KORAN.] MOHAMMEDANISM 005 The Koran not com- Hafsa, which, however, was soon afterwards destroyed by Merwan, the governor of Medina. The destruction of the earlier codices was an irreparable loss to criticism ; but, for the essentially political object of putting an end to con troversies by admitting only one form of the common book of religion and of law, this measure was necessary. The result of these labours is in our hands ; as to how they were conducted we have no trustworthy information, tradition being here too much under the influence of dog matic presuppositions. The critical methods of a modern scientific commission will not be expected of an age when the highest literary education for an Arab consisted in ability to read and write. It now seems to me highly probable that this second redaction took this simple form : Zaid read off from the codex which he had^ previously written, and his associates, simultaneously or successively, wrote one copy each to his dictation. It certainly cannot have been by chance that, according to sure tradition, they wrote exactly four copies. Be that as it may, it is impos sible now to distinguish in the present form of the book what belongs to the first redaction from what is due to the second. In the arrangement of the separate sections, a classifica tion according to contents was impracticable because of the variety of subjects often dealt with in one sura. A chrono logical arrangement was out of the question, because the chronology of the older pieces must have been imperfectly known, and because in some cases passages of different dates had been joined together. Indeed, systematic prin ciples of this kind were altogether disregarded at that period. The pieces were accordingly arranged in indiscri minate order, the only rule observed being to place the long suras first and the shorter towards the end, and even that was far from strictly adhered to. The short opening sura is so placed on account of its superiority to the rest, and two magical formulae are kept for a sort of protection at the end ; these are the only special traces of design. The combination of pieces of different origin may proceed partly from the possessors of the codices from which Zaid compiled his first complete copy, partly from Zaid himself. The individual suras are separated simply by the super scription " In the name of God, the compassionate Com- passioner," which is wanting only in the ninth. The additional headings found in our texts (the name of the sura, the number of verses, etc.) were not in the original codices, and form no integral part of the Koran. It is said that Othman directed Zaid and his associates, in cases of disagreement, to follow the Koraish dialect ; but, though well attested, this account can scarcely be correct. The extremely primitive writing of those days was quite incapable of rendering such minute differences as can have existed between the pronunciation of Mecca and that of Medina. Othman s Koran was not complete. Some passages are evidently fragmentary ; and a few detached pieces are still gxtant -which were originally parts of the Koran, although they have been omitted by Zaid. Amongst these are some which there is no reason to suppose Mohammed desired to suppress. Zaid may easily have overlooked a few stray fragments, but that he purposely omitted anything which he believed to belong to the Koran is very unlikely. It has been conjectured that in deference to his superiors he kept out of the book the names of Mohammed s enemies, if they or their families came afterwards to be respected. But it must be remembered that it was never Mohammed s practice to refer explicitly to contemporary persons and affairs in the Koran. Only a single friend, his adopted son Zaid (xxxiii. 37), and a single enemy, his uncle Abu Lahab (cxi.)^and these for very special reasons are men tioned by name ; and the name of the latter has been left in the Koran with a fearful curse annexed to it, although his son had embraced Islam before the death of Mohammed. So, on the other hand, there is no single verse or clause which can be plausibly made out to be an interpolation by Zaid at the instance of Abubekr, Omar, or Othman. Slight clerical errors there may have been, but the Koran of Othnuln contains none but genuine elements, though sometimes in very strange order. Of the four exemplars of Othman s Koran, one Avas kept in Medina, and one was sent to each of the three metro politan cities, Cufa, Basra, and Damascus. It can still be pretty clearly shown in detail that these four codices deviated from one another in points of orthography, in the insertion or omission of a wa ("and"), and such-like minutiae ; but these variations nowhere affect the sense. All later manuscripts are derived from these four originals. At the same time, the other forms of the Koran did Other not at once become extinct. In particular we have editions, some information about the codex of Obay. If the list which gives the order of its suras is correct, it must have contained substantially the same materials as our text ; in that case Obay must have used the original collection of Zaid. The same is true of the codex of Ibn Mas ud, of which we have also a catalogue. It appears that the principle of putting the longer suras before the shorter was more consistently carried out by him than by Zaid. He omits i. and the magical formulae of cxiii. cxiv. Obay, on the other hand, had embodied two additional short prayers, which we may regard as Mohammed s. One can easily understand that differences of opinion may have existed as to whether and how far formularies of this kind belonged to the Koran. Some of the divergent readings of both these texts have been preserved, as well as a con siderable number of other ancient variants. Most of them are decidedly inferior to the received readings, but some are quite as good, and a few deserve preference. The only man who appears to have seriously opposed Ibn the general introduction of Othman s text is Ibn Mas ud. Mas ud. He was one of the oldest disciples of the Prophet, and had often rendered him personal service ; but he was a man of contracted views, although he is one of the pillars of Mos lem theology. His opposition had no effect. Now when we consider that at that time there were many Moslems who had heard the Koran from the mouth of the Prophet, that other measures of the imbecile Othman met with the most vehement resistance on the part of the bigoted champions of the faith, that these were still further incited against him by some of his ambitious old comrades until at last they murdered him, and finally that in the civil wars after his death the several parties were glad of any pretext for branding their opponents as infidels ; when we consider all this, we must regard it as a strong testi mony in favour of Othman s Koran that no party, not even that of All, found fault with his conduct in this matter, or repudiated the text formed by Zaid, who was one of the most devoted adherents of Othman and his family. But tins redaction is not the close of the textual history of the Later Koran. The ancient Arabic alphabet was very imperfect ; it not history only wanted marks for the short, and in part even for the long of the vowels, but it often expressed several consonants by the same sign. text. Hence there were many words which could be read in very different ways. This variety of possible readings was at first very great, and many readers seem to have actually made it their object to discover pronunciations which were new, provided they were at all appro priate to the ambiguous text. There was also a dialectic license in grammatical forms, which had not as yet been greatly restricted. An effort was made by many to establish a more refined pronuncia tion for the Koran than was usual in common life or in secular literature. The various schools of "readers" differed very widely from one another ; although for the most part there was no im
portant divergence as to the sense of words. A few of them gradu-