B-Greek: The Biblical Greek Forum

FTR:

Koine pronunciation recommends unaspirated π, τ, κ, and fricative φ, θ, χ, as well as voiced fricative β, δ, γ.

PS:
If you use them for a few hundred hours you will begin to hear them.
The result is not always pleasant. In Hebrew it bothers me sometimes to hear חית pronounced as a fricative כית: thus אח pronounced like אך. My ears would say "ixsa" [slang for "not nice"]! But it is so common, one does live with it.

RandallButh wrote: ↑

May 23rd, 2020, 12:07 pm

Sorry for the delay in comment. We all have other commitments.

Hardly necessary, sir, when I, equally swamped, have responded so many weeks later! My apologies. Thank you for the discussion.

The quote (οδε κιτε "here lies") does, in fact, demonstrate merged/dropped length. First, ε is short while the etymological vowel αι was long. This cannot be attributed to pre-Euclidian spelling from 500-700 years (!) earlier because E was used for long-E ("proto-eta") not AI. Secondly, once ε demonstrates the drop of length, the  o for ω and ι for ει become confirmatory, a preponderance of evidence for this particular sentence.

Right, that's one explanation, and a good one, but not the only one from my point of view. Ancient spelling in Latin and Greek much more frequently follows vowel qualities rather than quantities, hence why Latin never had a standard system of showing vowel quantity (sporadic attempts such as vowel gemination as well as apices were variously employed, and even ei for /iː/ {which in 2cBC was used in direct imitation of contemporary Greek ει /iː/}), and after Euclid, Greek spelling only included (inconsistently) η and ω, and not long versions of the other vowels, because the qualities were different; their length was incidental to the reform.

The other issue I see in your argument is that you posit plain iota ι as the short vowel, and ει as the long counterpart. This, of course, is not true. The digraph ει merged phonetically with long ῑ, in words like ἀποκρί̄νεται, sometime in early Koine. So an ι can be written for ει and a long vowel still be indicated. You address this in your comments that follow, which I appreciate, but to me it seems plain that the example οδε κιτε relies on the ω > ο and αι > ε exclusively.

And of course you provide lots of excellent evidence, showing the onset of the loss of phonemic vowel length. And I counter with the Roman grammarians describing phonemic vowel length in Greek words, and the ample songs with pitch accent and phonemic vowel length maintained, as well as poetry, from the same period, circa 2cAD. Can we both be right, for dialects separated by geography and register?

I think so. I firmly believe there are probably some, if not very many, Greek speakers around the Empire who are pronouncing their language according to Buth Koine phonology (plus palatalization of velars before front vowels and with the obligatory retracted s), and if they are localized in the Judea region, so much the better since the goal of Buth Koine Pronunciation is to bring alive the languague of the New Testament. I fully support all of this, and laud all who practice it.

Lucian Pronunciation's goals are much broader. Buth Koine Pronunciation is less well suited to mid and late Koine texts that show pitch accent and phonemic vowel length (also, geminated consonants are a must through late Koine, unless you have evidence contrary to this argument?). Phonemic vowel and syllable length is essential for Ancient Greek poetry and music. Lucian Pronunciation's emphasis on this is historically valid for the Roman Empire, as previously demonstrated, and also for the Late Roman Republic (Graecia Capta), and thus all the rich literature of the first half of the period which is called Koine (Theocritus, Euclid, Aristotle, Thrax, Apollonius, Menander, Eudoxus, etc.) and even the great authors who follow during the Empire. If poems, rhymes, songs, puns, or other forms of literature are hampered by strict observance of phonemic vowel length — such as they are in much of Mediaeval Latin common songs and poems — then the phonemic vowel length skill must be turned off for those texts. Otherwise, it serves as an excellent vehicle to all Ancient Greek literature. Lucian Pronunciation does not seek to displace or supplant Buth Koine Pronunciation — which is more targeted in literary scope and in geographical dialect — but to coexist with it.

It does in this respect seek to replace Erasmian. Thus Lucian Pronunciation works as a convention even for Classical Attic literature, as well as texts written in other dialects or centuries prior (Sappho, Anacreon, Herodotus). And for those who want a more Classical Attic sound, the Archaic Variant of Lucian Pronunciation fits this nicely.

I assume that you have read the discussion on goals and criteria for pronunciation choice in the PDF at www.biblicallanguagecenter.com ?
https://www.biblicallanguagecenter.com/ ... n-2012.pdf

You assume correctly, sir! for many years. Your advice and appeals to certain compromises were the basis of our investigations.

As outlined in the Lucian Pronunciation video, the choice of a pronunciation when reviving an ancient tongue is a quadrivium of these factors:

science + art + pedagogy + politics

In your 2008 revised 2012 paper, Dr. Buth, the choices you make and compromises you advise are very much in line with that quadrivium. Lucian Pronunciation also acquires a good balance in the quadrivium, and is selected for based on it. Thus they can coexist.

In other words, fronted velars are adiafora, and not worth chasing, for the most part.

I find this extremely difficult to agree with, as fronting of velars is central to Buth Koine de facto, if not de jure, simply due to the fact that γ is treated in Buth Koine as in Modern Greek. I cannot find a strong linguistic argument why γε should be /ʝɛ/ but χε should be /xɛ/ and not /çɛ/ — the latter of course being the Modern Greek pronunciation too, which greatly adds to the authenticity of the sound, and its acceptability to the Modern Greek ear, a stated tenant in the 2008/2012 paper, and well synchronized with the politics factor in the quadrivium.

the system should not build an artificial pronuniciation in order to benefit "spelling" or "classical poetic scansion" or some otherwise laudable goal.

Right, as you have demonstrated plenty of Greek people of the first part of the Roman Empire, particularly those inhabiting Judea, to be losing phonemic vowel length, I have demonstrated others who retain it (Seikilos, 2cAD)

Students will not follow the rhetoric and word choices properly if they are using a system phonemically foreign to the Koine.

This is where I start to have a problem with terminology, while praising the goal you and I both aspire to. The Buth Koine Pronunciation is for a targeted, more specific dialect — let's qualify it as Biblical Koine. That I'm fine with that. But "a system phonemically foreign" to all of Koine over an 8-century year period in dozens of geographies? I find this use of "Koine" misleading. And it's utterly at odds with the first half of Koine literature demonstrably (Menander, Theocritus, who demand phonemic vowel length for correct reading, and spoke Koine) as well as the poetry even of the commoners (the Seikilos inscription of 2cAD which preserves both pitch accent and phonemic vowel length).

But you and I very much share the same goal of wanting to represent the voices of our authors with the greatest possible authenticity. I would merely caution against over-extending a Biblical pronunciation of Greek beyond certain boundaries of time and space to the whole of Koine literature 400 BC - 400 AD. That would be a dodgy move, in my estimation.

We don't use Chaucerian pronunciation for post-Great-Vowel-Shift Shakespeare.

But we do use the Shakespearean Original Pronunciation for Shakespeare, as I demonstrate in this video of mine: https://youtu.be/u-cTCAPFYOE

You might say that most people could object to the unfamiliar sound of OP in Shakespeare, and prefer 21st century Modern English pronunciation. Absolutely. But if that's the argument, it's equally valid to opine for Modern Greek pronunciation for all Ancient Greek. So that idea would need more definitive boundaries. Yet I share your sentiment.

You gave the analogy of the Greek system reordering itself according to a change akin to the English Great Vowel Shift, and it is a great idea. But just as Scots descends as a language with very different realizations after the Great Vowel Shift, there is no reason that the Greek Great Vowel Shift you propose was necessarily universal, or fast. My reading of the evidence is a many-centuries metamorphosis from start to finish, from 1cAD-5cAD — incidentally, the same centuries that Latin also gradually loses its phonemic vowel length. Intense language contact is probably a factor to be considered and researched.

So this all comes down to someone wanting to learn "Ancient Greek." What kind? Homeric? Classical Attic? Biblical? Non-Biblical Roman? All of the above?

If one is dedicated to Biblical study, one need not trouble oneself with phonemic vowel length and syllable quantity. Buth Koine Pronunciation is perfect for this.

However, if one wants to learn a pronunciation that gives access to all Ancient Greek literature 800 BC to 400 AD, then Standard Lucian Pronunciation works very well — its modified phonetic variant called the Evolved/Innovative Variant is nearly the same as Buth Koine Pronunciation, while its Archaic Variant is virtually the same as Restored Classical Attic of 5cBC, so individuals can choose one from among the three variants that is more to their preference for one of the reasons stated in the quadrivium (science + art + pedagogy + politics).

If you're like me, you enjoy the pleasure and challenge of learning several different historical pronunciations of an ancient language, changing according to the specific time periods and locations of the authors whose works you recite.

But if you're like most people, you'll want to choose one pronunciation that is an acceptable vehicle for all of the literary periods that interest you. That is the goal of Lucian Pronuciation: a convention for students of all Ancient Greek literature.

The quote (οδε κιτε "here lies") does, in fact, demonstrate merged/dropped length. First, ε is short while the etymological vowel αι was long. This cannot be attributed to pre-Euclidian spelling from 500-700 years (!) earlier because E was used for long-E ("proto-eta") not AI. Secondly, once ε demonstrates the drop of length, the o for ω and ι for ει become confirmatory, a preponderance of evidence for this particular sentence.

Right, that's one explanation, and a good one, but not the only one from my point of view.

I wouldn't try to push water uphill on these catacomb inscriptions. "Divide and conquer" doesn't work. The "short ε for the etymological long αι" makes it irrelevant to point out what MIGHT have caused the other parts of the example.

excellent evidence, showing the onset of the loss of phonemic vowel length.

"onset" sounds like a term to spin this as if it hadn't happened. But it did, it had happened. The Nazaereth inscription shows a mason/scribe in the employ of the government doing a length-less hatchet job on Caesar's edict.

I firmly believe there are probably some, if not very many, Greek speakers around the Empire who are pronouncing their language according to Buth Koine phonology (plus palatalization of velars before front vowels

Your argument seems to be that the ability of some literary poets to compose metrical poetry meant that the MAJORITY of the people spoke such Greek. Au contraire. The poetic evidence only shows that some people knew how to write metrical poetry, it does not show what the common people were speaking.

plenty of Greek people of the first part of the Roman Empire, particularly those inhabiting Judea, to be losing phonemic vowel length

That's better. Now it's "plenty of Greek people." Good. However, your "to be losing" seems to be tryng to avoid the evidence with an open-ended imperfective. They had LOST phonemic length, done. Those were not momentary, prophetic lapses. Again, misspellings by some shows what common people were pronouncing. Correct spelling by others at the same period only proves that some people were able to write etymologically. Of course, we know that some in the academy bemoaned what they were hearing. However, with literacy far far far from a majority, it does not make sense to use academy prescription for describing the majority. To do so would cultivate an artificial pronounciation and would pretend that it was not elitist. My Chaucer//Shakespeare analogy was saying that importing Chaucerian into Shakespeare would be incorrect. People continued to write "chaucerian" but that did not mean that they spoke it. In fact, people still write "right, night, and knight" but we only pronounce keniγχt as a joke. Likewise, Attic writing in the Roman period does not establish Roman koine pronunciation.

from 1cAD-5cAD

Much too late. Try 3rdBCE-1BCE as the correct period for loss of length.

RandallButh wrote: ↑

June 6th, 2020, 11:13 pm

"onset" sounds like a term to spin this as if it hadn't happened. But it did, it had happened. The Nazaereth inscription shows a mason/scribe in the employ of the government doing a length-less hatchet job on Caesar's edict.

Your argument seems to be that the ability of some literary poets to compose metrical poetry meant that the MAJORITY of the people spoke such Greek. Au contraire. The poetic evidence only shows that some people knew how to write metrical poetry, it does not show what the common people were speaking.

We have the direct attestation of Roman grammarians from the 2cAD — here, Velius Longus — asserting the length distinctions of the vowels: https://latin.packhum.org/loc/1374/1/1/745-761@1#1

The key note here: α ι υ are used "still today" for both long and short vowels.

Thus the Vowel Shift you propose is not universal at this time. It will be. But not yet. Since Roman grammarians frequently contrast the nature of Latin with that of not just a standard academic Greek, but of the current Greek dialects (as Velius Longus does in his book), this would have been the perfect time for Velius Longus to add: "however, no/few Greeks make the vowel length distinctions anymore."

Velius Longus's silence on the matter indicates the loss of phonemic vowel length was not a phenomenon known to him. It seems reasonable to conclude loss of phonemic vowel length was common in Egypt (as documented by Horrocks) and in surrounding regions by 3cBC or 2cBC. But we can surmise not universally in places such as Italy and likely not in Athens (where Romans would go to learn Greek fluently, among other outlying islands).

That's better. Now it's "plenty of Greek people." Good. However, your "to be losing" seems to be tryng to avoid the evidence with an open-ended imperfective.

I retort, with utmost respect, Dr. Buth, that your assertions seem to be avoiding the incontravertible evidence of thousands upon thousands of Greek words borrowed into Latin with their phonemic vowel length intact — not only in poetry, but in prose (Cicero's clausulae, for example).

And where Greek words borrowed from non-Attic dialects had different vowel lengths, those are retained as well — one example among many is πλατεῖα which comes into Latin as plătĕa, later plătĭa > platja > (Italian) piazza, (Spanish) plaza. The common Greek in Italy had *πλατεα.

They had LOST phonemic length, done.

"They" = all Greek speakers?
"lost" : when? If by 5cAD, we agree. But you state:

Much too late. Try 3rdBCE-1BCE as the correct period for loss of length.

This is very funny to me, as it's demonstrably false. Thousands of Greek words are borrowed into Latin from 3cBC-1cBC, and thereafter, with all their vowel lengths intact — until the end of the Western Roman Empire when this moraic phenomenon had left both languages for certain. And when pronunciation changed in contemporary Greek, borrowings into Latin reflected that pronunciation change. For example:

κωμῳδίᾱ > cōmoedia (circa 3cBC)
μελῳδίᾱ > melōdia (circa 2cAD (!)) https://latin.packhum.org/loc/1518/1/39/2345-2350@1#39 (here the word melōdia is borrowed in a poem about verse itself)

This is how we know iota-subscript is lost around the 2cBC, because Thrāx (Θρᾷξ !) states it directly in his text, and indirectly in the Latin spelling of his name.

The retained vowel lengths in both words above are a big deal, and cannot be easily dismissed. The change of iota-subscript shows the Roman sensitivity to the sound of Greek, as does their loyal employment of the contemporary vowel lengths of those Greek words.

I can cite Roman grammarians Varro, Cicero, Quintilian, Probus, Velius Longus, Aulus Gellius, Flavius Caper, etc., describing the contemporary Greek of their age that they came into contact with in terms of phonemic vowel length (and pitch accent too!). The great heaps of evidence in the form of Greek spelling mistakes you have wonderfully collected and presented form a compelling argument for loss of phonemic vowel length for those speakers. I am very much persuaded by your theory. But, it is a theory based on evidence, whereas we have direct attestation of Romans who spoke Greek with Greeks stating the contrary to your argument.

On οδε κιτε, Raphael Turrigiano pointed this out to me (W. S. Allen in Vox Graeca p.93-94):

"We have seen (p. 75) that the monophthong resulting from αι came often to be written as ε; but this need indicate no more than the quality of the monophthong in the absence of any other appropriate symbol (cf. Sturtevant, pp. 39, 103). The appearance of ει for short ι in the 2 c. A.D. need be no more than a graphic reflex of the use of ι ( = [ī]) for ει. Thus, whilst these phenomena could result from a loss of length-distinctions, they need not do so, and cannot therefore be relied upon as evidence. More suggestive is the confusion of ο and ω, which becomes common from the 2 c. A.D.; but since such confusion begins as early as the 3 c. B.C., it could again indicate a convergence of quality rather than duration, in which the considerations mentioned on p. 85 may be relevant."

In discussion with Raphael, we surmise that if all vowel length contrast had been neutralized in all colloquial Greek by the 1cBC, then all mispellings should look the same from that point on — but they don't. As the evidence for confusion of vowel length grows — and is especially present in certain dialects like Egyptian Koine — this indicates a system that is still evolving.

Allen says that the epigraphical phenomena that you have amassed, Dr. Buth, "cannot therefore be relied upon as evidence," but I don't concur — you have concluded that they do account for sufficient evidence, and I am persuaded by your argument — just limited to some extent in register or geography. Given that caveat, I support your conclusions. But not for the pan-Hellenic voice.

So, what's the solution to the seeming contradiction of historical evidence? I propose this as the solution: A division of Greek speakers, based on geographical or other lines, using phonemic vowel length versus those who don't.

Horrocks has previously demonstrated the phonemic vowel length loss in Egypt as early as 2cBC. And there is no reason to think it was necessarily isolated there. But the universal loss of phonemic vowel length is not clear until the late Western Roman Empire. I stand with Allen and Horrocks who date loss of phonemic vowel length becoming dominant in 3cAD.

Again, misspellings by some shows what common people were pronouncing.

This needs a caveat: a common people, but not all Greeks, in the Roman period, were pronouncing their language without phonemic vowel length. Common Greeks in Italy retained them, whence the vernacular borrowings found in Plautus, Terence, Horace, etc., where vowel length is maintained, plus the standard Greek learned by Romans abroad in Greece.

However, with literacy far far far from a majority, it does not make sense to use academy prescription for describing the majority.

It would be worthwhile to acknowledge that Buth Koine Pronunciation and Lucian Pronunciation have different aims, as I recalled in my previous post: Buth Koine Pronunciation is intended to appreciate Biblical Greek literature. Lucian Pronunciation is for all Ancient Greek. When it comes to the 2cAD, both can and probably did coexist in the same world, possibly even in the same geographies.

Can one use Buth Koine Pronunciation for all Ancient Greek, even Homer? Of course! It's merely a convention. People can and do speak Latin today with the Ecclesiastical Pronuncation, as well as traditional Polish and German Pronunciations, and appreciate Classical Latin texts with that set of sounds on a daily basis. It's not "historical," but who cares? It works for them. See the historical pronunciation quadrivium (science + art + pedagogy + politics). And just the same, one can appreciate Biblical literature with the Lucian Pronunciation.

I think it's perfectly acceptable to call Buth Koine Pronunciation — a key feature for which, as you assert, is loss of phonemic vowel length (but not geminated consonants, right?) — a historical pronunciation of the early Roman Empire, localized in Egypt and even through Judea, if palatalization of velars before front vowels is restored, geminated consonants retained, and retracted s is employed (though I can believe a Semitic substrate not using the last), becoming more historically plausible and geographically widespread the closer we approach Byzantine times. But to say Greek that lacks phonemic vowel length should be the historical pronunciation for all Greek in the Roman Empire (and Republic? no.) is contrary to direct attestation. You might as well claim that Classical Latin also lacked phonemic vowel length (and surely you don't claim that, do you? — but if you do I am sincerely very interested to hear the evidence!).

Thus, I see the following compromise. Hundreds if not thousands have learned Greek by the Buth Koine Pronunciation, and I support and defend every one who uses Buth Koine Pronunciation, and will for all my days, for they are employing an historically plausible, beautiful convention for Ancient Greek. And I vehemently support and promote Buth Koine Pronunciation over any variety of "Erasmian," including the kind found at the estimable Polis Instutite.

And yet, for those of us who are well acquainted with living languages that today have phonemic vowel length — Finnish, Japanese, Czech, Estonian, Serbo-Croatian, Hungarian, Slovak, etc. — whose literatures necessarily take account of phonemic vowel length in composing both poetry and prose, it is not desirable to dismiss phonemic vowel length in either Classical Latin or contemporary Koine Greek, because to do so robs us of the ability to appreciate the rhythm of verse and oration alike over a wide time period of literature.

We opine it should be learned rote for Latin and Ancient Greek, a skill easily turned off where it is not appropriate — Mediaeval Latin and Greek poems being a prime example.

Do you think, sir, we can agree to a compromise of coexistence?

Via Erasmeana sempiterna erit!

Sorry, just thought to inject a bit of humor (et Latine melius semper sonat).