Thread by @PhDniX

The consonantal dotting of the Arabic script is something about which quite a bit of confusion exists, from its origins, application in manuscripts, quranic or otherwise and finally anachronistic imposition of modern norms onto medieval manuscripts, a thread on Arabic dotting 🧵

It is often proclaimed that the consonantal dots were entirely absent in the Quran of Uthman, and some even say they have been invented some time after his lifetime. Neither of these statements are in line with the material evidence.

The earliest Arabic papyrus we have (PERF 558, dated 22 AH/643 CE) already has clear (albeit sparse) consonantal dotting, predating the reign of Uthman, so certainly predating the Uthman's standard text of the Quran.

There is nothing to suggest the practice wasn't pre-islamic.

It is often suggested that Uthman's text was undotted on purpose to allow different readings. There is no material evidence to corroborate this claim. Every early Quranic manuscript at least has some amount of consonantal dotting (see @Adam_Bursi's paper)
academia.edu/40653150/Conne…

It is in fact rather the later monumental kufic manuscripts that become extremely sparse in terms of dotting; in some cases they even seem to be entirely devoid of dots. An example of a fairly late manuscript that is seemingly entirely devoid of dots see:

https://twitter.com/CellardEleonore/status/1280840140622049280/

Now when it comes to dots themselves there is actually some amount of historical depth to them as well. Not all dotting practices seem to stem from the same time, and certain practices shift over time. Let's look at the consonantal dotting:

First, dotting was clearly invented to disambiguate letters that are otherwise homographous. This is in part due to the inheritance from the Aramaic alphabet (from which the Arabic script descends) which lacked signs for all the sounds of Arabic.

We can think of the Arabic letters in four categories:
1. Those that have a single sign and lack dots for disambiguation
2. Those that are ambiguous in all positions
3. Those that are ambiguous in final position

1. The letters that are always unambiguous (and undotted) are ʾalif, kāf, lām, mīm and wāw.

2. Those that are ambiguous in all positions without dots are: dāl/ḏāl rāʾ/zāy, sīn/šīn, ṣād/ḍād, ṭāʾ/ẓāʾ, ʿayn/ġayn.
All of these are distinguishes with a dot above (except šīn), and stem from an original deficiency of the Aramaic script (Aramaic has no ḍād or ẓāʾ sound)

The distinction between ḥāʾ/ḫāʾ is ambiguous in all positions and likewise distinguished with a dot above.
In non-final position (and later in all positions) they share a shape with jīm.

Likewise Bāʾ, tāʾ and ṯāʾ, who in non-finally share ambiguity with yāʾ and nūn too!

3. Final Nūn and yāʾ are perfectly distinct in shape from one another, but share a shape in non-final position with bāʾ/tāʾ/ṯāʾ

fāʾ and qāf are distinct word-finally, but in non-final position they are identical.

The fact that fāʾ and qāf are distinct in word-final position frequently surprises people, as in modern hand-writing they tend to be *very* close, But the further we go back in time the more distinct they are. Very obvious in Kufic manuscripts!

The fāʾ and qāf are somewhat of an unusual exception because because their dotting does not seem to have been fixed early on.

The two dots on top of the qāf seems to be a late invention. Early manuscripts mark these with dot above and below, but which one gets which is not fixed

So, for example, Or. 298 (dated 252 AH) has a dot above for fāʾ but dot below for qāf (and sometimes two dots above for qāf!)

yaqūlu & qad & fī

In the Maghreb the reverse practice of dot below for fāʾ and dot above for qāf remained prevalent

To my knowledge there are no examples in Hijazi or Kufi manuscripts where two dots above for qāf is original to the manuscript, always added by a later corrector, which clearly suggests that the two dot qāf solution only comes to be several centuries into the Islamic period.

As for the ǧīm: while in modern script, and (to my knowledge) all non-Quranic manuscripts its shape is merged with ḥā/ḫāʾ, I have shown that several early Quranic manuscripts still keep ǧīm distinct from the ḥāʾ/ḫāʾ. This continues pre-Islamic practice
academia.edu/40934697/A_New…

The hāʾ shape is used for both the hāʾ, and in word-final position the feminine ending also known as the tāʾ marbūṭah, which carries the two dots of the tāʾ to denote its pronunciation as a /t/ sound when in non-pausal position. This practice is much later than other dotting.

To my knowledge there is not a single Hijazi or Kufic manuscript that ever uses the two dots on top of the tāʾ marbūṭah. It is simply spelled identically to the hāʾ, and the difference must be derived from the context.

Nobody seems to have researched when it comes to be used, but it strikes me as a very worthwhile palaeographical research indeed. One thing is certain: it was not yet invented at the time of the earliest quranic manuscripts, but it existed by the year 1000...

al-ḍalālah

In modern orthography, if the yāʾ carries either a hamzah, or denotes the ʾalif maqṣūrah then it does not carry dots. This is a fairly recent innovation. It might predate the widespread use of printed Arabic, but not by much:

yastahziʾū & al-hudā in Ibn al-Bawwāb Quran

Dotted yāʾ is often taken as evidence for the absence of hamzah but ample examples of yāʾs with dot AND hamzah written explicitly exposes that this is wrong.

Expecting seats of the hamzah to be dotless is simply imposing print orthography onto medieval manuscripts: very wrong!

Likewise dotted ʾalif maqṣūrah is often taken as evidence of ʾimālah by researchers (as in modern standard arabic orthography dotless yāʾ is used for that) this is likewise anachronisitc. Vocalised manuscripts often leave no doubt that dotted ʾalif maqṣūrāt lack ʾimālah.

These last two really get under my skin (and those who have followed me a while have certainly seen me rant about it before). You cannot make linguistic arguments based on the assumption that what you learn in your schoolbook has always been true! Arabic (script) has history too!

If you enjoyed this thread and want me to do more of it, please consider buying me a coffee.
ko-fi.com/phdnix.
If you want to support me in a more integral way, you can become a patron on Patreon!
patreon.com/PhDniX