Thread by @PhDniX

I was asked both before and during the conference whether I would be posting my #IQSA18 talks on Twitter; With such enthusiasm why would I do anything but comply? Here is a thread.

The title of my talk speaks of a 'Quranic Archetype'; What I mean by this is that I think that all Qurans we have go back to a single written source -- a single book from which all Qurans were copied (with the notable exception the Sanaa Palimpsest).

So how do we go about showing that there was a single Archetype?
niʿmat aḷḷāh 'the grace of God' can be spelled in two different ways in Quranic orthography نعمت الله and نعمه الله.

There is no difference in meaning or context, just two ways of writing the same thing.

If we look in the Cairo Edition of the text we find that the distribution of these two spellings is essentially 50/50. 12x with hāʾ 11x with tāʾ.

This would suggest that its spelling is basically up to the whims of the scribe.

If the Qurans don't go back to a single written archetypal text, we would expect there to be no correlation between the spelling and the location it is found in across early Quranic manuscripts.

But we find the opposite: there's an almost perfect agreement across all text how the word should be spelled in each location. In this table we find the two spellings displayed. Columns are different manuscripts rows are locations in the Quran. Most lines have a perfect agreement

It is not just niʿmat aḷḷāh that has this strong correlation, there are literally thousands of very specific orthography practices and free variation on which there is perfect agreement across all early manuscripts. Hence: All Qurans go back to a single written text.

Some of these documents are so early that it's actually difficult to envision this single written text to be later than Uthman, so I think the evidence is consistent with the Muslim account of the standardization
[I have presented this argument before in other presentations here]

So is the Cairo Edition simply a perfect reproduction of the Uthmanic Archetype? It clearly got this Niʿmat aḷḷāh feature right; and it seems to get many other specific orthographic features right -- but does it get everything right?

The Cairo Edition is based not on early manuscripts but medieval *descriptions* of the consonantal texts of early Manuscripts, most notably Al-Dānī's muqniʿ.

If Al-Dānī get things right (which he does very often) the Cairo edition gets it right too. Sometimes Al-Dānī is wrong.

Let's look at a feature that is absent in the Cairo Edition but present in the Archetype first.

If we look at the spelling of fa-bi-ʾayyi in BnF Arabe 331, we find that it is not spelled as فباى as in the Cairo Edition, but as فبايى with two yāʾs.

Wetzstein II 1913 shows the exact same pattern.

As does BnF Arabe 333d, and in fact, any early manuscript one cares to examine. I have yet to find a single manuscript that follows the Cairo Edition as opposed to this spelling.

(This is part of a bigger trend of Al-Dānī saying double final yāʾ is spelled with a single yāʾ).

So to conclude: While the Cairo edition has فباى there can be no doubt that the original Quranic Archetype had فبايى.

There are also cases where the Cairo Edition has an unusual (non-Classical) spelling, where there is no early evidence for it in the manuscripts. Let's look at one of these.

In quran.com/40/50 duʿāʾu l-kāfirīna is surprisingly spelled دعوا الكفرين rather than the spelling that we might expect: دعا الكفرين. But if we look at Wetzstein II 1913, we just find the regular spelling.

The same is true for Saray Medina 1a.

And BnF Arabe 338a.

In Arabe 335 (which seems somewhat later than the rest) we see the first sign of the Cairo Edition spelling: But notice that it's a correction. It clearly originally read as دعا الكفرين but the first ʾalif was removed and replaced with a wāw, and a second ʾalif has been forced in

From this spelling we must once again conclude that the Cairo Editions دعوا الكفرين is not original to the Quranic Archetype and that rather the spelling دعا الكفرين was the spelling found there.

As there was an archetype, non-Classical features of the text can be examined in a new framework: "Weirdness" is not just the slip of the pen, but rather an integral part of the original text.

We find clear non-Classical features that can inform us about Quranic Arabic.

An oft-cited example of evidence for a Classical Arabic case system in the Quran are words that end in āʾ; When suffixed, the case vowel takes a 'carrier of the hamzah': wāw for ʾawliyāʾu-hum, yāʾ for ʾawliyāʾi-him. But is this really part of the archetype?

The answer is no: No matter where you look, ʾawliyāʾ + suffix is *never* inflected for case. the wāw and yāʾ are simply completely absent. Other words do have it, or sometimes have it, but this word is consistently uninflected in Quranic Arabic.

The Cairo Edition's rasm has been classicized in this regard. The Archetype only had اوليهم for all cases, whereas the Cairo Edition has اولياوهم، اولايهم، اولياهم.

This is discussed in "Case in the Quran" which I wrote together with Phillip W. Stokes
academia.edu/37481811/Case_…

Now for another feature: The Cairo Edition (like Classical Arabic) keeps final weak verbs and final hamza verbs perfectly distinct. Here we find naǧǧi-nā and nabbiʾ-nā, bot Stem II imperatives, one of a root nǧw the other of nbʾ. The hamzated verb has one denticle more.

But actually, if we look at early Quranic manuscripts, that denticle is consistently missing in all of them. Hence the verb is being treated like a final weak verb /nabbi-nā/ rather than a final hamzah verb /nabbiʾ-nā/.

The Cairo Edition, once again, has been Classicized. For the Archetype we should reconstruct نبنا not **نبينا.

The merger of these verb types is typical for the modern Arabic dialects as well as a pattern we see in early Islamic Arabic and seems related to the loss of the hamzah

There was a very early archetype. Copied from a written examplar to explain the consistent reproduction of orthographic features (no dictation!). We find clear deviations from Classical Arabic. We must compare across multiple early manuscripts to find these orthographic features.

Huge thanks to @ThomasMiloNL for giving me early access to his beautiful Naskh and Kufic fonts. The presentation has become significantly more beautiful because of it.