Audio for Classical Texts in Demotic

[Ursinus]

I was wondering if any of you have happened upon classical texts put to Modern Greek pronunciation. I suppose I am looking for anything, including Koine, post-classical, Byzantine, but I would prefer ante-classical texts. Obviously, I have the NT covered-- though if you knew where to get some LXX audio, I would be grateful. Finding some Xenophon would be great.

[rmedinap]

Yes, this guy has recorded some Xenophon, Plato and Homer using Modern Greek Pronunciation.

I'm currently recording a few good anthologies and easy texts I've found around using Modern Greek but you'll have to wait a while to see those.

[Dante]

isn't that like wanting to hear Shakespeare read by someone with a heavy Brooklyn accent?

[jeidsath]

Shakespeare is perfectly intelligible with modern accents, and the poetry carries through just fine. In fact, I don't actually know that Brooklyn accents are any farther from the English of Shakespeare's day than the BBC standard English is, or the American Midwestern drawl. These accents are probably all closer to each other than to Stratfordian 16th-century English.

[Timothée]

Shakespeare is perfectly intelligible with modern accents

It depends what is meant by “perfectly intelligible”. Much is lost in Shakespeare with modern accents. The most obvious deficiencies are puns that don’t work in modern accents: in S’s time, for instance, hour and whore were pronounced the same (a pun in As You Like It). The same is true of lines and loins, a pun in play in Romeo & Juliet’s prologue (l. 5: “from forth the fatal loins of these two foes”). Possibly of lesser importance is that many a rhyme doesn’t work anymore. More on the subject can be found e.g. on Youtube, where I got these examples.

It’d be interesting to hear for what purpose modern pronunciation of classical texts is felt needed.

[jeidsath]

Your examples are true, but how in how many lines of Shakespeare can we identify a problem like that? 1 in 100? 1 in 500? I listen to a fair amount of Milton/Spencer/Shakespeare and rarely have problems. By the time we get to Gibbon/Johnson, the difficulties are very rare.

On the other hand, Chaucer is very difficult, and rarely completely intelligible to me.

[rmedinap]

isn't that like wanting to hear Shakespeare read by someone with a heavy Brooklyn accent?

Yes, yes it is.

It’d be interesting to hear for what purpose modern pronunciation of classical texts is felt needed.

You're both right beyond any doubt in what you're trying to imply. Modern Greek Pronunciation is not the most accurate way to pronounce Ancient Greek... BUT the way most people pronounce the reconstructed pronunciation is even less accurate.

And I'm not talking about a beginner's clumsy attempt to read aloud like this one.

Ever heard of Stephen Daitz? As far as I know the guy was the only recognised authority of the reconstructed pronunciation (at least in the academic world cf. Allen's Vox Graeca 3ed page 129 or page 118 of the 2ed, Dillon page 323). He was known for his The Pronunciation and Reading of Ancient Greek and some recordings like this one of Homer. Here's a free sample.

I've never heard something so horrible, maybe you won't notice just how American he sounds because English is your mother tongue. The same way I've heard brilliant German scholars pronounce Greek as if they were reading the Faust (typical to hear is Εὐρώπη with the Εὐ pronounced as [oi]). I can complain about the same problem with French scholars, Spanish speaking scholars and Italian scholars. They all pronounce one way or another the same way they pronounce their mother tongue and they DON'T notice.

In 2011 I met a Greek Scholar in a congress, it took me a while to get him drunk enough so that he'd agree to read some Thucydides using the reconstructed pronunciation (He was quite reticent about it, he actually took it as an insult at the beginning, so I had to explain that I was just curious about how it would sound). It was like reading Umberto Eco's Foucault's Pendulum after watching the Da Vinci's Code movie, or hearing Anna Netrebko after hearing your mad sister scream. After that I became convinced that any legitimacy that the reconstructed pronunciation may have is completely lost if you continue to speak as if Ancient Greek were your mother tongue.

My arguments are:

1) We can know beyond any doubt how Modern Greek is pronounced and we can consequently imitate it accurately without getting into endless (and often pointless) debates about it. Afterwards we can take it as a starting point to learn the Reconstructed Pronunciation.

2) All the inaccuracies and lacunas and points of dispute among the linguists and philologists would be perfectly forgiveable if those who use the reconstructed pronunciation were to have a basic common ground. But not only do they pronounce however they like. They never seem to reach an agreement on some basic stuff. I heard people pronounce the iota subscript others don't.

Another example. Many Reconstructionists use a stress accent instead of a pitch accent despite knowing that the Greeks used a pitch accent. Allen himself (loc.cit. cf. Dillon 324) advises against using a pitch accent (“Melodic method”) despite his own evidence that Ancient Greek used a pitch accent. As we have seen the result is Daitz'... whatever that is.

And let's not forget that

Not only were there many different varieties of Greek spoken at any given point in time, but those varieties changed through time. Thus the problem of determining how to pronounce ancient Greek is both diachronic and synchronic both horizontal and vertical

McNeal page 99

If the argument to use the reconstructed pronunciation is not not lose whatever vocal qualities the language may have had, then I at least expect some coherence on its defenders and truly try and protect the language from "impurities" (I'm being sarcastic here) specially modern foreign ones.

Those and some other arguments have been discussed at length in an article by Matthew Dillon

Modern Greek must serve as the basis of any foundation for the sound of Ancient Greek. It is the only universally available standard, and recent research suggests it is not so far from the pronunciation of Ancient Greek, at least in the immediate post- classical period, as is usually assumed. This step is essential above all in order to master the art of συνεπέια, the flow and rhythm that is unique to any language, and in the case of Greek, very far from the cadence of English or most other European languages. If nothing else, the accentual pattern of Greek (which has remained absolutely stable over the millennia) will become instinctive, instead of remaining, as it does for many of us, a constant stumbling block. Even at the level of individual words, the problems presented by radically different phonetic values (e.g., iotacism, ευ, αυ, β, etc.) will be counterbalanced by a much improved pronunciation of other letters, especially "τ" and "π," which are as a rule fatally overaspirated by English speakers. Indeed, on the whole, Greek sits in the mouth rather lightly compared to English, and we need to adjust our vocal apparatus radically to imitate the overall effect.

Few classicists who take the trouble to learn a little Modern Greek will be content to pronounce Ancient Greek precisely as Modern. We must compromise to produce something like "la lingua antica in bocca moderna." The adjustments will vary from case to case and person to person; given the uncertainties and unknowns of Greek phonetics, we cannot expect general consensus. But the results should not be totally chaotic. If Modern Greek is recognized as a valid starting point, our differences will at least emerge from a basic common ground. A number of models might well emerge, but the underlying standard will be recognizably Greek.

Page 333

And notice that I'm not arguing AGAINST the reconstructed pronunciation (however much reticence and misgivings I may have about some of the evidence used to support it, or its interpretation), I'm arguing against those that either mispronounce it or are not coherent about their principles, and I'm offering a very valid alternative to teach an extant undisputed and available pronunciation that's easy to learn, in order to better learn a hypothetical and disputed pronunciation.

And I believe that Prof. Richard A. McNeal is right in pointing that the whole pronunciation debate is not REALLY about the pronunciation but about underling biases.

What is at issue here is a basic difference of attitude, one which needs to be more widely understood than it apparently is. It is not simply a matter of different facts, but rather of an entirely different habit of mind. There will probably never be a rapprochment between the contending sides because each group talks in terms which the other refuses to understand. The Neo-Hellenist glories in the beauty of a long-established tradition. The Erasmian yearns for an antiquity unsullied by medieval accretions. There is the difficulty in a nutshell, and no amount of logic is apt to disturb one's prior commitment to one of these two alternatives. It is really a matter for temperament rather than reason

Page 82

It is actually similar to the debate among German scholars about Goethe's German pronunciation. Many refuse to accept the very real possibility that Germany's iconic writer, born in Frankfurt, probably spoke the dialect of Hesse (Hessischer Dialekt) if not as his mother tongue at least as an everyday language and he also probably had a Hessian accent (the dialect of Hesse is the most hated of all German dialects and usually associated with poor rustic and uncivilised people), because, like all German aristocrats, he spoke French at court. It's all prejudice of course, but you won't meet any German that doesn't twist in agony if you dare recite the Faust with a Hessian accent.

I think the same applies to most of the defenders of the reconstructed pronunciation, it's not that they're wrong in their facts and arguments, but they choose that pronunciation because of some prejudice against the Modern Greek. And to be fair the same applies to Modern Greeks and their nationalistic prejudices against Western (German-English) Scholarship.

[jeidsath]

I don't have a problem with people sounding American or French or German or Modern Greek when they read ancient Greek, and I don't exactly see why anyone would prefer one to the other. But I do have a problem with unintelligibility and non-expressiveness.

The unintelligibility problem with modern Greek audio: the vowel system has collapsed from about 23 phonemes down to 8. Here is how our modern Greek pronounces the start of the Anabasis:

Δαρίου κέ Παρισάτιδος γίγνοντε πέδες δίο, πρεσβίτερος μὲν Ἀρταξέρξις, νεότερος δὲ Κίρος: ἐπί δὲ ισθένι Δαρίος κέ ιπόπτευε τελευτίν τού βίου, ἐβούλετο τό πέδε ἀμφοτέρο παρίνε.

It can be difficult to work out the grammar unless you already know what it says. The information just isn't there in many cases (famously, ὑμᾶς vs. ἡμᾶς are both just ιμάς in modern Greek pronunciation). It's Timothée's Shakespeare problem, but one that occurs every few words instead of once every few hundred lines.

And the non-expressiveness problem is just as bad for a modern Greeks as it is for Daitz. If you're just reading the syllables off the page because you don't understand the language well enough to understand them as you pronounce them, you're going to sound horrible. Anyone who has sat through lay members reading scripture at Church knows what this is like (and I mean native English speakers reading the text in modern English translations like the NIV). I actually like Lombardo's reading the of the Iliad for this reason, unlike some of the other members here. Yes he sounds like an American. Would it be an improvement if he sounded Belgian (or pick your favored nationality)?

(A Belgian friend of my acquaintance complains to me that when he visits Montreal, everyone switches to English as soon as they hear his accent. "But French is my native language!")

--

Ancient -> Modern Greek vowel phoneme collapse

a -- ᾰ, ᾱ, ᾳ
(av) -- αυ
e -- ε, αι
(ev) -- ευ
i -- ῐ, ῑ, ῠ, ῡ, η, ει, οι, υι, ῃ
(iv) -- ηυ
o -- ο, ω, ῳ
u -- ου

[rmedinap]

Would it be an improvement if he sounded Belgian (or pick your favored nationality)?

Actually I have no preferred national sound. I like all languages when spoken by their natives. But I'd be much more understanding and forgiving if those that condemned the Modern Greek weren't so vocal and radical with their criticism of the "ungreekness" or "inaccuracy" of the Modern Pronunciation while they themselves are no more accurate.

It's the incoherence what bothers, not their accent. I'm sure that I have a very strong Mexican accent when I speak English or German or French but I don't go around reciting Shakespeare with a restored English pronunciation and nagging innocent English majors about how inaccurate their modern English pronunciation is.

[jeidsath]

"No more accurate" may not be a claim that could be defended in detail. The "restored" pronunciation -- mostly Erasmusian with some fixes introduced in the late 1800s to Germany/England -- really is more accurate in some important ways. If nothing else, it has enough phonemes. And the ability to be able to read Greek quantitive poetry aloud so that you can hear the metre by ear is no small thing, even if very few people seem to be able to do it.

And your example about harassing people about Shakespeare might have been more apt if you had said "Beowulf" -- but also far less persuasive.

[rmedinap]

True, but like I said. I'm not saying we should abandon the restored pronunciation. I'm saying we should learn both pronunciations. First the Modern and then the restored one.

And you can still respect the metre with Modern Greek, it sounds a bit weird having so many long about but it's not impossible. And like you said. What difference does it make if the reconstructed pronunciation has more phonemes if none is going to recite poetry with the metre, cadence and accent that it's supposed to have? Isn't it a valid alternative to do it according to an extant and available pronunciation that you can accurately imitate until you learn to do it "right"?

And I'm not really convinced about the whole unintelligibility argument. French is similar example, its orthography does not always correspond to its phonemes (at least from the perspective of another romance language), and even if it takes some effort people still manage to learn it pretty good. Modern Greek and even sometimes English has the problem that many words are not written the way they sound, there's lots of orthographic rules that do not correspond with pronunciation, some even change depending on the variant you're speaking (like primer: British [ˈpraɪmər] American [prɪməʳ]). Isn't that the reason young children have spelling bees? In Spanish that is completely unnecessary, that's why it takes us much effort to learn the phonetics of English and French but we have little trouble with the phonetics of German or Italian.

[Tugodum]

In 2011 I met a Greek Scholar in a congress, it took me a while to get him drunk enough so that he'd agree to read some Thucydides using the reconstructed pronunciation

Was it anything like this reading of Plato's Apology https://www.youtube.com/watch?v=xcIvDUehsfE ? As far as I know, in the academia in Greece they use, when reading classical authors, Byzantine (=Modern) pronunciation only. It does not surprise me that they have no "unintelligibility" issues, as they perfectly understand each other's everyday speech, in which what is written as "και" is pronounced as [ke], etc.

[Tugodum]

If you're just reading the syllables off the page because you don't understand the language well enough to understand them as you pronounce them, you're going to sound horrible. Anyone who has sat through lay members reading scripture at Church knows what this is like (and I mean native English speakers reading the text in modern English translations like the NIV).

I heard, more than once, lay members reading even King James, and am not sure what you are referring to.

[jeidsath]

If you're just reading the syllables off the page because you don't understand the language well enough to understand them as you pronounce them, you're going to sound horrible. Anyone who has sat through lay members reading scripture at Church knows what this is like (and I mean native English speakers reading the text in modern English translations like the NIV).

I heard, more than once, lay members reading even King James, and am not sure what you are referring to.

Count yourself lucky. μακρά τινος ἀναγινώσκοντος καὶ πρὸς τῷ τέλει τοῦ βιβλίου ἄγραφόν τι παραδείξαντος "θαρρεῖτε," ἔφη, "ἄνδρες: γῆν ὁρῶ."

[Timothée]

I think I’m falling into a trap carefully set. Notwithstanding, a few points:

1) It’s never been a good argument that because we cannot be perfect, we shouldn’t try at all.
2) You are satisfied neither when tones are incorporated to the restored pronunciation nor when they are disregarded. Modern Greek has no tones, either. (E.g. Swedish has two, rising and falling.)
3) There is dialect variation in Modern Greek, as well. If we make this a problem (it really isn’t), we cannot use Modern Greek, either.
4) There is nothing wrong if, when pronouncing Ancient Greek, one can be recognised as, say, a Frenchman, Italian, Spaniard, or Englishman. If this is considered a problem, it won’t be solved with Modern Greek—they’ll still have an accent. This argument allows only native Greeks to pronounce Ancient Greek.
5) I just read for my (dis)pleasure the first 21 lines of the Iliad in Modern Greek pronunciation. I cannot make it scan, simply because thus it cannot be made scan: there isn’t the long and short vowel (and consonant) difference needed in Modern Greek, particularly in non-stressed syllables.

[RandyGibbons]

Great thread. I can say this for an absolute fact: Only two people will ever like the way you read ancient Greek, yourself and your mother. (Unless you're native Greek.)

I was taught Greek 101 in college using Crosby & Schaeffer. I suppose the pronunciation we used was "Erasmian," though the professor wasn't making a point of that. Before going on to graduate school, I read quite a bit of the Iliad on my own. I've always been an advocate of reading out loud, so as I orated impressively before the mirror, I suppose I was doing this in "Erasmian" too. On to graduate school, where, as I recounted in a different thread recently, I had quite a few native Greek friends. They found our pronunciations hilarious (they read ancient Greek the way I read Shakespeare). I learned the rudiments of modern Greek and for a while read my Homer with a modern Greek accent (such as it was). Later, I became a convert to "restored classical." I even tried my hand at pitch accent, but couldn't sustain it. More recently, I have had a friendly debate with a professor friend about how to read Greek poetry. I had bought into the argument that the ancient Greek would have, and we should, read the given accents, and that the underlying vowel quantities by their very nature produce the meter correctly. In other words, in Μῆνιν ἄειδε θεά, accent the initial α in ἄειδε; α is still short, ει long, and these by their nature produce the opening two dactyls. She won the debate on practical grounds. We were actually reading Sophocles, not Homer, and especially in the choral parts I simply couldn't sustain the style of reading I was advocating.

The lesson in hindsight for me: I've been reading Homer all these years in the native language! With so much more appreciation than in translation. Second, I'm grateful for just about anyone else's recordings. I personally still use my feeble attempt at "restored classical," but I'd be a fool to reject Bedwere's delightful recordings just because he happens to use Koine (he's the best contemporary reader I know). As another example, I find it amusing that people find Daitz's recordings "horrible" (do they mean by that "incorrect"?). Initially jarring, for sure. But for all we know, if Ion the rhapsode dropped out of the sky and recited Homer to us, we might find that jarring too. I've spent many a profitable 45-minute session on the elliptical at the gym listening to Daitz, and if someone else thinks they can do it better, or simply differently, I'll welcome that too.

Vive la différence!

[rmedinap]

1) It’s never been a good argument that because we cannot be perfect, we shouldn’t try at all.

Once again you're right. And I'm not saying that we shouldn't try. All I'm saying is that IF the argument against Modern Greek Pronunciation is going to be "inaccuracy", the "ungreekness" and even "impurity" (I heard that mostly from old professors). Then you have to be coherent and avoid sullying Greek with modern "impurities".

2) You are satisfied neither when tones are incorporated to the restored pronunciation nor when they are disregarded. Modern Greek has no tones, either. (E.g. Swedish has two, rising and falling.)

Same as with point one. I'm not dissatisfied with the tones, I'm dissatisfied with the lack of coherence of those that want to hear and speak Greek exactly the way Homer or Plato spoke it and yet neglect the tonality of the language or their own imperfect rendering of the reconstructed pronunciation.

On this point, maybe you've not had shy pupils that are too afraid to mispronounce something to read aloud, even the bold ones would stutter and lose concentration if I became adamant about the accent, that's why I first teach them the modern pronunciation where you do not have this problem, they can practice by themselves and hear some modern Greek on the radio, maybe we read a short Kavafis poem. Once they have some vocabulary and basic grammar under their belts I teach them the restored pronunciation. I explain the reasoning behind it and compare it to the modern one (like the Shakespeare example here mentioned). It's a pedagogical win-win, they learn their basics without the uneasiness of lacking a fluid and secure model, then they learn the restored pronunciation easily because they know the reasons for it, they can compare it themselves whenever they're in doubt, and most of them get rid of their native accent because they have the modern pronunciation as their starting point.

3) There is dialect variation in Modern Greek, as well. If we make this a problem (it really isn’t), we cannot use Modern Greek, either.

Same as with point one. IF accuracy is strong enough a reason to reject the modern pronunciation so decisively, then we should be careful not to recite Sappho or Alcaeus with an Athenian accent.

4) There is nothing wrong if, when pronouncing Ancient Greek, one can be recognised as, say, a Frenchman, Italian, Spaniard, or Englishman.

Once again you're right, but as I said in point one. I'm not arguing against the reconstructed pronunciation, I'm arguing against the "accuracy" or "purity" argument of those that reject the modern pronunciation.

If this is considered a problem, it won’t be solved with Modern Greek—they’ll still have an accent. This argument allows only native Greeks to pronounce Ancient Greek.

Actually I do believe that the only legitimate starting point we have available is the Modern Greek, but you do not have to be a native to do it. And like I said, I'm perfectly willing to turn a blind eye to native accents as long as people don't go preaching about the "impurity" or "innacuracy" of modern pronunciation.

Just like in Latin the valid starting points of comparison are the romance languages, however they may changed phonetically, it's still the closest rendering available to the original. A German friend once told me how jealous he was of my rendering of Cicero or Caesar (even when I used the German pronunciation), it didn't matter how much he practised, how competent a linguist he was or how smarter than me he was, my Spanish gave me an insurmountable advantage that he dind't have. He actually learned Spanish just to read his Cicero more fluently. And that's the thing about dead languages, we do not have a native to teach us the general cadence and fluidness of a living language so we have to adapt and make do with what we have at hand. That's why I find this whole "purity" or "accuracy" argument so unconvincing. It's like being on the Titanic but refusing to get on a boat because its colour does not match your dress.

5) I just read for my (dis)pleasure the first 21 lines of the Iliad in Modern Greek pronunciation. I cannot make it scan, simply because thus it cannot be made scan: there isn’t the long and short vowel (and consonant) difference needed in Modern Greek, particularly in non-stressed syllables.

It does take a bit of stretching vowels that are not meant to be long. I'm sure it'll come if you keep trying in good faith. But even if you don't, this argument goes in my favour. If I have a pupil whom I previously taught only the modern Greek pronunciation, he or she will notice that there's something in there that makes it odd to recite correctly with the pronunciation that he or she knows, I usually take that as a opportunity to teach them the restored pronunciation. Because I can then show them with clear examples that they understand how more accurate would it be to use the restored one.

As a teacher this kind of situations are invaluable for me, I cannot show them a recording of Homer himself and what's available I find it very questionable, the next best thing I can offer in all academic honesty is a comparison of pronunciations where they'll see why it would be more accurate to use the reconstructed pronunciation even for reciting purposes. Maybe they won't like my imperfect rendering or Daitz's but at least they'll know and understand the reasoning for it.

I've been lucky enough to have groups where one half adopted the reconstructed pronunciation and the other the modern one. (I never force anyone to adopt anything, I simply show what's available and let everyone do what they think best). You cannot imagine how well both pronunciations coexisted and how many interesting points of discussion were raised because of the comparison that would have otherwise gone unnoticed.

[Timothée]

You get a good enough approximation pronouncing vowels and consonants as they are in Allen (much much closer than with Modern Greek, and easier). Tones can be disregarded (as they are in Modern Greek pronunciation). Attic pronunciation will do (nor does Mod.Greek pron. differentiate between Homer, Sappho, Herodotus, Plato and Plutarch). But as is clear from your reply, you’re indeed simply trolling. My last post here.

[jeidsath]

I think that everyone here is being serious and are making valid points. This is an easy subject to get heated about, unfortunately, as is appropriate for a subject that has been argued for generations can never go anywhere.

Familiarity with Modern Greek may well have pedological advantages for learning the ancient language, and if rmedinap and his students are finding success with that, wonderful. A number of educators have made the same claims over the years.

And if people can listen to recordings of Homer, Plato and Xenophon, etc., with a Modern Greek pronunciation, and understand them, wonderful. Personally, I'm somewhat skeptical of this, given that I've listened to a fair amount of Modern Greek recordings of these authors, and have found that too much information has dropped out for easy listening. This is due, I think, to iotaization and lost quantities, as I outlined above. But rmedinap may not share that experience, and my arguments may be wrong.

I would suggest, however, that instead of debating with the ghosts of his old Professors, who apparently insisted on purity, etc., that he will engage with the arguments that people are actually presenting here.

[rmedinap]

I do not mind if someone disagrees with me. But I find it remarkably odd that you'd think I'm trolling. I meant every word.

Would you say that people in high academic positions like the quoted Matthew Dillon (“The Erasmian Pronunciation of Ancient Greek: A New Perspective.” The Classical World, vol. 94, no. 4, 2001, pp. 323–334) are also trolling as well?

What about Prof. McNeal (“Hellenist and Erasmian.” Glotta, vol. 53, no. 1/2, 1975, pp. 81–101)?

Maybe I was not clear enough, I'm not saying we should reject one pronunciation in favour of another, I'm saying that a very valid alternative is to learn both, as Prof. Dillon thinks.

Allen himself (page 129 of the 3rd edition of his Vox Graeca) had enough humour to be sarcastic about those that took his suggestions too seriously. This I also find disquieting in most reconstructionists: they're more papists than the Pope. They take Allen's Vox Graeca as if it were divine word. Some of Allen's claims have been questioned by Sven-Tage Teodorsson or even Allen's own pupils like Geoffrey Horrocks (cf. Dillon page 326). In this case I actually side with Allen, the opposing arguments I find unconvincing but worth taking into very serious consideration.

But as jeidsath said we should concentrate on the practicality of what I say. He's right in pointing that if one were to only learn the modern Greek pronunciation much grammatical information would be partially or totally lost in speaking, that's why I argue that we should teach both pronunciations. Modern Greek themselves have a lot of misspellings and orthography problems with old words precisely because of the inconsistency between their phonetics and their written language, yet they still learn to write nicely.

My experience with French has taught me that you can have a language that consistently silences word-endings (even verb endings) but as long as there's some regularity to it you won't have a problem in understanding and speaking it. It's just a matter of practice.

And I really think that we all agree on all the basic facts, I think I was clear enough every time I recognised someone was right. I'm really sorry if I offended anyone I did not mean to.

[opoudjis]

Gentle colleagues, a tangential contribution to this thread.

While on Quora, I often derided some of the attempts at reconstructed pitch accent—and I meant Stephen Daitz' in particular—as "yodelling Martians". You can google the phrase. I was also informed by a Christian friend that there's a whole lot of people pronouncing Koine with a pause. between. every. word, and. no. noone. actually. spoke. like. that. Unless you're Shatner. (Or T'Kuvma on Star Trek: Discovery, I suppose.)

This is not to vindicate the Reuchlinians, at all. (On the other hand, on the rare occasions I'm unfortunate enough to be trying to make sense of Thucydides, *of course* I'm going to mutter the passage to myself in Modern Greek pronunciation. I'm Greek, and I need all the help I can get.)

It is to point out that languages have cadence and prosody and pitch contours, and are still spoke by humans of flesh and blood. And the exaggerated renderings of pitch accent in poetry may have their paedagogical place, but I find it hard to credit that Greek was spoken with such exaggeration.

Not for the last time, I recommend the Greek of Ioannis Stratakis: https://www.podium-arts.com . His readings of poetry are yodelling as well, though certainly not to the extremes of Daitz. But his Herodotus, while sticking to the reconstructed phonology (to the extent of the odd digamma) sounds chatty and fluent and human.

And I don't think the fact that he's a native speaker of Greek is irrelevant to that. Not because Modern Greeks somehow have the key to the kingdom; but because Greeks, by recognising a critical mass of the words and structures (and yes, there are false friends) gain the confidence to rattle it off like a human language. And that counts for something.

Dante has brought up Shakespeare with a Brooklyn accent. I find a nice irony there: Ben Crystal has been plying his wares reciting Shakespeare with a reconstructed Early Modern English, and he's commented how much more visceral it sounds, and how much faster it comes out—trippingly on the tongue.

I don't think that's intrinsic to reciting Shakespeare like a Canadian pirate (which is what reconstructed Early Modern English sounds like to me—the /əɪ/ is quite striking.) I think it comes from *not* reciting Shakespeare in Received Pronunciation: letting go of the reverence for the mellifluous tones, and the notion of declamation as a sacred act—and actually getting on with it. Reciting Shakespeare in any dialect other than the official dialect, I suspect, would have the same effect. Because paradoxically, by using more accurate phonology, Ben Crystal has relaxed his intonation and prosody to something less reverential and more realistic.

And I think the same has happened with Stratakis.

[C. S. Bartholomew]

Never ceases to amaze me how passionate people become about issues like pronunciation. I know my pronunciation is wrong. I don't even get Erasmus right. Nobody is listening so what does it matter. I've listened to any number of erudite scholars reading ancient Greek texts. Some of them are actually pretty intelligible and that raises ones expectations a little bit. A lot of them are not.

[RandyGibbons]

Dear oroudjis, thanks for the link to Stratakis. I agree, I find his reading pleasant (and yes, I think helped by the fact that's he Greek) and indeed pretty much the sound I wish I could get.

But ah hah, you're Greek! Let's say I'm an American (or a Mexican, i.e., a non-Greek) teaching Greek 101 to a college freshman class. Which pronunciation would you recommend I teach?

Randy Gibbons

[Paul Derouda]

I strong agree with opoudjis about recommending Stratakis. My only real complaint with him is the use of background music. I don't find his poetry yodeling at all, quite the contrary – at least not his reading of the Odyssey (https://www.youtube.com/watch?v=MOvVWiDsPWQ). Perhaps it's your Greek background that makes his reading poetry sound unnatural to you? I've listened to quite many attempts at restored pronunciation, and he is the only one until now who, in my opinion, sounds like real language, and someone's native language at that. (Of course, I have very very minor disagreements like "spurious diphthong" ει pronounced as a real diphthong, sometimes he pronounces ϝ, and very occasionally he pronounces a long vowel short, but who cares) In short, he makes reconstructed ancient Greek sound like something that's not only possible but also quite easy and natural. I think it's quite pointless to recommend reconstructed pronunciation as long as people don't have any good models, but with Stratakis we have one.

I also very strongly recommend Stefan Hagel performing Homer (https://www.oeaw.ac.at/kal/sh/). He might have just a tiny German accent, but in addition we get, in his own words "an approach to the technique the Homeric singers used to accommodate melodic principles to the demands of the individual verse, guided by the accentual structure and sentence-intonation of the Ancient Greek language as well as by metrical structures." Homer with lyre and all! (I think Hagel was also one of the voices in the CDs that come with the Assimil Ancient Greek course (in French).)

Bwt, Opoudjis: I checked your Quora post: I think Greek pronounced with English concessions is much worse than either French or German.

[Paul Derouda]

Dear oroudjis, thanks for the link to Stratakis. I agree, I find his reading pleasant (and yes, I think helped by the fact that's he Greek) and indeed pretty much the sound I wish I could get.

But ah hah, you're Greek! Let's say I'm an American (or a Mexican, i.e., a non-Greek) teaching Greek 101 to a college freshman class. Which pronunciation would you recommend I teach?

Randy Gibbons

Note that many actual Greeks would happily crucify Stratakis for he has done! What he does might be easier for a native Greek, but it's definitely still quite different from modern Greek.

[opoudjis]

But ah hah, you're Greek! Let's say I'm an American (or a Mexican, i.e., a non-Greek) teaching Greek 101 to a college freshman class. Which pronunciation would you recommend I teach

I like what Robert Todd says about this question on Quora: whatever you pick, make sure it's internally consistent and practical. That's more important than anything else. Do ensure that your students know what the reconstructed phonology is, because it will help them make sense of the grammar. But it's not a crime to use Erasmian as opposed to Reconstructed when you're reading things aloud. It *is* a crime to use Reuchlinian (modern), but it's a misdemeanour.

[opoudjis]

Note that many actual Greeks would happily crucify Stratakis for he has done! What he does might be easier for a native Greek, but it's definitely still quite different from modern Greek.

And notice how careful he is to say that he worked out the reconstruction himself—i.e. that he's not being taught how to speak his own language by a Frank [Westerner] like Allen. I strongly suspect he feels he has to say that precisely in order to anticipate criticisms from Greeks. (Which are relentless, and in the case of YouTube... well, in the case of YouTube, the criticisms are of the caliber you'd expect from YouTube.)

[RandyGibbons]

Thank you oroudjis and Paul. Σύμφημι. And thank you, Paul, for making the connection to Stefan Hagel and the Assimil program. When I was resurrecting my Greek a few years ago, I enjoyed the (partial) recordings that accompany the Athenaze and Reading Greek programs, but most liked and tried to emulate the readings on the Assimil CDs.

I especially want to ditto the advice, whatever you pick, make sure it's internally consistent and practical.

[Tugodum]

it's not a crime to use Erasmian as opposed to Reconstructed

My apologies for being so slow: what is the difference between the two?

It *is* a crime to use Reuchlinian (modern)

Are you being sarcastic?

[jeidsath]

It depends where you learned your Erasmian from. There has been practically no difference at good schools in the UK since 1900. But it can be very hard to get some features right. Everywhere else is the wild west.

I don't share everyone's love for Stratakis -- I think that Lombardo is much more comfortable with his Iliad reading, though it's a shame that Lombardo murders so many of the long vowels. Lombardo also ignores word accent, which is a shame, but less of a shame than vowel length.

Stratakis regularly makes short vowels long when accented. He also tends to sound out words of more that 4 syllables. His intonation doesn't seem to indicate that he's reading ahead as he reads, or is especially familiar with the text/language. (I'm referring entirely to his free Herodotus video on Youtube, which I have just listened to, as it was specifically mentioned above.) I don't get the impression that he could look up from the text, and repeat the last couple of phrases/sentences.

However, he's obviously working on all of the right things as far as pronunciation, and I'd like to hear him again as he improves.

[Paul Derouda]

Stratakis regularly makes short vowels long when accented. He also tends to sound out words of more that 4 syllables. His intonation doesn't seem to indicate that he's reading ahead as he reads, or is especially familiar with the text/language. (I'm referring entirely to his free Herodotus video on Youtube, which I have just listened to, as it was specifically mentioned above.) I don't get the impression that he could look up from the text, and repeat the last couple of phrases/sentences.

Des goûts et des couleurs on ne discute pas - it's a matter of taste, I suppose. But I haven't noticed his accented short vowels being long. Mind you, it's quite possible that you're right, but that's not necessarily a problem. Finnish, my mother tongue, has distinctive vowel length both in accented and and unaccented vowel. Phonologists tell me that accented syllables (the main stress is on the first syllable in Finnish) are always longer than unaccented, whether long or short, but this is something that I'm unable to distinguish (but it might be easier for you, whose ear has not been conditioned similarly). At least in Finnish, phonemic vowel length is about relative length, not absolute measured duration, and the absolute true duration of short accented vowels is sometimes as long as that of long unaccented vowels. So, although Finnish has stress accent and ancient Greek had pitch, I think it's quite conceivable that something similar occurred in Greek as well.

Note that the Herodotus recording is almost 4 years old, and he's probably made further progress since then.

As for sounding out words, even natives do that - and besides, poetry is a lot easier than prose, where the meter makes it much easier to find a natural rhythm. Being able to read long stretches of prose naturally is, in my opinion, the ultimate test of reconstructed pronunciation. Stratakis isn't perfect I know, but he has certainly passed the Turing test, he could fool almost anyone into thinking that he's actually reading in his native language.

[RandyGibbons]

jeidsath makes my point: The only two people who will ever like your reading of (ancient) Greek are yourself and your mother!

[ἑκηβόλος]

I started undergraduate studies with a different pronunciation for Classical and Modern Greek, but ended up reading Classical Greek in Modern Greek pronunciation at some point during the 4 (or 5) years of the double major. Reading seperately or with a single pronunciation didn't make a big difference in terms of understanding at the beginning of Modern Greek studies, when the language was foreign, but as I acquired Modern Greek through practice there was as difference. It guess at the point when Modern Greek became the dominant lexical semantic reference in the vocabulary matrix in an emerging sub-coordinate L2 bilingualism language pair, it was a natural time to switch over and use the Modern pronunciation for other periods of the language too. Now, I tend to use Buth's system of pronunciation for pre-Modern Greek and Modern for Modern.

Changing pronunciation systems requires different strategies for changing the semantic lexical part of the word, than for changing the other morphological elements. In other words, it is not only about changing the phonemes, but the associations to experience and encylopedic knowledge that accompany lexical semantic knowledge need to be rewired to the new sounds on the one hand, and the other mophological - inflectional - components that are a mixture of automaticity and logic also need to be rewired in their own syntactic and discourse situations. Seemingly random errrors that occur in the change of pronunciation system are usually limited to the parts of the word representing the semantic data set or to those parts of the morphology conveying the relational meaning - it is not as simple as simply using new sounds.


Joannes,

About your initial question....

Just to make a quick practial suggestion for getting together a bulk of texts you can listen to (and follow along with reading too perhaps)

I have been playing synthesised speech as a way to guide me into reading texts, and as a counterpoint to text my progress in reading them for some time, but now with the release of the recent addition of Greek to the Google text-to-speech APP that produces output that sounds okay, that process has become a whole lot more meaningful.

Not having a better knowledge of the Android system, the steps that I follow to get an output are a little convoluted, but produce a workable result.

Step 1 - Conversion of polytonic to monotonic

• There is probably a better way, but at present:
  The polytonic unicode text is decomposed into Beta Code.
  The accents are changed to their monotonic counterpart and unwanted diacritics are removed
  The text is recomposed to unicode.

Step 2 - Use of the TTS

• The only way I know of doing this is in Google Translate. It seems that (perhaps to same space and processing power) lesser spoken languages don't "translate" Greek input text in any way - ie. the translated monotonic Classical Greek isn't modified in anyway to Modern Greek. Then the (unmodified) "translated" text can be played.

Step 3 - Recording of the output

• I record it using a screen capture APP. I haven't worked out how to mute the microphone on the phone without also muting the TTS output that goes to the screen capture APP, so I physically put my thumb over the in-built mic.

The method is currently cumbersome, and the result is just adequate.

[RandyGibbons]

Johannes,

A belated thanks for your references to R.A. McNeal's Hellenist and Erasmian and Matthew Dillon's The Erasmian Pronunciation of Ancient Greek: A New Perspective. I only just came to realize that one can register an account with JSTOR and get up to three free articles at a time on one's "shelf." With that realization and a weekend of otium, I just finished those articles and learned a great deal about Erasmus's dialog between the Bear and the Lion!

Dillon concluded,

Few classicists who take the trouble to learn a little Modern Greek will be content to pronounce Ancient Greek precisely as Modern. We must compromise to produce something like "la lingua antica in bocca moderna." ... My own efforts are directed to what I would tentatively call "Hellenistic" pronunciation.

I did note that per above quote, Dillon wasn't advocating the use of Modern Greek but Modern Greek as a starting point (you were clear about this, I wasn't). His argument that that is the only way to capture what Erasmus called the συνέπεια of speech (the flow of speech, or sermo in Erasmus's title), beyond one's chosen way of pronouncing the individual phonemes, is definite food for thought. Thanks again for the references.

Randy Gibbons

[RandyGibbons]

correction: I think credit for the references to Dillon and McNeal go to you, rmedinap.

[daveburt]

Step 1 - Conversion of polytonic to monotonic

• There is probably a better way, but at present:
  The polytonic unicode text is decomposed into Beta Code.
  The accents are changed to their monotonic counterpart and unwanted diacritics are removed
  The text is recomposed to unicode.

Step 2 - Use of the TTS

• The only way I know of doing this is in Google Translate. It seems that (perhaps to same space and processing power) lesser spoken languages don't "translate" Greek input text in any way - ie. the translated monotonic Classical Greek isn't modified in anyway to Modern Greek. Then the (unmodified) "translated" text can be played.

Step 3 - Recording of the output

• I record it using a screen capture APP. I haven't worked out how to mute the microphone on the phone without also muting the TTS output that goes to the screen capture APP, so I physically put my thumb over the in-built mic.

I've borrowed your main idea here (monophonize then TTS) and automated it for a project I'm working on. A link to the code is below, but the process is basically:

• Decompose Unicode (NFD)
• Delete breathing marks (I haven't yet implemented this, though it is not difficult)
• Convert all the remaining Combining Diacritical Marks into \u0301 (acute accent)
• Feed the result into `espeak` (free command-line software) to produce Greek audio files

https://github.com/dburt/duff_graded_gn ... uiz.rb#L16

[ἑκηβόλος]

I've borrowed your main idea here (monophonize then TTS) and automated it for a project I'm working on. A link to the code is below, but the process is basically:

• Decompose Unicode (NFD)
• Delete breathing marks (I haven't yet implemented this, though it is not difficult)
• Convert all the remaining Combining Diacritical Marks into \u0301 (acute accent)
• Feed the result into `espeak` (free command-line software) to produce Greek audio files

https://github.com/dburt/duff_graded_gn ... uiz.rb#L16

Try eSpeakNG. It seems to allow a degree of manipulation of the prosody. Prosody can be important in bringing a reading to life.

An example of that might be that the μήτι of Matthew 12:23 Μήτι οὗτός ἐστιν ὁ υἱὸς Δαυίδ; and the μή τι of John 21:5 μή τι προσφάγιον ἔχετε; even they may both have an interogative intonation, but in espeak (or another synthesising engine without SSML support) I guess they would sound the same. There is an element of the arbitrariness and subjective personal preference that has been discussed in this thread in assigning prosodic features to the same word in different syntactic functions, but as artistic as it may be, it really does help with the general level of intelligibility when a text is read. To further the μή example, it may not be the right interpretation of the differences, but μή with falling tone followed by a short space before participles, non-purpose infinitives, non-purpose subjunctives and indicatives, μή with a falling tone followed by a long pause for purpose infinitives, negated purpose subjunctives and imperatives, μή with a rising tone and a long pause when it introduces a question constitute a useful matrix of prosodic features which on the face of it seems to improve intelligibility. Another point of difference where pre-programmed Modern Greek prosody option such as eSpeak (original) or proprietry might not be as good as one that can programmed (perhaps offering SSML support) is relative pronouns. Modern Greek speakers seem to consistently "suffix" the relative pronoun (που) to the preceeding nominal element. That might not be the case for texts from the earlier periods. Basically, even the best Modern Greek pronunciation models are only going to work improperly for the Greek of an earlier period, because there are syntactic elements, which will not be recognised, and hence have no strategy for treatment in the decoding algorithm.

Don't forget the vertical stroke - iota subscript in addition to the breathings.

A point of difference in monotonic greek, as compared to polytonic Greek also comes up in one of those examples Μήτι οὗτός ἐστιν ὁ υἱὸς Δαυίδ; where οὗτός has two accents in the polytonic writing, but in the monotonic, it seems to only be written with the word's "own" accent. Where the word is going to be stressed when there is a choice of two will need to be a feature of your monothphongalising algorithm.

[daveburt]

Thanks for the tips. I'll check out espeakng. Aren't iota subscripts irrelevant to pronunciation? And the duplicated accents I threw at espeak in my testing seemed to come out OK for the low bar I set for it, but if I start paying attention to prosody, the bar will be getting higher!

[opoudjis]

Thanks for the tips. I'll check out espeakng. Aren't iota subscripts irrelevant to pronunciation? And the duplicated accents I threw at espeak in my testing seemed to come out OK for the low bar I set for it, but if I start paying attention to prosody, the bar will be getting higher!

By the time of the New Testament, if that is your interest, colloquial speech *probably* was like Modern speech in the following critical ways:

• No distinctive vowel length
• Therefore, no three-mora rule for enclitics: the οὗ- in οὗτός ἐστιν was no longer or shorter than the μέ- in μέρος ἐστιν, so that spelling with two accents was only historical. The three-mora rule would already have been substituted by the three-syllable rule that applies to Modern Greek: ο άνθρωπός μου (cf. ἄνθρωπός τις).
• Also, no pronunciation of iota subscript any more; that seems to have already disappeared towards the end of the Classical period.

Of course, ἑκηβόλος is also quite right about the difference in syntax and in function word meaning between Koine and Modern Greek affecting prosody; for example, Modern Greek που (derived from Ancient ὅπου, and subject of my doctoral thesis) as a relative pronoun is a proclitic, whereas the cognate ancient που is an enclitic; so they will be run in to different words (following vs preceding).



eSpeakNG, eh? I'll check it out. espeak was actually too Koine for me.

[jeidsath]

The theory that a demotic Greek with most of the features of modern Greek existed in the first century (or earlier, according to some) has been criticized. The positive evidence (spelling variation in papyrus and inscriptions) is not cut and dried. And it pushes thousands of years of change into a very short period. Most telling, to me, is that the ability to write quantitative poetry with few errors seems to continue for centuries. While an educated, Atticizing Greek, was a real phenomenon, it's hard to imagine an educated class recovering the prosaidic features of Attic Greek to the extent necessary to do write quantitative poetry without errors. They certainly weren't able to do that in later centuries, despite retaining a very high competency in Attic Greek. It's far easier for me to imagine dialectical variants of Greek that retained vowel quantity.

However the point about the long dipthongs having merged before the first centuries A.D. is more likely. The grammarians claim the same.