Thread by @PhDniX

New Article!

This article examines a famous passage in the Hadith that related the canonization of the Quran, where the Uthmanic committee has a disagreement on how to write the word for "Ark".

Insight into loan strategies elucidates the passage.

doi.org/10.1515/islam-…

In the Quran today the Ark of the Covenant is spelled التابوت and pronounced al-tābūt. This is a loanword from the Aramaic tēḇōṯ-ā, likely via Gəʿəz tābōt.

However, reports (which go back to Ibn Šihāb al-Zuhrī (d. 124/741-2)) tell us there was a controversy on how to spell it.

The Medinan Zayd b. Ṯābit wanted to spell it with a final hāʾ: التابوه, while his Quraši colleagues insisted it should be spelled التابوت.

They take it up with ʿUṯmān who says: the Quran was revealed in the Quraysh dialect, so it should be written according to it.

As a result, we now have the spelling التابوت.

But what exactly is going on here? If it was purely an issue of spelling, then "write it according to the dialect of Quraysh" doesn't make sense as an instruction. A spelling doesn't have a "dialect".

Thus the controversy presumably involved an actual dialectal difference. But what was it?

Already early on (e.g. al-Farrāʾ d. 209/822) the report was understood as referring to a pronunciation at-tābūhu, with a final hāʾ instead of at-tābūtu.

This is, however, not easy to make sense of linguistically. We have no evidence of stem-final -t to turn into -h in positions like that, and no other words show the same kind of controversy. Considering the Gəʿəz origin with -it we also don't expect this shift through borrowing.

So I take it from another angle. I look at how other loanwords in the Quran from Aramaic or Gəʿəz which end in word final -ūt or -ōt are borrowed. Here we find two distinct practices:

Either borrowed with -ūt (malakūt, aṭ-ṭāġūt), or with the long feminine ending -āh (miškāh).

As you can see, this feminine ending -āh, in all these words is spelled with ـوة rather than the later Classical spelling ـاة. This spelling has been suggested (by other and myself) to have originally indicated the pronunciation -ōh in pause and -ōt- in context.

Both -ūt- and -ōt-/-ōh# are excellent approximations within the phonology of Quranic Arabic to approximate the Aramaic -ūt/-ōt and especially the Gəʿəz -ōt. From the other loanwords we see that there were two strategies.

It therefore seems that in the process of borrowing tābōt from Gəʿəz, the Medinans (like Zayd b. Ṯābit) chose one strategy (like الصلوة, الزكوة, مشكوة) whereas Quraysh chose the other (like ملكوت, طاغوت).

Already by the late second century this was no longer understood by the specialists of language, which means that it was no longer transparent by that time, and phonetic memory of it was lost. This lack of transparency speaks to the archaicity of this detail in the report.

For some reason the article was not published Open Access, even though Leiden has an agreement with De Gruyter. I'll try to get that fixed ASAP. But as always, you can send me an e-mail and I'll happily share the PDF!