Muslims in general have a tendency to disarm any criticisms of Islam and in particular the Koran by asking if the critic has read the Koran in the original Arabic, as though all the difficulties of their sacred text will somehow disappear once the reader has mastered the holy language and has direct experience, aural and visual, of the very words of God, to which no translation can do justice.

In a letter to Mme. du Deffand, who wished to compare Virgil to Alexander Pope, Voltaire wrote "Vous le connaissez par les traductions: mais les poetes ne se traduisent point. Peut-on traduire de la musique?" ("You know him through translations: but poets are not translatable. Can one translate music?" May 19, 1754) As John Hollander remarks, Voltaire's opinion "seems to prefigure the views of a later century, in associating with music not the beauty, or decoration, but a strange sort of ineffable, incomprehensible, and (hence?) untranslatable core of pure poetry."" This, I think, captures the Muslim's almost mystical and rather irrational attitude to the untranslatability of the Koran very well.

Jackson Mathews also singles out another feature that is most difficult to translate: "Rhythm is the one feature of a foreign language that we can probably never learn to hear purely. Rhythm and the meaning of rhythm lie too deep in us. They are absorbed into the habits of the body and the uses of the voice along with all our earliest apprehensions of ourselves and the world. Rhythm forms the sensibility, becomes part of the personality; and one's sense of rhythm is shaped once and for all on one's native tongue."2 Thus, we can grant that in any translation, whatever the language concerned, there will be inevitable loss of melody and evocative power. However matters are, as we shall see, even more complicated when it comes to Arabic.

First, of course, the majority of Muslims are not Arabs or Arabicspeaking peoples. The non-Arabic speaking nations of Indonesia, with a population of 197 million; Pakistan, with 133 million; Iran, with 62 million; Turkey, with 62 million; and India, with a Muslim population of about 95 million, outnumber by far the total number of native Arabic speakers in about thirty countries in the world, estimated as 150 million. Many educated Muslims whose native tongue is not Arabic do learn it in order to read the Koran; but then again, the vast majority do not understand Arabic, even though many do learn parts of the Koran by heart without understanding a word.

In other words, the majority of Muslims have to read the Koran in translation in order to understand it. Contrary to what one might think, there have been translations of the Koran into, for instance, Persian, since the tenth or eleventh century, and there are translations into Turkish and Urdu. The Koran has now been translated into over a hundred languages, many of them by Muslims themselves, despite some sort of disapproval from the religious authorities.3

Even for contemporary Arabic-speaking peoples, reading the Koran is far from being a straightforward matter. The Koran is putatively (as we shall see, it is very difficult to decide exactly what the language of the Koran is) written in what we call Classical Arabic (CA), but modern Arab populations, leaving aside the problem of illiteracy in Arab countries,4 do not speak, read, or write, let alone think, in CA. We are confronted with the phenomenon of diglossia,5 that is to say, a situation where two varieties of the same language live side by side. The two variations are high and low. High Arabic is sometimes called Modern Literary Arabic or Modern Standard Arabic; is learned through formal education in school, like Latin or Sanskrit; and would be used in sermons, university lectures, news broadcasts, and for mass media purposes. Low Arabic, or Colloquial Arabic, is a dialect native speakers acquire as a mother tongue, and is used at home conversing with family and friends, and also in radio or television soap operas. But, as Kaye points out, "the differences between many colloquials and the classical language are so great that a fallah who had never been to school could hardly understand more than a few scattered words and expressions in it without great difficulty. One could assemble dozens of so-called Arabs (fallahin) in a room, who have never been exposed to the classical language, so that not one could properly understand the other."6

In the introduction to his grammar of Koranic and Classical Arabic, Wheeler M. Thackston writes, ". . . the Koran established an unchanging norm for the Arabic language. There are, of course, certain lexical and syntactic features of Koranic Arabic that became obsolete in time, and the standardization of the language at the hands of the philologians of the eighth and ninth centuries emphasized certain extra-Koranic features of the Arabic poetic koine while downplaying other, Koranic usages; yet by and large not only the grammar but even the vocabulary of a modern newspaper article display only slight variation from the established norm of classicized Koranic Arabic." 7

Though he does allow for some change and decay, Thackston it seems to me, paints a totally misleading picture of the actual linguistic situation in modern Arabic-speaking societies. He implies that anyone able to read a modern Arabic newspaper should have no difficulties with the Koran or any Classical Arabic text. Thackston seems totally insensitive "to the evolution of the language, to changes in the usage and meaning of terms over the very long period and in the very broad area in which Classical Arabic has been used."8 Anyone who has lived in the Middle East in recent years will know that the language of the press is at best semilit- erary,9 and certainly simplified as far as structure and vocabulary are concerned. We can discern what would be called grammatical errors from a Classical Arabic point of view in daily newspapers or on television news. This semiliterary language is highly artificial, and certainly no one thinks in it. For an average middle-class Arab it would take considerable effort to construct even the simplest sentence, let alone talk, in Classical Arabic. The linguist Pierre Larcher has written of the "considerable gap between Medieval Classical Arabic and Modern Classical Arabic [or what I have been calling Modern Literary Arabic], certain texts written in the former are today the object of explanatory texts in the latter." He then adds in a footnote that he has in his library, based on this model, an edition of the Risala of Shafi9i (died 204/820) that appeared in a collection with the significant title Getting Closer to the Patrimony. 10

As Kaye puts it, "In support of the hypothesis that modern standard Arabic is ill-defined is the so-called `mixed' language or `Inter-Arabic' being used in the speeches of, say, President Bourguiba of Tunisia, noting that very few native speakers of Arabic from any Arab country can really ever master the intricacies of Classical Arabic grammar in such a way as to extemporaneously give a formal speech in it."I I

Pierre Larcher12 has pointed out that wherever you have a linguistic situation where two varieties of the same language coexist, you are also likely to get all sorts of linguistic mixtures, leading some linguists to talk of triglossia. Gustav Meiseles 13 even talks of quadriglossia: between Literary Arabic and Vernacular Arabic, he distinguishes a Substandard Arabic and an Educated Spoken Arabic. Still others speak of pluri- or multi- or polyglossia, viewed as a continuum.14

The style of the Koran is difficult, totally unlike the prose of today, and the Koran would be largely incomprehensible without glossaries, indeed, entire commentaries. In conclusion, even the most educated of Arabs will need some sort of translation if he or she wishes to make sense of that most gnomic, elusive. and allusive of holy scriptures, the Koran.


According to Barbara F. Grimes15 of the Summer Institute of Linguistics, there are 6,703 living languages in the world. 16 These living languages, and the dozens of extinct languages whose structure are known and have been studied, are classified either typologically, that is, in terms of their structural properties (for example, according to the number and kinds of vowels they use, or according to the order of the subject, verb, and object in a simple sentence)17-or genetically-that is, on the basis of common origin.

Genetically related languages have developed or evolved from a common parent language. As scholar I. M. Diakonoff put it, "The only real criterion for classifying certain languages together as a family is the common origin of their most ancient vocabulary as well as of the word elements used to express grammatical relations. A common source lan guage is revealed by a comparison of words from the supposedly related languages expressing notions common to all human cultures (and therefore not as a rule likely to have been borrowed from a group speaking another language) and also by a comparison of the inflectional forms (for tense, voice, case, or whatever)."18


All the world's languages are classified into large groups or phyla (sometimes very loosely called "families"). Merrit Ruhlen19 classifies all languages into twenty independent groups, each group containing genetically related languages. Arabic belongs to the group (or family) now called Afro-Asiatic, though formerly it was called Hamito-Semitic, Semito-Hamitic, or even Erythraean. This family of genetically related languages can be subdivided into six primary branches, all descendants of the original parent language, namely: (1) Ancient Egyptian (from which Coptic, the liturgical language of the Monophysite Christians of Egypt, is descended); (2) Berber (widespread in Morocco and Algeria); (3) Chadic; (4) Omotic; (5) Cushitic; and (6) Semitic. Arabic, like Hebrew, Syriac, and Aramaic, is a Semitic language. The Semitic languages are further subdivided, sometimes into four groups and sometimes into two. I have chosen Robert Hetzron's and Merrit Ruhlen's classification (see language tree appendix D and E), which divides Semitic languages into two groups. As one can see, Arabic belongs to the Central Semitic group, which is further subdivided into two subgroups, Aramaic and AraboCanaanite (sometimes rather confusedly called South-Central Semitic; I have avoided this term to underline that Arabic does not belong to the same subgroup as South Semitic, containing Epigraphic South Arabian, Modern South Arabian, and Ethiopian or Ethiopic).

One of the distinctive features of all Semitic languages is the triliteral or triconsonantal root, composed of three consonants separated by vowels. The basic meaning of a word is expressed by the consonants, as well as different shades of this basic meaning are indicated by vowel changes, as well as prefixes and as suffixes. For example, the root ktb refers to writing, and the vowel pattern -a-i implies "one who does some thing"; thus katib means "one who writes"; kitab means "book"; maktub, "letter"; and kataba, "he wrote." The two genders, masculine and feminine, are found in Semitic languages, the feminine often indicated by the suffixes -t or -at. The Semitic verb is distinguished by its ability to form from the same root a number of derived stems that express new meanings based on the fundamental sense, such as passive, reflexive, causative, and intensive.20 The close relationship of the languages to one another in the Semitic family is attested by the persistence of the same roots from one language to another-slm, for example, means "peace" in Assyro-Baby- lonian, Hebrew, Aramaic, Arabic, and other languages.

Aramaic is the name of a group of related dialects once spoken, by various Aramaean tribes, for centuries in what is Syria today. There is evidence for it since the beginning of the first millennium B.c.E. As the Aramaeans moved into Assyria and Babylonia, their language spread to all of the Near East, replacing Akkadian, Hebrew, and other languages, eventually becoming the official language of the Persian Empire. In this period it is spoken of as Imperial Aramaic. Aramaic was itself replaced by Arabic after the rise of Islam in the seventh century C.E. Large parts of the biblical books of Ezra (Ezra 4:8-6:18; Ezra 7:12-26) and Daniel (Dan. 2:4b-7:28), and smaller parts of Genesis (Gen. 31:47) and Jeremiah (Jer. 10:11) are in Aramaic. Jesus' native tongue was Palestinian Aramaic; some words of Jesus in the New Testament (e.g., "Talitha cum" in Matt. 5:41) are Aramaic. On the cross, Jesus is said to have quoted Psalm 22:1 in Aramaic.

The Babylonian Talmud was written in Eastern Aramaic, a language close to Syriac, the language of the Christian city of Edessa (until the thirteenth century C.E.), still the liturgical language of the Nestorian and Jacobite Christian Churches.21

Edessa was an important center of early Christianity in Mesopotamia. (These early Christians gave the Greek name "Syriac" to the Aramaic dialect they spoke when the term "Aramaic" acquired the meaning of "pagan" or "heathen.") Edessene Syriac rapidly became the literary language of all non-Greek Eastern Christianity, and was instrumental in the Christianization of large parts of central and south-central Asia. Despite the fifth-century schism between the monophysite Jacobite Church in Syria and the Nestorian Church of the East, Syriac remained the liturgical and theological language of both these national churches. Syriac is still the classical tongue of the Nestorians and Chaldeans of Iran and Iraq, and the liturgical language of the Jacobites of Eastern Anatolia and the Maronites of Greater Syria. Missionary activity spread the Syriac language and script to India and Mongolia, and rather surprisingly, even the Mongolian script, though written vertically, is derived from the Syriac script.22

The importance of Syriac literature for our understanding of the rise of Islam was discussed by A. Mingana, J. B. Segal, Sebastian Brock, and Claude Cahen, and, of course, by Patricia Crone and Michael Cook.23 But Syriac also played an important role as an intermediary through which Greek learning and thought passed into the emerging Islamic civilization, since it was Syriac-speaking scholars who first translated late Hellenistic science and philosophy from Syriac into Arabic at the Dar al-Hikma in Baghdad.24 Other scholars such as Mingana, Margoliouth, and now Luxenberg want to argue that Syriac greatly influenced not only the vocabulary of the Koran, but its theological and philosophical ideas. How this happened is not yet clearly understood.

The oldest Syriac script, which dates back to the first century C.E., evolved from the Aramaic alphabet,25 which is also the ancestor of Arabic writing. Perhaps I should add here that in Hebrew, Arabic, and Syriac writing, vowels were at first omitted; symbols to indicate the vowels probably date from only the eighth century C.E.


The Arabic language, like any other language-and we must not forget that Arabic is like any other language, especially those in the Semitic group of the Afro-Asiatic family-has a history. It did not appear fully fledged out of nowhere, but slowly evolved over a period of time. Little is known about Old or Proto-Arabic. Early Arabic is the name given to the period from the third to sixth century C.E. "when over a large part of Arabia dialects quite distinct from Old Arabic, but approaching Classical Arabic were spoken, and during which Classical Arabic itself must have evolved."26 Hundreds of Aramaic loanwords entered the language during this period, through Jewish and Christian contacts.27

The earliest Arabic texts seem to have been Christian inscriptions, suggesting that the Arabic script was invented by Christian missionaries probably at Hira or Anbar.

It is probable that at least partial Bible translations into Arabic existed before Islam. Stylistic reminiscences of the Old and New Testaments are found in the Koran. A. Baumstark claimed a pre-Islamic date for the text of some Arabic Bible manuscripts. There is also a fragment of the Psalms in Arabic in Greek characters. Examination of this and two of Baumstark's texts shows a language slightly deviating from Classical Arabic towards the colloquials. This is typical for Christian-Arabic literature, for early papyri and for the language of scientific writing; it may be early colloquial influence, but also Classical Arabic not yet standardized by grammarians....

Wellhausen plausibly suggested that Classical Arabic was developed by Christians at al-Hira. Muslim tradition names among the first persons who wrote Arabic Zayd b.Hamad (ca. 500 A.D.) and his son, the poet 'Adi, both Christians of Hira. 'Adi's language was not considered fully fasi h, which may be taken as meaning that Classical Arabic was still in course of evolution.28

What we know as Classical Arabic was academically, and some would say artificially (because of its almost too perfect algebraic-looking grammar [root and pattern morphology]),29 standardized between the third Muslim/ninth Christian and fourth Muslim/tenth Christian centuries. "Its grammar, syntax, vocabulary and literary usages were clearly defined under systematic and laborious research." 30 We shall return to the issue of the evolution of Classical Arabic later.

Arabic words fully exhibit the typical Semitic word structure already mentioned (see above). An Arabic word is composed of the root of usually three consonants, providing the basic lexical meaning of the word, and the pattern, which consists of vowels and gives grammatical meaning to the word. This feature has been a positive boon to Muslim commentators, who have shown real genius in their inventivesness when confronted with an obscure word in the Koran in need of elucidation. They would often simply turn to the dictionary meaning of the root of an obscure word and try to employ an etymological interpretation of the word.31

Arabic "also makes use of prefixes and suffixes, which act as subject markers, pronouns, prepositions, and the definite article."

Verbs in Arabic are regular in conjugation. There are two tenses: the perfect, formed by the addition of suffixes, which is often used to express past time; and the imperfect, formed by the addition of prefixes and sometimes containing suffixes indicating number and gender, which is often used for expressing present or future time. In addition to the two tenses there are imperative forms, an active participle, and a verbal noun. Verbs are inflected for three persons, three numbers (singular, dual, plural), and two genders. In Classical Arabic there is no dual form and no gender differentiation in the first person, and the modern dialects have lost all dual forms. The classical language also has forms for the passive voice.

There are three cases (nominative, genitive, and accusative) in the declensional system of Classical Arabic nouns; nouns are no longer declined in the modern dialects. Pronouns occur both as suffixes and independent words.32

Arabic, also like any other world language, has its peculiar strengths and weaknesses. For Bernard Lewis, Classical Arabic is a precise and accurate vehicle of thought, a language of remarkable clarity and an almost peerless instrument of philosophical and scientific communica- tion.33 While according to Rabin, "Classical Arabic had an extremely rich vocabulary, due partly to the bedouin's power of observation and partly to poetic exuberance; some of the wealth may be due to dialect mixture. It was not rich in forms or constructions, but sufficiently flexible to survive the adaptation to the needs of a highly urbanised and articulate culture without a disruption of its structure."34 Here is how A. Schaade assesses the strengths of Arabic: "Comparing it first of all with the other Semitic tongues we notice that the possibilities of syntactic distinctions are in Arabic developed to a far greater extent and brought out with greater precision than in any of the others. Where other languages have to content themselves with simple co-ordination, Arabic commands a large number of subordinating conjunctions."35

Looking at the limitations of Arabic, Shabbir Akhtar, who taught for three years at the Malaysian Islamic University, contradicts Lewis: "In modern analytical philosophy, there is hardly anything in Arabic or any other Islamic tongue. Philosophical discussion is best conducted in English. Owing to the grammatical limitations of Arabic, it is impossible to express most philosophical claims with an acceptable degree of rigour and clarity. Moreover, Arabic is a devotional language lacking the vocab ulary requisite for detached discussion of controversial matters."36 Lewis and Akhtar are, of course, talking of two different historical periods; for Lewis does add the caveat, ". . . [Arabic's] only peer until modern times was Greek."37 Furthermore, Akhtar qualified his remarks a few months later, "... I concede that the attack on ... on the suitability of Arabic for philosophical discussion was unfair. Arabic, like Hebrew, has the capacity to generate novel words and expressions from existing roots ...... 38

Schaade points to other limitations:

In one respect however Classical Arabic as well as its sisters compares unfavourably with the Aryan languages: while for the noun it has created a great number of subtle distinctions which enable it to express even the most abstract concepts, the development of the verb has been one-sided. We seek in vain for a distinction between inchoative and per- mansive forms of expression: gama means "he was standing" and "he rose." Similarly the different grades of the simple meaning of the verb which we render by means of various auxiliary verbs, are frequently left unexpressed: yagra'u "he reads" and "he is able to read." The expression of the tenses also often lacks precision, in spite of the development of a number of verbal exponents with a temporal force (qad, kana, sawfa, etc.).39


What was the nature of Arabic before and after the rise of Islam, particularly between the third and sixth centuries, and then between seventh and ninth centuries? When did the break between the spoken and written language (the phenomenon of diglossia) take place? Out of what and when did Classical Arabic develop? In what language was the Koran written?

Let us begin with the last two questions. According to Muslims, the Koran was written in the dialect of the Quraysh of Mecca, and CA was born out of the Meccan dialect, which was considered the linguistic norm. The language of the Koran, which is identical to the poetical koine, is one of the two bases of CA; Muhammad, being from Mecca, could only have received the revelation in his original dialect, that of the Quraysh.

Noldeke seems to accept the traditional Muslim view that the Koran and pre-Islamic poetry (poetical koine) were the two sources of CA, and that the Koran was written in the Meccan dialect: "For me it is highly unlikely that Muhammad in the Koran had used a form of language absolutely different from the usual one in Mecca, that he would have used case and mood inflexions if his compatriots had not used them."40

However there are a certain number of objections to the Muslim view. First, it is unlikely that there existed a linguistic norm. Mecca, being an important commercial town and center of pilgrimage, must have been open to the linguisitic influence of the Yemen, Syria, and Najd. Second, Muhammad's preaching had at least Pan-Arab pretensions, but these pretensions would seem hardly realizable if he was using only his local dialect. Surely Muhammad's preaching in the urban language of Mecca would have had no meaning for the nomads, whose language was considered more prestigious.

For some Western scholars, like Blachere,41 CA was derived from pre-Islamic poetry and the language of the Koran. But for Blachere, the language of the Koran has nothing to do with the dialect of Mecca, but is rather the language of pre-Islamic poetry (the so-called poetical koine). As Schaade put it, "The earliest specimens of classical Arabic known to us are found in the pre-Islamic poems. The problem arises how the poets (who for the most part must have been ignorant of writing) came to possess a common poetical language,-either (perhaps with the object of securing for their works a wider field of circulation?) they used for their purposes a language composed of elements from all the different dialects, such as may have been created by the necessities of trade, and which it only remained for them to ennoble, or the dialect of any particular tribe (perhaps owing to political circumstances?) achieved in pre-historic times special pre-eminence as a language of poetry."42

Blachere certainly accepts the idea that diglossia is an old phenomenon going back to pre-Islamic times. That is to say, scholars like Blachere, Vollers,43 Wehr,44 and Diem,45 believe that the poetical koine, the language of pre-Islamic poetry, was a purely literary dialect, distinct from all spoken idioms and supertribal. This situation, in which two varieties, literary and spoken, of the same language live side by side, is called diglossia. Other scholars, like N6ldeke,46 Fiick,47 and Blau,48 agree with the traditional Arab view that diglossia developed as late as the first Islamic century as a result of the Arab conquests, when non-Arabs began to speak Arabic.

Karl Vollers upset many people when he argued at the beginning of the twentieth century that the Koran was written, without i`rab, inflection, or case endings, in a dialect of Najd, and was a result of editing and emendation carried out long after Muhammad with a view to harmonizing the sacred text with the language of so-called pre-Islamic poetry, which is that of Najd. Vollers is certain that the Koran as we have it today is not linguistically the revelation as it was received by Muhammad. One must take into account the numerous phonetic variants preserved in the commentaries and special treatises. These variants of a dialectal origin attest to the contrast between the speech of the Hijaz and that of Najd. The Koran preserves everywhere certain linguistic features maintained in Najd and on the way to disappearance in the Hijaz, according to Muslim grammarians; thus, the Koran represents the speech of Najd. The Koran is the result of adaptation, and issues from the emendations of the text by readers of Najdian atavism or influenced by the nomadic dialects of this region. As to the linguistic identity of the Koran and pre-Islamic poetry, it is explained by the fact that Muslim scholars unified them one by the other during the course of the establishment of the grammar. Vollers concludes that the Koran and pre-Islamic poetry are truly the two sources of CA, but with this reservation that the Koran is an adaptation of the Hijazi dialect to the norms of the poetical language.

Blachere contended that Vollers made too much of the putative contrast between the western dialect and eastern dialect. The contrast between the Hijaz and Najd is not as clear-cut as Vollers makes out. Vollers also seems to accept certain linguistic features as true of the time of Muhammad, but which, in reality, were the creations of much later Muslim philologists. If there had been harmonization of the Koranic text with the dialects of Najd, one would expect to find the essential character of these dialects, the taltala. One would find traces of this adaptation in the vocabulary and syntax.

Wansbrough has his own reasons for rejecting Vollers's theory: "The basic error lay in Vollers' adherence to an arbitrary and fictive chronology, though that may have been less important than his contention that the refashioned language of scripture could be identified as the CA of the Arabic grammarians. Neither from the point of view of lexicon nor from that of syntax could the claim be justified."49 In other words, the language of the Koran is not Classical Arabic.

However, Vollers's theory was revived in 1948 by Paul Kahle (chap. 3.3), who sees in a saying of al-Farra' promising reward to those reciting the Koran with i'rdb support for Vollers's view that the original Koran had no i'rdb.50

Corriente also makes the point in his classic papers' that the language of the Koran is not CA. For Corriente, CA was standardized by the grammarians in eighth and ninth centuries C.E., on the whole depending on a central core of Old Arabic dialects as koineized in pre-Islamic poetry and rhetoric, and the speech of contemporary Bedouins. Grammarians did not invent the i'rab system, which must have existed in the texts they edited. (I`rab is usually translated as "inflexion," indicating case and mood, but the Arab grammarians define it as "the difference that occurs, in fact or virtually, at the end of a word, because of the various antecedents that govern it.")52 They did come with their preconceptions about what constituted good Arabic, but they nonetheless respected what they learned from their Bedouin informants in order to standardize the language, and thus fix what came to be CA. However, some did reject certain utterances of the Bedouins as being incorrect.

Koranic Arabic is structurally intermediate between OA koine and Eastern Bedouin Arabic and Middle Arabic, and, of course, the Koran cannot have been written in CA since this was only finally standardized over a period of time during the eighth and ninth centuries.

Native tradition identifies two groups of dialects, Ancient West and East Arabian, neither of them identical to the OA koine. Corriente adds a third kind of Arabic, Nabataean, the immediate forerunner of the Middle Arabic of Islamic cities. It was very widespread indeed.

Finally, Corriente calls attention to the fact that Bedouin vernaculars themselves must also have been undergoing change under various socio- lingusitic pressures, a point perhaps overlooked by the romanticization of Bedouin speech by overeager Muslim grammarians.

All the above accounts rest on a number of assumptions that are not always either spelled out or subjected to rigorous questioning. For example, all our knowledge about the early dialects of Najd, the Hijaz, and the highland area of the southwest seems to have been gathered during the second and third Islamic centuries, when these dialects were already declining. Much of our data are preserved only in late works whose sources we cannot check.53 Second, these accounts also accept without hesitation the traditional Muslim chronology and the accounts of the compilation of the Koran. The first scholar in modern times to radically question these accounts is, of course, John Wansbrough, who wrote:

To draw from the same data conclusions about the origins and evolution of CA involves implicit acceptance of considerable non-linguisitic material often and erroneously supposed to be "historical fact." I refer to such assumptions as that of the isolation of speakers/writers of Arabic within the Arabian peninsula up to the seventh century, or that of the existence of ne varietur text of the Islamic revelation not later than the middle of the same century.54

Wansbrough points out that the Muslim accounts of the origins of CA have as their aim the establishment of the Hijaz as the cradle of Islam, in particular Mecca, and in the polemical milieu of eighth century C.E. Near East, to establish an independent Arab religious identity, with a specifically Arabic Holy Scripture.

Suppression of claims made on behalf of other tribal groups to the title afsah al-`arab [the most eloquent of the Arabs] is symbolized in the account ascribed to Farra' of how the inhabitants of cosmopolitan (!) Mecca (i.e. Quraysh) were in a position to recognize and adopt the best ingredients from each of the bedouin dialects in Arabia.55 Besides drawing attention to the role of Mecca as cultic and commercial center, this tradition, like the ones it eventually replaced, served to identify the northern regions of the Arabian peninsula as the cradle of CA at a date prior to the proclamation of Islam.56

Nor can we uncritically accept Muslim claims that the language spoken by the bedouins must be identical with that of the poetry called pre-Islamic. The bedouins were hardly disinterested referees. But more important, "for our purposes it is well to remember that the written record of transactions between bedouin and philologist dates only from the third/ ninth century, and is thus coincident with the literary stabilization of both Quranic exegesis and Muslim historiography."57

The polemical importance of "pre-Islamic poetry" for Muslims is also well explained by Wansbrough:

Whatever may have been the original motives for collecting and recording the ancient poetry of the Arabs, the earliest evidence of such activity belongs, not unexpectedly, to the third/ninth century and the work of the classical philologists. The manner in which this material was manipulated by its collectors to support almost any argument appears never to have been very successfully concealed. The procedure, moreover, was common to all fields of scholarly activity: e.g. the early dating of a verse ascribed to the mukhadrami poet Nabigha Ja`di in order to provide a pre-Islamic proof text for a common Quranic construction (finite verb form preceded by direct object), Mubarrad's admitted invention of a 'Jahili' [pre-Islamic] verse as a gloss to a lexical item in the hadith, and Abu 'Amr b. 'Ala's candid admission that save for a single verse of 'Amr b.Kulthum, knowledge of Yawm Khazaz would have been lost to posterity. The three examples share at least one common motive: recognition of pre-Islamic poetry as authority in linguistic matters, even where such contained non-linguistic implications. Also common to all three is another, perhaps equally significant feature: Ibn Qutayba, who adduced the verse of Nabigha to explain/justify Quranic syntax, lived at the end of the third/ninth century, as did Mubarrad; Abu `Amr, of whom no written works were preserved, lived in the second half of the second/eighth century, but this particular dictum was alluded to only in Jahiz (third/ninth century) and explicitly in Ibn 'Abd Rabbih (fourth/tenth century). Now, that pre-Islamic poetry should have achieved a kind of status as linguistic canon some time in the third/ninth century may provoke no quarrel. That it had achieved any such status earlier must, I think, be demonstrated. The fact that it had not, in one field at least, can be shown: the absence of poetic shawahid in the earliest form of scriptural exegesis might be thought to indicate that appeal to the authority of Jahili (and other) poetry was not standard practice before the third/ninth century. Assertions to the contrary may be understood as witness to the extraordinary influence exercised by the concept of fasahat al jahiliyya.58

In other words, the putative eloquence of pre-Islamic poetry became commonplace only in the third/ninth century; there are no references to preIslamic poetry in the early, pre-third-century works of Koranic exegesis.

There are even a number of scholars, such as Alphonse Mingana59 and D. S. Margoliouth,60 who think that all pre-Islamic poetry is forged, inspired by Koranic preoccupations. The Egyptian Taha Husayn, in Of Pre-Islamic Literature, 61 the second of his two famous books, concludes that most of what we call pre-Islamic literature was forged, though he seems to accept the authenticity of some poems, albeit a tiny number. This cautious acceptance of some pre-Islamic poetry as authentic seems to have been shared by several Western scholars, such as Goldziher, Tor Andrae, W. Marcais, and Tritton, who reject the total skepticism of Margoliouth, but shy away from the too generous credulity of Noldeke and Ahlwardt.62 Of course, if all pre-Islamic poetry is forged, then there was no such thing as a poetical koine, and the language of the Koran obviously could not owe anything to this fictive poetical language. We would have to look elsewhere for the origins of the language of the Koran.

If the Koran did not originally have i`rab, then the present rhyme scheme63 to be found in the Koran must be a later addition, since rhyme depends on i`rab, and the changes required in the Koranic text must have been considerable. The lack of original i'rab in the Koran, if true, also suggests that there is less of a relationship between poetry and the Koran than previously thought, and that the text of the Koran is primary.


Reading the Koran on its own terms, trying to interpret it without resorting to commentaries, is a difficult and questionable exercise because of the nature of the text-its allusive and referential style and its grammatical and logical discontinuities, as well as our lack of sure information about its origins and the circumstances of its composition. Often such a reading seems arbitrary and necessarily inconclusive.

G. R. Hawting

Ironically, far from increasing our understanding of the contents, as devout Muslims would have us believe, a look at the Koran in the original Arabic only increases the confusion. As Gerd-R. Puin said, "The Koran claims for itself that it is `mubeen' or `clear.'65 But if you look at it, you will notice that every fifth sentence or so simply doesn't make sense... . The fact is that a fifth of the Koranic text is just incomprehensible ...."66

As Hirschfeld once remarked, Why would the Koran need to superfluously repeat that it is written in clear or plain Arabic three times, if it had really been written in plain Arabic?67 Of course, there is much in it that is not Arabic at all, both in terms of the vocabulary, subject matter, and inspiration; further sources of obscurities are not only the large number of foreign words but the "new meanings pressed into service."68

Muslim scholars themselves are aware of the difficulties and obscurities of their sacred text. Fuat Sezgin lists no less than eighteen treatises by Muslim philologists, such as Aban b.Taghlib (died 758) and Niftawayh (died 859), for the period between the mid-eighth century and the midninth century entitled Gharib al-Qur'an, The Rare [i.e., strange] Expressions of the Quran.69

Muslim Exegetes divide the words of the Koran into four classes:70 Khass, words used in a special sense; 'Amm, collective or common; Mushtarak, complex words that have several meanings, and Mu'awwal, words that have several meanings, all of which are possible, and thus require a special explanation.

As an example of the latter class of words, Mu'awwal, we could look at two differing translations of Sura CVIII.2:

Sale: Wherefore pray unto thy Lord; and slay (the victims).

M. Ali: So pray to thy Lord and sacrifice.

The word translated "slay" is in Arabic inhar, from the root nahr, which has several meanings. The Hanafites, followers of Abu Hanifa (700-767) a leading fiqh scholar and theologian, translate it as "sacrifice adding the words "the victims" in parenthesis. However, the followers of Ibn Ash-Shafili (767-820) founder of the school of law named after him, say it means "placing the hands on the breast in prayers."71

The sentences ('Ibarah) of the Koran are divided into two classes, Zahir, obvious, and Khafi (or batin), hidden. Let us look at just the latter class. Khafi sentences are either Khaji, Mushkil, Mujmal, or Mutashabih.

Khaji sentences contain words that are understood to have hidden beneath their literal meaning a reference to other things or persons. The word "thief," sariq, for instance, has as its hidden references both pickpockets and highway robbers. Mushkil sentences are ambiguous, and hence, their meanings are very difficult to ascertain. Mujmal sentences may have a variety of interpretations, owing to the words in them being capable of several meanings. In this case, it is the tradition (hadith) that settles the meaning and must be accepted. Mujmal sentences may also contain rare words whose meaning is not at all clear. Hughes gives the following example of the first kind of mujmal sentence: "Stand for prayer (salat) and give alms (zakat)." Both salat and zakat are Mushtarak words. Muslims had recourse to tradition (hadith) for an explanation. According to the appropriate hadith, Muhammad explained that salat might mean the ritual of public prayer, standing to say the words "God is Great," or standing to repeat a few verses of the Koran; or it might mean private prayer. Whereas, zakat comes from the root word meaning to grow, zaka. Muhammad, "however, fixed the meaning here to that of `alsmgiving,' and said, `Give of your substance one-fortieth part.' "72

Mutashdbih sentences are "intricate" sentences, or expressions whose meaning is impossible for man to ascertain, though it was known to Muhammad. As Patricia Crone puts it,

The Qur'an is generally supposed to have originated in a social, cultural and linguistic environment familiar to the early commentators, whose activities began shortly after Muhammad's death and many of whom were natives of the two cities in which he had been active; yet they not infrequently seem to have forgotten the original meaning of the text. It is clear, for example, that they did not remember what Muhammad had meant by the expressions jizya an yad, al-samad, kalala or ilaf; indeed, the whole of Sura 106 (Quraysh) in which the ilaf occurs, was as opaque to them as it is to us; and the same is true of the so-called "mysterious letters." Kalala is a rather unusual case in that several traditions (attributed to `Umar) openly admit that the meaning of this word was unknown; more commonly the exegetes hide their ignorance behind a profusion of interpretations so contradictory that they can only be guesswork.

"It might," as Rosenthal observes, "seem an all too obvious and unconvincing argument to point to the constant differences of the interpreters and conclude from their disgreement that none of them is right. However, there is something to such an argument." There is indeed. Given that the entire exegetical tradition is characterized by a proliferation of diverse interpretations, it is legitimate to wonder whether guesswork did not play as great a role in its creation as did recollection; but the tradition is not necessarily right even when it is unanimous.73

The Koran itself admits to its own ambiguous passages whose meaning is known only to God: Sura 11.7: "It is He who has revealed to you the Book. Some of its verses are precise in meaning-they are the foundation of the book-and others ambiguous. Those with an evil inclination in their heart seek after what is unclear in it, wishing to trouble people's minds and wishing to interpret it. But no one but God knows its interpretation. Those who are firmly rooted in knowledge say: `We believe in it; it is all from our Lord.' "

We also have the curious phenomenon of a word that can have two contradictory meanings. For instance, at Koran XX.15: 'inna -s- sd`ata 'atiyatun 'akadu 'ukhfiha lituj za kullu nafsim bima tas'd.

Khafa is said to have the two opposite meanings, "to be hid" and "to reveal." M. Ali translates verse 15 above as: "Surely the Hour is coming-I am about to make it manifest-so that every soul may be rewarded as it strives."74

Pickthall has: "Lo! the Hour is surely coming. But I will to keep it hidden, that every soul may be rewarded for that which it striveth."75

I have gone through Bell's splendid two-volume commentary on the Koran, and have noted some of his comments and judgments on the various difficulties and obscurities of sense and reference. However, I have confined my search mainly to Sura II. I have also referred to Jeffery,76 Penrice's Dictionary,77 Lane's Arabic-English Lexicon,77 Blachere's French translation of the Koran,79 and the two articles in the present anthology by Margoliouth and Mingana.8° I have classified the difficulties into five fairly loose and sometimes overlapping categories in this way, and, of course, the lists make no pretensions of being complete:

[6.1 ] Individual words whose meaning is not certain.

[6.2] Phrases or sentences whose meaning is not clear, and passages whose reference is not clear (who or what putative historical event they refer to).

[6.3] Passages and words that are thought to be interpolations, insertions, or evidence for revisions.

[6.4] Sentences containing grammatical errors from the Classical Arabic point of view.

[6. 5] Phrases, sentences, and verses that do not seem to fit the context, and thus must have been transposed. These transposed or displaced verses are responsible for the disorder and incoherence that abounds in the Koran.

[6.1] Individual words whose meaning is not certain.

[6.1.1 ] LXXX.28. Qadb: meaning not certain, probably "green herbs" of some kind.

[6.1.2] LXXX.31. 'Abb: meaning not certain, probably "pasture." Cf. Hebrew: ebh; Syriac: 'ebba'; as Jeffery notes, "The early authorities in Islam were puzzled by the word as is evident from the discussion by Tabari on the verse, and the uncertainty evidenced by Zamakhshari and Baydawi in their comments, an uncertainty which is shared by the Lexicons (cf. Ibn Manzur, Lisan al-Arab, 20 vols. Cairo: A. H., 1300-1308 i, p. 199; Ibn al-Athir, Al- Nihaya ft gharib al-hadith, 4 vols. Cairo: A. H, 1322, i,10)."81

[6.1.3] IV.51. Jibt: no explanation has been found. As Jeffery observes, "the exegetes knew not what to make of it, and from their works we can gather a score of theories as to its meaning, whether idol, or priest, or sorcerer, or sorcery, or satan, or what not."82

[6.1.4] LXIX.36. Ghislin: according to Blachere,83 the Muslim exegetes do not know the meaning of this term. However, most translators, including Bell, seem to follow Ibn al-Kalbi in interpreting it as "what exudes from the bodies of the inmates of the Fire (i.e. Hell)." Blachere finds this unacceptable because of the use of the word ta`amun at the beginning of the verse, which reads, "Not any other food (= ta`amun) but ghislin." Ta`amun usually indicates solid food. Blachere thinks the word is of foreign origin.

[6.1.5] LXXXIX.7. Iram: as Jeffery says, the number of variant readings of this word "suggests of itself that [it] was a foreign one of which the exegetes could make nothing."84 It is perhaps the name of a city or country with which 'Ad was associated; usually taken to be of South Arabian origin. But as Blachere notes, "It is naturally impossible to know what this verse could have meant for Muhammad's generation."85

[6.1.6] XLVI.28. QurbanlQuraban: verse 28: "Did those help them, whom they had taken for qurban [as] gods ['alihatan] to the exclusion of Allah."

The word Qurban as it appears in 111. 182 and V.27 evidently means "sacrifice," but, according to Jeffery, here, at XLVI.28, it means "favorites of a prince."86 For Penrice, this word must be translated " `as a means of access to God,' the false deities there mentioned being sup posed to be on familiar terms with God, and therefore likely to act as intercessors with Him."87

Barth takes the word following qurban, that is, 'alihatan, meaning "Gods" as a gloss on qurban. This seems to be accepted by Wensinck88 and Bell, but is totally rejected by Blachere, though he does not say why. Blachere admits to being completely baffled by this term in this verse.89

Bell adds that another reading, quruban could be taken as a plural of garib, "neighbour," and accordingly his own translation reads: "Why helped them not those whom they had chosen apart from Allah as neighbours gods?" Bell adds as a footnote to neighbors, "i.e. patrons or inter- cessors."90

[6.1.7] 11.62. Sabian: as Bell says, this word has "baffled all investigators." Literally, it may mean "the baptizers." According to Bell, the whole verse is out of place, while Blachere believes that the words "the Christians, the Sabi'in" seem not to belong with the natural flow of the sentence; perhaps they were added later to fill out the expression "those who believe in God ... and do good." Some even hold that "the Sabian" must be a post-Muhammadan interpolation.91 It is unlikely to refer to the Sabaeans of Harran who were pagans and certainly did not practice "baptism," and cannot be considered the people of the Book. Perhaps the Mandaeans, a Judeo-Christian sect practising the rite of baptism, are meant.92

[6.1.8] 11.78. (a) 'ummiyyun

(b) 'amaniyya

(i) Bell: "Some of them are common people ['ummiyyun] who do not know the Book except as things taken on trust ['amaniyya] and who only cherish opinions."

(ii) Blachere: "While among them are the Gentiles who do not know the Scripture only chimaeras, and only make conjectures."

(iii) Dawood: "There are illiterate men among them who, ignorant of the Scriptures, know of nothing but lies and vague fancies."

(iv) Pickthall: "Among them are unlettered folk who know the Scripture not except from hearsay. They but guess."

(v) Muhammad Ali: "And some of them are illiterate; they know not the Book but only (from) hearsay, and they do but conjecture."

(a) 'ummiyyun

Dawood, Pickthall, and Muhammad Ali follow the Muslim tradition in translating 'ummiyyun (plural of 'ummi) as "illiterate," one who neither writes nor reads a writing.93

Bell thinks 'ummiyyan means belonging to the 'ummah or community, while Blachere translates it as "Gentiles," in the sense of "pagan." For the French scholar it is clear that the word ummi designates pagan Arabs, who, unlike the Jews and Christians, had not received any revelation and were thus living in ignorance of the divine law. Tabari does indeed quote some traditions that give this sense to the word ummi: according to Ibn Abbas, "'ummiyyan (refers to) some people who did not believe in a prophet sent by God, nor in a scripture revealed by God; and they wrote a scripture with their own hands. Then they said to ignorant, common people: `This is from God.' 1194 However, Tabari himself does not accept this interpretation, instead gives a totally unconvincing and improbable account of the derivation of this word: "I am of the opinion that an illiterate person is called ummi, relating him in his lack of ability to write to his mother (umm), because writing was something which men, and not women, did, so that a man who could not write and form letters was linked to his mother, and not to his father, in his ignorance of writing. . . ."95

There is even a series of traditions in Ibn SaId96 that show Muhammad himself writing his political testament. However, Muslim orthodoxy translates ummi as "illiterate" for apologetic reasons, to show that the Koran must have been of divine origin since it was revealed to an illiterate, who thus could not have plagiarized, as often accused, the Jewish or Christian scriptures.

(b) 'amaniyya

The meaning of 'amaniyya is not at all clear. For Baydawi it is the plural of 'umniyyah, from the root mny. But Bell prefers to derive it from the root 'mn, giving it the meaning "tradition, dogma, a thing taken on trust."97

[6.1.9] 11.89. yastaftahuna: the sense is not clear.98

[6.1.10] 11.243. 'uluf. probably plural of 'alf, thousand, but possibly an unusual form plural of 'ilf, "intimate friend."99

[6.1.11 1 11.260. sur in sur-hunna: variously pointed, but is usually taken as the imperative of swr, taken here to mean "cause to come," a very unusual meaning of the verb. Blachere translates it, intuitively, as "press" or "squeeze,"too Muhammad Ali as "tame," Arberry as "twist," while Mahmoud Ayoub, relying on Muslim exegetes, translates it as "cut into pieces." Scholars remain puzzled.

[6.1.12] 11.53, 185; 111.3; V11L29, 41; XXI.48; XXV.I: Furgan

First Jeffery:

In all passages save VIII.42, it is used as though it means some sort of a Scripture sent from God. Thus "we gave to Moses and Aaron the Furgan, and an illumination." (xxi.49), and "We gave to Moses the Book and the Furgan" (11.50), where it would seem to be the equivalent of the Taurah [Torah]. In 111.2, it is associated with the Taurah and the Injil [Gospel], and XXV 1, and I1.181, make it practically the equivalent of the Qur'an, while in VIII.29, we read, "if ye believe God, he will grant you a Furgan and forgive your evil deeds." In VIII.42, however, where the reference is to the Battle of Badr, "the day of the Furgan, the day when the two hosts met," the meaning seems something quite different....

The [Muslim] philologers, however, are not unanimous as to its meaning.' 01

Razi in his discussion of 11.53 goes through several possible meanings of the word Furgan:

The Furgan [separator, or that by which things may be distinguished] could be either the Torah as a whole or in part. It may also refer to something other than the Torah, perhaps one of the miracles of Moses, such as his staff, and so forth. It may mean relief and victory, as God said concerning the Apostle, "and what we sent down to our servant on the day of the criterion [Furgan], the day when the two parties met" (Koran VIII.41). The word Furgan may refer to the splitting [infiraq] of the sea, or as some have said, to the Qu'ran, which was also sent down to Moses.

Razi rejects the latter view as a false interpretation. He concludes, "The Furgan is that by which truth may be distinguished from falsehood. Thus it may either be the Torah or something external to it."102

[6.1.13] CV. 3.

'Ababil. Bell accepts without a great deal of enthusiasm'Ababil as the plural of 'ibbalah, meaning "a bundle," "flock." This verse is sometimes translated as "Did He not send against them flocks of birds ... ?" But the sense of this term is not clear, and the word is rare. Kasimirski and Montet see in it a proper name; hence Montet's translation reads, "Did He not send against them the birds Ababil." Lane, referring to al-Akhfash and as-Sijjani as his authorities, explains that verse 3 means "Birds in distinct, or separate, flocks or bevies: [or] birds in companies from this and that quarter: or following one another, flock after flock."103

As Jeffery points out, the long account in Ibn Manzur, Lisan al-Arab (xiii, 5), makes it clear that the philologers did not know what to make of the word.

Some have suggested that the word has nothing to do with birds but is another calamity in addition, connecting the word with smallpox. Whereas Carra de Vaux would take tayran 'Ababil (flock of birds) as a mistaken reading for tir babil, meaning "Babylonian arrows," which caused the destruction of the army. The word is very probably of foreign origin, though this origin is so far unknown.104

[6.1.14] Sijjil: XI.82; XV.74; CV.4.

Tabari and others seem to have derived it from the Middle Persian words sang, meaning "stone," and gil, meaning "mud."

It seems to designate stones resembling lumps of clay, fired or sun- dried,'05 and this is corroborated by sura LI.33-4 ". . . that we may loose on them stones of clay, marked by your lord for the prodigal."

As Tabari tells us, some took it to mean the lowest heaven, others connected it with the word kitab. Bayda wi points to those who took it to be a variant of sijin, meaning hell. More recently, F. Leemhuis106 has argued that sijjil is in origin a non-Semitic, apparently Sumerian word appearing in Akkadian as sikillu or shigillu, denoting a smooth kind of stone found in the Aramaic of Hatra, as sgyl or sgl, with a specialised meaning of "altar stone." From Mesopotamia, it must have entered the various Arabic dialects in Syria and elsewhere, but acquiring the meaning of "hard, flintlike stone."107

[6.1.15] Sijjin: LXXXIII.7,8.

Here is Vacca's account from the first edition of the Encyclopaedia of Islam:

Sijjin, one of the mysterious words of the Koran, "Verily the register of the wicked is surely in Sijjin. And what shall make thee understand what is Sijjin? A book written." Explained by commentators as a place where a record of the deeds of the wicked is kept, and also as that record itself. It is said to be a valley in Hell; the seventh and lowest earth, where Iblis is chained; a rock beneath the earth or the seventh earth; a place beneath Iblis, where spirits of the wicked are; a register comprising the deeds of the wicked, of the djinn and of mankind, or of the devils and unbelievers. Without the article it is a proper name of hell-fire. Also said to mean anything hard, vehement, severe, lasting, everlasting (interpretation influenced by the word's likeness to sijjil, [see above], erroneously connected with the root s -j-1).

Though [al- Suyuti's] Itgan classes it among non-Arabic words, no acceptable etymology is supplied ... ; ... lexicographers give it as a synonym of sijn, prison, and this last word has evidently influenced the prevailing interpretation of Sijjin by Muslim commentators as a place where the record of the wicked is kept, rather than as that record itself. The text of the Koran admits of both interpretations, and most European translators, following Maracci, have preferred the latter.108

[6.1.16] Sijill: XXI.104.

As Jeffery tells us, the meaning of sijill in this passage from XXI. 104 was unknown to the early interpreters of the Koran. Some took it to be the name of an angel, or of the Prophet's amanuensis, but the majority seem to be in favor of its meaning some kind of writing or writing material. Baghawi takes it to be an Arabic word, while others admit that it was a foreign word of Abyssinian or Persian origin. It is, however, derived from the Greek, 6tytX? ov, in Latin, sigillum, used in Byzantine Greek for an imperial edict.109

[6.1.17] kalala: IV. 12b.

The last five or so lines of Sura IV. 12 have been the source of much controversy among Muslim commentators. Tabari devotes seven pages to these few lines. As David Powers tells us: "Almost every word in the opening line of the verse is subject to dispute, and there may be as many as four or five different opinions, espoused by an even greater number of authorities, for every point in question." Powers shows that precise meaning of kalala also remains a subject of controversy, with Tabari citing twenty-seven separate definitions by various authorities. It is not clear if this word kalala refers to the deceased himself (al-mawruth) or to the heirs of the deceased (al-waratha).1io

It is of the greatest consequence as to how one reads this particular verse, and the above example shows that the uncertainties of meaning and the obscurities in the Koran are not a trivial matter. Powers himself gives his own novel interpretation, arguing that kalala was originally a kinship term referring to a female in-law.

[6.2] Phrases or sentences whose meaning is not clear, and passages whose reference is not clear (which historical person or what putative historical event they refer to).

[6.2.1] 11.27. "Who violate the covenant of Allah after making a compact with Him, and separate what Allah hath commanded to be conjoined, and cause corruption in the land; these are the losers."

Bell comments, "what is meant by `separating what Allah hath commanded to be conjoined' is not clear, but it may refer to their rejection of part of the Book (verse 85) or to their rejection of Muhammad while claiming to believe in Allah."

Ibn Kathir, however, explains this verse differently:

The covenant [`ahd ] is either the primordial covenant between God and humanity [mithaq] [Koran VII,172], the measure of the knowledge of God which He has implanted in the minds of human beings as proof against them or the reference may be to the Jews and Christians with whom the Prophet came into contact. That "which God commanded to be joined" means honoring the obligations of blood relationship or any relationship in general. I I 1

[6.2.2] I1.29. "He it is who created for you what is in the earth, as a whole, then straightened Himself up to the heaven, and formed them seven heavens; He doth know everything." As Blachere points out, the plural pronoun "them" in this verse has resisted all explanation.112 It is significant that certain translators find it hard to resist translating this passage as ". . . and He fashioned IT into seven heavens,"113 while others such as Arberry keep closely to the text and translate literally. Tabari gets out of the difficulty by insisting that samd' (heaven) is a collective noun which is to be treated as a plural.114


Verse 101: When a messenger has come to them from Allah con firming what is with them, a part of those to whom the Book has been given cast the Book of Allah behind their backs as if they did not know.

Verse 102: "And follow what the satans used to recite in the reign of Solomon. Solomon did not disbelieve, but the satans disbelieved, teaching the people magic and what had been sent down to the two angels in Babil-Harut and Marut; they do not teach anyone without first saying: `We are only a temptation, so do not disbelieve.' So they learn from them means by which they separate man and wife, but they do not injure anyone thereby, except by the permission of Allah; and they learn what injures them and does not profit, though they know that he who buys it has no share in the Hereafter; a bad bargain did they make for themselves, if they had known."

Verse 103: "If they had believed and acted piously, assuredly, a reward from Allah would have been better, if they had known."

Bell thinks that "what the satans used to recite in the reign of Solomon" may be a reference to the Rabbinic Law. Bell continues, "The mention of Babil may further suggest the Babylonian Talmud. But the whole verse is obscure. It has been extended to undue length by the insertion of clauses designed to obviate misconceptions:

Wa-ma kafara ... as-sihr [and Solomon did not disbelieve ... magic]

Wa-ma yu`allimani ... takfur [and they do not teach ... disbelieve]

Wa -ma hum ... Allah [but they do not ... Allah]

"Finally the verse [102] having perhaps given rise to misconceptions was discarded, and the short verse 103 substituted for it; this is shown by the repetition of the rhyme-phrase."lls

As Ayoub confesses, verse 102 "has been the subject of much controversy. Commentators have disagreed concerning every phrase and even word in it.- 116 I shall give just Tabar3's discussion of the meaning of "and what had been sent down to the two angels in Babil-Hdrut and Mdrut" (wa-ma 'unzila `ala 'l-malakayni bi-babila haruta wa-mdruta), though I suspect the reader will be confused rather than enlightened by the end of it.

Tabari117 gives several opinions about the meaning of ma at the beginning of this passage. According to one, it is a particle of negation, and the corresponding interpretation of this verse is: They follow the sor cery which the satans recited during the reign of Solomon; but neither was Solomon an unbeliever, nor did God send sorcery down to the two angels, rather satans disbelieved and taught sorcery to the people in Babil-Harut and Marut. In this case, Tabari tells us, the two angels are Gabriel and Michael, because Jewish sorcerers falsely claimed that God sent down sorcery to Solomon through Gabriel and Michael, and the Koran denies this; and Harut and Marut are the names of the two men to whom they taught sorcery in Babil. Tabari recounts a second opinion: ma means "that which" (or "what"), and thus Harut and Marut are the names of two angels to whom sorcery-different from that which the satans received-was sent down at Babil.

According to a third opinion ma means "that which," but it refers specifically to the knowledge of how to sunder a man from his wife. A fourth opinion allows ma as both a negative particle and as a relative pronoun. Tabari himself prefers the interpretation of ma as a relative pronoun, and of Harut and Marut as the names of the two previously mentioned angels. Tabari goes through further opinions, though he rejects them since they seem to create further difficulties in interpreting the rest of the verse.118

Many of the commentators took this opportunity to develop the story of Hdrut and Marut for a theological purpose, to prove or emphasize a point of Islamic law. Ibn Kathir and Tabarsi, for instance, give various traditions about Hdrut and Marut, the chief purpose of which seems to have been to show the evils of drinking wine. 119

Western commentators have not been idle, either: Geiger, Sidersky, Horovitz, and Wensinck have tried to show that the Muslims Commentators were inspired variously by the Babylonian Talmud, by an Ethiopian version of the Book of Enoch, and so on. Dumezil traces the origins of these myths to the Indian epic Mahabharata, while de Lagarde identifies Hdrut and Marut as the two secondary divinities associated with the cult of Mazda in the Avesta, the Zoroastrian scriptures: Haurvatat (Integrity) and Ameretat (the Undying).120

[6.2.41 II.114. "But who does greater wrong than those who bar the places of Allah's worship from having the name of Allah remembered in them, and who strive to destroy them? It was not for them to enter them but in fear. For them is (in store) humiliation in this life, and in the Hereafter a mighty punishment."

As Bell says, verse 114 is difficult to understand. Baydawi suggests that it refers to the Romans and their destruction of the temple at Jerusalem, or the Meccans who prevented the Muslims from visiting the Kalbah at the time of the Treaty of Hudaybiyya. It is typical of Muslim commentators to try to find links in such Koranic passages to putative events in the life of Muhammad. Conservative Western Orientalist scholars have followed suit. These attempts rest on vast assumptions about the reliability of the sources on which our knowledge of the rise of Islam is based. But as Lammens and other revisionists have tried to argue, many so-called events in the life of Muhammad were invented to explain obscure and difficult passages in the Koran. Similarly, once the largely fabricated story of the collection of the Koran was accepted by the Muslim scholars, the Muslim commentators set about trying to interpret in greater detail each and every general Koranic passage, amenable to every possible interpretation, within the framework of the traditional story of the rise of Islam, the life of the Prophet, and the compilation of the Koran. Baydawi's suggestions are a prime example of this activity.

But once again we have vastly divergent Muslim interpretations, each supposedly backed up by impeccable isnads (the chain of authority upon which a report is based); thus showing, once again, they did not have a clue as to what the verse really referred to, or what it really meant. Wah idi, relying on al-Suddi and Qatada, claims that Bukhtnassar (Nebuchadnezzar?) destroyed Jerusalem with the aid of some Byzantine Christians. Then, this time depending on the authority of Ibn Abbas, Wah idi reports that this verse was sent down concerning the associators of Mecca when they prevented the Muslims from visiting and worshipping at the Kalbah, perhaps at the time of Hudaybiyya. Nisaburi, also relying on the authority Ibn Abbas, tells us that "the King of the Christians attacked the holy house [the temple at Jerusalem], which he destroyed and desecrated with dead carcasses. He besieged the inhabitants of Jerusalem, killed them, and took their women and children captive. He also burned the books of the Torah. Jerusalem, moreover, remained in ruins until the Muslims rebuilt it during the time of 'Umar ibn al-Khattab. Thus the verse was sent down concerning the sanctuary of Jerusalem." 121

Tabarsi, on the other hand, claims that the people of Quraysh are being referred to. Finally, Bell, a rather conservative Western scholar, who, on the whole, accepted the traditional Muslim account of the rise of Islam, finds the use of the plural masajid, "places of worship," difficult to explain. Bells adds, "The Ka'bah is usually distinguished as al-masjid alharam, and it is doubtful if there was more than one definitely Muslim `mosque' in existence at this time. Masjid, however, is not limited to this, cf. XXII.40, and particularly XVII. 1. The reference might therefore quite well be to Christian churches in Jerusalem. Jerusalem was still the qiblah, but was in Persian hands, the Jews having aided them in its capture. Even this, however, seems far-fetched ...."122

[6.2.5] Here are some more verses where the references are not clear: 11.2: dhalika; I1.6-7; alladhina kafaru; 11.45; 11.80; 11.153-167; those who have done wrong; 11.175; 11.205; 11.210; 11.259.

[6.3] Passages and words that are thought to be interpolations, insertions, or evidence for revisions.

[6.3.1] II.105. According to Bell, the word "idolators" in verse 105 may be a later insertion; the grammar is uneven.

[6.3.2] 11.219. The latter part of this verse, according to Bell, is a formal rhyme-phrase, which was probably added at a much later revision.

[6.3.3] 11.221 ff. Bell observes, "verse 221 f. is not like the surrounding verses, an answer to a question, but it may belong to the same period. If, however, "idolators" includes Jews and Christians the verse must be large, but this is hardly correct. The rhyme clause is again formal and has no doubt been added later." 123

[6.3.4] II.217. The phrase in the middle of the verse wa-l-masjid ... minhu is an insertion from later date, argues Bell, when the duty of pilgrimage had been recognized, and the Meccan opposition was preventing the duty being fulfilled.124

[6.3.5] II.229. The clause 'ilia ... bihi is a later insertion, it shows a mixture of pronouns.125

[6.4] Sentences containing grammatical errors126 from the Classical Arabic point of view.

John Burton127 in a celebrated article, "Linguistic Errors in the Qur'an," points out that Muslim scholars have been aware of the grammatical lapses in the Holy Book. But "the errors have never been removed. Either they have been complacently explained away on this grammatical ground or that, or, at best, serious efforts have been made to justify them as actually conforming with the usage of the Arabs."128 Burton then quotes some hadith where the errors are recognized:

When the copies of the revelations which he had ordered to be made were submitted to him, 'Uthman noted several irregularities, "Do not change them," he ordered, "the Arabs will change (or will correct them) as they recite."129

Burton next quotes a version from al-Farra', where 'Urwah questions 'A'ishah about a number of verses, IV.162, V.69, [discussed below] XX.63, "cA'ishah replied: `That was the doing of the scribes. They wrote it out wrongly.' "130

[6.4.1 ] V.69. Bell agrees with Torrey that as-subicuna must be an interpolation here since it is grammatically out of order; after 'inna it should have been as-sabi'ina, i.e., in the accusative.131

[6.4.2] VII.160. "We divided them into twelve tribes" (wa gatta 'nahum- th natay `ashrata 'asbatan).

Strict grammar requires the singular, since the numerals from 11 to 99 are followed by the noun in the accusative singular, hence 'asbatan should read sibtan.132

[6.4.3] IV.162. al-muqimina is wrong grammatically; it should read al-mugimuna, i.e., in the nominative case, like the other preceding substantives in the nominative, al-rasikhuna, and al-mu'minuna, and those coming after it, al-mu'tuna and al-mu'minuna.

[6.4.4] VII.56. "Surely the mercy of God is nigh. . ." The Arabic word for "nigh," qarib, should agree in gender with the Arabic word for "mercy," rahmah, which is feminine, and thus should read qaribatun, and not garibun, as it is in this verse.

[6.4.5] XXII.19. These are two disputants who have contended about their Lord.

Hadhani khasmani-khtasamu fi rabbihim

There are three numbers in Arabic: singular (mufrad), dual (muthanna), and plural (jam`). The verb ikhtasamu should have the dual ending, and not the plural, since two individuals (or two parties) are involved, and thus should read ikhtasamu.

[6.4.6] IX.69. "... You plunged about (in talk) as they plunged about ... " (or, more literally, "... as they who plunged").

"Wa-khudttum ka-l-ladhi khadu."

The word "as" is a translation of the Arabic ka, "like" or "as," and the relative pronoun alladhi, "who, which, that," together forming kalladhi. But in Arabic, the relative pronoun is declined, and in this verse, it should be in the plural since it refers to a plural pronoun. Hence it should read kalladhina. instead of kalladhi.133

[6.4.7] LXIII.10. "0 My Lord, wouldst Thou not defer me a little while, that I may give alms, and become one of the upright?"

Rabbi lawla'a 'akh-khartani 'i 'ila 'ajalin qaribin fa 'assaddaga wa 'akun mina-s-salihina.

As Wright tells us in Grammar, the subjunctive mood occurs in subordinate clauses, and is governed by particles such as fa-, when this particle introduces a clause that expresses the result or effect of a preceding clause. The preceding clause must express a wish or hope. Hence the verb 'akun should be in the subjunctive, and should read, 'akuna.134

[6.4.8] XI.10. "If We cause him to experience prosperity after [ba'da] the dearth [darra'a] which has affected him, he will assuredly say: `The evil (deeds) have departed from me'; lo, he is rejoicing, boastful."

All prepositions (e.g., ba`da) are followed by the noun in the genitive, and thus darra'a should in fact be darra'i.135

[6.4.9] XXXVII.123-130. "Elias was surely one of those sent.... Peace be on Elias."

Many of the verses in this sura end with the rhyme -in. For the sake of this rhyme, the second instance of Elias (verse 130) is rendered Ilyasin, as though it were a plural; a good example of poetic license.

[6.4.10] XCV.1-3. "By the fig and the olive! And Mount Sinai! [sinin]

And this city made secure [al-'amin]

(Inflectional vowels at the end of a verse are disregarded for the sake of the rhyme.)

Similarly, in this verse grammar is sacrificed for the rhyme. Sinai (in Arabic, sina'a) is changed for sinin for the sake of preserving the -in ending; another example of poetic license.

[6.4.11111.80. "The Fire shall not touch us but for a few days."

... ilia 'ayyaman ma`dudatan.

Arabic has two forms of plural, the plural of abundance and the plural of paucity. The latter is used only of persons and things that do not exceed ten in number, while the former is used for the rest. In this case, clearly a small number of days is meant; the emphasis is on the "fewness." Thus the plural of paucity would seem to be appropriate, ma'dudat (when declined in the above verse it would then be ma'dudatin) rather than ma`duda (in the above verse, when declined, it is ma'dudatan).136

[6.4.12] 11.177. "Righteousness does not consist in whether you face the East or the West, but virtuous conduct is (that of) those who have believed in Allah and the Last Day and the angels and the Book and the prophets ... and practice regular charity, to fulfil [al-mafana] the contracts you have made, and to be firm and patient [as-sabirina] in pain and adversity...."

The whole verse is rather tortuous and inelegant; many of the verbs are in the past tense in the original Arabic ('amana, 'ata, 'agama), when the present would have been more appropriate. Indeed, the translations certainly read more naturally in English when the present tense is used.

Second, in the original Arabic the verse begins rather clumsily: "But the piety [al-birr] is he who believes ..."

Blachere and others prefer to read al-barru instead of al-birr, giving the more logical reading, "the pious man is he who believes ..."

There is, however, one undoubted grammatical error: as-sabirina is incorrectly in the accusative, it should, like al-mafana, be in the nominative, and thus should read: as-sabiruna.137

[6.4.13] III.59. Arberry: "The likeness of Jesus, in God's sight is as Adam's likeness. He created him of dust, then said He unto him, `Be' and he was."

Pickthall: "Lo! the likeness of Jesus with Allah is as the likeness of Adam. He created him of dust, then He said unto him: Be! and he is."

Pickthall translates more literally, and keeps close to the original Arabic tenses. However, it would be more consistent to use, as Arberry does, the verb "to be" in the past, "he was," to agree with the past tense of "he said. . ." The Arabic yakan (is) should thus be kana (was).

It is worth pointing out that another analysis of the above verse is possible. 138

[6.4.14] XII.15. "So, when they had taken him away, and agreed to place him in the bottom of the cistern and We suggested to him the thought: `Thou wilt certainly tell them of this affair of theirs, when they are not aware.' "

Fa-lamma dhahabu bihi wa 'ajma`u 'ny-yaj ` aluhu fighayabati-l- jubbi wa 'awhayna 'layhi latunabbi'annahum bi 'amrihim hadha wa hum la yash`uruna.

Bell comments: "In verse 15 there is no principal clause, unless we omit one of the connectives, either that before 'ajma`a, or that before 'awhayna; as the clause introduced by the latter breaks the narrative, and verse 16 is short, there has possibly been an insertion. Verse 16 being the original close of verse 15, and wa being added."139

6.5 Phrases, sentences, and verses that do not seem to fit the context, and thus must have been transposed. These transposed or displaced verses are responsible for the disorder and incoherence that abounds in the Koran.

[6.5.1.] XLVIII.8-9. "Surely We have sent you as a witness [shahidan], as a bringer of glad tidings, and as a warner: In order that you may believe in Allah and His Apostle [rasul], that you may assist [tu'azziruhu] and honor Him and celebrate his praises [tusabbihuhu] morning and evening."

Bell remarks, "Verse 9 cannot possibly be in its original form, for the -hu [him] in tusabbihuhu cannot refer to the rasul, while that in the tu'azziruhu most naturally would; the middle of the verse must therefore have been inserted later, probably to adapt the verses as an introduction to verse 10."140

There is hopeless confusion about the pronoun "him" throughout the verse.

[6.5.2] Even scholars who seem to, on the whole, accept the Muslim chronology and the traditional account of the compilation of the Koran admit to difficulties of sense and reference, and point to the frequent breaks in logic and coherence in the Holy Text. Goldziher, for instance, wrote:

Judgments of the Quran's literary value may vary, but there is one thing even prejudice cannot deny. The people entrusted, during the reigns of Abu Bakr and 'Uthman, with the redaction of the unordered parts of the book occasionally went about their work in a very clumsy fashion. With the exception of the earliest Meccan suras, which the Prophet had used before his emigration to Medina as liturgical texts, and which consist of self-contained pieces so brief as to make them less vulnerable to editorial confusion the parts of the holy book, and particularly certain Medinese suras, often display a disorder and lack of coherence that caused considerable difficulty and toil to later commentators who had to regard the established order as basic and sacrosanct. If scholars undertake one day "a real critical edition of the text, reflecting all the results of scholarly research"-a project recently urged in these words by Rudolf Geyer,-they will have to pay attention to the transposition of verses out of their original contexts and to interpolations.141 The fact of editorial confusion appears clearly from Noldeke's survey of the arrangement of individual suras.142

The assumption of inapposite interpolations can on occasion help us get around difficulties in understanding the text. I would like to illustrate this by an example. Sura 24 (from verse 27 on) deals with the way virtuous people visit one another,how they should announce themselves, greet the people of the house, how women and children are to behave on such occasions. The rules for such situations became confused because in verses 32-34 and 35-36 two digressions, only loosely related to the main theme, were interpolated.

Then in verse 58 the theme of announcing one's visit is reintroduced, and discussed through verse 60. Then verse 61 reads: "There is no restriction on the blind, no restriction on the lame, no restriction on the sick, nor on yourselves, if you eat in one of your houses, or the houses of your fathers, or the houses of your mothers, or the houses of your brothers, or the houses of your sisters, or the houses of your paternal uncles, or the houses of your paternal aunts, or the houses of your maternal uncles, or the houses of your maternal aunts, or in one whose keys you hold or in one belonging to your friend. It will not render you guilty of a sin, whether you eat together or apart. And when you enter houses, greet one another with a greeting from Allah, a blessed and goodly one."

In this passage Muhammad permits his followers to join their relatives at table without any restriction, and even to go as guests to the houses of female blood relations. One cannot fail to notice that the first words of verse 61, which extend this freedom to the blind, lame, and sick, do not fit the natural context very well. A writer on medicine in the Qur'an took this juxtaposition very seriously, and offered the critique that while the dinner company of the halt and the blind is unobjectionable, a meal in the company of a sick man may be dangerous for one's health; Muhammad would have done better not to combat the aversion to it. 143

On closer study we see that the passage out of place in this context strayed into it from another group of rules. Its original reference is not to taking part in meals at the houses of others, but to taking part in the military campaigns of early Islam. In Sura 48, verses 11-16, the Prophet inveighs against "the Arabs who were left behind," those who did not participate in the campaign just undertaken. He threatens them with severe divine punishments. He appends to this verse 17: "It is no compulsion for the blind (laysa ... harajun), no compulsion for the lame, and no compulsion for the sick"-the text agrees literally with 24:61- i.e., people handicapped in these or other serious ways may be excused if they abstain. This phrase was inserted into the other context, to which it is foreign. It evidently influenced the redaction of the verse, whose original beginning cannot be reconstructed with certainty. Muslim commentators too have attempted, naturally without assuming an interpolation, to explain the words in keeping with their natural sense as an excuse for the abstention from war of those bodily unfit for service, but they had to accept the rejection of such an explanation for the reason that if the words were so understood, "they would not be in harmony with what precedes and follows them."I'

[6.5.31 II.238 f.

As Bell argues, "verses 238 and following have no connection with the context. They seem designed for those on some military expedition." 45

[ II.243

Bell again

"Verse 243 is enigmatical; it is unconnected with the context, and the reference is unknown. Bayda wi gives two stories:

(a) that of the people of Dwardan, said to be a village near Wasit associated in legend with Ezekiel, who were stricken by a pestilence and fled; Allah caused them to die, but afterwards brought them to life;

(b) that of some of the Israelites who refused to fight when summoned to do so by their king; they were caused to die but restored to life after eight days.

The latter is evidently founded on a wrong interpretation of the verse, which has no connection with fighting, but is designed to enforce the doctrine of the resurrection .... 146


In September 1996 the Ibn Khaldun Society was launched in London as an independent forum for moderate Muslims. At the inaugural conference, the participants reached, among others, the following conclusions:

Muslims must become independent of tradition. Just as our forebears found their own way, Muslims today must find theirs. In the process, they need to re-evaluate the Islamic tradition.

The only reliable and relevant source of faith is the Qur'an. Muslims need new scientific research into the Qur'an, and a re-examination of the Qur'anic message and its meaning in the 21st century.... 147

All moderate Muslims would no doubt wholeheartedly endorse these laudable goals, but one wonders how many of them realize how much their putative understanding of the Koran rests entirely on Islamic traditions.

The Muslim tradition has woven a fantastic spiderweb around its holy scripture from which even modem scholarship has not managed to disentangle itself. For all Muslims, much of the Koran remains incomprehensible without the commentaries; indeed, that is the very reason there are so many Muslim commentaries. As Leemhuis put it, "... The more of the Qur'an that became obscure in the course of time, the more of it became provided with an explanation." 148 One would hardly need them if the Koran were truly mubeen, "clear." But, as all my examples above show, despite all the thousands of pages devoted to clarifying the text, the Koran still remains incomprehensible, even for those Western scholars who accept the traditional, specially chronological Muslim framework for the Koran.

Muslim Koranic exegesis of such influential scholars as Tabari tended to be tafsir bill-ma'thur (interpretation following tradition), rather than tafsir bill-rah (interpretation by personal opinion). Tabarri's great work, Jami` al-bayan 'an taw' it ay al-Qur'an, is full of exegetical hadiths, where the Prophet gives his explanation of various obscure verses. Similarly, Ibn Kathir advises that if we are unable to elucidate some passage of the Koran by some other Koranic passage, then one must examine the prophetic sunna, and if that fails, then one must have resort to the sayings of the companions of Muhammad.149

However, if we accept the negative conclusions of Goldziher, Schacht, Wansbrough, Crone, and Cook about the authenticity of hadiths in general, then we must be equally skeptical of the hadiths concerning exegesis of the Koran. In other words, we cannot separate discussions of the compilation and meaning of the Koran from the questions about the authenticity of hadith and the sirah, the life of Muhammad.150

It is Muslim tradition that has unfortunately saddled us with the fiction that such and such verse in the Koran was revealed at such and such time during Muhammad's ministry. As early as 1861, the Reverend Rodwell wrote in his preface to the translation of the Koran, "It may be considered quite certain that it was not customary to reduce to writing any traditions concerning Muhammad himself, for at least the greater part of a century. They rested entirely on the memory of those who have handed them down, and must necessarily have been coloured by their prejudices and convictions, to say nothing of the tendency to the formation of myths and to actual fabrication, which early shews itself, especially in interpretations of the Koran, to subserve the purposes of the contending factions of the Umayyads and `Abbasids." Even the writings of historians such as Ibn Ishaq are "necessarily coloured by the theological tendencies of their master and patron.... Traditions can never be considered as at all reliable, unless they are traceable to some common origin, have descended to us by independent witnesses, and correspond with the statements of the Koran itself-always of course deducting such texts as (which is not unfrequently the case) have themselves given rise to the tradition. It soon becomes obvious to the reader of Muslim traditions and commentators that both miracles and historical events have been invented for the sake of expounding a dark and perplexing text; and that even the earlier traditions are largely tinged with the mythical element." 1 s 1

The above passage is a remarkable anticipation of the works of not only Goldziher but also Henri Lammens. The former showed by 1890 the entirely spurious and tendentious nature of the hadith, and the latter that "on the fabric of the Koranic text, the hadith has embroidered its legend, being satisfied with inventing names of additional actors presented or with spinning out the original theme." It is the Koran, in fact, that has generated all the details of the life of the Prophet, and not vice versa: "one begins with the Koran while pretending to conclude with it." Muslim tradition has often been able to do this because of the often vague and very general way events are referred to, such that they leave open the possibility of any interpretation that the Muslim exegetes care to embroider.

Michael Schub shows that the traditional interpretation of sura IX.40 is suspect, and is more probably derived from the Old Testament, 1 Sam. 23:16 ff. "Faithful Muslims will forever believe that Quran IX.40: `If ye help him not, still Allah helped him when those who disbelieve drove him forth, the second of two; when they two were in the cave, when he said unto his comrade: Grieve not. Lo! Allah is with us. Then Allah caused His peace of reassurance to descend upon him and supported him with hosts ye cannot see, and made the word of those who disbelieved the nethermost, while Allah's word it was that became uppermost. Allah is mighty, wise' refers to the Prophet Muhammad and Abu Bakr, although not one word of the Quranic text supports this."152

Rippin has also argued that certain passages in the Koran that are traditionally interpreted as referring to Muhammad are not necessarily historical. Citing Sura XCIII, Rippin states that "there is nothing absolutely compelling about interpreting [Sura XCIII] in light of the life or the lifetime of Muhammad. The `thee' [in verse 3: `The Lord has neither forsaken thee nor hates thee'] of this passage does not have to be Muhammad. It certainly could be, but it does not have to be. (I might also point out that Arberry's translation also suggests the necessity of `he' as God [i.e., `He'] which is also not necessarily compelling.) All the elements in the verses are motifs of religious literature (and indeed, themes of the Qur'an) and they need not be taken to reflect historical `reality' as such, but, rather, could well be understood as the foundational material of monotheist religious preaching.""53 One of Rippin's conclusions is that "the close correlation between the lira and the Qur'an can be taken to be more indicative of exegetical and narrative development within the Islamic community rather than evidence for thinking that one source witnesses the veracity of another. To me, it does seem that in no sense can the Qur'an be assumed to be a primary document in constructing the life of Muhammad. The text is far too opaque when it comes to history; its shifting referents leave the text in a conceptual muddle for historical purposes. This is the point of my quick look at the evidence of the `addressee' of the text; the way in which the shifts occur renders it problematic to make any assumption about the addressee and his (or her) historical situation. If one wishes to read the Qur'an in a historical manner, then it can only be interpreted in light of other material."154


[8.1] Aramaic Alphabet155

The North Semitic alphabet, which was used in Syria from the eleventh century B.C.E. onward, is the direct or indirect ancestor of all subsequent alphabetic scripts (including the South Semitic scripts such as Ethiopic, though there is no scholarly consensus on this point).156 It gave rise to the Phoenician and Aramaic alphabets. The Aramaic alphabet was developed in the tenth and ninth centuries B.C.E.; the oldest inscription in Aramaic script dates from about 850 B.C.E. Both the language and the script were used as a lingua franca throughout the Middle East. The Aramaic alphabet has twenty-two letters, all indicating consonants, and is written, like Arabic and Hebrew, from right to left. "It is ancestral to Square Hebrew and the modern Hebrew alphabet, the Nabataean and modem Arabic scripts, the Palmyrene alphabet, and the Syriac, as well as hundreds of other writing systems used at some time in Asia east of Syria. Aramaic also has been influential in the development of such alphabets as the Georgian, Armenian, and Glagolitic [Slavonic]." 157

[8.2] Arabic Alphabet

The origins of the Arabic alphabet are still imperfectly understood. It very probably developed in the fourth century C.E. as a direct descendant of the Nabataean alphabet, which in turn comes down from Aramaic. Some scholars, however, think the Nabataean inscriptions found on a tombstone in Umm al-Jimal (see appendix J), and dated approximately to 250 C.E. are examples of at least proto-Arabic writing. Some scholars would claim that earliest example of Arabic script that we know of is a royal funerary inscription, found in Namara in 1901 (see appendix J), of the Nabataeans dating from 328 C.E. Others argue that this inscription, though it shows some of the characteristics of Arabic, is essentially Aramaic, and insist that the earliest extant example of Arabic writing is a trilingual inscription in Greek, Syriac, and Arabic discovered at Zabad (see appendix J), dating from 512 C.E.

John Healey158 sums up the two theories as to the origins of the Arabic script:

Basically the view that has become prevalent, despite some dissent, is that the early cursive Arabic script, evidenced in seventh-century papyri (mostly from Egypt and Nessana in the Negev), derived from the Nabataean script. I have argued159 that it derived specifically from the cursive [used for less formal everyday purposes] variety of the Nabataean script (a view for which the evidence is now strengthened by the publication on microfiches of more of the cursive Nabataean papyri). 160 ... The alternative view has sought a Syriac origin for the Arabic script. This view, associated especially with the name of the late Jean Starcky and advocated particularly by the French school, argues from the broader issue of the basic design of the Syriac and Arabic scripts, specifically the fact that both "sit" upon the line of writing, while the Nabataean script "hangs" from an upper line.161 This point is apposite, though somewhat weakend by the existence of Nabataean inscriptions and papyri in which the lower line seems to be more significant. It remains the fact, however, that a number of the Arabic letters could not have been derived from the Syriac.... It would seem, in fact, that there is a fairly even split in the Arabic inventory of letters: eleven of the Arabic letters could be either of Nabataean or Syriac origin, while ten are much more plausibly related to Nabataean are hard to explain from Syriac, formal or cursive. It may be also noted that none of the Arabic letters is impossible to explain from Nabataean.

It is very likely that both Hebrew and Arabic owe to Syriac their own system of vowel notation by supralinear and sublinear markings.162

The Arabic alphabet, written from right to left, has twenty-eight letters, twenty-two of them being those of the Semitic alphabet, from which it is descended; the remaining six letters represent sounds not used in the languages written in the earlier alphabet. All the letters represent consonants, and thus, as M. Cohen once put it, "the orthography always comprises an element of interpretation by the reader, an ideographic element."

The shape of each letter differs according to its position at the beginning, middle or end of the word (initial, medial and final respectively); a fourth form of the letter is when it is written alone (see appendix K).

Certain letters of the Arabic alphabet are identical in shape, and are only differentiated by the presence or absence of a dot, for instance, to distingush an r from a z; j, and an h, from a kh, other pairs are s and sh; d and dh; t and z; and so on. But as Beeston163 reminds us, in some cases the differentia tion is not simply by presence or absence of a dot, but between varieties in the number and placing of dots: initial and medial b, t, th, n, and y all have dots differing in number and placing, and in word-end position only n and y are distinctive without the aid of dots. Thus, a great many variant readings are possible according to the way the text is pointed (has dots added; these dots are usually called diacritic dots or even "diacritical points"-in Arabic, nuqat [see appendix L]) In the first two centuries of Islam, diacritical dots were hardly used at all. When they were eventually introduced, there were additional problems since many of the dots were often written at some distance above or below the letter itself. Thus it was often difficult to detect which of two adjacent letters the dotting was intended to affect.

Parallel to the problem of diacritics to differentiate the consonants was the problem of indicating the vowels. Following earlier Semitic script traditions, the earliest Arabic used "the letters w and y ambivalently, both as true consonants and as indicators of the long vowels a-, and i-; but long a- was noted (by an originally consonantal letter) only at the end of the word, hardly ever in the middle of the word.... As for the short vowels, these were ... normally omitted altogether in writing."164

In the very earliest Qur'an codices, and in inscriptions, coins and papyri, no marking at all is found for short vowels or for a- in the middle of a word. By the early second /eighth century, some Qur'an codices used coloured dots as indications of vowels, though only to a limited extent, where misreading was particularly likely.165

Short vowels eventually came to be represented by three orthographical signs-taking the form of a slightly slanting dash placed below or above the line, or a comma placed above the line. Using different vowels, of course, gave different readings. Compounding these problems was the lack of an adequate punctuation system. The Koran was indeed written in a scripta defectiva; scripta plena, which allowed a fully voweled and pointed text, was not perfected until the late ninth century.

Thus every Arabic text consists of three layers:

(1) the basic (unpointed) form, shape or drawing of the individual word; in Arabic, rasm.

(2) the diacritical points, in Arabic, nugat, the function of which is to differentiate letters of the basic rasm; there are seven letters which are the unmarked members of pairs where the other member has over-dotting.

(3) Signs for the short vowels, to be read with the consonants denoted by the basic drawing (rasm) and the diacritical points (nugat).

Gunter Luling gives the following example of the ambiguity of the unpointed Arabic script; the word rasm, if pointed and vowelled differently, gives at least six possible readings: zanaytum, "you have fornicated"; zayyantum, "you have adorned"; rabbaytum, "you have educated"; rannaytum, "you have delected"; rana'tum, "you have looked at, or you have walked heavily"; ra'aytum, "you have seen."

It should be clear by now that the ground layer of the Arabic script, that is, the rasm, or basic drawing, without the diacritical points and signs is very difficult to interpret, and very easy to misinterpret.'66

The traditional explanation of the existence of variants goes something like this. The problems posed by the scripta defectiva inevitably led to the growth of different centers with their own variant traditions of how the texts should be pointed or vowelized. Despite 'Uthman's order to destroy all texts other than his own, older codices must have survived. As Charles Adams says, "It must be emphasized that far from there being a single text passed down inviolate from the time of 'Uthman's commission, literally thousands of variant readings of particular verses were known in the first three (Muslim) centuries. These variants affected even the 'Uthmanic codex, making it difficult to know what its true form may have been." 167

Muslim scholars themselves, from the early days of Islam, have acknowledged the existence of variants. This tradition has led to the compilation of all variants in a mammoth work of eight volumes, Mu jam algira'at al-qur'aniyyah,168 edited in Kuwait recently. This dictionary lists over ten thousand variants, of which about a thousand are variants of or deviations in the rasm. Gerd-R. Puin, the German scholar most closely involved with the classification of the approximately sixteen thousand sheets or parchments of Koranic fragments discovered in Sanla', Yemen, has uncovered even more variants in the rasm that are not found in the above-mentioned eight-volume dictionary. By comparing the rasm of the Cairo Mushaf with a fragmentary Hijazi Mushaf consisting of eightythree sheets, which can be tentatively dated to the early eighth century C.E. on stylistic grounds, Puin discovered that the deviations in the Hijazi Mushaf by far outnumber the deviations that have been recorded by the Muslim authorities on the gira'at and which have been collected in the above-mentioned encyclopaedia. This observation is not specific to the Koranic manuscripts of Yemeni provenance, but it is true for more or less all of the extant manuscripts preserved in Hijazi style.

The Hijazi Korans show differences in the sytem of counting of verses from the two dozen or so schools of counting; even the sequence of suras is often at variance not only with the Standard Egyptian edition but with the sequence of suras as recorded for the Korans of Ibn Masud and Ubayy b. Kalb. These deviations cannot be dismissed as mere scribal errors (lapsus calami), since the so-called errors are repeated with the same word several times in several fragments studied by Puin. Thus, as Puin emphasizes, it makes common philological sense to look for a rationale. The recurrent deviations from the Standard Egyptian text must be taken seriously, and cannot be swept under the carpet and attributed to scribal inadequacy.

One of Puin's conclusions is that though there was an oral tradition (otherwise the Koranic text could not have been read at all), there were deliberate changes in the oral tradition of Koran reading/recitation. Thus this oral tradition was not very stable or elaborate-changes must have occurred as can be seen in the variant orthography to be found in the Hijazi manuscripts, in general.

As Guillaume says, the variants are not always trifling in signifi- cance.169 As an example of a variant reading on the level of vocalization though not of the rasm, we might cite the last two verses of Sura LXXXV, al-Buruj, which read: (21) huwa qur'anum majidun; (22) fi lawhim mahfuzunlin. The last syllable is in doubt. If it is in the genitive -in, it gives the meaning "It is a glorious Koran on a preserved tablet"-a reference to the Muslim doctrine of the Preserved Tablet. If it is the nominative ending -un, we get "It is a glorious Koran preserved on a tablet."

In IV. 117, the standard text ended in an obscure word: "They do not invoke in lieu of Allah other than ..."; the last word was usually read 'inathan ("females"). The problem is that many of the pre-Islamic deities were male. In XXIX.16, we find, "You only worship in lieu of Allah 'authdnan [idols]." Thus, an emendation gives us idols instead of females; however, the form 'authdnan involved the insertion of a letter, whereas the form uthunan was doubtful Arabic.170

Other examples include I11.11, where, in the account of the miracle of Badr, the nature of the miracle varies seriously according as we read "you saw them" or "they saw them."171

It is clear that many hundreds of variants, though not all, were invented by Muslim grammarians, philologists, and exegetes of the third and fourth Muslim centuries to explain all sorts of obscurities of the Koran, whether of sense or reference; Koranic grammatical aberra- tions;172 or, even more seriously, for doctrinal reasons to defend some particular theological position.173 A kind of ethics of variants had developed by the ninth century C.E., according to which only variants that were not too far from Islamic orthodoxy or doctrines, or not too ungrammatical, were to be accepted and preserved. Hence, if there had been startling deviations or variants, they would have been suppressed. Thus, the variants that do remain are not always very significant. But we need to make a distinction between the variants fabricated by the Muslim exegetes, and the variants to be found in the rasm in manuscripts such as those examined by Puin. The sheer number of variants in the orthography in the earliest manuscripts certainly cast doubt on the traditional account of the compilation of the Koran. The Hijazl fragments seem to suggest that, even in the eighth century C.E., the text of the Koran was yet to be defined, and the "reading" options that the meagre rasm allowed had to be limited by officially recognizing only a part of them as admissible gira'at.


Spotting contradictions in the Koran is something of a growth industry, particularly in the context of Muslim-Christian polemics, with Muslims desperately trying to keep their finger in the leaking dike.174

Contradictions do abound in the Koran, and the early Muslims were perfectly well aware of them; indeed, they devised the science of abrogation to deal with them. It is a very convenient doctrine that, as one Christian unkindly put it, "fell in with that law of expediency which appears to be the salient feature in Muhammad's prophetical career." 175 According to this doctrine, certain passages of the Koran are abrogated by verses revealed afterward, with a different or contrary meaning. This was supposedly taught by Muhammad at Sura 11. 105: "Whatever verses we [i.e., God] cancel or cause you to forget, we bring a better or its like." According to al-Suyuti the number of abrogated verses has been estimated from five to five hundred. As Margoliouth remarked, "To do this, withdraw a revelation and substitute another for it was, [Muhammad] asserted, well within the power of God. Doubtless it was, but so obviously within the power of man that it is to us astonishing how so compromising a procedure can have been permitted to be introduced into the system by friends and foes."176

Al-Suyuti gives the example of Sura 11.240 as a verse abrogated (superseded) by verse 234, which is the abrogating verse. How can an earlier verse abrogate a later verse? The answer lies in the fact that the traditional Muslim order of the suras and verses is not chronological, the compilers simply having placed the longer chapters at the beginning. The Muslim commentators, for whom the Koran and the Sira are necessarily and inexorably joined, have to decide the chronological order for doctrinal reasons. Western scholars, wedded to the traditional Muslim account, have also worked out a chronological scheme; though there are many differences of detail, there seems to be a broad-but by no means completeagreement as to which suras belong to the Meccan (i.e., early) period of Muhammad's life and which belong to the Medinan (i.e., later) period. It is worth noting how time-bound the "eternal" word of God is.

Let us take an example: everyone knows that Muslims are not allowed to drink wine in virtue of the prohibition found in the Koran (Sura 11.219), and yet many would no doubt be surprised to read in the Koran at Sura XVI.67, "And among fruits you have the palm and the vine, from which you get wine and healthful nutriment: in this, truely, are signs for those who reflect" (Rodwell). Dawood has "intoxicants" and Pickthall, "strong drink," and Sale, with eighteenth-century charm, has "inebriating liquor" in place of "wine." While Yusuf Ali pretends that the Arabic word concerned, Sakar, means "wholesome drink," and in a footnote insists that nonalcoholic drinks are being referred to; and then at the last moment concedes that if "sakar must be taken in the sense of fer mented wine, it refers to the time before intoxicants were prohibited: this is a Meccan Sura and the prohibition came in Medina."177

Now we can see how useful and convenient the doctrine of abrogation is in bailing scholars out of difficulties-though, of course, it does pose problems for apologists of Islam, since all the passages preaching tolerance are found in Meccan (i.e., early suras), and all the passages recommending killing, decapitating and maiming, the so-called Sword Verses, are Medinan (i.e., later); "tolerance" has been abrogated by "intolerance." For example, the famous Sword verse, ayat al-sayf, at Sura IX.5, "Slay the idolaters wherever you find them," is said to have canceled 124 verses that enjoin toleration and patience.178

Here are the supposedly early suras preaching tolerance:

CIX: "Recite: 0 Unbelievers, I worship not what you worship, and you do not worship what I worship. I shall never worship what you worship. Neither will you worship what I worship. To you your religion, to me my religion."

L. 45: "We well know what the infidels say: but you are not to compel them."

XLIII. 88,89: "And [Muhammad] says, `0 Lord, these are people who do not believe.' Bear with them and wish them `Peace.' In the end they shall know their folly."

The exceptions are to be found in Sura II, which is usually considered Medinan (i.e., late):

11.256: "There is no compulsion in religion."

11.62: "Those who believe [i.e., Muslims] and those who follow the Jewish scriptures, and the Christians and the Sabians, and who believe in God and the Last Day and work righteousness, shall have their reward with their Lord, on them shall be no fear, nor shall they grieve."

Unfortunately, as he gained in confidence and increased his political and military power, so the story goes, Muhammad turned from being a persuader to being a legislator, warrior, and dictator. Hence, the Medinan chapters such as Suras IX, V, IV, XXII, XLVII, VIII, and II reveal Muhammad at his most belligerent, dogmatic, and intolerant-that is, for those who want to closely link the Koran with the life of the Prophet.

XXII.19: "As for the unbelievers for them garments of fire shall be cut and there shall be poured over their heads boiling water whereby whatever is in their bowels and skins shall be dissolved and they will be punished with hooked iron rods."

The Koran also enjoins all Muslims to fight and kill nonbelievers:

XLVII.4: "When you meet the unbelievers, strike off their heads; then when you have made wide slaughter among them, carefully tie up the remaining captives."

IX.29: "Declare war upon those to whom the Scriptures were revealed but believe neither in God nor the Last Day, and who do not forbid that which God and His Apostle have forbidden, and who refuse to acknowledge the true religion [that is, the Jews], until they pay the tribute readily, being brought low."

IX.5-6: "Kill those who join other gods with God wherever you may find them."

IV.76: "Those who believe fight in the cause of God ..."

VIII.12: "I will instill terror into the hearts of the Infidels, strike off their heads then, and strike off from them every fingertip."

VIII.38-39: "Say to the Infidels: If they desist from their unbelief,what is now past shall be forgiven them; but if they return to it, they have already before them the doom of the ancients! Fight then against them till strife be at an end, and the religion be all of it God's."

It is a grave sin for a Muslim to shirk the battle against the unbelievers, those who do will roast in hell:

IX.39: "If you do not fight, He will punish you severely, and put others in your place."

Those who die fighting for the only true religion, Islam,will be amply rewarded in the life to come:

IV.74: "Let those fight in the cause of God who barter the life of this world for that which is to come; for whoever fights on God's path, whether he is killed or triumphs, We will give him a handsome reward."

We might give the following further examples of contradictions, though it seems rather doubtful if the doctrine of abrogation can deal with all of them.

The omnipotence of God is everywhere asserted in the Koran; man's will is totally subordinate to God's will to the extent that man cannot be said to have a will of his own. Even those who disbelieve in Him do so because it is God who wills them to disbelieve. This leads to the Muslim doctrine of predestination, which prevails over the doctrine of man's free will, also to be found in the Koran. As Macdonald says, "the contradictory statements of the Kuran on free-will and predestination show that Muhammad was an opportunist preacher and politician and not a systematic theologian." 19

"Taqdir, or the absolute decree of good and evil, is the sixth article of the Muhammadan creed, and the orthodox believe that whatever has, or shall come to pass in this world, whether it be good or bad, proceeds entirely from the Divine Will, and has been irrevocably fixed and recorded on a preserved tablet by the pen of fate."180 Here are some quotes from the Koran illustrating this doctrine:

LIV.49: "All things have been created after fixed decree."

111.145: "No one can die except by God's permission according to the book that fixes the term of life."

LXXXVII.2-3: "The Lord has created and balanced all things and has fixed their destinies and guided them."

VI1I.17: "God killed them, and those shafts were God's, not yours."

IX.5 1: "By no means can anything befall us but what God has destined for us."

XII1.31: "All sovereignty is in the hands of God."

XIV.4: "God misleads whom He will and whom He will He guides."

XVIII.101: "The infidels whose eyes were veiled from my warning and had no power to hear."

XXXII.13: "If We had so willed, We could have given every soul its guidance, but now My Word is realized-`I shall fill Hell with jinn and men together.' "

XLV.26: "Say unto them, 0 Muhammad: Allah gives life to to you, then causes you to die, then gathers you unto the day of resurrection ..."

LVII.22: "No disaster occurs on earth or accident in yourselves which was not already recorded in the Book before we created them."

But there are, inevitably, some passages from the Koran that seem to give man some kind of free-will:

LXXIV.54-55: "Nay, it is surely a Reminder. So whovever pleases may mind it."

LXXVI.3: "We have truly shown him the way; he may be thankful or unthankful."

LXXVI.29: "Surely this is a Reminder; so whoever will, let him take a way to his Lord."

X1I.17: "As to Thamud, We vouchsafed them also guidance, but to guidance did they prefer blindness."

XVIII.29: "The truth is from your Lord: let him then who will, believe; and let him who will, be an unbeliever."

Faced with this mass of contradictions, Muslim scholars, leaning on verses from Suras XVI.101, XXII.52, 11.106, LXXXVII.6 ff., devised the doctrine of abrogation by which the earlier Koranic passages were abrogated by chronologically later ones. Essentially, "abrogation [naskh] involved the suppression of a ruling without the suppression of the wording. That is to say, the earlier ruling is till to be found in the Qur'an, and is still to this day recited in worship, but it no longer has any legal force."] 81

Some Muslim scholars also postulated two further types of abrogation:

(a) where both the ruling and wording have been suppressed

(b) where the wording has been suppressed but the ruling is still in force (e.g., the famous stoning verse that condemns men and women to death by stoning for sexual immorality-zina')1 s2

It is very doubtful that the verses adduced to back the Muslim scholars' arguments really have anything to do with abrogation at all; on the contrary, the context indicates that the verses can interpreted very differently. Burton tried to show that the word aya in Sura 11. 106 refers to an individual ritual or legal obligation, and the verb yansakh means "modification." Thus, 11. 106 would refer to the modification of an earlier, Jewish ritual or legal regulation by a later, Islamic one.183

A second reason for scepticism about the classical theory of abrogation is that there has never been a consensus among jurists about which Qur'anic passages it affects. Az-Zuhri (d.742), an early authority on the subject, held that 42 ayahs [verses] had been abrogated. After his time, the number steadily increased until an upper limit was reached in the eleventh century, with Ibn Salama claiming that there were 238 abrogated ayahs, and al-Farisi claiming that there were 248. In subsequent generations, a reaction set in: the Egyptian polymath al-Suyuti (d. 1505) claimed that there were only 20, and Shah Wall Allah of Delhi (d.1762) whittled the number down to 5.184

We might add two other scholars whose calculation of the number of abrogated verses varies considerably: al-Nahhas, 138; Ibn al 1Ata'igi, 231.185

The sura lists of Muslim scholars purporting to indicate which belonged to the Meccan (early) period and which to the Medinan (later) seem at first promising.

Although no two lists are exactly the same, they all have a family likeness, and some of the lists are supported by isnads [chain of transmitters] ostensibly tracing them back to the period of the Companions. It seems probable, however, that these lists were compiled during the first quarter of the eighth century, at very earliest, and that they reflect the opinion of scholars whe were active at that time. The broad agreement amongst these scholars about which surah are Meccan and which Mad- inan is understandable, as in the majority of cases this can be deduced from the content. On the other hand, the differences of opinion about the precise order in which the surahs were revealed probably reflect rival views concerning the asbab al-nuzul the supposed occasions when such and such sura was revealed to Muhammad, see below] and abrogation. In short, there is insufficient evidence for holding that these lists are based on independent ancient traditions, although that possibility cannot of course be entirely ruled out.186

What of the so-called asbab al-nuzul, the occasions of revelation, when, according to Muslim tradition, such and such verse was revealed to Muhammad?

Surely, they settle definitively the chronology of the Koran, and decide which verses are Meccan and which Medinan?

In his Quranic Studies, John Wansbrough had expressed the view that asbab material had its major reference point in halakic works, that is to say, works concerned with deriving laws from the Koran. Andrew Rippin,187 however, examined numerous texts, and concluded that the primary purpose of the sabab material was in fact not halakic, but rather haggadic: "that is, the asbab functions to provide an interpretation of a verse within a broad narrative framework." This puts the origin of the asbab material in the context of the qussas: "the wandering storytellers, and pious preachers and to a basically popular religious worship situation where such stories would prove both enjoyable and edifying." He also notes that the primary purpose of such stories is to historicize the text of the Koran in order to prove that "God really did reveal his book to humanity on earth," and that in arguments over conflicting asbab reports isnad (chain of transmission) criticism was a tool that could be "employed when needed and disregarded when not."

As Hawting points out, "The very diversity of these `occasions of revelation' (asbab al-nuzul), the variety of the interpretations and historical situations the tradition provides for individual koranic verses, is an argument for the uncertain nature of the explanations that are provided. One often feels that the meaning and context supplied for a particular verse or passage of the Koran is not based on any historical memory or upon a secure knowledge of the circumstances of its revelation, but rather reflect attempts to establish a meaning. That meaning, naturally, was established within a framework of accepted ideas about the setting in which the Prophet lived and the revelation was delivered. In that way, the work of interpretation also defines and describes what had come to be understood as the setting for the revelation."188

I shall end this section with what I wrote in an earlier book:

Juynboll once said that Wansbrough's theories were so hard to swallow because of the obvious disparity in style and contents of Meccan and Medinan suras.189 There is indeed a difference in language, style and even message between the so-called Meccan and Medinan suras. But all that shows is that there are two quite different styles in the Koran, and of course, Muslim exegetes solved this problem by assigning one set to Mecca and the other to Medina, with considerable tinkering (verses from the "Medinese" suras assigned to Mecca and vice versa). But why should we accept the Medinan and Meccan labels? What is the source or sources of this difference? To accept these labels is simply to accept the entire traditional Muslim account of the compilation of the Koran, the biography of the Prophet, and the Rise of Islam. Again, this is precisely what is at stake: the reliability of the sources. The differences, if anything, point to a history far more extensive than the short life of Muhammad as found in the Sira, and they do not have to be interpreted biographically through the history of the life of Muhammad in Mecca and Medina. There is nothing natural about the Meccan/Medinan separation. It is clear from Lammens, Becker and others that large parts of the sira and hadith were invented to account for the difficulties and obscurities encountered in the Koran, and these labels also proved to be convenient, for the Muslim exegetes for the same reason. The theory of abrogation also gets the exegetes out of similar difficulties, and obviates the need to explain the embarrassing contradictions that abound in the Koran.190


The full implications of the sixty or so inscriptions found in and around Mecca, Saudi Arabia, have yet to be worked out.191 Some of these inscriptions, incised on white limestone, which have been dated to 80 A.H. and others to 84 A.H., 98 A.H., and 189 A.H., consist of what seem like quotations from the Koran. There are clearly recognizable phrases from the Koran but there is never a complete verse, and often one sentence is found to contain Koranic quotes from two different suras; others show considerable deviations from the Standard Edition. One could argue that they are not Koranic quotes at all, or that the "writer" has simply badly remembered the Koran. One could also argue that, once again, the Koran had not yet been standardized, or even reduced to a written form.


Given the above examples of some of the difficulties, any critical reading of the Koran should prompt the exasperated but healthy response, "What on earth is going on here?" The fact that so many, but thankfully not all, scholars of the last sixty years have failed to even ask this question, let alone begin to answer it, shows that they have been crushed into silence out of respect for the tender sensibilities of Muslims, by political correctness, postcolonial feelings of guilt, and dogmatic Islamophilia, and have been practicing "Islamic scholarship" rather than scholarship on Islam.

Some scholars did pose pertinent questions, and gave us important insights. I have tried to include their work in this anthology. And yet, so often their keen and just observations were vitiated by a faulty chronology, that is, they all accepted the traditional historical framework fabricated by Muslim tradition. It seems to me that their work makes far more sense within a broad revisionist structure provisionally constructed by Wansbrough and his disciples.

To give a plausible account of the rise of Islam we must put back the last of the three monotheist religions in its Near Eastern geographical, religious, historical, and linguistic (Hebrew, Aramaic, Syriac) context. Scholars have been well aware of the influences of Talmudic Judaism, heretical Christianity, and now even Essenians, on Islam, but relying on the fictive chronology of Muslim tradition has often meant the invention of ingenious but ultimately far-fetched scenarios of how Christian monks, Jewish rabbis, or Essenians fleeing Romans had whispered their arcane knowledge into the ears of an Arabian merchant.

So many scholars have also accepted totally uncritically the traditional account of the compilation of the Koran. But this account is, in the words of Burton, "a mass of confusion, contradictions and inconsisten- cies,"192 and it is nothing short of scandalous that Western scholars readily accept "all that they read in Muslim reports on this or that aspect of the discussions on the Qur'an."193

Given that so much of the Koran remains incomprehensible despite hundreds of commentaries, surely it is time to look for some more plausible historical mechanism by which the Koran came to be the Koran, and to restore the original text.

Despite Barth's pioneering work, there is still a reluctance to impugn the putative authenticity of the Koran, and talk of emendations. Bellamy, for instance, makes the following pertinent remarks,

... one seeks in vain for a systematic application of the techniques of textual criticism to the textual problems of the Koran, although classicists and Biblical scholars have for centuries made continuous efforts to improve the quality of the texts that are the bases of their disciplines.... Whatever the reasons, Western scholarship, with very few exceptions has chosen to follow the Muslim commentators in not emending the text. When faced with a problem, the Westerners have resorted to etymologizing and hunting for foreign words and foreign influences. They have produced a great deal of valuable scholarship important for our study of the Koran and the origins of Islam, but where they exercised their skill on corrupt texts, they, of course, produced only fantasies.194

And yet Bellamy ends his article with almost an apology: "It should not be assumed that in making these emendations, I am in any way trying to diminish the remarkable achievement of Zayd b. Thabit and his colleagues in producing the Uthmanic recension of the Koran." Even Bellamy, it seems, accepts the traditional compilation story.

Bellamy is quite right that among classicists emendations, and even the assumption of interpolations, practically comprise the definition of textual criticism. Here the typically trenchant remarks of the eminent classicist A. E. Housman195 are of the greatest relevance:

Textual criticism is a science, and, since it comprises recension and emendation, it is also an art. It is the science of discovering error in texts and the art of removing it. That is its definition, that is what the name denotes.196

... [T]extual criticism is not a branch of mathematics, nor indeed an exact science at all. It deals with a matter not rigid and constant, like lines and numbers, but fluid and variable; namely, the frailties and aberrations of the human mind, and of its insubordinate servants, the human fingers.197

... the amount of sub-conscious dishonesty which pervades the textual criticism of the Greek and Latin classics is little suspected except by those who have had occasion to analyse it. People come upon this field bringing with them prepossessions and preferences; they are not willing to look all facts in the face, nor to draw the most probable conclusion unless it is also the most agreeable conclusion.198

Interpolation is provoked by real or supposed difficulties, and is not frequently volunteered where all is plain sailing; whereas accidental alteration may happen anywhere. Every letter of every word lies exposed to it, and that is the sole reason why accidental alteration is more common. In a given case where either assumption is possible, the assumption of interploation is equally probable, nay more probable; because action with a motive is more probable than action without a motive. The truth therefore is that in such a case we should be loth to assume accident and should rather assume interpolation; and the circumstance that such cases are comparatively uncommon is no reason for behaving irrationally when they occur.199

Barth and Fischer's important work (translated here for the first time) on emendations and interpolations, though it did influence Richard Bell in the writing of his commentary on the Koran, was unfortunately not followed up. Even Bell, on the whole, is unwilling to accept emendations too readily, and most scholars seem to agree with Noldeke that the Koran is free of omissions and additions. But as Hirschfeld says, "Considering the way in which the compilation was made, it would have been a miracle, had the Qoran been kept free of omissions, as well as interpola- tions."200 Some scholars did question the authenticity of certain verses: Antoine-Isaac Silvestre de Sacy was doubtful about sura 111.138; Weil of Suras 111.182, XVII.1, XXI.35-36, XXIX.57, XLVI.14, XXXIX.30; and Sprenger, LIX.7.201 Hirschfeld questioned the authenticity of verses containing the name Muhammad, regarding it as rather suspicious that such a name, meaning "praised," should be borne by the Prophet:

... that name [Muhammad] could not have come into practical use until a period of the Prophet's life when the material of the Qoran was all but complete. Now it might be objected that the texts of the missionary letters which Muhammad commenced to send in the seventh of the Hijra to unconverted Arab chiefs, as well as to foreign potentates were headed by the phrase: "From Muhammad, the Messenger of Allah, to, etc." The authenticity of the majority of these letters ... is very doubtful, and besides, even if the genuineness of the texts of the documents be admitted, the superscription may have been added by the traditionists who took it for granted.202

Watt and Bell try to answer Hirschfeld by essentially assuming the reliablity and authenticity of the traditions: "[The name, Muhammad,] occurs, not only in the Qur'an but in the documents handed down by Tradition, notably the constitution of Medina and the treaty of al-Huday- biyya; in the latter the pagan Quraysh are said to have objected to the title rasul Allah, and to ar-Rahman as a name of God, but raised no question about the name Muhammad."203 This is an astonishingly naive and circular argument. First, it is the reliability of tradition that is the crux of the matter, and if tradition is capable of inserting the name Muhammad into the Koran, then it is equally capable of inventing the story where the Quraysh do not object to this name, for it is tradition that is our only source for the story of the reception of the treaty of al-Hudaybiyya. We do not have independent means of verifying the story. It is tradition that interpolated the name into the Koran, but it is also tradition that embroidered or spun out the details of the biography of the Prophet. As is so often the case, traditions contradict one another: Some traditions even claim that at birth Muhammad had received the name Qutham.204 It would seem arbitrary to pick on just one of them; as Burton said, "We must either accept all hadiths impartially with uncritical trust, or one must regard each and every hadith as at least potentially guilty of a greater or lesser degree of inherent bias, whether or not this is immediately visible to Western eyes."205

Another scholar who has dared to question the authenticity of the Koran is Paul Casanova, whose ideas are rather perfunctorily dismissed by Watt and Bell. Casanova finished his study Mohammed et la fin du Monde in 1921, but in recent years his work has been, I believe unjustly, ignored.206 I suspect one reason for this neglect has nothing to do with the force of his arguments or the quality of his scholarship, but the simple unavailability of all three volumes of his work; volume three, pages 169244, being particularly difficult to come by.207

Casanova wrote:

It is generally admitted that the text of the Koran, such as it has come down to us, is authentic and that it reproduces exactly the thought of Muhammad, faithfully gathered by his secretaries as the revelations gradually appeared. We know that some of his secretaries were highly unreliable, that the immediate successor of the Prophet made a strict recension, that, a few years later, the arrangement of the text was altered. We have obvious examples of verses suppressed, and such a bizarre way in which the text is presented to us (in order of the size of the chapters or surahs) shows well the artificial character of the Koran that we possess. Despite that, the assurance with which Muslims-who do not refrain from accusing Jews and Christians of having altered their Scriptures-present this incoherent collection as rigorously authentic in all its parts has imposed itself upon the orientalists, and the thesis that I wish to uphold will seem very paradoxal and forced.

I maintain, however, that the real doctrine of Muhammad was, if not falsified, at least concealed with the greatest of care. I shall set out soon the extremely simple reasons which led first Abu Bakr, then 'Uthman, to alter thoroughly the sacred text, and this rearrangement was done with such skill that, thenceforth, it seemed impossible to reconstitute the Ur-Koran or the original Koran. If however my thesis was accepted, it could serve as a point of departure for this reconstitution, at least for everything that concerns the original revelations, the only really interesting ones from my point of view, the only ones, moreover, that there was any advantage in reworking, by means of either very light changes of the text, or by deplacements. There is abundant evidence that the first Muslims, despite the undoubtedly powerful memories of the Arabs, were profoundly ignorant of the Koran, and one could, with Muhammad dead, recite them verses of which they had not, at their own admission, the slightest idea. A rearrangement which did not change the exterior forms of the verses was thus the easiest. Sprenger, who had had a vague intimation of the thesis that I advocate, accuses Muhammad of having thrown the incoherence into his text himself, in order to get rid of the trace of imprudent words.208 I say in fact that it is for a reason of this kind that the incoherence was introduced, but not by the authorby his successors.209

According to Casanova, Muhammad, under the influence of a Christian sect, put great emphasis on the imminent end of the world in his early proclamations. When the approaching end failed to take place, the early followers of the Prophet were forced to refashion or rework the text of the Koran to eliminate that doctrine from it.

Casanova provides some very convincing arguments for the presence of interpolations in the Koranic text, and further points up its general incoherence. Whether they prove what he wanted to prove is another matter. But it is certainly unfair of Watt and Bell to pronounce dismissively that Casanova's thesis is "founded less upon the study of the Qur'an than upon investigation of some of the byways of early Islam."210 Casanova has anticipated just such a criticism, and we can see the following as an implicit anwser to Watt/Bell-type accusations:

Already, at this period [Caliph, 'Abd al-Malik, reigned 685-705 c.E.] the book [Koran] was hardly understood. "If obscurity and lack of coherence with the context in our modem Koran ought to be considered as proof of non-authenticity, I fear that we ought to condemn more than one verse," says N6ldeke.211

I confess that as for me I accept these premises and this conclusion. Obscurity and incoherence are the reasons, not to deny absolutely, but to suspect the authenticity [of the Koran], and they permit all effort to restore a more clear and more coherent text.

Permit me some characteristic examples. I have collected them by a careful study of the Koranic text,212 I could have multiplied them but that would have uselessly padded out this book. Besides, in most cases, all the while feeling the strangeness and obscurity of terms, that the naive exegesis of the commentators only brings out the better, one is very perplexed to propose a rational solution, a credible restoration. I ought to be on my guard the more so because people will not fail to accuse me (that has already been done) of declaring falsified such and such passages because they go counter to my theories. To defend myself from this reproach, I shall add to this list of alterations a short analysis of those which have been noted before me by scholars totally unaware of my aforementioned thesis."213

There then follow examples of interpolations, displacement of verses, and so on; in other words, all the evidence of the general incoherence of the Koran.

Watt and Bell's defense depends completely on tightly linking the Koran to the biography of the Prophet, this linkage is,of course, entirely derived from Muslim tradition: "As to [Casanova's] main thesis, it is true that the Qur'an proclaims the coming Judgement and the end of the world. It is true that it sometimes hints that this may be near; for example, in XXI.1 and XXVII.71-73 f. In other passages, however, men are excluded from knowledge of times, and there are great differences in the urgency with which the doctrine is proclaimed in different parts of the Qur'an. All this, however, is perfectly natural if we regard the Qur'an as reflecting Muhammad's personal problems and the outward difficulties he encountered in carrying out a task to which he had set his mind. Casanova's thesis makes little allowance for the changes that must have occurred in Muhammad's attitudes through twenty years of ever-changing circumstances. Our acceptance of the Qur'an as authentic is based, not on any assumption that it is consistent in all its parts, for this is not the case; but on the fact that, however difficult it may be to understand in detail, it does, on the whole, fit into a real historical experience, beyond which we discern an elusive, but, in outstanding characteristics, intelligible personality."214

It requires little reflection to see, once again, the circularity of Watt and Bell's argument. If by "authentic" we mean that the Koran was the word of God, as passed on, either directly from God or through the intermediary of an angel, to a historical figure called Muhammad, supposedly living in Arabia, then clearly we need some independent confirmation of this extraordinary claim. We cannot say the Koran is authentic because "it does fit ... into a real historical experience."

For this circular reasoning would give us the following tautology: "the Koran is authentic, that is, it fits into a real historical experience, because it fits into a real historical experience."

Some have scholars have, of course, been trying to prise the Koranic text away from the supposed historical fit with the Sira, the life of Muhammad. Lammens,215 Tor Andrae,216 and more modestly Andrew Rippin,217 and Michael Schub.218 But perhaps the most radical thesis is that of Gunter Luling, who, argues very persuasively, that at least a third of the Koran predates Islam, and thus, of course, has nothing whatsoever to do with someone called Muhammad. A third of the Koran was originally a preIslamic Christian hymnody that was reinterpreted by Muslims, whose task was made that much easier by the ambiguity of the rasm, that is, the unpointed and unvowelled Arabic letters. Thus both Casanova and Luling point to the present incoherence of the Koranic text as evidence for its later editing, refashioning, emending, and "re-interpretation" and manipulation. It is interesting to note that though he finds Luling's evidence "unsound, and his method undisciplined,"219 Wansbrough nonetheless thinks that the "recent conjectures of Luling with regard to the essentially hynmic character of Muslim scripture are not unreasonable, though I [Wansbrough] am unable to accept what seems to me [Luling's] very subjective reconstruction of the text. The liturgical form of the Qur'an is abundantly clear even in the traditional recension, as well as from the traditional literature describing its communal uses. The detection of strophic formation is certainly not difficult, and the theological (as opposed to rhetorical) nature of orthodox insistence upon the absence from scripture of poetry and even (though less unanimous) of rhymed prose must be acknowledged."220

Luling is reviving a theory first put forward by H. Miiller,221 according to which it was possible to find in the Koran, as in the Bible, an ancient poetical form, the strophe or stanza. This form was present in seventeen suras, particularly Suras LVI and XXVI. For Muller, composition in strophes was characteristic of prophetic literature. Rudolph Geyer222 took up the theory, and thought he had proved the presence of a strophic structure in such suras as Sura LXXVIII. These ideas were dismissed at the time, but perhaps make more sense now, if we see, as Luling does, in the Koran pre-Islamic Christians texts.

Liiling's thorough grounding in Semitic languages enables him to show that we cannot hope to understand the Muslim tradition's reworking of the Koranic text without an understanding of Hebrew, Aramaic, and Syriac. Following in the footsteps of Mingana, Jeffery, and Margoliouth, but going way beyond them, is Christoph Luxenberg,223 who also tries to show that many of the obscurities of the Koran disappear if we read certain words as being Syriac and not Arabic. In order to elucidate passages in the Koran that had baffled generations of scholars, Muslim and nonMuslim, Luxenberg used the following method:

(1) He went carefully through Tabari's great commentary on the Koran, and also consulted Ibn al-Manzur's celebrated dictionary of the Arabic language, Lisan al-Arab, in order to see if Western scholars of the Koran had not omitted any of the plausible explanations proposed by the Muslim commentators and philologists. If this preliminary search did not yield any solutions, then

(2) he tried to replace the obscure Arabic word in a phrase or sentence that had hitherto mystified the Muslim commentators, or that had resulted in unconvincing, strained, or far-fetched explan- tions with a Syriac homonym, which had a different meaning (though the same sound), but which made more sense in the context. If this step did not yield a comprehensible sentence, then

(3) he proceeded to the first round of changes of the diacritical points, which, according to Luxenberg's theory, must have been badly placed by the Arabic readers or whoever was the original redactor or copier of the Koran, and which had resulted in the actual obscurity of the Koranic passage concerned. In this way, he hoped to obtain another more logical reading of the Arabic. If this also failed to give any results, Luxenberg

(4) then proceeded to the second round of changes of the diacritical points in order to eventually obtain a more coherent Syriac reading, and not an Arabic one. If all these attempts still did not yield any positive results,

(5) he tried to decipher the real meaning of the Arabic word, which did not make any sense in its present context, by retranslating it into Syriac to deduce from the semantic contents of the Syriac root the meaning best suited to the Koranic context.

In this way, Luxenberg was able to explain not only the so-called obscure passages, but a certain number of passages that he considers were misunderstood, and whose meaning up to now no one had doubted. He was also able explain certain orthographic and grammatical analomies which abound in the Koran.

This method allows Luxenberg, to the probable horror of all Muslim males dreaming of sexual bliss in the Muslim hereafter, to conjure away the wide-eyed houris promised to the faithful in Suras XLIV.54 and LII.20. According to Luxenberg, the new analysis yields "white raisins" of "crystal clarity" rather than doe-eyed and ever-willing virgins. Luxenberg claims that the context makes it clear that it is food and drink that is being offerred, and not unsullied maidens. Similarly, the immortal, pearllike ephebes or youths of suras such as LXXVI.19 are really a misreading of a Syriac expression meaning "chilled raisins (or drinks)" that the Just will have the pleasure of tasting in contrast to the "boiling drinks" promised the unfaithful and damned.

Luxenberg's work has only recently been published in Germany, and we must await its scholarly assessment before we can pass any judgments.


Credulity does not become an historian.

P. R. Davies224

The sources for that historical event [seventh-century Hijaz] are exclusively literary, predominantly exegetical, and incarcerated in a grammar designed to stress the immediate equivalence of word and world. Or, I might be inclined to add: all we know is what we have been told. With neither artifact nor archive, the student of Islamic origins could quite easily become victim of a literary and linguistic conspiracy. He is, of course, mostly convinced that he is not. Reason for that must be confidence in his ability to extrapolate from the literary version(s) what is likely to have happened. The confidence is certainly manifest; the methodological premises that ought to support, or, at least, accompany it, are less so.

John Wansbrough225

Surely it is time for a critical examination of the methodological assumptions that have gone totally unscrutinized for so long. Despite the fact that Wansbrough's literary analysis of the sources has undermined the traditional account of the origin of Islam, the Sira, and the coming into being of the Koran, scholars, who made their reputations from taking the Muslim account at face value, have carried on as if nothing has happened. Conveniently ignoring the full implications of Wansbrough's theories, these conservative scholars have not even seriously tried to answer him.226

But as P. R. Davies says, "it is not acceptable for an historian to trust the text or its unknown author. Credulity does not become a historian. Scepticism, rather, is the proper stance, just as in turn that historian's own text must earn trust too, and not demand credence."

In their Positivist classic of historical methodology, Langlois and Seignobos227 make a similar point:

For criticism is antagonistic to the normal bent of the mind. The spontaneous tendency of man is to yield assent to affirmations, and to reproduce them.... It takes a special reason to induce us to take the trouble to examine into the origin and value of a document on the history of yesterday; otherwise, if there is no outrageous improbability in it, and as long as it is not contradicted, we swallow it whole, we pin our faith to it, we hawk it about, and, if need be, embellish it in the process. Every candid man must admit that it requires a violent effort to shake off ignavia critica, that common form of intellectual sloth, that this effort must be continually repeated, and is often accompanied by real pain.

... [C]riticism is not a natural habit, it must be inculcated, and only becomes organic by dint of continued practice.

Historical work is, then, pre-eminently critical; whoever enters upon it without having first been put on his guard against his instinct is sure to be drowned in it.

While they warn against hypercriticism, Langlois and Seignobos make it clear that it is credulity that is the main enemy of scientific method. Certain historians "are content to examine whether the author was roughly contemporary with the events, whether he was an ocular witness, whether he was sincere and well-informed, whether he knew the truth and desired to tell it, or even-summing up the whole question in a single formula-whether he was trustworthy.

This superficial criticism is certainly better than no criticism at all, and has sufficed to give those who applied it the consciousness of incontestable superiority. But it is only a halfway house between common credulity and scientific method. Here, as in every science, the starting point must be methodical doubt. All that has not been proved must be temporarily regarded as doubtful .... 228

The historian ought to distrust a priori every statement of an author, for he cannot be sure that it is not mendacious or mistaken.... We must not postpone doubt till it is forced upon us by conflicting statements in documents, we must begin by doubting.229

An author may have any number of motives for violating the truth:

(1) He or she may seek to gain a practical advantage; the author knowingly gives false information, he or she has an interest in deceiving. This is the case with most official documents.

(2) The author was placed in a situation that compelled him to violate truth. This happens whenever he has to draw up a document in conformity with rule or custom, while the actual circumstances are in some point or other in conflict with rule or custom.

(3) "The author viewed with sympathy or antipathy a group of men (nation, party, denomination, province, city, family), or an assemblage of doctrines or institutions (religion, school of philosophy, political theory), and was led to distort facts in such a manner as to represent his friends in a favourable and his opponents in an unfavourable light."

(4) "The author desired to please the public, or at least to avoid shocking it. He has expressed sentiments and ideas in harmony with the morality or the fashion of his public; he has distorted facts in order to adapt them to the passions and prejudices of his time.... The purest types of this kind of falsehood are found in ceremonial forms, official formulae, declarations prescribed by etiquette, set speeches, polite phrases."

(5) "The author endeavoured to please the public by literary artifices. He distorted facts in order to embellish them according to his own aesthetic notions."230

Outside the more general need for methodological doubt and skepticism, there is an equally urgent, if more specific, necessity to put Islam firmly within the gradual development of Middle Eastern monotheism outside Arabia; that is, within the Judaeo-Christian sectarian milieu. This milieu necessarily includes not only the theological and polemical framework and assumptions of the various contending sects, but also the linguistic background. Arabic itself must be placed squarely back in its Semitic surroundings; its relationship to Hebrew, Aramaic, and Syriac must be reexamined. Placing the Koran in its Hebrew and Syriac milieu has already given us the startlingly new theories of Luling and Luxenberg.

What a new generation of biblical scholars, such as P. R. Davies,230 Keith Whitelam,232 N. P. Lemche,233 T. L. Thompson,234 J. van Seters,235 and G. Garbini,236 has achieved by its openness to the methods of other disciplines like anthropology, sociology, social history, linguisitics, and literary criticism is very instructive indeed. "In order to deal with the reports of seventh-century Arabia," Wansbrough "divided the field into `constants' and `variables': the former representing the `basic categories' common to most descriptions of monotheism; the latter representing `local components,' that give each version its special character.... The constants were prophet, scripture, and sacred language; the variables were the specifically Arabian features of these.... "237 What the new biblical scholars conclude about the Bible, history, and Ancient Israel is readily applicable to Islam, since these conclusions refer to "constants" common to most descriptions of monotheism. For instance, Lemche writes, "It is certainly not unusual for people to possess their own foundation myth. It is as a matter of fact quite common, almost universal phenomenon, that any group-ethnic, national, political, religious, and occupational-will be in possession of a narrative about its foundation known to and accepted by its membership.... The myth of the exile and return has a similar story to tell."238

History is one of the remedies open to the creators of ethnicity, and as has become conspicuous recently, it is of little importance whether this history is a real history or an invented one. History is written in order to create identity among the members of certain society, congregation, or whatever ethnic group we may speak about. The only important thing seen in the perspective of the author, who created this history, would be that it must be acceptable for its readership; its readers must be able to identify with the history as it has been told to them.

The biblical history about Israel ... is simply a reflection of the self-understanding of the people who created this history and for whom it was created. This community will have to be understood as a religious community, not an ordinary living organism such as a normal people; it is the people of God, now past its punishment and redeemed by its God. It is a community with a firm conviction of belonging to a specific place, which it alone is entitled to possess because it is the gift of its God, and because its membership are all one and the same family, the descendants of the patriarchs, Abraham, Isaac, and Jacob.239

Further on, Lemche240 writes that Israel in the Old Testament is an artificial creation, which has little in common with the Israel that existed once in Palestine. Similarly, Hawting241 has tried to argue that the preIslamic Arabia found in Muslim tradition is essentially a literary and ideological construct with probably little in common with the "real Arabia."

Wansbrough emphasizes two points whose full implications are perhaps too disturbing for most scholars to draw: first,

there is no Muslim literature which can be dated, in the form in which it is available to us, earlier than 800 C.E. (end of the second century of the Islamic era); the other is that Islam is a complex phenomenon the development of which must have taken many generations and occupied an extended geographical area before it attained a form resembling that which we know today.

Although it is true that there are a few traditional texts conventionally attributed to figures who died before 800 C.E. (notably Ibn Ishaq and Malik b. Anas), we only have those works in recensions made by Muslim scholars of later generations, and none of the works available to us were put into the form in which we know them earlier than the ninth century C.E. (the third century of Islam).We have no biography of Muhammad, no commentary on the Koran, no law book, no collection of Hadiths, no history of early Islam, etc., which can be said to predate, in the form in which have it, the beginning of the third Islamic century. And, given the impulse in traditional Islamic scholarship to attribute to great figures of the past texts which have been formed over a considerable period of time and which stabilized comparatively late, it may be suspected that the conventional attribution to "authors" living in the early ninth century of a number of important works may be too generous.

Wansbrough's work exhibits severe scepticism about these attempts to push our Muslim sources back earlier than the form in which we know them and he shows no interest in reconstructing or analyzing the isnads. His position seems to be that even were it possible to accept the accuracy and authenticity of the isnads (which seems doubtful for the most important, earliest, alleged links in the transmission), there would nevertheless be little possibility of assessing the transformation of the accompanying traditions as they were subject to the vicissitudes of transmission over many generations. Variant wording, the introduction of glosses, the removal of material from its original context, abbreviation, summary and expansion, incomplete transmission, and other features can all be assumed to have taken place. Above all, even though our earliest Muslim literature undoubtedly recycles and reworks material which originated much earlier, that material exists because it answers to the needs of the generations in whose work we finds it.242

Finally, Wansbrough can teach us one further lesson that places him directly within the revisionist tradition of biblical scholars mentioned above: The concept of Islam as an evolution from the sectarian monotheism of Mesopotamia in the wake of Arab migration and the establishment of Arab rule; the analysis of that evolution as a gradual elaboration of a series of ideas, practices, and institutions expressive of the independent identity of the community; and the understanding that an elaboration of an account of its own origins is a part of that evolution; these seem to me the especially liberating aspects of Wansbrough's approach.243

While modern biblical studies has made great progress building on the works of such pioneers as Wellhausen and Graf, Koranic studies is still lying contentedly, self-satisfied in the procrustean bed prepared by Muslim tradition more than a thousand years ago. As Wansbrough himself said, "As a document susceptible of analysis by the instruments and techniques if Biblical criticism [the Koran] is virtually unknown. The doctrinal obstacles that have traditionally impeded such investigation are, on the other hand, very well known. Not merely dogmas such as those defining scripture as the uncreated Word of God and acknowledging its formal and substantive inimitability, but also the entire corpus of Islamic historiography, by providing a more or less coherent and plausible report of the circumstances of the Quranic revelation, have discouraged examination of the document as representative of a traditional literary type."244

Rippin endorses Wansbrough's frustration:

... I have often encountered individuals who come to the study of Islam with a background in the historical study of the Hebrew Bible or early Christianity, and who express surprise at the lack of critical thought that appears in introductory textbooks on Islam. The notion that "Islam was born in the clear light of history" still seems to be assumed by a great many writers of such texts. While the need to reconcile varying historical traditions is generally recognized, usually this seems to pose no greater problem to the authors than having to determine "what makes sense" in given situation. To students acquainted with approaches such as source criticism, oral formulaic composition, literary analysis and structuralism, all quite commonly employed in the study of Judaism and Christianity, such naive historical study seems to suggest that Islam is being approached with less than academic candour.245

Conservative scholars such Watt or Welch have never given us an epistemologically or psychologically plausible, or even simply common sensical account as to how the Koran came into being. If they believe that the Koran is "authentic," how do they think Muhammad received his "revelations"? Do they believe that Muhammad literally went into a trance and somehow saw visions of angels who recited various verses to him, which he then revealed to his companions, who then wrote them down verbatim? Some of the passages and stories in the Koran are very long indeed. Are we to understand that Muhammad remembered several hundred lines of rhymed prose that were "revealed" to him in his trance? Do we assume that all his companions were literate, and able to write down his every word, all the time believing that their Prophet was in direct communion with an angel? What in fact is a revelation or revelation in general? How does it operate psychologically and epistemologically? "We do not yet possess a usable cross-cultural theory or typology of revelations. . . ."246

What exactly does "authentic" mean to non-Muslim scholars? Is there a coherent definition of "authentic"? Is there then a valid, i.e., noncircular, argument to show that the Koran is authentic?

These are crucial questions that have never been asked, let alone answered. Then what exactly are the implications of the research of scholars such as Geiger, Sidersky, Hirschfeld, Speyer, Katsch, Torrey, Schapiro, among a host of others,247 who have shown the various Judaic or Christian elements that have gone into the making of the Koran? Did Muhammad read the Babylonian Talmud in Aramaic? How did he then incorporate what he had read into his "revelatory trances" that were then written down "exactly as revealed" by literate companions, who were already aware that their leader was a prophet from God?

Even scholars skeptical of the sources of our knowledge of Islam are willing to accept the Koran as "authentic." I have already given the example of Watt and Bell arguing in a circle on this point. F. E. Peters is another very distinguished scholar who seems to want it both ways: "The Holy Book of Islam is text without context, and so this prime document, which has a very strong claim to be authentic, is of almost no use for reconstructing the events of the life of Muhammad."248 How can we know that the Koran is "authentic" if we cannot trust any of our sources for the rise of Islam and the life of Muhammad?249 It was Lammens who showed how the text of the Koran generated virtually every element that Muslim tradition attributes to the life of its prophet; as Lammens put it, "One begins with the Koran while pretending to conclude with it."250 Furthermore, Peters him self believes that "Lammens' critical attack has never been refuted."251 And yet Peters continues to talk in traditional terms of early Meccan suras and later Medinan ones, and seems confident we can "reconstruct to some degree what appears to be an evolution in Muhammad's own thinking about God."252 A little later Peters tells us that "Goldziher, Lammens, and Schacht were all doubtless correct. A great deal of the transmitted material concerning early Islam was tendentious-not only the material that was used for legal purposes but the very building blocks out of which the earliest history of Muhammad and the Islamic community was constructed." If this is true why take the traditional Muslim account seriously?253

One of the strongest arguments against the traditional account, or rather contradictory accounts, of the compilation of the Koran is what we have learned from biblical studies about the canonization process. Why and how are certain texts included in an anthology of texts and then elevated to the status of scripture? It is a long complex process and the Muslim account(s) of the Koran are far too simplistic: neither religions nor sacred texts are born fully-fledged.

It is also an extraordinary situation that in the twenty-first century we still do not have a definitive, scholarly text of the Koran. The situation is truly chaotic, with scholars content to work without specifying which manuscript or edition they are relying on, or more probably tacitly using the so-called Standard Egyptian Edition, sometimes also referred to as the 1342 Cairo text. However even the latter text, as Adrian Brockett pointed out, did not have an official status in northwest Africa or Iran: "In the last decade [Brockett is writing in 1984], for instance, even in central Muslim countries like Saudi Arabia and Qatar, texts differing considerably in orthography from the 1342 Cairo text have been printed under official approval."254 Brockett goes on to examine a number of printed Hafs copies available in the 1970s, and finds that they fall into five broad traditions: Iranian, Indian, Turkish, Egyptian, and northwestern. "The differences between these Traditions comprise script, orthography, recitative details and textual division.... In some respects the two outlying Traditions, the Indian and the northwest African, are makedly different from the other more centrally situated ones. They have also retained a few fossil elements of orthography lost from the central ones."255 Neither Western scholars nor ordinary Muslims have, it seems, something called the Koran, they all make do with a Koran.


The following is a selected list of books for further reading, but it does not include all the books cited or mentioned in my introduction and its footnotes.

Reference works

The Encyclopaedia of Islam. 9 vols. 2d ed. Leiden: E. J. Brill, 1954.

Hughes, T. P. Dictionary of Islam. Calcutta: Rupa & Co., 1988.


Wright, W. A Grammar of the Arabic Language. Cambridge, 1967.


Dozy, R. P. A. Supplement aux dictionnaires arabes. 2 vols. Paris, 1881. Reprint, 1960.

de Biberstein-Kazimirski, A. Dictionnaire arabe franpaise. 4 vols. Paris, 1860.

Lane, E. W. An Arabic-English Lexicon. 8 vols. 1863-1893. Reprint, Beirut, 1968.

Penrice, J. A Dictionary and Glossary of the Koran. 1873. Reprint, Delhi, 1990.

Wehr, Hans. A Dictionary of Modern Written Arabic. 1st Eng. ed. Ithaca, N.Y.: Cornell University Press, 1961.


Cook, Michael. Muhammad, Oxford, 1983

Hisham, Ibn. The Life of Muhammad. Edited and translated by A. Guillaume. London, 1955.

Warraq, Ibn, ed. The Quest for the Historical Muhammad. Amherst, N.Y.: Prometheus Books, 2000.

Watt, W. M. Muhammad: Prophet and Statesman. London, 1961.


Ayoub, M. The Qur'an and Its Interpreters. Vol. 1. Albany, 1984.

Beeston, A. F. L. Baidawi's Commentary on Sarah 12 of the Qur'an: Text Accompanied by an Interpretative Rendering and Notes. Oxford, 1963.

Bell, Richard. Introduction to the Qur'an. Edinburgh, 1953. Revised by W. M. Watt, 1970.

A Commentary on the Qur'an. Manchester, 1991.

Blachere, Regis. Introduction au Coran. 1958. Reprint, Paris, 1991.

Burton, John. The Collection of the Qur'an. Cambridge, 1979.

Cook, Michael. The Koran: A Very Short Introduction. Oxford, 2000.

Flugel, G. Corani Textus Arabicus. Leipzig, 1834.

Corcordantiae Corani Arabicae. Leipzig, 1842.

Goldziher, I. Muslim Studies. 2 vols. London, 1967, 1971.

Jeffery, Arthur. The Foreign Vocabulary of the Qur'an. Baroda, 1938.

Materials for the History of the Text of the Qur'an. Leiden, 1937.

Luling, Gunter. Uber den Ur-Qoran. Erlangen, 1993.

Margoliouth, D. S. Chrestomathia Baidawiana: The Commentary of el- Baidawi on Sura III Translated and Explained for the Use of Students of Arabic. London, 1894.

Noldeke, T. Geschichte des Qorans. 3 vols. 2d ed. Leipzig, 1909-1938.

Schacht, Joseph. The Origins of Muhammadan Jurisprudence. Oxford, 1950.

Sfar, Mondher. Le Coran est-il authentique? Paris, 2000.

Shaikh, A. Islam, Sex and Violence. Cardiff, 1999.

Wansbrough, John. Qur'anic Studies. Oxford, 1977.

The Sectarian Milieu. Oxford, 1978.

Warraq, Ibn, ed. The Origins of the Koran. Amherst N.Y.: Prometheus Books, 1998.

Translations of the Koran

Ali, Muhammad. The Holy Qur-an. Woking, 1917.

Arberry, Arthur. The Koran Interpreted. Oxford, 1964.

Bell, Richard. The Qur'an, Translated with a Critical Re-arrangement of the Surahs. 2 vols. Edinburgh, 1937.

Blachere, Regis. Le Coran. Paris, 1949-51.

Dawood, N. J. The Koran. Harmondsworth, 1990.

Palmer, E. H The Qur'an. Oxford, 1880.

Paret, Rudi. Der Koran: Ubersetzung. Stuttgart, 1962.

Der Koran: Kommentar and Konkordanz. Stuttgart, 1971.

Pickthall, Marmaduke. The Meaning of the Glorious Koran. New York, 1930.

Rodwell, J. M. The Koran Translated from the Arabic. London, 1861.

Sale, George. The Koran. London, 1734.

Yusuf Ali, A. The Holy Qur'an: Translation and Commentary. Lahore, 1934.

Arabic Commentaries

Baydawi. Beidhawii commentarius in Coranum. Edited by H. O. Fleischer, 2 vols. Leipzig, 1846-48.

Jalalain al-Mahalli, Jalal ad-Din, and as-Suyuti, Jalal ad-Din. Tafsir al-Jalalain, in Al futuha at al-ilahiyya bi-tau di h tafsir alJalalain li-daga'igal-khafiyya ta'lif Sulayman ibn `Umar al-'Uyayli ash-Shaft a. 4 vols. Cairo, 1337 A.H./ 1957-58.

Ibn Kathir, Tafsir al-Qur'an al-Karim. 7 vols. Beirut: 1385 A.H.

Al-Tabari. The Commentary on the Qur'an. Vol. 1. Translated by J. Cooper. Oxford, 1987.

Az-Zamakhshari. Tafsir al-Kashshaf. 4 vols. Cairo, 1373 A.H./1953-55.


1. J. Hollander, "Versions, Interpretations, Performances," in On Translation, ed. Reuben A. Brower (Cambridge: Harvard University Press, 1959), p. 208, where Hollander also quotes Voltaire.

2. Jackson Mathews, "On Translating Poetry," in On Translation, ed. Reuben A. Brower (Cambridge: Harvard University Press, 1959), p. 70.

3. See appendix, "Bibliography of Translations," in Arabic Literature to the End of the Umayyad Period, ed. Beeston, Johnstone, et al. (Cambridge: Cambridge University Press, 1983), pp. 502-20.

4. In Egypt, the rate of illiteracy is placed as high as 49.8 percent. See Information Please Almanac (Boston, 1997), p. 180.

5. Charles Ferguson, "Diglossia," Word 15, no. 2 (1959): 325-40; William Marcais, "La diglossie arabe," L'Enseignement public-Revue Peda- gogique 104, no. 12 (1930): 401-409; Alan S. Kaye, "Arabic," in The Major Languages of South Asia, the Middle East and Africa, ed. Bernard Comrie (London, Routledge, 1990), p. 181.

6. Kaye, "Arabic," p. 173.

7. Wheeler M. Thackston, An Introduction to Koranic and Classical Arabic (Bethesda, Md.: Iranbooks, 1994), p. xii.

8. B. Lewis, Islam and the West (Oxford: Oxford University Press, 1993), p. 65.

9. It is in fact becoming more and more westernized, i.e., de-Semitized under the influence of the international news agencies.

10. P. Larcher, "Les Incertitudes de la Poesie Arabe Archaique," La Revue des Deux Rives, no. 1 (1999): 129.

11. Kaye, "Arabic," p. 183.

12. P. Larcher, "La Linguistique Arabe d'Hier a Demain: Tendances Nou- velles de la Recherche," Arabica 45 (1998): 409-29.

13. Gustav Meiseles, "Educated Spoken Arabic and the Arabic Language Continuum," Archivum Linguisticum 11, no. 2 (1980): 118-42; quoted in Larcher, "Les Incertitudes de la Poesie Arabe Archaique."

14. A. S. Kaye, "Formal vs. Informal in Arabic: Diglossia, Triglossia, Tetraglossia, etc., Polyglossia-Multiglossia Viewed as a Continuum," ZAL 27 (1994): 47-66.

15. Barbara F. Grimes, Ethnologue, Languages of the World, 13th ed. (Dallas, 1996).

16. Everything, of course, depends on our definition of language and dialect; that is why estimates as to the number of languages in the world vary from five thousand to nearly seven thousand.

17. Merrit Ruhlen, A Guide to the World's Languages, vol. 1 (Stanford, 1991), p. 1.

18. I. M. Diakonoff, "Afro-Asiatic Languages," in Encyclopaedia Britannica [EB] (2000), at

19. M. Ruhlen, A Guide to the World's Languages, p. 380.

20. "Hamito-Semitic Languages," in The Columbia Encyclopedia, 6th ed. (New York, 2000).

21. R. Hetzron, "Semitic Languages," in The Major Languages of South Asia, the Middle East and Africa, ed. B. Comrie (London: Routledge, 1990), p. 162.

22. Wheeler M. Thackston, Introduction to Syriac (Bethesda, Md.: Iranbooks, 1999), p. vii.

23. S. Brock in G. H. A. Juynboll, ed., Studies in the First Century of Islamic Society (Chicago, 1982); Segal in Bernard Lewis and P. M. Holt, eds., Historians of the Middle East (London, 1962); Mingana in The Origins of the Koran, ed. Ibn Warraq (Amherst, N.Y.: Prometheus Books, 1998). Cahen in Arabica, vol. 1 (1954); Cook and Crone, Hagarism (Cambridge, 1977).

24. Thackston, Introduction to Syriac, p. viii.

25. "Syriac," in The Columbia Encyclopedia, 6th ed. (New York, 2000).

26. C. Rabin, "'Arabiyya," in EI2.

27. S. Fraenkel, Aramaeischen Fremdworter im Arabischen (Leiden, 1886).

28. Rabin, "`Arabiyya"; emphasis added by the author.

29. Kaye, "Arabic," p. 171.

30. Rabin, ""Arabiyya."

31. A. Rippin, "Epigraphical South Arabian and Qur'anic Exegesis," JSAI 13 (1990): 155-74.

32. "Arabic language," in EB (1999).

33. B. Lewis, Islam and the West (Oxford: Oxford University Press, 1993), p. 68.

34. Rabin, "'Arabiyya," p. 566b.

35. A. Schaade, "Arabia(e). Arabic Language. Classical Arabic," in Ell vol. 1, p. 394.

36. S. Akhtar, "Ex-defender of the Faith," Times Higher Educational Supplement [THES], August 22, 1997.

37. Lewis, Islam and the West, p. 68; emphasis added by the author.

38. S. Akhtar, in THES, February 13, 1998.

39. Schaade, "Arabia(e). Arabic Language. Classical Arabic," p. 394.

40. T. Noldeke, Beitrage zur Kenntniss der Poesie der alten Araber (Hanover, 1864), p. 2.

41. R. Blachere, Histoire de la Litterature Arabe. Des Origines a la fin du XVe siecle de-J.-C., vol. 1 (Paris, 1952), p. 79.

42. Schaade, "Arabia(e). Arabic Language. Classical Arabic," p. 393.

43. Karl Vollers, Volkssprache and Schriftsprache im alten Arabien (Strassburg, 1906).

44. Hans Wehr, "Review of Fuck (1950)," ZDMG 102 (1952): 179-84.

45. Werner Diem, "Die nabataischen Inschriften and die Frage der Kasus- flexion im Altarabischen," in ZDMG 123 (1973): 227-37.

46. T. Noldeke, "Zur Sprache des Korans," in Neue Beitrage zur semitischen (Strassburg: Sprachwissenschaft, 1910).

47. Johann Fuck, Arabiya: Untersuchungen zur arabischen Sprach-und Stilgeschichte [Abhandlungen de Sachsischen Akademie der Wissenschaften zu Leipzig, Philologisch-historische Masse 45/1] (Berlin, 1950).

48. J. Blau, "The Jahiliyya and the Emergence of the Neo-Arabic Lingual Type," JSAI 7 (1986): 35-43.

49. J. Wansbrough, Quranic Studies (Oxford, 1977), p. 102.

50. Rabin, ""Arabiyya," p. 566a.

51. F. Corriente, "From Old Arabic to Classical Arabic, the Pre-Islamic Koine: Some Notes on the Native Grammarians' Sorces, Attitudes and Goals," JSS 21 (1976): 62-98.

52. "I'rab" in E12, quoting al-Jurjani.

53. Rabin, ""Arabiyya," p. 565.

54. Wansbrough, Quranic Studies, p. 85.

55. [J.Wansbrough's note: "Suyuti, Muzhir i, 221; cf. Kahle, `Readers,' pp. 70-71: the story was pressed into the service of a number of distinct but related causes; for the literary effect of similar traditions see also above (Quranic Studies, pp. 42-3; pp. 69-70)."]

56. Wansbrough, Quranic Studies, p. 94.

57. Ibid., p. 95.

58. Ibid., pp. 97-98.

59. A. Mingana, Odes and Psalms of Solomon, ii, 1920, p. 125.

60. D. S. Margoliouth, JRAS (1925): 415-49.

61. T Husayn, Fi 1-adab al-jahili (Cairo, 1927).

62. For full bibliography, see Blachere, Histoire de la Litterature Arabe, vol. 1, pp. xviii-xxxiii.

63. The rhyme may have been there originally to aid memorization; the recording of rhyme depends on i`rab, but the use of rhyme does not.

64. G. R. Hawting, The Idea of Idoltatry and the Emergence of Islam: From Polemic to History (Cambridge, 1999), p. 48.

65. E.g., Sura XXVI.195; XLIII.1; XII.1.

66. T. Lester, What is the Koran? (New York: Atlantic Monthly Press, 1999); chap. 1.2 in this volume.

67. H. Hirschfeld, New Researches into the Composition and Exegesis of the Qoran (London, 1902), p. 6.

68. Ibid., p. 7.

69. Fuat Sezgin, GAS, band I, p. 24.

70. T. P. Hughes, Dictionary of Islam (1885; reprint, Calcutta: Rupa & Co., 1988), pp. 518 ff.

71. Ibid. p. 518.

72. Ibid., p. 519.

73. P. Crone, "Two Legal Problems Bearing on the Early History of the Quran," JSAI 18 (1994): 1-37.

74. M. Ali, The Holy Qur'an (Columbus, Ohio, 1995), p. 611.

75. M. Pickthall, The Meaning of the Glorious Koran (London, 1948), p. 318

76. A. Jeffery, The Foreign Vocabulary of the Qur'an (Baroda, 1938).

77. J. Penrice, A Dictionary and Glossary of the Koran (1873; reprint, Delhi, 1990).

78. E. W. Lane, An Arabic-English Lexicon (London, 1863-1893).

79. R. Blachere, Le Coran (Paris, 1949-51).

80. See chaps. 3.1 and 3.2, respectively.

81. Jeffery, The Foreign Vocabulary of the Qu'ran, p. 43.

82. Ibid., pp. 99-100.

83. Blachere, Le Coran, vol. 1, p. 61.

84. Jeffery, The Foreign Vocabulary of the Qu'ran, p. 53.

85. Blachere, Le Coran, vol. 1, p. 117.

86. Jeffery, The Foreign Vocabulary of the Qu'ran, p. 234 n. 4.

87. Penrice, A Dictionary and Glossary of the Koran, p. 117.

88. A. J. Wensinck, "Kurban," in Ell.

89. Blachere, Le Coran, vol. 2, p. 662.

90. Richard Bell, Translation of the Qur'an (Edinburgh, 1939), p. 511 n. 2.

91. Blachere, Le Coran, vol. 2, pp. 743-44 n. 59.

92. See J. D. McAuliffe, "Exegetical Identification of the Sabi'un," MW 72 (1982): 95-106; C. Buck, "The Identity of the Sabi'un: An Historical Quest," MW 74 (1984): 172-86.

93. Abu'l Qasim al-Husayn al-Raghib al-Isfahan, Al-Mufradat fi-Gharfb al-Qur'an (Cairo, 1324 A.H.).

94. Al-Tabari, The Commentary on the Qu'ran, trans. J. Cooper (Oxford: Oxford University Press, 1987), p. 410.

95. Ibid.

96. Ibn Sa`d, Kitab al-tabagat al-kabir, ed. Sachau et al. (Leiden, 192528), vol. 2, pp. 36-38.

97. Richard Bell, A Commentary on the Qu'ran (Manchester, 1991), vol. 1, p. 14.

98. Ibid., p. 17.

99. Ibid., p. 51.

100. Blachere, Le Coran, vol. 2, p. 809 n. 262.

101. Jeffery, The Foreign Vocabulary of the Qu'ran, pp. 225-26.

102. Fakhr al-Din al-Razi, al-Tafsir al-Kabir (Cairo: al-Matba`ah al- Bahiyah, n.d.), vol. 3, p. 77; quoted in M. Ayoub, The Qur'an and Its Interpreters (Albany, 1984), vol. 1, p. 102.

103. E. W. Lane, An Arabic-English Lexicon (London 1863), vol. 1, p. 9.

104. Jeffery, The Foreign Vocabulary of the Qur'an, pp. 43-44.

105. "Sidjdjil," in E12.

106. F. Leemhuis, "Quranic sijjil and Aramaic sgyl," JSS 27 (1982): 47-56.

107. "Sidjdjil," in EI2.

108. V. Vacca, "Sidjdjil," in Ell.

109. Jeffery, The Foreign Vocabulary of the Qur'an, pp. 163-64.

110. David Powers, Studies in Qur'an and Hadith: The Formation of the Islamic Law of Inheritance (Berkeley, 1986), pp. 22-23.

111. Quoted in Ayoub, The Qur'an and Its Interpreters, vol. 1, p. 70.

112. Blachere, Le Coran, vol. 2, p. 736 n. 27.

113. E.g., Dawood, Pickthall.

114. Tabari, The Commentary on the Quran, p. 201.

115. Bell, A Commentary on the Qur'an, vol. 1, pp. 18-19.

116. Ayoub, The Qur'an and Its Interpreters, vol. 1, p. 128.

117. Tabari, The Commentary on the Quran, p. 482.

118. Ibid., p. 483.

119. Ayoub, The Qur'an and Its Interpreters, vol. 1, p. 132.

120. Blachere, Le Coran, vol. 2, pp. 755-56.

121. Ayoub, The Qur'an and Its Interpreters, vol. 1, p. 145.

122. Bell, A Commentary on the Qur'an, vol. 1 p. 21.

123. Ibid., p. 46.

124. Ibid., p. 45.

125. Ibid., p. 47.

126. Students wishing to test their knowledge of Arabic grammar should perhaps see if they can spot the errors in 11.61; 11.83; 11.84; 11. 187; 11.238; 11.253; 111.146; IV.1; IV.13/14; IV.69; IV.78-79; IV.80; IV.136; IV.171; V.54; VI.25; VI.95; VII.178; IX.3; IX.107; X.92; XI.46; XI.46; XI.111; XII.30; XII.85; XV.51-52; XVI.13; XVI.69; XVI.101; XXIII.14; XXV.38; XXVI.16; XXX.30; XXXIII.63; XXXVII.6; XXXIX.21; XL.2-3; XLIII.81; LIII.20b; LIV.50; LV.39; LVI.13-14; LVII.18.

127. J. Burton, "Linguistic Errors in the Qur'an," JSS 33, no. 2 (1988): 181-96.

128. Ibid., p. 181.

129. Abu 'Ubayd al-Qasim b.Sallam, Fad'il al-Qur'an, MS., Tubingen, Ma, VI, 96, f.40b.

130. Yahya Ziyad al-Farra', Mauna al-Qur'an (Beirut 1955, 1980), vol. 1, p. 105; cf. Jalal al-Din al-Suyuti, al-Itgan fi 'ulum al-Qur'an (Cairo, 1354), vol. 1, pp. 182 ff.

131. W. Wright, A Grammar of the Arabic Language, 3rd. ed. (Cambridge, 1967), vol. 11, pp. 78-79.

132. Bell, A Commentary on the Qur'an, vol. 1 p. 255; Wright, A Grammar of the Arabic Language, vol. 1, p. 256D.

133. W. Wright, A Grammar of the Arabic Language, vol. 1, pp. 270-71.

134. Ibid., vol. 2, pp. 24-34; vol. 1, p. 60: On the subjunctive having a fatha.

135. Ibid., vol. 1, pp. 278 ff.: prepositions.

136. Ibid., p. 234.

137. Blachere, Le Coran, vol. 2, pp. 776-77 n. 172.

138. Indeed this verse is considered by many grammarians and linguists as a good example of aspect in Arabic. "Aspect" is a technical term used in the grammatical description of verbs referring primarily to the way the grammar marks the duration or type of temporal activity denoted by the verb; the contrast is often between perfective and imperfective, between the completion of an action, and duration without specifying completion. Blachere and GaudefroyDemombynes write: "He said [to Adam]: `Be,' and he was ... , that is to say: he started to exist and continued to live; the use of the perfect would have supposed an accepted fact, established, without the notion of duration." Blachere and Gaudefroy, Grammaire de l'arabe classique, 3 ed. (Paris, Maisonneuve et Larose, 1952), p. 254.

While the verb gala indicates that we are in the past, yakunu is presented as posterior (since it follows fa-) to the utterance of the imperative kun, whereas in French and English "he was" is coordinated to "he said" and presented as past.

139. Bell, A Commentary on the Qur'an, vol. 1, p. 378.

140. Ibid., vol. 2, p. 282.

141. Cf. August Fischer, "Eine Qoran-Interpolation," in this volume, chap. 6.2.

142. T. Noldeke et al., Geschichte des Qorans, 1st ed. (Gottingen, 1860), pp. 70-174; 2nd ed. (Leipzig, 1909-1938), pp. 87-234.

143. Karl Opitz, Die Medizin im Koran (Stuttgart, 1906), p. 63.

144. See Bayda wi, Anwar al-tanzil wa-asrar al-ta'wil, ed. H. O. Fleischer (Leipzig, 1846-1848), vol. 2, p. 6. I. Goldziher, Introduction to Islamic Theology and Law, trans. A. & R. Hamori (Princeton, N.J.: Princeton University Press, 1981), pp. 28-30.

145. Bell, A Commentary on the Qur'an, vol. 1, p. 49.

146. Ibid., pp. 50-51.

147. TransState Islam, Special Double Issue (spring 1997): 23.

148. F. Leemhuis, "Origins and Early Development of the tafsir Tradition," in Approaches to the History of the Interpretation of the Qu'ran, ed. A. Rippin (Oxford, 1988), p. 14.

149. J. D. McAuliffe, "Quranic Hermeneutics: The Views of al-Tabari and Ibn Kathir," in Approaches to the History of the Interpretation of the Qu'ran, ed. A. Rippin (Oxford, 1988), pp. 46-62.

150. See H. Berg, The Development of Exegesis in Early Islam: The Debate over the Authenticity of Muslim Literature from the Formative Period (London: Curzon Press, 2000).

151. Rev. J. M. Rodwell, The Koran Translated (1861; reprint London: E. P. Dutton, 1921), p. 7; emphasis added by the author.

152. M. Schub, "Dave and the Knave in the Cave of the Brave," ZAL 38 (2000): 88-90.

153. A. Rippin, "Muhammad in the Qur'an: Reading Scripture in the 21st Century," in The Biography of Muhammad: The Issue of the Sources, ed. H. Motzki (Leiden: Brill, 2000), pp. 299-300.

154. Ibid., p. 307.

155. "North Semitic Alphabet," "Aramaic Alphabet," "Arabic Alphabet," in EB (1999-2000).

156. I. J. Gelb, A Study of Writing, 2d, ed. (Chicago: University of Chicago Press, 1962), reflecting the state of research up to the end of the 1950s, seems to think that even South Semitic must be descended from the Phoenician syllabary, as much as Aramaic. Equally, Beeston thinks that the southern alphabets derive from the same stock as the Phoenician one: "Background Topics," in Arabic Literature to the End of the Umayyad Period, ed. Beeston, Johnstone, et al. (Cambridge: Cambridge University Press, 1983), p. 10. However, the article "Arabic Alphabet," in EB (1999-2000) seems to be more cautious.

157. "Aramaic Alphabet," in EB (1999-2000).

158. J. F. Healey, "The Early History of the Syriac Script, A Reassessment," JSS 45, no. 1 (spring 2000): 64-65.

159. J. F. Healey, "The Nabataean Contribution to the Development of the Arabic Script," Aram 2 (1990): 93-98; "Nabataean to Arabic: Calligraphy and Scropt Development among the Pre-Islamic Arabs," Manuscripts of the Middle East 5 (1990-91): 41-52.

160. E. Tov, ed., Companion Volume to the Dead Sea Scrolls Microfiche Edition (Leiden, 1995).

161. F. Briquel-Chatonnet, "De l'arameen a l'arabe: quelques reflexions sur la genese de 1'ecriture arabe," in Scribes et manuscrits du Moyen-Orient, ed. F. Deroche and F. Richard (Paris, 1997), pp. 135-49.

162. J. F. Healey, The Early Alphabet (Los Angeles, 1990), p. 51.

163. "Background Topics," p. 11.

164. Ibid., p. 12.

165. Ibid., p. 13.

166. G. Luling, On the Pre-Islamic Koran (Amherst, N.Y.: Prometheus Books, forthcoming), pp. 1-4.

167. C. J. Adams, "Quran: The Text and Its History," in ER, ed. M. Eliade (New York: Macmillan, 1987), pp. 157-76.

168. 'Abd al 'Al Salim Makram (wa-) Ahmad Mukhtar `Umar (hdad): Mu jam al-gira'at al-qur'aniyyah, ma`a maqadimmah ft l-gira'at wa-ashar al- qurra', I-VIII (Al-Kuwayt: Dhat as-Salasil 1402-1405/1982-1985).

169. A. Guillaume, Islam (Harmondsworth, Penguin Books, 1978), p. 189.

170. D. S. Margoliouth, "Textual Variations of the Koran," in The Origins of the Koran, ed. Ibn Warraq (Amherst, N.Y.: Prometheus Books, 1998), p. 159.

171. Ibid.

172. See A. Rippin, "Qur'an 21:95: A Ban Is Upon Any Town," JSS 24 (1979): 43-53: ". . . the variants still show traces of their original intention: to explain away grammtical and lexical difficulties. While obviously this is not true of all variant readings in the Qur'an, many variants being too slight to alleviate any problem, in Surah 21:95 and in many others the exegetical nature of Qur'anic variants is apparent" (p. 53).

173. See A. Rippin, "Qur'an 7:40, Until the Camel Passes through the Eye of the Needle," Arabica 27, fasc. 2, pp. 107-13. "Variants such as those for Surah 7:40 were created when polemically based pressures on the exegetes were the strongest and the attitudes towards the Qur'anic text less confining" (p. 113).

174. See especially Answering Islam: A Christian-Muslim Dialog, http://

175. Rev. T. Hughes, Dictionary of Islam (1885; reprint, Delhi, 1988), p. 520

176. D. Margoliouth, Mohammed and the Rise of Islam (London, 1905), p. 139

177. A. Yusuf Ali, The Holy Qur'an (Lahore, 1934), vol. 1, p. 673.

178. Ibn Salama, al-Nasikh wall-mansukh (Cairo, 1899), p. 184, referred to by D. Powers, "The Exegetical Genre nasikh al-Qur'an," in Approaches to the History of the Interpretation of the Qur'an, ed. A.Rippin (Oxford, 1988), p. 130.

179. D. B. Macdonald, "Kadar," in EII.

180. Hughes, Dictionary of Islam, p. 472.

181. Neal Robinson, Discovering the Qur'an (London, 1996), p. 65.

182. Ibid., p. 66.

183. John Burton, The Collection of the Qur'an (Cambridge, 1977), pp. 235-37.

184. Robinson, Discovering the Qur'an, p. 67.

185. D. Powers, "The Exegetical Genre nasikh al-Qur'an," in Approaches to the History of the Interpretation of the Qu'ran, ed. A. Rippin (Oxford, 1988), p. 123.

186. Robinson, Discovering the Qur'an, p. 75.

187. A. Rippin, "The Function of the asbab al-nuzul in Qur'anic Exegesis," BSOAS 51 (1988): 1-20, also in Ibn Warraq, ed., The Quest for the Historical Muhammad (Amherst, N.Y.: Prometheus Books, 2000), pp. 392-419.

188. G. R. Hawting, The Idea of Idolatry and the Emergence of Islam. From Polemic to History (Cambridge, 1999), pp. 31-32.

189. G. H. A. Juynboll, review of Quranic Studies by John Wansbrough, in JSS 24 (1979): 293-96.

190. Warraq, The Quest for the Historical Muhammad, pp. 74-75.

191. Al-Rashid, Sald `Abd al IAziz, in Kitabat islam iyyah min Makkah al- mukarramah (al-Riyad: Fahd al-Wataniyyah, 1416/1995).

192. Burton, The Collection of the Qur'an, p. 225.

193. Ibid., p. 219.

194. J. Bellamy, "Some Proposed Emendations to the Text of the Koran," in this volume, p. _.

195. A. E. Housman, Selected Prose, ed. John Carter (Cambridge, 1961), pp. 131-44.

196. Ibid., p. 131.

197. Ibid. p. 132.

198. Ibid., p. 135.

199. Ibid., pp. 144-45.

200. H. Hirschfeld, New Researches into the Composition and Exegesis of the Qoran (London, 1902), p. 137.

201. S. de Sacy, Journal des savants (1832), p. 535 sq.; G. Weil, Historisch- Kritische Einleitung in den Koran, 2d ed. (Bielefeld, 1878), p. 52, A. Sprenger, Das Leben and die Lehre des Mohammad (Berlin, 1861-65), vol. 3, p. 164.

202. Hirschfeld, New Researches into the Composition and Exegesis of the Qoran, p. 139.

203. W. M. Watt and Richard Bell, Introduction to the Qur'an (Edinburgh, 1970), p. 53.

204. Ibn al-Jawzi, Wafa, p. 32a; idem Talqih (ms. Asir effendi, Istanbul), II, p. 3a; Anonymous, Sira (Berlin, no. 9602), p. 155a; al-Barizi (Berlin, no. 2569), p. 81b; Magrizi, Imta, III; Sibt ibn al-Jawzi, Mirat at az-zaman, II (ms. Kuprulu, Istanbul), p. 149b.

205. Burton, The Collection of the Qur'an, pp. 233-34.

206. I hope to publish extracts in English in an anthology in the near future.

207. I was lucky enough to obtain a photocopy of the third volume at New York Public Library. Two of the greatest modern scholars of the Koran did not possess the third volume, and were happy to receive a photocopy from me. What I have called volume three is, in fact, Notes Complementaires II of Deuxieme Fascicule.

208. A. Sprenger, Das Leben and die Lehre des Mohammad, 2d ed., p. 533.

209. P. Casanova, Mohammed et la Fin du Monde (Paris, 1911-21), pp. 3-4.

210. Watt and Bell, Introduction to the Qur'an, pp. 53-54.

211. Noldeke, Gesch. des Q., p. 202.

212. My emphasis.

213. Casanova, Mohammed et la Fin du Monde, pp. 147 ff.

214. Watt and Bell, Introduction to the Qur'an, pp. 53-54.

215. H. Lammens, "Koran and Tradition," in The Quest for the Historical Muhammad, ed. Ibn Warraq (Amherst, N.Y: Prometheus Books, 2000), pp. 169-87.

216. T. Andrae, "Die Legenden von der Berufung Muhammeds," Le Monde Oriental 6 (1912): 5-18.

217. A. Rippin, "Muhammad in the Qur'an: Reading Scripture in the 21st Century," in The Biography of Muhammad: The Issue of the Sources, ed. H. Motzki (Leiden, Brill, 2000), pp. 299-300.

218. M. Schub, "Quran 9.40, Dave and the Knave in the Cave of the Brave," ZAL 38 (2000): 88-90.

219. J. Wansbrough, The Sectarian Milieu (Oxford, 1978), p. 52.

220. Ibid., p. 69.

221. H. Muller, Die Propheten in ihrer urspriinglichen Form (Vienna, 1896).

222. R. Geyer, "Zur Strophik des Qurans," WZKM 22 (1908): 265-86, chap. 8.1 in present volume.

223. C. Luxenberg, Die Syro-Aramaische Lesart des Koran (Berlin, Verlag Das Arabische Buch, 2000).

224. P. R. Davies, In Search of "Ancient Israel" (1992; reprint, Sheffield, 1999), p. 13.

225. John Wansbrough, Res Ipsa Loquitur: History and Mimesis (Jerusalem: Israel Academy of Sciences and Humanities, 1987), p. 10.

226. David Hall, "History, Literature and Religion," in New Humanist (September 2000): 13.

227. C. V. Langlois and C. Seignobos, Introduction to the Study of History, trans. G. G. Berry (London, 1898), p. 69.

228. Ibid., p. 156; italics in original.

229. Ibid., p. 157; italics in original.

230. Ibid., pp. 166-70.

231. Davies, In Search of "Ancient Israel. "

232. K. W. Whitelam, The Invention of Ancient Israel: The Silencing of Palestinian History (London and New York: Routledge, 1996).

233. N. P. Lemche, The Israelites in History and Tradition (London: SPCK, 1998).

234. T. L. Thompson, Early History of the Israelite People: From the Written and Archaeological Sources (Leiden: E. J. Brill, 1992).

235. J. van Seters, Prologue to History: The Yahwist as Historian in Genesis (Westminster: John Knox Press, 1992).

236. G. Garbini, History and Ideology in Ancient Israel (London: SCM Press, 1988).

237. Wansbrough, Res Ipsa Loquitur, p. 11.

238. Lemche, The Isrealites in History and Tradition, p. 88.

239. Ibid., p. 96.

240. Ibid., p. 165.

241. G. R. Hawting, The Idea of Idolatry and the Emergence of Islam. From Polemic to History (Cambridge, 1999).

242. G. R. Hawting, "John Wansbrough, Islam, and Monotheism," in The Quest for the Historical Muhammad, pp. 516-17.

243. Ibid., p. 521.

244. J. Wansbrough, Quranic Studies (Oxford, 1977), p. ix.

245. A. Rippin, Muslims. Their Religious Beliefs and Practices, Vol. 1:The Formative Period (London, 1991), p. ix.

246. W. J. Hanegraaf, New Age Religion and Western Culture: Esotericism in the Mirror of Secular Thought (Brill, 1996), pp. 25-26.

247. A. Geiger, "Judaism and Islam," in The Origins of the Koran; H. Hirschfeld, Judische Elemente im Koran (Berlin, 1878); Beitrage zur Erklarung des Koran (Leipzig, 1886); New Researches into the Composition and Exegesis of the Qoran; A. Katsch, Judaism in Islam (New York, 1954); D. Sidersky, Les Origines des legendes musulmanes dans le Coran (Paris, 1933); H. Speyer, Die Biblischen Erzahlungen im Qoran (Hildesheim, 1961); B. Heller, "Recits et personnages bibliques dans la legende mahometane," REJ 85 (1928): 113-36; "La legende biblique dans I'Islam," REJ 98 (1934): 1-18; P. Jensen, "Das Leben Muhammeds and die David-Sage," Der Islam 12 (1922): 84-97; I. Schapiro, Die haggadischen Elemente im erzalenden Teil des Korans, vol. 1 (Leipzig, 1907); H. Schwarzbaum, "The Jewish and Moslem Versions of Some Theodicy Legends," Fabula 3 (1959-60): 119-69; C. Gilliot, "Les 'informateurs'juifs et chretiens de Muhammad. Reprise d'un probleme traite par Aloys Sprenger et Theodor Noldeke," JSAI 22 (1998): 84-126; C. C. Torrey, The Jewish Foundation of Islam (New York, 1933), reprinted in The Origins of the Koran.

248. F. E. Peters, "The Quest of the Historical Muhammad," in The Quest for the Historical Muhmmad, p. 455.

249. Hall, "History, Literature and Religion," pp. 10-14.

250. H. Lammens, "Koran and Tradition-How the Life of Muhammad Was Composed," in The Quest for the Historical Muhammad, p. 455.

251. F. E. Peters, "The Quest of the Historical Muhammad," p. 458.

252. Ibid., p.455

253. Hall, "History, Literature and Religion," p. 12.

254. A. Brockett, "Studies in Two Transmissions of the Qur'an" (Ph.D. disc., University of St. Andrews, 1984), p. 13.

255. Ibid., p. 19.

If you find an error please notify us in the comments. Thank you!