Notwithstanding the efforts of writers like Desai and Siddique to maintain the hypothesis of the Qur'an's perfect compilation it must surely be obvious from all that we have considered that the Qur'an went through a number of stages during which actions were taken to limit the variations in the written text and in its verbal recitation to establish, as far as each intervener could, a single text for the whole Muslim community. A mushaf waahid was the goal of the redactors, it was not their possession by divine preserve. The Hadith records testify consistently to the imperfection of the Qur'an text and what has come down through the ages to a single text can only be regarded as relatively authentic.

Some Muslim scholars are well aware that it is impossible to maintain the popular sentiment against the records in the Sirat, Hadith and Tafsir literature which testify quite unambiguously to the contrary. The shortcomings and inadequacies of the writings of apologists like Desai and Siddique are all too obvious. So these scholars take another line. By rejecting the Hadith records, they maintain that the Qur'an itself testifies to its own compilation and that this testimony is sufficient to prove that the Qur'an text, as it now stands, is absolutely authentic.

This is the theme of the article by Abdus Samad Abdul Kader titled How the Quran was Compiled referred to in the Introduction and it seems appropriate, in summing up our study of this subject, to begin with a review of his argument and the verses he quotes from the Qur'an to support it.

Right from the start Abdul Kader expresses the notion that indirectly underlies all Muslim studies on this subject. It is the assumption that, if the Qur'an was the Word of God revealed to Muhammad, then it must have been preserved to perfection throughout the ages since its deliverance. The fear is that, if it can be proved that the Qur'an has in any way been amended, or that passages have been lost, or that there is some confusion as to exactly what the original readings were, then the Qur'an's divine origin and authenticity in consequence must fall to the ground and be discounted. We have already seen that this is the motivating consideration behind Desai's booklet and Siddique's article and it explains why their approach to the subject is so sensitive, subjective and, at times, highly irrational. Abdul Kader expresses the conviction directly when he says in his article:

In these statements the author gives sufficient proof that the doctrine of the Qur'an's perfect preservation arises not from a scholarly study of the history of the text but from a popular sentiment that is imposed upon it, a presupposition that has to be maintained at all costs. "It was necessary", he says, to preserve the text; such a scripture "should be complete, perfect"; it "has to be protected from being interpolated". This is the language of presupposition, it is the spirit of hypothesis, it indicates that, before the scholar has even come to a study of his subject, he has already decided long in advance what his findings and conclusion will be. No matter what directions the evidences may lead, the matter is predetermined. It is hardly necessary to say that such an approach is subjective in the extreme and will not yield a balanced or accurate perspective.

The Muslim approach to this whole subject is hard to understand for, if a book never was the Word of God in the first place, no amount of proof that it has been absolutely and perfectly preserved will make it the Word of God. Conversely, if a book was indeed the Word of God at the time when it was first inscribed, the later existence of a few suspect passages and variant readings which do not affect the overall content of the text would not negate its original divine authenticity. Nevertheless, having thus briefly considered the emotional Muslim approach to the subject, let us return to it at a purely factual / interpretational level so that we may conclude with a balanced perspective on what the history of the Qur'an text really was and the extent to which the text, as it stands today, can be regarded as authentic.

Abdul Kader quotes the following verse in proof of his contention that the Qur'an testifies to its own completion and attendant perfection:

Even a superficial study of the text will show that the completion spoken of is not the Qur'an as a book but rather the extent of the words of God in truth and justice. Arberry translates this verse "Perfect are the words of thy Lord in truthfulness and justice" and Yusuf Ali gives the same application: "The Word of thy Lord doth find its fulfilment in truth and in justice". The key word here is tammat, meaning "to be fulfilled", and it is clear that the subject of the perfection spoken of is the truth and justice of God's words and not the text of the Qur'an as a book. The word appears yet again in Surah 11.119 where it is said "the Word of thy Lord shall be fulfilled (tammat): 'I will fill Hell with jinns and men all together'". The context makes it quite clear that we are dealing with a fulfilment of God's words and not of the completion of a text.

As the Qur'an was still in the process of compilation at the time when this verse (Surah 6.116) became a part of its text it is hard to see in any event how it can testify to the Qur'an's supposed perfect compilation. The book was very much incomplete at this point and it is well-nigh impossible to see how this text can be manipulated to prove that the Qur'an was eventually perfectly compiled and preserved to the last dot and letter.

Although Abdul Kader concedes that the Qur'an was being delivered piecemeal over a number of years and is aware that there were many loose parchments and other materials upon which it was being inscribed, to draw the conclusion that the Qur'an was, in fact, perfectly preserved in a single text he argues that the following text testifies to a collection of these parchments into a single book:

The text, like the other one quoted, is very general in its description and it requires no small amount of imagination to make it testify to the perfection of the Qur'an text. Yet, when it is studied in its context, it will be seen that the kitab (translated by Abdul Kader "a Book") spoken of is not the Qur'an at all but one of the five signs of the coming Day of Judgment. The whole context reads:

Once again we see that the passage has nothing to do with the actual compilation of the text of the Qur'an at all and it soon becomes very obvious that Abdul Kader is devoid of evidences for the perfection of the Qur'an in the Hadith records and, in consequence, finds himself constrained to force texts of the Qur'an to yield meanings never intended by the author of the book to provide the required proofs. He concludes by claiming that the Qur'an, in the following verse, actually testifies to a "master copy" of its text that was being preserved:

What is the original Arabic word in this text which Abdul Kader translates as "preserved"? It is maknuun which comes from the root word kanna, meaning "to hide". From this word come the following words used in the Qur'an: aknaan, meaning "a refuge" or hiding-place in the mountains (Surah 16.81); akkinah, meaning "veils" or coverings upon men's hearts (Surah 6.25, etc.); and akanna, meaning "to hide" something in the heart (Surah 2.235). Thus the clear underlying meaning of any form of this word is to conceal or to hide, and Arberry translates Surah 56.78 as "a hidden Book". It would appear that what the Qur'an is really saying of itself is that it is "a concealed scripture" without explaining what this means. In any event it once again is very hard to see how this can be distorted into a testimony to the Qur'an's textual perfection and completion at the end of Muhammad's life. We again have a general and rather vague statement taken right out of context to support a cherished hypothesis.

Ultimately it is the gradual compilation of the Qur'an during Muhammad's lifetime that is the strongest argument against any evidence in the Qur'an (were any to exist) regarding its own completion and perfection. Surah 56.78 and Surah 80.13-16, which is also quoted by Abdul Kader and says no more than that the Qur'an texts were being written on suhuf (parchments) by pious scribes, both come from the very early Meccan period. This was at the time when the Qur'an text was only just beginning to take shape and there is no way that such passages can be adduced in support of the Qur'an's ultimate supposed textual completion and perfection. We find it strange that it should be argued that a book which throughout Muhammad's final years was still being supplemented by additional passages and texts can, in the middle of its course, suddenly testify to its own exactness and completeness!

As long as Muhammad lived there was always a possibility that further passages might be added to the text and the Qur'an nowhere draws the curtain upon itself. There is no verse in the Qur'an stating that the text had been completed and that no further passages could be expected. As we saw early in this book, more was being added to the Qur'an just before Muhammad's death than at any other time during his mission. It was the death of Muhammad that fixed the extent of the Qur'an text, it was this event alone that brought the compilation of the book to a sudden conclusion. Throughout Muhammad's life the Qur'an continued to expand and we must therefore conclude that the Qur'an cannot possibly testify to its own completeness or the extent of the preservation of its text.

There is only one place in the Qur'an where the word jama'ah (to compile or collect together) is used in connection with the text of the book itself, namely in Surah 75.17 where Allah is quoted as saying "It is for Us to collect it and to recite it". It is surprising that Abdul Kader overlooked this verse altogether in his article as it is the closest the Qur'an comes to saying anything significant about its own compilation. Nonetheless it makes Allah speak of collecting the Qur'an before it is recited from heaven to Muhammad, so it too cannot be adduced as evidence for the collection of the text after the time of its deliverance.

It is our opinion that none of the texts quoted by Abdul Kader even remotely testify to the supposed textual perfection and completion of the Qur'an as compiled by his companions at the end of his life. As said already, a book that at all times during its composition was still being supplemented by fresh material cannot possibly give evidence as to the completeness of the final product.

Abdul Kader's whole argument centres on the compilation of the Qur'an during the lifetime of Muhammad and understandably so, for the Qur'an could not testify historically in advance to the course of the text after Muhammad's death. Yet it is precisely this restriction to his lifetime that renders the Qur'an an incompetent witness to the state of the text at the time of its completion. That completion only came upon the death of Muhammad and it is to independent historical records of the text thereafter that we must turn for the evidences we require, namely the series of Hadith records we have already considered.


In sharp contrast to the records we have been studying throughout this book on the development and collection of the Qur'an text we find Abdul Kader declaring that a "Master copy of the Qur'an" was kept by Muhammad and that all other texts of the Qur'an in written form were copied from this original text. He says:

He goes on to allege that the copies made from this master copy were transcribed "under the personal supervision of the Prophet". These are all allegations of fact and yet the writer, like Desai, gives no documentation or authority for his claims. The Qur'an itself nowhere states that a perfect copy of its text was being kept in a safe in the masjid an-nabi of Medina near a column named after it, so Abdul Kader must have obtained this information from another source, but he neglects to substantiate his statements with disclosures of his sources and his claims therefore cannot be tested or critically analysed.

We have seen already that materials upon which the Qur'an was being written were being kept in Muhammad's house at Medina (as-Suyuti, Al-Itqan, p.137) but there are express statements in the same compilation of early records of the Qur'an text which make it plain that the Qur'an had not been brought together into a single location during Muhammad's lifetime, whether in his own home or anywhere else (as-Suyuti, Al-Itqan, p.96). Abdul Kader's statements are set right against the evidences furnished in the Hadith records and other historical sources we have mentioned and, as his claims have no factual basis in the Qur'an, it would be most interesting to know where he obtains his information. His silence on these sources would appear to us to be most significant.

All that he has shown is that, if the Hadith records of the compilation of the Qur'an text are not accepted, there is really no other source to consult. The Qur'an furnishes virtually no useful information at all about its own codification and collection into a single text and, in fact, when one considers the nature of the Qur'an itself, one finds that it is a most improbable witness to the completeness or otherwise of its text.

There is no chronological sequence of any kind in the Qur'an. The surahs have generally been arranged from the longest to the shortest so that the earliest passages appear at the end of the book and the later passages at the beginning. There is nothing of historical foundation in the Qur'an in that no event recorded in the book is ever dated and no regard is paid to any kind of historical sequence in the book.

If the Qur'an does not serve as a good history book, then nor does it offer much of geographical value either. Only one place is mentioned by name in the Qur'an - Mecca in Surah 3.96 (where it is named Bakkah) - and nothing else is given any sort of location in the book. No one reading the Qur'an alone could place any event it records at any point in history or give a specific geographical placement to any locality it mentions or otherwise speaks about.

Many of the longer surahs are made up of passages dating from both Muhammad's mission at Mecca and at Medina and within these composite surahs we find the subject of the text varying from legal restriction to prophetic narratives, from ethical teaching to praises to God, etc., coupled with numerous catch-phrases. More often than not the different subjects of the longer surahs have no connection with each other at all.

The Qur'an is, in these respects, a quite disjointed book. As it stands today it is a collection of fragmentary texts and passages compiled into an unharmonious whole without respect to sequence or theme. It is hardly the kind of book that can offer useful testimony to its own textual accuracy or completion. It has no definite beginning or conclusion and there is no way that a study of the Qur'an text alone can assist one to determine whether it has been completely preserved, nor is there anything in the book to prove that nothing has been omitted from its pages or modified in the process of compilation.

It is only in the Hadith records that we find any evidence as to how the Qur'an really was originally compiled. The science of the study of the Hadith literature has often centred on the reliability or otherwise of the Hadith texts and some Muslim scholars have rejected the Hadith records of the Qur'an's compilation as unreliable because it was well-known that, in the early days of Islam, some Hadith material was fabricated and was handed down alongside material that was authentic.

Such inauthentic hadith records were usually related to opposing schools of law or political issues. The rivalry between the Umayyads and the Abbasids resulted in many records being fabricated to favour the one or the other and as the fiqh (jurisprudence) of Islam developed, so traditions were invented to provide authority for different maxims of law. Many of these can be recognised as fabrications merely through a cursory study of their contents, but to determine the reliability of the rest of the Hadith literature various means were applied to each specific tradition. How sound was its isnad (its chain of transmitters)? How many independent records of the same tradition existed - was it an isolated (ahad) record, a generally accepted text (mashur), or was it widely attested (mutawatir)? Then again, after a consideration of these issues, could it be classed as sahih (genuine), hasan (fair) or da'if (weak), or should it be discounted entirely as mardud (to be rejected)?

This science of classification has rarely been applied to the traditions setting out how the Qur'an was compiled. The earliest records of the collection of the Qur'an were generally taken at face value as this subject was not one which spawned any motivation for fabrication, although John Burton argues to the contrary in his book The Collection of the Qur'an, suggesting that many of the verses said to be missing from the Qur'an were invented after Muhammad's death to give support and authority to the legal maxims of those who made them up. He applies the same argument to some of the recorded variant readings of the Qur'an. None of the three writers who wrote articles in reply to my earlier notes on the compilation of the Qur'an text, however, raised such a possibility, nor did they make any attempt to define which traditions could be accepted and which should be rejected.

There is no standard by which those early records can really be distinguished. Any scholar seeking to separate them into those which can be approved and those which cannot will have to rely almost exclusively on his own initiative and his findings will have to be purely subjective and speculative.

One cannot dispense with some of the Hadith records on this subject without eventually doing away with them all as they give an overall impression of how the Qur'an was codified into a single text and, as we shall see in the last section of this chapter, they are far more consistent in giving a general picture of what actually occurred than some scholars are willing to admit. The fact is that, without these records, there is no evidence as to how the Qur'an was compiled. If they are to be rejected, then nothing authoritative whatsoever can be said about the manner in which the Qur'an was compiled into what it is today. The record of the codification of the Qur'an text as found in the early Sirat, Hadith and Tafsir literature is the only historical source in Islam to consult - without it there is only a void and nothing authoritative really can be said. No other thesis about the original collection of the Qur'an can be documented or grounded in historical evidences. Let us press on in closing to a review of the history of the text as we have thusfar set it out.


We are left with a sharp contrast between sentiment and reality in Islam on the subject of the authenticity of the Qur'an text. Popular sentiment opts for the claim that the Qur'an text has been perfectly preserved by divine authority without so much as an alteration in the text of any kind whatsoever. Reality, however, testifies to a far more mundane and predictable history of the text with much evidence as to passages that are now missing from the Qur'an, substantial variant readings that existed in the earliest codices, and other variants in dialect which have survived more than one attempt to establish a universally accepted single text. Yet another typical testimony to the loss of portions of the Qur'an in the early days can be mentioned here.

In his short section on the codex of Abdullah ibn Umar, in speaking of differences in reading between Abdullah and the other companions of Muhammad, Ibn Abi Dawud quotes Abu Bakr ibn Ayyash as saying:

What sort of evidence would have been required to substantiate the Muslim hypothesis of a perfect text? Firstly, there would most certainly have had to be a complete silence in the Hadith records regarding missing passages, variant readings and the like. The historical sources of Islam apart from the Qur'an itself would have had to support the theory of an absolutely perfect text instead of contradicting this theory as consistently as they do. We would have required sound evidence that the Qur'an was carefully inscribed in a single text during Muhammad's lifetime and that this text had survived his death and been carefully looked after as the sole authority from which other copies alone could be made. This is very much what Abdul Kader alleges as the actual history of the text but his claim is directly contrary to the evidences which show that it was only after Muhammad's death that any attempt was made to collect the Qur'an into a single text.

As pointed out already, Abdul Kader furnishes no proofs, evidences or documentation for his theory and it appears that the wish has become father to the thought. He rejects the Hadith not because they are unreliable but because he finds them unacceptable in that they solidly undermine the theory he cherishes so much. Instead, being aware of the sort of evidences that would have been required, he summarily sets forth the ideal as historical fact without offering any source material that can be checked or critically reviewed.

A very different history of the text of the Qur'an would have had to be recorded than the one that the heritage of Islam has preserved for us to support the case for a text absolutely free of alteration, omission or variation. We would have required very strong evidences that only one text of the Qur'an ever came down through those early years of Muslim history and these evidences would have had to show quite convincingly that the whole text, verse for verse, is precisely the same today as it was then. There would also have had to be no evidences to show that other codices, differing from the standard text, had ever existed. Such is the kind of proof we would require to entertain seriously the claim that the Qur'an text had been preserved to perfection without variations of any kind. Our study shows that such proof and the evidences required therefor quite simply do not exist.

The evidences that do exist for the history of the Qur'an text on the contrary ruin the claim for the Qur'an's textual perfection and relegate such a claim to the realms of popular sentiment and wishful thinking. These evidences, in their broad outline, give us a very reasonable picture of the development of the text and, in fact, allowing for the unusual nature of the Qur'an as a book, yield very much the kind of history that we would have been inclined to expect. Instead of a case for divine preservation we find a very mundane and predictable course.

The Qur'an was compiled piecemeal, was not compiled in a single book during Muhammad's lifetime, was recited by many companions and was read at the time by Muslims with varying Arabic dialects. The course of the text thereafter down to the present day is largely what one would have expected and is generally consistent with itself, most certainly in its broad outline.

After Muhammad's death passages of the Qur'an were lost irretrievably when a number of reciters died at the Battle of Yamama. This incident together with the Qur'an's automatic completion as a book once its mediator had passed away inspired a number of companions to compile their own codices of the text. These were basically consistent with each other in their general content but a large number of variant readings, many seriously affecting the text, existed in all the manuscripts and no two codices were entirely the same. In addition the text was being recited in varying dialects in the different provinces of the Muslim world.

During the reign of Uthman a deliberate attempt was made to standardise the Qur'an and impose a single text upon the whole community. The codex of Zaid was chosen for this purpose because it was close at hand and, having been kept in virtual seclusion for many years, had not attracted publicity as one of the varying texts as those of Abdullah ibn Mas'ud and Ubayy ibn Ka'b had done. The other codices were summarily destroyed and Zaid's text became the textus receptus for the whole Islamic world as a result.

Numerous records were retained, however, showing that key passages were missing from this text. It also had to be reviewed and amended to meet the Caliph's standard for a single approved text. After Uthman's death, however, al-Hajjaj, the governor at Kufa, made eleven distinct amendments and corrections to the text.

As the early codices were only written in consonantal form, however, the varying dialects survived largely unaffected by Uthman's action and it was only three centuries later that a scholar, Ibn Mujahid, managed to limit these to seven distinctly defined readings in accordance with a tradition which stated that the Qur'an originally came in seven different readings although the tradition itself made no attempt to define these readings.

Over the succeeding centuries the Qur'an continued to be read in seven different forms until five of them largely fell into disuse. Eventually only those of Hafs and Marsh survived and, with the introduction of a printed Qur'an, the text of Hafs began to take almost universal prominence.

The Qur'an text as it is read and printed throughout the Muslim world today is only Zaid's version of it, duly corrected where necessary, later amended by al-Hajjaj, and read according to one of seven approved different readings. This is the reality - a far cry from the popular sentiment which argues for a single text right from the time of Muhammad himself. The reality, however, based on all the evidences available, shows that the single text as it stands today was only arrived at through an extended process of amendments, recensions, eliminations and an imposed standardisation of a preferred text at the initiative of a subsequent caliph and not by prophetic direction or divine decree.

The Qur'an is an authentic text to the extent that it largely retains the material initially delivered by Muhammad. No evidence of any addition to the text exists and, in respect of the vast number of variant readings and missing passages that have been recorded, there does not appear to be anything actually affecting or contradicting the basic content of the book. In this respect one can freely assume a relative authenticity of the text in the sense that it adequately retains the gist and content of what was originally there. On the contrary there is no basis in history, facts or the evidences for the development of the text to support the cherished hypothesis that the Qur'an has been preserved absolutely intact to the last dot and letter.

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