother. She was an emancipated woman; she headed
a women's organization of sorts, whose aim seems to
have been Mu'tazilite evangelization. As we know, the
Mu'tazilites not only quarreled among themselves,
but also directed considerable aggression toward the
"others," the hashwiyya, the uneducated, the rabble. Is
that because for at least a century the Mu'tazilites
had been the elite, the intellectual "orthodoxy" of
their time? The majority of theologians were still
Mu'tazilites when al-Tawhidi was writing. He con-
cluded with a somewhat critical remark: "I do not
really understand what takfir means for them! Why
can't we put an end to that ordeal [fitna]?"
2
Which of the two positions corresponded to the
typical attitude of Muslim society during the classi-
cal period-the takfir practiced by certain intellectu-
als or al-Tawhidi's discomfort with that behavior? I
do not know whether I shall answer that question
satisfactorily. But let me briefly consider the term
used at the time, kuft; or unbelief. Christians would
have preferred "heresy," and today we speak of tol-
erance and intolerance. My treatment will pay par-
ticular attention to word choice. There are always
realities behind language, however-the context, the
social structures or institutions (if they exist), and
IO
THEOLOGY IN ITS OWN EYES
the founding doctrines: in this case, scripture, the
Prophetic tradition, and historical myths.
The word for heresy in the European languages
comes from the Greek. But in antiquity, hairesis
meant simply "choice"; the word had no negative
connotations. Any school of philosophy could be
called a hairesis, and it was not considered scandalous
or blameworthy to prefer one hairesis to another. It
was only the Church Fathers who used the word to
mean a reprehensible or foolish choice, a bad im-
pulse-a bawdy as the Arabs and the Koran would
later say. In the religions of antiquity no mention was
made as yet of aberration or schism. Rather than re-
ject a foreign faith, they incorporated and trans-
formed its elements, a process known as syncretism.
The only exception was Judaism, which introduced
a new factor, revelation. At the time, Jewish! theolo-
gians were wont to speak of their covenant with God.
But the Israelites were not alone in doing so. The
Christians understood the birth of their religion as a
"new covenant," and the Muslims too possessed their
'ahd or m'ithaq, with a new book and new command-
ments. The words "pact" and "covenant" are legal no-
tions; they imply the existence of an obligation, a
taklif. Heresy or aberration can therefore become
I I
THE FLOWERING OF MUSLIM THEOLOGY
apostasy, irtidad in Arabic, a term stemming from a
particular historical situation, the ridda, the rebellion
of Arab tribes against islam-that is, against the com-
munity as it existed immediately after the death of
the prophet Muhammad.
Like kufi> the concept of apostasy refers to a typi-
cally Islamic way of seeing things. Nevertheless, the
problem is characteristic of all three Abrahamic reli-
gions. From a doctrinal and institutional point of
view, the religion that proved to be the most rigid
and the best "armed" in that respect was not Islam;
it was without a doubt Christianity. Given the pre-
cariousness of its intermediate position in history,
Christianity had to prove its supremacy over the
other monotheistic religions by stressing that Juda-
ism had come before it and that Islam had come after
it. It was only in Christianity that orthodoxy was de-
fined by dogmas, some of which were summed up in
a profession of faith, a "symbol." That symbol was
even incorporated into the Christian rite, as the Ni-
cene creed uttered during Mass. Dogmas were for-
mulated and confirmed by councils, and the counCils
were in turn legitimated by an institution, the
Church. The Church dispensed salvation; anyone
who repudiated it forfeited redemption. As a result,
heretics could be denied a place within the Church
12
THEOLOGY IN ITS OWN EYES
and could even be denied the use of the word "Chris-
tian," as was the case with the Gnostics. Eventually,
the Church laid claim to what was called the secular
arm. In 1215, the fourth Lateran council obliged secu-
lar authorities to take measures against the Cathari.
The noun "Cathari," in fact, is the source of the Ger-
man word for heretic, Ketzer.
Neither Islam nor Judaism has ever had a church.
Above all, both lacked the concept of redemption, a
founding concept for the Christian Church. In Islam,
there exists no special category of individuals, no spe-
cial profession, whose task it is to dispense salvation;
all Muslims are laypersons. There is also no universal
creed other than the shahada, which consists in recit-
ing, "There is no god but God, and Muhammad is
His messenger." From time to time, theologians or
muhaddithun (specialists in the traditions O sayings
of the Prophet, hadith) did write professions of faith
('aqa'id) that can be compared to the Christian creed,
but these texts entailed no obligation and remained
valid only for a circumscribed time and place. No of-
ficial institution ever constrained secular authorities
to persecute heretics, though that did not prevent
particular governments from becoming guardians
of orthodoxy. It must be conceded that everything I
have said so far has no direct and necessary conse-
13
THE FLOWERING OF MUSLIM THEOLOGY
quence for the analysis of actual practices. It is per-
fectly possible to think dogmatically without having
dogmas, and I will not venture to argue that lay-
persons are any less fanatical than members of the
clergy. Islam as well as Christianity has an impressive
record of executions, pogroms, and burned books.
The systematic and structural differences I have de-
scribed are real, but they were eclipsed by the preva-
lence of a trait characteristic of all Abrahamic reli-
gions: the devotion demanded by revelation. All these
religions belonged to the same family, as it were. Still,
we must always keep the differences in mind. In what
follows, therefore, I shall speak of the small but sig-
nificant difference.
What is perceived as a systematic difference is only
the result of a historical evolution. Neither the Gos-
pels nor the Koran establishes a system. The content
of the faith was defined later, through the work of ex-
egesis. Although, when compared to the New Testa-
ment, the Koran is much more explicit in its rules
prescriptions, it does not contain a list of canon-
ical doctrines. Even when it comes closest to an enu-
meration of prescribed truths, in the verse known as
Ayat al-birr (sura 2:177), there is a different emphasis:
"Righteousness [birr] does not consist in whether you
14
THEOLOGY IN ITS OWN EYES
face towards the East or the West. The righteous man
is he who believes in God and the Last Day, in the an-
gels and the Book and the prophets; who, though he
loves it dearly, gives away his wealth to kinsfolk, to
orphans, to the destitute, to the traveller in need and
to beggars, and for the redemption of captives; who
attends to his prayers and renders the alms levy."3
This verse does not read like a doctrine, an 'aqlda) but
is rather a nice mix of statements of faith and chari-
table practices. The scripture emphasizes "the differ-
ence between the formal act of turning toward God
and the acceptance of the existential consequences of
that act."
4
The doctrinal details mentioned in this
verse-God, the Last Judgment, the angels, the Book,
the prophets-were never placed in doubt, but they
were also never a focus of interest. From the begin-
ning, theology was concerned with differ-
ent problems: free will and predestination, the attri-
. butes of divinity, justification by faith. On the whole,
taken in relation to Christianity, Islam did not treat
new problems; it treated the same problems differ-
ently.5
Nevertheless, it would be wrong to say there was
no doctrinal progress in Islam. That is far from the
case. Over the long term, however, that progress
occurred in jurisprudence rather than in dogmatics.
IS
THE FLOWERING OF MUSLIM THEOLOGY
That is the most important thing Muslims learned
from their sacred text: how to conduct themselves in
an honest and upright manner. The Prophet himself
had the opportunity to lay the social and political
foundations of his community in Medina. To that
end, he was able to make use of what the Koran of-
fered him-namely, a law.
Here again, we find the minor difference. For Is-
lam, orthopraxy is more important than orthodoxy.
At the level of action, in the liturgy and in daily life,
details counted a great deal. Anyone who wanted
to emphasize the unity of the faith by designating
Muslims as such, irrespective of their denomination,
called them ahl al-salat, that is, "all who pray in the
Muslim manner," or ahl al-qibla) because they all
turned in the same direction (qibla) to pray.
6
The law,
like Christian dogma, presupposes a distinction be-
tween true and false. But a jurist often has to con-
front the fact that truth presents itself as a matter of
circumstance. There are times when the qibla cannot
be determined, yet the prayer remains valid. The ju-
rist always knows that the truth is not easy to find
in a concrete case. Judgment, as his colleagues the
fuqaha) said, can attain only probability, never cer-
tainty. In principle, this could be seen as an appeal
for caution.
16
THEOLOGY IN ITS OWN EYES
Nonetheless, scripture also asserted that disagree-
ment was harmful. The Jews and the Christians had
sown division in the world; that is why the new reve-
lation was necessary. In the beginning, humankind
had been a "single community (umma wahida)) and
God had sent his messengers to maintain its cohe-
sion. But no one followed his advice, and Islam came
into being to restore the primordial state of the
community. "Had your Lord pleased, He would have
made all humanity a single community. They are still
at odds, except for those to whom your Lord has
shown mercy. To this end He has created them" (sura
n:u8; [translation modified]). The early Muslims
pondered that verse a great deal. They were obviously
the ones "to whom your Lord has shown mercy'; that
is why he had spared them antagonism, division.
But reality quickly took a different course. With the
.first civil war, the "Great Discord" in Hishem Djai:t's
expression, the community split apart. As yet, this
event had nothing to do with the doctrine oflslam: it
was a political dispute. The first major conquests,
however, had convinced the Muslims that they would
always be victorious, thanks to the unity God willed
for them. Did the fitna now mean that, in spite of ev-
erything, Muslims were "created" to remain divided,
just as the Jews and Christians had been? Initially,
17
THE FLOWERING OF MUSLIM THEOLOGY
people thought they could arrest that pernicious ten-
dency. "Do something about this community before
it turns against itself over scripture, as happened to
the Jews and Christians," Hudhayfa b. al-Yaman is
supposed to have told 'Uthman, a caliph who would
be murdered precisely because he had failed to re-
spond to the rift. The statement is not authentic; it
was formulated after the fact, when Muslims realized
they were condemned to live in a state of discord. As
a result, they made a virtue of necessity and discov-
ered a kind of inter-Muslim pluralism. The Koran did
not shed much light on that question. But they were
able to turn to the Prophet, whose famous hadith
was beginning to circulate: "In my community, dis-
agreement is a sign of divine mercy"; or, to restore
its original sense: "In my community, disagreement is
an opportunity for divine mercy." To this we ought to
add, "in contrast to earlier religious communities." It
is no accident that this maxim repeats exactly the key
words found in the Koranic verse previously cited:
"disagreement," "community," "mercy." Hence,
though the Muslims were no doubt "created" to be
in disagreement, just as all other communities had
been, in Islam that same disagreement could be un-
derstood in the sense of"diversity." It was something
good.?
18
THEOLOGY IN ITS OWN EYES
Nevertheless, the hadith I quoted was accepted
only in the field of jurisprudence. Divine mercy had
manifested itself in the plurality of legal solutions
that the sahaba, or Companions of the Prophet, had
proposed. The fact that this first generation had
failed at the political level was ignored. Nevertheless,
even in the legal realm, people were not always in-
clined to accept the disagreements among the Com-
panions as something positive. The Mu'tazilite theo-
logian an-Nazzam did not hesitate to treat their legal
positions with caustic irony.
8
But the next stages of
Muslim law culminated in such a glorification of the
ancestors that they seemed to have become incapable
of serious error. This led to the birth of the sunnah
nabawiyya, the Prophetic tradition, which always left
.choices to the individual. The space that would later
be occupied by legal options or schools, madhahib,
was created only by virtue of the Companions' dis-
sent. Islam never had a uniform law, and a jurist
could, at least theoretically, always allow himself to
turn away from received opinion and adopt one of
the minority views held during the first generation.
Muslim law is a law of jurists; that is why it took a ca-
suistic form, and why there was never a universally
constraining code enforced by a sovereign.
* * *
I 9
THE FLOWERING OF MUSLIM THEOLOGY
Here again we are faced with one of those minor
differences. The Justinian Code and the Codex Juris
Canonici taught Christianity that, in jurisprudence as
in theology, there was a single fixed system and that
it could no longer be called into question. Conversely,
in Muslim law, even during the Ottoman Empire,
there existed only an intention, a plan, a professorial
exegesis of certain fundamental texts carefully cho-
sen according to the consensus of a school. Let us
recall in this context that most theologians were
trained as jurists; they may therefore have come to
the realization that even in the field of theology it
was better not to restrict choice excessively. Quite
early on, one of them, 'Ubaydallah al-'Anbari, judge
and governor of Basra, expressed that idea in the
form of a maxim. Kull mutjahid mus'ib) he said: "Who-
ever forms a reasoned opinion is right." He meant
not only that the judgment he pronounced as a qad'i
Uudge) was valid in all cases and allowed for no judi-
cial review but also that anyone who expressed his
own opinion in matters of faith was free to defend
it-as was any detractor who reached a different con-
clusion.9 It must be admitted, however, that in theol-
ogy this liberal position did not prevail. That is be-
cause theology, unlike jurisprudence, was concerned
with eternal truths and not with everyday matters.
20
THEOLOGY IN ITS OWN EYES
For that reason, theological disagreement would very
quickly take the form of scandal.
This is brought home to us once again in a saying
of the Prophet. It is certainly apocryphal, like the one
previously quoted, but it responds to the same ques-
tion: "My community will have experiences similar to
those of the Israelites ... The Israelites were divided
into seventy-two sects, but my community will be di-
vided into seventy-three sects, and all but one will go
to hell." No trace remains of the idea that disagree-
ment could be a sign of divine mercy. Rather, the
tone of the hadith is one of resignation, as if the
truth belonged only to a minority. The desire to pro-
tect the community from decadence is obvious. Mus-
lims were therefore glad to be able to quote the
Prophet, who claimed: "My community will never
reach agreement on an error." The truth will p1.1evail-
such was the message. For that reason, Muslims were
told, "Abide by the majority view!" The truth, it had
to be admitted, was only one among seventy-three
opinions, and yet the majority seemed able to distin-
guish it from the many errors. to
At this stage, no sign of the small difference men-
tioned earlier is in evidence. Muslims had learned
that heresy had many faces, yet it was considered ab-
errant and isolated. For some time, in fact, events
21
THE FLOWERING OF MUSLIM THEOLOGY
proceeded along what would seem to be the normal
course, with trials and executions. They were not
set in motion by jurists but by the rulers: the caliph
and his representatives. The sources tell us whom
these measures were directed against: Ghaylan al-
Dimashqi, Ja'd b. Dirham, Jahm b. Safwan, and later,
al-Hallaj. In one particular case, that of Bishr al-
Marisi, they even tell us how and where the trial took
place: in the courtyard of a mosque, with the parti-
cipation of a large, probably raging, crowd (even
though the qadi allowed the accused to escape).1
1
But
in many cases, the essentials are not mentioned. The
first trials seem to have been political; but who would
be surprised about that? What emerged was the
temptation associated with any conviction, religious
or not, that is held by the majority-namely, to use vi-
olence and to secure the support of the government.
The minority sects did not have that opportunity.
The Ibadites, for example, and later the Druze, could
not execute heretics; they excommunicated and repu-
diated them until the heretics publicly repented and
were again accepted into the community. For the pe-
riod I am discussing, recourse to violence was not yet
the most notable feature. The most important thing
for our purposes is that those who staged the trials
came to realize what Islam meant to them. There was
22
THEOLOGY IN ITS OWN EYES
as yet no orthodoxy; those who would later be called
Sunnis still had to determine the content of the
'-sunna, the established custom. Hence the heresy tri-
- als also marked stages in the self-definition process
of the faith.
As an example, let me cite an execution that took
place in Damascus near the middle of the second
century AH. The victim was one Muhammad b. Sa 'Id
al-Urdunni, a traditionalist (muhaddith) who had
questioned the irrevocability of Muhammad's
prophethood. As is well known, the Prophet consid-
ered himself a link in a chain of messengers, someone
who confirmed a revelation that God had "brought
down" several times since Creation. With his own
preaching Muhammad placed his seal (khdtam) on
that ancient revelation-that is, he bore witness to its
truth and authenticity. That does not automt.Ltically
mean he was the last to place his seal on it. The chain
could have continued, and in fact in the early genera-
tions, certain people claimed the right to prophesy
or were considered prophets. Prophecy had been a
unique event on the Arabian Peninsula, but its pass-
ing with the death of Muhammad also gave rise to a
sense of frustration. For a short time, even 'Umar
could not manage to believe it had ended. But proph-
ecy always led to a change-one accomplished by an
23
THE FLOWERING OF MUSLIM THEOLOGY
irrefutable authority-and Islam would have contra-
dicted itself had it given credence to the idea that
someone in the future would act toward Islam as
Muhammad had acted toward Judaism and Chris-
tianity. For that reason, the Koranic formulation that
called Muhammad the seal of the prophets was ac-
companied by the statement that there would be no
prophet after him (Ia nabryya ba'dahu). The Dama-
scene theologian, as it happens, now added in sha'a
'llah (if it be God's will). That is why he was crucified.
In a sense, his reaction was altogether logical: it is
surely up to God to decide whether he wants to send
another messenger. But that objection could not
eradicate the scandal.
12
The sources claim that the word used to designate
heresy in this case was zandaqa) a term that became
widespread shortly thereafter. But the term was at
first reserved for a very specific heresy, dualism. The
zanadiqa were initially Manichaeans or Muslims at-
tracted to that Iranian doctrine. In reality, the attrac-
tion Manichaeanism exerted over Muslims had little
to do with dualism. What Muslim intellectuals ap-
preciated above all was its scientific modernity. That
may seem astonishing today. The image of Mani-
chaeanism that emerges from the Greek texts, the
24
THEOLOGY IN ITS OWN EYES
works of Saint Augustine, and the original Mani-
chaean writings discovered in the meantime is totally
different, that of a bizarre mythology and a harsh
and rigid vegetarian asceticism. But in the Near East,
Manichaeanism came to be known for its cosmolog-
ical speculations. For the Manichaeans, the world
originated from the mixing of two eternal principles,
light and shadow; therein lay the Iranian solution to
the problem of creation. The category of mixture was
reminiscent of an element of Hellenistic philosophy:
the Stoic idea of krasis di' holou) t a d a k h u ~ as the
Mu'tazilite an-Nazzam would have said, an idea that
had permeated Iranian thought long before.l
3
This
was a sort of natural science whose hypothetical
character made it ideal for educated people inclined
to speculate without examining things closely. The
fact that this system was defended by duali'!>ts was
not important in itself. Muslim intellectuals, Mani-
chaeans, and thinkers with similar leanings (dualists
who embraced Bardesanes's system, alchemists, and
so on) all belonged to the same circle. But outside
that circle, the curiosity they displayed for such ideas
left a bad impression. The simple people, who under-
stood Islam differently, and the scholars in the tradi-
tionalist camp, the ashab al-hadtth) took a dim view
of such speculations. Under Caliph al-Mahdi, the
25
THE FLOWERING OF MUSLIM THEOLOGY
government responded, conducting investigations
and dismissing officials, since most of the "crypto-
Manichaeans" worked for the government.I4
The sources have somewhat more to say in this
case. Nevertheless, details of the events elude us. We
have no trial transcripts, and the accounts are rarely
reliable or contemporary with the events. In general,
it appears that heretics were not sentenced to death.
But for a certain period of time a new position was
created, that of inquisitor (sahib al-zanadiqa). The sit-
uation was thus taken seriously. From that point on,
it became clear that Manichaeanism was not com-
patible with Islam. Even during the persecution, the
term zandaqa, dualist heresy, had been used by intel-
lectuals to slander one another.
15
Shortly thereafter,
the same word became a legal term. Henceforth,
anyone who committed zandaqa merited the death
sentence. The only question remaining open was
whether the condemned ought to have an opportu-
nity to repent before being executed. In legal terms,
that procedure was more or less logical and consis-
tent, since it respected the profession of faith la ilaha
illa 'llah (there is no god but God) of the shahada,
which is taken from the Koran (sura 37=35). The sec-
ond half of that profession of faith is also in the Ko-
ran (sura 48:29). We do not know at what point the
26
THEOLOGY IN ITS OWN EYES
two expressions were brought together, but it was
fairly early, long before the trial in Damascus in any
case. Coins minted under 'Abd al-Malik ibn Marwan
are already inscribed with the shahada. Henceforth it
would become the eternal symbol oflslamic identity.
The foreign origin of zandaqa was obvious to any
Arab who heard the word pronounced. Its morphol-
ogy left no doubt that it was Persian. Zindiq is derived
from zand, the commentary on the Avesta. By the
Sassanid era, the Manichaeans had acquired the du-
bious reputation of developing a very personal exege-
sis of Mazdean doctrine.I
6
At first, that did not mat-
ter much to Muslims. The Umayyads had tolerated
the Manichaeans and had even used them to oppose
the old Sassanid bureaucracy. But when the Abbasids
moved the center of the Syrian empire to Iraq and
founded the new capital of Baghdad near tb.te ruins
of ancient Ctesiphon, things took a different turn,
and in the ensuing persecutions the defamatory term
the Sassanids had invented was put to use. Most
Arabs knew nothing of its etymology, and soon the
sources applied it to every sort of heretical deviation.
In the end, however, the word came to designate
something more than a simple error-namely, unbe-
lief. For that reason, the punishment had to be the
one reserved for apostasy.
27
THE FLOWERING OF MUSLIM THEOLOGY
It was at that point that theologians joined the
fray. It was primarily jurists who were responsible for
meting out punishment, whereas theologians were
supposed to be concerned with defining heresy and
refuting the arguments of its proponents. The au-
thorities wanted to know when exactly mere error be-
came unbelief. Quite often, heresiography was not
merely theoretical. By producing a list of deviations,
it offered categories that could be used to establish
measures for prevention and persecution. We can as-
sume that, when these events were occurring under
Caliph al-Mahdi, theologians played the role of at-
tack dogs for Islam, with the approval if not the
encouragement of the government. We know that
several of them, still young and ambitious, wrote
pamphlets against the supporters of zandaqaY Early
on, they also began to make vicious attacks on one
another. At first they played fair, engaging in debates
and disputations, but gradually they became more
aggressive. Those who built a protective wall around
Islam were obliged to indicate what would remain
outside it, and as a result monotheism and Muham-
mad's prophecies were no longer understood as the
only criteria for truth. Theologians believed in the
power of reason, since they could not conduct a de-
bate with by appealing to the Koran or to
28
THEOLOGY IN ITS OWN EYES
tradition. In addition, they were nearly certain they
would prevail in these interreligious controversies.
Since they were "in the right," the possibility of fail-
ure was practically ruled out. Unprovable or inexpli-
cable facts, "mysteries" as a Christian would have
called them, do not seem to have existed. That is
why they had no scruples about arguing, even within
their own camp, that their way of understanding Is-
lam was the only valid one. They were rationalists
and as such were always sure of themselves. Hence,
when the opinions expressed by one of their col-
leagues became too personal, theologians tended to
call them "unbelief" as well. In a sense, these were
only words. They simply meant: Whoever believes in
such things will certainly be damned. Unlike the
cases previously mentioned, these verbal attacks did
not generally put in jeopardy the physical wr,l.l-being
of those being criticized. But the polemics contained
the seed of radicalization. Indeed, the theologians,
both Mu'tazilites and others, sometimes tried to
hand the dissidents over to the police. That is what
happened to the Mu'tazilite al-Mu'ammarls and later
to his colleague Ibn al-Rawandi.J9
Here, then, was the notorious takfir (anathema). But
what was the source of the aversion toward that atti-
29
THE FLOWERING OF MUSLIM THEOLOGY
tude? One answer is that it is already found in the
Koran, in the words Ia ikraha fi 'l-d'in (no compulsion
in religion) of sura 2:256. But that argument was
never made in classical Islam, and the verse cited did
not play a notable role. The decisive factor in the re-
straint of medieval Muslims was instead a historical
event: Kharijism. The Kharijites were the first to prac-
tice the anathema and, for that reason, bad memo-
ries were always associated with them. They consid-
ered themselves the only true Muslims. They did so
in the first place for reasons more political than reli-
gious, but they were intent on reproducing the
Prophet's exemplary migration (hijrah), and they
withdrew to regions where they could live alone, in
the desert or on the periphery of the Muslim world.
That schism resulted from the claim to exclusive
sanctity. Hence the Kharijites abhorred intermarriage
with non-Kharijite Muslims. They also battled their
coreligionists everywhere they could. They believed
they were dealing not with Muslims oflesser quality
but quite simply with unbelievers, who, moreover,
had knowingly rejected the true faith. Because of that
postulate, they were able to apply to their coreligion-
ists everything that had been said in the Koran about
the pagans. As a result, not only were they convinced
that all other Muslims would go to hell, but they
30
THEOLOGY IN ITS OWN EYES
even felt justified in conducting a jihad against them.
The majority of believers considered them not only
extremist schismatics, but also thugs (bughat) and
terrorists. That image of them persisted even after
they had become peaceful communities or had com-
pletely disappeared from entire regions. From that
time on, it was the norm to identify the practice of
takfir with exclusivism and extremism.
The court theologians I mentioned, Mu'tazilites
for the most part, took that fact into consideration.
They carefully refrained from cutting themselves off
from the community; they wanted only to define the
creed more precisely. The doctrines they hated most
were anthropomorphism and predestination. These,
they thought, crossed the line into unbelief. But they
inevitably found themselves in disreputable company
because of the language they used. Their adv;rsaries
saw their suspicions confirmed when the Mu'tazilites
became involved in a campaign of religious perse-
cution, the mihna, which somewhat served the
Mu'tazilites' goals. This inquisition had been un-
leashed by the caliph. I will not comment further on
it.
20
What matters in this context is simply that the
theologians were on the side of the persecutors. They
added a certain arrogance to the state's action. They
were shameless in pointing out to the victims that
31
THE FLOWERING OF MUSLIM THEOLOGY
they owed their situation to their stupidity, which
had led them astray. No one was executed, and as
for the circumstances that led Ahmad b. Hanbal, the
most distinguished victim-at least in the light of
later tradition-to be flogged during a disputation,
they are still much debated.
2
1 The caliph did, how-
ever, try to paralyze the entire "system" of opponents,
by barring their jurists from practicing their profes-
sion. Not only were the judges dismissed, but the
muftis, who were consulted by devout commoners in
religious matters, were forbidden to practice. That
decision was considered scandalous because those
who worked in the legal system, including profes-
sional witnesses-indispensable for the verification of
all official documents-were notables, probi homines.
The common people took the victims' side; the per-
secution was perceived as an ordeal, a tribulation
(mihna). And the court theologians completely lost
touch with the masses.
Matters took a bad turn. Fifteen years later, the
government measures had to be revoked. That event
would be a key experience for the Sunni world, and
it is still commemorated in many popular accounts.
The people had won a victory over the authorities, or
at least that was the impression. The devout victims
were considered martyrs for a just cause. The caliph
32
THEOLOGY IN ITS OWN EYES
had lost the battle, and henceforth any religious
p o l i c ~ conducted by the government was unlikely to
gain the necessary support. From that time on, spe-
cialists in the Prophetic tradition, ashdb al-had'ith,
came to form a separate bloc. Their role was to inter-
pret divine law and to decide whether to cooperate
with the government. Theology was discredited; its
representatives had abused reason by exerting pres-
sure on those who held views contrary to their own.
The hateful aspect of anathema (takfir), the anxiety it
produced, is mentioned in the historical accounts. It
was associated with a negative view of all theological
and dialectical argument. Once more, what matters is
the perception of history, the Geschichtsbild, the image
engraved in collective memory. It took time for that
evolution to occur, and rational theology never com-
pletely died out. But for many, dialectical thought
now had a bad reputation. It could be understood as
eristic, and anathematizing was assimilated to arro-
gance and presumptuousness. As a result, theology
was not deemed the loftiest of the sciences, as it was
in the Christian Middle Ages-philosophy even less
so. It was jurisprudence that became the premier dis-
cipline. Here again we are faced with the small but
significant difference. In Islam, religious expertise re-
mained in the hands of those who were later called
33
THE FLOWERING OF MUSLIM THEOLOGY
the ulema, a group of middle-class scholars who ex-
plained the faith on the basis of an exegesis of the
Koran to others of the same class and who solved ev-
eryday problems with their legal advice. As it hap-
pens, the criterion they used to reestablish order was
consensus, a term I use in its sociological sense and
not in the limited sense it has assumed in Islamic re-
ligious law.
That social consensus was never institutionalized. It
was not officially sanctioned by synods or councils,
for there was no one to invoke such things. The "sys-
tem"-ifwe may call it that-was extremely informal.
Every properly trained jurist had the right to pro-
nounce legal opinions based on certain presuppo-
sitions, and the minority view was a sign of divine
mercy, as explained earlier. Naturally, the margin of
freedom in the decision was limited by the growth of
consensus, but given that a legal consensus, i j m a ~
was established only after the fact, it did not exert its
full power to clinch matters until after society itself
had given its approval. The question of which move-
ment constituted a sect and which did not was de-
cided not by dogma but by the views of the local ma-
jority or the minority's lack of success. Persecution
was a matter not of definitions and laws, but of cir-
34
THEOLOGY IN ITS OWN EYES
cumstances and temperament. Nevertheless, the situ-
ation was far from ideal. The Hanbalites, former vic-
tims of the inquisition and their successors, gave
their adversaries a taste of their own medicine. Their
self-defense organizations ('aiyyarun, ahdath) carried
their reign of terror into enemy neighborhoods. With
all the more reason, deviationists in their own camp
were forced to retract their opinions.
But social tolerance was preserved nonetheless.
The jurists were interested primarily in what some-
one had done, not in what someone had thought.
Until evidence to the contrary surfaced, they felt
obliged to hold a good opinion (husn al-zann) of every
coreligionist, because any sin committed remained a
matter between the sinner and God, so long as it did
not harm anyone. Naturally, the danger always ex-
isted that a theologian or jurist would stiw up the
masses against someone who had a bad reputation-
a reprobate, an adversary-or against a dissident
group. But when the government wanted to make an
example of someone, it had first to establish an i j m a ~
a consensus of jurists, and normally that did not
happen unless several fatwas had been pronounced
by different legal schools (as in the case of al-Hallaj).
In general, the secular authorities were interested less
in suppressing heresy than in avoiding the popular
35
THE FLOWERING OF MUSLIM THEOLOGY
uprisings that resulted from them, and it was very
rare for the ulema, who were responsible for the spiri-
tual realm, to ask for help from the secular arm, as
the Christian Church did.
It is no simple matter to describe the mechanisms
at work here. The important thing to remember is
that almost nothing was officially regulated. Some-
one whose "orthodoxy" was considered suspect was
treated with a certain coolness. Fathers were unlikely
to allow a daughter to marry him; neighbors would
often not visit him when he was sick; few would at-
tend his funeral. Above all, people were hesitant to
greet such a person, for the greeting always took the
form of a blessing, and the devout tended to save
their blessings for those who shared their faith. As a
matter of fact, that was also the case with the Jewish
"shalom." So long as someone was not considered a
member of one's own faith, one simply returned the
greeting, and not always in a friendly way. A scholar
who was not orthodox could be boycotted; when he
no longer had any students, his books would no
longer be copied. When an unpopular preacher was
dismissed or driven out, his pulpit was sluiced out to
cleanse it. In addition, one could show respect by
praying behind a particular person, for there were
enough mosques to allow each Muslim to choose his
36
THEOLOGY IN ITS OWN EYES
own imam. It was only on Friday that everyone went
to the main mosque, where the caliph or his repre-
sentative led the prayer. Only at that time were some
people liable to act against their personal convic-
tions. During funerals, the social rank of the person
who uttered the prayer over the body revealed what
was thought of the orthodoxy of the deceased. Even
in the major cities, personal reputation did not lose
its importance; people lived in segmentary societies
where modern anonymity was still unknown. As for
scholars, the value judgments applied to them have
been preserved in a rich body of biographical litera-
ture, and the criteria were always religious.
22
These value judgments were conservative. It could
hardly be otherwise in a society founded on consen-
sus. Communal harmony was more important than
originality. In doctrine and in personal ,Pehavior,
Muslims leaned toward a generally accepted middle
ground. Against that backdrop, both the major
Islamic denominations chose terms to designate
those who deviated from the norm. They were "exag-
gerators" (ghuldt), Shiites said, when they no longer
wanted to be associated with the idiosyncrasies of
their own past.
23
They introduced innovations
( ahddth, bida '), said the Sunnis, contrasting the here-
tics' attitude to their own traditionalism. In a sense,
37
THE FLOWERING OF MUSLIM THEOLOGY
the term "innovation" captures the mindset of any
religion that, on account of revelation, has its apogee
at the beginning. The Byzantines spoke of neoterismos
and Tertullian alluded to the novellitas of the sectari-
ans. The Latin text that provided Christian theology
with its popular definition of orthodoxy, Vincent of
Lerins's Commonitorium, bears the subtitle "Defense
of the Antiquity and Universality of the Catholic
Faith against the Impious Innovations of All Heretics."
The notion contained a good dose of criticism of the
present and nostalgia for the good old days, an atti-
tude one finds everywhere and in every era. That said,
the Sunnis were to a certain degree predisposed to
such an attitude. Over against the notion of bid'a was
always its opposite, sunna, or ancient custom, which
was legitimated by the Prophet's authority. That is
why the term "innovation" ultimately designated not
only doctrinal aberration but also, and to an even
greater degree, reprehensible practices: additions to
the ritual prayer, immoderation in reciting the Ko-
ran, popular customs practiced during Ramadan or
religious holidays, excesses in expressing grief, or
women's participation at funeral ceremonies.
24
In typically Muslim fashion, people paid more at-
tention to orthopraxy than to orthodoxy. That is also
why what was new one day could be old the next.
38
T
THEOLOGY IN ITS OWN EYES
Dogmas can become ossified and resist any change in
language or point of view; certain phrases from the
Christian creed are a good example. The innovations
Islam spoke of were something else again: at one
time, coffee and tobacco were innovations and were
attacked with many theological and legal arguments.
A little later, however, everyone accepted their use.
Hence, people began early on to differentiate between
innovations that were good and those which were
bad or, to express it in the jurists' refined vocabu-
lary, between innovations that were deplorable, those
which were permitted, and those which were recom-
mended. It would have been impossible to use the
terms "good" or "bad" with respect to heresies. As
a result, even the most unacceptable innovation
stemmed not from rebellion against God but from
mental confusion. It could and ought to nfeet with
disapproval, but no legal measures should be taken.
In this context, it may be advisable to cite a great
theologian, al-Ghazali (AH 450-505; 1058-nn CE). In
his Tahdfut al-folasifa, al-Ghazali had indicated the
point at which the philosophers' doctrines degener-
ated into unbelief; he thus knew whereof he spoke.
But a few years later, after a spiritual crisis that made
him doubt the omnipotence of reason and of intel-
39
THE FLOWERING OF MUSLIM THEOLOGY
lectual squabbles, he wrote a treatise on the distinc-
tion between Islam and zandaqa in which he warned
his readers against hasty use of the term kufr (her-
esy).2S Unbelief, he said, is a purely legal category. A
Christian would never have said that. Al-Ghazali
could defend that formulation because he knew that
unbelief entails certain consequences. For example,
no one can inherit from an unbeliever, and an unbe-
liever can be killed if not under the protection of a
Muslim authority. But, added al-Ghazali, that ap-
plies only to Muslim apostates, and one commits
apostasy only when one denies the essential dogmas:
monotheism, Muhammad's prophecy, and the Last
Judgment. That short list once again conforms to the
shahada. The only thing that has been added is the
Last Judgment, which naturally implies acceptance of
the divine commandments. But only the person who
entirely rejects the Law, that is, the antinomian, is
an unbeliever. Al-Ghazali wrote a treatise against the
antinomian current, or ibahiyya.
In that book, Faysal al-tafriqa, al-Ghazali also took
the opportunity to argue in favor of reconciliation
and restraint. He knew that Sunnis and Shiites had a
hard time coexisting, but he recommended that any-
one accused of heresy not return the insult, for God
is generous toward human beings and one must al-
40
THEOLOGY IN ITS OWN EYES
ways follow his example. Of course, everything de-
pends on the way one interprets scripture; false exe-
gesis, t a ' w z ~ is at the root of every deviation and
aberration. But one must always abide by unanimous
agreement, and if someone does not bow to it, one
must remember that consensus does not include
everyone. It is not fixed, and it is not the same every-
where and always. Granted, only one of the seventy-
three Muslim sects will achieve salvation, according
to the famous Prophetic tradition quoted earlier. But
that hadith, al-Ghazali added, is not reliable; a vari-
ant of it is to be preferred. According to that variant,
only one of the seventy-three groups that constitute
Islam will perish. Al-Ghazali thus went a long way in
his defense of tolerance and peace. He did not hesi-
tate to change one of the Prophet's sayings that
seemed to justify those who thought they hjj-d a mo-
nopoly on salvation, and he did so in such a way as to
express the hope that almost all Muslims would ulti-
mately be saved. In al-Ghazali's time, the majority of
Muslims, those we now call Sunnis, had stopped be-
lieving that any of the groups of the ahl al-qibla would
be eternally damned; rather, all would ultimately be
saved by the shahada. Even heretics and innovators
- had only a kind of purgatory to fear. In addition,
even if a true unbeliever sinned because he was igno-
41
THE FLOWERING OF MUSLIM THEOLOGY
rant of Islam, he would be forgiven by God, said al-
GhazalL
We may wonder to what extent al-Ghazali, despite
his renown, represents Islam as such. It is beyond
question that what he said represented first and fore-
most a postulate-theory and not reality. And yet in
the tradition in which al-Ghazali was trained, al-
Ash'ari is reported to have asserted at the time of
his death: "I attest that I anathematize none of my
brothers (the ahl al-qibla), for they all call on the same
God; the differences are only verbal." It is possible
that these words are apocryphal; others were also at-
tributed to the dying al-Ash'arL Nevertheless, the fact
that they are cited is in itself revealing, for al-Ash'ari,
like al-Ghazali, represents the middle ground and
avoids the extremes of the two major persecutions:
the anathemas of the Mu'tazilite inquisitors and
those of their Hanbalite victims. That attitude be-
came the guiding principle for later Sunni theology.
26
Ricoldo da Montecroce, a Dominican monk who
traveled to the Orient near the end of the thirteenth
century, was outraged that Muslims did not follow
the maxim that narrow is the way to salvation, be-
cause, he said, "even when they do no more than re-
cite that 'there is no god but God and Muhammad is
his messenger,' for them that means they will surely
42
r-
THEOLOGY IN ITS OWN EYES
achieve salvation. All the Saracens agree with the
view that a Saracen need only say that and he shall be
saved, even if he has committed all the sins in the
world. In spite of the fact that their law, the Koran,
has established a goodly number of prohibitions, the
sinner will suffer no punishment .... We can there-
fore call their law permissive. Satan in his malice saw
to it that those who do not achieve bliss by taking the
straight and narrow way will take the wide road to
hell."
27
There is reason to wonder whether, in the end, that
wide road was not the better one. Strictly speaking,
Islam had no religious wars like those in Europe. But
Islam can be as militant as Christianity; the danger of
making religion into an ideology exists everywhere.
Those who assault the dignity of the Prophet, even
without denying him directly-the Baha'Is.and the
Ahmadiyya because they postulate another divine
messenger after him, Salman Rushdie because he de-
fames the person of Muhammad through literary
parody-are sure to incur the wrath of the majority.
It seems that "heresy" is now defined in relation to
Muhammad's prophethood and sanctity, rather than
by the tawh!d-that is, the concept of God's oneness.
More emphasis is placed on the second half of the
shahdda than on the first. We may wonder about the
43
THE FLOWERING OF MUSLIM THEOLOGY
reasons for that change in perspective. For al-Ghazali
in his Tahafut al-falasifo) that was not yet the case. The
fear of heresy has certainly not disappeared; we are
witness to new orthodoxies created by the media,
Saudi money, and politico-religious movements. But
that phenomenon will have to be dealt with else-
where. With regard to classical Islam, we can say that
in spite of everything the jurists, however suspect
they may have become for us, treated the question
with more flexibility than we would expect from the-
ologians or, to speak of our own time, from politi-
cians. The "innovations" they felt obliged to stigma-
tize did not condemn anyone to be burned at the
stake. But like "deviations" and "counterrevolutions"
in our own day, the "unbelief" that was the object of
theologians' polemics was a much mare unpleasant
companion.
44
2
THEOLOGY AND THE KORAN
The Mi'raj and the Debate on Anthropomorphism
THE ASCENSION TO HEAVEN was an experience
shared by Jesus and Muhammad, that is, by the reli-
gious imagination of Christianity and of Islam. But
the two instances have totally different contexts. For
Christians, Jesus' ascension was the logical conse-
quence of the resurrection. The Koran, however, does
not speak of resurrection. According to it, Jesus rose
to heaven directly from the cross, without 4ying on
it. And when Muhammad ascended to heaven, he did
not remain there, but returned to earth. His ascen-
sion marked the beginning of his career; it was then
that he learned what task he was to perform, and the
success of that task would be obvious to everyone in
his community. By contrast, at the time of his ascen-
sion,Jesus had completed his earthly mission and, in
fact, had apparently failed. In returning to heaven, he
showed that, despite his crucifixion, he belonged to
45
THE FLOWERING OF MUSLIM THEOLOGY
the kingdom of God. Let us not forget that it is only
in Western languages that the term "ascension" is
applied to both events. In Arabic, there are two: for
the Prophet, the word used is mi'rdj) whereas Jesus is
"lifted" to heaven, rafa'aha )lldh ilayhi) the Koran says
(sura 4:158; compare T55). Muslims never compared
Muhammad to Jesus in that regard. That fact is all
the more glaring, given that comparisons in general
were not lacking, at least in the first phase of Islam,
as attested in hadith. But these comparisons, which
were based on purely Koranic ideas, referred to other
prophets: Abraham, it was said at the time, was
the friend of God (kha!Jl Allah)) Moses the one to
whom God had spoken on Mount Sinai (kalim Allah))
and Muhammad, finally, the one who ascended
to heaven, where he saw God in person. Th'is was
more than a comparison, it was a tripartite schema
of increasing intimacy: being close to God (Abra-
ham), hearing his voice (Moses), seeing his coun-
tenance (Muhammad). That schema presented the
situation in a new form, which also raised specific
problems.
I will not deal with accounts of the mi'rdj as such.
They are well known, and for those who do not re-
member the texts, the Persian miniatures that evoke
them stand as a reminder.
1
Let me emphasize o.ne
46
THEOLOGY AND THE KORAN
point, however. As I have said, Muhammad did not
remain in heaven. He was not reunited with God af-
ter being sent to earth for a time. In reality, what hap-
pened was that he had an audience with God. And
that audience had a particular aim: God prescribed
the number of prayers his community was supposed
to say every day. At first, the Almighty was quite de-
manding: he spoke of fifty prayers. Muhammad had
to bargain, and he managed to reduce the number to
five. This reminds us of the scene in the Old Testa-
ment in which Abraham also bargains on behalf of
the righteous who live in Sodom and Gomorrah and
who are spared by God when he lets "loose . . . a
shower of claystones," as the Koran says (sura rr:82).
It was Moses who advised the Prophet to bargain;
Muhammad had encountered him during his jour-
ney through the heavenly spheres. Moses aJOtd Abra-
ham were the last two prophets Muhammad saw,
those closest to God in their cosmic role-and those
with whom he was compared in the tripartite schema
I mentioned.
We can imagine the fascination these tales exerted on
the public at the time. They had one major defect,
however. At first glance, nothing of what they related
could be found in scripture proper. In the Koran, the
47
THE FLOWERING OF MUSLIM THEOLOGY
word mi'raj appears only once, in sura 70, where God
is called dhu 'l-ma'arij, Lord of the Ladders. The text
continues: "The angels and the Spirit will ascend to
Him in one day: a day whose space is fifty thousand
years." But that is something other than the mi'raj;
the Prophet is not mentioned. As a result, exegetes
had to search for other evidence, and in fact they
never relied on the passage cited, but rather on two
others, also found in the Koran. First, there is the fa-
mous verse in sura q: "Glory be to Him who made
His servant go by night from the Sacred Temple to
the farther Temple [masjid al-aqsa] whose surround-
ings We have blessed, that We might show him of our
signs." It is an isolated verse, and what follows of-
fers no further clarifications. But the word "signs"
( ayatina) could be interpreted as an allusion to ascen-
sion or to vision. And given the qualifier, the masjid
a l - a q s ~ the "final place of veneration," found in a
place "whose surroundings we have blessed" could be
located in the Holy Land, in Jerusalem. Under the
Umayyad dynasty, a mosque built in that city came to
be called al-Masjid al-aqsa. Muhammad had traveled
by night to Jerusalem-that was the conclusion exe-
getes drew. It was not yet an ascension in the strict
sense, for Muhammad's journey was horizontal, not
vertical. True, he went to the place where Jesus had
48
THEOLOGY AND THE KORAN
ascended to heaven, but Muslims had no desire to
compare him to Jesus, as we have seen.
Let us now consider the second passage in the Ko-
ran. It is longer, but just as mysterious. It acquired its
importance by virtue of the fact that it seemed to
contain an allusion to a meeting with God. In the be-
ginning of surat al-Najm (53), the Koran speaks of
two visions by the Prophet. These are unusual texts
'
since normally Muhammad does not see, but hears.
The Koran resulted from instances of hearing. The
passage is well known:
1. By the declining star,
2. your compatriot is not in error, nor is he
mad!
3 His is no language of passion.
4 This is only revelation revealed to him.
5 He is taught by one who is mighty
6. and wise. He stood
7- on the supreme horizon;
8. then, drawing near, he hung suspended
9 within two bows' length or even closer,
Ill
IO. and revealed to his servant that which he re-
vealed to him.
II. His heart did not lie about what he had
seen.
49
THE FLOWERING OF MUSLIM THEOLOGY
Immediately thereafter comes the account of the sec-
ond vision:
I3. He had seen him come down another time
I4. at the sidra tree, beyond which no one may
pass.
I5. Near it is the Garden of Repose.
I6. When that tree was covered with what cov-
ered it [him),
I7. his eyes did not wander, nor did they turn
aside:
IS. for he saw one of his Lord's greatest signs.2
Once again, then, there is a "sign," as in the case of
the masjid al-aqsa. That may have led exegetes to com-
bine the two passages. The location of the event is
even vaguer than before. But the visions as such are
fairly concrete, described in a way that cannot fail to
excite curiosity. One striking fact: the account is in
the third person. It may be surmised that God spoke
in his own voice; he seemed to reveal a secret that no
one but the Prophet could have known.
It was not known exactly whom the Prophet saw,
however. Those hearing or reading the passage had to
decide for themselves. Modern exegetes usually insist
on the idea that the Prophet saw Gabriel at that time.
But it is practically certain that those responsible for
50
I
. ~
I
I
!
THEOLOGY AND THE KORAN
incorporating these verses into the traditions of the
mi'rdj were convinced that Muhammad had seen God
rather than Gabriel. The sight of God was the apogee
of the tripartite schema; the audience granted to the
Prophet was not limited to a mere perception of the
divine voice, as Moses had experienced it on Sinai.
The only difficulty was that the concept of God im-
plied by that idea was liable to cause a scandal, since
the vision seemed to imply tashbih1 anthropomor-
phism, and the problem of anthropomorphism had
concerned Islam ever since Muslims had begun to
think about it in theological terms. In Christianity,
the question was neutralized by the idea of incarna-
tion. It would be an exaggeration to say that this
change in perspective made it easier to understand. It
is no accident that Tertullian, the African Church Fa-
ther, said in his treatise De carne Christi (On ~ e flesh
of Jesus Christ): Certum est quia impossibile1 "[the incar-
nation] is a certainty because it is impossible." A Mus-
lim would never have put it that way. In any case, it
was not easy to find a rational explanation for God's
visibility.
The internal contradictions are pointed out in the
Koran itself. A later passage found in surat al-Takwir
contains an allusion to one of the two visions in
surat al-Najm: "No, your compatriot is not possessed.
51
THE FLOWERING OF MUSLIM THEOLOGY
He saw him on the clear horizon" (81:23). There is the
same lack of precision: "He saw him." But in a previ-
ous verse the person seen is identified as a "venerable
and mighty messenger, held in honor by the Lord of
the Throne" [translation modified]. It absolutely can-
not refer to God; a "venerable messenger" is normally
an angel.
Is our problem solved by that parallel? Perhaps it is,
for a reader these days, but it was not for the exegetes
we are discussing. And for us the question arises
on the historical level. Jewish thought already had a
tendency to eliminate the scandal of anthropomor-
phism by taking statements about God to be state-
ments about an angel. The angel Metatron plays that
role in certain Talmudic passages. The Kabbalah
takes the same tack. The angel could assume the
functions of the Creator; it was like a kind of demi-
urge, "one who is obeyed," m u t d ~ as the Muslims
said.
3
In fact, the term muta< appears in surat al-
Takwir (81:21). The fundamental conditions had
changed. The verses in surat al-Takwir were certainly
later than those of surat al-Najm, for now the vi-
sion is no longer described in detail, but simply men-
tioned as something already known. The public ad-
dressed in surat al-Takwir may have lived in Medina,
where the Jewish community listened to the Proph-
52
THEOLOGY AND THE KORAN
et's revelations with a critical ear. In any case, the
later passage does not rule out the possibility that
previously, in surat al-Najm, it was God that Muham-
mad thought he had seen. At the end of the first
vision, in fact, the text says, of the person who ap-
proached the Prophet, that he "revealed to his ser-
vant that which he revealed to him." "His servant"
(abdhu) can only be understood here as "the servant
of God," that is, Muhammad. If that is true, the one
who reveals is no longer a "venerable messenger" but
God himself, and God would therefore also be the
object of the vision.
Nevertheless, we must take one other thing into
account: According to the Koran, the Prophet did
not see God while the Prophet was in heaven but
when he was somewhere on earth. He saw him "on
the supreme horizon," and he had seen him, itt is said
at the beginning of the second vision, "come down
another time" (nazlatan ukhrd). Hence, it is not the
Prophet who rises to God but God who descends to
Muhammad. Moreover, it is immediately added that
he saw "the sidra tree, beyond which no one may
pass" (sidrat al-muntahd) and that "near it is the Gar-
den of Repose." That sounds like code for paradise,
with "Garden of Repose" (janndt al-ma'wa) designat-
ing the place where the blessed will rest during or af-
53
THE FLOWERING OF MUSLIM THEOLOGY
ter the Judgment (see sura 32:19). As for the "sidra
tree," it marks the place beyond which no one may
pass, the outer boundary of the holy of holies, where
God himself dwells. Nevertheless, God could "come
down" to that point, since people still thought at the
time that paradise was located on earth. It is there-
fore not necessary for us to embrace the idea earlier
defended by a number of Orientalists (from Grimme
and Caetani to Richard Bell and Regis Blachere) who
saw the Garden of Repose simply as a plantation near
Mecca, perhaps a villa, a kind of Monrepos for well-
off city folk, and the "sidra tree, beyond which no one
may pass" as a tree of some sort found on the border-
line of the Mecca sanctuary. In any case, Muslim exe-
gesis never questioned that the encounter had taken
place in paradise, though that place was on earth.
The "sidra tree, beyond which no one may pass" thus
became a kind of emblem of the Prophet's ascension.
Even accounts of the ascension that take nothing
else from surat al-Najm use that mythical tree to
mark the ultimate threshold, seventh heaven. It is
there that is located the source of the four rivers of
paradise. The exegetical situation was thus becoming
rather complex. Two competing alternatives surfaced:
God or angel, heaven or earth. That created four pos-
sibilities. In addition, we must not forget we are deal-
54
THEOLOGY AND THE KORAN
ing with two visions, not just one. For a single ascen-
sion, one of them would have sufficed; but if the
Prophet had to bargain with God concerning the
number of prayers, two are hardly enough.
Any exegesis whatsoever ratifies a theological deci-
sion. Nevertheless, in the present case, as in all the
others, it is not the product of a free-floating imagi-
nation. Its general orientation comes from models
worked out within a local tradition. We know that
J the motif of the heavenly journey was widespread in
the ancient world; an entire body of secondary litera-
ture is devoted to it. But this is not the place to elabo-
rate on the question of "influences." What matters is
the exegete's existential decision, his Vorentscheidung,
or prejudgment. Those who were put off by anthro-
pomorphism would soon come to believe <that the
Prophet had seen Gabriel, and in this world. Muham-
mad, it was thus supposed, learned of the task he was
to perform and received his first revelation from the
angel. That is also the reason, as the exegete in ques-
tion might have gone on to say, that this event is wor-
thy of mention in the Koran. Indeed, at that mo-
ment, Gabriel appeared to the Prophet for the first
time, and he also revealed his true angelic nature; he
was not disguised in human form. The only question
55
THE FLOWERING OF MUSLIM THEOLOGY
still needing to be resolved, then, is why Muhammad
saw Gabriel twice. Moreover, at the philological level,
one had to address the difficulty of the personal pro-
noun in the sentence: "He revealed to his servant
what he revealed." In spite of this, that view of things
was accepted by the majority. It is already found in
Ibn Ishaq's biography of the Prophet, composed in
Medina in the first half of the second century AH.
Those who adopted that interpretation could not,
strictly speaking, reconcile the account of surat al-
Najm with the mi'rdj. Many were not inclined to ac-
cept this circumstance, for no other support for the
mi'rdj appears in the Koran, and hadith on its own
had not yet acquired sufficient authority. But those
who insisted on taking the two passages from surat
al-Najm as proof of a vision of God could not help
wondering how and in what form the Prophet had
seen God. The responses given constitute the first
evidence we have of theological reflection. They are
found in the Prophetic tradition, hadith. The texts
reported there are frequently nothing but a sort of
disguised exegesis. A chain of transmitters an isndd
' ,
precedes each tradition and guarantees its reliability.
The example I will cite is of particular interest be-
cause the people constituting this isndd are incorpo-
rated into a sort of frame story:
56
THEOLOGY AND THE KORAN
Yahya reports: I asked Abu Salama: "What part
of the Koran was revealed first?" He replied:
"You that are wrapped up in your cloak!" (sura
74). I said: "[But] I was told it was "Recite in the
name of your Lord" (sura 96)." Abu Salama re-
plied: "I myself asked the question of]abir b.
'AbdAllah, and he said: "You that are wrapped
up in your cloak," and I, like you, replied that I
had been told it was "Recite in the name of your
Lord." Uabir] then replied: "I can only tell you
what the Messenger of God himself said,
namely: 'I had withdrawn to Mount Hira' to
meditate Uawartu fi Hira']. After completing my
retreat, I descended to the valley floor. Then I
heard a voice calling me. I looked around me,
ahead of me, behind, to the left and to the right;
but [only when I looked upward] ... There he
Ill
was, seated on his throne, between heaven and earth. I
immediately returned to Khadija and told [her]:
Wrap me up and pour cold water over me!'
Then [the revelation] came down to me: 'You
that are wrapped up in your cloak! Arise and
give warning! Magnify your Lord."'
The story ends with the first verses of sura 74,
God's call, which prepared Muhammad for his pro-
57
THE FLOWERING OF MUSLIM THEOLOGY
phetic mission. That call is preceded by the vision.
But the crucial problem lies elsewhere. Scholars were
divided on the identity of the first revelation that
"came down" to Muhammad. Was it sura 74 or sura
96? Each camp defended its position by invoking
Companions as its authorities; that is why the opin-
ions of each could only take the form of a hadith.
The chain of transmitters shows us that the dispute
occurred in Medina. There, in the city where the
Prophet died, it was thought that the information on
his life and experiences was more accurate than else-
where.
The Prophet thus saw God in all his might and
glory seated on his throne. Historians of religion are
hardly surprised by that, for we are familiar with that
scenario from the Bible and the ancient Orient.
When someone wanted to depict God in his majesty,
he was shown seated on his throne. But as an inter-
pretation of the vision of surat al-Najm, the image is
somewhat surprising, for although the Prophet sees
God "on the supreme horizon" ("between heaven and
earth," as the hadith says in its exegetical reformula-
tion), according to the Koran he is apparently not
seated. Rather, istawd wa-huwa bi'l-ufuq al-a'ld, "he
stood on the supreme horizon." At this point, we
must turn to philology. Translations are always inter-
58
T ...... /'
:_,
THEOLOGY AND THE KORAN
pretations and, where the Koran is concerned, they
still frequently rely on medieval exegesis. The verb
istawd, "to stand," is ambiguous. It can also signifY a
seated posture; when the verb is used in the Koran
with reference to God, it normally appears in the ex-
pression istawd 'ala 'l- 'arsh, "God sits upright on his
throne." In surat al-Najm, it could be understood the
same way. Conversely, the throne was no longer ap-
propriate when the vision was attributed to Gabriel;
an angel does not sit on a throne, he stands. He re-
ceives orders, he is ready to be sent out as a messen-
ger, but he is not a lord. When he is standing, Ibn
Ishaq tells us, "his feet are next to each other" on the
supreme horizon, sort of like an Egyptian statue's,
and then he draws near. Yet according to the hadith
we are considering, it is rather the throne that draws
near. The throne comes down, hovering (tadctJla), like
a pail ( dalw) into a well, until it is two bows' lengths
or even closer (qdba qawsayni aw adnd). In that inti-
mate situation, the Prophet, his ear close to God's
mouth, receives his first revelation.
More precisely, according to the text, Muhammad
did not receive the revelation until he was home, cov-
ered by a cloak. But the content of the message re-
ceived at home corresponds exactly to what he might
have previously heard from God's mouth: "Arise and
59
THE FLOWERING OF MUSLIM THEOLOGY
give warning! Magnify your Lord." He is not yet
obliged to transmit a particular message; he is quite
simply initiated. He must henceforth magnify the
Lord, and he learns of that mission after returning
home and wrapping himself in his cloak, in order to
concentrate. At the time the vision occurred, he was
not yet prepared to receive the message; he heard a
voice, but he was so dazzled by the apparition that he
did not realize the meaning of what he heard.
Let us take this opportunity to pause and recapitu-
late. Everything that has been said up to now re-
mains hypothetical. The sources are contradictory,
and the secondary literature offers incompatible so-
lutions. The first hypothesis is the easiest to defend:
In the initial phase of Koranic exegesis, there were
people who believed that, consistent with the ac-
count of surat al-Najm; Muhammad had seen God,
and not Gabriel. As I have pointed out, that contra-
dicts the image of the Prophet presented in his ca-
nonical biography. But at the time, Ibn Ishaq's
Sira was not yet a canonical book; the positions the
author supported were not always identical with
those of his contemporaries. One need only consult
al-Tabari's Tafsir; or commentary, to realize this. Al-
Tabari himself did not like anthropomorphism. His
60
r
t
THEOLOGY AND THE KORAN
skepticism, in fact, led to his being persecuted by the
Hanbalites of his time.
4
He therefore interpreted the
visions in surat al-Najm as being those of an angel-
Gabriel, to be precise. Nevertheless, he cites 'Abdallah
b. 'Abbas and Anas b. Malik among the Companions
of the Prophet who were inclined to believe it was a
vision of God, and he also cites 'Ikrima, Ibn 'Abbas's
slave and disciple. He even mentions a saying by Ka'b
al-Ahbar, who maintained that the Prophet had in
fact seen God twice) adding, with all the authority of
an expert in Judaism (even though he contradicted
the Bible in this particular case), that Moses had also
spoken with God twice.
5
This reminds us once again
of the old tripartite schema: Abraham-Moses-Mu-
hammad. In fact, that schema was attributed, pre-
cisely, to Ibn 'Abbas. A vision is the most profound
experience of the divine, and this vision of G ~ d is fi
ahsani suratin) as Ibn 'Abbas would have said, the vi-
sion of a benevolent and merciful God, very different
from the terrifying aspect he would have at the Last
Judgment.
But there were also conflicting traditions, result-
ing from a transcendentalist spirit that continues to
this day. Consider Muhammad's wife 'A'isha who, in
hadith, is frequently the interpreter of her husband's
most personal experiences. 'A'isha emphatically de-
61
THE FLOWERING OF MUSLIM THEOLOGY
nied that the Prophet had ever seen God. Elsewhere,
we find compromises of every kind. It was said, for
example, that the ascension, the vision, and the
Prophet's night journey (isra') had been only a
dream: "He saw God" (ra'a Allah), but "he saw in his
sleep" (ra'a fi nawmihl). Some also claimed that Mu-
hammad had not really seen God "with his eyes,"
but only bi'l qalb, "in his heart." In essence, that last
explanation also amounted to a dream, a truthful
dream. As it is said in the tradition, "the Prophet's
eyes are sleeping, but his heart is not sleeping." The
Koran taught that the soul returns to God during
sleep, while the body remains in bed. In addition,
the account of the vision ends with the sentence ma
kadhaba 'lfu'adu ma ra'c; "His heart did not lie about
what he had seen."
6
Finally, there were efforts to de-
materialize the object of the vision. Let us refer once
more to a hadith. A certain 'Abdallah b. Shaqiq al-
'Uqayli recounts: "I told Abu Dharr (al-Ghifari): 'If
I had met the Prophet, I would have asked him
a question.' 'Which one?' 'I would have asked him
whether he saw his Lord.' 'Well, I asked him exactly
the same question.' 'And what did he reply?' 'He re-
plied: "Light. How could I have seen him?"' [ nur, anna
arahu]."
The last sentence is somewhat difficult to inter-
62
i
1
t
I
I
I
I
THEOLOGY AND THE KORAN
pret, and variants do exist. But the intention of the
story is clear. Light has no form: thus, God does not
assume any shape (sura), and the vision is reduced to
a sort of bedazzlement. In a sense, that was the phi-
losopher's stone, a vision that did not interfere with
transcendence.
Yet all these speculations also show that no one
wanted to give up the vision as such; an angel was not
enough. The staking out of that position was not at
all an isolated or marginal phenomenon. We find evi-
dence of it everywhere. I will limit myself to a single
example, which might seem exotic but which mirrors
the vast agreement buttressing the idea. In about
AH r6o, a strange figure, a "heretic," according to the
Muslim sources, led a revolt against the Abbasid gov-
ernment in eastern Iran and Central Asia. This was
the Muqanna', a man who veiled himself <jp.d to
whom miracles were attributed; Jorge Luis Borges
mentions him in one of his essays? The Muqanna'
believed that God embodied himself in the proph-
ets-the first time in Adam when he created him 'ala
suratihl, in his image, and of course in Jesus as well,
and finally in Muhammad. He is said to have entered
Muhammad's body during the Prophet's vision, for
at that moment he was "two bows' lengths or even
closer," according to surat al-Najm. The heresio-
63
THE FLOWERING OF MUSLIM THEOLOGY
graphical text regarding the Muqanna' himself, in
which he revealed that he was an incarnation of God,
renders this as "as close to him as an arrow is to its
bow."S The scandalous aspect of his doctrine was pre-
cisely that idea of incarnation. To make it acceptable,
the Muqanna' cited an exegesis of surat al-Najm that
was clearly accepted by his audience. If he had in-
vented it, he would not have been persuasive.
That leads directly to the second hypothesis. Mu-
hammad saw God seated on his throne. That is a
more difficult assertion to make, since it goes against
not only the general view held by Muslims but also
against the Western study oflslam. It is true that for
philological reasons Orientalists generally assume
that there is a vision of God in surat al-Najm, but
they believe Muhammad saw God standing on the
horizon. The Muqanna' was of a different opinion.
The heresiographical passage I quoted says he be-
lieved that God returned to his throne after each
manifestation or "incarnation." His conception of
Muhammad's vision therefore implied that God had
left the throne to draw near to the Prophet and to
enter his body.
Once again, it is easy to quote other texts in sup-
port. But there is no need to do so, since we can refer
to a witness who mentions the idea within the pre-
64
THEOLOGY AND THE KORAN
cise context of the mi'rdj. This is a hadith, a non-
canonical and clearly apocryphal text that in its long
version (about twenty pages) is mentioned only by al-
Suyilti in his La' all al-masnu'a fi '1-ahdd'ith al-mawda'a. 9
But al-Tabari already quotes part of it, because the
chain of transmission once more begins with Ibn
'Abbas.l
0
He is followed by Dahhak b. Muzahim, a
popular preacher (qdss) and exegete who continued
the tradition of Ibn 'Abbas in Transoxiana. Dahhak
b. Muzahim lived in Balkh, and it seems to have been
from there that he transmitted the text in the region
of ancient Bactria, where religions and civilizations
had mingled and where a small Arab and Muslim
community maintained commercial ties with Central
Asia. Here are the essential lines:
I looked toward him (God) with my heart unt'fl I
knew he was truly there, that I was really seeing
him (hattri athbattuha wa-athbattu ru'yataha). And
there he was, drawing back his veil and seated
(mustawi) on his throne, in all his dignity and
grandeur ... In his majesty, he leaned slightly
toward me and had me draw near [to him]. Such
is his word in his Book, when he told you how
he dealt with me and glorified me: " ... one who
is mighty and wise. He stood on the supreme
65
THE FLOWERING OF MUSLIM THEOLOGY
horizon; then, drawing near, he hung suspended
within two bows' length or even closer, and re-
vealed to his servant that which he revealed,"
that is, the task he had decided to entrust
to me.
The task the text mentions is obviously the pro-
phetic mission. There were some (Anas b. Malik and
those who referred to him) who were even more pre-
cise: at that time, God revealed to Muhammad the
fifty prayers he wanted to impose on his community.
But Dahhak b. Muzahim placed the emphasis differ-
ently: on the vision of God. He did so cautiously. Mu-
hammad had to close his eyes, and Gabriel himself
(who, of course, was very familiar with the effect of
that vision) covered his face with his hands. For an
instant, however, the Prophet was assaulted by the
full force of the light, and that is what he now saw in
his heart: God seated on his throne, drawing near to
touch him and to transfer his revelation in a truly
corporeal manner. "He placed one of his hands be-
tween my shoulder blades, and for a time I felt the
cool of his fingers touch my heart."
That leads to a third hypothesis. The vision of the
throne represents the starting point and the point of
reference in the theme of the ascension. If the Koran
66
I
-T--
1
I
THEOLOGY AND THE KORAN
seemed to confirm a vision of God seated on his
throne, it was fairly easy to imagine that Muhammad
had risen to heaven to see God on the throne. The
new conception, however, included prominent and
previously unmentioned features; suddenly there was
the possibility of uniting the two visions into a single
point of view. During his journey, Muhammad might
have seen God first "on the supreme horizon," and
then "near the Garden of Repose." In addition, the
direction of the movement had changed. It was not
God but the Prophet who had moved; and he was ris-
ing, not descending, as God or the angel did in the
Koran. God never moves, he is immutable, according
to a view established in antiquity and adopted soon
thereafter by the transcendentalist theologians. He
remains immutable in his majesty and receives the
Prophet as a visitor, in an audience. But evan the
transcendentalists eventually overcame their distaste
for the theme of the ascension. At the same time,
it was now easier to salvage the Koranic passage on
the isrd'. There again, it was the Prophet who moved,
though horizontally, toward Jerusalem.
In principle, however, we need to distinguish be-
tween the night journey and the mi'raj. Despite the
fact that the two themes were melded together, they
were fundamentally different. In Ibn Ishaq's biogra-
67
THE FLOWERING OF MUSLIM THEOLOGY
phy of the Prophet, the two accounts are still sepa-
rate. Ibn Sa'd even assigns them different dates. From
the beginning, however, two things contributed to
their being amalgamated. In the first place, the
throne of God was said to be not only in heaven
but also in Jerusalem. In that city stood his earthly
throne, from which he had created the world and
from which he would dispense justice at the end of
time. And second, the night journey was sometimes
understood as a rapture, a "mystical" translation to
al-bayt al-ma 'mar of sura 52:4, the "Visited House,"
which people tended to see as the heavenly prototype
for the Kaaba in Mecca and, as such, a "very distant
place of veneration" (masjid al-aqsa)) a heavenly Jeru-
salem, so to speak. That opened up a completely new
dimension, which we cannot consider here. We are
compelled to note, however, that as a result the ac-
counts of the night journey sometimes also ended
with a vision of God, and this vision was described
in the same way as in the mi'raj. The Prophet meets
God in Jerusalem, in a garden of the haram al-shari
in a hortus conclusus (haz'ira) as it was called-that is,
within the walls surrounding the ancient Temple. It
was there that Muhammad had seen God seated on
his throne, in the form of a young man wearing a
crown of light; and God touched him with his hand,
68
THEOLOGY AND THE KORAN
between his shoulder blades, as a sign of friendship,
and spoke to him.
Now that I have considered the history of the theme,
let me conclude by attempting to establish a chronol-
ogy. First, we need to emphasize that we have been
dealing with exegesis and not reality. Whether the
Prophet saw God or only an angel is a matter of faith;
it cannot be resolved on the basis of texts. The Koran
is clear in that regard in surat al-Takwir, but not in
surat al-Najm. The verses in surat al-Najm were cer-
tainly prior to those of surat al-Takwir, but they do
not seem to be the expression of an immediate expe-
rience simultaneous with the telling, for in that case
the sura would not speak of two events and two vi-
sions at the same time. In addition, the text is homo-
geneous in appearance; the rhyme scheme is idefitical
throughout the sura, with the exception of the six
verses at the end (57-62). That is why we find no jus-
tification for the argument that the two visions were
combined in a later revision, under 'Uthman or ear-
lier. Under such circumstances, we are compelled to
hypothesize that the beginning of surat al-Najm does
not describe a unique event that took place a mo-
ment earlier, but rather refers to two separate events,
which by virtue of their singularity may have served
69
THE FLOWERING OF MUSLIM THEOLOGY
to confirm the truth of something else. What, then,
was the Sitz im Leben of surat al-Najm?
This topic remains to be examined. I cannot un-
dertake a study of it here. Nevertheless, allow me to
digress briefly again to al-Tabari. In his Tafslr and
in his Ta'rikh, he expresses fairly clear ideas on the
question. Accordingto him, what lies behind the very
first part of surat al-Najm is the scandal of "Satanic
verses."
11
The three pagan goddesses are mentioned
in verses 19-20, immediately after the account of the
second vision. The method they used to exert their
influence, namely, intercession with Allah, shafd'a, is
the theme of verse 26. Is the beginning of surat al-
Najm, therefore, simply a speech, a sermon even,
given by the Prophet when, as the accounts cited by
al-Tabari suggest, the Companions who had emi-
grated to Ethiopia returned to Mecca, ill informed
about everything that had taken place in the mean-
time and desirous of knowing the Prophet's position
on the rumors circulating? The Prophet, then, may
have used all his authority and eloquence to defend
himself against criticism. To do so, he evoked his en-
counters with the "numinous" (as we would now
say). That is why the text says: "Your compatriot is
not in error, nor is he mad! His is no language of pas-
70
THEOLOGY AND THE KORAN
sion. This is only revelation revealed to him. He is
taught by one who is mighty and wise." The impor-
tant thing was the encounter (or encounters) as such.
In that regard, the question of whom he encountered
was secondary. But the claim of having met God nat-
urally carried more weight than that of having met
Gabriel. For the moment, I cannot advance beyond
that hypothesis; I am speaking of exegesis and not
reality.
As far as the exegesis is concerned, we are faced
with a totally different situation. The Koran had be-
come scripture-that is, the canonized record of all
the statements uttered by Muhammad as revelation.
No one had forgotten that surat al-Najm might be
somehow related to "Satanic verses." But this was no
longer so important, for these verses were never in-
corporated into the final version of the Kora:n. The
two visions of surat al-Najm were presented in a new
light. Instead of being an allusion by the Prophet to
a thing already known, to prove his truthfulness to
an audience that may have briefly lost confidence in
him, these visions were now understood as direct .
testimony of his encounter with the divine force. At
the same time, the second Koranic reference to the
event-namely, the statement from surat al-Takwir-
71
THE FLOWERING OF MUSLIM THEOLOGY
began to exert its influence. When the Koran became
scripture, a complete and unchangeable text, com-
parisons could be made, a network of references and
cross-references created. At that point, theological re-
flection became a factor.
Under such circumstances, one can only be struck
by how well accepted the anthropomorphic interpre-
tation of surat al-Najm was in the early days of Is-
lam, though we cannot be sure that the discussion on
the subject actually goes back to the generation of
the Companions. Much of what was said about the
Urgemeinde, the early community, is derived from
later projections rather than reality. Recent research-
ers have viewed Ibn 'Abbas as a sort of mythological
figure.l
2
As for the accounts attributed to 'A'isha, the
supposition has been that she became aware of her
husband's doubts concerning the interpretation of
his visions. But her commentaries are mingled with
polemical remarks against certain Shiite currents
and are therefore apocryphal, or have at least been re-
formulated. As for the controversy as such, it is clear
it did not begin before the end of the first century AH.
During the caliphate ofHisham b. 'Abd al-Malik, be-
tween AH 105 and AH rzo,Ja'd b. Dirham was executed
in Iraq because, it was said, he denied that Abraham
72
THEOLOGY AND THE KORAN
was the friend of God and that Moses had heard the
voice of God on Sinai. He had apparently rejected the
tripartite schema attributed to Ibn 'Abbas and, as a
result, the vision of God accorded to Muhammad.l
3
It was also in Iraq that accounts of the mi'rdj seem
initially to have spread. There, the idea of a heavenly
journey was profoundly rooted in Hellenistic Gnos-
ticism and in an apocalyptic and mystical form of
Judaism. In Mecca and Medina, scholars proved cau-
tious. Ibn Ishaq, who could hardly conceive of the
idea that the Prophet had seen God in person, ac-
cepted only an attenuated variant of the account of
the ascension. Syrians preferred to speak of a night
journey; Jerusalem was the focus of their interest.
They were not against the idea that Muhammad had
seen God and had even been touched by him, but
they did not see the need for an ascension. Niverthe-
less, the theme took root almost everywhere, proba-
bly because Iraq was, under the Abbasids, becoming
the political and intellectual center of the Muslim
world. From Iraq, the motif spread to Iran, where
Iraqi troops had established themselves from the
Umayyad caliphate onward. The longest text comes
from Balkh. If it can legitimately be attributed to
Dahhak b. Muzahim, it too was written in the first
73
THE FLOWERING OF MUSLIM THEOLOGY
century AH.
14
Nevertheless, the first evidence of the
theme as such that can be reliably dated comes from
Iraq. Yet this account is unrelated to Muhammad. Its
hero is a heretic who claimed to be a prophet himself:
Abu Mansur al-'Ijli.
The figure Abu Mansur is complex and difficult to
interpret.l
5
Suffice it to say that he belonged to the
g h u l d ~ the lunatic fringe of ancient Shiism. He was
executed on the order of Yusuf b. 'Umar al-Thaqafi,
governor of Iraq between AH 120 and 126, after rebel-
ling against the Umayyad regime. To acquire legiti-
macy as a prophet, he claimed that he had been rapt
up to heaven ('urija bihi ila 'l-samd'). God is said to
have invited him to draw near and to have spoken to
him in Persian, calling him ya pisar, my son. He then
sent Abu Mansur back to earth with a mission to
preach his word. Abu Mansur seems thus to have
considered himself the son of God; his followers
called him Logos (al-kalima) and took oaths using
that word. God had passed his hand over Abu Man-
sur's head, and as a result Abu Mansur called himself
the Messiah ( al-masih ). In Arabic, the word masih sig-
nified something other than what it meant in the
Aramaic language from which it was borrowed. The
verb masaha did not originally mean "to anoint," as
did mdshah in Hebrew and meshah in Syriac. Masih is
74
-l
THEOLOGY AND THE KORAN
not the "Christ," the Anointed, but "one over whose
head the hand has been passed." Abu Mansur's fol-
lowers heard that word with Arabic ears. It was a very
vivid way of appropriating the ancient theme. It
seems that Abu Mansur wanted to compete with
Muhammad, but the model he followed was that of
Jesus.
The important thing for us is that he failed, for the
only thing his revolt accomplished was the occupa-
tion of a mosque. After that, no one could embrace
with impunity the idea of ascension. Only a prophet
could visit God. Meanwhile, the majority of Muslims
came to agree that Muhammad had been the last
prophet, the khatam al-nabiyyin. In many respects, the
Shiites disagreed with the majority view; for a time
they held on to the earlier ideas. They also continued
to embrace anthropomorphism, even after .,the era
when prophets had arisen in their own community.
But eventually they too changed their views and.con-
verted to the transcendentalism of the Sunnis. That
change was due to a theological movement called
Mu'tazila, whose systematic approach has deter-
mined Muslim thought even up to our own time. In
the Mamluk era, when someone in Cairo claimed he
had ascended to heaven and had seen and heard God,
he was, after a jurist had been consulted, quite simply
75
THE FLOWERING OF MUSLIM THEOLOGY
placed in an insane asylum. Conversely, the Prophet's
mi'raj) the founder's exaltation, had a long and glori-
ous history in art and literature, even in a Latin text,
Liber Scalae Machometi. In general, Muslims no longer
accepted the idea that the Prophet had seen God; he
had only heard his voice from behind a curtain. But
they also no longer perceived the event as a mere
dream. It was a real event, a miracle.
Nevertheless, the vision of God did not entirely
lose its importance. The Sunnis engaged in long con-
troversies over whether after the Last Judgment the
ru'ya bi'l-absdr (the "beatific vision" of the Christians)
would be part of the joys of paradise, and finally they
decided that such would be the case. All Muslims
would see God in paradise-as if in a theater when
the curtain rises-as a manifestation of divine grace.
But by then the world would have reached its end,
and the vision would be the apogee of eternal bliss.
The Prophet, conversely, still had a task to carry out
and for that reason returned to earth. His glorificat-
ion marked the beginning and not the end. He be-
came the symbol of Muslim identity, and in that re-
spect his uniqueness was felt more than ever. But in
our own time that uniqueness has been defined in
the first place by the categories of the world here be-
low, his role as lawmaker and spiritual guide. In the
76
THEOLOGY AND THE KORAN
end, his supernatural encounter with God remained
an isolated event. In that regard, Islamic thought, by
comparison with Christian theology, has always re-
spected the limits of the human condition. That may
be an advantage.
77
3
THEOLOGY AND SCIENCE
Mu 'tazilite Atomism
THERE IS SOMETHING very modern-sounding
about the word "atomism."
1
But appearances can be
deceiving. The Muslim theologians I wish to discuss
were atomists, but their atomism was different from
our own. Like the Greeks, they dealt with it in a
purely speculative manner. However imaginative they
may have been, they never felt the need to verify the
existence of the atom; they simply posited., it as a
theoretical necessity. For them as for the Greeks, an
atom was what the literal meaning of the word indi-
cates: an indivisible entity, juz' la yatajazza'. It did not
occur to them to split it. They were not precursors,
they were heirs. But as heirs, they displayed original-
ity; their approach was unique. In that respect, they
were modern in their own society and for their time
in history.
Theirs was a complex inheritance. Several points of
79
THE FLOWERING OF MUSLIM THEOLOGY
view and models could be drawn from it. The Greek
past was remote, but it was not entirely forgotten.
People's brains were still filled with ancient notions
and ancient problems. Despite the emergence of Is-
lam (and before it of Christianity), the program for
research-by which I mean the mental program, the
human computer code, the categories of thought,
and not the education system-had not changed a
great deal. That was true in the Orient as well as in
Constantinople and Rome. But in the place and time
at issue, the Iraq of the Abbasids, traces of antiquity
mingled with other influences: those of Sassanid Iran
and even India (in medicine, for example, but also, fa-
mously, in arithmetic). So much freedom of choice
soon brought trouble. Hence our first question: Why
did Muslim theologians, mutakallimun as they were
called, opt for atomism?
Atomism was far from a passing hypothesis. The
model had been used by a few pre-Socratic philoso-
phers and revived and reinterpreted by Epicurus.
Enormous resistance was mounted against it by Aris-
totle in his Physics, by the Stoics, and by the Neopla-
tonists. David Furley has spoken in this context of
the cosmological crisis in classical antiquity; what
he means by that is precisely the profound difference
80
THEOLOGY AND SCIENCE
of opinion regarding the origin of the world and of
things.
2
That difference was exacerbated by new axi-
omatic positions. Philosophers had been replaced by
theologians. Henceforth, not only the world but God
as well was in play. We could even wonder whether
there was really a place for God within atomism, a
system that, at least since the famous controversy be-
tween Rene Descartes and Pierre Gassendi, has repre-
sented the epitome of materialism. Nevertheless, the
Stoics in their works include the chapter Peri the6n,
"On the Gods," in the part on physics; theology and
cosmology were never systematically separated. Even
Aristotle's Metaphysics, the ilahiyyat of Arabic philoso-
phers, did not initially go as far "beyond" (meta)
physics as we think. It was not yet conceived of as
something transcendent but was a subject addressed
"after" the Physics in the curriculum of thr, Peripa-
tetics' lyceum. But it might have been easier for a
theologian to find inspiration in Aristotle's Unmov-
ing Mover or in the One (hen) of the Neoplatonists.
The Church Fathers decided to move in that direc-
tion. They despised atomism because they saw it as
an attempt to explain the world in a wholly mechani-
cal way, without metaphysical principles, simply as a
result of chance, with no creation and no God. And
8 I
THE FLOWERING OF MUSLIM THEOLOGY
they were wary of the Epicureans, whom they consid-
ered vulgar people who demonstrated their material-
ism through their lifestyle.
The most ancient Muslim theologians may have
been of the same opinion. Some researchers have hy-
pothesized that thinkers during the Umayyad caliph-
ate, Jahm b. Safwan in Iran or Ja'd b. Dirham in
Mesopotamia, for example, followed a Neoplatonist
model that led to a profoundly apophatic theology, a
theology of tanzih, uniqueness, that was little con-
cerned with the world. This was a "theology" in the
true sense of the word, a logos about God (theologia)
with no digressions into physics or cosmology.3 Un-
fortunately, the few shreds of information that have
come down to us are not enough to form a clear and
unequivocal idea about them. But if that hypothesis
is correct, we can imagine that Harran, in Upper Mes-
opotamia, capital of the last Umayyad caliph but also
seat of a Neoplatonist school and bastion of a pa-
gan astral religion, played a role in the transmission
of these ideas. Ja'd b. Dirham lived in Harran, and
Jahm b. Safwan may have studied there.4 In any case,
atomism was not present from the start in Muslim
culture. Rather, it was discovered and developed at
the apogee of the Abbasid dynasty by a generation
that included Mu'ammar and Abu '1-Hudhayl, who
82
T
' l
THEOLOGY AND SCIENCE
taught during the caliphates ofHarun al-Rashid and
al-Ma'mun. Shortly after they had constructed their
new approach, the anti-atomist resistance coalesced
once again. The form ofNeoplatonism championed
by al-Kindi and his team of Christian translators put
the works of Plotinus and Proclus into circulation,
evidently under the label "theology in the manner of
Aristotle," that is, metaphysics.
5
The Christian theo-
logians of the time who wrote in al-
Basri, Abu Ra'ita, Theodore Abu Qurra-also based
their thip.king on a Neoplatonist vision. The transla-
tors could thus feel they were being true to their own
religious tradition. Conversely, the path the mutakal-
limun were to take is much more difficult to discern.
Mu'ammar and Abu '1-Hudhayl had no translations
to rely on. No text by Democritus or Epicurus was
ever available in Arabic. And though the of
the Bayt al-Hikma, or House of Wisdom (the famous
library of the Abbasids), dated back to the caliphate
of Harlin al-Rashid, the translations it produced do
not seem to have made an impression on the major
figures of atomism. Abu '1-Hudhayl and Mu'ammar
were trained in Basra and not Baghdad, and they
do not seem to have moved to the capital until al-
Ma'mun's reign.
* * *
83
THE FLOWERING OF MUSLIM THEOLOGY
Nevertheless, atomism did not come from nowhere,
and its formation can be traced. Before Abu '1-
Hudhayl and Mu'ammar, there was Dirar b. 'Amr, a
thinker from al-Kufa who visited Baghdad during
the reign of the Barmakids. He was the first Mu'tazil-
ite to develop a set of coherent notions about physics,
and his approach was marked by the hypothesis that
reality consists of parts that combine to form a body.
That is the atofhistic principle, but without the vo-
cabulary normally associated with it, since the word
Dirar uses to designate the "parts" (ba 'd) is different
from the word adopted for "atom" (juz'). And its
meaning is also fundamentally different, for when he
spoke of "parts," he was not thinking of particles or
tiny bits of matter, but of phenomena by virtue of
which any body whatsoever is perceived. He called
these parts "accidents" (a 'rad) and described them as
such: color and temperature, weight and state of the
surface, animation or inertia, and so on. It is through
these accident-parts that the body presents itself, and
it is only in the form of these "accidents" that we can
perceive it. Qualities exist only in combination; there
is no color or temperature as such. But there is no
substance either. The body is only a conglomeration
of phenomena that form the parts of the image or
the evidence we form of the body. The aim of that
84
THEOLOGY AND SCIENCE
theory is less to explain reality than to describe our
idea of it. Reality is what we perceive of it, a quantity
of details that are all accidental but that, by combin-
ing, present themselves to us in the form of coherent
objects. Richard Sorabji has spoken in this context of
"bundle theory" and has found many examples of it
in late antiquity. 6
Abu '1-Hudhayl seems to have been inspired by
that model. He recognized its theological advantage:
the "bundles" need someone to assemble them, and
human beings can play that role only from the epis-
temological angle, as subjects of knowledge. If, how-
ever, one considers reality-that is, if one asks not
only by whom the bundles are perceived but also by
whom they are created-the answer can only be God.
It is God who puts the parts together, and since real
parts, not to say material parts, are now in qvestion,
it is better to use the term "atom." The phenome-
nological aspect has become secondary; for Abu '1-
Hudhayl, the "parts" Dirar spoke of are merely acci-
dents that appear in general only after the fact. The
atom, by contrast, is also a substance. God wills
atom-substances to form bodies; that is called cre-
ation. To achieve the result he has in view-that is, to
join together or assemble a quantity of atoms-God
adds the accident of juncture, assemblage (ta'lif); he
85
THE FLOWERING OF MUSLIM THEOLOGY
thereby creates dimension and corporeality. Natu-
rally, he can also take away that accident of cohesion
after some time; that is called disintegration, decom-
position, death. And he can even add the accident of
juncture a second time. That is what will happen at
the resurrection, when God will bring human beings
back to life and will create a new earth. He reassem-
bles the bones, so to speak. But the metaphorical as-
pect of that image has now become obvious; the vi-
sion of Ezekiel, where that conception appears for
the first time in the Old Testament, is replaced by a
scientific construct. 7
Atomism as ancilla theologiae> subject to the will of
God, is what Abu '1-Hudhayl wanted to achieve. He
changed a materialist model into an instrument of
monotheism, at first in a wholly speculative way, but
in a manner perfectly consistent with Koranic reve-
lation. He had satisfied the postulate of oneness,
tawhld: Everything that is made depends ultimately
on God, except human free will. There is no nature in
charge of things, and atoms, particles of being, do
not act on their own or by chance. Fundamentally,
they are only symbols of divine power. In fact, God
does not even use them as building blocks. Rather, he
calls things into being through his creating word, the
kun> the "So be it" recorded in the Koran, and he does
86
THEOLOGY AND SCIENCE
so all at once, without physical genesis.
8
On that mat-
ter, Dirar held the same view. But his approach was
too sensualistic; his epistemology implied a concept
of being that made the existence of God difficult to
prove. In fact, it was Abu '1-Hudhayl who formulated
the first proof of the existence of God in Islam.
9
With
what had originally been a materialist model, he
overcame materialism. He was able to satisfy the re-
quirements ofhis time.
The two distinctive elements in that development,
epistemological sensualism and the hypothesis of the
atom, indicate the likely site at which Muslims were
able to revive that tradition: Iranian cosmology. In-
deed, it is only in relation to that cosmology that
Arabic sources mention these two elements as some-
thing living and contemporary. Christian theology
does not breathe a word about them, either in f>.rabic
or in Syriac; Greek philosophy was known only
through literary allusions, and only imprecisely at
that. Aristotle's Physics had not yet been translated,
and the doctrines of the pre-Socratics were still un-
known.lo Although we have few accounts of Iranian
cosmology, there is no doubt that the Muslims were
very familiar with it, especially in Basra, where a large
part of the population spoke Persian. Doxographers
summarized it and theologians vigorously attacked
87
THE FLOWERING OF MUSLIM THEOLOGY
it. Those against whom they directed their polemic
were not phantoms; they lived in their region, and
the two camps ran into each other in the homes of
middle-class city dwellers.l
1
People had taken to find-
ing fault with the Manichaeans for their sensualism,
that is, their belief that only what can be perceived
by the senses actually exists. The criticism was false
and was frequently used to demonstrate that those
Muslims who were spiritual descendants of the
Manichaeans-namely, the zanadiqa, learned men and
intellectuals who did not directly reject the principles
oflranian cosmology-were not only heretics but also
libertines. In relation to the earthly world, however,
the remark was probably justified. The Manichaeans
of that time-and along with them many other dual-
ists-presented their religion in a scientific form
without appealing to irrational convictions. They
wanted to explain the world, and in that respect their
sensualism or empiricism was shared by an entire
generation of Muslim theologians. In addition to
Dirar, for example, there was also his Basrian con-
temporary al-Asamm.
12
The principle remained valid
so long as the concept of God was strictly apophatic;
if God is the Other par excellence, he is beyond our
reason just as he is beyond our senses. The world can
88
~ ~ - - --- --- --
~
l
THEOLOGY AND SCIENCE
be analyzed through phenomena, but God is acces-
sible only through revelation. That view is still ex-
pressed by Dirar, but no longer by Abu '1-Hudhayl.
The atomism defended by Abu '1-Hudhayl was less
characteristic of the Iranian tradition than was sen-
sualism. The atomistic model appears there only as
an option, and not even the first option. The key no-
tion was rather that of mixture. In all the dualist sys-
tems known at the time, the two first principles, light
and shadow, had to mix in order for the world to
form. To separate that process from mythology, one
had to invoke the traditional Stoic philosophical
category of krasis di' holou rather than the atomism
of Democritus and Epicurus. That was not roo per-
cent true, however; there were atomists among the
zanadiqa, as the doxographical sources tell us.B The
phenomenon they sought to explain with t\l.at ap-
proach was primarily motion.
The mechanics of motion had always attracted atten-
tion. For Aristotle, motion was one of the six kinds of
change-namely, change in relation to place brought
about through locomotion. There were other sorts of
change-change in relation to substance, for example,
which manifested itself in the phenomena of genera-
89
THE FLOWERING OF MUSLIM THEOLOGY
tion and corruption, or change in relation to acci-
dents: a body gets hot after being cold, and so on.
For the dualists, these two sorts of change were ex-
plained by mixture. But motion was different, fleet-
ing and difficult to pin down. In contrast to mate-
rial changes, the body in motion always remains the
same; the result of motion is manifested only in dis-
placement, after locomotion has ended. Aristotle had
interpreted motion as a coherent process; he thought
in continuist terms, as do we. But it was only a hy-
pothesis, an axiom that was rather vague at the time
and difficult to prove. What happens during motion?
To make the problem more transparent, one can al-
ways imagine that the interval separating the begin-
ning of the motion from the end of it is shortened to
such a point that a hypothetical body found some-
where at a given moment is found immediately after-
ward, "at the second moment" (fi '!-hal al-thanl)> in a
different place. Since only a minimal amount of time
has elapsed, the second place ought to be contiguous
with the first. And if, again in the interest of greater
clarity, one imagines that the moving body is also re-
duced to minimum size, one arrives at a model in
which atomic units move over a surface that consists
of compartments entirely occupied by these units.
90
? J: I;
I
I
~ ~ - ~ ! ~ - ~
THEOLOGY AND SCIENCE
Naturally, actually existing bodies consist of a quan-
tity of these atoms and their motion is more reminis-
cent of the way a millipede walks. But one can imag-
ine an ant or a mite instead. At every instant in its
movement, it is faced with a contiguous compart-
ment in a discrete space. In fact, the modern Arabic
word for "atom" is dharra> which originally meant a
small, undefined insect-a mite, if you like.
We do not know whether the dualists-or more
precisely, those among them "(ho were working from
the atomistic hypothesis-understood the model in
that way; the documentation is missing. But we are
certain that Abu '1-Hudhayl conceived things in
those terms.
14
Even in his case, however, the sources
are sparse: we have no original texts, only the ac-
counts of doxographers. But Abu '1-Hudhayl's view of
things caused a great stir. The theory he acllfanced
was repeated and discussed for centuries. In the
fourth and fifth centuries AH, more than a century
after his death, the theory was erected into a rigid
scholastic system. For this period, texts abound, rela-
tively speaking. The present difficulty stems from the
fact that these texts have become available only re-
cently; some have not been published yet, and the
editions that do exist are far from reliable. We must
9 I
THE FLOWERING OF MUSLIM THEOLOGY
always be wary of reconstructing a later period on
such a basis. Up to now, no one has considered
what happened during the phase separating Abu
'1-Hudhayl from his epigones.
For Abu '1-Hudhayl himself, it is self-evident that
every "substance," whether a body or a simple atom,
is found somewhere; it occupies a concrete place.
That is what he calls kawn) "location."lS Alongside
the kawn) there is the makan) the place that can be
filled by an object, and the m a h a l ~ an accident's sub-
stratum. During motion, the object changes its kawn)
but before arriving at its new location it can be found
in several amkina (the plural of makan) or spatial
points, points at which it does not pause and which
remain devoid of its presence. Abu '1-Hudhayl's set of
axioms does not provide for eternal motion, a pre-
clusive decision that, coming from a theologian, is
readily understood. According to Abu '1-Hudhayl,
even the blessed will stop moving in paradise eventu-
ally, and the motion of the stars and spheres will
cease with the end of the world.JG In addition, the
points-or rather the compartments-touched by an
object as it moves are all of equal value. Nothing pos-
sesses a place proper to it and toward which it tends,
an idias tapas) as Aristotle would have said. Neverthe-
less, the compartment occupied by an atom is re-
92
THEOLOGY AND SCIENCE
served for it thanks to the kawn; in no case could
two atoms share the same place. That characteristic
would later be called tahayyuz) the quality of filling a
place.
For subsequent generations, that tahayyuz was an
accident-the only accident, moreover, that was indis-
pensable to the atom.J7 Conversely, Abu '1-Hudhayl,
who was unfamiliar with that term, conceived of
the kawn) in which the tahayyuz was preformed, as it
were) only as something indefinable but individual
and perceptible-a m a ' n ~ it was called at the time.
What interested him most was not the atom pure
and simple but the ta'li the accident by virtue of
which atoms form something coherent by joining to-
gether, that is, by filling two or several contiguous
compartments at the same time. As a result, the
problem of motion was raised in a different.,form.
For a complex body, motion was an accident whose
substratum was divisible, and the question arose
whether that accident was inherent in all the atoms
of the substratum or whether it occupied only a few
of them. That question would soon be of great im-
portance. For the moment, however, no one knew
whether the system would have a chance to develop,
since Abu '1-Hudhayl found himself facing a criticism
that went to the heart of the matter. It came from his
93
THE FLOWERING OF MUSLIM THEOLOGY
nephew al-Nazzam and is the reason that the discus-
sion for a while focused only on the principle-that
is, on the possibility of proving the atomistic hypoth-
esis pure and simple.
Al-Nazzam represents the other panel in the cosmo-
logical diptych that Islam inherited from antiquity.
His network of axioms is anti-atomistic and is based
rather on concepts formerly known and accepted by
the Stoics. But once again, the resulting philosophy
was filtered through the Iranian tradition. In addi-
tion, the cluster of categories he had received from
the dualists was much clearer and broader in scope
than it was for Abu '1-Hudhayl. His governing idea
was mixture, though he refrained from using that
term too often. For him, bodies are not aggregates of
atoms in which conflicting qualities alternate in the
form of accidents. Rather, these qualities themselves
are bodies that intermingle. An apple, for example, is
a body that consists of other bodies, namely, color,
. odor, taste, and so on, which intermingle-"interpen-
etrate," as al-Nazzam said. There are simple bodies
and mixed bodies, and the simple bodies (colors,
tastes, and so on) are the aspects as which mixed bod-
ies (the apple) present themselves. In a sense, the old
sensualism was not yet forgotten, and Dirar's theory
94
, __ ..,
T
. ........ .." ..
I
THEOLOGY AND SCIENCE
was once more on the ascendant. What had in the
past been called "accidents" were simply replaced by
"bodies." But alongside actualized and visible quali-
ties, the mixed body also contains unperceived poten-
tial qualities. However hot an object may be, within it
is always simultaneously a certain quantity of cold-
ness; otherwise it would burn. Different qualities can
alternate in a mixed body only because opposites are
hidden inside. As soon as one of the opposites comes
to the surface, the body changes in appearance. Nev-
ertheless, it continues to exist as a coherent entity
only by virtue of the balance of opposite qualities. If
one of them abolished the other completely, the com-
posite body would disintegrate (through being con-
sumed by fire, for example). All the qualities thus ex-
ist simultaneously and, in a sense, in the same place
as well. Similarly, all the qualities were created .91-t the
same time. It is God who put them together in a
mixed body and who maintains their balance despite
the fact that they are opposites. The only accident
in this whole mixture is motion, but motion under-
stood in the broadest sense of the word, as the princi-
ple of change.
18
For al-Nazzam, motion in the restricted sense-
that is, motion understood as locomotion-occurs in
a continuous and infinitely divisible space. According
95
THE FLOWERING OF MUSLIM THEOLOGY
T
to him, then, the moving object does not proceed --t---
from one compartment to another, and it is not I
always associated with a certain makan. Rather, it I
"jumps" from time to time, to cross infinity. The the-
ory is both original and complicated; I shall not de-
scribe it in detail here.1
9
But we must keep in mind
the difference between the models. Al-Nazzam's con-
tinuous space is that of Euclid and Aristotle; by con-
trast, the atomists declare that space is discrete. It is
possible that Abu '1-Hudhayl had not yet clearly per-
ceived the incompatibility between the two concepts;
perhaps he had not completely realized that atom-
istic matter presupposes an atomistic space. Even a
theologian such as Abu Hashim, who lived more
than a century later, did not accept that conclusion.
In any case, al-Nazzam's merit was to have put his
finger on the problem areas. These were not totally
unknown: together, Abu '1-Hudhayl and al-Nazzam
sometimes repeated the commonplaces of an inher-
ited philosophical discourse. Both, for example, tried
to make use of Zeno's paradoxes, or what had be-
come of them. But all this can hardly overshadow the
fact that it was only at this moment that discussion
began in Muslim theology and that a number of ele-
ments in that discussion were new. If the atom fills a
compartment and touches others, it must have sides.
96
THEOLOGY AND SCIENCE
Why, then, wonders al-Nazzam, is it said that the
atom acquires extension, length, and later width and
depth only in contact with another, through junc-
ture (ta'lif)? An entity that has sides will also always
have dimensions.
20
And why would it not be possible
to imagine an atom that stands at the juncture of
two other underlying atoms? That superimposed
atom would then "claim" both at once, but only half-
way, which amounts to saying that the two would
be divisible.
21
The arguments are problematic, even
false; they show that al-Nazzam had trouble abstract-
ing from sensory data and accepting the axioms of a
noncontinuist geometry. Hence, he sometimes stops
short of his conclusion. He constructs an interesting
argument concerning the diagonal in a square, for
example, but does not say that in a square the length
of the diagonal is always an irrational number (the
square root of r8, 32, so, and so on), which can never
be represented by discrete, atomistic units.
22
He was
not a mathematician, and his capacity for abstraction
was generally less well developed than his uncle's. But
he forced the atomists to make their theory more ex-
plicit.
It appears that Abu '1-Hudhayl assumed that all at-
oms are equal and homogeneous_ They have neither
97
THE FLOWERING OF MUSLIM THEOLOGY
hooks nor rings, by contrast with what Democritus
had said; and as Epicurus suggested, they can appear
in different forms. Mu'ammar, a contemporary and
colleague of Abu '1-Hudhayl, had imagined that eight
of them form a cube, four on the bottom and four on
the top. That is conceivable only if one assumes that
every individual atom also presents itself in the form
of a cube, a hexahedron.
23
Yet it would be a distortion
to say that they are cubes, for a dimensional body, ac-
cording to Mu'ammar, results only from eight atoms'
being placed together. Two atoms form only a single
dimension: length; and even when two others are
added to them, the result is only length and width,
without the depth necessary to any corporeal entity.
Similarly, it would be unwise to propose that two at-
oms form a line, two lines a surface, and two surfaces
a body; only dimensions are at issue. That reasoning
becomes clear when it is compared with that of Abu
'1-Hudhayl, who claimed that .six atoms were suf-
ficient to form a body: two for the left and right, two
more for the front and back, and finally two for the
top and bottom. That never results in a geometri-
cal body; even if one imagines a cluster, a coherent
group without regular sides, a seventh atom would
be needed in the middle. The six atoms mark only
spatial directions.
24
In addition, the number of at-
98
- ---
T
1
THEOLOGY AND SCIENCE
oms was later reduced even further. Abu '1-Qasim al-
Balkhi would say that four atoms suffice; one need
only imagine that an unchanging central atom is
paired with a second and that together they mark out
left and right, then with a third to mark out front
and back, and finally with a fourth marking top and
bottom.
25
In all these models, the atoms amount to "min-
ima" (elakhista) understood in Epicurus's sense, but
located in a discrete space that has the structure of a
three-dimensional grid. They fill a compartment, and
one can visualize that compartment as a square, like
one on a chessboard. But as to their form, one must
always keep in mind that the cube is only a meta-
phor. The atoms resemble cubes, as a later theologian
said. To be real cubes, they must be bodies. They are
not points; the term "point" is part of Euclid' con-
tinuist geometry. But we still need to resolve the ques-
tion of whether they have extension. In his pioneer-
ing book published more than half a century ago,
Shlomo Pines responded in the negative; in fact, we
know of no text that provides a response where Abu
'1-Hudhayl is concerned.
26
There is no doubt, how-
ever, that fourth-century theorists replied in the af-
firmative, in invoking the category of tahayyuz. Abu
'1-Hudhayl had only kawn, "location," and he did not
99
THE FLOWERING OF MUSLIM THEOLOGY
yet conceive of kawn as an accident. But though he
had not yet clarified his categories, the internal dy-
namic of his system pushed it in the direction his
successors were to pursue. These successors, haunted
by al-Nazzam's objections, hastened to make that dy-
namic explicit.
But was the theory adequate to explain reality? Let
us consider horses at a racetrack, for example: What
happens at the atomic level? They run at different
speeds and can accelerate. Perhaps the example is too
complex; a motion executed by a living thing is also
an action, and that adds a new dimension. Yet the
same phenomena can be observed in an inanimate
object-an arrow, for example. When an arrow moves
after it is shot, it does so at a constantly decelerating
speed. To explain its motion, one would have to in-
troduce the factor of time. Abu '1-Hudhayl does not
seem to have thought of that. The doxographers did
transmit his definition of time-which is a rare thing
in the sources available to us-and he even developed
it from an atomistic foundation, speaking of the in-
stant (waqt) instead of continuous time (zaman). But
his intention is anthropological rather than physical;
he seeks to explain the relation between time and hu-
man acts.
27
He also knew that motion always unfolds
IOO
THEOLOGY AND SCIENCE
in time, proceeding from a first instant (waqt awwal)
to a second (waqt thanf). But in considering the phe-
nomenon of acceleration and changing speed, he
limited himself to an altogether material context. He
said that motion as an accident is connected not sim-
ply to the compact body as a whole but to its particu-
lar atoms. It is not necessary for every atom of the
body (of the arrow) to be occupied by a unit of mo-
tion; even if only some serve as a substratum for that
accident, the concrete body begins to move or does
not completely stop moving. The other atoms are oc-
cupied by units of pause, waqafat, as Abu '1-Hudhayl
put it; the relation between units of motion (harakat)
and units of pause ( waqafat) determines the speed.zs
Abu '1-Hudhayl's system differs from Epicurus's in
its details. Epicurus had explained the phenomenon
of acceleration by the fact that at first the atolll6 of a
moving body do not all move in the same direction.
It is only after a certain time that motion becomes
uniform. Every directed motion thus passes through
a stage of inertia. Conversely, Abu '1-Hudhayl tried to
quantity the process. It seems; moreover, that his dis-
tinction between harakat and waqafat corresponded
better to his concept of discrete geometry. All the
same, his so!ution raised particular problems. It was
necessary to take into consideration that in general
IOI
THE FLOWERING OF MUSLIM THEOLOGY
the accident of motion is conferred on the moving
object by an external force, often a human being. And
sometimes the force is exerted by two people at once
-if, for example, two people want to move a stone.
We must then imagine that each of the two acts on
only part of the atoms; the people divide them up be-
tween themselves, so to speak, each according to his
or her individual strength: But let us suppose the at-
oms are odd in number; in that case, the two people
could never apply their force equally.
2
9
We do not know whether Abil '1-Hudhayl would
have been impressed by such an argument. He might
still have said that for a slow-moving stone there are
always atoms not yet occupied by an accident of
motion. Since we do not know their number, we do
not know whether that number is odd or even. The
argument cited, in fact, was not directed against Abil
'1-Hudhayl but against one of his contemporaries,
Bishr b. al-Mu'tamir. The concession the atomists
would eventually mal(e on that question was to
maintain that even accidents-motion, but also color,
and so on-are divisible. Something like accidental
atoms exist, a consequence that corresponded poorly
with the spirit of the system.
30
But that' is no more
than a detail. The most important thing is that once
again al-Nazzam had considered the question. The
102
THEOLOGY AND SCIENCE
difference lies simply in the fact that in his polemic it
is always the theory of the "jump" that comes into
play. He posited, for example, the dilemma of a man
walking on a moving ship. In that case, the man is
not going faster than the ship; nevertheless, he tra-
verses a greater distance, for in the time it takes the
ship to travel twenty yards, let us say, the man can ad-
vance from the stern to the bow-that is, another
twenty yards. In relation to the surface of the water,
he has thus traveled forty yards, whereas the ship has
traveled only twenty. That has nothing to do with the
number of atoms composing the man and the ship;
he must therefore have jumped. We encounter the
same difficulties in analyzing the process in relation
to the surface of the water on which the two motions
occurred. If we assume that the ship touched every
compartment of that surface-and these compart-
ments are atomistic by nature-the passenger must
have touched only half of them while traveling twice
the distance.
31
Let us take another example, a rotat-
ing millstone. The atoms move in concentric circles,
and their number increases as these circles approach
the periphery. Yet the distance traveled by the periph-
eral atoms during a complete rotation is much
greater than that traveled by the atoms near the axis;
they must therefore jump.3
2
103
THE FLOWERING OF MUSLIM THEOLOGY
The last theorem reminds us of another famous
paradox known as Aristotle's Wheel, a problem dis-
cussed for the first time in Pseudo-Aristotle's Mechan-
ics. Hero of Alexandria speaks of it as well. (It is inter-
esting to note that his text survives only in Arabic
translation.) In short, we are nearly certain that the
roots of that whole multiform discussion go back to
antiquity. Similarly, we find parallels to al-Nazzam's
"jump" in the writings of the Neoplatonist Damas-
cius and in other late Greek thinkers.
33
But the de-
tails of the argument are unknown to us, as are those
of the generations immediately following the dis-
pute, the theologians of the third century AH. Once
again, we must resign ourselves to extrapolating from
fourth- and fifth-century texts. Apart from a few re-
marks by doxographers, we have nothing precise on
the first reactions to al-Nazzam's criticism. The later
texts show us, however, that in the end the theoreti-
cal basis of atomism was refined and subtly modified.
The difference in speed was no longer explained by
the quantity of atoms occupied through an accident
of motion. It was now said that motion could be in-
terrupted by moments of pause or rest (sakanat), i n ~
finitely small and imperceptible moments that cause
the speed to abate.
34
In the millstone, for example,
104
THEOLOGY AND SCIENCE
that is the case for the atoms located near the axis,
by contrast with the peripheral atoms, which simply
continue their course. These instants of rest were
even observable on a cogwheel, more or less as they
are on modern train station clocks or in a film when
it is slowed down. By the end, the model, with that
adjustment, had completely incorporated temporal
atomism; the instants of rest could be conceived only
as minima in Epicurus's sense. And the phenomenon
of acceleration was now explained like any other mo-
tion, in relation to the underlying space and not in
relation to the moving body. The theory had thus
achieved greater coherence. The new hypothesis ac-
cepted the postulate of the usual model-namely,
that an isolated atom (which can carry only a sin-
gle accident of motion) advances, within one unit of
time, only by a single compartment in spaae. Con-
versely, the explanation Abil '1-Hudhayl had given
could apply only in cases where there were at least
two atoms. Al-Nazzam had certainly contributed to-
ward clarifying the argument. But essentially, the di-
lemmas he had brought to light remained insolu-
ble; at the level of axioms, the opposing parties did
not speak the same language. Was it truly possible to
posit that a millstone is composed merely of rings
lOS
THE FLOWERING OF MUSLIM THEOLOGY
of atoms? It is a coherent object, and if at different
points certain atoms pause, while others continue
their course, the millstone shatters.
35
Of course, atomism as conceived by Abu '1-Hudhayl
was not designed to be used by physicists. In fact, the
little we know of the discussion by his earliest succes-
sors shows that at first it continued to focus on axi-
oms. Abu '1-Hudhayl's heirs were tremendously pre-
occupied by the relation between the atom and the
body or-and this mattered a great deal to a theolo-
gian-by whether it is possible, in a fundamentally
materialist model such as atomism, to distinguish
between the inanimate world and living beings. The
minds of the mutakallimun were shaped by legal cate-
gories, and that is why they persisted in construct-
ing a definition before solving a problem. As we have
seen, they were constantly reducing the number of
atoms necessary to form a body: eight, according to
Mu'ammar, six according to Abu '1-Hudhayl, and
finally, four for Abu '1-Qasim al-Balkhi. Hisham al-
Fuwati, one of Abu '1-Hudhayl's disciples, had set out
in the opposite direction, by postulating molecules
with thirty-six atoms, but his idea met with no suc-
cess. Some people were even ready to set more strin-
gent limits: three atoms (Qalanisi), two, or even
106
THEOLOGY AND SCIENCE
one.
36
Indeed, if one imagines a body in light of its
definition, it is nothing but a composite (al-mu'allaf)
formed by the ta'lif, the juncture or cohesion that
holds its atoms together. And for a simple juncture,
two atoms suffice. But the ta'lifis an accident; where,
then, is its substratum? Can one really assume there
is an accident inherent in two substrata at a time-
that is, in the two atoms that have combined? Why
not postulate that the ta'lifis located as a potentiality
in each atom, while all the rest is only a verbal prob-
lem, given that one begins to speak of juncture only
after it has actually come about? Atom and body
would then be virtually identical, and it would be
possible to forgo any discussion of the different con-
notations of the term jawhar, "substance."37 The con-
sequences were enormous for the general context of
the theory. An atom that is joined only to a single
other to form a body no longer needs sides; if the
two touch, they become wholly indistinguishable.
The man who defended that thesis, a certain Salih
Qubba, had studied with al-Nazzam; he knew that
the idea of an atom with sides but no dimensions
bordered on the absurd.
38
But he remained an at-
omist, for he postulated that every atom occupies a
space and that two atoms occupy more space than
one, despite the fact that the two are now found in
107
THE FLOWERING OF MUSLIM THEOLOGY
the same place. His theory contradicts the evidence
of the senses, but on the speculative level it was not
lacking in coherence. Other people, though attacking
its internal contradictions, embraced the basic ap-
proach. Al-Ash'ari himself believed that two atoms
sufficed to form a body, and many Asha'rites fol-
lowed him.
The second problem raised from the start be-
longed to what we would now call microbiology: the
atomic structure of living beings. Nevertheless, as
one might expect, the emphasis was placed on an-
other dimension: anthropology. Dirar b. 'Amr had
hypothesized merely that the human being, like any
being open to perception, is composed of sensible
and accidental phenomena that can be observed. In
other words, one can perceive his warmth, his com-
plexion, his smell. But his person is limited to that
"bundle"; he has no soul, at least no soul indepen-
dent of the body. The human being is reduced merely
to the way he functions. His personality is manifest
in his actions, which can be called good or bad in ac-
cordance with religious law. Dirar was a judge, and
the categories he employed were not those of a
philosopher. Nevertheless, it was some time before
they came to be considered inadequate. True, Abu
108
~
THEOLOGY AND SCIENCE
'1-Hudhayl was quicker to realize that there was
something that characterized the human being, be-
yond the bundle of phenomena that confronts us
everywhere: there was also life. He thought about the
human soul, and he may have seen the problem. But
though the sources are even more imprecise here
than usual, there is no doubt that he believed that
the soul dies with the body. And when he found him-
self obliged to define the human being, he confined
himself to such expressions as "the living body (jasad)
that eats and drinks" or "the figure (shakhs) that has
two hands and two feet."
The discovery of the immortal soul in Muslim the-
ology is attributable to al-Nazzam (and in a sense
to a Shiite theologian under whose influence he fell,
Hisham b. al-Hakam). Al-Nazzam introduced the no-
tion of rah, or pneuma, which he included am9ng the
interpenetrating "bodies" that combine to form the
human being. And he proved man's immortality with
arguments Plato had developed in his Phaedrus. For
him, life results from pneuma, and man and pneuma
are, strictly speaking, identical. The body is no longer
anything more than an envelope of flesh, corrupt-
ible and perishable, a h a y k a ~ a Gehduse. The atomists,
conversely, never found so clear a term to describe
109
I
I
THE FLOWERING OF MUSLIM THEOLOGY
human singularity. They were convinced that the var-
ious parts of the human body do not act indepen-
dently; rather, as they said, these form a totality, a
specifically structured set (jumla) in relation to which
life, knowledge, action, and so forth, present them-
selves as accidents or conditions-states ( ahwal), to
use Abu Hashim's terminology. But fundamentally,
the vocabulary remained inadequate because the sys-
tem could not accommodate such complexity.
39
Nevertheless, the majority of Mu'tazilite theolo-
gians turned away from al-Nazzam's ideas. This rejec-
tion is somewhat surprising, for from a modern per-
spective they were not without merit and can even
be considered more advanced. But controversies once
bore a different emphasis. Al-Nazzam did not pro-
vide a satisfying concept to explain creation. It
seemed that the elements that, according to him,
combine in existing objects had been created all at
once and were subsequently simply hooked together
indefinitely, as in a rail yard.
40
His theory of motion
offered only one original idea, the "jump." But the
fundamental axiom, that of motion as something
continuous, could be demonstrated only after the in-
vention of infinitesimal calculus by Pascal and Leib-
niz.41 The pneuma hypothesis ran the risk of leading
IIO
THEOLOGY AND SCIENCE
to the doctrine of a universal soul; in fact, certain dis-
ciples of al-Nazzam believed in metempsychosis.42 At-
omism, despite the problems specific to it, seemed to
be the better alternative.
But at least in the long run, those who embraced
it had to take into consideration the growth of phil-
osophical and doxographical knowledge. Since the
time oflshaq b. Hunayn (d. AH 289), a translation of
Aristotle's Physics had been available, and fragments
of Epicurus's philosophy were known, in the Arabic
version of Pseudo-Plutarch's Placita philosophorum.
That additional information, a new burst of accul-
turation, so to speak, seems to have turned theology
in the direction of physics, toward the real aspects of
the theory. It was only then that people began to
reflect that an atomistic world necessitates the ex-
istence of a vacuum. Abu '1-Hudhayl had noli men-
tioned its existence, nor had al-Nazzam. This is odd,
for Democritus had raised the question from the
outset. In fact, it is difficult to understand how at-
oms can move in a discrete space if no compartments
are unoccupied. But the originality of the theolo-
gians had consisted precisely in separating atomism
from the operation of this sublunary world. Where
the Greeks had spoken of physis, nature, the theolo-
II I
THE FLOWERING OF MUSLIM THEOLOGY
gians always spoke of God. Given that they already
had a knowledge of Aristotle's thought-al-Nazzam
claimed to have refuted it-they might have referred
to the fact that Aristotle had seen no need to pos-
tulate anything like a vacuum. The Stoics had be-
lieved in the existence of a vacuum, but they imag-
ined it only outside the universe. A few mutakallimun
thought that way as well. But a vacuum outside the
universe has no importance for the theory of motion.
A vacuum within the world is a more serious matter;
why accept something nonexistent and incommen-
surable, in the order of the universe or in God's cre-
ation? Strato of Lampsacus had already demon-
strated that if a stone is thrown into a receptacle
filled with water, the stone merely changes places
with the water. Instead of occupying an empty place,
the stone puts itself in the place of another material.
Abu '1-Hudhayl seems to have assumed that where
there is no object, space is filled with air. It is the air
in a room that prevents the walls from moving to-
gether and joining. As Abu '1-Hudhayl said, air is
the place (makan) for bodies. In addition, he did not
speak of the space in which a particular object is lo-
cated, but of a place where it is found, and that place
is only an accident of the object itself It is in the
same spirit that Abu '1-Qasim al-Balkhi began to do
112
THEOLOGY AND SCIENCE
physics experiments similar to those done by Strato
more than a millennium earlier.
It was precisely these experiments that excited cu-
riosity and criticism. Abu Hashim observed that it
is impossible for animals to live in a deep well and
explained that fact by the absence of air, that is, by a
vacuum. Others were inspired by passages from the
Physics in which Aristotle mentioned identical experi-
ments, which he attributed to Anaxagoras. Hero of
Alexandria also exerted an influence. Gradually, views
changed. Qadi 'Abd al-Jabbar, a Mu'tazilite master at
the turn of the fifth century AH (eleventh century CE),
considered the existence of a vacuum indispensable,
and his disciple Abu Rashid proved it, in opposition
to the Baghdad "naturalists" (ashab al-tabaY)> an am-
biguous term that seems to conceal nothing less
than the Abu '1-Qasim al-Balkhi school, to.,which
Abu Rashid had previously belonged. The so-called
Basrian school that he later joined managed to ex-
tend its influence beyond the Mu'tazilites, and the
existence of a vacuum was accepted by the majority
of the Ash'arites.43
Atomistic theory had reached its apogee. In the
following periods, its contours gradually blurred. We
do not know when atomism died out. In the long
run, the theologians, facing growing antipathy,
113
THE FLOWERING OF MUSLIM THEOLOGY
tended to dispense with the preliminaries of physics.
Simultaneously, the enormous success of Ibn Sina's
philosophy offered an invitation to adopt a new
infrastructure. Ibn Sina (Avicenna) did not like atom-
ism; he refuted it in a long chapter of his Physics. Ulti-
mately, he did not put much store in the mutakal-
limun. In that respect, he saw no difference between
al-Nazzam and his colleagues. Yet that general rejec-
tion was somewhat overstated; those who wanted to
remain mutakallimun continued to reflect on al-
Nazzam's arguments. Fakhr al-Din al-Razi, though a
good commentator on Ibn Sina, enumerated them
all, only to refute them later in one of his philosophi-
cal works, the Mabahith al-mashriqiyya. Al-Biruni, a
contemporary oflbn Sina-he was seven years his se-
nior-proved even more skeptical toward the new
wave. He was familiar with Indian atomism and won-
dered somewhat caustically, in his famous correspon-
dence with his young colleague, why Aristotle had
treated the atomists so superciliously. Even in the
Mongol period, certain theologians still followed the
old system; the Shiite Maytham b. 'Ali al-Bahrani is a
good example.
44
That is why the old ideas were still
visible in a different garb. Baha' ul-din Valad, father
of Jalal ul-Din al-Rumi (Rumi), used atomistic con-
cepts to convey his mystical psychology.45 It is uncle-
II4
THEOLOGY AND SCIENCE
niable that the model developed by Abu '1-Hudhayl
was much more resistant to the passage of time than
the systems that exist today. I am not altogether sure
that that is a merit; but for what it is worth, the fact
deserves at least to be pointed out.
..
I I 5
l
4
THEOLOGY AND HUMAN REALITY
Historical Images and Political Ideas
BoTH THE REFORMIST and the fundamentalist cur-
rents of modern Islam take their inspiration from a
vision of history that favors the beginning over the
end, the past over the future.
1
Such a view unques-
tionably posits a utopia of the ideal beginning, so
to speak. That sort of backward-looking utopian
thought is fairly common. In nineteenth-century Eu-
rope, it took the form of nationalism; there .too, a
mythical past was constructed in an effort to forge
an identity, and that mythical past was reconstituted
through a slanted reading of the historical texts. We
know, in fact, that utopias are never completely alien
to reality. The problem stems from the fact that
reality is violated by the interpretation imposed by
the utopian view. For Muslims, a further element has
been added-namely, revelation, which marks the be-
ginning of historical reality and therefore forms an
117
THE FLOWERING OF MUSLIM THEOLOGY
indelible part of the utopia. The divine message has
been adapted with unexpected success to the contin-
gencies of this world. It is perfectly logical to appre-
hend the Prophet as the founder of his community,
who in his capacity as lawmaker created a new soci-
ety. And it goes without saying that one can infer
from this understanding the imperative to rebuild
today's society in accordance with categories devel-
oped at the time. But it must also be recognized that
that is not the only way to interpret the Prophet's
role. True, the early generations oflslam knew, just as
we do, that the new revelation had brought about a
fundamental change. Compared with the pre-Islamic
period, the jahiliyya) Islam marked a radical turning
point. But the Prophet was considered a spiritual
guide above all, the model of perfect behavior at the
ritual and moral level. That was the main content of
the sunna. In the political realm at the time, it was
instead the first caliphs who staked out the future,
through their military campaigns beyond the Ara-
bian Peninsula and their victories over the two great
empires of the Old World, but also through their in-
ternal quarrels, the schism of the first civil war. It was
then that the community definitively distanced itself
from eschatological ideas and exchanged the ideal of
IIS
THEOLOGY AND HUMAN REALITY
Bedouin freedom for the concept of a structured gov-
ernment-a "state," if you like.
In principle, the era of the Prophet's successors oc-
cupied center stage, historically speaking: no one was
looking backward just yet. The feeling that some-
thing had been lost, that a break had occurred mark-
ing the end of an ideal era, a golden age, so to speak,
developed only gradually and eventually took shape
with the concept of the "rightly guided" caliphs, al-
khulafa) al-rashidun. That concept was reinforced by
the canonical number four. It did not appear for a
long time-not before the first half of the third cen-
tury, when Ahmad b. Hanbal was one of the first to
defend it, and even then only at the end of his life.
Why four, precisely, and not two or three? The Shiites
never accepted that construct. One's view of history
is always selective; it contains a certain elemeQt of ar-
bitrariness. Therefore, we cannot refrain from asking:
What else could have been incorporated into this
view of history, and why was the exemplary reality we
find in books (and, alas, only in books) finally de-
cided on? Three of the first four caliphs were assas-
sinated, so the reality was hardly as idyllic as the
schema supposed. But these murders were nearly for-
gotten, with the exception, perhaps, of'Uthman's. As
I I 9
THE FLOWERING OF MUSLIM THEOLOGY
for 'Umar's, Muslims exonerated themselves on the
pretext he had been killed by a non-Muslim and had
thus died a martyr. The same was true for 'Ali: al-
though it was not an unbeliever who had murdered
him, it was surely a heretic, a Kharijite.
'Uthman, by contrast, had been killed by true be-
lievers. In addition, his successor, 'Ali, was suspected
of having had a hand in the affair, or at the very least
of having benefited from it. In the long run, it was
not the murder as such that was so shocking, but
the resulting schism. The unity of the community
collapsed; henceforth, there were the followers of
'Uthman, the future Sunnis, and the followers of'Ali,
the future Shiites. Theoretically, it should not have
happened that way, for division, in the Koran, was
imputed to infidels, especially Christians. They had
destroyed one another through their theological dis-
putes. Islam, conversely, had reconstituted the pri-
mordial revelation. That development was a sign of
divine mercy, said the Koran. Why, then, was there
discord in the Muslim ranks, why were there division
and bloodshed that would be perpetuated in the gen-
erations to come? God's commandment had clearly
been broken, and the guilty party had to be found.
People tried to justify themselves, or they blamed
themselves; and so historiography came into being.
120
THEOLOGY AND HUMAN REALITY
They wondered how good Muslims ought to have
acted, and so political theory began. They constantly
evaluated events by religious criteria, and so theology
entered the picture. They spoke not of misconduct
but of sin, and, as it happened, those who had killed
one another were the Companions of the Prophet,
and as a result the very ones who would become
models for later generations.
Such is the constellation I would like to take as my
starting point: the explosive mixture of historiogra-
phy, theology, and political thought. The chronology
is especially important. These three dimensions ap-
pear at the same time, and so early that they are part
of the search for identity that was a driving force in
the early generations and continues to be so today.
That explains the efficacy of these ideas, an4 why
people are still reinterpreting that past and rereading
the texts that refer to it. These repeated rereadings
also determined the fate of the sources available to
us. Many interpretations soon lost their value, and
when they were no longer in fashion, the texts in
which they had been expressed no longer got copied.
The loss was enormous. In the field of historiography
we sometimes have the opportunity to reconstruct a
lost work; for political theory, we have only doxog-
121
THE FLOWERING OF MUSLIM THEOLOGY
raphy.2 The summaries we find are briefand inade-
quate, and a great deal of imagination must be mar-
shaled to restore their original plasticity.
The schism has continued into our own time. Ecu-
menical efforts began very early, but they all failed.
Nevertheless, they influenced the shape of history
and in turn produced schools of thought, sects, as
they are generally called, alongside the two major
faiths. Those who wanted to reconcile the camps
were all of the opinion that it was no longer possible
to impute the sin to a single guilty party or to a single
group. In one way or another, they wanted to forget
the whole affair. As a consequence, the Murji'ites,
or "postponers," ultimately recommended deferring
judgment; hence the origin of their name. At first,
the doctrine of i r j a ~ postponement, was formulated
in relation to 'Ali and 'Uthman. The Murji'ites em-
phasized that no one knew any longer which of the
two had committed the "sin." Initially, then, that
view was a call for political moderation.
3
Later, when
even the Murji'ite current was no longer in a position
to remain neutral, the political principle was trans-
formed into a theological doctrine: No one must call
the faith of a coreligionist into question: God alone
would decide who would be punished and who
122
THEOLOGY AND HUMAN REALITY
would escape punishment. In this world, all Muslims
remain united by Islam.4
That position, which dates from the last quarter of
the first century AH, integrated the lesson of the first
two civil wars. Muslims were weary and hoped to re-
establish peace by declaring a general amnesty. What
followed was the opposite, a third fitna) fiercer and
more widespread than the two previous ones. It cul-
minated in the Abbasid revolution. In the climate of
general confusion that reigned during the last decade
before the advent of the new dynasty, a new group of
theologians arose and tackled the same problem,
though in a much more disillusioned manner. These
were the Mu'tazilites. They addressed the contempo-
rary situation, but spoke of it only indirectly. Instead,
they evoked the earlier era-on which the Murji'ites
had already taken a position-except that now they
emphasized the dimension of sin much "more
strongly than the Murji'ites had. Sin had destroyed
the 'adala) that is, the integrity and reliability of
those responsible for the battles during the first civil
war. As witnesses, therefore, all of them were compro-
mised. But since we do not know which of them was
wrong, the situation resembled a case of adultery in
which the partners engage in mutual repudiation
123
THE FLOWERING OF MUSLIM THEOLOGY
(li'an). The marriage is voided, but the question of
guilt remains unresolved. The testimony of both
partners continues to be accepted, except when they
testify against each other. The argument had thus
taken a legal turn, which is characteristic of the
kalam in general and is not as surprising as one
might think. Nevertheless, we might wonder why
they were so preoccupied with the testimony of peo-
ple who had died two or three generations earlier. It
was because they were thinking of hadith they had
transmitted. It was by reference to those traditions
that the schism was usually justified and reinforced.
The 'adala in question is thus integrity as conceived
by specialists in hadith.
5
In a sense, the ashab al-had'ith learned their lesson.
They developed criteria to exclude overly one-sided
traditions. But they applied these criteria only to
others, "heretics." Among themselves, they did not
manage to differentiate hadith from historiography.
What emerged was the concept of the four "rightly
guided" caliphs, in which the 'adala of the first ca-
liphs was almost frozen. In the long run, even the
Mu'tazilites embraced the concept. At first, only
three caliphs from the first generation were recog-
nized, and in .. fact it was not always the same three:
Abu Bakr, 'Umar, and 'Uthman in Medina and Basra,
124
THEOLOGY AND HUMAN REALITY
but Abu Bakr, 'Umar, and 'Ali in al-Kufa. In the late
second century AH, a few muhaddithun from al-Kufa
made a concession, accepting Abu Bakr, 'Umar, 'Ali,
and with some reservation 'Uthman. After all, 'Uth-
man had been one of the Companions (sahaba) or,
more precisely, one of the 'ashara al-mubashshara) the
ten the Prophet had promised would go to paradise.
Other muhaddithun then felt impelled to reestablish
the historical sequence-that is, to reverse the places
of 'Ali and 'Uthman, in privileging 'Uthman and
finally elevating him to the same rank as the others.
The glorification of the four khulafa) rashidun had
thus begun; it coincided with the canonization of the
Companions.6
But what are we to make of the Battle of the Camel
and the Battle of Siffin? After all, those who fought
during these crises were Companions. A
was made claiming that those responsible for the
murder of'Uthman and for the battles that followed
were not the most highly regarded Companions, but
people on a lower plane, Bedouins or Shiite extrem-
ists, who were called the Saba)iyya at the time. Sayfb.
'Umar was one of those who propagated that histori-
cal myth.7 But the man who made it a truly scientific
doctrine was the Mu'tazilite Hisham al-Fuwati, a
disciple of Abu '1-Hudhayl who lived during al-
125
THE FLOWERING OF MUSLIM THEOLOGY
Ma'mun's reign. For him, the Battle of the Camel was
only a misunderstanding, a disaster caused by certain
unreliable elements and not by protagonists such as
Talha or Zubayr.
8
It is possible that Hisham al-
Fuwati was the sort of man who closed his eyes to
reality. But the ashab al-hadtth also began to favor a
solution of that kind. Walid b. Aban al-Karabisi, a
theologian from the Iraqi town of Wasit who was
a contemporary of Hisham al-Fuwati but not a
Mu'tazilite, thought that 'Ali and his opponents, in
joining the battle, were only obeying their ijtihad,
their independent reasoning. And in the ijtihad there
is always room for an error that goes unpunished be-
cause kull mujtahid mus'ib: Whoever forms a reasoned
opinion is right.
9
That legal variant was later ac-
cepted, albeit with a few modifications, by many
theologians from the school of Ibn Kullab and al-
Ash'ari, Shafi'ites especially, but also by Malikites
such as al-Baqillani.lo There was a universal need to
put the past to rest.
Hisham al-Fuwati's position was certainly ideologi-
cal. To shore it up, it was necessary to manipulate
the sources, but the position had the advantage of
neutralizing a discussion that had become fruitless.
Hence, people could turn their attention to other
126
THEOLOGY AND HUMAN REALITY
subjects, to questions dealing with process pure and
simple-succession, for example, or the political sys-
tem as such. There were many problems to settle:
Who would become caliph? Should he be elected,
and if so, by what means? Could the power conferred
on him later be taken away-could a caliph be de-
posed? Was government power, the state, an absolute
necessity after all? Al-Mawardi was by no means the
first to provide answers to these questions. From the
beginning, theologians and jurists debated them at
length.
As always and everywhere, however, care was taken
to differentiate between theory and practice. Theory
develops freely and takes its time; it usually comes
fairly late. Practice has to be decisive; it creates reali-
ties. On the question of taking power, the practice of
the "rightly guided" caliphs was far from uniform.
Abu Bakr had been named by acclamation, but that
acclamation was not planned in advance, it was ar-
ranged somewhat unexpectedly-foltatan, as 'Umar
later declared: without a regular procedure. As for
'Umar, he was designated by his predecessor; he thus
owed his power to Abu Bakr's will. 'Uthman was
elected by a council, a six-person shura, not by the
people, as is sometimes now said, but by an electoral
committee of which he was a member. 'Ali's position
!27
THE FLOWERING OF MUSLIM THEOLOGY
as caliph was approved by the bay'a> the oath of alle-
giance taken by community leaders, except that, in
his case, that act of allegiance was not unanimous.
One entire group had backed out of it. He and his
followers did not attach a great deal of importance
to that. They believed that power belonged to the
Prophet's family and that those who did not bow to
it were simply wrong. Fundamentally, those who were
their enemies and who defeated them, the Umayyads
and the Abbasids, were of the same opinion: as much
as possible, they kept power within their own fami-
lies. They all shared the conviction that power was in-
herited or given by God. For them, election was not
an act, but an innate quality.
For 'Ali's followers, the future Shiites, that would
soon become pure theory. After 'Ali was assassinated,
they did not regain power until the advent of the
Fatimids in North Africa and later in Egypt. In real
terms, the major problem confronting their model
was that 'Ali had many descendants. Rifts among
the different branches were almost inevitable and in
fact occurred immediately. In theoretical terms, their
model collided with the fact that 'Ali had not been
the first caliph, but the fourth. His followers, since
they could not change history, changed the percep-
tion of it, the Geschichtsbild> to use the German expres-
128
THEOLOGY AND HUMAN REALITY
sion. From that standpoint, 'Uthman's case was
clear: According to them, he was a good-for-nothing.
But how were they to judge Abu Bakr and 'Umar?
The Shiites began to write treatises about the day
Abu Bakr was named caliph,yawm al-saqifa> and they
wondered how someone like him, who had no claim
to the caliphate, could have been legitimately elected.
The diversity of opinion was great, ranging from
those who considered the election an excusable error
to those ~ h o saw it as the result of a conspiracy. At
one point, the theologians introduced a distinction
between the rule of the "righteous" man, that is, the
one most fit to govern because of his personal merit,
imdmat alfadil> and the rule of the man of inferior
merit, imdmat al-mafdul> who had nevertheless been
elected by a majority. As always, the schema was ap-
plied only to the exemplary past, the classical jlge of
the first caliphs. Abu Bakr was oflesser merit, mafdul>
but he still had the right to be caliph. Seen in that
light, the schema already had a Shiite bias: Abu Bakr
was of lesser quality, inferior in merit, compared to
'Ali. The two terms may have been developed by the
Mu'tazilites, but those among them who used the
terms, Bishr b. al-Mu'tamir for example, belonged to
the Zaydi current, which in this case means moder-
ate Shiites who accepted the government of the
129
THE FLOWERING OF MUSLIM THEOLOGY
Abbasids.ll Presented in that way, the schema can be
seen as the product of a compromise.
As is only natural, opposition came from both par-
ties to the compromise. It came first from those who
considered Abu Bakr the preeminent caliph. They
were in Basra, a city that had always supported
'Uthman. But it would also come from Shiite extrem-
ists who completely rejected Abu Bakr, and 'Umar
along with him. For them, both those caliphs, like
'Uthman, had been usurpers. That idea was especially
widespread in al-Kufa; those who embraced it were
called the rawafid.12 At first, they were merely one
group among others. But they soon came to repre-
sent the majority, and their attitude was adopted by
those who began to be called Imamis (Imamiyya), or
later Twelvers (Ithna 'ashariyya). As the theologians
argued over the respective qualifications of the candi-
dates, they compared them with one another. To es-
tablish criteria, they developed catalogues of virtues.
It was then that for the first time they spoke of the
qualities necessary or desirable in a ruler. That is
what Bishr b. al-Mu'tamir did, for example, introduc-
ing a theme that would be developed later by such
Sunni authors as al-Mawardi. The Rafidites took a
different path; they did not like shades of gray. For
I30
THEOLOGY AND HUMAN REALITY
them, it was not a matter of combining qualities, of
whatever kind. If the power to rule-the imamate as
it was called at the time-was passed down through a
last will and testament or statement (nass), the can-
didate's qualifications were secondary. In addition,
there was no reason to consider the Companions
of the Prophet examples of excellence. From the
Rafidite perspective, they were, with rare exceptions,
all discredited by the fact that they had collaborated
in the plot that allowed Abu Bakr and 'Umar, who
were not part of the Prophet's family, to assume
power at 'Ali's expense. Truth was not a monopoly
held by the majority; the Shiites never accepted the
principle of ijma' practiced by the Sunnis.
What the Rafidites had was only a theory of refusal.
But there were other resistance movements tha' had
nothing to do with the Shiites. They had their own
ideology, based on different arguments. They relied
on the Koran much more than the Shiites did. They
liked to lay claim to the "Book of God and of His
Prophet's sunna," and what that motto expressed
was their desire to bring true Islam into being. They
had a tendency to identifY their propaganda with the
obligation to command the good and to forbid evil,
I 3 I
THE FLOWERING OF MUSLIM THEOLOGY
al-amr bi'l-ma'ruf wa'l-nahy 'an al-munkar, a principle
also taken from the Koran. That was naturally all a
matter of exegesis; the obligation to command the
good lent itself to different interpretations. The
Mu'tazilites, for example, appropriated the motto in
their formative phase, when they were still in a po-
sition to oppose the government. Later, when they
were admitted to the courts of the Abbasid caliphs,
they hastened to give it a different meaning. Indeed,
the government also had an interest in "command-
ing the good and forbidding evil," and in doing so in
a way consistent with the Book of God and of his
Prophet's sunna. The problem had a long history.
Traces can already be found in hadith of a discussion
by scholars on that subject. How does one command
the good? By the sword (bi'l-sayf)) by exhortation (bi'l-
lisdn)) or simply by internal disapproval (bi'l-qalb)?
Who has the right to act, and what action should be
performed? The responses varied, but they date from
as early as the Umayyad dynasty.
13
When Muslims did take up arms, a verse from the
Koran entered the picture, shifting the focus. It is
surat al-Hujurat 9: "If two parties ofbelievers take up
arms the one against the other, make peace between
them. If either of them commits aggression against
132
THEOLOGY AND HUMAN REALITY
the other, fight against the aggressors till they sub-
mit to God's judgment. When they submit, make
peace between them in equity and justice." Of course,
it was not always possible to know which was the ji'a
al-baghiya) the aggressor or rebellious party; regard-
less, the verse recommended reconciliation, resolu-
tion of the conflict, and not the militant pursuit
of confrontation until the brutal outcome. Revolt
therefore remained an irregularity, an exception,
which the whole community had a duty to prevent.
Even rebels who thought they were fighting for a just
cause were in the wrong if they did not agree to nego-
tiate or if the other party, generally the government,
seemed to be making concessions. The discussion
was broached with the 'Ali case. As a legitimate ca-
liph, he had been able to treat his adversaries at the
Battle of the Camel as a dissident group, the ~ a al-
baghiya. In Siffin, he again had right on his side. Ac-
cording to a well-known hadith, the Prophet had pre-
dicted to 'Ammar b. Yasir that he would be killed
by the ji)a al-baghiya) that is, by Mu'awiya's troops.
But 'Ali had also agreed to arbitration, so that peace
could be reestablished. The peace that followed was
not altogether the peace he had wanted, and it was
ultimately Mu'awiya who prevailed. Nevertheless, the
133
THE FLOWERING OF MUSLIM THEOLOGY
majority view was that 'Ali had made a good decision.
In the long run, his religious prestige was greater
than Mu'awiya's, even among the Sunnis.
For theological discussion, the bughat (rebels) par
excellence were the Kharijites. They too had been re-
pressed by 'Ali, but they had revolted again after his
death and had sown terror with fanaticism and bru-
tality. In the eyes of the middle class in Iraqi cities,
they were extremists that the government was right
to combat. The line between resistance and terror-
ism was increasingly blurred. Al-Asamm, a second-
century Basrian Mu'tazilite, taught, moreover, that
the ruler could not demand his subjects' loyalty
against the ahl al-baghy unless he himself was "just";
the ahl al-baghy could be opposed only by the ahl al-
'adl. After all, the verse in surat al-Hujurat ends with
the sentence "Make peace between them in equity
and justice; God loves those who exercise justice"
("fa-aslihu baynahuma bi'l'adl wa-aqsitu. Inna llaha
yuhibbu '1-muqsitina," 49:9). But in the Fiqh absat, a
Murji'ite treatise written by an Iranian Hanafite ju-
rist in Balkh, the dilemma was settled in a completely
different way. Subjects were obliged to persevere in
their loyalty toward the government, even when that
government committed injustices against the ahl al-
baghy. 'Abdallah b. al-Mubarak (d. 797) defined the
134
l
I
THEOLOGY AND HUMAN REALITY
ahl al-sunna-he was one of the first to use the term in
a technical sense-as those who rejected insurrection
in all cases and who prayed behind the caliph's repre-
sentative, whether or not he was devout. Jurists were
no longer concerned with the question of which was
the fi'a al-baghiya in a particular case; the bughat were
simply the rebels. The government had the monop-
oly on power.
All this means in essence that the theoretical basis
of a right to resistance was rather weak. Of course,
there was the hadith "No obedience to a created
being with respect to sin!" (la ta'ata li-makhluqin fi
ma'siyati 'l-khaliq). But in his Risala fi 'l-sahaba, Ibn al-
Muqaffa' had already warned the caliph against a
false exegesis of it. The Qadarites had used that say-
ing of the Prophet; their propaganda had made it
a fixed part of collective memory. Indeed, the pmay-
yads had claimed that their governmental power had
been a gift of God, a rizq, predetermined and irre-
versible, and the Qadarites had protested, saying that
power was worthless as long as one did not prove
worthy of it. Their attitude was noble and justifiable,
but it was linked to the heresy of free will. Even the
militant fundamentalists of our own time prefer to
define their resistance against the state as a resistance
against the infidel. For them, it is not an intracom-
135
THE FLOWERING OF MUSLIM THEOLOGY
munity rebellion, a rebellion of Muslims against
other Muslims, but a war, a jihad against the kuffar,
infidels merely claiming to be Muslims. One does not
make war against one's coreligionistsJ4
Yet power came ultimately from God and not from
the ruler. The Koran had not singled out the Prophet
as head of the community; God had addressed believ-
ers as such through him. For a century, minted coins,
the most visible symbol of power, did not bear the
caliph's name. Until the reign of the Abbasid caliph
al-Mahdi, only religious mottos appeared on them.
Poets did tend to address the caliphs as khalifat Allah)
"vicar of God"; but theologians generally emphasized
that he was only the khalifat rasul Allah) the Prophet's
successor when it came to earthly authority. It was
the overarching importance of the community that
led Sunni theologians and jurists, including many
Mu'tazilites, to recommend the elective model even
though the political reality did not directly favor it.
One might think that they debated about the con-
sultation, the shara
1
a great deal in that context. But
such is not the case; we must not be misled by mod-
ern concerns. Of course, the principle of the shard
continued to play a role, but only under certain con-
ditions. The shura was not an "election" but a "con-
136
THEOLOGY AND HUMAN REALITY
sultation." Taken in that sense, the term could even
become a revolutionary postulate. The opponents of
Walid b. Yazid-that is, the Qadarites who supported
Yazid b. al-Walid-had used the term to justifY their
cause, and Jahm b. Safwan, the Iranian theologian,
had used it at the end of the Umayyad dynasty to at-
tack Nasr b. Sayyar, governor of Hisham b. 'Abd al-
Malik.15 All of them, in fact, wanted not only to par-
ticipate in the exercise of power but actually to hold
power. Although mentioned in the Koran, the shura
does not trace its origin to the sacred text; it was
rather a legacy of tribal society and was always prac-
ticed oligarchically. For that reason, the Mu'tazilite
al-Asamm, who had on the whole a great deal of sym-
pathy for the way the procedure had been applied by
'Umar, did not find it satisfactory in his own time.
This was under the Abbasids, and he demanded in-
stead the unanimous agreement of all Muslims, a
true to warrant the naming of a caliph.'6 We
are not sure he believed that such an ijma' had been
achieved in his time. But we know he believed that
unanimous agreement, once established, was irrevo-
cable. The by contrast, could on principle al-
ways be replaced by another consultation. In his en-
thronement speech, Yazid b. al-Walid promised the
Damascene Qadarites who had elected him that he
137
THE FLOWERING OF MUSLIM THEOLOGY
would resign if they no longer found him satisfac-
toryJ7 Fortunately, he died before that situation
arose. But with his caliphate the third civil war be-
gan, leading to the Abbasid revolution.
The Abbasids drew the necessary conclusion. The
concession made by Yazid b. al-Walid was never men-
tioned by them. Al-Asamm, who formulated his doc-
trine during the reign ofHarun al-Rashid, had to ar-
gue in an altogether different context, for the first
Abbasids used a Shiite vocabulary to define their le-
gitimacy. That is not surprising if we recall that it
was the alliance with the Shiites that facilitated their
revolution and guaranteed its success. Like the Shi-
ites, they thought in terms of succession and felt they
were legitimate heirs because, like them, they were ahl
people of the house (that is, of the family of
the Prophet). They even had an advantage over 'Ali;
'Abbas had been the Prophet's uncle, whereas 'Ali
owed his legitimacy to his wife. That particularity
was convenient for them when the alliance was bro-
ken, with the insurrection of al-Nafs al-Zakiyya un-
der al-Mansur. They could then claim that, according
to divine law, when there was no son, the uncle had a
greater right to the inheritance than the daughter.
Legally speaking, it was an ingenious theory; the Shi-
ites had some difficulty refuting it and as a result
138
THEOLOGY AND HUMAN REALITY
ultimately changed their entire legal system, giving
women many more advantages in their law of succes-
sion than the Sunnis did. In spite of that, the
Abbasids abandoned the theory fairly early, under
Harun al-Rashid. The advent of the Barmakids led to
a change of strategy. The Shiites had been defeated
on the battlefield and no longer constituted an im-
mediate danger. Given that situation, reasons of state
dictated a rapprochement with the Sunnis, the ma-
jority of the population not only in Iraq but every-
where in the empire. The jurists were thus free to re-
turn to the elective model. But in accordance with
their logic, they continued to understand election
primarily as a quality and not as an act. Instead of
asking who was going to elect the caliph, they won-
dered which of the candidates was eligible.
The majority said: "They are of the Quraysm tribe"
("al-a'imma min Quraysh").I
8
Abu Yusuf held that
view. He was the most eminent jurist in the court
of Harun al-Rashid, the first supreme judge, qadi 'l-
known to us. In supporting that thesis (he was
not its author), he limited the number of candidates
but did so realistically. The postulate that granted
the Quraysh the privilege of being the only Arabs to
assume power may already have been used in naming
Abu Bakr. In fact, all the caliphs, both the Umayyads
139
THE FLOWERING OF MUSLIM THEOLOGY
and the Abbasids, belonged to that tribe. But there
was also a minority opinion formulated by a con-
temporary of Abu Yusuf-namely, Dirar b. 'Amr, the
Mu'tazilite theologian mentioned earlier, who was
himself a practicing jurist. He was a qadi not at the
court or in the capital but in al-Kufa, and as a result
among a population that was far from favorable to-
ward the government. The objection formulated by
Dirar against the privilege of the Quraysh was based
on the observation that if irregularities or com-
plaints arose, it was difficult to depose a member of a
powerful clan. In saying that, Dirar was thinking of
'Uthman and the protests directed against him be-
cause of his ahdath, his "innovations." If one is to
avoid civil war, it is better to have a ruler who does
not come from an influential family. Fundamentally,
an Arab had no more right to the caliphate than a
NabatJ, that is, a peasant from the Iraqi countryside
who spoke Aramaic and did nothing but plow his
fields and pay his taxes.19 Of course, no one seriously
thought of installing a Nabataean caliph. That was
obviously a utopian idea. But Dirar was not wholly
isolated; his idea was adopted by Thumama b.
Ashras, and later even by al-Jahiz. He had predeces-
sors: the Iraqi Qadarites in Basra, though they ac-
cepted the prerogatives of the Quraysh, did so only
140
THEOLOGY AND HUMAN REALITY
on condition that the Quraysh prove worthy of them
by behaving in a fair and exemplary manner. That
relation to the Basrian Qadarite movement which
' '
through the figure of 'Amr b. 'Ubayd, lay at the root
of Mu'tazilism, explains why Dirar conceded at least
implicitly that the caliph could be deposed by the
community.
That was theory, as I have said. Dirar knew that for
caliphs and government officials, kuttab, the matter
was seen in a different light, and he was prepared to
accept it, for he was not a rebel. When he came to
speak of his own time, his tone changed. The ruler
guarantees the integration and harmony of society,
he said; that is why subjects owe him obedience.zo
That ideal of well-intentioned government, of the en-
lightened ruler, so to speak, was apparently devel-
oped under al-Ma'mun's reign in particular . .,Dirar's
Mu'tazilite successors, Abu '1-Hudhayl and al-Naz-
zam, seem to have offered additional nuances. Like
his predecessors, al-Ma'mun considered himself the
heir to the Prophet. In addition, he liked to play the
role of teacher, of pastor keeping watch over his flock
(ra'ryya)-that is, over the community of believers.
The Risalat al-Khamfs refers to him as the people's
guide, imam al-huda, inspired by divine instruction.
That risala, or treatise, appeared at a timely moment.
141
THE FLOWERING OF MUSLIM THEOLOGY
It was composed in AH 198, hence shortly after the
death of al-Ma'mun's brother al-Amin, and in the of-
fice of al-Fadl b. Sahl, who was al-Ma'mun's vizier in
Marv, Khorasan, where the caliph was residing at the
time. The ruler did not reign by force, therefore; he
influenced his subjects by virtue of his superior judg-
ment. It was precisely his qualifications as a teacher
that placed him just below the Prophet, in the posi-
tion of heir, so to speak. It seems that the Mu'tazilite
theologians who developed these ideas were influ-
enced by the Iranian tradition. Education and ex-
hortation were already central categories for Ibn al-
Muqaffa'; he expressed them in his Risala fi 'l-sahaba.
And the ta'dlb al-'amma, the duty to lead the people to
their salvation-the islah, as it was called-played a
role in the mirrors of princes that were translated
from Persian. Consider in particular the Testament
of Ardashir, 'Ahd Ardashlr; which al-Ma'mun chose as
a basic text for educating his nephew al-Wathiq.zl
The mihna soon showed how easy it was to abuse
that ideal to justify violence. After the experience of
the inquisition and persecution, the premise that
formed the basis for pedagogical optimism was more
clearly recognized: the conviction that the man in
the street cannot decide for himself The people-the
masses, we would now say-are rapacious and igno-
142
THEOLOGY AND HUMAN REALITY
rant, al-Jahiz tells us, and that is why prophets and
rulers are needed. In a similar vein, al-Qasim b.
Ibrahim, the Zaydi imam who was nearly a contem-
porary of al-Jahiz, said that if by chance people do
not express their rage against one another, it is not
because they restrain themselves, but because God
in his wisdom never leaves them without a teacher.
Bishr b. al-Mu'tamir compares them to wolves always
seeking their own advantage; this is close to the
theme of homo homini lupus familiar to us through
Thomas Hobbes. Bishr's pessimism can be explained
by his personal experience. He had tried to introduce
the Mu'tazilite doctrine to groups of ordinary peo-
ple. He had wanted to evangelize them, particularly
with popular poems, and his audience would not
hear of it.
22
Thus he was disappointed. The Mu'tazil-
ites thought they had a civilizing mission; they felt
II
obliged to fight the taqlld, conformity, of the masses
and, like many others, they failed. They had projected
their ideal onto the caliphs, and the caliphs too had
failed to live up to it. Both groups, the caliphs and
the court theologians, were soon disenchanted. "A
hundred years of tyranny are better than one day
of civil war": that was the only positive lesson the
masses were able to draw from all those efforts at
education. The maxim was mentioned by Ya'qubi, a
143
THE FLOWERING OF MUSLIM THEOLOGY
Shiite author who compromised with a Sunni regime
that espoused views opposed to his.
In fact, the gap between the government and the
masses widened. Instead of looking to the court, or-
dinary people increasingly relied on the ones who
had always been generous with the advice they
needed in their everyday affairs, namely, the ulema.
Al-Ma'mun, it seems, was the last Muslim ruler in the
classical age to have tried to form society to fit his
own ideas. After him, the government no longer felt
sure of its competency. In later generations, caliphs
such as al-Qadir, for example, or later, al-Nasir, did
take action, but they could do so only with the sup-
port of jurists. That did not prevent people from rec-
ognizing the authority of the government as such. In
general, it was accepted as a natural fact or as an in-
stitution desired by God. Even the Shiites decided to
collaborate or at least to adopt a quietist attitude.
But on the far left of the doctrinaire spectrum, so to
speak, there had always been the idea that the caliph
was wholly unnecessary because the people could
govern themselves.
At first, that was the attitude of the Kharijites in
particular. But the idea did not remain confined to
the periphery. In the late second century AH, it took
!44
THEOLOGY AND HUMAN REALITY
hold at the center of the community, following the
conflict between al-Amin and al-Ma'mun. During the
unrest that broke out in Baghdad after al-Amin's
death, when al-Ma'mun was still in Marv, the popula-
tion of the capital had had to take matters into their
own hands. Under the order of a certain Sahl b.
Salama, they had even formed a private army, a sort
of middle-class militia paid primarily by the mer-
chants, who were always the first victims of pillaging.
Sahl b. Salama seems to have been linked to the
Mu'tazilites.
23
In fact, the theoretical model corre-
sponding to that reality was presented by theologians
of the Mu'tazilite current. In utopian terms, a leader-
less community, an Islam without a caliph, could al-
ways be imagined, and utopia was gaining strength at
a time when disappointment with reality was spread-
ing. After the fall of the Umayyads, the Abbasfds had
claimed they were inaugurating a new era by means
of a dawla1 a turning point, a revolution in the original
sense of the term. But in the long run the question
remained whether they had really been successful.
Dirar had toyed with the idea of replacing the caliph
of the Quraysh tribe with a weaker candidate. Some
of his colleagues developed even more radical mod-
els. The thesis at the time was that the community
did not need a leader, an imam, at all, except in case
145
THE FLOWERING OF MUSLIM THEOLOGY
of emergency-during a war for example. For the
most part, it could do without one. Believers simply
had to abide by the Koran. Scripture offers enough
instruction to build a strict society. That position has
fundamentalist overtones; it implies that the com-
munity, the believers, are capable of interpreting the
Koran and that all of them would do so in the same
way. But that idea was formulated by a theologian, al-
Nazzam, who sought to establish ties with the court
of al-Ma'miln.24 It seems, therefore, that he hoped
that the community would confer power on the ones
who knew how to interpret scripture-that is, the
ulema.
In fact, he was not the first to propose that idea,
and it was not only the growing ambitions of his pro-
fession that pushed him in that direction. It would
be somewhat bold to maintain that he was still in-
spired by a tribal model in which the sheikh is the
primus inter pares; for if there was anyone in Bagh-
dad society at that time who represented the typi-
cal middle-class man and urban intellectual, it was
certainly al-Nazzam. Still, the model found support
among the Kharijites. The followers ofNajda b. 'Amir
believed that the community did not need an imam,
provided it followed the commandments of the Ko-
ran. In Iraq, it was the Ibadiyya in Basra that, with its
146
THEOLOGY AND HUMAN REALITY
senate, the jama'at al-muslimln; showed how a com-
mercial empire could be managed by a group of not-
ables who were both merchants and scholars. Al-
Nazzam's immediate source of inspiration, however,
seems to have been al-Asamm, a Mu'tazilite who lived
in Basra. As we have seen, he demanded that the ruler
be supported by the consensus, or i j m a ~ of the whole
community. He added that if absolutely necessary-
when that unanimous agreement could not be
achieved-the umma could do without a common
leader.zs
Al-Asamm also recognized that this model-a Pres-
byterian model, so to speak-suited small and local
communities rather than an empire such as the
Abbasid. But the conclusion he drew was true to
form: he recommended decentralization. He may
have done so because he had witnessed the Qftvision
of the empire under Hariln al-Rashid. The caliph's
decision had certainly been disputed. The harmful
consequences of such a policy-that is, the civil war
between al-Amin and al-Ma'miln-appeared only af-
ter al-Asamm was at the end of his career and close to
death. A century earlier, the Medinese jurist Sa'Id b.
al-Musayyab had also refused to swear an oath of al-
legiance, a bay'a) to two sons of'Abd al-Malik during
their father's lifetime. Al-Nazzam, for his part, lived a
147
THE FLOWERING OF MUSLIM THEOLOGY
generation after al-Asamm; he thus experienced the
unrest and anarchy that erupted in Baghdad after al-
Amin's murder. During those years, the jurists ad-
ministered justice without being officially recognized
by the state. In essence, people at the time were living
in a society without a caliph.
That unusual situation-organized anarchy, so to
speak-did not last long. The jurists apparently justi-
fied their independent decisions as the application of
the "obligation to command the good." Al-Ma'miin
put an end to that development after he returned to
Baghdad in AH 204. Basically, he had a great deal of
respect for the scholars; he invited them to his court
and liked to hold discussions with them.
26
But he did
not grant them any particular role in government af-
fairs. During the mihna they were forced to follow his
directives. It may be that at a certain point al-Wathiq
granted them the role they wanted, that of keeping
order. As the Mu'tazilite sources tell us, he may have
proposed adding religious counselors to his group of
officials, especially to those who were responsible for
collecting taxes. Hence, he may have tried to make
the kuttab and the ulema cooperate with each other.27
If that is not a legend, it was certainly a futile wish.
Over time, even the Mu'tazilites distanced them-
selves from the government. The tide had turned,
148
THEOLOGY AND HUMAN REALITY
and the ascetic element associated with their
thought since the beginning grew stronger.
That current reached its peak with a group that
one of the sources calls the sufiyyat al-Mu 'tazila. zs
They rejected not only the state in the sense of a cen-
tral government, but also the "world"-that is, com-
merce or any sort of profitable activity. The power-
money combination was suspect to them. This was
an urban movement. When these Mu'tazilite Sufis
spoke of profit, they were thinking of merchants and
tradespeople, perhaps also of artisans, but not of
peasants. Satan plants his banner in the marketplace,
according to a hadith. As a result, not only was the
caliph's role reduced to nothing but establishing or-
der; order as such was corrupt. The dar al-islam was
no longer a sound, intact world; mores had changed
so much that it could only be called dar al-kufr or dar
al-fisq. Power was merely the result of a usurpation.
At first, the Sufis were probably merely following
their w a r a ~ their fear of dealing with impure and du-
bious things. The state, al-sultan) was counted among
the shubuhat, suspect things. The Sufis did not accept
a salary from the government; they did not serve it as
soldiers; they even abstained from consuming food
from the princes' lands and from frequenting baths
built with government money on a usurped piece of
149
THE FLOWERING OF MUSLIM THEOLOGY
land.29 Those who had such scruples were not neces-
sarily Mu'tazilites. On the contrary: Ibn Hanbal tells
similar stories in his K. a l - W a r a ~ as does Muhasibi in
his K. al-Makasib. But it is also not possible to say that
the Mu'tazilites remained apart from these ascetic
currents. They were not the indifferent rationalists
we imagine them to be. Ja'far b. Harb, a Baghdad
Mu'tazilite who had served in the army, had a spiri-
tual crisis and went to be purified in the Tigris River.
Then he dressed in new clothes he had received from
the man who had converted him to his new life,
Murdar, also a Mu'tazilite. He refused his father's in-
heritance because the father had been a government
employee.3o After the mihna failed, and under the
reign of al-Mutawakkil, the Mu'tazilites also had to
adjust to seeing the government form pacts with
non-Mu'tazilite forces-hence, with those they con-
sidered heretics. As al-Jubba'I said in the late third
century, Baghdad and Egypt had become a land of
heresy, dar kufr; because no one could live there with-
out professing the eternal nature of the Koran and
determinism. He himself had left the capital and Iraq
to live in Iran, in 'Askar Mukram, where the majority
of the population still followed his doctrine.
Let me stop there. The attitude of the sufiyyat al-
Mu 'tazila had no future. That radical refusal led to an
ISO
THEOLOGY AND HUMAN REALITY
impasse and in Iraq Mu'tazilism in general ceased to
exert a consistent influence. The majority of Muslims
agreed that government (al-imam); despite its imper-
fection, was indispensable for the operation of soci-
ety. In the long run, absolute refusal was practiced
only by itinerant dervishes, qalandar.3l As far as the-
ory was concerned, the theologians soon lost the ini-
tiative to the philosophers. Thinkers such as al-
Farabi introduced completely new models. And over
time the new models were confronted with new expe-
riences: the weakening of the caliphate by the Buyids
and the Seljuks, which could be explained only by
a dualist model in which authority and power were
carefully distinguished; the annihilation of the ca-
liphate itself under the Mongols, the first period in
the eastern world when domination was seized from
Muslim hands completely; and finally the miltttary re-
gimes of the Mamluk period. It is easy to see that in
one way or another almost all these models have par-
allels in the contemporary world. But that is not my
subject here. The telling fact is that even today, when
people have hopes, they turn to the past, as the early
Mu'tazilite theologians once did. Islam began with a
great political success, and it remains forever tied to
that success.
I 5 I
5
THEOLOGY AND ITS PRINCIPLES
Hermeneutics and Epistemology
THE RELATIONSHIP-the contradiction-between
faith and knowledge has always preoccupied Chris-
tianity. One believes things one cannot prove. "Cre-
dibile est quia ineptum est," said Tertullian of the in-
carnation: it is believable because it is absurd. And he
added, "certum est quia impossible," it is certain be-
cause it is impossible. Of course, we must take into
consideration the fact that the Roman Church Fa-
.,
ther was a rhetorician and not a philosopher. The
famous credo quia absurdum does not appear in the
sources before Kierkegaard, who was the first to in-
troduce the cult of the unknowable, which was taken
up by Albert Camus and existentialism.
1
But it is
clear that where the incarnation was concerned,
Muslims always had the impression that their Chris-
tian brothers were clinging to an illusion. Christian-
ity speaks of the "mysteries" of faith; Islam has noth-
153
THE FLOWERING OF MUSLIM THEOLOGY
ing like that. For Saint Paul, reason belongs to the
realm of the "flesh"; for Muslims, reason, 'aql, has al-
ways been the chief faculty granted human beings by
God. Of course, this was not the independent rea-
son characteristic of the Enlightenment period, but
rather an intelligence subject to the will of God
and to the order established by him. Still, that divine
gift was accepted and appreciated everywhere, even
among ascetics and mystics. Al-Harith al-Muhasibi,
who supplies one of the first definitions of intelli-
gence, bases his psychological analysis on it, an analy-
sis that has earned the admiration of Sufis even in
our own time.
2
Only the application of intelligence
was subject to debate. Al-Muhasibi did not like dis-
putations (munazardt); conversely, they were the driv-
ing force of the Mu'tazilites' activities.
Let us acknowledge that some theologians
doubted the validity of the methods employed by rea-
son. Sometimes they went so far as to cultivate a sort
of irrationalism. The first to be catalogued in the
sources were Sufis-but also, surprisingly, Mu'tazil-
ites. They belonged to the sufiyyat al-Mu 'tazila I have
already mentioned, and their social criticism was di-
rected at the theologians themselves, at their intel-
lectual arrogance. This arrogance overlooked the
fact that fundamentally no theological speculations
154
THEOLOGY AND ITS PRINCIPLES
could surpass the simple faith of the masses. How
does one know what one knows? they asked, espe-
cially given that arguments for and against a particu-
lar assertion often contradict each other. All proofs
have the same value-this is the famous equivalence
among proofs (takafu' al-adilla), which Abu Hayyan
al-Tawhidi still refers to with some sympathy. The
term, and the practice associated with it, seem to
have originated in the isostheneia ton logon of Greek
skepticism, whose apogee was marked by the figure
of the "archheretic" Ibn al-RawandP But that wave
of anti-intellectualism was not representative. The
sufiyyat al-Mu 'tazila vanished without a trace, and Ibn
al-Rawandi, despite the emphatic rebuttal directed at
him over several generations, did not exert any influ-
ence. Essentially, Muslims had their own methods
and relied on those.
The Mu'tazilites were convinced that Wasil b. 'Ata'
had already developed a short "discourse on
method." A summary of it, consisting of only a few
sentences, was preserved by Qadi 'Abd al-Jabbar and,
with a few variants, by his contemporary Abu Hilal
al-'Askari in his Kitab al-Awa'il.
4
We cannot rule out
the possibility that the text is fiction, a projection de-
rived from a later position, and in certain places it
155
THE FLOWERING OF MUSLIM THEOLOGY
seems to have been reworked, but these revisions
themselves lead us to think that fundamentally it is
an authentic kernel of a discussion held in the early
days of Muslim theology.
5
Wasil begins with the cri-
terion for truth. In the first place, truth is truthful-
ness. It is presented in the form of trustworthy prop-
ositions. Any proposition is worthy of trust when it is
uttered by several people who could not have agreed
on it in advance. Mter that, one is obliged to weigh
the content of the sentence recognized as true; it may
be either general or particular. That distinction was
derived from the legal exegesis of the Koran, and
hence of a text for which the question of trust did
not arise, at least not for a Muslim. For that reason,
Wasil added a remark on abrogation ( naskh ), a proce-
dure that at the time concerned only scripture. His
text defines it in a way that would henceforth be con-
sidered a given and which might have already been in
use beforehand-namely, that only verses that are le-
gal in nature can be abrogated, not those which
speak of the world beyond or of the historical past.
6
The passage of this text that would later become the
most important comes at the end: it is a brief enu-
meration of sources of knowledge. Among these,
Wasil lists the Koran first, inasmuch as it is precise in
meaning (muhkam) and not ambiguous (mutashabihj
156
THEOLOGY AND ITS PRINCIPLES
compare sura 3:7). Next, he lists the propositions or
reports (akhbar) that according to the criterion men-
tioned above have the value of argument. And finally,
he lists "sound" reason-that is, judgments not based
simply on something given in advance but obtained
independently through personal reflection. Wasil
says simply, hi-'aql salim. He does not specifY the pro-
cedure as being "argument" (or "reflection"), nazar
(as later theologians would say), or ijtihad, indepen-
dent reasoning (as the jurists would have said). In the
enumeration in question, those two realms were not
yet separate.
The list-assuming it is in fact that old-influenced
both theologians and jurists. Among the Mu'tazil-
ites, for example, al-Jahiz borrowed from it and re-
formulated it; among the jurists, Shafi'I is our best
witness.
7
He never cites Wasil, of course; but when
Shafi'I came to Baghdad to make his career, he
could no longer ignore the Mu'tazilites, the ahl al-
kalam, as he called them in his treatises. The chief
disciple he acquired in the capital, Abu 'Abd al-
Rahman al-Shafi'I, was a student of Abu 'l-Hudhayl
and followed Mu'tazilite doctrine. As an employee of
Ibn Abi Duwad, he played an active role in the inqui-
sition, the mihna.
8
We cannot overlook Shafi'I, be-
cause the list marks the beginning of the concept of
157
THE FLOWERING OF MUSLIM THEOLOGY
usa! al-fiqh (the foundations of jurisprudence), which
he is generally considered to have originated. What
Wasil and Shafi'I had in common was that they disre-
garded consensus, ijma'; Shafi'I moved away from his
predecessor in his definition of khabar.
9
For Wasil,
khabar designates any proposition received from an-
other person, whereas Shafi'I, as a jurist, is interested
in the khabar only as hadith and sunna.l
0
Later, al-
Jahiz maintained a position between the two. He
knew that hadith was indispensable for jurists, but as
a Mu'tazilite he did not like it. In place of khabar;
which in the meantime had become too ambiguous,
he spoke of a "sunna accepted by all," al-sunna al-
mujma' 'alayha. Although he introduced the notion
of i j m a ~ consensus, with that expression, he did so
only verbally, without granting it the status of an in-
dependent notion.
In the end, theology used other criteria. The Koran
never constituted its central evidence. Muslims were
living in a pluralist society, and non-Muslims could
not be persuaded by quotations drawn from Koranic
revelation. Theology had an apologetic task, among
other roles, and that task could be performed only
through reason. And the way to deal with the Koran,
the way to apply the hermeneutics of commentary
158
THEOLOGY AND ITS PRINCIPLES
(tafs'ir) to it, was well known. Several Mu'tazilite theo-
logians wrote commentaries on the holy book, espe-
cially those belonging to the generation preceding al-
Ash'ari. Al-Jubba'i and Abu '1-Qasim al-Balkhi did so,
as did his contemporary Abu Muslim al-Isfahani,
but, even earlier, so did al-Asamm, and before him
'Amr b. 'Ubayd, who collected Hasan al-Basri's
courses in exegesis.l
1
From the beginning, the
method was realistic, sober, and entirely exempt from
the fanciful allegories of Origen, for example. The de-
sire was to reconstruct the historical situation of the
revelation, the asbab al-nuzal) or to take into account
the precise implications of a rule extracted from the
law. The practical approach can be explained, on one
hand, by the desire to organize an entire complex so-
ciety in accordance with the commandments of God
and, on the other, by the fact that Muslims hadwnly
one book of scripture and not two like the Chris-
tians. The Church Fathers were always confronted
with the problem of submitting the Old Testament
to the demands of the New. Muslims, by contrast,
though they respected the Bible, did not read it; it
was "abrogated" and set aside for the "people of the
Book," that is, for those who did not want to accept
the new version brought forth by the Prophet.
There were two words for what Muslims were do-
159
THE FLOWERING OF MUSLIM THEOLOGY
ing: tafsir and ta'wu. Both were in the Koran; scripture
had reflected upon itself.
12
Koranic metalanguage
was possible because the community knew the proce-
dure from the past. Tafsir was derived from the Ara-
maic pishra and the Hebrew pesher.
1
3 Ta'wil was a dif-
ferent case: the word was of Arabic origin but was
determined by its Koranic context. The line from
sura 37, where the notion is developed, says that "no
one knows its exegesis except God" [translation
modified]: that is, no one knows how to interpret the
ambiguous and complicated passages (mutashabihat)
previously mentioned. This could be understood as a
warning: Anyone trying to interpret them will fall
into heresy. As a result, the word acquired a negative
connotation: ta'wilat was equated with bida' (innova-
tions) or ahwa) (aberrations, vagaries).1
4
But that was
not always the case. Al-Maturidi called his commen-
tary of the Koran Ta'wilat ahl al-sunna (The interpreta-
tions of the people of the sunna). Later, interpreta-
tion was associated with the allegorical speculations
of people such as the Ismailis, who sought a hidden
meaning (batin) in the Koran. The Shiites always had
a certain predilection for exegetical "secrets"; they
were ill used by historical reality and sought to justify
their utopian ideas.
* * *
I60
THEOLOGY AND ITS PRINCIPLES
Hadith was different. The Mu'tazilites were scrip-
ruralists, like the Kharijites; in their view, the tradi-
tion of the Prophet could only introduce chaos into
the sacred text. In fact, hadith, the "oral Torah" (tara
shel be-pe)) so to speak, was not canonically fixed; it
was part of an oral tradition teeming with contradic-
tions. Dirar b. 'Amr, and al-Nazzam after him, col-
lected striking examples of them.ls It was Ibn
Qutayba who finally managed to resolve the contra-
dictions in his treatise Differences among the Hadith
(Ta'wu mukhtalif al-hadtth). Unlike the Koran, hadith
always suffered from a lack of reliability. The "tradi-
tion of the Prophet" (sunnat al-nabi) did enjoy enor-
mous success later on, but it owed its victory to its
pure and unavoidable necessity. The theologians,
who did not need the tradition as much as the ju-
rists, continued to require that the criteria of truth
II
be applied to it. As we have seen, Wasil had asked
that those reporting a saying not come to a prior
agreement with one another. Abu '1-Hudhayl had in-
troduced a numerical postulate: The number of wit-
nesses guaranteed certainty. Quantity was thus trans-
formed into quality. But he immediately recognized
the inadequacy of that rule. The number had to be
specified, and no one could do that without being
challenged. The solution could only be an act of will.
I 6 I
THE FLOWERING OF MUSLIM THEOLOGY
Abu '1-Hudhayl decided on twenty people. As evi-
dence, he simply modified a Koranic verse that spoke
of jihad rather than of knowledge: "If there are
twenty steadfast men among you, they shall vanquish
two hundred."
16
That analogy was extremely weak,
being founded on an arbitrarily chosen resemblance,
a qiyas al-shabah, as the jurists would have said. Above
all, a hadith cannot be transmitted by just any multi-
tude; it is a sacred text and belongs to the tradition of
the umma. The twenty people must therefore be Mus-
lims, "friends of God," or, as Abu '1-Hudhayl seems to
have said, "candidates for paradise" (min ahl al-janna).
We do not know whether he thought he would find
these "candidates" among his contemporaries or in
the generation of the Companions of the Prophet;
the sources are too vague to allow us to determine
that. But the profound skepticism in his words is ob-
vious. He spoke of certainty, and for him the only
things that were certain were truths such as the exis-
tence of God, prophecy, and the experience of the
senses. Words alone would never be capable of
achieving certainty; they could only be probable. Re-
garding that probability, Abu '1-Hudhayl was much
more generous: four people were sufficient. That
placed the question within the domain of jurists. He
162
THEOLOGY AND ITS PRINCIPLES
may have been thinking of the four witnesses re-
quired in cases of adultery.J7
The advocates of hadith could not be satisfied
with that arrangement. Many legal rules were based
on a hadith attested by a single chain of transmitters,
a khabar al-wahid. In the long run, those unique
hadith, ahad as they were called, became a problem.
The jurist 'Isa b. Aban (d. AH 221/836 cE), a disciple of
Shaybani and an influential man in the court of
Harlin al-Rashid, seems to have written the first
monograph on the question.
18
Shafi'I considered
them indispensable; he devoted a long chapter of his
Risala to them.
19
But even by the looser criteria of a
jurist, they were difficult to justifY; to be valid, testi-
mony normally had to be supplied by at least two
people. An additional criterion had to be agreed on,
therefore: the transmitter's integrity, or 'adala
11
The
Mu'tazilites accepted that criterion from the begin-
ning; Wasil had spoken of it. For them, however, it
was primarily a category of public life, regulating the
"cohabitation" of people in politics, for example. To
consider someone a person of integrity meant that
you trusted him. Nevertheless, despite the existence
of trust, verification was better. So it was that in place
of the witness's reliability, al-Nazzam proposed
163
THE FLOWERING OF MUSLIM THEOLOGY
verification by context, concomitant facts ( qara'in)
that lent support to the veracity of a report. In doing
so, he was still thinking within the framework estab-
lished by Wasil, for his examples had to do with news
of the day: for example, one heard of a neighbor's
death and also saw a coffin set out in front of his
house.zo Like many others, al-Nazzam did not yet
take the trouble to point out what was unique about
hadith. There, context was less important because
the reports regularly resurfaced over time along the
chain of transmission, the isnad. It was his disciple
Jahiz who first focused on hadith. In pursuing his
master's thinking, he added that the collective expe-
rience of a community accumulates in the tradition.
Unfortunately, that experience deteriorates as a re-
sult of transmission, and in the end God finds him-
self obliged to send a new messenger.
21
Al-Nazzam was the archetype of the rationalist, as
his reactions constantly show. He did not believe in
the existence of jinni, ahd he rejected the popular in-
terpretation of dreams and omens.
22
But he was also
the first to prove the prophethood of Muhammad by
predictions found in the Koran. He believed that the
Prophet was gifted with a miraculous knowledge of
the occult and of the future (ghayb ). Indeed, the
events the Prophet had foretold were borne out in ac-
164
THEOLOGY AND ITS PRINCIPLES
counts. Some of these were historical, such as the
short-lived triumph of the Byzantines over the
Sassanids alluded to in sura 30:1-3, whereas others
were literary, such as the stoning of devils by shoot-
ing stars or comets (sura 6?:5, 15:16-17, among other
verses), for which proof was sought in pre-Islamic po-
etry.23 In opening the door to miracles halfway, he set
off an avalanche. The next generation ofMu'tazilites
would concern themselves with countless narratives
on the subject, and not only those about Muham-
mad's exploits mentioned in the sacred text itself.
There were also many in Ibn Ishaq's Stra and in other
texts. 'Abbad b. Sulayman felt he could no longer ig-
nore all that material. Miracles had become the prin-
cipal tool for demonstrating Muhammad's truthful-
ness, just as they had always been for Jesus among
the Christians. 'Abbad claimed that these events. that
surpassed the grasp of reason were true because they
were sometimes recounted by people as irreproach-
able (ma'sum) as the Prophet himself. That argument
allowed him more confidently to adopt the postulate
of the people of paradise, ahl al-janna, formulated by
Abu 'l-HudhayP
4
The people to whom he was allud-
ing were probably the Companions. As already men-
tioned, his master, Hisham al-Fuwati (who had him-
self been a disciple of Abu '1-Hudhayl), took an
165
THE FLOWERING OF MUSLIM THEOLOGY
important step toward canonizing the Companions,
the sahaba. He did so in political theory but antici-
pated the consequences of that move for historiog-
raphy.
In essence, ijma' bore only a marginal relation to epis-
temology. In its epistemological manifestation, Aris-
totle included it within the "famous" or well-known
opinions accepted by everyone, the mashhurat in Ibn
Sina's later terminology.
25
In Islam, the concept origi-
nally belonged rather to the political arena. Ijmd' was
closer to consultation, shurd. When, according to a
text in Waki"s Akhbdr al-quddt, 'Umar b. 'Abd al-'Aziz
recommended that his Basrian governor 'Adi b. Artat
consult competent, discerning people in cases where
the Koran, the sunna, and the practice of caliphs no
longer provided any solution, 'Umar was speaking in
principle only of a consensus, i j m a ~ of local schol-
ars.26 As early as the pre-Islamic era, a tribal chief
could not make a decision simply on his own whim;
he was obliged to follow procedure by eliciting a con-
sensus. Al-Asamm, who was the first of the
Mu'tazilites to grant a key position to i j m d ~ recom-
mended it primarily as a political instrument to es-
tablish the validity of the oath of allegiance, or
bay'a.
27
166
THEOLOGY AND ITS PRINCIPLES
But just as they are in our own societies, political
decisions were open to judicial review, at least after
the fact. Practice had to submit to the test of theoret-
ical validity. It was thus integrated into a new con-
text-that, indeed, of epistemology. In jurisprudence,
ijma' more or less corresponded to the sunna under-
stood in the sense of "local custom"; we need only
compare the consensus of Medinese scholars in
Malik b. Anas's writings to be assured of that. It is
known in the same form in Shafi'I's writings, as we
have seen; Shafi'I considers ijmd' an additional confir-
mation and not an independent criterion. For him,
the word generally means only agreement among
specialists on the interpretation of a text, and this
text is often a sunna.
2
s Argument, therefore, always
depends on the sunna. Compared to the omnipres-
ent authority of one of the Prophet's sayings, l!ie au-
thority of an ijma' was rather limited. The opinion of
Shafi'I was shared outside Medina by many other an-
cient jurists-Awza'I, for example, or Abu Yusuf.29 For
the Mu'tazilites who did not value hadith, the situa-
tion appeared in a different light. For them, ijmd'
could replace the sunna rather than simply confirm
it. Dirar b. 'Amr, for example, considers it the only
criterion besides the Koran. In a sense, Dirar even
granted it a higher place than the Koran itself, for it
167
THE FLOWERING OF MUSLIM THEOLOGY
was not clear, given the ambiguity of scripture and
the problem of abrogation and mutashabihat, on
which passage of scripture a judgment ought, with-
out the unanimous support of the community, to be
based. People would simply set one auctoritas against
another, as they did in hadith.
30
Later, when the
Mu'tazilites had lost their battle against the Pro-
phetic tradition and had to accept it as a source, they
called ijmd' "the proposition of the community,"
khabar al-umma, as opposed to the proposition of the
Prophet, khabar al-nabi-that is, hadith.
31
We therefore arrive at a rather paradoxical result:
i j m d ~ which later became the Sunni principle par ex-
cellence, was first propagated absolutely and without
restrictions by those who would be considered here-
tics in later centuries-that is, by the Mu'tazilites.
The situation was complicated by the fact that the
parties opposing ijma' were also recruited from
groups that were unorthodox in their time, the
Kharijites and the Shiites. But these cases were differ-
ent; unlike the Mu'tazilites, they had always been ex-
cluded from the political coalition. In addition, the
Kharijites had severed themselves from the commu-
nity before it had even acquired a coherent shape. As
a result, they were more literalist than the Mu'tazil-
ites. They did not punish adultery with stoning, they
168
THEOLOGY AND ITS PRINCIPLES
did not practice mash 'ala 'l-khuffayn, the practice of
rubbing the shoes instead of the feet during ritual
cleansing, and they inexorably cut off thieves' hands
without recognizing the exceptions through which
jurists attenuated the rigor of the Koran. That, at
least, is what 'Abd al-Qahir al-Baghdadi said.
32
The
reality may have been more complex than it ap-
peared. They wanted to have nothing in common
with those they had abandoned. But they, of course,
established solidarity among themselves; there is rea-
son to believe that "the community of believers" of
Basrian Ibadites, the jama'at al-muslimin, proceeded
by consensus of a sort, though it was still on a politi-
cal level. As for the Shiites, they were disappointed
because they had been in the minority since the time
of the Companions. During the caliphate of Abu
Bakr and 'Umar, the consensus of the community
had not favored 'Ali. In jurisprudence, the Shiites al-
ways indicated their preference for ijtihad, indepen-
dent reasoning. There, the vote of the majority lost
much of its value very early on.
Among the Mu'tazilites, criticism emerged only
with al-Nazzam. He was accused of having been in-
fluenced by the Shiites, but in reality he was reacting
against the "people of hadith," ashab al-hadith, who
had begun to base the ijma' on a hadith: "My com-
169
THE FLOWERING OF MUSLIM THEOLOGY
munity will never reach agreement on an error."
33
That saying was demonstrably apocryphal. Shafi'I
does not cite it.3
4
Its content reminds us of the role
played by the Holy Spirit in Christianity. "The
Church of Rome does not make mistakes," Pope
Lucius I is supposed to have said, according to the
Pseudo-Isidorian Decretals. By disseminating the
hadith cited, Muslims reassured themselves that they
were orthodox. Dissension, ikhtila had always been
considered characteristic of sects and infidels. In ad-
dition, the infidels not only disagreed among them-
selves, they also filled their imaginations with collec-
tive errors. Christians, for example, all agreed with
the assertion that Jesus had died on the cross-which
was clearly false, according to the Koran. The Jews
were convinced that the revelation transmitted to
Moses had never been abrogated, a belief that was re-
futed by the existence of Islam. But is it true that
such errors had never occurred among Muslims?
Jahiz, following in al-Nazzam's footsteps, noted that
during the caliphate of al-Mansur, in Bahrain (which
at the time covered the entire eastern coast of the
Persian Gulf), the whole population prayed as a com-
munity on Thursdays-hence, the day before the pre-
scribed day. Al-Nazzam adopted a sarcastic tone in
expressing his view on the subject: if a group of blind
170
THEOLOGY AND ITS PRINCIPLES
people are brought together, they see no better than
they did before. All Muslims believe, for example,
that Muhammad was the only prophet sent to the
whole world. Yet reason shows us that that is false;
for all prophets-Jesus, Moses, and the others-
proved their authenticity through miracles, and a
miracle is perceived by the senses and is thus ad-
dressed to human beings as such. The ijma' is, then,
nothing but an illusion. But that sort of ruthless
skepticism not only destroyed i j m a ~ it had conse-
quences for hadith. By analogy, a report cannot be-
come reliable by the mere fact of having had several
chains of transmission ( mutawatir ps
Al-Nazzam was intelligent but was found to be a
bit too capricious. Jahiz, the first of the Mu'tazilites
to distance himself from his master's doctrines, de-
scribes his character.
36
Khayyat, who lived a genera-
tion later, claims that in Baghdad no one shared al-
Nazzam's views any longer. Yet another generation
later, Abu Hashim accepted the hadith mentioned
above, which al-Nazzam had ridiculed. He was deter-
mined to accept the i j m a ~ and no "authority" sup-
ported it, except this hadith and a few passages from
the Koran, which were however much too vague.37 As
the time of open options receded into the past, peo-
ple increasingly felt the need for an agreement on
171
THE FLOWERING OF MUSLIM THEOLOGY
true custom. Al-Nazzam had still made fun of the
Companions of the Prophet and of their internecine
quarrels, but immediately after his death political
correctness took hold in regard to the sacred past,
even in the Mu'tazilite schooP
8
To al-Nazzam, the
disputes among the Companions still seemed nor-
mal; discord was life, and human beings had received
intellect to help them find their own bearings. After
him, that attitude became untenable. The Mu'tazil-
ites never stopped praising reason, but now limits
were placed on the breaking of taboos. Qadi 'Abd al-
Jabbar would later say that even though the umma's
infallibility cannot be proved, one must not conclude
that it will necessarily commit errors.
The "sound reason" Wasil had spoken of was not yet
al-Nazzam's arrogant rationalism. What Wasil had in
mind was the good sense of the jurists. People re-
membered that among these jurists were virtuosi
who had in their time dazzled the masses with their
wisdom and subtlety: Iyas b. Mu'awiya, for example,
or, before him, Sha'bl_39 These men had a gift for
judging a situation intuitively, by firasa, with perspi-
cacity. But they did not yet have a method, and even
less a discourse on it. They employed analogies but
relied in doing so on their knowledge of human be-
172
THEOLOGY AND ITS PRINCIPLES
ings, without invoking any explicit rule. At the time,
jurisprudence was at the forefront, having come into
its own partly thanks to the instrument of the fatwa.
No one was writing manuals yet, but people were try-
ing to analyze specific situations that had not been
clarified. Each community, each "sect," had its own
specialist. In Mecca, there were specialists on pil-
grims. Texts are available to help us reconstitute that
forgotten culture: the al-jawabat (Responses) by Jabir
b. Zayd al-Azdi (d. AH 93/712 CE), or the al-Aqwal
Qatada, summaries of juridical decisions made by
Qatada b. Di'ama (d. AH II7?/735 CE?).
40
Unfortu-
nately, they were neglected because they were "sectar-
ian," alien to the dominant trend in later centuries.
The "method" revealed in them is ray (literally,
"opinion"), a mode of thought that followed the
logic of the situation without constantly appealing
to an independent authority, but which was opposed
to pure whim and arbitrary decisions (hawa).
The verb used in combination with ray was
ijtahada, "to exert oneself"; ijtahada rayaha meant "to
form an opinion about something."
41
In the long
run, the term ray lost its positive connotations. It
was associated with the al-Kufa school, the followers
of Abu Hanifa who had stayed closest to the princi-
ples of ancient tradition in Iraq. In Basra, a town with
173
THE FLOWERING OF MUSLIM THEOLOGY
a complex intellectual outlook, the term ijtihad re-
placed ray, and people began to concern themselves
with the specific reason, 'illa, that justified the proce-
dure applied.42 Those who did so came from different
"disciplines." 'Uthman al-Batti was a contemporary
of Abu Hanifa and corresponded with him. Abu
Hanifa sent him his famous Epistle, Risala.
43
'Uthman's response was that of a jurist, 'Amr b.
'Ubayd's that of a theologian, and the response of
'Abdallah b. Abi Ishaq al-Hadrami, who was a genera-
tion older than the two others (d. AH II7/735 CE), that
of a grammarian.44 In the beginning, however, the
Bastian ijtihad was taken to extremes, just as the ray
was in al-Kufa. 'Ubaydallah b. al-Hasan al-'Anbari (d.
AH r68/785 cE), a disciple oflyas b. Mu'awiya, formu-
lated the following maxim: "Whoever forms an opin-
ion (and is capable of doing so) is right," kull mujtahid
musfb.4S Muslims had suddenly arrived at pluralism-
and, it appeared, at relativism.
The expert's independence was never so con-
fidently asserted as at that time. 'Anbari did not dis-
tinguish between jurisprudence and theology. He
chose his examples from both fields at once. Mash 'ala
'1-khuffayn, shoe-rubbing, could be either accepted or
rejected, he said; but he also maintained that some-
one who defends free will is just as right as someone
174
THEOLOGY AND ITS PRINCIPLES
who believes in predestination. He always chose
problems that had not been definitively solved by the
Koran, the only authority 'Anbari seems to have ac-
cepted. For theology, that position turned out to be
untenable. Eternal truths could not be subject to the
caprices of the human intellect (though they were in
reality). In the fiqh, the human attempt to understand
divine law, it was a different situation. It was soon
recognized that the judgment of a mufti or qadi never
led to more than a certain probability (ghalabat al-
zann). Manuals devoted to the hermeneutic founda-
tions of the fiqh (usa! al-fiqh) never fail to address the
subject.
46
There too, Muslims wanted to avoid indeci-
siveness. 'Anbari had touched on a sensitive and im-
portant point, as the absence of appeals courts in the
classical judicial system demonstrates.
47
Yet jurists were not ready to resign themselves to
uncertainty. Shafi'I did not yet see the situation as re-
lating to verisimilitude. According to him, the be-
liever, in following the law, is doing something that is
either objectively correct or subjectively permitted.4B
It was the Mu'tazilites' rationalism that slowed down
the process. Both al-Asamm and Bishr al-Marisi, a
very influential jurist in the court of al-Ma'mun and
the instigator of the mihna, as well as al-Nazzam at a
later time, continued to hope that basing all legal ar-
175
THE FLOWERING OF MUSLIM THEOLOGY
gument on reason would be possible.
4
9 As for reason-
ing by analogy (qiyas), with which a casuistic system
cannot dispense, al-Asamm and Bishr al-Marisi
thought they could manage the situation through a
method more solid than the 'illa.
50
At that point,
however, al-Nazzam, with his usual skepticism, sug-
gested that they were not on the right path. Not even
God applied analogy-the Koran was the proof of
that. According to sura 24:31, no one is allowed to see
the hair of a free woman. Analogy would dictate that
that is also the case for a slave woman, especially if
she is beautiful (or more beautiful than the free
woman). But the opposite is true, as verse 33:59
shows. 5
1
The Sharia is rife with such contradictions.
For example, the traveler, ibn a l - s a b i ~ who is exempt
from prayer and fasting, must make up for only the
second of these obligations, not the first. 5
2
Chaos is
intentional, and analogy would only increase it. The
Koran is beyond criticism and can only be accepted.
Apart from it, the only authority is reason. It does
not tell us to proceed by analogy but to take scrip-
tural commandments literally. If a husband repudi-
ates his wife, he must do so with the exact words
mentioned in the Koran or with a sentence contain-
ing the word talaq (divorce); intention alone will not
176
THEOLOGY AND ITS PRINCIPLES
suffice. In taking that position, al-Nazzam aban-
doned the majority opinion and arrived at a literalist,
or zahirite, view. 5
3
Unlike his two predecessors, he was
not a practicing jurist.
Analogy, or qiyas, was considered a normal expression
of ijtihad, and was included among the four basic
principles of jurisprudence, the usul al-fiqh. It was not
a "source" but a method; it belonged to the realm of
form and not of matter. In jurisprudence, its func-
tion was to support casuistic probabilism. In theol-
ogy, that was not sufficient. Theologians used the
same term, but what they meant by it was qiyas al-
gha'ib 'ala 'l-shahid. Rather than an analogy, it was a
conclusion (qryas) based on what was before one's
eyes, shahid (that is, what was present and well
known), as compared to the hidden, gha'ib--.that is,
God and the hereafter. They did not admit that there
could be an analogy between God and the world. Is-
. lam has never developed the doctrine of analogia entis,
and the Mu'tazilites, despite all their theories about
the attributes of God, always had the utmost respect
for the affirmation of his transcendence, tanzih
(aphairesis for the Greeks).
54
For the Mu'tazilites,
analogy was a manifestation not of ijtihad but of
177
THE FLOWERING OF MUSLIM THEOLOGY
istidlal, a term they associated with the word for
"proof," dalzl. Dal!l was nothing other than a "sign,"
an indication, and qiyas was any ratiocination what-
soever, preferably a deduction. It is for that reason
that Aristotle's translators, usually Christians, took
the liberty of applying the term qiyas to the syllo-
gism, which in the Organon indicated the method for
drawing conclusions kat' exokhen. The fact that qiyas
was only a formal element in the jurists' tool kit facil-
itated the transition to philosophy, where the prob-
lems addressed were similar to those of theology.
With regard to the other three "sources" oflaw, such
an affinity did not exist. 5
5
Theological knowledge was however not only a cer-
tainty; it was also-and for the same reason-a duty.
The Koran emphasized that God provided the
"signs" (ayat) so that humankind could recognize
God's existence. In a sense, a knowledge of signs was
even "necessary," for they were perceived with the
eyes, and sense perception was inevitable-daruri;
given a priori. What was "acquired" (iktisabz) in that
process was only the conclusion, the result of moving
from the semeion ( dal!l) to the semeioton ( madlul ), from
sign to signified; but that conclusion could be drawn
by anyone. One had only to desire it. What was ac-
quired could also rightly be called ikhtiydri; depen-
178
THEOLOGY AND ITS PRINCIPLES
dent on the choice (ikhtiyar) of the individual. As are-
sult, knowledge of God became an obligation by
virtue of divine law, an element of taklif. It was a hu-
man act like any other. People could disagree on the
way to define that act. On the subject of free will
(ikhtiyar), for example, Dirar b. 'Amr had argued for a
synergism that divided the elements of the human
act between God and man. He thought that human
beings are agents by virtue of the fact that they per-
form the action and "acquire" it for their own benefit
(kasb or iktisab), whereas God creates it in that he
makes it occur. 56 The Mu'tazilites did not adopt that
model, but Abu '1-Hudhayl was one of the first to re-
fute it.57 He spoke of 'ilm i k t i s a b ~ knowledge "ac-
quired" by theological speculation (nazar).
58
And the
kasb, "acquisition," remained forever a "cipher," a
metaphor for "responsibility." 59
11
The appeal to the taklif produced new difficulties.
If the "acquisition" of religious knowledge was oblig-
atory, one ought not to postpone it. Every instant
spent without knowledge of God and Islam was
wasted. Moreover, knowledge required time; as a re-
sult, Muslims wondered what respite (muhla) people
could have once they had reached adulthood. In that
regard there was no fundamental difference between
believers and pagans. Of course, Muslims took ad-
179
THE FLOWERING OF MUSLIM THEOLOGY
vantage of the fact that their "respite" was normally
taken up by religious education, but the obligation
itself was also valid for non-Muslims, and insofar as
they fulfilled it, they were following God's command-
ment, just as believers were. Naturally, it was possible
that they might remain unbelievers in spite of every-
thing; in that case, their knowledge of truth would be
only partial, and they would have fulfilled their duty
without the intention that ought to accompany it.
Abu '1-Hudhayl was the first to discuss the problem.
There are works of obedience, he said, by which one
does not "will" God. That can also happen to Mus-
lims. They find themselves in this predicament in the
first instant of their knowledge of God, for they do
not yet know that what they are doing is a meritori-
ous act.
60
That theorem was rather far removed from reality.
Normally, education introduced the concept of merit
from the beginning, alongside the first religious
ideas. But the aporia hidden within it fascinated
Muslims. It may therefore be possible, they said, to
fulfill, without realizing it, an obligation established
by God. That eventuality seems to be universal: there
is no reason to limit it to the infidels, for to know
that one is obliged to know God, one must already
I80
THEOLOGY AND ITS PRINCIPLES
know him. One might conclude that the notion of
God is an a priori concept. That was for a long time
the conclusion; the word for "a priori" was fitra. 61
Abu '1-Hudhayl believed it, despite the fact that he
had gone to a great deal of trouble to develop proofs
of the existence of God.
62
In his view, these proofs
were designed only to provide reassurance a posteri-
ori. But in that case one would have to concede that
pagans too possessed that prior knowledge of God,
and should one truly concede that pleasure to them?
Certain theologians argued the reverse. The conse-
quence seemed to be that unbelievers, at least those
who had never had the opportunity to hear of God or
of Islam, were innocent. They also did not deserve
paradise, of course, but henceforth they could not be
condemned to hell. The Koran offered a third alter-
native: They would return to dust after their d ~ a t h s .
According to sura 78:40, the polytheists of Mecca
wanted to be treated that way after their resurrection,
but to no avail. By comparison with hell, it was a
more agreeable fate. But for those who had known
the truth and rejected it, such a fate was out of the
question.
Thumama b. Ashras, councilor and "minister
without portfolio" to the court of al-Ma'mun, made
I 8 I
THE FLOWERING OF MUSLIM THEOLOGY
the same argument, and al-Jahiz came to embrace his
viewpoint.63 But al-Jahiz took the idea further by as-
serting, in a psychological approach that was revolu-
tionary for the time, that despite the "presumption
of innocence," it may still be true that everyone, Mus-
lim or not, arrives at the knowledge of God without
being initiated in advance. One must simply concede
that such knowledge cannot be controlled and that it
shares that distinctive character with all other types
of knowledge. Above all, it need not be "produced"
(muwallad, mutawallid) by something; it can result
naturally from an act of reflection, but that is not a
condition sine qua non. Our brains work in a differ-
ent way; our speculations are not calculable. Hence,
there is no obligation to know God.
These ideas were ingenious and infused with a sub-
tle tolerance, but they also indicated the bankruptcy
of Mu'tazilite rationalism. In reacting against that
apparent flaw, al-Jubba'I, with his usual scholasti-
cism, tried to prove why human beings, despite the
arguments advanced by al-Jahiz, must necessarily feel
the obligation to know God. No one, he said, lacks
the experience of depending on a power that governs
him and to which he is beholden. To show gratitude
toward this power, however, a person must know
whom to address. Knowledge of God is thus as it
182
THEOLOGY AND ITS PRINCIPLES
were the personalization of the experience of the "nu-
minous," of an indefinite divine power.64 That theory
too was only a hypothesis at first. A Muslim did not
need that confrontation with the numinous, nor did
the "people of the Book," the ahl al-kitdb, nor the ma-
jority of pagans, for they were all conditioned by the
religious ideas of their society. To support the theory,
therefore, an example had to be found or con-
structed. It was soon discovered in the figure of an in-
dividualliving alone on an island, the Robinson Cru-
soe motif. Al-Jubba'I himself seems to have used it.
Later on, it was primarily Shiite authors who
adopted it: Kulini (d. AH 328/939 cE) and Ibn Baba-
wayh (d. AH 381/991 CE). The idea was later developed
by Ibn Tufayl. Finally, Ibn al-Nafis granted it promi-
nence and pushed it to the point of caricature in his
Risdla al-Kamiliyya.
65
In his narrative, the hero.,discov-
ers not only theological and metaphysical truths by
virtue of his intellect, but also, through the rational
necessity of prophethood, the course of human his-
tory. The location of the "Robinson Crusoe" (who is
called Kamil, "Perfect," in Ibn al-Nafis and Hayy b.
Yaqzan, "Living son of Awakened," in Ibn Tufayl), was
naturally a marginal detail. The Orient had various
remote places to offer: the desert (Ibn Babawayh), a
mountaintop (Kulini). Only Juwayni, an Ash'arite,
183
THE FLOWERING OF MUSLIM THEOLOGY
and Muhammad b. Ibrahim al-Kindi, an Ibadite,
spoke of an island.
66
Rationalists trusting too much in certain models
may become presumptuous, if not blind and fanati-
cal. Over time, the Mu'tazilites stopped being the
friends of the masses; they hated the incapacity of
the latter to understand true doctrine. Even Bishr b.
al-Mu'tamir, who had tried to win over the common
people with his didactic poetry, criticized their taqlid,
the intellectual indolence they showed-especially in
not following his views.
6
7 His disciple al-Murdar de-
veloped the same attitude, at least for a while, until
he converted to a more accessible didactic style and
began to compose books intelligible to common
mortals.Gs His ascetic simplicity linked him to the
sufiyyat al-Mu 'tazila, the only ones to take a stand
against intellectual arrogance. Outside the Mu'tazil-
ite circles, Ibn Kullab and his friends, who had en-
dured the brutality of the mihna, also opted for more
moderation. For them, someone incapable of articu-
lating on his own why he is a believer can neverthe-
less be considered one. But even they expressed reser-
vations. They admitted that such a person obeyed
God with his faith, but they remained convinced that
in spite of everything he was committing a sin be-
184
THEOLOGY AND ITS PRINCIPLES
cause he neglected theological reflection. The
muqallid, the one who, through his sacrificium
intellectus, fails to find the right path, must therefore
be likened to the "prevaricator," fasiq. 69 Except that
according to them the fasiq was not in an intermedi-
ate position (manzila bayna 'l-manzilatayn), as he was
for the Mu'tazilites, but could place his hope in God,
who in his mercy might spare the fasiq the sufferings
ofhell.7
Intellectualism turned inward led to a certain
quarrelsomeness. It was instigated most forcefully by
public disputation, mundzara, where arrogance
joined forces with competitiveness. The warping of
the profession that resulted was typical of the
mutakallimun. Let us compare other kinds of religious
activity. Hadith was simply transmitted. Mysticism
was limited at the time to intimate contact bef!Ween
master and disciple. Greek philosophy and science
were taught at home. Conversely, theology, because
of its apologetic nature, favored public debate from
the start,' and that attracted curiosity-seekers and
produced emotional reactions. Winning and losing
were always at stake. As in the marketplace, skill and
speed of reaction often prevailed over circumspection
and sincerity. Al-Jahiz saw clearly that that propen-
sity could devolve into charlatanry. He himself,
185
THE FLOWERING OF MUSLIM THEOLOGY
though a good writer, was far from a good public
jouster. His physiognomy did not lend itself to that,
and the rules of public display may have been the
same as they are in modern democracies. Al-Jahiz
also recognized that in the heat of argument the line
between antagonistic positions could vanish. That
was not serious so long as the positions were fictive,
as they were in belles lettres, adab (in his Book of Ani-
mals, he himself had il).vented a disputatio between a
cock and a dog).71 But theology was entirely different.
One does not play around with the truth, and it is
not enough to be right, especially by virtue of one's
rhetorical skill. In the long run, the image of the
mutakallimun increasingly came to resemble that of a
star lawyer in an American trial today. The Mu'tazil-
ites were aware of the disadvantages. They found
themselves facing a wave of resentment and antipa-
thy. Even now, the inclination is to take the word
kaldm, from which 'ilm al-kaldm and mutakallimun are
derived, to mean "pointless talk," "prattle." The sim-
ple folk had that reaction, but so did the scholars
close to them, the ashab al-hadtth, many jurists, and
finally the philosophers themselves. Philosophers,
however, reacted that way for a different reason: they
thought that the style of the kaldm corresponded to
Aristotle's dialectic, a genre the Greek master had
186
THEOLOGY AND ITS PRINCIPLES
ranked second, clearly below the apodictic art repre-
sented by syllogistic logic.72
How do we know what we know? Why are we right
when we are right? Men of later generations, al-
Ghazali, for example, but also Fakhr al-Din al-Razi,
were fascinated by syllogism and had great hopes for
it. But its promise proved illusory. Ultimately, syllo-
gism yields only what has been put into it. The con-
clusion depends on the premises, but who will verifY
the premises? If the premises had been taken from
the Koran, the question would have been
superfluous. That was not generally the case; theolo-
gians always proceedecj. differently, as we have seen.
Knowledge comes from a conviction (i'tiqad) sup-
ported by proofs, they said. Abu '1-Hudhayl had al-
ready made that assertion, adding that such <t1 pre-
condition is not valid for every sort of knowledge.
Some forms of knowledge are not "produced" be-
cause they are "necessary."73 But when they are pro-
duced, that is, when truth emerges at the end of our
reflection process, we sense it. Truth does not mani-
fest itself through reality testing, which in any case is
not always possible in theology. Rather, it imposes it-
self through a subjective criterion: peace of mind
(sukun al-qalb ). To think, to reflect, is a movement of
187
THE FLOWERING OF MUSLIM THEOLOGY
the soul, said al-Nazzam, and that movement sub-
sides when one arrives at knowledge.7
4
Nevertheless,
it is illusory to believe that the piece of knowledge is
true for that reason; perhaps we are simply the vic-
tims of a natural reflex. Indeed, those who err also be-
lieve they are right. Al-Jahiz exhorted people not to
forget that; he remembered the Greek skeptics' objec-
tions,75 It was reported that one of these skeptics, a
"sophist," had badgered Thumama b. Ashras, in in-
sisting that every thought is only conjecture (hisbdn).
As a result, al-Jahiz added an objective criterion to
al-Nazzam's approach. There are sentences, he said,
that are recognized as true not only subjectively but
also objectively, by virtue of their correspondence to
reality, their adaequatio intellectus et rei, as the Chris-
tian Scholastics would have said. He had discovered
the need for that adjustment when considering false
propositions. Although it is acceptable to say that
something is true because one believes it, one cannot
say that the same thing is false because one does not
believe it. There thus exist not only statements that
are true and others that are false, but also statements
that are neither true nor false, either because they do
not correspond to reality, despite the fact that some-
one believes them, or, conversely, because someone
says of a thing that it is objectively true, without be-
188
THEOLOGY AND ITS PRINCIPLES
ing convinced of it. The first case is that of "inno-
cent" pagans, the second that of "hypocrites,"
mundfiqun.7
6
Al-Jahiz knew that his master al-Nazzam
had already referred to a Koranic verse in which a
similar case was described by God himself "When
the hypocrites come to you [the Prophet] they say:
'We bear witness that you are God's apostle.' God
knows that you are indeed his apostle; and God bears
witness that the hypocrites are surely lying."77 At a
certain point, then, the mundfiqun had said, but with-
out believing it, something that was true; for that
reason, they were liars. Truth and sincerity were not
the same thing, nor were lies and error. Unfortu-
nately, at the time, the Arabic expression kadhabta
could mean both "you are lying" and "what you're
saying is false." Like the word pseudos in Greek, the
Arabic word does not distinguish between the subjec-
tive and the objective meaning.7s But even though
there is no difference in the language, there certainly
is one in reality. Once again, al-Jahiz took the Koran
to witness. Sura 34:8 tells us that the polytheists of
Mecca could not decide whether to consider the
Prophet a liar or a man possessed by jinni. Yet in this
context, "possessed" could only allude to someone
who says false things without knowing it.
The first to refute the "sophists," or rather, the ar-
189
THE FLOWERING OF MUSLIM THEOLOGY
guments of the Greek skeptics, was another student
of al-Nazzam, Muhammad b. Shabib al-Basri. He
knew that to achieve that goal, he would have to ex-
plain sensory illusions.79 We are familiar with the
skeptics' arguments through Sextus Empiricus's
treatises and through other texts. In the Muslim
world, these books were still unknown, and their
"sophisms" were instead spread via a "diffuse tradi-
tion" that disseminated many Greek ideas never of-
ficially translated. Those responsible for it may have
been physicians: the "empirical" school that refused
to base diagnoses on syllogisms had links with skep-
ticism. But fundamentally, theologians were not
dealing with sensory illusions; the problem was reve-
lation. No one would ever succeed in explaining it
and replacing it with reason. If all the elements of
faith could be discovered through human reflection,
why would God have spoken? That is what Ibn al-
Rawandi said, and he was not altogether wrong. If he
had heard the reply to his argument that his antago-
nists later gave-namely, that God did so to spare his
creatures a little trouble-he would have only smiled
sarcastically.
80
It is true that Muslims have avoided
absurdity for all time;
81
even for them, however, har-
mony between faith and reason remained an inacces-
sible ideal.
190
NOTES
INDEX
il
NOTES
Introduction
1. Or "a rationality" ('aql). Taha Husayn, Mustaqbal al-
thaqdfa ft Misr; 2 vols. (Cairo, 1939), 1:23, fourth and
fifth lines from the bottom.
2. Ibid., 63, lines 2-3.
3. Mustafa 'Abd al-Raziq, "L'Islam et l'Occident," Ca-
hiers du Sud 34 (1947): 19.
4 A. von Kiigelgen, Averroes und die arabische Moderne:
Ansdtze zu einer Neubegrundung des Rationalismus im
Islam (Leiden, 1994).
5 Such "conversations," takallama and kallama (con-
nected with the accusative form), were probably
modeled on the Greek dialegesthai (peri tinos); see my
Theologie und Gesellschaft im 2. und 3 ]ahrhundert
Hidschra, 6 vols. (Berlin, 1991-1997), 1:48-49. Hereaf-
ter cited as TG.
6. TG, 3:66, 307.
7 Ibid., 299-300.
8. Ibn al-Faqih, Akhbar al-buldan (Frankfurt, 1987), in
Majmu 'ft 'l-jughrafiya, publications of the Institut
193
NOTES TO PAGES 9-32
fur Geschichte der Arabisch-Islamischen
Wissenschaften, C 43, p. 105, lines 4-5.
I. Theology in Its Own Eyes
I. See TG, 3:470-47r, 4:676.
2. TG, 4:675; for the sister, see also 722-723.
3 Passages from the Koran are taken from the Pen-
guin translation by N.]. Dawood (New York,
I995).-Trans.
4 According toT. Nagel, Der Koran (Munich, r983),
r38.
5 TG, 4:353-360.
6. Ibid., 68o-683.
7- Ibid., 657-660.
8. Ibid., J:390-392.
9. Ibid., 2:r55-r64.
ro. Ibid., 4:656-657.
II. Ibid., 3:176, 4:691.
r2. Ibid., r:r36-r37.
I3. Ibid., I:4I6-443, J:33I-342.
I4. Ibid., 3:20-22.
I5. Ibid., I:443-450.
r6. Ibid., 4r7-4I8.
I7 Ibid., 4:976, index, s.v. "Radd 'ala '1-zanadiqa."
r8. Ibid., 3:65.
I9. Ibid., 4:269, 299-304.
20. Ibid., 3:446-481.
21. Ibid., 4:46r-465.
194
NOTES TO PAGES 37-50
22. Ibid., 676-678, 68o-68I.
23. For the origin of the term among the Shiites, see
TG, r:377 and 385; compare suras 4:I7I and 5:77.
24. TG, 4:678-679, 687.
25. Abu Hamid al-Ghazali, Paisa[ al-tafriqa bayna 'l islam
wa'l zandaqa, written before the Mustasfd and the
Munqidh min al-daldl, but after the Tahdfut.
26. TG, 4:677-678; for the Mamluk dynasty, see E.
Geoffroy, Le soufisme en Egypte et en Syrie (Damascus,
I995), 380-385.
27. ]. C. M. Laurent, Peregrinatores medii aevi quattuor
(Leipzig, r864), I35, chap. 30, nos. 4-r2.
2. Theology and the Koran
I have discussed the subject of this chapter in more detail
in Le voyage initiatique en terre d'islam, ed. M. A. Amir-
Moezzi (Louvain, I996), 27-56; compare TG, 4:373-424,
esp. 387-391. For the problem of anthropomorphisp in
general, see D. Gimaret, Dieu a !'image de l'homme: Les
anthropomorphismes de la sunna et leur interpretation par les
theologiens (Paris, I997). Gimaret cites evidence from
hadith.
I. See especially Le livre de l'echelle de Mahomet (Paris,
I99I);Jamel-Eddine Bencheikh, Le voyage nocturne de
Mahomet (Paris, r988).
2. Translation has been significantly modified to con-
form to the French. The author is following, with
his own modifications, Jacques Berque's French
195
NOTES TO PAGES 52-70
translation, Le Coran, un nouvel essai de traduction
(Paris, 1995).-Trans.
3 A later example was al-Ghazali in his Mishkat al-
anwdr.
4 TG, 2:642, 3:450-452, 4:215.
s. Al-Tabari,jdmi' al-baydn 'an ta'wll dy al-Qur'dn
(Cairo, AH 1373/1954 CE), 27:44-45.
6. Ibn 'Abbas may have already used that argument
(al-Tabari, Tafsir, 27:48, lines 3-4);fu'ad and qalb are
synonyms.
7 Jorge Luis Borges, El tintorero enmascarado Hakim de
Merv, in Obras completas (Buenos Aires, 1974), 324-
325.
8. W. Madelung and P. E. Walker, An Ismaili
Heresiography: Tbe "Bah al-shaytdn" from Abu
Tammam's Kitab al-shajara (Leiden, 1998), 76-77
(Arabic text), 75 (translation). On this text, see also
P. Walker, An Isma 'ili Version of the Heresiography of the
Seventy-two Erring Sects, in Medieval Isma 'ili History
and Thought, ed. F. Daftary (Cambridge, 1996), 161-
162.
9 Al-Suyuti, Al-La' ali al-masnii 'a fi 1-ahddith al-mawdii 'a,
2 vols. (Cairo, n.d.), 1:63-81; compare TG, 2:509.
ro. Al-Tabari,jdmi' al-baydn 'an ta'wil ay al-Qur'dn, 27:48,
18-19 and before.
n. Al-Tabari, Ta'rikh, r:r192, lines 3-4.; al-Tabari,jdmi'
al-baydn, 17:186-187, but related to sura 22:52 and
not to the passages in surat al-Najm.
196
NOTES TO PAGES 72-87
12. See C. Gilliot in Arabica 32 (1985): 62-63.
13. TG, 2:452-453.
14. Regarding that attribution, however, see the remark
by C. Gilliot in Israel Oriental Studies 19 (1999): 67-68.
15. See TG, 1:317, 4:393-394, and I027 (index).
3 Theology and Science
r. As a general reference for this chapter, see TG,
4:459-477
2. David Furley, in Proceedings of the Boston Area Collo-
quium in Ancient Philosophy, vol. 2, ed. J.-J. Cleary
(Lanham, 1986), 1-2.
3 TG, 2:499-SOO, 454-455
4 Ibid., 442-449, 450-452, 495, sao.
s. I am thinking of the famous "Theology of Aris-
tode," but also of other texts such as the later Liber
de Causis; see F. Zimmermann in Pseudo-Aristotle in
the Middle Ages, ed. Jill Kraye et al. (London, 1 ~ 6 ) ,
IIO-III.
6. TG, B7-44
7- Ibid., 224-225; compare 67-70.
8. Ibid., 229-230, 280-282.
9 Ibid., 231-232, 4:446-447.
ro. At the time, the pre-Socratics were recast as Neopla-
tonists in an apocryphal text attributed to
Ammonius Saccas; see U. Rudolph, Die Doxographie
des Pseudo-Ammonios: Bin Beitrag zur neoplatonischen
Uberlieferung im Islam (Stuttgart, 1989).
197
NOTES TO PAGES 52-70
translation, Le Coran, un nouvel essai de traduction
(Paris, I995).-Trans.
3. A later example was al-Ghazali in his Mishkdt al-
anwdr.
4. TG, 2:642, 3:450-452, 4:215.
s. Al-Tabari,jdmi' al-baydn 'an ta'wil ay al-Qur'dn
(Cairo, AH I373/I954 CE), 27:44-45.
6. Ibn 'Abbas may have already used that argument
(al-Tabari, Tafsir, 27:48, lines 3-4);jit'ad and qalb are
synonyms.
7 Jorge Luis Borges, El tintorero enmascarado Hakim de
Merv, in Obras completas (Buenos Aires, I974), 324-
325.
8. W. Madelung and P. E. Walker, An Ismaili
Heresiography: The "Bah al-shaytan" from Abu
Tammam's Kitab al-shajara (Leiden, I998), 76-77
(Arabic text), 75 (translation). On this text, see also
P. Walker, An Isma 'ili Version of the Heresiography of the
Seventy-two Erring Sects, in Medieval Isma 'ili History
and ed. F. Daftary (Cambridge, I996), I6I-
I62.
9 Al-Suyuti, Al-La'dli al-masnii'a fi l-ahddlth al-mawdu'a,
2 vols. (Cairo, n.d.), I:63-8I; compare TG, 2:509.
IO. Al-Tabari,]dmi' al-baydn 'an ta'wil dy al-Qur'dn, 2]:48,
IS-I9 and before.
II. Al-Tabari, Ta'rikh, I:II92, lines 3-4.; al-Tabari,jdmi'
al-baydn, I7:I86-I87, but related to sura 22:52 and
not to the passages in surat al-Najm.
196
NOTES TO PAGES 72-87
I2. See C. Gilliot in Arabica 32 (I985): 62-63.
I3. TG, 2:452-453.
I4. Regarding that attribution, however, see the remark
by C. Gilliot in Israel Oriental Studies I9 (I999): 67-68.
IS. See TG, I:3I7, 4:393-394, and 1027 (index).
3. Theology and Science
I. As a general reference for this chapter, see TG,
4:459-477
2. David Furley, in Proceedings of the Boston Area Collo-
quium in Ancient Philosophy, vol. 2, ed. ].-]. Cleary
(Lanham, I986), I-2.
3 TG, 2:499-500, 454-455
4 Ibid., 442-449, 450-452, 495, soo.
s. I am thinking of the famous "Theology of Aris-
totle," but also of other texts such as the later Liber
de Causis; see F. Zimmermann in Pseudo-Aristotle in
the Middle Ages, ed. Jill Kraye et al. (London,
IIO-III.
6. TG, B7-44
7 Ibid., 224-225; compare 67-70.
8. Ibid., 229-230, 280-282.
9 Ibid., 23I-232, 4:446-447.
w. At the time, the pre-Socratics were recast as Neopla-
tonists in an apocryphal text attributed to
Ammonius Saccas; see U. Rudolph, Die Doxographie
des Pseudo-Ammonios: Bin Beitrag zur neoplatonischen
Uberlieferung im Islam (Stuttgart, I989).
197
NOTES TO PAGES 88-104
II. TG, !:4!8-443
12. Ibid., 2:398-400.
13. Ibid., 1:442.
14. Ibid., P33-234, 314-316.
15. Ibid., 234-235.
r6. Ibid., 255-263.
17. For that later phase, see A. Dhanani, The Physical
Theory ofKaldm: Atoms, Space, and Void in Basrian
Mu 'tazili Cosmology (Leiden, 1994).
r8. TG, 3:331-352.
19. Ibid., 309-323.
20. Ibid., 309.
21. Ibid., 6:19-20 (n. 33).
22. Ibid., J:3I6-3I7.
23. Ibid., 67.
24. Ibid., 225-227.
25. Ibid., 227; see the diagram in Dhanani, Physical The-
ory ofKaldm, 135.
26. Schlomo Pines, Beitrage zur islamischen Atomenlehre
(Berlin, 1936).
27. TG, 3:241-243.
28. Ibid., 233-236.
29. Ibid., 236, rzr (with correction 5:307).
30. Ibid., 4:473-474, 479
31. Ibid., 6:23n38, 3:319.
32. Ibid., 6:21-22, 22n36.
33 Ibid., 3:320, 323-324.
198
NOTES TO PAGES 104-125
34 Ibid., 236-237.
35 Dhanani, Physical Theory ofKaldm, 177-r8o.
36. TG, 4:4-5, 467-468.
37 Ibid., 137-138.
38. Ibid., 3:425-426.
39. Compare ibid., 4:479-480 and 5!3-517, plus the ref-
erences indicated there.
40. Ibid., J:367-369.
4L Ibid., 317-
42. Ibid., 428-445.
43 Ibid., 4:468-470.
44 Ibid., 476-477, 558.
45 F. Meier, Baha'-i Walad: Grundzuge seines Lebens und
seiner Mystik (Leiden, 1989), 436-439.
4 Theology and Human Reality
I. Here I am summarizing TG, 4:695-717.
2. At least for the theologians: the writings CO!liposed
by the kuttab ('Abd al-Hamid b. Yahya, Ibn al-
Muqaffa', and so on) and the mirrors of princes
had a happier fate.
3 TG, I:I69-I7I, 175-!76.
4 Ibid., 183-184.
5 Ibid., 2:271-273.
6. Or to use Arabic terminology, with the condemna-
tion of sabb al-sahaba, polemicizing against the
Companions: see TG, 1:236-237, 2:436, 3:451.
199
NOTES TO PAGES 125-151
7 See E. Landau-Tasseron in Der Islam 67 (I990 ): 2-3.
8. TG, 4:I4-I5.
9 Ibid., 3:439.
IO. Ibid., 4:700.
II. Ibid., J:I29-I30.
I2. Ibid., I:3o8-312.
I3. Ibid., 2:387-391.
I4. Ibid., 4703-706.
I5. Ibid., I:87, 2:493. Jahm was a follower ofHarith b.
Surayj, who had risen up against the Umayyad re-
gime with the watchword irjd'.
I6. Ibid., 408-409.
!7 Ibid., I:86-88.
IS. Ibid., 4709-710.
I9. Ibid., 3:55-57.
20. Ibid., 55
2!. Ibid., 47II.
22. Ibid., 3:I09-II2.
23. Ibid., I73-I75
24. Ibid., 4I6.
25. Ibid., 2:408-409, 47I4-7I5.
26. Ibid., 2:388, 3:I74, I99-2oo.
27. Ibid., 4:46.
28. Ibid., 3:I30-I33, 4:88-94.
29. Ibid., 47I6.
30. Ibid., 69-70.
31. For them, see A. T. Karamustafa, God's Unruly
Friends (Salt Lake City, I994).
200
NOTES TO PAGES 153-158
5 Theology and Its Principles
I. See G. Sohngen in Lexikon fur Theologie und Kirche, II
vols. (Freiburg, I957-I967), 3:89. A theological evalu-
ation is given in H. Schiitze-Eichel, "Credo quia ab-
surdum est?" Trierer Theologische Zeitschrift 84 (I975):
I56ff.
2. TG 4:205-206, I98.
3 Ibid., 92-93, 295-296; see also my "Skepticism in Is-
lamic Thought," Al-Abhath 2I (I968): Iff.
4 From 'Askar Mukram, where there were many
Mu'tazilites.
5 A German translation and commentary appear in
TG) 5:I6I-I62.
6. Ibid., I:34-35.
7 See quotation, ibid., 5:I63.
8. Ibid., p92-293.
9 Up to now, there has been a tendency to read.,Shafi'I
through the eyes oflater theoreticians of his
school, who projected their quadripartite schema
onto his Risdla (Epistle). But]. E. Lowry, "Legal-
Theoretical Content of the Risala of Muhammad b.
Idris al-Shafi'i" (Ph.D. diss., University of Pennsyl-
vania, I999), has clearly shown that the concept of
haydn dominating Shafi'I's thought turns essen-
tially on correspondences between the Koran and
the sunna. For the i j m d ~ see 43ff. and 426ff.
IO. For Wasil, see TG) 2:279-280 and 4:649-650.
201
NOTES TO PAGES 159-164
n. For Al-Jubba'i, see D. Gimaret, Une lecture mu'tazilite
du Coran: Le Tafs!r d'Aba 'All al-Djubbd't (m. 303/915)
partiellement reconstitue a partir de ses citateurs
(Louvain, 1994). For Abu '1-Qasim al-Balkhi (d. AH
319/931 cE), see Encyclopaedia Iranica, ed. Ehsan
Yarshater (London, 1985-), r:36ob, s.v. "Abu '1-
Qasem al-Ka'bi." For Abu Muslim al-Isfahani (d. AH
322/934 CE), see TG, 1:430-431. For al-Asamm, see
2:403-404, and for 'Amr b. 'Ubayd, 298-299.
12. Suras 25:33 and 3:7.
13. Daniel27 and 5:12. When in Arabic the verb ap-
pears in its second form (fassara ), this is also Syriac
(Brockelmann, Lexicon Syriacum, 6r5a).
14. See TG, 4:984, index of terms.
15. Ibid., 3:51-52, 384; see the text I published in Der
Orient in der Forschung: Festschrift 0. Spies, ed. W.
Hoenerbach (Wiesbaden, 1967), 17off.
r6. Sura 8:65.
17. TG, 3:266-267, 4:650-651; see in general, and also for
what follows, my "L'autorite de la tradition
prophetique dans la theologie mu'tazilite," in La
notion d'autorite au Moyen Age: Islam, Byzance,
Occident (Paris, 1982), 2nff.
r8. TG 3:6o, 4:652.
19. Shafi'I, Al-Risdla, ed. Ahmad Muhammad Shakir
(Cairo, AH 1358/1940 CE), 169ff., 998-1308; com-
pare Lowry, Legal-Theoretical Content, 26rff.
20. TG, 3:383.
202
NOTES TO PAGES 164-174
2!. Ibid., 4:II3-II4.
22. Ibid., J:306.
23. Ibid., 4I0-4II.
24. Ibid., 4:42-43.
25. See my Die Erkenntnislehre des 'Adudaddtn al-!ct
(Wiesbaden, 1966), 400.
26. TG, 2:133-134.
27. Ibid., 4:654.
28. Lowry, Legal-Theoretical Content, 427ff.
29. TG, 4:655.
30. Ibid., 3:5r.
31. Ibid., 4:657.
32. 'Abd al-Qahir al-Baghdadi, Usa! al-Dtn (Istanbul,
1928), 19, lines 6-7.
33 TG, J:385-386.
34 Lowry, Legal-Theoretical Content, 436.
35 TG, J:384, 4:656-657.
36. See, for example, his opinion on the tafra, ibjd.,
4:419; for the description of Al-Nazzam's character,
see 3:306.
37- Ibid., 4:657.
38. Ibid., B90-39L
39 For Iyas b. Mu'awiya, see ibid., 2:154-155. For Sha'bi,
see G. H. A. Juynboll entry, Encyclopaedia of Islam, 2d
ed., (Leiden, 1960-2004), 9:162-163.
40. TG, 2:191, 143-144.
41. Ibid., 4:664.
42. Shafi'I would later use the term ma'nd instead of
203
NOTES TO PAGES 174-176
'illa-see W. Hallaq in Der Islam 64 (1987): 45-but
that does not necessarily mean that elsewhere 'illa
was not already in use. Shafi'I was never influenced
by Basrian thought. Lowry, Legal-Theoretical
translates ma'na as "policy reason" (zn-212).
43 Al-Risala ila 'Uthman al-Batti, ed. Muhammad Zahid
al-Kawthari, with the Kitab al-'alim wal-muta'allim
(Cairo, 1368/r949), 34ff.; compare TG, r:192ff.
44 Compare the two texts in TG, 5:171-172, nn. s-6; see
also 2:302, 4:662. In grammar, the search for the
'illa, the cause of the evolution oflanguage and its
irregularities, remained characteristic of the
Basrian "school."
45. Ibid., 2:161-162; see also my "Laliberte du juge dans
le milieu basrien du VIlle siecle," in La notion de
liberte au Moyen Age, zsff.
46. See the texts mentioned in TG, s:n8-n9.
47 The mazdlim courts are different; see Encyclopaedia
of Islam, zd ed., 6:933-934.
48. Lowry, Legal-Theoretical 359. The essential
point is that in Islam everyone must apply the law
in his or her own ritual duties.
49 TG, 3:176ff.
so. See, for Bishr, ibid., 187; for al-Asamm, ibid., 2:415-
416; see also the summary in 4:662-663, where their
position is compared with Shafi'I's.
51. Ibid., J:387.
52. Ibid., 6:190, text 255c. The case of the ibn al-sabtl was
204
NOTES TO PAGES 177-179
also discussed by Shafi'I, Risala 352; compare
Lowry, Legal-Theoretical 450.
53. That attitude would be propagated openly only two
generations later, by Dawud b. 'Ali al-Isbahani (d.
AH 270/884 CE). He was often in contact with the
Mu'tazilites (see TG, 4:223-224); he also studied
with Abu 'Abd al-Rahman al-Shafi'I, the disciple of
Abu 'l-Hudhayl, but Dawud's father was a Hanafite.
54 On this term, see my article in Encyclopaedia of Is-
lam, zd ed., !0:341-342, s.v. "Tashbih wa-tanzih."
55 The only possible exception is the ijma'. Schacht
has noted its affinity to the opinio prudentium of Ro-
man law. Greek philosophers spoke of koinonia ton
anthropon. But the resemblances are deceptive, and
direct influence is impossible to prove (see TG,
4:655). The equivalent of qiyas al-gha'ib 'ala 'l-shahid
is analogismos, which in the Arabic version of Galen
ofPergamum's Peri tes iatrikes empeirias was tranili
lated as qiyas bi'l-zahir 'ala 'l-khaft (see TG, 4:665). On
the question whether the influence of the Stoic
semeion, semeioton, semeiosis can be seen in the no-
menclature madlal, istidlal, see my "The Logical
Structure ofislamic Theology," in Logic in Classical
Islamic Culture, ed. G. E. von Grunebaum (Wies-
baden, 1970), 21-22.
56. TG, 3:45-46, 4:502-503.
57 The first was probably Bishr b. al-Mu'tamir (ibid.,
5:231, "Widerlegungen," B1); Abu '1-Hudhayl espe-
205
i
NOTES TO PAGES !79-!86
cially attacked Hafs al-Fard, who was Dirar's most
eminent disciple (ibid., 2729-730, 3:4S, 276).
sS. Bishr b. al-Mu'tamir did the same, but the ant-
onym he used was not dariiri but ibtidd'I, "given in
advance" (ibid., 3:nS).
59 Mas'uliyya is a recent term: R. Dozy, Supplement aux
dictionnaires arabes, 1:621b, recognizes the word only
through Butrus al-Bustani's Muhit al-Muhlt, with
the meaning "obligation"-despite the fact that
mas'ul is Koranic.
6o. TG, 3:252-353.
6r. Ibid., 4:361-362.
62. Ibid., 3:231-232.
63. Ibid., 16S-169, 4:99-roo.
64. See my Die Erkenntnislehre, 329-330; see also my
"Early Islamic Theologians on the Existence of
God," in Islam and the Medieval West, ed. K. I.
Semaan (New York, 19So), 64ff.
65. Published by M. Meyerhof and]. Schacht under the
title The Theologus Autodidactus of Ibn al-Nafis (Ox-
ford, 196S).
66. TG, 4:669-670.
67. See, for Bishr b. al-Mu'tamir, ibid., 3:109-IIO.
6S. Ibid., 13S-139.
69. For this term, see D. Gimaret, La doctrine d'al-Ash'ari
(Paris, 1990), 470n2.
70. TG, 4:671.
71. This was not purely imaginary. Rather, he tells of
206
NOTES TO PAGES !87-!90
episodes in a middle-class salon in Basra, where al-
Nazzam had taken the side of the dog while a
Mu'tazilite colleague had defended the cock.
72. TG, 4727-72S.
73 Ibid., 3:254. According to him, that was especially
true of sense perceptions. Later, the universal laws
of thought, such as the principle of the excluded
third term, were added. These laws exist in us a pri-
ori (see my Die Erkenntnislehre, 164ff.). The sequence
zann-i'tiqaa-'ilm by which knowledge advances to
ward certainty is reminiscent of the Kantian hierar-
chy Meinen-Glauben-Wissen (Kritik der reinen
Vernunft, B Sso-Transzendentale Methodenlehre,
2:3).
74 TG, J:3So.
75 Ibid., 4:ro2-I03.
76. Ibid., 97-9S.
77 Sura 63:1.
7S. See Wiirterbuch der klassischen arabischen Sprache,
1:90a, 91b.
79 TG, 4:12S.
So. See, for the details of the discussion, ibid., 320ff.
Sr. See my "Gottliche Allmacht im Zerrbild
menschlicher Sprache," Melanges de l'Universite Saint-
Joseph 49 (1975-1976): 653ff.
207