Superfluous variants in the readings of the Quran

By Avnar Sidiche

Related article: Dialogue variants in the canonical Qirāʾāt readings of the Qurʼān

This article looks at another class of canonical variants at the other end of the spectrum. One thing I found very noticable while looking through variants for the earlier work was the large number that do not convey any additional meaning (though exegetes sometimes try). Human poets may typically vary their recitations in such ways, though they also bring with them the usual cost of looking like human errors or changes creeping into the transmission. These are some of the common categories.

1. Plural-singular variants

A few examples are given below.

21:104 The Kufan readers (except Šuʿba’s transmission of ʿĀṣim) here read li-l-kutubi (for the books), as indeed did Ibn Masʿūd, reportedly. The others read li-l-kitābi (for the book). The same readers who use kitāb singular here readily use kutub plural in other verses.

Bridges translation Corpus Coranicum Tafsīr al-Jalālayn (they give interpretations of the verse, but none that explain any value of the variant).

59:14 Ibn Kaṯīr and Abū ʿAmr read ǧidārin (a wall singular), whereas the others read ǧudurin (walls plural). There is no reason to have both.

Bridges translation Corpus Coranicum Tafsīr al-Jalālayn

2. Active-passive

This kind of variant is also common, very often adding no additional meaning. Here are a couple of typical examples:

4:140 ʿĀṣim and Yaʿqūb read nazzala (It has come down), while the others read nuzzila (it has been sent down). The active is already implied by the passive.

Bridges translation Corpus Coranicum Tafsīr al-Jalālayn

23.115 Yaʿqūb and the related Kufan readings of Ḥamza, al-Kisāʾī and Khalaf read tarǧiʿūna (return), while the others read turǧaʿūna (be returned).

Bridges translation Corpus Coranicum Tafsīr al-Jalālayn

3. More-less intensive verb form

Also commonly, the more intensive/causitive verb form II is used, with a meaning that already implies the form I reading. One example will suffice:

21.96 Ibn ʿĀmir, Abū Jaʿfar, and Yaʿqūb read futtiḥat (has been opened wide; Arabic verb form II), whereas the others read futiḥat (has been opened). The first reading already implies the other. The same variants appear in different reader permutations in other verses in the context of gates opening (39:71 and 73, and 78:19).

Bridges translation Corpus Coranicum Tafsīr al-Jalālayn

4. Qul-Qāla

There are several verses where we have the variant reading qāla (He said) instead of qul (Say). There is no additional information from the qāla variant because it already follows from Muhammad’s recitation of the qul variant that, in so doing, “he said” it. One example will suffice:

72:20 ʿĀṣim, Ḥamza and Abū Jaʿfar read qul, whereas the others read qāla.

Bridges translation Corpus Coranicum (No comment on the variant in Tafsīr al-Jalālayn)

5. Wa, fa

The word wa (and) is sometimes added or omitted for no apparent reason, or instead is read as fa (and then) as in 91:15. An example of the former is:

2:116 Ibn ʿĀmir reads qālū (They say) whereas the others readers have wa qālū (And they say). Like the other types of variants in this article, we could well ask why we have the variance in some places but not other verses if they are not transmission errors.

Bridges translation Corpus Coranicum Tafsīr al-Jalālayn

This and some other cases of this type were regional variants (scribal errors) in the Uthmanic rasm, or consonantal skeleton of the original four codices sent to certain regions, as are (though this is controversial) some instances of Qul-Qāla.

Incidentally, any supposed deliberate purpose of these variants existing in the original Uthmanic codices encounters a number of difficulties. Not only are they very minor compared to the variants read by companions in certain places (so why bother with these?), but the rasm variants as reported by early scholars have been found to form a copying stemma, or tree (Michael Cook), confirmed in early manuscripts (Hythem Sidky), and were only gradually documented rather than there being any sign the Uthmanic committee were aware of the differences (Sidky). It is also notable that canonical readings supposedly must comply with the Uthmanic rasm or the regional variants thereof, yet at the same time the readings chosen as canonical got away with transgressing it sometimes anyway (such as 19:19 mentioned in my first article).

6. Variants overkill

40:26 The canonical readers give a total of four permutations of Pharaoh’s words here, both on whether he said or ʾau ʾan (or that) or wa-ʾan (and that) and on whether he said yuẓhira fi l-ʾarḍi l-fasāda (he will cause corruption in the land) or yaẓhara fi l-ʾarḍi l-fasādu (corruption may appear in the land).

Bridges translation Corpus Coranicum Tafsīr al-Jalālayn

A similar example occurs in Q19:25 where in the sentence “And shake toward you the trunk of the palm tree; it will drop upon you ripe, fresh dates”, Yaʿqūb reads “it” meaning the trunk, while the other readings have three other forms of the word where “it” refers to the palm tree, with or without consonant doubling to emphasise a great dropping of dates — why this strange overkill, not to mention several further non-canonical variants such as “We will cause to drop”? (Corpus Coranicum). What would be the point in deliberately reciting so many variants in such sentences?

Some theological questions that spring to mind

Though the Qurʼān itself (3:7) famously suggests that some verses are unclear / allegorical, it doesn’t mention the existence of reading variants. While the meaning is usually clear in each of these pairs of variants in this article, any benefit or reason to have both of them is not. What would be their divine purpose? These superfluous variants fit right into the general pattern of variants that look like they arose during transmission.

Similarly, why does each category arise only in certain places and not others? The same could be said for some other variants that do slightly change meaning, where in some places you have the pair of variants and others just one of them. This again naturally looks like changes arising unintentionally during transmission, or a sometimes loose approach to recitation by the Prophet.

Exegetes sometimes strive to layer on some meaning in these kinds of variances, yet no real difference is apparent from the Qurʼān itself. If there are canonical variants that bring no or negligible additional meaning, and at the same time come at the cost of looking very plausibly like human error or indifference, it is especially straightforward to explain them as being so, whether they arose from the Prophet or later transmission.