Thread by @PhDniX

New Open Access Article: "Hišām's ʾIbrāhām: Evidence for a Canonical Quranic Reading Based on the Rasm."

This paper brings together different parts of study that I've been focusing on (Qirāʾāt, Manuscripts), happy it's finally out!

A short summary thread
doi.org/10.1017/S13561…

While most Quranic readers are in agreement that the pronunciation of the prophet Abraham is ʾIbrāhīm, the two transmitters of the Syrian reader Ibn ʿĀmir -- Hišām and Ibn Ḏakwān -- use a second pronunciation besides ʾIbrāhīm, namely, ʾIbrāhām. Ibn Ḏakwān reads it as such in Q2

This makes intuitive sense if one looks at the Cairo Quran. The normal spelling we find is ابرهيم throughout the Quran, except in Q2 where it is spelled ابرهم which is indeed more naturally read as ʾIbrāhām (other readers read a "virtual yāʾ" there.
e.g. quran.com/2/124

Hišām, however, reads ʾIbrāhām with a much more haphazard distribution. He reads ʾIbrāhām in 33 distinct places. Including all of Surah 2, 14, 16, 19, 42, 51, 53 and 57. In other surahs both pronunciations occur. In quran.com/60/4 both occur in a single verse!

So what is going on? Well it is well-known that in early Quranic manuscripts the spelling that denotes ʾIbrāhām, i.e. ابرهم is spread fairly haphazard throughout the Quran as well. So I wondered: is there any relationship between Hišām's reading and manuscripts spelling?

And sure enough, there is! While manuscripts are not always perfect matches, there is a clear correlation between where Hišām reads ʾIbrāhām and the manuscripts spell ابرهم.

So how do we make sense of this correlation?

Yasin Dutton already noticed this correlation for Or. 2165 and the Codex Parisino-Petropolitanus (CPP). Both of these are Syrian in terms of their rasm variants. In this Dutton saw confirmation that these manuscripts were "written in the reading of Ibn ʿĀmir."

There is a chronological problem here: Ibn ʿĀmir lived a bit late for the age that we believe these manuscripts to be (but you can make it work).
There is also a practical problem: early manuscripts are not written "in" a reading. If that were the case we'd lose archetypal signal

Early manuscripts retain minute orthographic differences which are faithfully transmitted from one manuscript to the next. The transmission of the rasm of the Quran was a purely written one. On this see my (Open Access) "Grace of God" article: doi.org/10.1017/S00419…

Those objections aside though -- both of which could be challenged in a variety of ways -- the prediction Dutton's explanation makes is that all manuscripts that have the correlation of ʾIbrāhām to ابرهم should be of the Syrian manuscript tradition.

As I discuss in this other thread, there are about 40 rasm variants that allow us to identify from which regional master copy a manuscript descends. Both Or. 2165 and the CPP descend from the Syrian type.

I examined seven additional mss for region.

https://twitter.com/PhDniX/status/1218669152371650560

As it turns out not all manuscripts that have the ʾIbrāhām to ابرهم correlation are of Syrian descent. In fact, each region is represented in my sample. As such the spelling is *not* a reflection of the Syrian manuscripts and the Syrian reading.

How can we understand that Hišām's reading and the spellings correlate so closely? The only explanation is that the rasm wasn't changed to accommodate his reading; he based his reading on the rasm. If the copy he recited from had ابرهيم he read ʾIbrāhīm; with ابرهم , ʾIbrāhām.

Hišām's reading tradition is clearly dependent on the written form of the text, and he shaped his reading towards it. This reveals an intimate relationship between the written and the oral Quran in the early Islamic period, much in the same way as with the Hebrew bible.

[END]
Two small typos that I didn't catch in the proofs that you can chose to edit in your own copy if you like.

Pg. 17: (B1) should've read لله for the non-Basran mushafs, not الله
Pg. 19: W's (S1) should've read MBK, not just BK.