1. The Masoretic Text of the Hebrew Bible contains words and phrases that are written in one way (called the ketiv) but recited in a different way (called the qere). In Bible codices such differences between the ketiv and qere are generally indicated by providing a marginal note to indicate what the qere – the recitation – is. Out of reverence for the stable Masoretic Text, this text itself was not changed to accommodate the deviating reading tradition.
2. However, in the 10th and 11th centuries CE we find a striking evolution of Hebrew Bible productions among the Qaraites. They argued for the primacy of the reading tradition, and did not place the idiosyncratic written form of the text on a pedestal, even going as far as to completely abandon the Hebrew alphabet in some manuscripts. Instead, we find that Qaraite scribes developed a very precise system of transcribing the reading tradition of the Hebrew bible using the Arabic script in the nasḫ style, vocalized with the traditional Hebrew vowel signs.1
3. An example of such a Qaraite Bible can be seen in the picture below (fig. 1). It is an excerpt from the manuscript T-S Ar. 41.18, which transcribes Leviticus 25:10. The Arabic script is used to maximum effect by using plene vowel spelling for long vowels that are spelled defectively in the Masoretic Text. These defective spellings in the Masoretic Text renders some vowel lengths ambiguous, e.g. hā-ḥămiššīm spelled هاحمشيم (hāʾ–ʾalif–ḥāʾ–mīm–šīn–yāʾ–mīm) effectively representing the difference in vowel length of the two ḥīrēqs that the vowel signs in the Masoretic Text cannot distinguish. Note also the word هي (spelled hāʾ–yāʾ) which in the Masoretic text has the ketiv as הוא (spelled hē–vå̄v–ʾå̄lɛ̄p̄) which conflicts with the qere. This fragment spells the qere of this word, hī, phonetically.
4. Geoffrey Khan has noted that the 10th-century Qaraite author al-Qirqisānī gives a number of explicit statements concerning the primacy and infallibility of the reading tradition of the Bible, as opposed to the defective written Bible.2 Al-Qirqisānī affirms that there is “consensus of the entire nation” (ʾijmāʿ al-ʾummah) upon the reading tradition, whereas there is the deviant written text is not known in the same broad way. In their Arabic script Bible manuscripts, the Qaraites found a way to facilitate the accurate recitation of the Bible, abandoning the fixed consonantal skeleton in favour of a phonetically more accurate transcription that followed the reading tradition.
5. To anyone who knows the discourse around the preservation and transmission of the Quran, from medieval times up until today, this kind of argumentation is strikingly familiar. Like the Masoretic Text, the standard Quranic text – the so-called ʿUthmānic Text – is a highly stable text with archaic, and sometimes idiosyncratic spelling. A number of formal parallel reading traditions of this text existed, which do not always obviously match the fixed written text. Muslims, even today, argue for the primacy of the oral tradition, citing the uninterrupted chain back to the Prophet Muḥammad of these reading traditions as evidence for the infallibility and accuracy of the oral transmission – even if it is at odds with, or not elegantly represented by the written form of the text. By the 10th century, Ibn Mujāhid (d. 936) writes a formal description of the reading tradition of seven specific readers. These seven came to be considered ‘canonical’, and they are still considered to be that today. In the centuries that follow we see the emergence of a flurry of specialized works focused on the precise formal description of the reading traditions and the orthoepic tradition that belongs to it (tajwīd).
6. This specific focus on the oral transmission of the reading tradition as against the written form of the text is roughly contemporaneous with similar developments among the Qaraites, and Khan is right to draw attention to these parallels. There is another parallel in this ‘oral turn’ of the Qaraites and Muslims in the 10th and 11th centuries that is not generally acknowledged, however. Copyists of the Quran – at least in the Islamic east, i.e. Egypt eastwards – likewise started to abandon the strict adherence to the standard ʿUthmānic Consonantal Text, in favour of a spelling that more precisely represented the phonetic details of recitation, and to match Classical Arabic orthography. In order to demonstrate this phenomenon, I will compare Kufic Quranic manuscripts from the 8th and 9th centuries to the masterful copy of the Quran by Ibn al-Bawwāb (d. 1022 CE), written in the year 1000 CE (Dublin, Chester Beatty Library, Is. 1431).
7. The first notable thing we can see when comparing such manuscripts is that the Ibn al-Bawwāb no longer follows the highly archaising Kufic script, which barely uses any of the consonantal dots necessary for reading the Quran, and instead adopts a fully dotted and elegant nasḫ style, which closely parallels the non-Quranic writings of its time. This may be compared to the abandoning of the highly archaising Hebrew square script in favour of likewise the nasḫ style Arabic script by the Qaraites. Besides this, Ibn al-Bawwāb employs new orthoepic signs that are not attested in the Kufic manuscripts from the preceding centuries, such as a dedicated hamzah and maddah signs, which denote the glottal stop and the orthoepic pronunciation of overlong vowels respectively.
8. Besides this, Ibn al-Bawwāb changes the orthography to fit the standard Classical Arabic orthography. For example, long ā was consistently spelled without the ʾalif in plural active participles such as ṣādiqīna صدقين (ṣād–dāl–qāf–yāʾ–nūn). The Ibn al-Bawwāb Quran has full plene spelling of such words صادقين (ṣād–ʾalif–dāl–qāf–yāʾ–nūn).
9. But it is not just the spelling of vowels that were previously spelled defectively that are upgraded orthographically towards a more phonetic spelling. We also find that certain orthographic idiosyncrasies, such as the spelling of the worth bi-ʾāyāti “with the signs of”, which in the ʿUthmānic text was curiously written with two yāʾs باييت (bāʾ–ʾalif–yāʾ–yāʾ–tāʾ). Ibn al-Bawwāb steps away from this unusual orthographic practice, and simply spells this word in the Classical Arabic spelling بايات (bāʾ–ʾalif–yāʾ–ʾalif–tāʾ).
10. All of these interventions discussed so far serve to make the written form better reflect how the text is to be recited, much in line with the orthographic shift from the defective Masoretic Text towards the Arabic script transcriptions of the Karaites.
11. But we even find interventions to the spelling of words when we are confronted with the Quranic equivalent of a ketiv–qere situation: places where the reading tradition disagrees with the reading suggested by the ʿUthmānic text. While such cases are rare, Kufic manuscripts maintain the same Masoretic conservatism, and do not intervene in the consonantal text. They may sometimes add extra signs to the red-ink vocalisation later, but the black consonantal text is sacrosanct. Not so for Ibn al-Bawwāb. At this point it is important to note that the Ibn al-Bawwāb Quran is written in accordance with the canonical reading tradition of ʾAbū ʿAmr (d. 770 CE) in the transmission of al-Dūrī (d. 860 CE). ʾAbū ʿAmr has several variant readings which deviate from the ʿUthmānic text,3 and in these cases Ibn al-Bawwāb opts for an accurate representation of the reading tradition, rather than reproducing the ʿUthmānic text.
12. Two cases can be used for illustrative purposes. The first one is found at Q19:19, where ʾAbū ʿAmr read li-yahaba “so that he may give”, as opposed to the majority reading li-ʾahaba “so that I may give” which matches the spelling of the ʿUthmānic text لاهب (lām–ʾalif–hāʾ–bāʾ). Ibn al-Bawwāb intervenes and makes the text match the reading ليهب (lām–yāʾ–hāʾ–bāʾ).
13. Another case where ʾAbū ʿAmr deviates from the ʿUthmānic text is at Q63:10 where he read ʾakūna “I will become” in the subjunctive, while all other readers read ʾakun in the jussive which matched the spelling of the ʿUthmānic text اكن (ʾalif–kāf–nūn). Once again, Ibn al-Bawwāb changes the text to match the reading rather than the ʿUthmānic Text: اكون (ʾalif–kāf–wāw–nūn).
14. The Ibn al-Bawwāb Quran is in no way unique. From the 11th century onwards, Qurans in the Islamic east almost completely abandoned the ʿUthmānic Text in favor of spellings that better reflect the phonetics of the reading traditions. Geoffrey Khan first highlighted the similarity in argumentation for the primacy of the reading tradition used by both Qaraites and Muslim scholars in the 10th and 11th centuries. In this blog post I hope to have shown that, as with the Qaraites, this prioritization of the reading tradition over the written text led to extensive reforms in the written form of the Quran. Considering how these developments take place in the same time and place among the Qaraites and Muslims, it seems likely that this is the result of cultural cross-pollination.
- For an excellent introduction as to the various reasons why the Qaraites adopted this Arabic script transcription of the Bible see Geoffrey Khan, Karaite Bible Manuscripts From the Cairo Genizah (Cambridge: Cambridge University Press, 1990), pp. 20–21. See also idem, The Tiberian Pronunciation of Biblical Hebrew, 2 vols (Cambridge: Cambridge University Press, 2020). ↩︎
- Khan, Geoffrey, “Al-Qirqisānī’s Opinions concerning the Text of the Bible and Parallel Muslim Attitudes Towards the Text of the Qurʾān”, The Jewish Quarterly Review, 81/1–2 (1990), pp. 59-73. ↩︎
- Marijn van Putten, “When the Readers Break the Rules: Disagreement with the Consonantal Text in the Canonical Quranic Reading Traditions”, Dead Sea Discoveries, 29/3 (2022), pp. 438–62. ↩︎
Marijn van Putten is assistant professor at the Leiden University Centre for Linguistics and the Leiden University Institute for Area Studies. His research focuses on the linguistics, transmission and history of the Quranic text and the Quranic reading traditions. Besides this, he also researches the linguistic history of Arabic and Berber. He is currently the PI of the ERC Consolidator project QurCan: The Canonization of the Quranic Reading Traditions. Marijn runs his own blog at https://phoenixblog.typepad.com/blog/.
Suggested Citation: Marijn van Putten, “The ‘Oral Turn’ in Qaraite and Quranic Manuscripts in the 11th Century”, Munich Research Centre for Jewish-Arabic Cultures Blog, 22 January 2025, URL: https://www.jewisharabiccultures.fak12.uni-muenchen.de/the-oral-turn-in-qaraite-and-quranic-manuscripts-in-the-11th-century/. License: CC BY-NC 4.0.
