Koran Birmingham: VARIANT READINGS – by GABRIEL SAID REYNOLDS

VARIENT READINGS – by GABRIEL SAID REYNOLDS
The Birmingham Qur’an in the context of debate on Islamic origins Published: 5 August 2015

A section of a Qur’anic manuscript in the Hijazi script held by the University of Birmingham. Experts in radiocarbon analysis at the University of Oxford have dated the parchment to between the years 568 and 645

On July 22 the BBC published an article on its website entitled “‘Oldest’ Koran Fragments found in Birmingham University” that sparked great excitement among academics and Muslim

believers alike. At the heart of the article is the news – mentioned in last week’s issue of the TLS – that two leaves of a Qur’an manuscript studied by Alba Fedeli and held at Birmingham’s university library had been carbon dated to somewhere between AD 568–645 (carbon dating allows only for a range of years, and not a precise date). Several academics and Muslim leaders are quoted in the article, and they agree that this finding reveals something of immense importance about the origins of Islam. Muhammad Isa Waley, a manuscript expert from the British Library, declares that the news will have Muslims “rejoicing”. David Thomas, Professor of Christianity and Islam at Birmingham University, explains that whoever composed this manuscript might “have heard the Prophet Muhammad preach”, and insists that the dating of the manuscript shows that the Qur’an “has undergone little or no alteration”.

However, the BBC article – like a subsequent New York Times article (also July 22) – misses the most significant point about the dating of this Qur’an manuscript (which contains only a small section of the text: parts of chapters 18, 19, and 20). Islamic tradition reports that Muhammad received revelations from the angel Gabriel between the year 610, when he was forty years old, and his death in 632. But according to Islamic tradition, he did not write down these revelations. Instead, his proclamations were preserved only on various scraps (one tradition speaks of palm leaves, parchment and the shoulder blades of camels), or in versions which some of his companions composed. An official text of the Qur’an was only recorded around 650, during the reign (644–656) of Uthman (the third Caliph, or successor, of the Prophet Muhammad). According to a well-known Islamic tradition Uthman had his “official” text of the Qur’an prepared by a committee, and all variant versions destroyed by fire: “Uthman sent to every Muslim province one copy of what they had copied, and ordered that all the other Qur’anic materials, whether written in fragmentary manuscripts or whole copies, be burnt”.

Most traditional Muslim scholars believe that Uthman’s version of the Qur’an was reliable, although there are interesting traditions according to which certain companions of Muhammad resisted the Caliph’s orders, notably one Ibn Masud (a companion who is said to have been commended by the Prophet for his knowledge of the Qur’an) in the city of Kufa. Moreover, many early Shiite scholars doubted the reliability of the Qur’anic version of Uthman (someone who, from a Shiite perspective, usurped the rightful place of Ali at the head of the Islamic community).

Yet the very early dating of the Birmingham manuscript (568–645) – almost certainly before the reign of Uthman – casts doubt on the traditional story. The Birmingham

manuscript does not appear to be a scrap, or a variant version kept by some companion, which somehow escaped the Caliph’s burning decree. It appears to be the standard Qur’an which Muslims attribute to Uthman. In other words, the dates of the Birmingham manuscript are not simply early. They’re too early. Instead of rejoicing, the news about this manuscript should lead to head-scratching.

Moreover, the extremely early date range of the Birmingham text (most of which is before even the date when Muhammad is said to have begun his preaching) seems to confirm the early dating of other manuscripts. Among the manuscripts that were discovered in 1972 when repair work was being done on the ceiling of the Great Mosque of Sanaa in Yemen was a rare Qur’anic palimpsest – that is, a manuscript preserving an original Qur’an text that had been erased and written over with a new Qur’an text. This palimpsest has been analysed by a German husband and wife team, Gerd and Elisabeth Puin, by Asma Hilali of the Institute of Ismaili Studies in London, and later by Behnam Sadeghi of Stanford University. Sadeghi benefited from the use of X-Ray fluorescence imaging to render certain leaves of the lower (that is, original) text of the Qur’anic palimpsest visible. What all of these scholars have discovered is remarkable: the earlier text of the Qur’an contains numerous variants to the standard consonantal text of the Qur’an.

Now here it is important to explain that ancient manuscripts of the Qur’an tend not to represent all of the vowels, but only a skeletal form of Arabic consonants. Later manuscripts vary in terms of the marks they add to the consonantal “skeleton” of the Qur’an to indicate vowels and consonants. Indeed, throughout most of Islamic history there were open discussions about variant readings of the Qur’an. Things changed only in the early twentieth century. In 1924 a committee organized by the Egyptian ministry of education produced a text of the Qur’an for use within the country (and had competing editions sunk in the Nile River). This Egyptian text (slightly revised later in 1924, and again in 1936, the first year of King Farouk’s reign, for which reason it became known as the King Farouk Qur’an) has now become the standard Qur’an text. Today this text is so widespread it might lead one to conclude that the Qur’an has never had any variants. Yet this reflects the success of the Egyptian project, and not the history of the Qur’anic text.

Nevertheless, while the history of Qur’anic variants has long been a topic of academic discussion, it has also long been thought that at least the Qur’an’s consonantal skeleton was unchanging. Before the Sanaa palimpsest, no early manuscript was known to vary significantly in terms of that skeleton. The basic form of the Qur’anic text, in other words, was thought to have been more or less perfectly preserved. Yet the Sanaa manuscript, which is almost certainly the most ancient Qur’an manuscript known to us, contains a surprising number of variants, including completely different words, and presents the chapters (known as suras) of the Qur’an in a different order.

What made this discovery all the more exciting was the dating of this manuscript. When Sadeghi sent out a sample of parchment of the palimpsest for radiocarbon dating (performed at a laboratory in Arizona) the result came in that it had a 75.1 per cent chance of dating before 646. Now carbon dating estimates when the animal was slaughtered to make the parchment, not when the text itself was written, but it is thought that not much time would pass between the two (in theory it is possible to date the ink on a manuscript, but it is difficult to get enough ink, and to avoid contamination from the parchment, to do so). Sadeghi’s colleague at Stanford, Uwe Bergmann, announced elsewhere that this manuscript likely dates to the lifetime of Muhammad himself.

Meanwhile, two fragments of this same manuscript were sent out for dating by a French scholar of early Arabic, Christian Robin, to a laboratory in Lyon. Results came back which indicated that the manuscript is older still: one fragment was dated to 543–643 and the other to 433–599. There has been a lot of discussion of these early dates over social media. Some scholars have held that they are so early that the job had been botched. However, still further tests (not yet published) on additional fragments of this manuscript have been done which have also yielded early results. In any case, the Birmingham results suggest that Lyon might not have botched the job after all. Intriguingly, the first date range from Lyon (543–643) corresponds rather closely to the date range given (from a laboratory in Oxford) for the Birmingham manuscript (568–645).

Now the Sanaa manuscript has so many variants that one might imagine it is a vestige of an ancient version that somehow survived Uthman’s burning of all versions of the Qur’an except his own. The problem with this idea is that the variants of that manuscript do not match the variants reported in medieval literature for those codices kept by companions of the Prophet. Sadeghi argues that this must have been the codex of some unknown companion. This is an interesting, although speculative, idea. For now all we know is that our most ancient manuscript of the Qur’an does not agree with the standard text read around the world today.

It is also important to remember that the carbon dating of parchment is an imprecise science (something indicated by the large range of possible dates given for the various fragments). Scholars have long debated, for example, the carbon 14 tests that have been carried out on the Dead Sea scrolls found around Qumran. And indeed, the date ranges for the Scrolls vary widely. For example, the famous Isaiah scroll of Qumran has been dated (with a 95 per cent probability) variously either to 351–295 BC, or 230–53 BC (or, according to another laboratory, 351–296 BC, or 203–48 BC).

Thus the Dead Sea Scrolls dating allows for a range of several hundred years (and even then many scholars argue that the palaeographic dating – that is, dating based on the script – of the Scrolls is more reliable than carbon dating). What is more, the dating of the Dead Sea Scrolls might be considered more accurate than the dating of the Qur’an manuscripts, since fragments from many different samples of the scrolls – and even samples from other materials found at Qumran (including a piece of leather and a scrap of linen) have been tested. This allows scientists to calibrate their measurements more precisely. Such calibration has not yet been possible for Qur’an manuscripts. We are not even sure of the precise original location of the manuscripts (the Birmingham manuscript may have been located in Fustat, Egypt at some point, but this does not mean it was written there).

Nevertheless, the early dating of the Sanaa manuscript is telling in light of the extremely early dating of the Birmingham manuscript. Almost all of the date range (568–645) given for that manuscript is before the reign of Uthman and most of it is even before the traditional dates of Muhammad’s preaching. This is all the more startling because the Birmingham manuscript has the appearance of a more developed text. Both the lower text of the Sanaa palimpsest and the text of the Birmingham manuscript include certain features – such as dividers between suras, and certain dashes to distinguish consonants – which may represent a later stage of writing (which means it is possible that we will still find even earlier, and more primitive, Qur’an manuscripts). However, the Birmingham manuscript largely conforms to the standard text of the Qur’an. It seems, in other words, to represent a distinctly later stage in the history of the Qur’anic text. Now one might assign the Birmingham manuscript to the very end of the date range given for it (568–645) because of what we think we “know” about the traditional story

of the Qur’an’s origins, in an attempt to make it fit in to the reign of Uthman (644–656). Alternatively, it might be time to consider again what we think we know.

The upshot of all of these early dates is that the Qur’an may very well date earlier than Uthman, possibly much earlier. It may be time to rethink the story of the Qur’an’s origins, including the traditional dates of Muhammad’s career. In other words, what observers have celebrated as something like evidence of the traditional story of Islam’s origins (the New York Times article argued that the manuscript “offered a moment of unity, and insight, for the world’s 1.6 billion Muslims”) may actually be, when considered carefully, evidence that the story of Islam’s origins is quite unlike what we have imagined.

Nevertheless, the extremely early dating of these manuscripts is helpful for the way it helps clarify something which has troubled scholars of early Islam. Many elements of the Qur’an are difficult to understand. For example, twenty-nine of its suras begin with a series of disconnected letters, yet the origin and meaning of those letters remains a mystery (for which reason they have been dubbed the “mysterious letters”). To give another example: in two passages (2:62 and 5:69) the Qur’an speaks of a group called the Sabi’un who seem to be promised entry into heaven (along with “the believers”, Jews and Christians). Yet no one is sure exactly who these Sabi’un are. Indeed, we find that Muslim scholars, even the earliest Muslim scholars, do not understand the “mysterious letters” and cannot identify the Sabi’un. In other words, somehow the meaning of these things had been lost by the time the text reached them.

The early dating of these Qur’an manuscripts helps us make sense of this (even if it won’t tell us the meaning of the “mysterious letters” or the identity of the Sabi’un). It seems that by the time the Qur’an reached these scholars (whose work begins to be written in the second half of the eighth century) it was already a very old text which was no longer understood well. This is a hypothesis raised by Michael Cook at the end of his work The Koran: A very short introduction (2000). Now that the Qur’an appears to be older than imagined, the hypothesis seems more likely than ever.