Thread by @PhDniX

This cool thread by @joumajnouna of Arabe 6430 discovered a couple of cool things, including a marker for aw/ay and a marker for long final -ū. Neither I had seen before.

She suggests the red ink marks recitation marks rather than true vocalisation.

https://twitter.com/joumajnouna/status/1126492629817073664

This seems to be correct. Black vocalisation marks the "base word" while red vocalisation mostly marks the pronunciation that is specific to its context. For example:
min miṯli-hī assimilated nūn to mīm in recitation yielding: mim-miṯlihī marked by red Šaddah.

The same thing happens with tanwīn

Here we have fī raybin mimmā where tanwīn is marked as assimilated to the followīng mīm in red: fī raybim-mimmā

Not only does it mark some recitation cues, it also marks variant readings.

Here black marks the reading ʾinnī (Q2:33: ʿĀṣim, Ḥamzah, Ḫalaf, al-Kisāʾī, Ibn ʿĀmir, Yaʿqūb) and red marks ʾinniya (Ibn Kaṯīr, Nāfiʿ, ʾAbū ʿAmr, ʾAbū Ǧaʿfar).

And here we find either a secondary reading or indeed a reading instruction. Red marks the "major assimilation" of ʾAbū ʿAmr:

qāla rabbuka > qār-rabbuka

Here yet another interesting assimilation:
Red puts a miniature yāʾ on top of
man yufsidu

This probably is not an attempt to render Ḫalaf ʿan Ḥamzah's may-yufsidu, but rather the more general assimilation to a nasalized /y/: maỹ-yufsidu.

ʿan-kum in proper tajwīd pronounces the nūn befor a kāf with a velar nasal [ŋ] (the ng sound in sing), a form of ʾiḫfāʾ. This red semi-circle above the nūn clearly marks this ʾiḫfāʾ: ʿaŋkum.

Incidentally much of what we've run into so far reminds one of ʾAbū ʿAmr's reading -- one that is said to have been exceptionally popular in the past -- and here is the Shibboleth of Abu Amr. He is the only one of the canonical readers that reads ʿalayhimi before alif al-waṣl.

An upward facing semi-circle is employed to write the imālah or taqlīl of, e.g. al-kēfirīna, ʾuḫrē, ad-dunyē, al-ʾabṣēri

We get so obsessed with early manuscripts that we never quite get to enjoy the really smart and interesting orthographic solutions that later manuscripts started employing to really precisely mark recitation.

This kind of stuff is mostly unexplored, and should be explored more!

ADDENDUM: The confusing thing here is that while ʿalay-himi before waṣl is THE Shibboleth of ʾAbū ʿAmr, the black reading ʾinnī at Q2:33 rather than ʾinniya does *not* belong to the canonical ʾAbū ʿAmr. So if this is ʾAbū ʿAmr, we're dealing with a non-canonical transmission.

While it should not surprise us to find non-canonical readings before the establishement of the canon by Ibn Mujāhid (d. 324 AH), this manuscript postdates that establishment by centuries! Yet here we are, with what looks like a non-canonical reading.

We need more people working on the readings or I need more time. We currently have absolutely no idea how large the percentage of Quranic manuscripts out there belongs to canonical readings, and which don't. But I keep on running into non-canonical readings all the time...