Sources of the Quran

Revised by Montgomery Watt




Contact Truthnet

Sources of the Quran


Chapter 1
The Historical Context

Chapter 2

Mohammed's Prophetic Experience

Chapter 3

The History of the Text

Chapter 4

The external form of the Quran

Chapter 5

The features of Quranic style

Chapter 6

The Shaping of the Quran

Chapter 7
The Chronology of the Quran

Chapter 8
The names of the revealed message

Chapter 9
The Doctrines of the Quran

Chapter 10
Muslim Scholarship and the Quran

Chapter 11
The Quran and Occidental Scholariship



Bell's Introduction to the Quran
Revised by Montgomery Watt

Chapter 3: The History of the Text

1 . The 'collection' of the Qur'n

abu bakr

(a) The 'collection' under Ab-Bakr.

There is a widespread report, found in many slightly differing forms, telling of a 'collection' of the quran Qurn in the caliphate of abu bakr Ab-Bakr (632-4). According to this report 1 umar Umar ibn al khattab ibn-al-Khab (who succeeded as caliph in 634) was perturbed by the fact that in the battle of yamama Yamma during the 'wars of the apostasy (ridda)' many of the 'readers' of the quran Qurn were killed. Since these were the men who had parts of the quran Qurn by heart, umar Umar feared that, if more of them died, some of the quran Qurn would be irretrievably lost. He therefore counselled abu bakr Ab-Bakr to make a 'collection' of the quran Qurn. At first abu bakr Ab-Bakr hesitated to undertake a task for which he had received no authority from muhammad Muammad, but in the end he gave his approval and commissioned zayd ibn thabit Zayd-ibn-Thbit. The latter, who had been one of muhammad Muammad's secretaries, had no illusions about the difficulty of the task, but at length agreed. As mentioned above (p. 32), he then proceeded to 'collect' the quran Qurn 'from pieces of papyrus, fiat stones, palm-leaves, shoulder-blades and ribs of animals, pieces of leather and wooden boards, as well as from the hearts of men'. The last passage to be found was 9.128/9f.-the two closing verses of sura 9. Zayd wrote what he 'collected' on sheets (suhuf uuf) of equal size and gave them to abu bakr Ab-Bakr. On the latter's death they passed to umar Umar, and on umar Umar's death to his daughter hafsa afa, a widow of the prophet.

This tradition is open to criticism on a number of grounds. For one thing it seems to assume that up to the time of muhammad Muammad's death there had been no authoritative record of the revelations and no attempt to bring some order into them but it has been already shown that this is unlikely. Then there are many discrepancies between this tradition and others and between the different versions of this tradition. Thus there is no unanimity about the originator of the idea of collecting the quran Qurn; generally it is said to have been umar Umar, but sometimes abu bakr Ab-Bakr is said to have commissioned the 'collection' on his own initiative. On the other hand, there is a tradition which says umar Umar was the first to 'collect' the quran Qurn and completely excludes abu bakr Ab-Bakr. 2 Again, the reason given for the step, namely the death of a large number of 'readers' in the battle of yamama Yamma has also been questioned. In the lists of those who fell in that campaign, very few are mentioned who were likely to have had much of the quran Qurn by heart. 3 Those killed were mostly recent converts. Besides, according to the tradition itself, much of the quran Qurn was already written in some form or other, so that the death of some of those who could recite it from memory need not have given rise to the fear that parts of the quran Qurn would be lost.

Perhaps the weightiest criticism of the tradition is that an official collection of this kind might have been expected to have had wide authority attributed to it, but of this we find no evidence. Other 'collections' of the quran Qurn seem to have been regarded as authoritative in different provinces. The disputes which led to the recension of the quran Qurn under uthman Uthmn could hardly have arisen if there had been an official codex in the caliph's possession to which reference could have been made. Again the way in which umar Umar himself is represented elsewhere as insisting that the verse of stoning 4 was in the quran Qurn, is hardly consistent with his having in his possession an official collection. Lastly, and most significant of all, the suhuf uuf on which Zayd wrote the quran Qurn were, at the time when the revision came to be made, in the keeping of hafsa afa. Now hafsa afa was umar Umar's daughter, and we are apparently to assume that since umar Umar had become caliph by the time Zayd finished his work, the suhuf uuf were handed to him, and from him passed to his daughter. If Zayd's collection was an official one, however, it is hardly probable that it would pass out of official keeping, even into the hands of the caliph's daughter. That hafsa afa had a copy of the quran Qurn on suhuf uuf seems certain; but it is unlikely that it was an official copy made in the official way that tradition asserts.

It seems practically certain, then, that no complete 'collection' of the quran Qurn was officially made during the caliphate of abu bakr Ab-Bakr. The traditional account so far considered was doubtless gradually elaborated to avoid the awkward fact that the first 'collection' of the quran Qurn was made by uthman Uthmn, who was greatly disliked. On the other hand, there is no good ground for doubting that hafsa afa possessed a quran Qurn written on suhuf uuf whether this was written by herself, by Zayd, or by someone else.


(b) The 'collection' under Uthmn.

The traditional account of what led to the next step in the fixing of the form of the quran Qurn implies that serious differences of reading existed in the copies of the quran Qurn current in the various districts. During expeditions against Armenia and Azerbaijan, we are told, disputes concerning the reading of the quran Qurn arose amongst the troops, who were drawn partly from Syria and partly from Iraq. The disputes were serious enough to lead the general, hudhayfa udhayfa, to lay the matter before the caliph, uthman Uthmn (644-56), and to urge him to take steps to put an end to these differences. The caliph took counsel with senior Companions of the Prophet, and finally commissioned Zayd thabit ibn-Thbit to 'collect' the quran Qurn. With Zayd were associated three members of noble Meccan families, abd allah Abd-Allh ibn-az-Zubayr, said Said ibn al as ibn-al- and abd ar rahman Abd-ar-Ramn ibn al harith ibn-al-rith. One of the principles they were to follow was that, in case of difficulty as to the reading, the dialect of Quraysh, the tribe to which the Prophet belonged, was to be given the preference. The whole quran Qurn was carefully revised and compared with the suhuf uuf which had been in hafsa afa's keeping and which were returned to her when the work was finished. Thus an authoritative text of the quran Qurn was established. A number of copies were made and distributed to the main centres of Islam. As to the exact number of these standard codices, and the places to which they were sent, the account varies; but probably one copy was retained in Medina, and one was sent to each of the towns, Kufa, Basra and Damascus, and possibly also to Mecca. Previously existing copies are said to have been then destroyed, so that the text of all subsequent copies of the quran Qurn should be based upon those standard codices.

This traditional account of the 'collection' of the quran Qurn under uthman Uthmn is also open to criticisms, though they are not so serious as in the case of abu bakr Ab-Bakr's 'collection'. The most serious difficulties are those connected with the suhuf uuf of hafsa afa. Some versions of the story suggest that the work of the commissioners was simply to make a fair copy, in the dialect of Quraysh, of the material on these leaves. Some important material, however, has come to light since the publication of Friedrich Schwally's revised edition of the second volume of Nldeke's Geschichte des quran Qurns in 1919. In particular there is a story of how the caliph marwan Marwn when governor of Medina wanted to get hold of the 'leaves' of hafsa afa to destroy them, and eventually on her death persuaded her brother to hand them over. 5 marwan Marwn was afraid lest the unusual readings in them might lead to further dissension in the community. On the whole it is unlikely that this story has been invented, for it implies that the 'leaves' of hafsa afa were unsuitable as a basis for the official text. The 'leaves' are not to be confused with a codex of the new official text said to have been given to hafsa afa. The most likely solution of the problem is to hold that, while hafsa afa may well have had 'leaves' on which she had written down many suras sras, hers was in no respect an official 'collection'. It is perhaps specially mentioned to link up this account with that of the first 'collection' under abu bakr Ab-Bakr. On the whole, then, it seems unlikely that the 'leaves' of hafsa afa were of primary importance. They cannot have contained more than what had been arranged in the 'book' by muhammad Muammad at the time of his death; and they can hardly have been the sole or main basis of the uthmanic Uthmnic text.

Other criticisms are minor. There are various lists of the persons who helped Zayd. Schwally shows that the suggested names are all improbable. 6 He also questions the instruction to write the revelations in the dialect of the Quraysh (the tribe of Mecca) on the ground that the quran Qurn is in a partly artificial, literary language. 7 Perhaps the function of the commissioners was to help to 'collect' revelations from sources known to them. Schwally dismisses this possibility on the ground that the commission was mainly concerned to produce a fair copy of hafsa afa's 'leaves'; but since the new material shows that hafsa afa's 'leaves' were unsuitable as a basis for the new edition, Schwally's objection falls. Indeed, there is no reason now for rejecting two points in the traditional account: (1)the commissioners were to collect all the pieces of revelation they could find; (2) where men had remembered it with dialectal variations of the literary language, they were to make the Meccan forms standard.

This establishment of the text of the quran Qurn under uthman Uthmn may be dated somewhere between 650 and his death in 656. It is the cardinal point in what may be called the formation of the canon of the quran Qurn. Whatever may have been the form of the quran Qurn previously, it is certain that the book still in our hands is essentially the uthmanic Uthmnic quran Qurn. uthman Uthmn's commission decided what was to be included and what excluded; it fixed the number and order of the suras, and the 'outline' of the consonantal text (that is, its shape when the dots distinguishing letters are omitted). If we remember that to preserve every smallest fragment of genuine revelation was an ineluctable requirement, the commission under Zayd must be adjudged to have achieved a wonderful piece of work.


2. The pre-Uthmnic codices

While uthman Uthmn's effort to obtain uniformity throughout the caliphate in the quranic Qurnic text must on the whole have been successful in practice, the uthmanic pre-Uthmnic or non-canonical readings were by no means forgotten. Most of the larger commentaries on the quran Qurn such as those of tabari a-abari and zamakhshari az-Zamakhshar refer to such non-canonical readings from time to time. One or two Muslim scholars in the early tenth century made a special study of the early masahif maif (sing. mushaf muaf). One such work, the kitab Kitb masahif al-maif of ibn abi dawud Ibn-Ab-Dwd (d. 928), survives and was published in 1937. It contains a note to the effect that he uses mushaf muaf in the sense of 'reading' or 'set of readings', but the modern editor, Arthur Jeffery, thinks that this is an interpolation, and that when the author speaks of 'the mushaf muaf of N' he means an actual written codex. 8 From the information given by ibn abi dawud Ibn-Ab-Dwd and from other sources Jeffery has drawn up a list of fifteen 'primary codices' and almost as many 'secondary codices', and from most of these he has collected at least a few non-canonical readings. The presumption is that at an early period certain Muslims began to write down as much as they could of the quran Qurn. At first these written collections would not necessarily be of interest to the Muslims in general, since men accustomed to the dominance of oral tradition tend to be suspicious of writing and some Muslim scholars said the phrase 'to collect the quran Qurn' simply meant 'to remember the whole of the quran Qurn', In the course of time, however, some of the written collections came to have special authority in various great centres of the Islamic world. In particular that of abd allah Abd-Allh ibn masud ibn-Masd was held in high regard in Kufa, and that of Ubayy kab ibn-Kab in most parts of Syria. 9

No copies exist of any of the early codices, but the list of variant readings from the two just mentioned is extensive, running to a thousand or more items in both cases. masud Ibn-Masd (d. 653) was for a time a personal servant of muhammads Muammad's, but eventually settled in Kufa where he became an authority on religious matters on account of his interest in the subject and his close association with the Prophet. Ubayy kab ibn-Kab was a Muslim from Medina who frequently acted as secretary for muhammad Muammad. The variant readings in the codices of both these men chiefly affect the vowels and punctuation, but occasionally there is a different consonantal text. For both, too, we have lists of the suras, and it is noteworthy that these differ from each other and also from the uthmanic Uthmnic list in the order in which the suras are arranged. On the whole the longer suras come first as in the standard order. The names of the suras, too, are mostly the same as those normally used although in many cases alternatives exist (as will be explained in the next chapter); but it is conceivable that this uniformity in names is due to later transcribers of the lists who substituted the common names for unusual ones. There is, of course, no way of being certain that the contents represented by the names are identical; but on the other hand there is no indication that the contents were different except in respect of the variants noted.

Questions of the omission or addition of suras are also specially connected with these two early codices. The lists giving the order in which they placed the suras are not complete, but it would be rash to infer from this that suras absent from the list had been omitted from the codex. There are explicit statements, however, that masud Ibn-Masd omitted altogether the last two suras [113, 114], the muawwidhatan Muawwidhatn or suras 'of taking refuge with God'; but these are a kind of charm or prayer of commendation, and may not originally have been regarded as part of the quran Qurn. It is also doubtful whether masud Ibn-Masd included the first sura or fatiha Ftia. This also is a prayer, whose function is not unlike that of the Lord's Prayer in Christianity. Some scholars have argued that, if it were part of the quran Qurn it should have been preceded by the word qul, 'say', that is, a command to use it as a prayer. Ubayy seems to have included the three suras mentioned, and also to have had two other suras which are not in the standard text of the quran Qurn. The text of these suras has been preserved by some Muslim scholars. They are short prayers and, as in the case of the fatiha Ftia, one might have expected them to be preceded by the word 'say'. Short as the text of them is, there are a number of points where the linguistic usage is not paralleled in the quran Qurn. Schwally, while noting this, thought they might nevertheless go back to muhammad Muammad, 10 but this is extremely doubtful. It is conceivable that they were used by Muslims in muhammad Muammad's time, but they cannot have been part of the quran Qurn. Of a different character are omissions of parts of the text for dogmatic reasons. Such is the declaration of the kharijite Khrijite sub-sect of the maymuniyya Maymniyya that the sura of Joseph (12) was not part of the quran Qurn. 11 Their reason for this, however, seems to have been that it was not fitting that a love-story should be included in the quran Qurn. This declaration, then, hardly contributes to our knowledge of the history of the text.

Thus on the whole the information which has reached us about the uthmanic pre-Uthmnic codices suggests that there was no great variation in the actual contents of the quran Qurn in the period immediately after the Prophet's death. The order of the suras was apparently not fixed, and there were many slight variations in reading, but of other differences there is no evidence. The modern scholar, familiar with the way in which textual studies have elucidated the stages in the development of early European literary texts, would like to achieve some thing similar in the case of the quran Qurn, but for this the available information is insufficient, except in respect of the relation of the secondary codices to the primary codices.


3. The writing of the Qurn and early textual studies

While the promulgation of the uthmanic Uthmnic text was a major advance towards uniformity, its importance may easily be exaggerated. For one thing, knowledge of the quran Qurn among the Muslims was based far more on memory than on writing. For another thing the script in which the quran Qurn was originally written was what is referred to as a scriptio defectiva in contrast to the scriptio plena in which it is now written. The nature of the early scripts is fairly well known from the study of early quran Qurns and fragments in some of the great libraries. 12 In the earliest examples only consonants are written, and even these are not adequately distinguished from one another, since the same written shape may sometimes indicate either of two consonants. One might say, then, that this scriptio defectiva was little more than an elaborate mnemonic device. It presupposed in the 'reader' some degree of familiarity with the text. A man with no knowledge of the quran Qurn but who understood the script would have had great difficulty in deciphering the writing, though not so much as a person unaware of the structure of Arabic words might suppose. Certainly, however, extensive memorization is presupposed, and this is the background of the improvement of the writing and the growth of textual studies. There was a special class of men, the qurra qurr or quran reciters Qurn-reciters (sometimes called 'readers'), who specialized in memorizing the sacred text. As the centuries passed their social character changed; eventually we find that this study of the text is chiefly associated with philology, and is a regular part of higher education.

By the time of the caliph abd al malik Abd-al-Malik (685-705) the inadequacy of the existing script was clear to leading Muslims and improvements began to be made. The problem of the incorrect copying of the defective script had also to be dealt with. The traditional accounts of the passage to the scriptio plena do not tally with one another, nor with the findings of palaeography. It is virtually certain that the scriptio plena did not come into existence all at once, but only gradually by a series of experimental changes. One of the more probable traditional accounts ascribes the introduction of diacritical marks and vowel points to the initiative of hajjaj al-ajjj, probably during his governorship of Iraq (694-714). The actual work is said to have been done by scribes such as Nasr asim ibn-im (d. 707) and yahya Yay yamur ibn-Yamur (d.746). It is hardly possible that the scriptio plena should have been introduced all at once by abu aswad Ab-l-Aswad ad duali ad-Dual (d. 688), as is sometimes suggested. Existing copies of the quran Qurn illustrate different methods of obviating deficiencies of the script; e.g. dots of different colours for the vowels instead of the signs now in current use. The chief matters to be dealt with were: (a) distinguishing between consonants with a similar shape; (b) the marking of long vowels, which eventually was mostly done by adding the consonants ahif waw, ya y; (c) the marking of short vowels; (d) certain other matters such as the doubling of consonants and the absence of a vowel after a consonant.

The process of improving the script was completed towards the end of the ninth century. It now became possible to enforce a greater measure of uniformity than was conceivable with the original script. It is not surprising, then, to find in the early tenth century a series of moves to ensure a measure of uniformity. These are chiefly associated with the name of mujahid Ibn-Mujhid (859-935). 13 He was not, of course, the first to concern himself with securing uniformity in the text. Malik ibn-Anas (d. 795), the great scholar of Medina and founder of the Malikite legal rite, had explicitly stated that the performance of the worship behind someone who used the readings of masud Ibn-Masd was invalid. 14 The more precise script, however, enabled mujahid Ibn-Mujhid to make more exact regulations. As a result of his studies he wrote a book entitled 'The Seven Readings' (qiraat Al-qirt saba as-saba). He based himself on a Tradition to the effect that muhammad Muammad had been taught to recite the quran Qurn according to seven ahruf aruf, interpreted to mean seven sets of readings', though ahruf aruf is the plural of harf arf which is properly 'letter'. 15 His conclusion was that the set of readings of each of seven scholars of the eighth century was equally valid, but that these seven sets alone were authentic.

The conclusions of the scholar were made effective by the action of the courts. In 934 a scholar called Ibn-Miqsam 16 was forced to renounce the view that one was entitled to choose any reading of the consonantal outline that was in accordance with grammar and gave a reasonable sense. This decision was tantamount to an insistence that only the seven sets of readings were valid. In April 935 (about four months before the death of ibn mujahid Ibn-Mujhid) another scholar, ibn shannabudh Ibn-Shannabdh, was similarly condemned and forced to retract his view that it was permissible to make use of the readings of masud Ibn-Masd and Ubayy kab ibn-Kab. Up to this time some scholars had apparently been in the habit of making some use of these readings in commenting on and elucidating the quran Qurn. The readings of ali Ali ibn abi talib ibn-Ab-lib were also rejected by mujahid Ibn-Mujhid.

The seven sets of readings accepted by mujahid Ibn-Mujhid represented the systems prevailing in different districts. There was one each from Medina, Mecca, Damascus and Basra, and three from Kufa. For each set of readings (qiraa qir'a), there were two slightly different 'versions' (sing. riwaya riwya). The whole may be set out in tabular form. 17



District             Reader First Rw       Second Rw 


Medina Nfi (d. 785)  Warsh (812)     Qln (835)    

Mecca  Ibn-Kathr (737)         al-Bazz (854) Qunbul (903)   

Damascus         Ibn-mir (736)           Hishm (859)   Ibn-Dhakwn (856)     

Basra   Ab-mir (770)         ad-Dri (860)   as-Ss (874)

Kufa     Aim (744)     af (805)       Shuba (809)  

Kufa     amza (772)    Khalaf (843)     Khalld (835)  

Kufa     al-Kis'i (804)              ad-Dri (860)   Ab-l-rith    (854)   


While ibn mujahid Ibn-Mujhid's system of seven readings came after a time to be generally accepted in theory, only one of the fourteen versions, that of hafs af from asim im, is now widely used in practice. The new standard Egyptian edition reproduces this version and thus gives it a certain canonical supremacy. The restriction to seven readers was not immediately approved by all Muslim scholars. Some spoke of ten readers (with two versions each), while others had fourteen, though with only one version of at least the last four. The' three after the seven' were:

Medina             abu jafar Ab-Jafar (d. 747)

Basra               yaqub Yaqb hadrami al-aram (820)

Kufa                 Khalaf (also rawi rw of hamza amza) (843).

The 'four after the ten' were:

Mecca              muhaysin Ibn-Muayin (740)

Basra               yazidi al-Yazd (817)

Basra               hasan al-asan basri al-Bar (728)

Kufa                 amash al-Amash (765).

These different lists are a reflection of fierce discussions among scholars of different schools and the struggle of divergent tendencies in the Islamic community; but a detailed history of these matters from a modern standpoint remains to be written. There have been Muslim scholars who prided themselves on knowing the quran Qurn according to every one of the seven readings. The existence of variants, however, has been found inconvenient, especially in modern times. The ordinary Muslim is mostly unaware of the existence of the seven sets of readings; and the modern heretical sect of the ahmadiyya Amadiyya appears to deny, in the interests of propaganda, even the existence of the uthmanic pre-Uthmnic variants.


4. The authenticity and completeness of the Qurn

If one asks what guarantee there is that the quran Qurn as 'collected' in the caliphate of uthman Uthmn is a correct record of the revelations as they were originally received and proclaimed by muhammad Muammad, the modern scholar will seek an answer first of all in the quran Qurn itself and in a comparison of its contents with what he takes to be reliably known about the Prophet's life.

It may be noted to begin with that uthman Uthmn's revision was based on written documents previously existing. The official collection by express authority of the caliph abu bakr Ab-Bakr is, as has been seen, somewhat doubtful. A mass of written documents of some kind, however, was in hafsa afa's possession. If we reject the assumption that they were an official collection made by Zayd, we must find some other explanation of what they were. It is clear that they were regarded as authoritative and were used in producing uthman Uthmn's quran Qurn. Other 'collections' of the quran Qurn were in existence, and there must have been a considerable number of people who knew these, or parts of them, by heart. If any great changes by way of addition, suppression or alteration had been made, controversy would almost certainly have arisen; but of that there is little trace. uthman Uthmn offended the more religious among Muslims, and ultimately became very unpopular. Yet among the charges laid against him, that of having mutilated or altered the quran Qurn is not generally included, and was never made a main point. The shia Shia, it is true, has always held that the quran Qurn was mutilated by the suppression of much which referred to ali Ali and the Prophet's family. This charge, however, is not specially directed against uthman Uthmn, but just as much against the first two caliphs, under whose auspices the first collection is assumed to have been made. It is also founded on dogmatic assumptions which hardly appeal to modern criticism. On general grounds then, it may be concluded that the uthmanic Uthmnic revision was honestly carried out, and reproduced, as closely as was possible to the men in charge of it, what muhammad Muammad had delivered.

Modern study of the quran Qurn has not in fact raised any serious question of its authenticity. The style varies, but is almost unmistakable. So clearly does the whole bear the stamp of uniformity that doubts of its genuineness hardly arise. The authenticity of a few verses has indeed been questioned. The great French scholar Silvestre de Sacy expressed doubts regarding 3.144/38. 18 This speaks of the possible death of muhammad Muammad, and is the verse said in a well-known tradition to have been quoted by abu bakr Ab-Bakr, when umar Umar refused to believe the report of the death of the Prophet, which had just taken place. Gustav Weil extended these doubts to a number of other passages which imply the mortality of the Prophet: 3.185/2; 21.35/6f.; 29.57; 39.30/1. 19 abu bakr Ab-Bakr, however, is hardly likely to have invented 3.144/38 for the occasion; nor does the statement that umar Umar and others professed never to have heard such a verse, weigh very much. The complete quran Qurn was not circulating among muhammad Muammad's followers in written form for them to study, and a verse once delivered might easily have been forgotten in the course of years, even by one who happened to hear it. If the verse does not fit smoothly into the context, that is probably because it is a substitution for the one which follows, as the recurrence of the same rhyme-phrase suggests. It fits admirably into the historical situation, for it is a reference, put into an address delivered before uhud Uud and re-delivered after the defeat, to the report which had spread during the battle and had no doubt contributed to the rout, that muhammad Muammad had been killed. There is no reason to question the authenticity of a verse so suited to the circumstances.

As for the other verses which imply the mortality of the Prophet, Schwally 20 has pointed out how they fit well into their contexts and are quite in accord with the rest of the quran Qurn. The humanity and mortality of the Prophet were part of the controversy between him and his opponents, and to take that out of the quran Qurn would be to remove some of its most characteristic portions.

Weil 21 also questioned the authenticity of the famous verse in which reference is made to the night journey to Jerusalem [17.1]. He argued that there are no other references to such a night journey in the quran Qurn, that it is contrary to muhammad Muammad's usual claim to be simply a messenger and not a wonder-worker, that so far as there is any basis for the later legend in muhammad Muammad's life, it is merely a dream or vision, and that the verse has no connection with what follows. As matters of fact these arguments are correct; but they hardly bear the inference based on them. If we take the verse by itself, without the structure of later legend built upon it, there is nothing in it very much out of keeping with other claims made for muhammad Muammad; and there are so many unconnected verses in the quran Qurn that we can hardly make that an argument against this one in particular.

Finally, Weil 22 questioned 46.15/14 on the ground that Tradition makes it refer to abu bakr Ab-Bakr and that presumably it was invented in his honour. No one who knows the traditional exegesis of the quran Qurn, however, will pay much attention to such a statement. Tradition is full of guesses about the particular person to whom a verse refers. This verse is quite general, and simply develops an injunction several times repeated in the quran Qurn.

Hirschfeld 23 has questioned the authenticity of certain other verses, in which the name muhammad Muammad occurs, on the ground that this was not the Prophet's real name but was bestowed upon him later. There may be something suspicious in such a name, meaning 'Praised', being borne by the Prophet; but even if it were an assumed name, it might have been adopted in his own lifetime. It occurs, not only in the quran Qurn but in documents handed down by Tradition, notably the constitution of Medina, 24 and the treaty of hudaybiya al-udaybiya; 25 in the latter the pagan Quraysh are said to have objected to the title rasul rasl allah Allh, and to ar rahman ar-Ramn as a name of God, but raised no question about the name muhammad Muammad. Further, though it does not appear to have been common, there is evidence that muhammad Muammad was in use as a proper name before the time of the Prophet. There is therefore no reason to doubt that it was his real name.

The most serious attack upon the reliability of the book and the good faith of the collectors was that made by the French scholar, Paul Casanova, in his book, muhammad Mohammed et la fin du monde (Paris, 1911-24). His thesis is a development of the view that muhammad Muammad was moved to undertake his mission by the impression made on him by the idea of the approaching Judgement. Casanova thinks that he must have come under the influence of some Christian sect which laid great stress on the near approach of the end of the world. He considers that this formed the main theme of his early deliverances and was an essential part of his message from beginning to end of his prophetic activity; but that when no event occurred to substantiate his prophecy, the leaders of early Islam so manipulated the quran Qurn as to remove that doctrine from it, or at least conceal its prominence.

This thesis has not found much acceptance, and it is unnecessary to refute it in detail. The main objection to it is that it is founded less upon study of the quran Qurn than upon investigation of some of the byways of early Islam. From this point of view, the book still has value. When Casanova deals with the quran Qurn itself, however, his statements often display incorrect exegesis and a failure to appreciate the historical development of quranic Qurnic teaching. As to his main thesis, it is true that the quran Qurn 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 21.1 and 27.71/3f. 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 quran Qurn. All this, however, is perfectly natural if we regard the quran Qurn as reflecting muhammad Muammad's personal problems and the outward difficulties he encountered in carrying out a task to which he had set his hand. Casanova's thesis makes little allowance for the changes that must have occurred in muhammad Muammad's attitudes through twenty years of ever-changing circumstances. Our acceptance of the quran Qurn 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.

The question whether the quran Qurn, as we have it, contains all that muhammad Muammad delivered, is more difficult to answer. It is difficult to prove a negative and we cannot be certain that no part of the quran Qurn delivered by muhammad Muammad has been lost.

The quran Qurn itself speaks of the possibility of God causing muhammad Muammad to forget some passages [87.6f.]; and further states that when this happens other verses as good or better will be substituted for those forgotten [2.106/0]. It should be noted, however, that some Muslims found difficulties in such an interpretation of 2.106/0, and tried to avoid these by adopting other readings and interpretations. 26 There would seem, however, to be no good reason for rejecting the standard reading and the obvious interpretation of it, and this course has the advantage of giving an assurance that no revelation of permanent value has been omitted. There is also a Tradition which describes how muhammad Muammad heard a man reciting the quran Qurn in a mosque, and realized that the passage recited contained a verse (or verses) which he had forgotten. 27

Tradition again gives a number of verses as belonging to the quran Qurn although they do not stand in our present book. 28 The most famous of them is the 'verse of stoning', a verse in which stoning is prescribed as punishment for persons of mature age guilty of fornication. The caliph umar Umar is said to have been very positive that this was laid down in the quran Qurn, until he was convinced of the contrary by lack of evidence to support his opinion. The verse is assigned either to sura 24 or to sura 33; but the rhyme does not fit sura 33, while the prescription of stoning contradicts 24.2 where flogging is ordered. On the whole it seems unlikely that the punishment of stoning was ever prescribed in the quran Qurn, since in certain tribes in pre-Islamic times loose forms of polyandry appear to have been normal practice, and it would have been difficult to distinguish some of these from fornication. The story about umar Umar and certain Traditions about muhammad Muammad himself are probably attempts to meet the criticism that the quran Qurn differs from the Old Testament on this point. 29 Interesting also is the addition to 98.2 said to have been read by Ubayy, which began 'religion in God's sight is the moderate hanifiyya anfiyya. 30 It is noteworthy that in 3.19/17, which normally runs 'religion in God's sight is Islam', masud Ibn-Masd read hanifiyya 'anfiyya' instead of 'Islam'. 31 Since there appears to have been a time when a follower of muhammad Muammad was called a hanif anf by preference, and his religion the hanifiyya anfiyya, it may well be that these readings reflect an older form of the text. 32 Of the other verses preserved by Tradition, two or three may simply be variants of verses in the standard text; but apart from such variants there are no good reasons for thinking any of these verses from Tradition belonged to the quran Qurn, while there are grounds for holding that some did not belong.

In a different category are the so-called 'satanic verses' two (or three) verses which came after 53.19, 20 when these were originally proclaimed in public in the precincts of the kaba kaaba Kaba at Mecca. muhammad Muammad is said to have been hoping for a revelation which would have led the Meccan merchants to accept his religion, when there came to him the passage:

Have you considered lat al-Lt and uzza al-Uzz

and manat Mant, the third, the other?

These are the intermediaries exalted,

whose intercession is to be hoped for.

Such as they do not forget. 33

Later-but it is not clear how much later-muhammad Muammad realized that this could not have come from God, for he received an emended revelation in which after the first two verses there came the passage beginning

Is it the male for you and the female for him?

That would then be a crooked division.

The first passage permitted intercession to the local deities, presumably regarded as a kind of angelic being who could plead with the supreme God on behalf of their worshippers, while what was substituted was an argumentum ad hominem against the belief that such deities were 'daughters of God' and was understood as making such intercession impossible. In essentials it would seem that this account is true, since no Muslim could have invented such a story about muhammad Muammad. The story has also some support from the quran Qurn, since 22.52/I (which is said to refer to this incident) states that God 'never sent messenger or prophet before (muhammad Muammad) but that, as he desired, Satan threw (something) into his formulation', though the satanic addition was afterwards abrogated by God.

Whatever view is taken of the collection and compilation of the quran Qurn, the possibility remains that parts of it may have been lost. If, as Tradition states, Zayd in collecting the quran Qurn was dependent on chance writings and human memories, parts may easily have been forgotten. Yet the conjunction of apparently unrelated verses at certain points in the quran Qurn suggests that the editors preserved absolutely everything they came across which they had reason to believe had once been part of the quran Qurn. The hypothesis that muhammad Muammad had some way of obtaining a revised form of a revelation would lead one to suppose that he might then have discarded the older form; and something similar might be inferred from the quranic Qurnic phrase about God causing him to forget. In this way some revealed passages might be altogether lost. There is no reason, however, to think that anything of importance has gone astray. The very fact that varying and even contradictory deliverances have been preserved is strong proof that, with perhaps minor exceptions, we have the whole of what was revealed to muhammad Muammad.