A study of the compilation of the Qur'an text must begin with the character of the book itself as it was handed down by Muhammad to his companions during his lifetime. It was not delivered or, as Muslims believe, revealed all at once. It came piecemeal over a period of twenty-three years from the time when Muhammad began to preach in Mecca in 610 AD until his death at Medina in 632 AD. The Qur'an itself declares that Allah said to Muhammad: "We have rehearsed it to you in slow, well-arranged stages, gradually" (Surah 25.32).

Furthermore no chronological record of the sequence of passages was kept by Muhammad himself or his companions so that, as each of these began to be collected into an actual surah (a "chapter"), no thought was given as to theme, order of deliverance or chronological sequence. It is acknowledged by all Muslim writers that most of the surahs, especially the longer ones, are composite texts containing various passages not necessarily linked to each other in the sequence in which they were given. As time went on Muhammad used to say "Put this passage in the surah in which so-and-so is mentioned", or "Put it in such-and-such a place" (as -Suyuti, Al Itqan fii Ulum al-Qur'an, p.141). Thus passages were added to compilations of other passages already collected together until each of these became a distinct surah. There is evidence that a number of these surahs already had their recognised titles during Muhammad's lifetime, as from the following hadith:

At the same time, however, there is also reason to believe that there were other surahs to which titles were not necessarily given by Muhammad, for example Suratul-Ikhlas (Surah 112), for although Muhammad spoke at some length about it and said its four verses were the equal of one-third of the whole Qur'an, he did not mention it by name (Sahih Muslim, Vol. 2, p.387).

As the Qur'an developed Muhammad's immediate companions took portions of it down in writing and also committed its passages to memory. It appears that the memorisation of the text was the foremost method of recording its contents as the very word al-Qur'an means "the Recitation" and, from the very first word delivered to Muhammad when he is said to have had his initial vision of the angel Jibriil on Mount Hira, namely Iqra - "Recite!" (Surah 96.1), we can see that the verbal recitation of its passages was very highly esteemed and consistently practised. Nevertheless it is to actual written records of its text that the Qur'an itself bears witness in the following verse:

There is evidence, further, that even during Muhammad's early days in Mecca portions of the Qur'an as then delivered were being reduced to writing. When Umar was still a pagan he one day struck his sister in her house in Mecca when he heard her reading a portion of the Qur'an. Upon seeing blood on her cheek, however, he relented and said "Give me this sheet which I heard you reading just now so that I may see just what it is which Muhammad has brought" (Ibn Ishaq, Sirat Rasulullah, p.156) and, on reading the portion of Surah 20 which she had been reading, he became a Muslim.

It nonetheless appears that right up to the end of Muhammad's life the practice of memorisation predominated over the reduction of the Qur'an to writing and was regarded as more important. In the Hadith records we read that the angel Jibril is said to have checked the recitation of the Qur'an every Ramadan with Muhammad and, in his final year, checked it with him twice:

Some of Muhammad's closest companions devoted themselves to learning the text of the Qur'an off by heart. These included the ansari Ubayy ibn Ka'b, Muadh ibn Jabal, Zaid ibn Thabit, Abu Zaid and Abu ad-Darda (Sahih al-Bukhari, Vol. 6, pp. 488-489). In addition to these Mujammi ibn Jariyah is said to have collected all but a few surahs while Abdullah ibn Mas'ud, one of the muhajirun who had been with Muhammad from the beginning of his mission in Mecca, had secured more than ninety of the one hundred and fourteen surahs by himself, learning the remaining surahs from Mujammi (Ibn Sa'd, Kitab aI-Tabaqat al-Kabir, Vol. 2, p.457).

Regarding the written materials there are no records as to exactly how much of the Qur'an was reduced to writing during the lifetime of Muhammad. There is certainly no evidence to suggest that anyone had actually compiled the whole text of the Qur'an into a single manuscript, whether directly under Muhammad's express authority or otherwise, and from the information we have about the collection of the Qur'an after his death (which we shall shortly consider), we must rather conclude that the Qur'an had never been codified or reduced to writing in a single text.

Muhammad died suddenly in 632 AD after a short illness and, with his death, the Qur'an automatically became complete. There could be no further revelations once its chosen recipient had departed. While he lived, however, there was always the possibility that new passages could be added and it hardly seemed appropriate, therefore, to contemplate codifying the text into one harmonious whole. Thus it is not surprising to find that the book was widely scattered in the memories of men and on various different materials in writing at the time of Muhammad's decease.

Furthermore we shall see that the Qur'an itself makes allowance for the abrogation of its texts by Allah and, during Muhammad's lifetime, the possibility of further abrogations (in addition to a number of verses which had already been withdrawn) would likewise preclude the contemplation of a single text.

Still further, there appear to have been only a few disputes among the sahaba (Muhammad's "companions", i.e., his immediate followers) about the text of the Qur'an while Muhammad lived, unlike those which arose soon after his demise. All these factors explain the absence of an official codified text at the time of his death. The possible abrogation of existing passages, and the probable addition of further ayat (the Qur'an nowhere declares its own completeness or that no further revelations could be expected) prevented any attempt to achieve the result desired very soon thereafter by his closest companions. It also appears that new Qur'anic passages were coming with increasing frequency to Muhammad just before that fateful day, making the collection of the Qur'an into a single text at any time all the more improbable.

At the end of the first phase of the Qur'an, therefore, we find that its contents were widely distributed in the memories of men and were written down piecemeal on various materials, but that no single text had been prescribed or codified for the Muslim community. As-Suyuti states that the Qur'an, as sent down from Allah in separate stages, had been completely written down and carefully preserved, but that it had not been assembled into one single location during the lifetime of Muhammad (as-Suyuti, Al-Itqan fii Ulum al-Qur'an, p.96). All of it was said to have been available in principle - Muhammad's companions had absorbed it to one extent or another in their memories and it had been written down on separate materials - while the final order of the various verses and chapters is also presumed to have been defined by Muhammad while he was still alive.


If Muhammad had in fact bequeathed a complete, codified text of the Qur'an as is claimed by some Muslim writers (e.g. Abdul Kader - cf. Chapter 6), there would have been no need for a collection or recension of the text after his death. Yet, once the primary recipient of the Qur'an had passed away, it was only logical that a collection should be made of the whole Qur'an into a single text.

The widely accepted traditional account of the initial compilation of the Qur'an ascribes the work to Zaid ibn Thabit, one of the four companions of Muhammad said to have known the text in its entirety. As we shall see, there is abundant evidence that other companions also began to transcribe their own codices of the Qur'an independently of Zaid shortly after Muhammad's death, but the most significant undertaking was that of Zaid as it was done under the authority of Abu Bakr, the first Caliph of Islam, and it is to this compilation that the Hadith literature gives the most attention. It also became the standard text of the Qur'an during the caliphate of Uthman.

Upon Muhammad's death a number of tribes in the outer parts of the Arabian peninsula reneged from the faith they had recently adopted, whereupon Abu Bakr sent a large number of the early Muslims to subdue the revolt forcibly. This resulted in the Battle of Yamama and a number of Muhammad's close companions, who had received the Qur'an directly from him, were killed. What followed is described in this well-known hadith:

Zaid eventually expressed approval of the idea in principle after Umar and Abu Bakr had both pressed the need upon him and agreed to set about collecting the text of the Qur'an into one book. One thing is quite clear from the narrative - the collection of the Qur'an is said quite expressly to have been something which Allah's Apostle did not do.

Zaid's hesitation about the task, partly occasioned by Muhammad's own disinterest in codifying the text into a single unit and partly by the enormity of it, shows that it was not going to be an easy undertaking. If he was a perfect hafiz of the Qur'an and knew the whole text off by heart, nothing excepted, and if a number of the other companions were also endowed with such outstanding powers of memorisation, the collection would have been quite simple. He needed only to write it down out of his own memory and have the others check it. Desai and others claim that all the huffaz of the Qur'an among Muhammad's companions all knew the Qur'an in its entirety to perfection, to the last word and letter, and Desai himself goes so far as to suggest that the power of thus retaining the Qur'an in the memory of those who learnt it by heart was no less than supernaturally acquired:

He goes on to describe the memorising of the Qur'an as "this divine agency of Hifz" (p.26). If we are to take this assumption to its logical conclusion, we must conclude that the collection of the Qur'an would have been the easiest of tasks. If Zaid and the other qurra (memorisers) each knew, by divine assistance and purpose, the whole Qur'an to the last letter without any error or omission - this is the Muslim hypothesis - we would hardly have found him responding to the appeal to collect the Qur'an as he did. Instead of immediately turning to his memory alone he made an extensive search for the text from a variety of sources:

We saw earlier that the Qur'an, at the death of Muhammad, was scattered in the memories of men and on various written materials. It was to these that the young companion of Muhammad duly turned when preparing to codify the text into a single book. The two primary materials, amongst the others mentioned, were ar-riqa'a - "the parchments" - and sudur ar-rijal - "the breasts of men" (as-Suyuti, Al-ltqan fii Ulum al-Qur'an, p.137). He looked not only to human memory but also to written materials, consulting as many of the latter as he could find no matter what their origin (i.e., white stones, etc.). It was to many companions that he turned and to all kinds of material upon which fragments of the Qur'an had been written.

His was not the action of a man believing he had been divinely endowed with an infallible memory upon which he could exclusively rely but rather of a careful scribe who was going to collect the Qur'an from all the possible sources where it was known to be, from scraps, fragments and portions. This was the action of a man conscious of the wide dispersal of the text who would assemble as much of it as he could to produce as complete and authentic a text as was humanly possible.

The earliest traditions of Islam make it quite clear that the search was widespread, though one finds later writers claiming that all the written materials Zaid is said to have relied on - the shoulder-blades of animals, parchments, pieces of leather, etc. - were all found stored in Muhammad's own household and that they were bound together to ensure their preservation. Al-Harith al-Muhasabi, in his book Kitab Fahm as-Sunan, said that Muhammad used to order that the Qur'an be transcribed and that, whereas it was indeed in different materials, when Abu Bakr ordered it to be collected into one text, these materials "were found in the house of the messenger of Allah (saw) in which the Qur'an was spread out" (as-Suyuti, Al-ltqan fii Ulum al-Qur'an, p.137). They were thereafter gathered together and bound so that nothing could be lost.

The earliest records of Hadith literature, however, make it quite plain that Zaid conducted a wide search for the parchments and other materials upon which portions of the Qur'an had been inscribed. Desai also argues for a more limited field of research on the part of Zaid to collect the Qur'an, stating that Zaid was the only companion to be with Muhammad on the last occasion when Jibril went over the Qur'an with him (The Quraan Unimpeachable, p.18) and that he only looked for those pieces of leather and other materials already mentioned upon which the Qur'an had been written under "the direct supervision of Rasulullah (saw)" (p.27). He states that although there were other texts of the Qur'an available, these had not been written down under Muhammad's supervision but by his companions relying on their memories. No evidences or documentation of any kind is given by Desai to show his sources for all these claims, in particular to prove that they are based on the earliest records available. In fact we have already. seen that, in respect of Muhammad's last recitation of the Qur'an with Jibril, the fact that it was recited twice by him was a secret divulged only to his daughter Fatima (Sahih al-Bukhari, Vol. 6, p.485). This would hardly have been a secret if Zaid had been present on that occasion.

Likewise the earliest records of the collection of the Qur'an under Abu Bakr make no distinction between portions of the Qur'an written directly under Muhammad's supervision and those that were not, nor do they suggest that Zaid relied on the former alone. As we in due course shall see, this is a relatively modern interpretation of the research done by him to maintain the hypothesis that the Qur'an was perfectly compiled, but one without foundation in the earliest records.

There are traditions that show that, upon receiving a portion of the Qur'an, Muhammad would command his scribes (of whom Zaid was one) to write it down (Sahih al-Bukhari, Vol. 6, p.481), but there is nothing in the very earliest works to support the idea that the whole Qur'an, as written under Muhammad's supervision, was already assembled in his own home.

There are a number of traditions in the Kitab al-Masahif of Ibn Abi Dawud which suggest that Abu Bakr was the first to undertake an actual codification of the text, each of which reads very similarly to the others and follows this form:

Even here, however, we find clear evidence that there were others who preceded him in collecting the Qur'an texts into a single written codex:

This Salim is one of only four men whom Muhammad recommended from whom the Qur'an should be learnt (Sahih al-Bukhari, Vol. 5, p.96) and he was one of the qurra (reciters) killed at the Battle of Yamama. As it was only after this battle that Abu Bakr set out to collect the Qur'an into a single text as well, it goes without saying that Salim's codification of the text must have preceded his through Zaid ibn Thabit.


At this stage we have a clear trend emerging. Official tradition focuses on the collection of the Qur'an by Abu Bakr as the first, foremost and, at times, only compilation of the text made upon Muhammad's death. Later writers have endeavoured to strengthen this view by suggesting that Zaid was the only man qualified for the task, that the whole Qur'an, no matter in what form, was found in Muhammad's apartments, and that it was to written portions inscribed under Muhammad's supervision alone that the redactor turned to compile his codex. Contemporary Muslim opinion goes even further to claim that the Qur'an, as thus compiled, is an exact record with not so much as a dot, letter or word added or lost - of the script as it was delivered to Muhammad.

On the other hand an objective analysis of the initial collection of the Qur'an, based on a rational assessment of the evidences without regard to sentiment or presupposition, can only go so far as to conclude that the text as compiled by Zaid, which later became the model for Uthman's standardised text, was simply the final product of an honest attempt to collect the Qur'an insofar as the redactor was able to do so from a wide variety of materials and sources upon which he was obliged to rely.

It is the very character of these sources that we should at this stage assess and reconsider. Zaid relied on the memories of men and various written materials. No matter how much those early companions sought to memorise the text perfectly, human memory is a fallible source, and, to the extent that a book the length of the Qur'an had been committed to memory, we should expect to find a number of variant readings in the text. As we shall shortly see, this anticipation proves to be well-founded.

The reliance on a host of portions of the Qur'an scattered among a number of companions must also lead to certain logical expectations. There exists a clear possibility that portions of the text may have been lost - the loose distribution of the whole text in many fragments and portions as opposed to a carefully maintained single text is adequate ground to make such an assumption and, as we shall see, the expectation again proves to be well-founded when the evidences are considered and assessed.

A typical example worth quoting at this point is found in the following hadith which plainly states that portions of the Qur'an were irretrievably lost in the Battle of Yamama when many of the companions of Muhammad who had memorised the text had perished:

The negative impact of this passage can hardly be missed: lam ya'alam - "not known", lam yuktab - "not written down", lam yuwjad - "not found", a threefold emphasis on the fact that these portions of the Qur'an which had gone down with the qurra who had died at Yamama had been lost forever and could not be recovered.

The very fact of such a wide distribution of the Qur'an texts, however, appears to negate the possibility that anyone could have added anything to the text after Muhammad's death. Not being collected into a single text but spread among many companions, there exists a strong possibility that some of the text may have been lost, but at the same time there appears to be no such possibility that it could have been interpolated in any way. The retention of so much of the Qur'an in the memories of Muhammad's companions is a sure guarantee that no one could have added to it in any way and gained acceptance for his innovations.

Lastly, in considering the sources, we should not be surprised to find that other codices of the Qur'an text were being compiled in addition to that being executed by Zaid. Once again we look to the evidence that a number of companions had an extensive knowledge of the Qur'an and it is only to be expected that these would soon seek to preserve, in single codices, what was at that time still fresh in their memories and loosely transcribed on a selection of different materials. Once again we shall find our expectations fulfilled and will discover that the evidences strongly support the conclusions one would draw naturally about the compilation of a book such as the Qur'an rather than the hypothesis that the book was divinely preserved, to the last dot and letter, without loss or variation.

The possibility that part of the text may have been lost is strengthened by evidences in the Hadith literature which show that even Muhammad himself occasionally forgot portions of the Qur'an. One of these traditions reads as follows and is taken from one of the earliest works of Hadith:

The translator has a footnote to this tradition, stating that Muhammad had not forgotten these verses of his own accord but had been made to forget them by Allah as a teaching for the Muslims. Whatever the purpose or cause, it is quite clear that Muhammad had occasion to forget passages that had been, as he proclaimed, revealed to him. The suggestion that Muhammad's oversight of such texts was not of his own doing but brought about through Allah's decree is based on the following text of the Qur'an:

The word ayat is the word consistently used in the Qur'an for its own texts and the word nunsihaa comes from the root word nasiya which, wherever it appears in the Qur'an (as it does some forty-five times in its various forms), always carries the meaning "to forget".

Let us conclude this section. Zaid, quite obviously one of the companions of Muhammad who had an outstanding knowledge of the Qur'an, set about collecting its text so as to produce as genuine and authentic a codex as he possibly could. His integrity in this undertaking is not to be questioned and we may accordingly deduce from all the evidences he consulted that the single Qur'an text he finally presented to Abu Bakr was a basically authentic record of the verses and suras as they were preserved in the memories of the reciters and in writing upon various materials.

The evidences, however, do not support the modern hypothesis that the Qur'an, as it is today, is an exact replica of the original, nothing lost or varied. There is no evidence of any interpolation in the text and such a suggestion (occasionally made by Western writers) can be easily discounted, but there are ample evidences to indicate that the Qur'an was incomplete when it was transcribed into a single text (as we have already seen) and that many of its passages and verses were transmitted in different forms. In the course of this book we shall give more detailed consideration to these evidences and their implications.


Before closing our study on the collection of the Qur'an during the caliphate of Abu Bakr it is important to study the brief mention made by Zaid of the two verses which he said he found only with Abu Khuzaimah al-Ansari. The full text of the hadith on this subject reads as follows:

Insofar as the text speaks for itself without further enquiry, we can see quite plainly that, in his search for the Qur'an, Zaid was dependent on one source alone for the last two verses of Surat at-Tauba. At face value this evidence suggests that no one else knew these verses and that, had they not been found with Abu Khuzaimah, they would have been omitted from the Qur'an text. The incident suggests immediately that, far from there being numerous huffaz who knew the whole Qur'an off by heart to the last letter, it was, in fact, so widely spread that some passages were only known to a few of the companions - in this case, only one.

This ex facie interpretation of the narrative naturally undermines the popular sentiment among Muslims of later generations that the Qur'an was preserved intact because its contents were all known perfectly by all the sahaba of Muhammad who had undertaken to memorise it. A more convenient explanation for the hadith had to be found and we find it expressed in the following quotation from Desai's booklet:

Although the hadith as recorded by al-Bukhari makes no mention of this, Desai claims that the statement that Abu Khuzaima alone had the last two verses of Surat at-Tauba (Bara'a) means that he was in fact the only one who had them in writing under Muhammad's direct supervision. He goes on to say:

The maulana gives no evidences whatsoever in support of these statements. Nowhere in the earliest records of the Hadith literature is there any suggestion that hundreds of Muhammad's companions knew these verses and that others had them in writing, and that what Zaid intended to say was that Abu Khuzaima alone had them in writing directly from Muhammad. Desai's omission of any documentation for his statement is, in the circumstances, most significant.

Siddique, in his article in Al-Balaagh (p.2), also claims that when Zaid said "I could not find a verse" he actually meant he could not find it in writing. As said before, there is nothing in the hadith text itself to yield such an interpretation. From what source, then, do these learned authors obtain this view? It is derived from the following extract which is taken from the Fath al-Baari fii Sharh al-Bukhari of Ahmad ibn Ali ibn Muhammad al-Asqalani ibn Hajar, the translation appearing in Burton's The Collection of the Qur'an on pages 127 and 128:

The source from which Desai and Siddique derive their opinions is not from the earliest records of the compilation of the Qur'an but a much later commentary on the Sahih al-Bukhari done by the famous Muslim author al-Asqalani ibn Hajar who was born in 773 A.H. (1372 A.D.) and died in 852 A.H. The earliest source for the interpretation that Zaid was looking for the verses only in authorised written sources thus dates no less than eight centuries after Muhammad's death by which time, as is the case to this day, it had become fashionable to hold the view that the Qur'an had been widely known to perfection by all the companions of Muhammad who had memorised it. It is, therefore, a convenient interpretation read into the text of the hadith to sustain a more recent supposition. There is nothing in the text of the hadith itself, however, to support this interpretation. The extract continues with some very interesting comments:

While Desai boldly states that it was known "beyond the slightest shadow of doubt" that the last two verses of Surat at-Tauba were part of the Qur'an and that they were known by "hundreds of Sahaabah" in their memories and by others who had recorded them in writing, his source only goes so far as to suggest that it is "probable" that when Zaid produced them from Abu Khuzaima, the other companions recalled having heard them. A cautious suggestion that the others may have recalled having heard the verses has been transformed by Desai into a bold declaration that they were known by hundreds of them without the aid of recollection "beyond the slightest shadow of doubt".

Here is clear evidence that modern Muslim writers are out to establish a cherished hypothesis - the unquestionable perfection of the Qur'an text - instead of objectively assessing the factual evidences as they stand. Desai's source is only a comparatively recent work of interpretation and yet, even here, he cannot resist the temptation to expand it into wholesale allegations of fact.

Ibn Hajar goes on, on the same page, to say "al-Da'udi commented that Abu Khuzaima was not the sole witness. Zaid knew the verse. It was thus attested by two men", an indication that it was believed by other Muslim scholars that Zaid's statement was not to be manipulated into a claim that the verses were not found in writing but should rather be given its obvious meaning, namely, that no one else knew these verses at all.

What makes the convenient claims of Ibn Hajar, as repeated by Desai and Siddique, even less acceptable is the fact that there is a record in one of the very earliest works of tradition showing in greater detail what Zaid's statement really meant. The narrative reads:

This narrative implies that the incident took place during Uthman's reign and not at the time of the collection of the Qur'an under Abu Bakr, but it is clearly the same event that is under consideration. (Siddique in fact states that the records showing that Zaid also missed a verse at the time of the recension of the Qur'an under Uthman actually apply to the last two verses of Surat at-Tauba. We shall say more on this when discussing Uthman's recension shortly).

The significant feature of this narrative is that Zaid and the others are said to have missed these verses completely when transcribing the Qur'an. In fact the statement that Zaid only found them with Abu Khuzaima is hare stated to mean that it was only on the latter's initiative that the verses were recorded at all. He found it necessary to draw the compiler's attention to them - it was not Zaid's search for two verses he already knew that occasioned their inclusion. In fact the text goes on to say that Abu Khuzaima was asked where they should be inserted in the Qur'an and he suggested they be added to the last part of the Qur'an to be revealed, namely the close of Surat at-Tauba (Bara'a in the text).

When one considers this tradition with the relevant hadith in the Sahih al-Bukhari, certain facts cannot be avoided. The verses were missed completely, they were only recalled and thereafter included upon Abu Khuzaimah's initiative, and it was left to him to advise where they should be included. It is only by taking the word tilqiyya ("directly") to mean that he was the only companion who had these verses in writing under Muhammad's supervision that Muslim writers have been able to sustain the hypothesis that the verses were known to many of Muhammad's companions. It is surely quite obvious, however, that the word tilqiyya was used by Abu Khuzaima purely in the sense that he had the verses first-hand from Muhammad, thereby justifying their inclusion. What he was really saying was that he had not learnt them from a secondary source but from Muhammad himself and, therefore, they had to be included in the Qur'an. There is no warrant for the interpretation that he alone had them in writing under Muhammad's authority.

This convenient interpretation, in any event, goes right against the contents and implications of the narratives. If the verses had been well-known, Zaid would hardly have overlooked them. It was precisely because they were not known or remembered that Abu Khuzaima was obliged to point out the oversight. One cannot help asking these modern Muslim authors, on the basis of their own interpretation, whether Zaid would have included these verses in his redaction of the Qur'an if they had not been found "in writing under Muhammad's supervision" even though they were supposedly known in the memories of hundreds of the sahaba and were recorded In writing from other sources.

Our study shows that the collection of the Qur'an by Zaid under Abu Bakr was a gathering together of the texts of the Qur'an from widely divergent sources and materials where the Qur'an was scattered, so divergent that at the Battle of Yamama some passages were irretrievably lost and, in another case, only one of Muhammad's companions was aware of the text. "I searched for the Qur'an", Zaid declared, indicating that he did not expect to find all the texts of the book in the memory of any one man or on written materials in any one place.

The Qur'an thus compiled was the product of a widespread search for what was known in the memories of many men and had been inscribed upon various materials. This type of source-material hardly supports the notion and claim that the Qur'an, as eventually collected, was perfect to the last dot and letter. The Muslim hypothesis is the product of wishful sentiment, it is not based on an objective and realistic assessment of the facts contained in the earliest historical records of the initial collection of the Qur'an.

Jam' Al-Qur'an: Table of contents
Answering Islam Home Page