The Qurʾān is believed by Muslims to have been revealed to the Prophet Muḥammad over the course of roughly two decades from 610 to 632 CE. The scripture is considered to have been preserved both orally - through the meticulous memorisation of the text - and in written form. While the Prophet is reported to have dictated the text to scribes, the traditional sources nonetheless indicate that the Qurʾān was not compiled in the form of a codex (mushaf) during the lifetime of the Prophet. Instead, it was compiled during the reign of Prophet's successor Abu Bakr [632 - 634] and officially standardised during the reign of the third caliph ʿUthmān b. ʿAffān. All modern copies of the Qurʾān trace back to the original codices prepared during ʿUthmān's reign.
How reliable is the traditional narrative presented above? This article aims to answer this question in three parts. The first part of this article will present evidence that affirms the historicity of the standardisation of the Qurʾān by ʿUthmān. This section shall show that the Qurʾān we have today faithfully preserves the rasm (i.e. the undotted text) of the codices prepared by ʿUthmān. Even if the Qurʾān remained unchanged after the time of ʿUthmān, can the same be said for the period between the Prophet's death and ʿUthmān's reign? This question is addressed in the second part of this article, where I shall present evidence that strongly suggests that the Uthmanic text preserves the basic wording of the Prophet's revelations. Finally, the third part of this article will discuss the transmission of the Qurʾān after ʿUthmān. I shall argue that the existence of variant readings can be justified with reference to the seven aḥruf ḥadīths, which suggest that the Prophet Muḥammad approved of some flexibility in the recitation of the Qurʾān as long as the basic meaning of the verses is preserved.
The majority of scholars affirm that the Qurʾān was likely fixed by the mid 7th century based on the following evidence:
Several Qurʾān manuscripts can be dated to the 1st/7th century. Some of these have even been radio-carbondated to within two or three decades of the Prophet's death. These include the famous Birmingham manuscript (C14: 568–645 CE), Codex DAM 01-25.1 (C14: 543–643 CE) and Ms Qāf 47 (C14: 606–652 CE). Caution is still necessary, however, as some scholars have expressed concerns about the reliability of radio-carbon dating.
Nonetheless, even if we set aside radio-carbon dating, we can still date most of these manuscripts to before ʿAbd al-Malik's reign. The earliest manuscripts are generally written in the ḥijāzī script, which is known for its slanting of the letters (and alif in particular) to the right. The shapes of letters also vary considerably from one scribe to another and even a single scribe may write the same letter in different ways. This changes in the late 1st/7th century, when scribes trained in highly regular "Kufic" calligraphy styles become more common. Some Qurʾānic manuscripts are written in calligraphy styles which are quite similar to the ones used in the Dome of the Rock as well as two inscriptions from Makkah dated to 80 AH. As these styles are generally considered to be later developments of the ḥijāzī script, manuscripts written in the latter are likely even earlier.
The manuscripts written in the ḥijāzī script also contain archaic spelling features. Most notably, the verb qāla is almost always spelled without an alif. The Kufic manuscripts, on the other hand, generally spell the verb with the alif. Thus, paleography (i.e. writing style), orthography (i.e. spelling features) and radio-carbon dating all point in the same direction: we have manuscripts of the Qurʾān that pre-date ʿAbd al-Malik's reign.
Some of these manuscripts attest a substantial portion of the Qurʾān. The Paris codex (actually known as the Codex Parisino Petropolitanus) and the British Library Qurʾān each attest more than 40% of the text. Marijn van Putten has also shown that the earliest manuscripts clearly trace back to a common written exemplar. Considering how early some of these manuscripts are, he concludes that this evidence is perfectly consistent with the traditional account of ʿUthmān's canonisation.
The Qurʾān does not explicitly refer to the political and theological disputes that affected the Muslim community after the death of the Prophet Muḥammad. As several scholars have noted, this suggests that the text most likely already reached closure prior to these events. Nicolai Sinai, for instance, states:
If the Quranic rasm did not reach closure until c. 700, it does seem odd that it should nowhere engage with the major developments that defined Islamic history between 630 and 700, in particular the unprecedented speed with which an alliance of “barbarian” tribes from the fringes of the Byzantine and Sasanian empires established themselves as the masters of an immense territory, and the bitter disputes and civil wars that soon wreaked havoc on the unity of the conquerors
Everyone agreed that ʿUthmān was responsible for uniting the Muslims upon one standard Qurʾān. Muslims from different parts of the empire with varying political and theological commitments traced the text to ʿUthmān. As Behnam Sadeghi notes, the fact that the Islamic tradition does not preserve any traces of dissenting views on this suggests that we are dealing with a genuine memory of a historical event:
It is not possible to envision an Umayyad caliph establishing the standard version. To assume so would require explaining not only how the Umayyads managed to erase their empire-wide intervention from the memories of their subjects, whose statements have reached us in large numbers, in various cities, tribes, and sects, but also how it was possible to induce the same false memory about ʿUt̠mān among all, including the dissidents who preferred ʿAlī to ʿUt̠mān.
In the previous section, we discussed the evidence in favour of the historicity of ʿUthmān's standardisation. The traditional sources inform us that ʿUthmān felt compelled to standardize the Qurʾān as Muslims from different parts of the empire disagreed with one another concerning its recitation. What led to this variation? One possible source is the Prophet Muḥammad himself reciting the Qurʾān in different ways. This view would appear to be supported by a group of ḥadiths which assert that the Qurʾān was revealed in seven "modes" (aḥruf):
Various interpretations have been put forward regarding the meaning of the term seven aḥruf. One common interpretation is that they refer to seven dialects, but this explanation would appear to be challenged by the fact that both ʿUmar and Hishām spoke the same dialect. Moreover, many of the variant readings attributed to Companions cannot be explained as dialectical variation. An alternative view would be to view the aḥruf as a divine concession allowing for some flexibility in the recitation of the Qurʾān - especially for those who find it difficult to recite or memorise the text. This view finds some support in the early Islamic tradition. Ibn Masʿūd, a Companion of the Prophet, is reported to have stated:
Even in the 2nd/8th and 3rd/9th centuries, we hear of a few scholars as renowned as Ibn Shihāb al-Zuhrī and al-Shāfiʿī justifying the transmission of hadiths and supplications in a non-verbatim manner based on the fact that permission was given to transmit the Qurʾān similarly. Thus, regardless of how we interpret the seven aḥruf hadiths, it seems quite likely that there may have existed some flexibility in the transmission of the text during the Prophet's lifetime and shortly after. Nonetheless, we should not exaggerate the extent of this flexibility. There are, in fact, good reasons to believe that the Uthmanic text faithfully preserves the structure and basic wording of the "original" revelations.
The lower text of the Sana'a Palimpsest provides us with the opportunity to better understand the variation that was existed in the codices of the companions. The manuscript has undergone radio-carbon dating several times, the results of which seem to suggest that it was produced before 650 CE. Even if the manuscript were written as late as 690 CE, hypothetically speaking, it could still represent an older tradition which pre-dates 650 CE. This possibility would seem to be supported by Elenore Cellard's observation that the scribes who produced the lower text alternated with one another even in the middle of a verse - a phenomenon which suggests that the manuscript was copied from an earlier written exemplar.
Unlike all other known Qur'anic manuscripts, the lower text represents a companion codex which differs from the Uthmanic text in two ways: (1) the wording of the verses and (2) the arrangement of chapters. Nonetheless, it agrees with the Uthmanic text as far as the structure of each chapter is concerned. A sample portion of the lower text containing Q. 9:123-9 is presented in the table below. Not only does the lower text preserve the same verse order, but its wording also closely corresponds to that of the Uthmanic text.
| Uthmanic Text | Sana'a Lower Text |
|---|---|
| يأيها الذين آمنوا قتلوا الذين يلونكم من الكفر وليجدوا فيكم غلظة واعلموا أن الله مع المتقين وإذا ما أنزلت سورة فمنهم من يقول أيكم زادته هذه إيمنا فأما الذين آمنوا فزادتهم إيمنا وهم يستبشرون وأما الذين في قلوبهم مرض فزادتهم رجسا إلى رجسهم وماتوا وهم فسقون أولا يرون أنهم يفتنون في كل عام مرة أو مرتين ثم لا يتوبون ولا هم يذكرون وإذا ما أنزلت سورة نظر بعضهم إلى بعض هل يرىكم من أحد ثم انصرفوا صرف الله قلوبهم بأنهم قوم لا يفقهون لقد جاكم رسول من أنفسكم عزيز عليه ما عنتم حريص عليكم بالمومنين روف رحيم فإن تولوا فقل حسبي الله الذي لا إله إلا هو عليه توكلت وهو رب العرش العظيم | يأيها الذين آمنوا قتلوا الذين يلونكم من الكفر وليجدوا فيكم غلظة واعلموا أن الله مع المتقين وإذا أنزلت سورة فمنهم من يقول أيكم زادته هذه إيمنا فأما الذين آمنوا فزادتهم إيمنا وهم يستبشرون وأما الذين في قلوبهم رجس فزدتهم رجزا إلى رجسهم وماتوا وهم كفرون أولا يرو أنهم يفتنون في كل عام مرة أو مرتين ثم لا يتوبون ولا يتذكرون وإذا أنزلت سورة نظر بعضهم إلى بعض هل يرىنا من أحد فانصرفوا فصرف الله قلوبهم ذلك بأنهم قوم لا يفقهون ولقد جاكم رسول منكم عزيز عليه ما عنتكم حريص عليكم بالمومنين روف رحيم فإن تولوا عنك فقل حسبي الله لا إله إلا هو عليه توكلت رب العرش العظيم |
The lower text, as well as other companion codices known only from literary sources, allow us to make some tentative conclusions regarding the early history of the Qurʾān. As these codices generally have the same passages within the surahs (with minor variations in wording), it is likely that the surahs were fixed during the Prophet's lifetime. Some caution is, of course, still necessary as only half of the 80 folios of the palimpsest have been edited so far. Nonetheless, these folios attest portions of a wide range of surahs such as Q2,5,8,9,15,16,19,20,22,24,33,62,63 and 89 (as well as some other surahs attested in very fragmentary folios). The companion codices also allow us to gain insight into the supposedly "missing" verses of the Qurʾān. According to some reports, Surat al-Ahzab used to equal Surah al-Baqarah (the longest chapter) in length while Surah al-Tawbah too was around 4 times longer than its present length. Several scholars have also noted that the Uthmanic text as well as the companion codices seem to arrange the surahs in order of decreasing length. The lower text of the Sana'a Palimpsest seems to adhere to this principle as well. As surat al
The codices prepared during ʿUthmān's reign contained very few diacritical marks and no vocalisation. Thus, at some places, they could be read in more than one way. A qirāʾāh (pl. qirāʾāt) refers to a specific way of vocalising the entire Uthmanic codex. Today, ten qirāʾāt are generally considered to be authentic. The variations amongst the ten qirāʾāt are of two types. The first concerns linguistic principles which are generally applied throughout the Qurʾān and have no effect on the meaning. The second concerns disagreements at specific points in the Qurʾān. Some examples of variants of the latter type are given below:
1. At Q1:4, four of the ten reciters read: maliki yawmi l-din while the remaining six reciters read: māliki yawmi l-din. Both of these readings fit the Uthmanic rasm which sometimes omitted the long ā vowel:
2. At Q18:26, one reciter (the Syrian Ibn ʿĀmir) read: wa-lā tushrik fī ḥukmihi aḥadan (And do not ascribe any partners in His judgement) while the other 9 readers read: wa-lā yushriku fī ḥukmihi aḥadan (And He does not share in His judgement with anyone). The two readings share the same rasm.
3. At Q71:23, two of the reciter (Nāfiʿ and Abū Jaʿfar) read : wa-lā tadharunna wuddan, while the remaining reciters read: wa-lā tadharunna waddan.
The ten reciters nonetheless recite the majority of the Qurʾān in the same way. This strongly suggests the presence of a shared oral tradition as there are several places where these reciters agree with one another despite the fact that an alternative reading is possible. In a recent article, Hythem Sidky has further developed this line of reasoning by examining ten such cases. He observes that the reciters' agreement cannot be the result of written transmission as early manuscripts are usually left undotted at these places - or if they do contain dots, they sometimes attest the alternative non-canonical variant! The fact that the ten reciters agreed with one another, therefore, is highly significant:
Manuscript examination also confirms that written transmission was not responsible for this common core. Therefore, this common core must have been transmitted by way of a living oral tradition parallel to the written text. In other words the reading traditions are not exclusively the result of independent attempts to decipher the undotted skeletal text of the Quran
How early can we date the shared oral tradition? As one of the canonical readers (Abū Jaʿfar) is reported to have already been teaching the Qurʾān prior to the battle of Harra in 63/683, Sidky dates the oral tradition to "no later than the middle of the first/seventh century". If one is more sceptical of such reports, however, then the latest possible date for the oral tradition would be the late 1st/7th century. It is important to emphasise that these are the latest possible dates. In my opinion, the shared oral tradition is actually a common memory of how the Prophet Muḥammad himself recited the Qurʾān. This shared memory of the Prophet's recitation may have been adapted slightly to fit the Uthmanic rasm.
To summarise, the ten readers agree with one another regarding the vocalisation of the majority of the Qurʾān. This agreement likely traces back to a shared oral tradition which dates at least to the mid 7th century (if not, earlier). But what about the disagreements amongst the reciters? Most Western scholars have suggested that these variations arose later in the 1st century, due to transmission errors, grammatical/exegetical concerns and/or attempting to read an ambiguous written text. According to the dominant Sunni traditional perspective, on the other hand, even these variations trace back to the Prophet Muḥammad. This is difficult to prove, however. Another issue concerns variants in the rasm itself. The codices which ʿUthmān dispatched to the provinces of the empire were not identical. There were roughly 40 or so minor variations between them, generally involving the addition/omission of a single letter. For example, at Q. 2:116, the Medinan and Syrian codices read: If we adopt this interpretation, the existence of variant readings may be justified with reference to the "divine permission model" proposed recently by Dr. Yasir Qadhi. That is to say, even if the Prophet himself did not recite these specific variants, they remain valid as they do not significantly alter the meaning of the verse.
In Q. 15:9, God states that He will preserve the Scripture that He has revealed. Classical scholars often interpreted this verse as a divine promise to protect the Qurʾān from manipulations by devils. This is certainly a plausible interpretation. In verse 6 of this surah, we learn that the Prophet's opponents accused him of being possessed by a jinn (majnūn) and ask him why he does not bring angels to them instead. Verse 9 may be a reassurance to the Prophet that the transmission of the scripture from God to His Prophet was secure. In other words, the central issue here was the process of revelation to the Prophet, not the transmission of the Qurʾān after the revelation. Of course, God's promise would extend to both periods but the point that I wish to make here is that the passage is concerned with malicious distortion of the text. It is not necessarily a promise to preserve every single letter and sound of the prophetic revelation.
In my opinion, this interpretation of the passage is consistent with the early transmission of the Qurʾān. There is no evidence that politically or theologically motivated changes were made to the Qurʾān after the Prophet's death. In fact, as we had seen earlier, the silence of the Qurʾān on issues that affected the early Muslim community after the Prophet's death is often taken as evidence that the text was closed at a very early date. While the latest possible date for this closure is the reign of ʿUthmān, the lower text of the Sana'a Palimpsest as well as other internal features of the Qurʾān allow us to go even further.
BnF Arabe 330f: The plene spelling of qāla is attested at Q. 12:89, 12:90, 34:31, while the plene spelling of qālu is attested at Q. 4:141 (twice).
© Mohammed Al-Firas