I have visited some sources and I can't finally understand that Eta (η) in Ancient Greek pronounced like 'eɪ' (delay), or like 'eə' (hair). In fact, should we say an 'ɪ' at the end of pronouncing η, or it is pronounced only like long 'e'?
As Asteroides has said, the "classical" pronunciation (from the fifth century BCE or thereabouts) is reconstructed as
/ɛː/, a longer version of the vowel in English "red".
However, there are quite a lot of different ways to pronounce Ancient Greek—and unlike with Latin, the "reconstructed classical" version hasn't caught on as thoroughly. One of the more prominent systems is the "Erasmian" system, which was originally based on the works of Erasmus but has evolved in all sorts of different ways in different places and different traditions.
In the traditional English "Erasmian" system, Ε is pronounced as in "red" (
/ɛ/), and Η and ΕΙ both as in "raid" (
/ej/). This isn't especially accurate to how Plato and Aristophanes would have spoken, but is not uncommon among modern scholars.
(Another system pronounces it all like Modern Greek, so Ε and ΑΙ are
/e/ and Η, ΕΙ, Ι, Υ, ΟΙ, ΥΙ are all
/i/ as in "reed". This is arguably more accurate than the "Erasmian" system, because it's how the Greek language actually evolved over time, but can make it difficult to distinguish words because so many vowels merged together.)
1My understanding of IPA is sadly very limited. "a longer version of the vowel in English "red": would that make it like the vowel in the English word reed or is it not that long? Also, for what it's worth, there is no
/i/as in "reed" in modern Greek, all i versions (ι, η, ει, οι and υ) are pronounced the same and sound like the i in the English words pit or sit. (I am not a linguist, just a native Greek speaker; all I can say is that I'm certainly not aware of any other form of i sound in any of the dialects I am familiar with).– terdonNov 29, 2020 at 19:58
2@terdon No, "longer" just means literally more time/breath, it doesn't change the sound. (That's also what the IPA
ːindicates.) I don't think there is a way to notate that in english since "ee" is a completely different sound. Nov 30, 2020 at 4:07
1I was reading "Technopaegnion" by Ausonius earlier today, and came across this line: "Hτα quod Aeolidum, quodque ε valet, hoc Latiare E." just sharing, for no specific reason, that's all :)– Alex B.Dec 7, 2020 at 20:35
2@AlexB. Oh, fascinating! Looking up the context, it sounds like Ausonius equates the quality of Latin Ē with "Aeolic eta" and (Attic?) epsilon [e], Latin Ĕ with "Doric epsilon" [ɛ], Latin Ō with a sound not found in Attic [o], and Latin Ŏ with (Attic?) omicron and omega [ɔ]? Or am I misreading that?– Draconis ♦Dec 7, 2020 at 20:58
1@AlexB. Of course, Ausonius is a millennium after Homer, and over 750 years after Plato. It's certainly useful as evidence for how Greek changed, but it likely has no bearing on how Classical Attic, Doric, and Aeolic were pronounced.– cmw ♦Feb 11, 2022 at 15:58
Sources that say Ancient Greek eta was pronounced like the vowel in English "delay" or "hair" are only providing a loose approximation of the vowel.
Greek η was always a monophthong
The pronunciation changed over time. The standard reconstruction for Classical Attic Greek η is [ɛː], a long front mid-low monophthong. In modern Greek, the vowel has been raised to [i] (a front high monophthong), and between [ɛː] and [i], we reconstruct the intermediate stage [eː] or [e] (a front mid or mid-high monophthong) for Koine Greek.
For more information on Greek pronunciation and the timeline of changes such as [ɛː] > [eː~e] > [i], see the answers to Resources for pronouncing Greek.
There is no evidence that I know of for η ever being pronounced as a diphthong by Greek speakers.
English speakers often diphthongize vowels
In some accents of English spoken in southern Britain, hair is pronounced with a long monophthong [ɛː] (the symbol "eə" is a non-phonetic conservatism in transcriptions of these accents). For a speaker of this kind of English accent, the hair vowel is a close approximation to the oldest reconstructed pronunciation.
However, in many other dialects of English, the contrast between short [e] and long [ɛː] does not come very naturally, and cannot easily be described by reference to the pronunciation of English words like "late", "delay", or "hair". Since most English speakers who learn ancient Greek do so for the purpose of reading texts, and none are learning for the purpose of speaking to Ancient Greek speakers (which is obviously not possible), many pronunciation systems in common use by English speakers don't emphasize accuracy to the phonetic reconstruction of Ancient Greek. Some pronunciation choices or recommendations that you find in texts may instead made primarily for the purpose of making it easier for English speakers to remember the spelling of Greek words.
For example, Mitchell L. Holley writes in "Greek Pronunciation: The Pedagogical Pertinence":
The Erasmian system possesses great heuristic value for the new Greek student. By using the system of Erasmus, students learn a pronunciation system that allows them to quickly memorize the spelling of words, which is a vital part of learning an inflected language.
For the sake of this goal, it doesn't matter whether an English-speaking student distinguishes epsilon and eta as [e] and [ɛː], [e] and [eə], [ɛ] and [ɛː], or even [ɛ] and [eɪ]. But there is no reason to think that Ancient Greek speakers shared English speakers' tendency to diphthongize long vowels.
youtube.com/watch?v=amvLd1mnXH4. For example here αρχη is pronounced like αρχει. Nov 28, 2020 at 22:07
2@AliNikzad: There's no reason to regard a Youtube video like that as a useful source about pronunciation. Nov 28, 2020 at 22:08
1You are correct, But for another example in "Learn To Read Greek, part 1", Yale University Press, 2012, p. 5, it is written that η is pronounced like 'late'. It seems to me that it is 'eɪ'. Nov 28, 2020 at 22:13
1@AliNikzad That's an approximation designed more to be readily familiar to an English speaker than an accurate representation of the actual vowel. True distinctions between short and long vowels are rare or absent in English.– chepnerNov 29, 2020 at 15:51
These issues in rendering non-English languages into words of similar sounds in English, is only made worse by English being such a prominent research language. For a Germanic or Slavic language speaker, for instance, /e/ to /e:/, /i/ to /i:/ etc., is no problem at all. I wish such explanations would rather rely on explanations along the lines of ‘a long vowel: like the e in let, but long.’ Jan 18, 2021 at 12:51