Thanks so much, Nathan, for sharing our video and article! It's quite enthralling to see how many people are interested not only in revitalizing Ancient Greek as a spoken language, but that they are interested in investigating historical phonology as well.
RandallButh wrote: ↑
May 18th, 2020, 4:35 pm
Briefly, the λουκιανὴ προφορά is a serious endeavor and moving in the right general direction.
It's very exciting to have caught your attention, sir. Raphael Turrigiano and I are great fans of your work. As polyglots of many modern languages, we can't imagine one ever attempting to truly read
material without having first attained basic fluency. And in your talks and papers, you have correctly insisted that one must learn how to speak
in order to be able to read well. The Buth Koine Pronunciation is an excellent vehicle to achieve that, as demonstrated by Ben Kantor at KoineGreek.com and many others who have used your system to great effect. Buth Koine Pronunciation is elegent, simple, and close enough to Modern Greek to help potentially bring in more living Greeks.
So, yes, in the authors' terms I am "guilty" of permissiveness. He said he didn't understand my permissiveness, so maybe we have a difference of personality and pedagogical emphasis.
Actually, ὦ κύ̄ριε μου, I was critiquing my own
permissiveness, hehe, not yours. Before collaborating with Raphael, I was playing around with something a bit more archaic than Buth Koine Pronunciation, but I wanted both /f/ for φ and /au̯/ for αυ, which as we know is an extremely unlikely coaeval historical development. And then Rahpael helped guide me back to the intermediate stages of both phonemes, which we decided was preferable for Standard Lucian Pronunciation. But our Evolved Variant of Lucian Pronunciation in most respects is virtually identical to Buth Koine Pronunciation; they dovetail together nicely I think.
The only recommendation we suggested was that palatalization, which you excellentely demonstrated to have been incipient quite early (I am currently thinking it was common back into the Classical period), ought to be applied not just to γ before front vowels, but to all velars. If that is added to it, then Koine Buth Pronunciation becomes completely historically valid in my opinion at least for Late Koine and much of the Byzantine era, plausibly even representing the sound of at least some speakers back to the year 1 AD (or even much before!).
2. The Λουκιανή system misses the merging of length in the Koine system. αι=ε, οι=υ, ω=ο, ει=ι, along with its effects on dipthongs ευ, αυ, ηυ and the accent system. The great Greek vowel-shift was apparently triggered by EI=I: and driven by a loss of length.
This point 2 has a small caveat. Ancient academics were aware of length and preserved it in some schools. But the populace had merged long and short vowels all around the Mediterranean. Length dropped out of the system post-Alexander. Read the Jewish catacomb inscriptions [οδε κιτε ΑΒΓ "here lies so-and-so" ωδε κειται ΑΒΓ] in the West, and the Babata texts in the Judean desert in the East. When someone walked out into the street from the academy, they had to understand Greek that no longer contained phonemic length, with all that that entails.
To give you an idea of where I come from, I specialize in Latin phonological changes across the centuries (q.v. my talk on Latin and Greek changes from 500 BC to 500 AD https://youtu.be/c_Giy_LHAlU
, which is on the whole surface level but fun). And thanks to the research by J.N. Adams, we know when the Latin phonemic vowel quantity starts to erode (1cAD Pompeian incriptions show final syllable open long vowels shortening) until the system is mostly gone for all speakers (around 5cAD). By coincidence, or by cross-pollination of a sort, Latin and Greek evolve remarkably in step, showing evidence of /β/ or /v/ both in the 1cAD, and both seem to lose their phonemic vowel quantities during the Western Roman Empire.
Because of the elegance and simplicity of the Buth Koine Pronunciation, I am very much tempted to accept its validity even as early as you posit here, specifically than phonemic vowel length had dropped out after Alexander the Great. But this cannot be reconciled, I'm afraid, with the borrowings of Greek words into Old Latin (300 to 100 BC) and later into Classical Latin (100 BC to 200 AD). The Roman ear demonstrates particular sensitivity to vowel changes in Greek. For example, Latin cōmoedia < κωμῳδίᾱ
— this word was first used in the time of Plautus (pre-Classical or Old Latin) and retained to the present day in spoken Latin, whereas later rhapsōdus < ῥαψῳδός
was borrowed later. The loss of the long diphthongs based on these sorts of spellings is placed at 150 BC to 50 BC.
But the constant is that Greek words borrowed into Latin retain
the same phonemic vowel lengths as they did in the contemporary Greek. If Romans were merely slavishly imitating Classical Attic or 4cBC authors, then they would have retained the long diphthongs too. But that's not the case. Moreover, Classical Latin grammarians frequently compare Latin long vowels and dipthongs to the Greek ones, including αε and οι. The 2cAD Latin tombstone inscription "foinix" < φοίνιξ
shows that the inscriber there adopted the contemporary /f/ or /φ/ sound of φ, but the vowel is not written y
, which we would expect in a merger. Was this Latin ingraver hearing /øy̯/ or /øː/? It could be either. But οι and υ were not the same sound for him. And moreover, that sepulchral inscription was written in verse, so we know it had a heavy weight — it was a long vowel. What of αι
as /ɛː/ ? Absolutely! I can confirm that Latin ae
was also pronounced /ɛː/ by many Romans in the Classical Latin Period, and as common as it was, it was not the only dominant pronunciation as evidenced by descendents into the Romance languages. So from my investigations I see Latin ae and Greek αι behaving as either diphthong or monophthong, and these differences must have been diatopic, diaphasic, or any number of possibilities. The point I would emphasize is that we as modern revivalists get to choose from either for historical reconstructions over a broad span of a few centuries. Our Evolved Lucian Pronunciation, in a yet more Innovative variant, also recommends monophthongs for αι and οι.
You have shown highly compelling pieces of evidence for the losing of phonemic vowel length during the Koine Period, but οδε κιτε
for ωδε κειται
doesn't happen to be one of them, first because ι
could be short or long, such as ἀποκρί̄νεται
, and also because the spelling of ο for ω, just like ε for η and ει before the 403 BC Euclidean spelling reform, was not universal, or so it seems to me based on the usage of Greek orthography in Magna Graecia, where older spelling customs were dominating. And both Greeks and Romans consistently show more attention to vowel quality
than quantity when writing, hence the importance of η and ω being introduced because their qualities were sufficiently different from ε ο by the 4cBC, yet α ι υ don't get a similar update (thus it was not an update for the sake of vowel length, but for quality). Classical Latin puns also shows this; Cicero explains why we say nōbīscum
and not cum nōbīs
because, due to assimilation, cum nōbīs
sounds like cunnō bis
(!) and therefore is highly offensive to the ear. And as assiduous as the Classical Latin speakers were about vowel length, the long 'i' in no way prevented the listener from hearing the word "bis" with a short 'i'.
But I in no way dispute the idea that phonemic vowel length is disappearing
1cAD-5cAD! Indeed, the very reason the diacritical marks we use today for pitch accent were invented in Alexandria was because the common Egyptian Greek speakers had lost phonemic vowel length — the loss of phonemic vowel length drives the loss of the pitch accent system. Thus, this large population of Greek speakers in Egypt fully justify our speaking Koine without phonemic vowel length, if we choose to do so.
As for pitch accent, the Roman grammarians explain the pitch accent of contemporary Greek with high and low notes in the same way that the Classical Greeks did. Thus I opine that the Greeks with whom the Romans were in contact had pitch accent as well as vowel length (and probably stress accent growing under the pitch accent, as Serbo-Croatian). This in no way elimates the possibility of many millions of Greeks (such as those in Egypt) who had lost these features.
I would be hesitant to say a majority of Greeks had lost geminated consonants, however. Geminated consonants survive in many Greek dialects today. I believe thay should be prescribed in the pronunciation of any form of Ancient Greek.
My conclusion based on the contradictory evidence is that the some Greeks retained phonemic vowel length and others did not, some retained pitch accent and others did not. I don't think it's an either-or situation.
When we have ample evidence of every type of sheet music for voice from the Roman Empire with both pitch accent and phonemic vowel length as well as geminated consonants, such as the 2cAD epitaph of Seikilos whose piece I do a rendition of here (do forgive my mediocre singing voice!)
— it then seems reasonable to prescribe the teaching of phonemic vowel length as we do for Latin literature of the same period, because otherwise we are robbed of the intended sound of Greek poetry and song. Was such a pronunciation limited to higher registered speakers and composers? Possibly. Since the epitaph of Seikilos was written by common folks, and was meant to be appreciated by common folks who walked by it, I tend to think that the phonemic vowel length question was a diatopic
rather that diastratic difference.
PS: It is also interesting that they have named their system "Lucian." Lucian wrote at the end of the second century and is noted for being extremely witty as well as "high register." I've enjoyed many a good story that he has passed on. His "True Story/History" is one of the first sci-fi short stories, replete some full Athenian Greek, in the height of the Second Sophistic. He would be a good choice for an artificial high-register pronounciation scheme. In fact, I think that Lucian wrote the Syrian Goddess. A great read where we have a 2-3c author writing an artificial Herodotian/Ionic Greek. When we run our 9-mo. Greek ulpan/σχολη in ἐν τοῖς ᾽Ιεροσολύμοις (2022?), the Syrian Goddess will be the introductory reading to Ionic Greek, just like the epigrams will provide an introduction into full Greek poetry, and Nonnus' Homeric epic-verse rendition of the gospel of John will be the intro to Homeric verse.
I am pleased you liked the name of "Lucian" !
Yes, we chose that very deliberately, not only because Lucian flourished in the 2cAD, but because he was probably an L2 Greek speaker like us, and was archaizing. In that way he seems like a great poster-child for us: using as ancient form of an ancient language, foreign speakers of Greek, seeking a join the disparate threads of literature through the centuries. So, if we have erred in our historical phonology, we might be no worse than Lucian who surely also had a a foreign accent when he spoke.
Effectively, we wish to see Lucian Pronunciation replace Erasmian some day, which is unsuitable for historical reconstruction. And we also see future users of Lucian Pronunciation interacting harmoniously with Buth Koine Pronunciation users. Indeed, Lucian Pronunciation's most Innovative Variant is essentially the same as Buth Koine Pronunciation, and the Archaic Variants are virtually indistinguishable from Restored Classical Attic. Thus Lucian Pronunciation is designed to be rolled forwards or backwards per the preference of the speaker, in a systematic way that permits access to Ancient Greek literature of all periods — and best of all, permits spoken
communication with any other Ancient Greek speaker, no matter which system that person might also use.
Πάλιν, ὦ κύ̄ριε Βουθ, ἐγὼ καὶ Ῥᾱφᾱὴλ ἔχομέν σοι χάριτας πολλά̄ς — we are very flattered you took the time to look at our work, and reiterate our utmost admiration for what you have done and continue to do for Spoken Koine.