It is six weeks since the news first broke of a fragmentary Qur’an manuscript whose radiocarbon dating (95% accuracy) is between 568 and 645. Early reports declared how much the discovery supported traditional Muslim accounts of the origins of the Qur’an, but this week the news media went abuzz with claims that the Qur’an may in fact predate Muhammad and debunk Islam’s account of its own origins. (N. B.: The article was first published in the Times of London, but linking to paywalls is unhelpful.) Some scholars (such as Georgetown Prof. Jonathan Brown) have now published critiques of the counter-claims, and the buzz continues apace.
Before we get too carried away, it’s important to remember that this is a single manuscript. Statisticians derisively refer to the conclusions which can be reached when the number of data points N = 1. Such a datum is not “statistically significant.” While statistical significance is not the only measure of historical evidence (and indeed, most ancient and early medieval history does not achieve statistical significance), it does suggest that we might usefully remember the limits of what we are looking at here.
This discovery consists of two leaves of a single manuscript, containing a small portion of text which we recognize as a portion of what has (later) been labeled the Qur’an, without varying from the text of later standardized Qur’ans. This manuscript is not “the Qur’an” or “the Qur’anic text,” but rather “a Qur’an” and “a Qur’anic text.” It does not tell us about the range of Qur’anic texts available at the time it was written, but only that that range included one exemplar, part of whose text (for the whole does not survive) does not vary from what later became the standard text. It does not tell us that the order of the suras was already standardized, only that what we now call Sura 20 (or at least, what we now call it’s opening part) immediately followed a break after what we now call the close of Sura 19. It does seem to indicate that, for the scribe of this codex at least, there existed a notion of ayah and sura divisions (“verses” and “chapters”) which could be marked by punctuation, and that in the small portion represented here, the divisions match what became the standard version. But it says nothing about what might once have existed elsewhere in the codex, or what other codices might have contained. Indeed, because the carbon dating applies only to the writing material, it does not guarantee that the writing is as old (although paleography confirms that this is an “early” script, but we lack reliably dated comparanda for how Arabic was written in codices in the 640s or earlier). In other words, this codex tells us very little, and is compatible with a wide range of theories about the origins of the Qur’an.
It is perhaps useful to review what has become the most common traditional Muslim view of the formation of the Qur’an: Muhammad, receiving the revelation from God through the angel Jibril (Gabriel), recited it in the hearing of his Companions, but did not write it down himself. (According to one widespread interpretation of a difficult phrase in the Qur’an, Muhammad was illiterate.) Parts were written down, but after Muhammad’s death (usually dated c. 632) his successor Abu Bakr (d. 634) desired to compile a single codex containing all the revelations, which was done by Zayd b. Thabit, as well as by other Companions of Muhammad. Muhammad’s third successor ‘Uthman (reign usually given as 644-656) noticed differences among these codices, and had a committee create a standardized text. Variations persist in a certain number of standardized recitation traditions, but the written text is surprisingly standardized among Qur’ans surviving today. (The lower writing of the Sana’a Qur’an palimpsest is, I believe, still the only Qur’an manuscript known not to belong to the standardized text type of the Qur’an.)
This manuscript does not confirm the traditional Muslim view any more than any single manuscript could confirm it. If the traditional Muslim view is correct (and if the pre-645 date is correct), this manuscript would probably be a Companion codex, one of those which varied so much from each other that ‘Uthman decided a standardization was necessary. The fact that the surviving portion matches ‘Uthman’s standardization merely indicates that ‘Uthman’s standardization was (at least in the portions attested here) not a creation ex nihilo, which of course Muslim tradition already believes that it was not. (There is just the barest sliver of a chance that this codex is in fact what remains of Uthman’s first standardized codex, but that would be even more surprising.) The logic which trumpets this codex as proof that the standard Qur’an existed before Uthman would make Uthman’s standardization unnecessary, and the traditional Muslim account unintelligible.
On the other hand, this codex could also fit with various revisionist theories. Revisionism in religious history is not exclusively a party line, nor need it presume that all religious autobiographies are false, but rather revisionism includes any attempt to tell a different story from the same sources. This is what all historians do, to varying degrees. Some revisionists clump together based on their preferred models, while others are eclectic. But revisionist theories which suggest that the Qur’an was not associated with Mecca (by Patricia Crone), for example, find no counter-evidence in this codex. The codex does not even lay to rest John Wansbrough’s excessively late dating of the final version of the Qur’an, since what we have before us is not a complete text. (Note: I am not endorsing Wansbrough’s model, merely pointing out a lack of logical inconsistency.) From Wansbrough’s model, the Birmingham codex would not represent the entire “finished” Qur’an, but merely one part of one exemplar in an earlier stage in the development of the Qur’an, indicating that the transition from Sura 19 to Sura 20 eventually standardized in the way represented in this early codex.
The small portion of text represented in the Birmingham Qur’an, if the carbon dating is accurate for the ink as well as the parchment, is consistent with traditional Muslim as well as revisionist accounts of the origins of Islam and its most sacred book. In other words, this discovery neither confirms traditional Muslim accounts nor debunks them. It is a very welcome addition to the small range of early Qur’anic texts available for study, and its evidence needs to be suitably weighed and incorporated into scholarly syntheses. Such syntheses will likely maintain the full range of views from traditionalist Muslim to radical revisionist (and everywhere in between), but the new evidence will fill in the various competing pictures more. But remembering that the Birmingham Qur’an does not represent the Qur’an, only a Qur’an, will sell fewer headlines, even as it leads to more correct conclusions, because more cautious.
[Image credit: By Anonymous (http://www.bbc.co.uk/news/business-33436021) [Public domain], via Wikimedia Commons]