In my blogpost Is Thamud a triptote? I suggested that the name of the people ṯamūd was a triptote in the language of the QCT, rather than a diptote as it is in Classical Arabic. A triptotic inflection of this noun would be Nom. ṯamūdun <ṯmwd> Gen. ṯamūdin <ṯmwd> and Acc. ṯamūdan <ṯmwdʔ>. The diptotic inflection would be Nom. ṯamūdu <ṯmwd>; Gen. ṯamūda <ṯmwd> ; Acc. ṯamūda <ṯmwd>. Essentially the two forms can therefore only be consistently distinguished by the accusative orthographically. The word ṯamūd occurs in the Accusative 4 times in the Quran, while it occurs 12 times in the genitive (where, if it were diptote, the two should have been spelled identically but are note). If you want more detail on the arguments for a triptotic ṯamūd, please read the linked blogpost.
The fact that the orthography is perfectly consistent with the triptotic flection, leaves little to no reason to doubt that it was a triptote in the language of the QCT. This was my argument in my previous blogpost, but today I want to get into the reading traditions of the Quran. There are 7 authoritative reading traditions of the Quran, which mostly differ in minor points of pronunciation, and only occasionally read words, which are written ambiguously, differently.
In the Cairo Edition of the Quran, which is based on the Ḥafs reading tradition, ṯamūd is consistently treated as a diptote, as per Classical Arabic grammar. This is in conflict with the Quran's orthography, which has an extra <ʔ> on the accusative form of this noun. A sukūn is placed on top of this alif to 'acquiesce' it, so it will the word will not be read as a triptote.
The Ḥafs reading tradition is at odds with the language of the QCT, and, to my mind, cannot possibly directly reflect the original composition of the text in the way it was pronounced. Looking at this further, I wondered whether there were any other reading traditions that would read the word ṯamūd in the accusative as the expected ṯamūdan, and indeed, there are several traditions that do this:
The Nāfiʕ reading tradition as transmitted by Qālūn and Warš, for example, reads ʔinna ṯamūdan in Q11:68 rather than the ʕāsim tradition as transmitted by Ḥafs ʔinna ṯamūda. This is exactly what we would expect from the orthography. However, in the same verse ṯamūd is repeated, but this time in the genitive case, here we would expect li-ṯamūdin if the nouns was really treated like a triptote, but instead, even in these same reading traditions, in the genitive of the noun is vocalized as a diptote li-ṯamūda. This nouns, however is in pause, and pronounced li-ṯamūd, which would be consistent in either reading as a pausal pronunciation.
In Q11:73 however, the noun is once again in the genitive, and this time not in Pause, and again we find it read as ʔilā ṯamūda rather than the expected ʔilā ṯamūdin.
This present us with a problem: In a reading tradition that clearly treats the accusative as a triptote, we find that the genitive (and also the nominative) is treated as diptote. There are three imaginable solutions to this:
- ṯamūd was diptotic, but the person, or persons, who recorded the Quranic Consonantal Text were not aware of this, and recorded it as a triptote. i.e. Ḥafs' reading is correct.
- ṯamūd is a special class word, which is the only in its kind, which has a triptotic accusative and a diptotic nominative and genitive. i.e. Nāfiʕ's reading is correct.
- ṯamūd was triptotic, and the QCT reflects the correct composition.
The first scenario, seems to me, untenable. No other diptotes in the whole Quran are ever 'mistakenly' treated as triptotes, and the basis for this mistaken transmission seems to be extremely difficult to motivate.
The second interpretation is also very difficult to motivate. No such category is ever attested in Classical Arabic, and there seems to be absolutely no reason to think that such a 'unique' paradigm would develop, not just for some noun, but specifically for this proper name.
With the other scenario's dismissed as being essentially impossible, it seems to me that the third explanation must be taken as the most plausible.
But what does this mean for the reading traditions? The fact that the reading traditions reflect such a 'confused' reading, clearly suggests that they have been changed over the ages to adhere to a rather artificial grammatical model of Arabic and cannot possibly be the form in which the text was first pronounced by the prophet. While different reading traditions have come to different conclusions, it seems unlikely to me that such artificial readings were already in place when the Eponymous original readers, Nāfiʕ al-Madani and ʕāsim ibn Abī an-Nuǧūd of the respective reading traditions.
This ties into the bigger question of case endings in the language of the QCT. If case can clearly be artificially replaced to be 'correct' in the reading tradition, according to one, or the other grammatical insight, this places into doubt the whole question of whether the QCT was read with case vowels at all. The Arabic orthography of the QCT clearly represents a dialect where only the indefinite accusative of masculine nouns was receiving a case ending -ā.
While it is essentially possible that there indeed was a disconnect between the orthography, and the actual language of the QCT (much in the same was this must be the case for later Classical Arabic compositions), the fact that we can see that some of these unwritten case vowels have been changed at some later time, causes one to wonder whether maybe all of these final short case vowels are later artificial additions of a Classical Arabic framework being imposed over the original Quranic composition.
This raises a variety of questions: Why was a 'Classical Arabic' framework imposed over the original Quranic composition? Where did this Classical Arabic-like dialect come from? How big would this "linguistic conspiracy" have to be, to plausibly not show up at all in the traditional Muslim historical narrative? Are we not seeing the "conspiracy", because we have not looked at the historical texts with this question in mind, or has it really been covered up?
A more charitable stance towards the traditional historical narrative presented by Muslims, would allow us to assume that the readings, with all their Classical Arabic-like case vowels, are 'more correct' than the QCT orthography. Even if we assume this, we must still admit that the confused case of ṯamūd as presented here, is a later artificial intrusion. If the readings are 'correct' while the QCT somehow represents a much more advanced Arabic dialect that lost its final short vowels, one might imagine why ʕabdullāh ibn Masʕūd was critical of the canonized Quranic Consonantal Text commissioned by the third Caliph Uthman, which would be the reflection of a spoken dialect very far removed from the Arabic dialect in which the Quran had been revealed. 
This approach raises many questions: Why would the Caliph have allowed the Quran to be written down in such a vulgar reading? Was it an attempt to make the Quran more accessible to the 'lay man'? Why did this canonical text gain such traction, while the language of the popular readings remained squarely opposed to the orthography presented in the canonical text? Why was the dialect in which the QCT was written down so different from the real dialect of the Quran, while it must have been canonized mere decades after the death of the prophet?
Either scenario brings along a lot of enormous questions, which I don't think will be answered any time soon. But the fact remains, that the fundamental disconnect between the language of the reading traditions, and the language of the Quranic Consonantal Text needs to be explained in some way. I hope my previous articles on the disconnect between the QCT and the modern readings are felt, by my readers, also to be so fundamental as to require an explanation e.g. as the developments and loss of *ʔ, the diptosy of the feminine ending, the loss of nunation, retention of ē, (and also ō?) etc.
 As a sidenote, it should be noted here that there are two other diptotes which are, apparently treated as triptotes in the QCT. Both times, however, this seems to be motivated purely by the rhyme, where the final a was lengthened to ā to accommodate the rhyme. Q76:15-16 (it is for the rhyme in verse 15, 16 seems to simply copy the orthography of the word that was used just the verse above it). We find the same trhing happen a few times with the short a of the definite accusative, to accommodate the rhyme: Q33:10, 66-67; It seemingly also happens in Q33:4, although it is not marked in the QCT.
 Even such an interpretation is at odds with the history as narrated by the Muslims themselves. ʕāsim was said to not accept ibn Masʕūd's Quranic text, and exclusively based on the canonized Uthmanic Quran.