[thaumasiotes]

> Regarding the "αυ" sound: Same as all the original diphthongs such as "ου", "αι", "οι" etc. sounds, the "ι", "υ" etc were mostly supplemental/modifiers to the first vowel (also υ sounded more like e.g. modern German ü or indeed modern German y Ι guess).

> EDIT: I think the transformation of some diphthongs had already started by the time the Roman empire conquered Greece, so ta-oo might have been closer to the pronunciation at that time.

Vox Graeca says the opposite in regard to upsilon:

>> In both αυ and ευ the υ preserved its original quality as a back [u], i.e. it was not fronted to [ü] as elsewhere

https://archive.org/details/allen-vox-graeca-the-pronunciati... [page 76]

(The reason this statement isn't also applied to ου is not that the upsilon is fronted - it's that in ου, the upsilon is lost.)

Upsilon by itself began as [u] and developed into [y], but diphthongs ending in it didn't follow that development. Tau starts with the pronunciation /tau/ and stays that way.

[gschizas]

Technically, the υ in both αυ and ευ diphthongs were approximations of the w (waw) sound. Which is a sound between oo and w. The sound still exists in isolated dialects, e.g. Tsakonic dialect which is a descendant of Doric dialect.

Obviously saying it's stayed that way is wrong on its own, since it had converted to taf as early as 500 CE, in the branch that led to modern Greek. The branch that followed Latinization and Anglicization (much later) converted the unpronounceable waw sound to plain "oo"

[thaumasiotes]

> Technically, the υ in both αυ and ευ diphthongs were approximations of the w (waw) sound. Which is a sound between oo and w.

What? There is no sound "between" /u/ and /w/; they are the same sound. We call that sound /u/ in the nucleus of a syllable and /w/ in the onset.

> Obviously saying it's stayed that way is wrong on its own, since it had converted to taf as early as 500 CE

Early?? You were talking about ancient Greek. Are you calling 500 AD "ancient Greek"? At that point the ancient, classical, Hellenistic, and Roman periods have all concluded. You can call it koine. It's the end of koine.

[gschizas]

> What? There is no sound "between" /u/ and /w/;

Huh? "oo" is a very different sound than "w". One is a vowel, one is mostly/technically a consonant, I'd say mostly related to the modern γ/gamma sound! One is spoken with open mouth, for the other you put the back of your tongue on the roof of your mouth.

> Early?? You were talking about ancient Greek. Are you calling 500 AD "ancient Greek"? At that point the ancient, classical, Hellenistic, and Roman periods have all concluded. You can call it koine. It's the end of koine.

I think I might have confused CE with BCE. In my defense, I was writing on mobile. I mean 500 years *before* the birth of Christ. It was certainly taf in early medieval Greek of course, but my point was indeed that it was also taf in some Ancient Greek dialects.

I'm saying that the notion that "tau" was always and forever ta-oo (two syllables, open /u/ sound) is wrong. It might be right on the branch of pronunciation that was mixed with Latin and later English (and only AFTER it was converted for use from Latin speakers), but not in the branch that started with archaic Greek and led to modern Greek.

The fact of the matter is that there aren't two distinct languages, Ancient Greek and Modern Greek. The language evolved naturally and gradually (granted, with a very accelerated pace in the Hellinistic/Koine period). The distinction of Ancient and Modern Greek is a modern educational tool. There is a branch that led to "ta-oo" and "pie" (that's apparently only the English pronunciation), but since it's been converted to be used by Latin (and English) speakers, it's really a Latin adaptation of Greek.

[thaumasiotes]

> Huh? "oo" is a very different sound than "w". One is a vowel, one is mostly/technically a consonant, I'd say mostly related to the modern γ/gamma sound! One is spoken with open mouth, for the other you put the back of your tongue on the roof of your mouth.

Neither is spoken with open mouth. That would be a low vowel, something more like /a/. And neither is spoken with tongue contact to any other part of your mouth, such as the roof.

They're both spoken with the same tongue position and the same lip position, because they're the same sound. Compare wikipedia:

> The close back rounded vowel, or high back rounded vowel,[1] is a type of vowel sound used in many spoken languages. The symbol in the International Phonetic Alphabet that represents this sound is ⟨u⟩.

So much for /u/ - it's a vowel in which the tongue is pulled close to the roof of your mouth, and back to the rear of your mouth, while your lips are rounded.

( https://en.wikipedia.org/wiki/Close_back_rounded_vowel )

What about /w/?

> The most common labiovelar consonant is the voiced approximant [w]. It is normally a labialized velar, as is its vocalic equivalent [u]. (Labialization is called rounding in vowels, and a velar place is called back).

It's a consonant that is pronounced with the same tongue position as /u/. And the same lip rounding as /u/. And it's the consonant that is the equivalent of the vowel /u/. What makes them equivalent? Well, mostly it's the fact that there are no differences.

( https://en.wikipedia.org/wiki/Labialized_velar_consonant )

> I mean 500 years before the birth of Christ. It was certainly taf in early medieval Greek of course, but my point was indeed that it was also taf in some Ancient Greek dialects.

Well, from the same page of Vox Graeca:

>> at a later date, which cannot certainly be determined,³ the second element (which could alternatively be analysed as a semivowel /w/: cf. p. 5) developed a fricative pronunciation [v], so that in modern Greek the value of these digraphs is [av] and [ev]

>> ³ The Jewish catacombs at Rome still indicate a diphthongal value in the 2-3 c[entury] A.D.

Greek: A History of the Language and Its Speakers states in a footnote that there is evidence for frication of diphthong-final upsilon having "gone through" by 150 BC. The detailed evidence for this development is given as a habit in Egypt of misspelling αυ and ευ as αου and εου; it is inferred that the "ου" must represent a consonant or at least something a bit more consonantal than an ordinary vowel. It goes on to note that "spellings with β become increasingly common in late Roman and early Byzantine documents."

( https://archive.org/details/horrocks-2/page/168/mode/2up )

A date of 500 BC definitely cannot be supported. I also found a comment on reddit stating that this change still hasn't occurred in Pontic Greek today.

---

> "pie" (that's apparently only the English pronunciation)

Yes, that's related to the Great Vowel Shift, a specific English phenomenon. You're right about "pie" for π. You're way off on "tau" for τ; pretty much every language that isn't modern Greek has it as the equivalent of /tau/. Check out the list of translations for "the letter Τ / τ in the Greek alphabet" here: https://en.wiktionary.org/wiki/tau#Translations .