There are several letters of the prophet to several heads of state, which have been recorded in literary sources.
There are some documents out there, which are said to be the actual letters mentioned in these sources
Scholars rightly take these to be forgeries. Here's why:
First, in some orthographic aspects, it is *much* too modern. All throughout the early Islamic papyri there was only one spelling of salām:
سلم، السلم
NEVER سلام، السلام. 1. Tracing of the Munḏir 'letter' 2. 65 AH papyrus. 3. ~60 AH papyrus 4. CPP (first century Quran)
The shape of the rāʾ is wrong. In the early first century this is consistently a small semi-circle that ascends above and descends below the baseline. In these forgeries it has the modern shape. 1. Tracing of the Munḏir letter 2. PERF 558 (22 AH) 3. 60s AH papyrus 4. CPP
The final dāl is too 'Kufic'. Early manuscripts have much less broad dāls. The 'uptick' is also missing. 1. The munḏir letter. 2. 22 AH hāḏā 3. 42 AH ḏakara 4. 1st c. Quran muḥammad
The forger seems to be unaware of the fact that word-final kāf is different from word final dāl and writes it in the same "hyperkufic" manner. It should have an upward stroke in final position. 1. Munḏir ʾilayka 2. 60s AH [fa-]ḏālika 3. 25-30 AH ʿalayka 4. 1st c. Quran ʿalayka
And this one is funny: We would be required to assume the prophet spoke Fuṣḥā with a Turkish accent. He writes al-munḏir as المنزر!
He slips up again for allaḏī which he writes as الزى Oops! This can probably give us an idea where the forger was from.
A final reason to be skeptical about these forgeries is that they are *verbatim* the letters as we find in the literary sources. It is unlikely that the literary sources retained the letters (if they existed, and they may have) reproduced them down to the last letter.
So from this it should be clear that we *do not* have letters from the prophet. These are clearly modern forgeries. This does not mean that the letters mentioned in the literary sources are fake: They may have existed, but we only have those sources as proof of them.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
Now that my monograph that I've worked on for the past years is finally out (and free for anyone to download!: brill.com/view/title/615…). I thought it would be nice to do a series of threads, writing accessible summaries on what my book is actually about. Today Chapter 1! 🧵
So the main question my book sets out to answer is: "What is the language of the Quran?"
There's an easy but unhelpful answer: "the language that you find in the Quran."
But what kind of language is that? Obviously Arabic. What kind of Arabic? What were its linguistic features?
As I set out to answer this question, I became more and more amazed by the fact that, somehow, this question hardly ever has gotten asked in scholarship. And the answer has typically been: "It's Classical Arabic".
And then nobody went on to define "Classical Arabic".
One of the frustrating things about Afro-Asiatic is that, despite all the great morpheme comparisons, it is really hard to reconstruct vocabulary, at all.
I keep a list of cognates that I myself find compelling, between the two families I know best, Berber & Semitic.
Within the Islamic context, where adopting Islamic names is common, it is not unthinkable isəm is actually an early loan from Arabic. But indistinguishable from a cognate.
This root is of course one of the few really well-attested cognates in Afro-Asiatic. The Berber conjugation is highly irregular, no other verb behaves like it.
A beautiful classical example of assimilation of parallels in Q67:11 of Saray medina 1a.
The canonical text reads فاعترفوا بذنبهم fa-ʿtarafū bi-ḏambihim "So they acknowledge their sin", and that's what the manuscript currently reads, but clearly not what it always read! 🧵
First there is an unusually large gap between the ḏāl and the nūn, and you can see traces of removed text.
Moreover, the denticle of the nūn appears to have been added later (not quite as obvious, but obvious enough).
What the scribe obviously originally wrote is not بذنبهم ḏambihim in the singular, but rather بذنوبهم bi-ḏunūbihim "their sinS". That's not how any canonical readers recite it, nor have I found evidence for non-canonical readers of this kind. But it is an easy mistake to make.
A fascinating and, likely, extremely early rendering of Sūrat al-ʾIḫlāṣ, both remarkable for its not-quite-canonical wording AND its pre-Islamic spelling practices.
A thread on what information can be gleaned from it 🧵
The basmalah is unremarkable, but the first verse is different from from the canonical reading. Rather than:
qul huwa ḷḷāhu ʾaḥadun قول هو الله احد "He is Allah, the one" the text reads: الله لا احد, which, at first blush might look like it says: God, not one ?
Is this verse espousing an anti-monotheistic version of al-ʾIḫlāṣ? No. In pre-Islamic inscriptions, and occasionally in early Arabic manuscripts the asseverative particle la- before a word with a hamzah is, for some reason written with لا.
The Quran has a written form and recited forms. Its written form remained more or less unchanged. But the recited forms were sometimes at odds with what is written in the text.
A thread on what scribes did to alleviate these conflicts, in early Quranic manuscripts.🧵
Conflicts between the written and the recited should be familiar to those who know the Hebrew Bible, which shows a peculiar interplay between the standard written text (ktiv), and its recitation (qre) which are not infrequently at odds with one another.
Such differences are marked with marginal ktiv-qre notes. Notes that point out that the word written is to be recited differently.
In Josh 13:16 the written באדם "at Adam", has a ktiv-qre note in the margin to point out it should be read מאדם "from Adam".