Progress in the Study of the Koran Text

Arthur Jeffery

PERHAPS IT WOULD BE AS well to begin with a word about the origins of the Plan. It grew out of a suggestion made to the writer in Madras many years ago by the late Canon Sell, that one of the most pressing needs of Islamic studies was a reliable critical commentary on the Koran. Orthodox Muslim commentaries we have in great abundance, and both rival schools of the Ahmadiyya movement have been busy providing us with apologetic commentaries intended to make the Koran more palatable to people with a Christian background. What we needed, however, was a critical commentary which should embody the work done by modern Orientalists as well as apply the methods of modern critical research to the elucidation of the Koran.

No great advance had been made, however, in the collection of material for such a commentary, when it became evident that a necessary preliminary was a volume on the theology of the Koran, which would provide us with some account of the development of the teaching of the Koran, and make possible a reliable scheme according to which would be arranged the masses of material that were available to be digested into the commentary. Now no such volume existed, for Grimme’s “System der Koranischen Theologie” in the second part of his Mohammed (Münster, 1895) was quite unsatisfactory, and Sacco’s work Le Credenze religiose di Maometto, when it appeared (Rome, 1922), proved to be the work of a writer inadequately prepared for the task. It was necessary, therefore, to make an attempt to fill this need.

Originally published in The Muslim World 25 (1935): 4—16. Reprinted with permission.

Here also it did not take long to discover that such a work could not be written until we had a satisfactory lexicon to the Koran, such a lexicon as the older Grimm-Thayer, or the more recent Milligan-Moulton lexicon to the New Testament. In other words, no adequate examination had ever been made of the Koranic vocabulary. Very little work at the native commentaries is sufficient to demonstrate how much at sea they were with regard to much of the technical religious vocabulary of the Koran, and how much more inclined they were to interpret the Koranic verses in the light of the theological and juristic controversies of their own day, than to work back to discover the original meaning of the passages.1 The native lexicons are also as a rule far from helpful when it comes to Koranic vocabulary, so that without an independent investigation it is almost impossible to know what exactly Muhammad meant by the terms he used. Before we can discuss the development of the theological ideas of the Koran, therefore, we must make an exhaustive investigation of the vocabulary of the Koran.

A preliminary study along this line was made by the writer in 1925-26 in a thesis on the foreign vocabulary of the Koran, which unfortunately, owing to the printing costs, it has not yet been possible to publish. This study took up all the technical and cultural terms in the Koran of Hebrew, Aramaic, Syriac, Ethiopic, Iranian, and other origins, i. e., the non-Arabic element in the Koranic vocabulary. It made clear, however, that before we can tackle the study of the Arabic element in any scientific manner, we must have a critical text of the Koran. There is little use in making a lexicon to the standard text of the Koran when it is known that it is but one form of tradition as to the text, and that there are thousands and thousands of textual variants, many of them representing a much older type of text than the one we have in our hands.

It is an amazing fact that up to the present we have no critical text of the Koran. What type of text it was that was published in 1530 or thereabouts at Venice by Paganini Brixensis we cannot tell, as this first European edition of the text of the Koran was entirely destroyed at the command of the pope.2 The editions of Abraham Hinckelmann in 16943 and Ludovico Marraccio in 16984 were monuments of industry at that date, but naturally not constructed on any critical basis. Flügel’s edition of 1834,5 which has been frequently reprinted, is still the text used by practically all Western scholars, but while it is a beautiful and carefully printed text, it is useless for critical purposes. It has no critical apparatus, but, worse than that, the text represents no consistent Oriental tradition either in orthography or reading, and if Flügel used some critical principle in constructing his text, no one to this day has been able to discover what it is.

Oriental lithographs of the Koran are legion. The vast majority of them give with more or less accuracy the text-tradition of Hafs from ‘Asim, i.e., the best known of the three traditions from the Kufan School which may be said to represent the textus receptus in Islam. In North Africa the lithographs usually follow the text tradition of Warsh from Nafi’, i.e., the tradition of the Medinan School. I have heard of Korans written according to the text tradition of ad-Duri from Abu ‘Amr, i.e., the tradition of the Basran School. This tradition used to be followed in the Sudan up to a generation ago,6 but apparently no texts were ever lithographed according to this tradition. With the passage of time the lithographs of the Hafs text had become progressively normalized to the customary orthography of the day, and thus could not be said to represent that tradition accurately. In 1923, however, the Egyptian government issued a standard edition under the editorship of a board of Muslim savants, and in this edition an attempt is made to reproduce as faithfully as possible the original Hafs tradition. Owing to the use of modern qira’at works instead of going back to the oldest sources of the tradition, the editors did not quite succeed in their endeavor, but their edition is nevertheless the nearest approach to a critical edition ever produced in the Orient.7 It, however, contains no apparatus criticus. In some of the Kazan lithographs published in the middle of the last century there are given on the margin the variant readings of the Seven, and these appear also in certain Indian lithographs which are derived from the Kazan text. These readings, however, are given very imperfectly and frequently inaccurately. There appeared at Teheran in 1323 A.H. a lithograph giving a reproduction of a considerable portion of an old Kufic text, along with the Hafs text, and on the margin a not always accurate attempt at reproducing the variants of the Seven.8

In brief outline, the history of the Koran text is this. When Muhammad died there was no collection of his revelations in any official form. It is possible that he meant to make such a collection to clear up finally what had been abrogated and what not, but he died before this was done. Portions of revelations had been written down during his lifetime by various persons in his community, portions had been memorized and indeed some portions had apparently been used liturgically in the community. After his death several of his followers made the attempt to collect all the known revelations and write them down in codex form. Codices of Ibn Mas‘ud, Ubai b. Ka‘b, ‘Ali, Abu Bakr, Abu Musa al-Ash’ari, Miqdad b. al-Aswad, etc., were known and referred to. With the spread of Islam and the founding of metropolitan centers, the necessity of having the Koran in written form was urgent, and we find growing up what we may call metropolitan codices in the centers of Mecca, Medina, Damascus, Kufa, and Basra.

As the differences between these various codices threatened to become a scandal and a disrupting force within Islam, the third caliph, ‘Uthman, made an attempt to canonize the Medinan codex. He had copies of this written out and sent to each of the metropolitan centers, and ordered all other codices to be burned. ‘Uthman seems to have been very successful in his destruction of the codices, for all the old Kufic manuscripts that have survived to us9 apparently represent the same type of text with insignificant variants, which can usually be put down to the vagaries of the scribes. ‘Uthman’s work was intended to standardize the consonantal text, yet for long after ‘Uthman’s time there is evidence that variant traditions as to the consonantal text survived among the learned, and we can gather a great mass of material as to the readings in the text of Ubai or Ibn Mas’ud.10 As only one type of text continued to be copied, however, it is clear that only such readings from the older codices as had some philological or theological interest would be likely to survive.11

This consonantal text, however, was unpointed. That is, there was nothing to distinguish a b from a t, a th, an n or a y at the beginning or middle of a word. Similarly f and q; j, h and kh; s and d ; r and z; s and sh; d and db, t and z were indistinguishable. Thus even in the ‘Uthmanic standard text there was possible a great variety of variant readings according to the way in which the skeleton text was pointed. Besides this there was the vowelling, for even when the consonants (huruf) were settled, they might be vowelled very differently. Thus there grew up in the different metropolitan centers what one might call schools of readers who developed variant traditions as to how the pointing and vowelling of the text should be done. It was generally the work of some great teacher, whose system was followed by generation after generation of pupils. Pupils were required to memorize two things—first a tradition as to the huruf and then a tradition as to the vowelling. Of many authorities of the second and third Islamic centuries we read that so and so had an ikhtiyar fi’l huruf, i.e., a tradition of his own as to what the huruf should be.

The curious thing is that in many cases we find that this ikhtiyar does not confine itself to a choice among possible pointings of a standard consonantal text, but frequently represents a different consonantal text altogether. Sometimes this represents a correction of some mistake or inelegancy of the ‘Uthmanic text,12 but in the majority of cases suggests that in the different metropolitan centers some of the readings of the older codices were strong enough to survive in the schools alongside the ‘Uthmanic readings. The authority of the standard text, however, naturally grew as standard texts always do,13 and these variants from it came to be recorded merely as curiosities. Before the end of the third century it was to all intents and purposes fixed, and the main differences between the schools were in the matter of vowelling the consonantal texts.

Gradually these variations crystallized themselves into the traditions of the great schools of Koranic learning at Medina, Mecca, Damascus, Basra, and Kufa. Homs for a time seemed about to develop a school of its own, but it soon coalesced with that of Damascus. Later still, these traditions attached themselves to the names of certain famous teachers. In the year 322 A.H. the great Koranic authority Ibn Mujahid at Baghdad succeeded through his influence with the wazirs [viziers], Ibn ‘Isa and Ibn Muqlah, in having an official fixing ne varietur of the permissible readings of the text. Not only was there a definite canonization of one system of huruf (supposedly the ‘Uthmanic) with a prohibition of the use of any other ikhtiyar, but also a limiting of the variations in vowelling the text to the systems of the seven.14

The seven chosen by Ibn Mujahid were Nafi‘ of Medina († 169), Ibn Kathir of Mecca († 120), Ibn ‘Amir of Damascus († 118), Abu ‘Amr of Basra († 154), ‘Asim of Kufa († 128), Hamza of Kufa († 158) and al-Kisa’i of Kufa († 182). The curiosity in this list is in the choice of three readers from Kufa. The choice of al-Kisa‘i seems to have been due to some personal predilection of Ibn Mujahid. Al-Kisa’i had a great, though hardly deserved, reputation as a grammarian, but his Koran readings in general follow those of Hamza, and where he differs from Hamza his variation is rarely of any importance. There was a considerable body of opinion that Ya‘qub of Basra († 205) ought to have been chosen in his place, while some were in favor of Khalaf of Kufa († 229), or of Abu Ja’far of Medina († 130). Thus in spite of Ibn Mujahid there continued to be memorized and handed on the tradition as to the ten. Besides these there survived, with varying degrees of authority, the traditions of Ibn Muhaisin of Mecca (t 123), al-Yazidi of Basra († 202), al-Hasan of Basra († 110), and al-A’mash of Kufa († 148), making the fourteen.

We now have, then, two classes of variants to the Koran text, the canonical, consisting of the variants of the seven canonized by Ibn Mujahid, and with lesser degree of authority those of the ten, and uncanonical (technically known as shawadhdh) consisting of all other variants—those of the four who make up the fourteen coming nearest to recognition as canonical. As was natural, we soon had differences of tradition within the schools as to the readings of each of the seven, but by the next century two lines from each of the seven had been chosen as the orthodox tradition, and when the traditions came to be digested into book form only these two lines were given. Thus we have the system15

Nafi‘ of Medina according to Warsh († 197) and Qalun († 220).

Ibn Kathir of Mecca according to al-Bazzi († 270) and Qunbul († 280).

Ibn ‘Amir of Damascus according to Hisham († 245) and Ibn Dhakwan

(† 242).

Abu ‘Amr of Basra according to ad-Duri († 250) and as-Susi † 261).

‘Asim of Kufa according to Hafs (t 190) and Abu Bakr († 194).

Hamza of Kufa according to Khalaf († 229) and Khallad († 220).

Al-Kisa‘i of Kufa according to ad-Duri († 250) and Abu’l-Harith († 261).

Of these, as we have already indicated, only the systems of Warsh from Nafi‘, of Hafs from ‘Asim, and of ad-Duri from Abu ‘Amr seem to have gained any wide acceptance, and for some reason which has not yet been fully elucidated, the system of Hafs quickly gained such an ascendancy over all the others as to have become the textus receptus of Islam, being used today everywhere except in the stretch of North Africa from Tripoli to Morocco.

Early in the fifth Islamic century the systems of the seven were digested into book form. The outstanding work of this kind was the Taisir of the Andalus scholar ad-Dani (t 444) who himself followed the system of Warsh. The Taisir was versified by ash-Shatibi,(t 590) in a work called Hirz al-Amani (also commonly known as Ash-Shatibiya, which has been the subject of innumerable commentaries and is still the chief source used by Muslim savants in their studies of the Koran text.

The task of preparing a critical edition of the Koran, therefore, is twofold—first that of presenting some form of tradition as to the text itself, and secondly that of collecting and arranging all the information scattered over the whole domain of Arabic literature, concerning the variant readings both canonical and uncanonical. The writer had begun to collect variant readings years ago, when he first became interested in the Koran, but in 1926 began the task of consistently working through all the Arabic commentaries, lexicons, and philological works to collect the various readings recorded. That same year Professor Bergsträsser published the first fascicule of his Geschichte des Qorantexts and it was evident that our studies were interlocking. We met at Munich in 1927, and agreed to collaborate on a much bigger plan of assembling all the material that would assist in some day making it possible to elucidate fully the history of the Koran text. I was to go on with my task of collecting the variants and preparing an edition of the text, while Bergsträsser was to commence gathering material for an archive of photographs of all the oldest Kufic manuscripts of the Koran, a collation of which he hoped would throw light on the history of the text. Then we were to pool our resources with a view to a large volume dealing with the variants.16

Meanwhile there remained much supplementary work to do. A large number of source books that we needed were still in manuscript, and some indeed had yet to be discovered. A pupil of Bergsträsser’s, Dr. Otto Pretzl, was at work already on a critical edition of the famous Taisir of ad-Dani, of which we possessed as yet only a very poor lithograph from Hyderabad. This Taisir was published in the Bibliotheca Islamica in 1930,17 followed in 1932 by an edition in the same series of the Muqni’ of ad-Dani,18 both edited by Dr. Pretzl. Meanwhile at Damascus in 1927 had appeared the two large volumes of Ibn al-Jazari’s Kitab an-Nashr fi’l-Qira‘at al-‘Ashr, so that for the first time in our hands a sound basis for the study of the seven and the ten. That is, with regard to the canonical readings, we were now in a position to go back as near as we are ever likely to get to the original sources. In the winter of 1928 Professor Bergsträsser spent some months in Cairo and commenced the photographing of the Kufic codices preserved here. Also he was able to make considerable progress with the task of editing the text of Ibn Khalawaih’s work on the uncanonical readings,19 which we have reason to think represents Ibn Mujahid’s own teaching on this subject, as the Taisir of ad-Dani represents his teaching on the canonical variants. Also he was able to extract from the manuscripts of the Muhtasab of Ibn Jinni, that important philologist’s references to uncanonical variants.20 Before returning to Germany he also made arrangements for a Cairo edition of the Tabagat al-Qurra’ of Ibn al-Jazari, a large work giving biographies of all the early Koranic authorities, and what is more important, the isnads showing through what lines of tradition their readings reached them.

Another pupil of his, Dr. Eisen, was set to work on an edition of the Kitab Fada ‘il al-Koran of Abu ‘Ubaid, which will be published shortly. Meanwhile the search for older sources went on. Following on Bergsträsser’s discovery of the work of Ibn Khalawaih, the present writer has discovered two manuscripts of the lost Kitab al-Masahif of Ibn Abi Dawud, and more recently has unearthed a beautiful complete manuscript of the I‘rab al-Qira’at ash-Shadhdha of al-‘Ukbari. In the autumn of 1930, Dr. Pretzl visited Stambul [Istanbul] and carefully examined the many libraries there, unearthing many manuscripts of great interest for the text history of the Koran, and providing material for many years of work.21 Other source material has also appeared in the Orient. In 1932 the firm of al-Halabi in Cairo completed tbe printing of the commentary of the Yemenite scholar ash-Shawkani, which is very rich in uncanonical variants, and particularly useful because the author apparently had access to sources which are no longer available to us. Also the same firm published an edition of the famous old commentary of Abu Sham on the Shatibiya, viz., Ibraz al-Ma‘ani min Hirz al-Amani, on the margin of which are two works by a living Koranic scholar ad-Dabba,22 one a supercommentary to the Shatibiya and the other an explication of the readings of the three extra who make up the ten. Bergsträsser, indeed, was very hopeful of enlisting the interest of modern Muslim savants to publish a whole corpus of qira’at works, at least of those which from their point of view would be considered orthodox. Cf. the Jami’ al-Bayan, which is ad-Dani’s biggest work, the Lata’if al-Isbarat of al-Qastallani, the Kitab al-Hujja of Abu ‘Ali al-Farisi, the Kitab at-Tajrid of Ibn Fahha m, the Ma’ani al-Qur‘an of al-Farra’, the Suq al-‘Arus of Abu Ma’shar at-Tabari, etc., etc.

In the summer of 1933 while I was in Oxford, working in the Bodleian Library on one little item of our plan, Bergsträsser was lost in a mountaineering accident in the Bavarian Alps during his summer vacation. His death is a tremendous loss to Islamic studies in general, and to Koranic studies the loss is irreparable. On my way back to Cairo in the winter I stopped at Munich to spend some time with Dr. Pretzl, who is at present carrying on Bergsträsser’s work, and we have made the best arrangements we can for carrying out the plan. Dr. Pretzl is finishing the third fascicule of the Geschichte des Qoran-texts. The edition of Ibn Khalawaih is finished, and my annotations thereto, which were originally to have been a Nachtrag [appendix] to the volume, will now appear separately in one of the journals, possibly in Islamica. Dr. Pretzl will continue to gather the material for the archive of Kufic codices and other material for the history of the text.23 I shall produce the text with apparatus criticus and also attempt the volume of notes on the variant readings, though how this latter can ever be adequately done without the assistance of the massive learning and critical acumen of Professor Bergsträsser is more than I can see.

This, then, is how the plan stands. For the text it had been my original intention to commence with Flügel’s text as a basis, and endeavor to reconstruct the earliest possible type of text. There were two strong objections, however, to this: in the first place, by reason of the condition of the sources it would be largely a subjective piece of work, and thus perhaps not any better than Flügel’s, and secondly, it would be entirely unacceptable to Muslims. The only reasonable plan with regard to the text is to print consistently one type of Oriental tradition, and the obvious one among them is that of Hafs, which is so generally acclaimed as the textus receptus. This text will be constructed according to the oldest sources we have concerning the tradition of Hafs, but will be printed according to saj‘ and the Kufan verse numbering, with Flügel’s numbering, however, also given for convenience of reference. Pausal signs and the ajza’ will be noted, and on the margin a selection of marginal references such as those in a reference Bible, which will facilitate reference to parallel passages.

At the foot of each page will be the apparatus criticus. All the thousands of variants gleaned from the commentaries, lexicons, works of traditionists, theologians, and philologers, and even from some of the Adab books, will be given with symbols indicating the Reader or Readers who are quoted for each variant. It is hoped by means of different types to indicate in these symbols whether the authority concerned is earlier than the canonical seven, of the circle of the canonical readers, or more recent. It may also be possible to arrange some symbolical way of indicating from which school or schools the reading in question comes. It cannot be hoped that this apparatus criticus will be complete, for one finds variant readings noted in the most unexpected places, and a complete collection would involve the superhuman task of combing through the whole of Arabic literature printed and unprinted. All the more important sources that are available, however, will be utilized.

To the text it is hoped that some day there will be a volume of introduction, to provide for English readers what German readers already have in the second edition of Nöldeke’s Gerchichte des Qorans. It will certainly be accompanied by a volume of annotations, which will be in the nature of a commentary to the apparatus criticus. The bare citation of the reading with the symbol for the Reader in the apparatus will be sufficient for the Koranic expert in most cases, but the vast majority of students who use the apparatuswill want more. It is for the purpose of explaining these readings, discussing the origins, provenance and signification, that the annotations are provided, and also in cases where there is dispute over a reading to give scholars the necessary additional information that will enable them to reach their own conclusions as to the value of the various lines of tradition. A fourth volume is planned to contain a Koranic lexicon.

Apart from these four volumes it is planned, if time and money are available, to issue another series of volumes. Prof. Bergsträsser had thought of editing a series of Studien zur Geschichte des Korantexts, in which would appear material such as his already mentioned work on Ibn Jinni and Ibn Khalawaih. The necessity for such a series still exists. The manuscripts of Ibn Abi Dawud and Al-‘Ukbari recently brought to light by the present writer, the relevant section of the Berlin manuscript of the Mabani, Ibn al-Anbari’s Waqf wa Ibtida’, and similar works, must be published, and the intensive search now being made for some of these lost qira’at books will certainly have some success in recovering to us texts that will demand publication. It is also possible that the archive of Kufic codices may hold surprises that will call for early publication. The plan is, therefore, to look toward the issue of a series of Studies in the Text of the Koran, where such material, as it becomes available, can be placed in the hands of students.

This then is where the Koran plan stands at present. Whether it will be possible to carry it through in whole, or even in part, is of course another question.

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