In Search of the Original Koran: The True History of the Revealed Text (13 page)

In summary, at the present time there remains no trace of the state of the Koran as it existed before the end of the first century of the Hijra, or a little later, which brings us back to the reign of Walid ibn Abd al-Malik (705-715), the era of the last shaping of the Koran attributed to the governor of Iraq, al-Hajjaj.

Lacking the ancient manuscripts of the first century of the Hijra, the historian must be content with the testimony that has come down to us. But here again, a new disappointment awaits us, for practically all the works that flourished shortly before the end of the Umayyad dynasty dealing with the differences between the compendiums (ikhtilaf al-masahi) of the Koran have in fact disappeared. These were comparative studies of the state of the Koranic text as it was practiced (especially orally) in the great regions of the Muslim Empire: Arabia, Syria, and Iraq. One of the oldest books of which we know was written by Ibn Amir al-Yahsubi (died in 118 H/736 CE); one of the last works on the Korans was by Ibn Ashta al-Isfahan (died in 360 H/970 CE). However, only the book by Ibn Abi Dawud (died in 316 H/928 CE) has come down to us. Our knowledge of these works actually comes via citations that later authors inserted into either their commentaries on the Koran, into works on Koranic readings (gira'at), or else into treatises on grammar, and so on.

In addition, Tradition has arranged things so that what has filtered down to us are only minor variants. This was admitted by a writer of the fourth century, Abu Hayyan: he had not cited in his work those Koranic variants that were too far removed from the "Uthmanian" text." This historiographic fact is of the highest importance, if one links it to the information we do have on the persecutions suffered by all those who obstinately continued to use noncanonical variants, as was the case with Ibn Shanabudh (245-328).

The traditionalist thesis about the Uthmanian collection is only a fantastic reconstruction, hiding a reality that people sought to erase from human memory: the Koran is multiple because its text has a history and thus presents an evolution, as well as variations over time. And this history was only possible because it was in the nature of the redaction of the text that would eventually become the Koran to take the routes of elaboration, composition, stylization, and rectification. In short, it was the product of a historical elaboration (divine or human, it does not matter), and not a dictation carried out on the basis of a preexisting text, definitive and ready to be published. From the time of the prophecy, the divine word had to be shaped, a task that was incumbent on scribes, and it is this operation that later generations have tried to erase, in order to give a simplified and more-reassuring image of the Koranic message, that of a text composed by God in person.

The principal and the most decisive lesson that one should draw from the information we do have on these ancient Koranic corpuses claimed to be noncanonical is that they are essentially comparable to the text of our vulgate called "Uthmanian": the works do not contain much variation except for details, neither in their structure, their content, or in the number of surahs. For one thing, the texts in each corpus are the same from one corpus to another, with small variants. They seem (still according to the descriptions we have of them) to carry a total number of surahs that differs very marginally from the vulgate, depending on whether or not one adopts certain very short surahs, like the Fatiha, or the two final surahs. The only other visible difference, without real importance anyway, is in the order of surahs, which sometimes varies significantly from one corpus to another.

From these observations, we may conclude that the ensemble of the known noncanonical works by ancient authors belong to one and the same generation, and were constituted at a well-determined stage in the evolution of revealed texts: the same one that witnessed the present constitution of both the content and the number of surahs.

There is a paradox here that will not have escaped the reader. The most ancient works in the corpuses of the Korans, like those of Ubayy and Ibn Masud, which were supposed to have been formed independently of the recension realized by the first caliphs AN Bakr and Uthman, all resemble the latter like two drops of water, and they therefore all belong to the same collection. This situation presents us with an alternative: either the definitive collection was realized in the time of the Prophet-but in this case, a Uthmanian recension never existed, or else this collecting took place after the death of the Prophet (in this case it could not have been realized by some companion of the Prophet like Zayd, Uthman, All, Abu Bakr, Ibn Masud, Ubayy), in which case would have resulted from the slow evolution of an oral tradition relying on writings that reproduced the partial compendiums of texts revealed in the Prophet's time. This is the most likely hypothesis, con sidering the structure of the text that we have studied, and the oldest traces of the Koran that are still divergent.

Moreover, the definitive version of the Koran leaves no doubt about the absence of a veritable systematic and voluntary collection that would have taken place under a definite authority. When we observe the division of surahs in the Koran, we quickly see that they did not obey any criterion of composition at all. This can be well perceived in the extreme disparity that exists between the long surahs: fifty-some pages for the longest, "The Heifer," as opposed to one and a half lines for the shortest. In fact, this longest surah is the equivalent in length to the 75 last surahs combined (out of a total of 114 surahs in the Koran). It is clear that, if there had existed even a slight desire for a systematic gathering and definitive shaping of the Koranic text, then the Koran would not today present a disequilibrium so pronounced in so important a facet as that of surahs.

After about a half century, this state of affairs ended up settling, thanks to the collections composed by anonymous scribes, into a model corpus that was made official under the Umayyad caliph Abd al-Malik or under his son Walid I, that is to say, perhaps in the time of al-Hajjaj's rule.

The manuscript of Samarkand is a composite, containing portions written by different scribes at different dates. It is not complete, since dozens of pages are missing at the beginning and at the end, and some in its middle. It would assuredly merit the deeper study that we are cruelly lacking for ancient manuscripts. It is now located in Tashkent, after long peregrinations that brought it in 1485 to Samarkand, then in 1868 to St. Petersburg, ending up in 1917 in Tashkent. In 1905, Czar Nicholas II had Dr. Pissarev make fifty reproductions of it under the title "Kufic Koran of Samarkand," of which some copies can still be found in Western libraries.

There are some particularities of this manuscript that are of interest to us here.' It has omissions that are due, some of them, to a scribe's negligence; some have been rectified by corrections in the margins. Some variants are worth remembering. The name "Allah" is sometimes replaced in the Samarkand manuscript by the pronoun hu (-wwa), as in verses 2:284 (ms. of S. 90) and 3:78 (ms. of S. 109). Sometimes, it is omitted, as in 2:283 (ms. of S. 89) and 5:119 (ms. of S. 252). In 3:37, it is the expression "inna Allah" that disappears from the manuscript (p. 92), without harming the comprehension of the text. Note also the variant of verse 3:146 (ms. of S. 134) that speaks of "what [Muhammad] (ma asabahu) underwent" in the course of the battle of Uhud, whereas the vulgate speaks of "what [his companions] underwent (ma asabahum)." It seems we have in this manuscript the first version of the text that alludes to the severe wounds the Prophet received during this battle, a vulnerability that would have provoked doubts in the minds of a certain number of his companions (see verses 3:146-47).

Moreover, this manuscript gives us interesting clues concerning the history of the composition of the verses. Like the most ancient manuscripts, our manuscript does not demarcate verses in certain parts, and elsewhere only for sets of five or ten verses. These hesitations are very probably due to the fact that the manuscript is composed of fragments from different dates. But a quick comparison of the manuscript with our vulgate shows us a phenomenon worthy of interest. In three cases of verses-6:91, 6:128, and 7:25-where the vulgate signals the end of the verse by adding the conjunction wa (and) at the start of the following verse, this is absent in our manuscript. On the other hand, we find an inverse case (3:113), where it is the vulgate that suppresses the wa at the start of the verse, although it is part of the phrase in the manuscript. We also find some of the Seven "Readings" also omitting the wa at the start of certain verses, contrary to the vulgate. This relation between the presence and absence of the sign of the end of the verse and of the coordinating conjunction would merit a study that would shed more light on the history of how the Koranic texts came to be enduringly divided into verses. But we may deduce from the preceding that this division of the text into verses necessitated a readjustment of the text to adapt it better to its new recitatory dynamic.

Can Muslim orthodoxy continue to claim a literal and compositional authenticity for the Koran? The Koran itself, as we have seen, invalidates such a claim. Koranic doctrine in this respect is quite clear: the divine message has no single form but corresponds to two distinct realities, the original and the copy. The original is designated in the Koran by the precise term kitab, and this writing was consigned to a heavenly tablet (lawh), well guarded by and near to God. As for the copy, it is an emanation and an extract from this original and results from a chain of transmitting agents: Gabriel, Muhammad, and the scribes, as well as secretaries assigned to the shaping of the revealed statements. Above all, we have seen that it was by an act of inspiration (wahy) that the passage of the kitab to the qur'an was realized, and not by a simple reproduction of a text.

At no time has the Koran claimed a literal identity between the revealed text and its divine source. This is so evident that even the two myths imagined by orthodoxy in order to prove the authenticity of the Koran remain scarcely convincing. It was effectively claimed that the archangel Gabriel had adopted the habit, every year during the month of Ramadan, of taking stock of the revealed text: correcting it, putting into order the verses revealed during the preceding year, and also eliminating verses said to be abrogated. In short, here we have work of a scribal nature that people have attempted to impute to a divine authority in order to legitimate it. The other myth attributed the communication of the revealed text to two successive stages: first, the text of the Koran came down in its entirety by traversing the seven heavens; arriving at the lowest heaven, it remained there, conserved in a place called beit al-'izza. Second, from there, Gabriel drew verses that he communicated regularly to the Prophet during the whole period of Koranic revelation, meaning over more than twenty years. It is clear that this latter myth invented by Muslim orthodoxy also tries to blot out the contingent character of the revealed texts, which relates to the fact that they are manifestly linked to the long history of the Prophet and to the vicissitudes of his struggle against the Qurayshites.