How would Marcus Aurelius have pronounced his Latin?

Question

It is my understanding that Julius Caesar, Cicero, Octavian (Augustus) would have pronounced Latin in a manner that is decidedly Classical, characterised by:

• "v" as /w/
• "c" and "g" always hard (i.e., /k/ and /g/)
• both long and short vowels
• "ae" diphthong as /ae̯/
• consonantal "i" as /j/

(All these points distinguishing it from Vulgar and Ecclesiastical Latin pronunciation.)

I also understand that Vulgar Latin pronunciation started to develop around the first century BC as well, but took a long time to become the standard for the upper classes. I have heard it was standard even among the patricians and senators by the early 5th century, but quite possibly earlier.

My question relates to Marcus Aurelius (121–180), and how he and others of the higher echelons of Roman society would have spoken Latin. Of course, Marcus Aurelius grew up in Rome, so he would most likely have had a very orthodox pronunciation/accent for the time, despite his provincial heritage (from southern Hispania). Would his pronunciation likely have been totally Classical, like that of Caesar or Augustus, or would it already be showing signs of "vulgarisation"? How late did this "correct" Classical pronunciation persist amongst the highest classes during the Imperial period?

(Incidentally, I presume that Marcus Aurelius' usage of grammar and vocabulary – and probably spelling too – would all have been very much Classical/Ciceronian in any case.)

Answer 1

For the most part, the upper classes in Rome still spoke Classical Latin in the 2nd century AD.

Features in common with Classical Latin

• c - hard, as /k/. The softening came much later.
• g - hard, as /g/ (although whether it was still /ŋ/ before "n", or had become a simple /g/ there too by this time, is unclear).
• h - the educated élite probably still pronounced it, at least in careful speech.
• i - still /j/ as a consonant.

Likely change: consonantal u

• It's likely that by this point, consonantal "u" was no longer /w/ and was realised either as [v] or more likely [β].

Likely change: vowels

• By sometime in the 2nd c. AD Velius Longus observed a qualitative difference between long and short "i". This suggests that Latin short "ĭ" and long "ē" had already started their convergence towards Romance [ẹ], while "ī" remained [i].

Possible change: diphthongs

• ae - unclear whether it was still /ae/ or had become /e/ at this point.
• au - would have become /a/ in unstressed syllables, but might still have been /au/ in stressed syllables.

The above observations are deduced or inferred from Allen, Vox Latina.