Eta had the value of a long mid front vowel (long "e"). German "See" is very close, but for English, the closest approximation is probably "say", which is a diphthong. This refers to Ancient classical pronunciation, before the Hellenistic period.

It is definitely not a hight front vowel ("see"), which is the value it had in Koine and has in modern Greek.

The pronunciation used in schools for didactic purposes has only a modest relationship to authentic historical pronunciation. Pronouncing eta as [i] is an influence from modern Greek - in ancient times it was [ɛː], as described by the author (and as used in my secondary school Ancient Greek class - when we switched teacher to one who pronounced it [i], he conformed to our standard rather than vice versa). The bLogicarian has an article [1] which goes into more detail with some recordings, or you can pick up a copy of Vox Graeca by W. Sidney Allen.

[1]: https://blogicarian.blogspot.com/2012/10/ancient-greek-food-...

I wonder how close it is to how the ancient Greek pronounced it. Would our Greek scholars sound like non-native speakers to the ancient Greek? Or maybe barely even intelligible?

With extremely strong accents and sometimes probably hardly intelligible, depending on which time and place they land in. One primary problem is that there was no "Ancient Greek" pronunciation. Greece was not a single country back then but consisted of many different regions and kingdoms, who frequently went to war with each other (which I always found rather ridiculous. Compare the Kingdom of Rhode Island going to war with the City State of New York).

So Greek was what is called a pluricentric language and had IIRC at least five major dialects (Attic, Ionic, East Greek, West Greek etc.). Of course these changed significantly over the centuries, too.

There is obviously a whole field of research into Ancient Greek phonology and they have various points of reference to determine possible pronunciation. There are also a number of surviving texts that do discuss pronunciation of some words.

However, to me the most interesting tidbit was always that they found out how to pronounce the letter "η" (which in modern Greek is similar to "ee" like "meet") when a fragment of an ancient comedy by Aristophanes was discovered in which a guy is about to be sacrificed by a priest (I think?) and the priest tells him to behave like a sheep (as a kind of proverbial "sacrificial lamb" I suppose) and imitate sheep noises (difficult to put in English but something like "baa").

This was rendered as "βῆ βῆ" in the fragment and because Greek sheep have not changed their accent over the millennia, we were able to get some insight in to how that must have sounded. ;-)

> because Greek sheep have not changed their accent over the millennia

I find that really hard to believe. Surely animal sound onomatopoeia changed over 3000 years?

There were quite a lot of differences for sure, and many more modifiers on vowels than modern Greek, but I never heard that particular example. That particular modifier doesn't make it sound like that especially.

Think of it similar to Latin where real pronunciation was quite different to how you hear it quoted in media today, e.g. Veni, Vidi Vici sounding more like Weni, Widi, Wiki.

> real pronunciation was quite different to how you hear it quoted in media today, e.g. Veni, Vidi Vici sounding more like Weni, Widi, Wiki.

What you call "real pronunciation" is the pronunciation in the Classical dialect of Latin[0].

What you hear pronounced most often is the Liturgical[1] dialect, which is certainly how I learned to sing Latin in choir in college (a secular school) and I'm pretty sure the Liturgical dialect is what's most commonly taught in Catholic schools. (I'm not Catholic and I didn't attend a Catholic school, but that's my understanding.)

Given the scientific revolution's roots in the Renaissance[2], I presume many scientists use pronunciation somewhere between Liturgical and Classical Latin.

[0] https://en.wikipedia.org/wiki/Classical_Latin

[1] https://en.wikipedia.org/wiki/Ecclesiastical_Latin

[2] https://en.wikipedia.org/wiki/Latin#Renaissance_Latin

> Veni, Vidi Vici sounding more like Weni, Widi, Wiki.

That's Liturgical vs. Classical dialects of Latin.

I don't know where this idea that the "ci" group is pronounced "ki" came from (or "cae" / "ce" is pronounced "ke"). Given that both Italian and Romanian preserved them as "tchi" / "tche" (so Caesar is "Tche-zar", not "Ke-zar" as I read somewhere), I have some strong doubts about it.

Much like the majority of the Latin-speaking world, Italian underwent a sound change where [k] softened before a front vowel. The original [k] can be found in Sardinian and the now-extinct African Romance.

Romanian adopted the Latin alphabet in the 19th century and the spelling was intentionally based on Italian. So it's not a good representative for this particular argument — of course it spells it this way, because it was intentionally chosen to spell it this way by importing Italian spelling rules for these sounds.

There's some idea of ancient greek pronunciation based on phonetic misspellings found scattered about. I studied /some/ ancient greek in undergrad, and when I visited Greece it did, indeed, seen like a while different language, with large shifts in both pronunciation and meaning of words. My favorite was agathos, which is noble and good in ancient greek, but means native or simple minded in modern greek.

Also worth keeping in mind that broadcast and recordings do a lot of work to reduce language drift today, over time but especially over geography. A mountainous, watery place like Greece almost certainly would have a lot of drift in accent and informal language.

[On edit: That feeling when you swipe out a long comment, hit submit, and see a fresh neighboring long comment making the same points better.]

Agathos reminds me, nimrod in English used to mean a mighty hunter and after Looney Toons used it sarcastically to make fun of Elmer Fudd now means dimwit.