CHAPTER 5:

SAB'AT-I-AHRUF: THE SEVEN DIFFERENT READINGS

1. THE SAB'AT-I-AHRUF IN THE HADITH LITERATURE.

While writers like Siddique seek to gloss over the wealth of evidence in the early historical records of Islam showing how the Qur'an was eventually standardised against a background of variant readings, missing passages and texts which had been lost altogether, others like Desai duly acknowledge the evidences and admit the many differences that existed in the earliest manuscripts and codices. On the other hand we find Desai, for example, nonetheless determined to maintain the popular hypothesis that the Qur'an has been perfectly preserved and is intact to the last dot and letter. We have already seen how he overcomes the difficulty with the passages said to be missing from the Qur'an - he conveniently declares them all to have been abrogated by Allah during Muhammad's lifetime. How does he evade the implications of the numerous variant readings in the earliest texts and codices? He claims that they resulted not from uncertainty about the text or partial confusion about the actual wording of each passage but rather that each and every variant was in fact part of the original Qur'an text as delivered by Allah to Muhammad! He says that "the 'differences' in the recitals of various people were all official, authorized and divine forms which were taught by Rasulullah (saw) to the Sahaabah who in turn imparted their knowledge of Qira'at to their students" (The Quraan Unimpeachable, p. 13) and goes on to quote the following statement of Muhammad in support of his interpretation:

The statement concludes a tradition which informs us that Umar one day heard Hisham ibn Hakim reciting Suratul-Furqan in a way very different to that which he, Umar, had learned it. Umar struggled to control himself and intended to spring upon him but, when Hisham had finished, Umar confronted him and accused him of being a liar when he stated that he had learned it so directly from Muhammad himself. When they came before the Prophet of Islam he confirmed the readings of both companions, adding the above statement that the Qur'an had been revealed alaa sab'ati ahruf - "in seven readings". A similar tradition stating that the Qur'an originally came in seven different forms reads as follows:

We are further informed that Ubayy ibn Ka'b recalled an occasion where Muhammad reported that Jibril had come to him one day and told him Allah had commanded that the Qur'an be recited in only one dialect, to which Muhammad replied that his people were not capable of doing this. After much going back and forth the angel finally decreed that Allah had allowed the Muslims to recite the Qur'an in seven different ways and that each recital would be correct (Sahih Muslim, Vol. 2, p.391).

Further than these records there is no evidence in the Hadith literature as to what these seven different readings were. The narrative in the Sahih of Al-Bukhari, also recorded in Vol. 6, p.481, does not tell us how Hisham's recital of Suratul-Furqan differed from Umar's, nor whether the differences were purely dialectal as is suggested in the traditions from the Sahih of Imam Muslim.

There are no other records in the earliest works of Hadith and Sirat literature to give any indication as to what the seven different readings actually were or what form they took. Were there ultimately seven different forms in which the whole Qur'an could be recited? Or was it purely a question of different dialects in which the text could be recited? There is nothing in the earliest records giving any idea of what the sab'at-i-ahruf were or what form they took other than the clear indications in the traditions quoted from the Sahih of Muslim that they were confined to dialectal variants. No more is said than that the Qur'an had actually been revealed in seven different ways in which it could be recited.

In the As-Sunanul-Kubra of Abu Dawud we find the compiler recording up to forty variant readings of the Qur'an under the heading Kitab al-Huruf wa-al Qira'at ("The Book of Dialects and Readings"). We shall mention some of them later in this chapter, but here let it suffice to say that in each one of the readings he quotes, only one variant is mentioned and in each case it is purely a variation of dialect or pronunciation that is involved. There is no suggestion that these variant readings were authorised as part of the original text or that they formed part of the seven different readings but, if they did, they were confined to dialectal variants alone.

As a result of the paucity of evidence as to exactly what the sab'at-i-ahruf originally were a host of different explanations of the relevant hadith have been suggested. Some say that as the Arab tribes had divergent dialects the Qur'an came in seven different forms for their convenience while others say that the seven different readings were distinct forms conveyed to the centres of Islam by approved readers in the second century after Islam. Thus Abu 'Amr is said to have taken one of the readings to Basrah, Ibn Amir took one to Damascus, Asim and two others took theirs to Kufa, Ibn Kathir took one to Mecca and Nafi retained one in Medina (Sunan Abu Dawud, note 3365, Vol. 3, p,1113). What they were in each case is anyone's guess. There are numerous other explanations which we need not consider here. From what we have already considered it is quite clear that nothing certain can be said about the seven different readings except that they were confined to differences in dialect and pronunciation alone.

Desai constantly talks about "all the authorized 'variant readings'" which were "revealed and part of the Qur'an" and, as said already, he simply catalogues all the different readings of the Qur'an that can be found in the earliest records as part of the sab'at-i-ahruf and as therefore divinely sanctioned. The key difficulty here, however, which Desai conveniently overlooks, is that those records show that the differences between Zaid ibn Thabit's codex and those of Abdullah ibn Mas'ud, Ubayy ibn Ka'b and others relate not only to dialectal variants but also to real variations of the text itself. We have quoted numerous examples in this book of words, clauses and even whole verses that were said to have differed radically between the different codices.

It has been amply proved already that these differences were not purely dialectal but at times related to the basic content of the Qur'an text itself. It must be said again that if all these differences had been purely in the pronunciation of the text according to the various dialects of the Arab tribes, they would not have appeared in the written text, especially when we remember that those early codices had only consonants and did not include the relevant vowel points upon which the different dialects invariably turned.

Uthman would never have ordered the wholesale destruction of all the codices other than Zaid's if the differences of reading were only in the verbal expression of the text. There are, as we have seen, many different explanations of the sab'at-i-ahruf, yet it is invariably claimed that these related solely (or almost exclusively) to dialectal variants. If we accept this interpretation we must at the same time conclude that these seven different readings have nothing or very little to do with the extensive textual variants which existed between the codices of Ibn Mas'ud, Zaid, Ubayy, Abu Musa and others before Uthman ordered the destruction of all but one of them. While Desai endeavours to give divine sanction and authority to all the variant readings that existed at that time, whether textual or dialectal, by claiming that they were all part of the sab'at-i-ahruf, the unanimous opinion of the early Muslim scholars was that these seven readings consisted solely of dialectal differences and the learned maulana has no justification for seeking to apply them to those instances where there were real distinctions in the actual text of the Qur'an in the various codices.

We are clearly dealing with two different types of "variant" reading. On the one hand we have the substantial differences between the early codices which covered the addition of whole clauses such as wa salaatil'asr in Surah 2.238, the inclusion of expressions such as yawmal-qiyamati in Surah 2.275 in Ibn Mas'ud's codex, the extra clause wa huwa abuu lahum in Surah 33.6 in the codices of Ibn Mas'ud, Ubayy ibn Ka'b, Ibn Abbas and others as well as the numerous other actual textual variations we have mentioned.

On the other hand we have finer points of distinction in pronunciation and dialect which were not nearly as distinct in the written text as the other variants. It is only to these variants that the sab'at-i-ahruf can be applied if, as is generally held, the seven different readings related only to dialectal variants.

We know that Uthman was concerned about both serious textual differences and dialectal variants. To eliminate the former he simply chose Zaid's text in preference to the others which he ordered to be destroyed. To remove the latter we know that he was not satisfied that Zaid's text itself adequately represented the Quraysh dialect and he therefore ordered Sa'id ibn al-As and two others from the Quraysh to amend Zaid's text where necessary. The following impression of Uthman's action is very informative:

Uthman was thus concerned not only to standardise the Qur'an into a single text but also to establish the Quraysh dialect as the standard medium of expression at the same time. He achieved the first objective by burning the other codices, the second by employing three of the Quraysh to revise the dialect of Zaid's codex insofar as it affected the written text (which effect could only have been negligible as most of the dialectal variants would have been reflected solely in the use of vowel points which were not at that stage included in the transcribed text).

The sab'at-i-ahruf were regarded as affecting only the second concern, that is, dialectal variants. The ahruf (readings) referred to were, therefore, only those affecting the different lugaat (dialects) of the Arab tribes. There is no suggestion anywhere in those early records that the traditions which stated that the Qur'an had been revealed in seven different readings had anything to do with the large number of substantial variant readings in the actual text that were found in the codices of Zaid-ibn-Thabit, Abdullah ibn Mas'ud and the others written out before Uthman's action to standardise the text. Thus the sab'at-i-ahruf had nothing to do with Uthman's first concern, namely the authorisation of a single written text at the expense of the others, and indeed there would have been no need to burn them if the differences had been purely dialectal as the seven different readings were said to be.

Thus Desai is wide of the mark when he tries to explain away all the textual differences that were found in the early codices as being part of the divinely authorised seven readings. These related solely to different dialects and the maulana errs when he tries to make them cover the real textual distinctions we have mentioned in this book and in the booklet which he set out to refute. It may suit his cause considerably to claim that all those variant readings in the different codices were divinely authorised as part of the sab'at-i-ahruf, but, to reach this conclusion, he has had to blur the distinctions between the two types of variant reading we have considered - textual and dialectal - with the seven different readings applying only to the latter.

It is clear that the hypothesis that the Qur'an has been perfectly preserved to the last dot and letter cannot be sustained in the light of the many textual differences that existed in the early codices. Desai could find no way of getting around this difficulty other than to take hold of just one hadith record - the statement of Muhammad about the sab'at-i-ahruf - and apply it to those differences against the clear indications that these readings were confined to dialectal variants alone.

2. THE PERIOD OF IKHTIYAR: THE "CHOICE" OF READINGS.

We have shown that there were two different types of variant reading at the time of Uthman's recension, both of which the Caliph sought to eliminate as part of the accepted text of the Qur'an. It is intriguing to discover that he succeeded in almost totally eliminating the first type - the substantial differences in the text of the Qur'an itself that were found in the various codices - but did not succeed in eliminating the second type, namely the variations in dialect and pronunciation that were widespread among the early Muslims and which continued to be read as part of the Qur'an text. This was chiefly because the codices which Uthman sent out to the various provinces had no diacritical points or vowel marks but represented only the consonantal text of the Qur'an. Unlike our alphabet which has vowels and consonants, the Arabic alphabet only has consonants and in the early days the alphabet was limited to only seventeen letters so that one consonant could reflect one of two or more letters. It was only in the later generations that vowel marks above and below the letters were introduced to give an exact representation of the vocal text and diacritical points were then also added above and below the relevant consonants to achieve the same result.

It was because the dialectal variants were reflected primarily in the vowelling of the Qur'an text that Uthman's official manuscripts, written in consonantal form alone, were unable to bring about a uniform reading of the text in the single Quraysh dialect. Thus we find that in spite of his recension variant readings of the text continued to remain widespread among the Muslims but were generally confined to differences in dialect alone. Throughout the first three centuries of Islam there was a period of ikhtiyar, a time of "choice" when Muslims were considered free to recite the Qur'an in whichever dialect they chose on the strength of the hadith text which stated that Muhammad had taught the Qur'an had been revealed in seven different ways in which it could be recited.

During this period until the year 322 A.H. (934 A.D.), all the scholars of the Qur'an taught that these dialectal variations constituted the sab'at-i-ahruf of which Muhammad spoke. Thus the "seven readings" became confined to variations in dialect and pronunciation alone and were not considered to be applicable to the very real differences that occurred in the earliest days of the development of the Qur'an text, many of which we have mentioned in this book and which Uthman sought to eliminate in the interests of establishing a single text.

We do have sound evidences, however, to show that, even after Uthman's recension was complete, his text was still considered to be imperfect over and above the fact that it was largely a reproduction of Zaid ibn Thabit's original compilation. During the caliphate of Abd al-Malik in the first century of Islam the governor of Iraq, al-Hajjaj ibn Yusuf, took steps to correct Uthman's text. He is said to have made eleven direct changes to the Qur'an text as it stood in its consonantal form, all of which are reflected in the Qur'an as it stands today.

Under the heading Baab: Ma Ghaira al-Hajjaaj fii Mushaf Uthman ("Chapter: What was Altered by al-Hajjaj in the Uthmanic Text") Ibn Abi Dawud lists these specific amendments and his narrative setting them out begins as follows:

The whole section continues to name each one of the amendments made by al-Hajjaj so that the Qur'an text as we have it today is not only the Uthmanic text but also a subsequent minor recension of it by the Iraqi governor. It is interesting to find that one of the alterations mentioned by Ibn Abi Dawud was originally the reading of Ubayy ibn Ka'b as well. Surah 12.45 is said to have originally read anaa aatiikum but was amended to read anaa unabbi'ukum and we are informed that the former reading, as originally read in the Uthmanic text, was also the reading of Ubayy ibn Ka'b and al-Hasan (Jeffery, Materials, p.138). It is probable that Zaid and Ubayy agreed on the original reading but that it was widely acknowledged by the other companions after Uthman's recension that this was a variant reading and that the correct reading was that which al-Hajjaj eventually put in its place.

In addition to these eleven changes to the Qur'an text there are evidences that a few further variant readings in the actual consonantal outline of the Qur'an still remained. All but two of these related to a single letter alone but in Surah 9.100 we find that the word min ("from") was read between the words tajrii tahtihaa, and in Surah 56.24 the pronoun huwa was known to be added as an extra word. Desai, in recording some of the variant readings of the Qur'an in his booklet (p.15), acknowledges the first variant mentioned here and also points out that other variants took the form of different word placements, diacritical points, attenuations and tenses. All these, however, relate to variants still known to have been freely recognised after the recension by Uthman. Throughout his booklet, however, there is no mention of any of the substantial variants that existed in the actual text of the Qur'an which led to the other codices being destroyed.

In this book and in my booklet Evidences for the Collection of the Qur'an which Desai set out to refute I have given a wealth of examples of such variant readings which went far beyond the question of dialects and pronunciation. The issue here was not one of different forms of qira'at (reading) but of the actual content of the text itself. Expressions were found in some codices that were omitted in others (such as yawmal-qiyaamati in Surah 2.275), single words were likewise confined to some codices and were not found in all of them (such as mutataabi'aatin in Surah 5.91) while whole clauses only appeared in some of the texts (such as wa huwa abuu laahum in Surah 33.6).

It is hard to tell at times which variant readings Desai is in fact admitting in his booklet. He makes no specific mention of these substantial differences and all the variants he does refer to can be categorised in the sab'at-i-ahruf, the dialectal variants which survived Uthman's recension.

In my previous booklet, however, I recorded a number of the major textual variants that existed in the other codices before they were destroyed and Desai took no issue with any of them. His admission of the existence of the variant readings has to be taken against the background of his express purpose to respond solely to my booklet and it must therefore be presumed that he was acknowledging the authenticity of the early textual variants. In his response, however, he deals only with the second class of variants, the sab'at-i-ahruf, and conveniently glosses over the others. He then uses this second class alone to support his contention that all the variant readings of the Qur'an were divinely authorised and it appears that he was fully aware that he could not expressly acknowledge the authenticity of the substantial textual variants without at the same time conceding that the Qur'an had not been perfectly preserved to the last dot and letter. It became convenient, therefore, to blur the distinction between the two and make an overall admission about the variant readings of the Qur'an while citing only the dialectal differences in support of his defence that the Qur'an had been revealed in seven divinely authorised forms. One cannot help feeling that the learned maulana is guilty of a degree of casuistry in his argument.

In closing let us consider some of the variants recorded by Abu Dawud in his Kitab al-Huruf wa al-Qira'at, all of which relate to dialectal distinctions alone and do not affect the consonantal record of the written text. They thus all form part of the second type of variant reading and can be regarded as part of the sab'at-i-ahruf of which Muhammad spoke. We shall mention just three of these readings that the compiler records to illustrate the point:

In each case the variant is found solely in the vowelling of the text and would not have been reflected in the consonantal text transcribed by Uthman as the standard form of the Qur'an for the whole Muslim community. This explains why so many of these dialectal variants survived Uthman's recension while the substantial textual variants were duly eliminated from the actual recitation of the Qur'an text. Let us press on to the time when the period of ikhtiyar, the time of "free choice", closed and the sab'at-i-ahruf, the seven readings of the Qur'an, were defined more exactly. Thereafter we shall close with a brief analysis of the actual character of these readings.

3. IBN MUJAHID'S FINAL DEFINITION OF THE SEVEN AHRUF.

It was not until the fourth century of Islam that an attempt was made to actually define the seven different readings. As said earlier there is nothing in the earliest works of Sirat and Hadith literature giving any indication as to what these readings actually were except for a statement attributed to Muhammad that they were all a part of the Qur'an as revealed by Allah. By the fourth century after Muhammad's death, therefore, the decision as to what these seven readings were was at the discretion of whoever sought to determine and define them.

In 322 A.H, the well-known authority on the Qur'an at Baghdad, Ibn Mujahid, took it upon himself to resolve this issue. He had considerable influence with Ibn Isa and Ibn Muqlah, two of the wazirs in the Abbasid government of the day (the equivalent of a cabinet minister in a contemporary regime), and through them he managed to establish an official limitation on the permissible readings of the Qur'an. He wrote a book titled Al-Qira'at as-Sab'ah ("The Seven Readings") based on the hadith which stated that there were seven divinely authorised ahruf of the Qur'an and he established seven of the current readings as canonical and declared the others in use to be shadhdh ("isolated", that is, non-canonical).

The seven readings established have already been mentioned in this book, namely those of Nafi (Medina), Ibn Kathir (Mecca), Ibn Amir (Damascus), Abu Amr (Basra), Asim, Hamzah and al-Kisai (Kufa). In each case there were certain recognised transmitters who had executed a recension (riwayah) of their own of each reading and two of these, namely those of Warsh (who revised the reading of Nafi) and Hafs (who revised that of Asim), eventually gained the ascendancy as the others generally fell into disuse and were no longer read in the major parts of the Muslim world.

Ibn Mujahid's determination to canonise only seven of the readings then in circulation at the expense of the others was upheld by the Abbasid judiciary of his day. Very soon after his action a scholar named Ibn Miqsam was publicly forced to renounce the widely-held opinion that any reading of the basic consonantal outline that was in accordance with Arabic grammar and made common sense was acceptable. This decision virtually validated the seven sets of readings chosen by Ibn Mujahid as the only officially acceptable qira'at. Not long after this another scholar, Ibn Shannabudh, was forced in a similar way to retract the view that it was permissible to use the readings of Ibn Mas'ud and Ubayy ibn Ka'b (meaning only those variants confined to dialectal differences which were attributed to them and not the substantial variants which Uthman had eliminated from the recitation of the Qur'an).

Over the centuries most of the seven canonical readings also fell into disuse until only those of Nafi and Hafs became widely used in practice. Warsh's riwayah of Nafi's reading has long been used in the Maghrib (the western part of Africa under Islam's rule, namely Morocco, Algeria, etc.), mainly because it was closely associated with the Maliki school of law, but it is the riwayah of Hafs that has gradually gained almost universal currency in the Muslim world, especially since the printing of the Qur'an came into vogue. Virtually all the lithographed editions of the Qur'an that have been printed in the last two centuries have followed the reading of Asim through Hafs. The fully vocalised printed editions of the Qur'an that are in the possession of millions of Muslims in the world today reflect the reading of Hafs and in time this version is likely to become the sole reading in use in the whole world of Islam.

The period of ikhtiyar closed with Ibn Mujahid. He did to the vocalised reading of the Qur'an what Uthman had done to the consonantal text. Just as the latter had standardised a single text for the whole Muslim community by destroying the other codices that existed, so Ibn Mujahid established seven fixed canonical readings by outlawing all the others that were in current use. Just as the text standardised by Uthman cannot be regarded as a perfect reproduction of the Qur'an exactly as it was delivered by Muhammad because it did no more than establish the codex of just one man, Zaid ibn Thabit, at the Caliph's personal discretion, so the seven readings canonised by Ibn Mujahid cannot be accepted as an exact reflection of the sab'at-i-ahruf spoken of by Muhammad, once again precisely because they were simply the readings of later reciters arbitrarily chosen by the redactor at his own personal discretion.

4. REFLECTIONS ON THE UNIFICATION OF THE QUR'AN TEXT.

Thusfar we have dealt with the seven different readings as they were treated during the first centuries of Islam. The time has come, however, to consider this subject from a more critical perspective. Can we summarily accept that all the variant readings of the Qur'an, even if we consider only the dialectal variants and not the substantial textual differences, can be regarded as divinely authorised simply on the basis of the statement attributed to Muhammad that the Qur'an came originally with seven different readings? We know what those readings eventually became: three centuries after Muhammad's death Ibn Mujahid at his own discretion simply chose seven of the many different readings that prevailed at his time and declared them to be the divinely authorised readings. No objective scholar of the Qur'an text can accept such a unilateral and arbitrary approach as even remotely authoritative, however, and Ibn Mujahid's action can only be regarded as an ambitious attempt to make the different readings of the Qur'an in his day fit the concept of seven original readings. The action by this fourth-century redactor is something of a red herring across the path of the real issues in respect of this subject.

The key question is: what actually were those seven different readings at the time of Muhammad? What were they originally supposed to have been? We have virtually given the answer already: no one can possibly say. Nothing more is indicated in the earliest hadith records mentioning these readings than that they were generally confined to variations in dialect and rarely affected the actual consonantal text.

We have on the one hand a tradition about seven different readings, on the other a vast number of examples of actual variant readings which cannot be made relevant to the tradition in any definite way. Desai claims that Uthman eliminated six of the readings and retained just one in the interests of standardising a single text of the Qur'an. On whose authority he reduced the Qur'an to just one of seven different forms in which it was said to have been revealed Desai does not say, but to circumvent the obvious conclusion that six of the divine forms of the Qur'an have thereby been lost and eliminated he claims that the variant readings were nonetheless at the same time separately preserved. He says in his booklet:

As usual no documentation in support of this allegation is given and the maulana's readers are, it appears, once again obliged to simply accept what he says without further enquiry. He tells us nothing of these so-called separate compilations nor does he give the source for his claim that Uthman ordered that they be put together. Such an action on the part of the Caliph can only be considered grossly improbable in the light of the fact that it was his express purpose to entirely eliminate the variant readings that existed in the interests of maintaining a single text.

The whole argument of the maulana, however, can be shown to be extremely fragile from another consideration. If, as he claims, the other six readings were so carefully retained, what were they? Can Desai transcribe for us today seven different Qur'an texts fully vocalised, showing all the variant readings that existed at the time of Uthman's recension which were said to have been divinely authorised and duly set them out in seven different forms? Even if he could, we would yet have to ask on what authority he would expect us to accept that his proposed seven different forms of the Qur'an as thus defined were in fact precisely what Muhammad was speaking about.

A study of the earliest readings, both dialectal and substantial, will soon show that such an undertaking is an impossible task. These readings are sometimes said to have come from one companion, sometimes from another, at times from a number together. No indication of the actual division of all these variants into seven distinct forms is even hinted at in the earliest records. It is quite impossible to authoritatively define what those seven different readings were supposed to have been.

Thus the hadith records about the sab'at-i-ahruf are really quite meaningless. They cannot, without a considerable degree of speculation and pure guesswork, be applied to the variant readings of the Qur'an that have been preserved through the centuries. The figure "seven" has, thus, no relevance at all to what we are considering. All that has happened is that we have, alongside the single text of the Qur'an in consonantal form that was standardised by Uthman, a vast number of passages that are said to have been lost, a host of variant readings of specific texts, together with finer distinctions in the vowelling of the text. These evidences strongly contradict the popular sentiment that the Qur'an has been perfectly preserved to the last dot and letter, nothing lost, varied or amended.

The vague statement about seven different revealed forms of the Qur'an has become a convenient blanket to cover all the readings that are known to have existed so as to give them divine authorisation. This is the whole theme of Desai's booklet - every variant that can be produced is summarily declared to have been divinely revealed as one of the seven readings even though the maulana could not possibly hope to define exactly what the seven readings were supposed to have been, to which one of the seven each respective reading belongs, least of all produce any evidences to substantiate such a definition and say on what authority he draws his conclusions. The tradition about the sab'at-i-ahruf has become an expedient licence to claim divine authority for any variant that can be produced - thus the maulana maintains the popular sentiment, the hypothesis that nothing of the Qur'an has been lost or varied by anything other than divine decree.

A very good example of the confusion caused in subsequent generations about the supposed seven different readings and the total inability of the early Muslim scholars to categorise the variant readings that were all at hand into seven distinct forms is clear from the following quote:

This statement shows how impossible it was to define the seven different readings in terms of the recital of the Qur'an as it was actually being read in its different forms in the Muslim community and how the two could not plausibly be related to each other in any way whatsoever. Any good reading was automatically considered to be one of the seven authorised readings, not because it could be proved to belong to one of them, but because it became acceptable through other considerations - its isnad, its consistency with the single Uthmanic consonantal text, and its compliance with proper Arabic grammar.

Other Muslim writers like Siddique have an easier way of getting around the problem. They simply declare that such variants never affected the written text of the Qur'an at all, notwithstanding the clear evidences to the contrary in the exhaustive summaries of the evidences for the compilation of the Qur'an in the Itqan of as-Suyuti and the Kitab al-Masahif of Ibn Abi Dawud, both of which Siddique alludes to briefly with complete approval in his article.

There is a further thrust in Desai's argument that proves defective upon closer analysis. His reasoning that Uthman's "measure of eliminating all other authorized and true versions of the Qur'aan Majeed" (p.32) meant that only one form of qira'at was standardised to ensure uniformity at the expense of the other six goes against the whole character of what Uthman actually did. The maulana seems to overlook the fact that Uthman only standardised the consonantal text of the Qur'an and, in sending out manuscripts which did not have diacritical points or vowel marks, he hardly affected the dialectal variants of the text that were said to have made up the sab'at-i-ahruf (cf. the traditions quoted earlier on the seven readings in the Sahih of Muslim). Thus there came the period of ikhtiyar when the Qur'an was freely recited in numerous different dialects until Ibn Mujahid arbitrarily chose seven of them at his own discretion to represent the readings of which Muhammad had spoken.

Uthman never had it in mind to eliminate six divinely authorised readings in the interests of standardising one of them for the purposes of uniformity as the maulana claims. He believed all along that there never was nor should have been more than one single text of the Qur'an and he viewed the evidences that the Qur'an was beginning to be divided up into all sorts of different readings with alarm, fearing that if this continued the original text might be lost altogether. He thus took the drastic step of ordering the destruction of all but one of the codices to outlaw variant readings of the Qur'an precisely because he considered such a practice to be an unauthorised deviation from the original text.

Desai constantly claims that Uthman's purpose was to establish one of the seven different forms of qira'at at the expense of the others but, as said already, he is missing the point. Uthman's action had very little to do with qira'at, in fact it centred primarily on masahif which were restricted to representations of the consonantal text of the Qur'an alone. The vast number of distinctions in qira'at that would have been reflected solely in vowel points thus escaped his action completely. Uthman only standardised the consonantal text of the Qur'an - its basic form - and the sab'at-i-ahruf were always regarded by the early scholars of Islam to have thus survived his action and for three centuries the Qur'an was officially recited in all sorts of different dialects. In fact all that Ibn Mujahid did thereafter was to standardise seven of these as officially acceptable and they too continued to survive as part of the authorised qira'at. Thus what was eliminated by Uthman was only the class of variant readings that affected the actual written text of the Qur'an and not its many forms of qira'at that would have been reflected solely in different vowel points.

The sab'at-i-ahruf, in conclusion, cannot be considered in any way relevant to the wealth of variant readings that have come down alongside the Qur'an in the heritage of Islam. There is nothing in the records of these variants or the different forms of dialect that actually existed that can be related to seven specific forms of reading as stated in the relevant tradition. Writers like Desai merely seek to force an identification between the two so as to give divine sanction to all the variants known to have existed, but no objective scholar of the history of the Qur'an text can possibly find a direct connection between the two. In the next chapter we shall give our own impressions on the real causes of the variant readings and missing passages of the Qur'an.