The miscanonisation of the Qurân

Aziz al-Azmeh offers a summary of one scholarly faction's understanding of the Canonisation of the Qur'ān. I think it's for an encyclopaedia. It starts well... but then it dumps this on its readers:

Canonisation is this process of literarisation, whose rapid cumulative emergence is reflected in the chronology of the text. This moved from rather indistinct references to sheets or tablets (SuHuf) sent to Abraham and Moses (Q 87:17, 19; 53:36; 20:133), followed by generic references to a book (Q 52:2; 50:4), followed in turn by the Book, clearly a full scripture, sent down to Moses (Q 46:12; 40:53; 29:27; 28:45, 25:35)—a generic book of phatic delivery, a notion that was to persist through Muslim history, along with other senses, after the Qur'ān came to be considered, exegetically and otherwise, as a canonical text (cf. Madigan, 52, 56).

Given the Arabic text's internal interrelations, we cannot rule out that a later sura collected to itself earlier language, through quotations, which it might not prefer for itself. Sura 20 is the example I'd single out here; and I must admit, I too thought sura 20 was one of the earliest, until 2005 or so when I started piecing together The Arabs and Their Qur'an (and I still think it's pre-40s / 660s). For sura 50 and sura 52, I'd say that they consciously affect earlier language of the primordial Text.

We are agreed upon suras 53 and 87 though. We are also agreed on suwar 25, 28, 29, 40, and 46. I would list these five in ascending order - because, by happy coincidence, I do (now) think that this was these five's order of composition.

But I would go further. Q. 10:94 is the example Azmeh wants for the book to Moses. Up to sura 3, the qurrâ understood "the kitâb" as still the scripture of Jews and Christians, with the current Revelation being a Divine corrective on that. It was sura 6 which presented the official collection of suwar for the Umma, as the new Book. Sura 25 reïterated sura 6's project and the rest, so to speak, is commentary.

All this may be known to anyone willing to sort out which quoted what, through the same analytic tools used in every other textual discipline ... excepting, to this day, this one.

Azmeh instead assumes such a prior order of composition as reserves sura 25 et al. for Mecca and postpones (especially) sura 3 for the Madîna. He cites Watt and Robinson, who IIRC rely upon Nöldeke if not upon classical Islam's sîra. This is what makes his summary incoherent.

Once Azmeh moves on to the canonisation of the Qurân over the 20s/640s and beyond, he relies on palaeographic-parallel and carbon-dating. I do not consider such evidence to be strong; and other scholars (not me) have questioned this, which Azmeh ignores. Also brought is later Islam's insistence that al-Hajjaj's recension was almost wholly orthographic again, without taking (say) de Prémare seriously.

Azmeh does not summarise a consensus; he does apologetic, and presents it as consensus. When he's not actively poisoning the well.