Following the Uthmanic attempt to constrain readings to a single rasm (c. 30/ 650), and Ibn Mujāhid’s (d. 324 / 936) selection of seven readings a few centuries later, the task of fully defining the canonical scope was yet incomplete.
5th-6th (11th-12th centuries) Stage 3 canonisation: Two canonical transmitters per reader
Each of the canonical readings had a large number of transmitters. Gradually, a system of two canonical transmitters per reader gained acceptance. This two-Rāwī canon for the seven readings, as Nasser explains in one of his papers, was popularised and de-facto canonised as a result of a simplified student Qirāʾāt manual by al-Dānī (d. 444), and later ryhming version of the same by al-Shāṭibī (d.590). As per my article compiling expert comments, they were selected from diverse groups of transmissions and all differ to a greater or lesser extent from one another (especially the transmissions of Hafs and Shu’ba from ‘Asim).
At the next level, many ṭuruq (transmission ways) proceed from these transmitters of the eponymous readers and Ibn al Jazarī chose two primary ṭuruq per canonical transmitter, the others now considered non-canonical (more on this below).
9th (15th) century Stage 4 canonisation: Ten readings canonised by Ibn al-Jazarī (d.833 / 1429)
Ibn al-Jazarī obtained a brief fatwa (translated in this article) from Taj ad-Din as-Subki that the seven and three more well known readings that had been popular since the time of the seven were all fully mutawātir, necessary knowledge and not to be questioned. He says that the ten are accepted everywhere (Munjid chapter 2) and he knows of no-one who denied them, even if some were unaware of the three after the seven (chapter 3). All had appeared in various books of readings of various sizes from an early time (most notably beginning with Ibn Mirham (d.381), who gave the same set of ten), though these had competed with smaller numbers of works ranging from 8 to 14 readings (Nasser’s Transmission of the variant readings of the Qurʼān , p.64 and Melchert’s paper, The relation of the ten readings to one another p.75). Ibn al Jazarī also limited the primary ṭuruq for each transmitter to two. Now the scope of the canonical Qurʼān had reached its final definition.
These three readings were those of ʾAbū Jaʿfar, Yaʿqūb, and Khalaf.
A significant observation is that, aside from Khalaf, who is known to have amalgamated the readings of two of his fellow Kufan readers (Ḥamza and al-Kisāʾī), the other two each have variants that were not found in Ibn Mujāhid’s seven. From a quick look, these were the first examples I found, and are typical of the kinds of noted in my superflous variants article:
- ʾAbū Jaʿfar verse 82:9 “No! But you deny the Recompense. ” becomes “No! But they deny the Recompense. ” Corpus Coranicum
- ʾAbū Jaʿfar verse 21:104 “The Day when We will fold the heaven …” becomes “The Day when will be folded the heavens…” (passive instead of active voice) Corpus Coranicum
- Yaʿqūb verse 19:25 “And shake toward you the trunk of the palm tree; it will drop upon you ripe, fresh dates” (“it” meaning the trunk — the other readings have three other forms of the word where “it” refers to the palm tree, with or without consonant doubling to emphasise a great dropping of dates — why this strange overkill, not to mention several further non-canonical variants?). Corpus Coranicum
Hence, five centuries were allowed to pass during which these three readings with their distinct variants were not definitively afforded the same status as the seven (eight centuries from the founding of Islam). This is not to say that everything outside of the canon was considered inauthentic, but rather it was a matter of defining what was to be treated as authentic. Nevertheless, these readings did suffer in the meantime.
Ibn al-Jazarī complained that when they heard any reading differing from the seven, the masses mistook them as shādhdh (anomalous) because of Ibn Mujāhid’s decision to choose just seven and its coincidence with the number of the seven ahruf (Nasser’s Transmission p.64). An unfortunate chain of events for divine revelation. Ibn al Jazarī felt it necessary to argue extensively (and citing earlier major scholars) to establish that the three were on a par with the seven. While the ten were the most accepted set even before Ibn al-Jazarī, some scholars after Ibn Mujāhid (who himself regarded Yaʿqūb’s reading as inferior to Abū ʿAmr’s in Basra — Nasser’s Transmission pp.63–64) considered any beyond his seven readings as shādhdh in terms of level of transmission, or expressly placed the three at a lower transmission level than the seven (Ahmad ‘Ali Al-Imam’s Variant readings of the Qur’an pp.126-133).
Even now one finds statements that the three are mashhūr (widely known) and a level below the mutawātir seven. As recently as the 20th century, Arthur Jeffrey remarked that “Islamic scholarship is still divided over the question as to whether seven only or all ten are canonical” (The Old Codices, p. 2 footnote 1).