As I have essentially finished my series on the Quranic orthography for now, I thought it would be a fun test to look at some of the definitions of Middle Arabic defined by Joshua Blau in A Handbook of Early Middle Arabic (2002).
Middle Arabic, as a term, is in my opinion grossly underdefined, and badly applied to documents, which don't unambiguously point to Middle Arabic. Middle Arabic is a literary style that combines both forms of Classical Arabic and "Neo-Arabic". Neo-Arabic is a rather awkward catch-all term for precursors of the modern Arabic dialects, which certainly aren't monogenetic, which makes a term referring to all of them at once, at least, a little misleading.
Especially document from the period before the Classical Arabic grammatical tradition (that is before Sibawayh around 790AD), can hardly be considered Middle Arabic, simply because the very definition of Middle Arabic is dependent on the unambiguous existence of a diglossia between Classical Arabic and some vernacular dialect.
Blau (2002: 29-56) has a list of features typical for Middle Arabic, he marks some of these features as NA 'Neo-Arabic' while outhers are markes as PcA 'Post-Classical Arabic'. Many of the features he discusses are actually found in the Quran, making both the labble NA and PcA rather ill-suited, e.g.
1. PcA+NA "The elision of short vowels in open unstressed syllables". Examples given are tetġaṭṭā 'you will cover yoursef', cf. Cl. Ar. tataġaṭṭā.
This elision is well-attested in Quranic Arabic, and is, in fact, one of the features where Quranic Arabic clearly differs from canonical Classical Arabic, seen most clearly in the stem V verbs (tafaʕʕala) and stem VI verbs (tafāʕala) verbs, where the t prefix assimilates to the following dental/alveolar consonant. This assimilation could only have taken place if there had been a syncope of the vowel in the ta- prefix:
yaddabbarū <ydbrwʔ> 'they ponder', not yatadabbarū
yaṣṣaddaqū<yṣdqwʔ> 'they remit (as) charity', not yataṣaddaqū
iṯṯāqaltum<ʔṯʔqltm> 'you cling heavily', not taṯāqaltum
The assimilation seems to apply quite irregularly. It might be worth looking into the distribution one day.
2. PcA Sometimes la-'indeed' is spelled لا.
To my knowledge, this happens only once in the Quran, but that certainly qualifies as 'sometimes'.
la-ʔaḏbaḥanna-hū <lʔʔḏbḥnh> 'I will certainly slaughter him' (Q27:21).
3. NA ʔalif maqṣūra bi-ṣūrat al-yāʔ may be represented by ʔalif, e.g. ḥattā
Does not occur in the modern Quranic edition, but very common in the Kufic Quran of Samarkand, and the Sanaa Palimpsest.
4. Pca+NA Weakining of glottal stop, orthographic complete Elision of Hamza.
I discussed this at length in previousposts on the Quranic orthography. I think the evidence in favour of massive glottal stop loss in the language of the Quran is basically undeniable. Even words like marʔatun are written <mrh> perfectly following Middle-Arabic spellings of the word, rather than the classical orthographic practice that writes <mrʔh>.
5. NA tāʔ marbūṭa in construct is speltت
In fact quite common in the Quran. The examples cited by Blau are raḥmatu llāh and ḥayyātu mājidah. The former, is one of these phrases that is spelled with construct t, basically regardless of what register it is written in. While clearly a 'Neo-Arabic' feature, it seems obvious that it was a feature that was around already in Pre-Islamic Arabic times.
6. PcA The spelling of يا 'O' as يَ connected with the next word.
This is literally the only way the vocative particle yā is ever written in the Quran. Calling it a Post-Classical feature therefore seems rather silly.
7. PcA Contrary to the script classical orthography, all forms of the relative pronoun allaḏī may be spelt with one lām.
There are not many instances of the relative pronoun in the Quran that would require a double lām spelling, but the two that are found, do not have the double lām.
allaḏāni <ʔlʔḏʔn> and allaḏayni <ʔlʔḏyn>
8. NA Because of the weakening and elimination of glottal stop raʔā developed into را.
This is the regular Quranic Arabic form, I talked about its historical development recently.
So there you have it, 8 features common in Quranic Arabic, which are "Middle Arabic" or Post-Classical Arabic features. Many of the morphological features discussed by Blau, might also apply to the Quran, but simply cannot be evaluated because we don't have any certain vocalisation of the Quran.
8 features seem to be very little, but one needs to realise that some of these features are quite commonly used to argue that, especially early-Islamic Arabic (in my definition, from 622-750AD), is "Middle Arabic". This to me is highly anachronistic, and by this criterion as we see the Quran would become "Middle Arabic" too.
It seems more prudent to conclude that both Quranic Arabic and Early-Islamic Arabic deviate in the same way from the later established Classical Standard. This is of course not surprising, as the Quran was composed and written down in the Early-Islamic period, and therefore we would expect the document to share important orthographic and linguistic similarities to other linguistic documents of the time.