In the narrations that relate the establishment of the standard text of Uthman, usually an episode is mentioned where disagreement develops on the spelling of التابوت ‘chest, ark’. The version as related in Jāmiʿ al-Tirmiḏī[1] goes as follows:
فاختلفوا يومئذ في التابوت والتابوه فقال القرشيون التابوت وقال زيد التابوه. فرفع اختلافهم إلى عثمان فقال اكتبوه التابوت فإنه نزل بلسان قريش.
“[the committee that standardized the Quran, consisting of Zayd b. Thābit and three Qurashi men] disagreed upon [the spelling] of التابوت or التابوه. The Qurashis said that it is التابوت whereas Zaid said it was التابوه. So their disagreement was brought to Uthman, and he said: “write it as التابوت, for [the Quran] was revealed in the speech of the Quraysh [bi-lisāni Qurayš]”[2]
In modern print editions these words are usually vocalised as at-tābūtu and at-tābūhu. From a historical linguistic perspective a shift of ūt to ūh, in a non-pausal position is not very attractive. There are, to my knowledge, no reports of t shifting to h, except when it concerns the feminine ending, in which case we have evidence for a pausal shift from -at to -ah, but no evidence from the grammarians of -ūt to -ūh.
The interpretation of the hāʾ as a tāʾ marbūṭah is certainly not an option. It seems quite clear that the conflict here between Zayd and the Qurashis was a linguistic one, and not an orthographic one. Had both spellings concerned an identical pronunciation, there would not likely have been a conflict, and it certainly would not have been resolved by referring to the authority of the speech of the Quraysh. A tāʾ marbūṭah is moreover not attractive, because tāʾ marbūṭah is otherwise never used to write any other feminine ending but -at-, i.e. with a t that is preceded by a. Nouns that end in -ūt are invariably written with a tāʾ maftūḥah, e.g. عنكبوت ʿankabūtun‘spider’, الطاغوت aṭ-ṭāġūt‘false idols’, ملكوت malakūt ‘kingdom’.
Therefore, the linguistic reality that this conflict seems to be based on is currently unclear. To get further insight into this, we must first examine the etymology of التابوت. It has long been recognised that this word must derive from Aramaic tēḇūṯā‘chest, ark’[3] most likely through the intermediary of Classical Ethiopic tābot ‘ark of Noah, ark of the Covenant’[4] on account of the fact that both Arabic and Ethiopic unexpectedly have ā as an initial vowel rather than Arabic ay or Classical Ethiopic e that one maybe would have expected from its Aramaic origins.[5] Several Aramaic words that end in -ūṯ- and -ōṯ- are consistently borrowed into Ethiopic with -ot, but Classical Arabic displays two different strategies, such words are either borrowed with -ūtun or with the long feminine ending -ātun (-āh in pause). The table below gives an overview, and includes one Ethiopic loanword that ends in -ot which displays a similar strategy.
Aramaic | Ethiopic | Classical Arabic |
malḵūṯā ‘kingdom’ | malakot ‘Godhead’ | malakūtun ‘kingdom’[6] |
ṭāʿūṯā‘error; a spirit’ | ṭāʿot ‘idol, ungodliness’ | ṭāġūtun‘false idol’[7] |
ṣlōṯā ‘prayer’ | ṣalot ‘prayer, vow, intercession’ | ṣalātun ‘prayer’[8] |
zāḵūṯā ‘merit, benefit’[9] | – | zakātun ‘alms’[10] |
– | maskot ‘window’ | miškātun ‘niche’[11] |
I believe that these two competing resolutions of the Aramaic -ūṯ-/-ōṯ-, and Classical Ethiopic -ot can help us understand the linguistic nature of the conflict referred to in the account discussed above. ṣalātun and zakātun nouns that in Classical Arabic end in the long feminine ending -ātun are written with ـوة early Arabic orthography (including Quranic orthography), i.e. الصلوة, الزكوة. It is not just Aramaic loanwords that are spelled this way, but also native Arabic words occur with this spelling, such as النجوة an-naǧāti ‘refuge’ (Q40:41), الحيوة al-ḥayātu ‘life’ (passim), منوة manāta ‘the pre-Islamic goddess Manāt’ (Q53:20), بالغدوة bi-l-ġadāti ‘in the morning’ (Q6:52; Q18:28). These words were originally pronounced differently from the Classical Arabic pronunciation, namely with a rounded vowel -ōh, thus aṣ-ṣalōh, az-zakōh, al-ḥayōh, etc.[12]
Considering that we have evidence from the Quran that Aramaic -ūṯ-/-ōṯ and Ethiopic -ot are either borrowed with -ūt (malakūt, ṭāġūt) or -ōh (zakōh, miškōh) it now becomes possible to better understand the conflict between التابوت and التابوه, evidently these two strategies were applied differently by different Arabic speakers, and thus there were two competing pronunciations at-tābūt (التابوت) and at-tābōh (التبوة). These two options, naturally lead to the scribes having a difference of opinion as to its pronunciations and by extension its spelling. It must be the existence of these different borrowing strategies that lead to the conflict reported in these narrations that ultimately go back to al-Zuhrī, if not earlier.
[1] Tirmiḏī, Muḥammad b. ʿĪsā al-. Al-Jāmiʿ al-Kabīr. Edited by Baššār ʿAwwād Maʿrūf. Beirut: Dār al-Ġarb al-ʾIslāmī, 1996, vol. 5, p. 182.
[2] Harald Motzki has conducted an in-depth study of the narrations of the canonization of the Quranic text, showing that these narrations are well-established, and the common link is shown to be Ibn Šihāb al-Zuhrī (d. 124/741-2). See Motzki, Harald. “The Collection of the Qurʾān. A Reconsideration of Western Views in Light of Recent Methodological Developments.” Der Islam 78 (2001): 1–34.
[3] Sokoloff, Michael. A Dictionary of Jewish Babylonian Aramaic of the Talmudic and Geonic Periods. Baltimore & London: Johns Hopkins University Press, 2002, p. 1203; Sokoloff, Michael. A Dictionary of Jewish Palestinian Aramaic of the Byzantine Period. Ramat-Gan, Israel: Bar Ilan University Press, 1990, p. 580.
[4] Leslau, Wolf. Comparative Dictionary of Geʿez (Classical Ethiopic). Wiesbaden: Harrassowitz, 1987, p. 570.
[5] Nöldeke, Theodor. Neue Beiträge Zur Semitischen Sprachwissenschaft. Strassburg: Karl J. Trübner, 1910, p. 49.
[6] Sokoloff, Jewish Babylonian Aramaic, p. 681; Leslau, Geʿez, p. 344; Nöldeke, Neue Beiträge, p. 33.
[7] Sokoloff, Jewish Babylonian Aramaic, p. 509; Leslau, Geʿez, p. 584; Nöldeke, Neue Beiträge, p. 35.
[8] Sokoloff, Jewish Babylonian Aramaic, p. 964; Leslau, Geʿez, p. 557; Nöldeke, Neue Beiträge, p. 36.
[9] Sokoloff, Jewish Babylonian Aramaic, p. 412.
[10] Jeffery, Arthur. The Foreign Vocabulary of the Qurʾān. Leiden & Boston: Brill, 2007 [1938], p. 153.
[11] Leslau, Geʿez, p. 365; Nöldeke, Neue Beiträge, p. 51.
[12] Sībawayh reports such readings as being acceptable in the recitation of the Quran, stating explicitly that the people of the Hijaz would pronounce aṣ-ṣalāt, az-zakāt, and al-ḥayāt with an ʾalif at-tafḫīm“a backed ā”, i.e. something akin to ō. See Sībawayh, Abū Bišr ʿAmr. Al-Kitāb. Edited by ʿAbd al-Salām Muḥammad Hārūn. Cairo: Maktabat al-Ḫānajī, 1988, vol. 4, p. 432. For a discussion on the historical origins of these forms see Al-Jallad, Ahmad. “Was It Sūrat Al-Baqárah? Evidence for Antepenultimate Stress in the Quranic Consonantal Text and Its Relevance for صلوه Type Nouns.” Zeitschrift Der Deutschen Morgenländischen Gesellschaft 167, no. 1 (2017): 81–90 and Putten, Marijn van. “The Development of the Triphthongs in Quranic and Classical Arabic.” Arabian Epigraphic Notes 3 (2017): 47–74.