Thread by @PhDniX

Leemhuis in his 1977 PhD thesis, later to be published as a book, conducted a study on the semantic distinction between the faʿʿala (D stem) and ʾafʿala (C stem) stems in Quranic Arabic. He takes the distinction between nazzala and ʾanzala 'to send down' as a case study. 🧵

He observes that for the morphological stem distinctions, the Sibawayh mentions that the distinction between the D and C (which Leemhuis calls H) stem can be related to the plurality of the object.
ʾaġlaqtu l-bāba "I closed the door"
ġallaqtu l-ʾabwāba "I closed the doors"

A bit later on that same page, Sibawayh informs us that ʾAbū ʿAmr used to make a distinction between nazzala and ʾanzala -- considering the placement it seems to imply: distinguished this type of distinction dependent on the plurality of the object.

The ʾAbū ʿAmr being referred to here is none other than the Basran canonical reader of the Quran. But Leemhuis puts the hypothesis to the test not using ʾAbū ʿAmr's own reading, but by using the "Egyptian vulgate", i.e. in the reading of Ḥafṣ ʿan ʿĀṣim!

He then concludes that there is in fact no relation between plurality of the object and the verb stem.

This is a fairly egregious case of #hafsonormativity especially because ʾAbū ʿAmr specifically differs from Ḥafṣ in this reading of these verb forms...

Whenever the consonantal text is ambiguous between these two verb forms (that is: whenever it appears in the prefix conjugation), ʾAbū ʿAmr reads the stem as -unzil- (where Ḥafṣ reads -unazzil-). With the exception of Q6:37 and Q15:21.

Ḥafṣ thus operates on a general undifferentiated rule: whenever the consonantal text is ambiguous, read -unazzil- It is therefore not much of a surprise that no clear pattern would emerge for Ḥafṣ (Leemhuis does not appear to have noticed this...).

But we should be testing the claim against ʾAbū ʿAmr, of course, because it is for him the claim of distinction is made! He however also has an almost generalized pattern, with two exceptions, so even finding the pattern claimed by Sibawayh would be a little surprising.

So let's have a look at the cases Leemhuis cites with plural objects:
D-stem
Q6:111 = fine
Q15:8 in AA's reading is neither D nor C-stem: But tD stem tanazzalu l-malāʾikatu. So should be excluded.
Q16:2 = yunzilu, so should be under C stems with plural objects.

Q25:25 = fine (note that Ibn Kathir has a non-passive form with rasm variant nunzilu here).
Q20:80 = fine
Q57:9 = yunzilu so should be under C stems
Q31:34 = yunzilu so should be under C stems
Q42:28 = yunzilu so should be under C stems
Q30:49 = yunzilu so should be under C stems

Q24:43 = yunzilu so should be under C stems

Those were all D-stems with with plural objects. As you can see only three of them are actually attested with a D-stem

No need to go to the C-stems since we know that all the ones he cites are going to be differentiated by the rasm.

We can know this, because Ḥafṣ does not have any cases of the C stem in the case of ambiguous rasm.

So ʾAbū ʿAmr the majority of plural objects are just marked with the C-stem. But some plural stems are marked with the D-stem.

But this is to be expected, as we saw above in the quote of Sibawayh ʾaġlaqtu l-ʾabwāba "I closed the doors" is fine. It is rather **ġallaqtu l-bāba that would have to be an issue. So we should rather look whether we find the D-stem in places where the object is singular.

Q2:23: nazzalnā, implied multitude of things.
Q2:97: nazzalahū, singular object.
Q2:176; Q3:3 nazzala [...] l-kitāba, singular object.
Q4:47: ambiguous, could be 'signs'.
Q4:136: nuzzila read in the passive, singular subject.
Q4:140: nuzzila read in the passive, singular subject.

As we see when the rasm is unambiguous the D-stem has singular object quite often too.

So what about those two cases where ʾAbū ʿAmr moves away from his general pattern of reading ambiguous rasms as C-stems? Is that where he made the distinction?

Not really. In Q6:37 the object is clearly singular, and the same is true for Q15:21. So the two times that ʾAbū ʿAmr decides to deviate from his general rule it is explicitly not to make a plural object distinction...

So for ʾAbū ʿAmr use of the D stem clearly does not correlate with plurality of the object. So what do we make of Sibawayh's statement that ʾAbū ʿAmr used to make a distinction between the two stems for that verb? Has something gone wrong in the transmission of his reading?

The earliest systematic recording of ʾAbū ʿAmr's reading that we have happens centuries after Sibawayh's time... But it would actually kind of surprise me. But the comment is difficult to reconcile with the reading we have.

Either way, Leemhuis' attempt to plumb the depths of the distinction between the D and C stems n-z-l run into many more problems because he failed to recognise that Ḥafṣ has a generalized rule for the prefix conjugation (and that others had the opposite form in all places).

Thus Leemhuis tries to show that because in Q24:43 lifeless mountains are 'sent down' uses the D stem whereas Q33:26 which involve people being *made* to go down uses the C stem, there is a difference between a factitive (make something X) and a causative (force someone to do X).

But this comparison falls flat, since Q24:43 the rasm is ambiguous, and thus Ḥafṣ reads with a generalized rule, and thus cannot tell us anything about the semantics, unless we assume by some stroke of bizarre luck any case of prefix conjugation in the quran is always factitive

And if we accept *that* already rather unlikely scenario, we would have to conclude that Ibn Kathir and ʾAbū ʿAmr didn't get the memo.

There are of course cases where the rasm isn't ambiguous and there Leemhuis' observations are valuable.

Though one must wonder about the reliability of the method, if it detects "ghost distinctions" like Q24:43/Q33:26.

In either case: let this be a lesson. Sometimes #hafsonormativity is not such a big deal, but at other times it will totally undermine your argument.

And *definitely* when your sources are explicitly citing a reading of the Quran other than the ubiquitous Ḥafṣ, it probably makes sense to judge its implications by that reader himself.

I'd like to think that we are now also slowly transitioning into a proper paradigm shift, where the "Egyptian vulgate" is taken as the default, and other qira'at as "weird stuff people sometimes do on the side", rather than equally authoritative forms of the Quran.