Learning ancient-greek with modern-greek pronunciation?

Greek Learning Greek
61

Not a problem. I’m happy to discuss :slight_smile: The question of quantitative verse is indeed puzzling. I’ve been studying it for some time. Initially I was of the opinion that quantitative distinctions were phonemic in Ancient Greek, but now I’m not so sure. If we assume that quantitative poetry was based on features present in actual speech (i.e. vowel length) we come up against some bizarre phenomena.

For example, take syllables long by position between adjoining words. Compare the following:

ὃς ἤθελεν

ὃς θέλει

In the first instance, ὃς is short, in the second it’s long, even though in both cases there’s the same word boundary. Are we really to believe that when an Ancient Greek said ὃς θέλει, he lengthened the ο in ὃς in anticipation of the θ in θέλει, but when he said ὃς ἤθελεν he left the ο short?

The same can be said about correption:

θέλω εἰπεῖν

θέλω λέγειν

When conversing, did the Greeks really shorten the ω in anticipation of the ε in εἰπεῖν in the first case, but leave it long in anticipation of the λ in λέγειν, even though the word and word boundary is identical?

Or what about the so-called δίχρονα? According to the rules, the ο in ὄψις can be either long or short because it’s followed by the compound sound ψ. Well, which one was it? Did the Ancient Greeks sometimes say ὄψις with a long ο and other times with a short ο? By what mechanism does ψ condition this?

The same applies to the letters I, Y, and A: sometimes length is predictable, sometimes in varies. For example, the A in 'Aπόλλων can be either long or short. The α’s in ἀθάνατος can be whatever the poet chooses. Ιn the Iliad 1.583 Homer uses the word ἴλαος, with a long ι and α, but in Iliad 9.639 with a long ι and a short α. If this is simply poetic licence, why doesn’t it happen also with η or ω? I’ve never heard of any language where variable vowel-length is associated to specific vowel sounds.

Another problem is typological. Languages which have phonemic vowel length tend either to have a very weak accent (like Sanskrit or Japanese), or a fixed accent (like Hungarian and Finnish, where it’s on the first syllable, or Farsi, where it’s generally on the last). As far as we can tell, though, accent was a central and very prominent element in the phonology of Ancient Greek. What’s more, it’s placement is unpredictable based on the underlying vowel-length. A word like λεπτός has a long penultimate and a short ultimate, but it’s stressed on the ultimate, while a word like κέρδος, with the exact same length structure, is stressed on the penultimate.

Then there’s the whole question of the tri-syllabic rule: the accent cannot fall further back than the antepenultimate syllable. For example, one says ὄνομα but ὀνόματα. This rule applies regardless of the length of the word so that a monstrosity like ἐννεακαιεικοσικαιεπτακοσιοπλασιάκις (Plato Republic 9, 587E) has only one accent. This would suggest a syllable-timed language.

Of course, there is then the issue of the long ultima which affects the accent placement. One says ὀνόματα but ὀνομάτων. The way this is usually explained is through morae: a long vowel counts for two morae, a short vowel for one. The accent is supposed to fall on the third-to-last mora.

But this scheme also raises problems. Why then is ἄνθρωπος stressed on the fourth-to-last mora? Or words like ἔχουσιν, πραότητι, κάθημαι, δέσποινα, etc? Why not ἐχοῦσιν, πραοτῆτι, καθῆμαι, and δεσποῖνα?

If vowel length was such an important element to Greek phonology, how did it come to disappear at the height of Classical culture?

Finally, there’s the fact that the ancients did not conceive of their poetry in terms of simple alternations between long and short vowels, but as a rhythmical pattern. And this rhythm was understood as akin to accent.

Aristotle , Rhetoric:

It is natural for us to imitate both harmonies and rhythms – for it is clear that the meters are a part of rhythm. (1448b)

Now all things are limited by number, and the number belonging to the form of diction is rhythm, of which the metres are divisions. Wherefore prose must be rhythmical, but not metrical, otherwise it will be a poem. (1408b)

Longinus, “Prolegomena” to Hephaestion’s Treatise on Metrics:

For the father of meter is rhythm and god: meter derived its beginnings from rhythm, while god articulated it into being.

Aristoxenus, On Harmony:

Starting from these definitions and classifications we must seek to indicate in outline the nature of melody. We have already observed that here the motion of the voice must be by intervals; herein, then, lies the distinction between the melody of music and of speech — for there is also a kind of melody in speech (λογῶδες τι μέλος) which depends upon the accents of words, as the voice in speaking rises and sinks by a natural law. (1.18)

Quintilianus Aritides, De Musica:

…τὸν μὲν ῥυθμὸν ἐν ἄρσει καὶ θέσει τὴν οὐσίαν ἔχειν, τὸ δὲ μέτρον ἐν συλλαβαῖς καὶ τῇ τούτων ἀνομοιότητι. (1.23)

Tαύτης [τῆς μουσικῆς] δὲ μίμημα λέξις…αὕτη δὲ ὀξύτητάς τε καὶ βαρύτητας προσλαβοῦσα μετὰ διαστημάτων συγκεχυμένη μὲν ἐγέννησεν ἁρμονίαν, λόγοις δὲ τοῖς συμφώνοις τεταγμένη ῥυθμόν. (2.7)

Plato, Laws

The order of motion is called “rhythm,” while the order of voice in which acute and grave are blended together is termed “harmony,” and to the combination of these two the name “choristry” is given. (665A)


Interestingly, in Hungarian and Finnish which possess phonemic vowel length, the oldest folk songs and poems are accentual, not quantitative-based.

All this to say that I have no idea what is going on and that Ancient Greek meter is a profound mystery :wink:

62

Hi Evangelos96, I’m entering this conversation late and I’m not an expert in the area at all. I’ve started working through your arguments in the thread from the beginning (I haven’t looked at the others yet), and would appreciate your thoughts on the below:

Argument 1: “Take a look at these Imperial-era mosaics (1st to 3rd centuries AD) … Ἀλκιβιάδης is spelled Ἀλκηβειάδης, Ἴκαρος is spelled Εἴκαρος, and Δαίδαλος is spelled Δέδαλος. So clearly, η = ι = ει and αι = ε”.

I see your argument 1 as an enthymeme with a suppressed premise 1.2:

1.1 (premise): The vowel η and digraph ει were substituted for the classic spelling ι, and the vowel ε for the classic spelling αι, in mosaics made at a certain time.
1.2 (suppressed premise): If vowels or digraphs were substituted for a classic spelling at that time (i.e. when the mosaics were made), they were homophones at that time.
1.3 (conclusion): Therefore, the vowel η and digraph ει were homophones of ι, and the vowel ε of the digraph αι, at that time (i.e. when the mosaics were made).

1.2 appears to be needed for the inference in 1.3, but it is a questionable premise. Threatte, in The grammar of Attic inscriptions (1980), volume 1: phonology notes: ‘it is often difficult to decide whether a spelling anomaly is due to some purely mechanical cause, i.e. a graphic mistake such as a simple copying error or accidental omission, or whether there is in fact some phonological basis for it. Again, if phonological causes can reasonably be assumed, their precise nature may be disputable, or the implications of certain spellings as evidence for the pronunciation of contemporary Attic may be highly debatable’ (p. ix). It seems to me that this caution would apply equally to trying to decide homophones from the mosaics linked above.

I think the conclusion 1.3 therefore needs separate support, as premises 1.1 and 1.2 don’t support the inference in 1.3.

Argument 2: " What about the diphthong OI? According to Thucydides, when plague befell Athens during the Peloponnesian War, the people remembered an old oracle. However, they were unsure if the oracle had in fact predicted a λοιμός (plague) or a λιμός (famine): http://www.perseus.tufts.edu/hopper/tex … apter%3D54. If οι was sounded as two distinct vowels, then there never would have been any confusion."

I read your argument as an enthymeme with a suppressed premise 2.2:

2.1 (premise): There was a dispute at a certain time as to whether an oracle had predicted a λοιμός or a λιμός.
2.2 (suppressed premise): If there was a dispute at a certain time as to whether an oracle predicted X or Y, then X and Y were homophones at that time.
2.3 (conclusion): Therefore, λοιμός and λιμός were homophones at that time (i.e. when the people disputed).

2.2 is, however, questionable: it leaves out another explanation, that the people’s memory simply failed; this alternative interpretation might be supported by the following wording in the passage: οἱ γὰρ ἄνθρωποι πρὸς ἃ ἔπασχον τὴν μνήμην ἐποιοῦντο. Another issue is that the conclusion 2.3 might even run counter to what Thucydides says in the next sentence of the passage: ἢν δέ γε οἶμαί ποτε ἄλλος πόλεμος καταλάβῃ Δωρικὸς τοῦδε ὕστερος καὶ ξυμβῇ γενέσθαι λιμόν, κατὰ τὸ εἰκὸς οὕτως ᾄσονται: this seems to imply that the verse would be recited differently if λιμός were substituted, but how could they be recited differently if they were homophones? The quantity is locked in through the verse form, and so only the quality could differ, but this complies that they are not homophones: I think the words οὕτως ᾄσονται here are problematic for your argument. The conclusion in 2.3 would therefore need further argumentation in order to stand.

Argument 3: “The great antiquity of this monophthongal pronunciation of οι is further proved by Homer, who uses the Ionic word ξυνός, meaning κοινός (common). The fact that this variant existed means that οι was homophonous to υ.”

I read your argument as containing two suppressed premises and a suppressed sub-conclusion:

3.1 (premise): Homer used ξυνός, which has the same meaning as the word (which Homer does not use) κοινός.
3.2 (suppressed premise): If Homer used a word X which has the same meaning as word Y (which Homer does not use), then X and Y were homophones at [Homer’s time? The later time when word Y is used?].
3.3 (suppressed sub-conclusion): Therefore, ξυνός and κοινός were homophones at [Homer’s time? The later time when word Y is used?].
3.4 (suppressed premise): If two words were homophones at a certain time, then their corresponding syllables were homophones at that time.
3.5 (conclusion): Therefore, οι and υ were homophones at [Homer’s time? The later time when word Y is used?].

Premises 3.1 and 3.2 seem questionable. On 3.1, the range of meanings of ξυνός (according to Cunliffe: common, impartial) is not the same as the LSJ entry for κοινός, and a quick check of Chantraine’s etymological dictionary has these words under different entries: I’m not sure whether you’re saying that they are the same word spelled differently at different times, having the same sound? 3.2 would lead to extremely odd results where Homeric and non-Homeric dialects use different words for the same meanings. Also, since Homer did not use κοινός, I can’t work out which time you are referring to when these were homophones: at the time that Homer used ξυνός, or at the time that others used κοινός (or both?).

Grateful for your thoughts on the above before I look at the remaining arguments: if I’ve accidentally misrepresented your arguments, it’s not because I am trying to argue for a certain position in this debate (I don’t mind how people pronounce ancient Greek nowadays), but due to my own carelessness. Many thanks in advance.

Cheers, Chad

63

Evangelos, thanks for the reply, and I’m sorry if I failed to make the question clear. I was asking about vowels, not about syllables, a linguistically elementary distinction that you appear to ignore or misapprehend.

For instance, you start by comparing ὃς ἤθελεν with ὃς θέλει, and rhetorically ask
“Are we really to believe that when an Ancient Greek said ὃς θέλει, he lengthened the ο in ὃς in anticipation of the θ in θέλει, but when he said ὃς ἤθελεν he left the ο short?”
The answer, of course, is No. In both cases, as in all cases, omicron is short. It’s the syllable that’s affected by its local environment, not the vowel itself.

You then move to correption. Comparing θέλω εἰπεῖν and θέλω λέγειν, you ask
“When conversing, did the Greeks really shorten the ω in anticipation of the ε in εἰπεῖν in the first case, but leave it long in anticipation of the λ in λέγειν, even though the word and word boundary is identical?”
This time the answer is basically Yes. The ω of θέλω εἰπεῖν was shortened by virtue of its juxtaposition with the immediately following vowel (a diphthong in this case, ει- rather than ε-). Correption is a well attested phonological phenomenon, not exclusive to Greek, and you must know that adjacent vowels are liable to be affected across word boundaries. (Compare elision and crasis.)

Next you ask “what about the so-called δίχρονα?” and very oddly proceed to adduce ὄψις, where there is no question of a δίχρονον. You write
“According to the rules, the ο in ὄψις can be either long or short because it’s followed by the compound sound ψ. Well, which one was it? Did the Ancient Greeks sometimes say ὄψις with a long ο and other times with a short ο?”
This is hopelessly muddled. Nobody says the ο in ὄψις can be either long or short, or if they do, they’re equally muddled. The ο is consistently short (while οψ is consistently heavy). This is like the first instance above. Again you’re failing to distinguish vowels from the syllables that house them.

I’ll leave it at that. —No, not quite. In light of the rest of your post I’ll add a recommendation to read W.S.Allen’s Accent and Rhythm, and Devine and Stephens’ The Prosody of Greek Speech, and some of the extensive body of work largely in Italian on Aristoxenus’ Ῥυθμικά and the interrelation of meter and rhythm. In the meantime I’m afraid your position simply does not merit consideration. So I have to stand by my post of four years ago.

Sorry to step over Chad’s intervening post.

64

How the syllable is perceived by the ear at least. This was my only real complaint about the otherwise very enjoyable videos posted by Evangelos earlier (in this thread or another, I forget). They simply lengthen the vowel in the long position every time, and it really sticks out and levels away the prosody. Fix just this, please, and any other complaint about them would be quibbling. I am sure that there are plenty of educated Greeks who can scan, and who could simply go through and mark up the text for the performers.

65

Joel of course I mean in aurally perceptual terms, not in measurably acoustic ones. It’s all relative. In verse the perception is more strongly regulated by the conceptual conditioning: it conforms to the abstract metrical template. (This is the answer to the διχρονα problem.) Ordinary speech will have been less stringently constrained, but still, as you know, there’’s ample reason to think that the ο/ω distinction held good for quite a long time and was only gradually eroded. In the videos (which I haven’t listened to) apparently the short/long differentiation has shifted from syllable to vowel, hence Evangelos’ confusion I imagine.

66

Hi Chad, thank you for that very interesting and thorough comment. As with all things there are assumed premises :slight_smile:. I tried to base my conclusions on the most parsimonious solution and taking into account the complete context of the corpus. When you look at things individually, there might be some doubt, but when the same phenomenon reoccurs repeatedly the evidence starts pointing a certain direction.

Argument 1. I agree that it is possible that irregular spelling may be due to a “simple copying error or accidental omission.” However, such copying errors would not show the regularity which they do if there were not an underlying phonological cause:

  • H and EI alternate with I from the 6th century BC in Attic inscriptions. There are even a few cases on vase paintings of E alternating with AI, e.g. ΠΙΑΙ for ΠΙΕ. In contemporary Boetian inscriptions (just across the border), ΑΙ was commonly substituted by H, which stood for the old E. For example, KAI was commonly written KH.

  • Plato (Cratylus 418 B-C), says that the older generation used to pronounce the word ἡμέρα with an ι as ἱμέρα, but then people changed the sound to an ε (ἑμέρα), before finally changing it to an η in his day (ἡμέρα). Given that it is unlikely that the pronunciation of the word changed 3 times in such a short time-span, the confusion suggests that the old sound ε (spelled η after the orthographic reform of the archon Euclid in 403 BC) was undergoing or had already undergone a shift to ι by Plato’s time.

  • In the Hellenistic papyri, ι for η and ει, and ε for αι are extremely frequent.

  • In classical Latin, Greek EI is consistently transliterated as I, e.g. Ἀριστείδης → Aristides. In the Septuagint (3rd century BC), Hebrew E is often transliterated as AI, and I as EI: Bethel → Bαιθήλ (Genesis 12:8), Teman → Θαιμάν (Genesis 36:15), Pithom → Πειθώμ (Exodus 1:11).

  • Dionysius of Halicarnassus and Strabo (1st century BC) transliterated the Latin names Titus Herminius, Aricini, and Scipio as Τῖτος Ἑρμήνιος, Ἀρικηνοί, and Σκηπίων respectively. Σκηπίων is actually the standard form found in Plutarch (1st-2nd centuries AD), Lucian (2nd century AD), and Cassius Dio (3rd century AD).

-There is also the evidence of the grammarians. The earliest surviving orthographical treatise we have are the Ἐπιμερισμοὶ of Aelius Herodian (2nd century AD), which is essentially a word-list organized by homophonous sounds. For example, words beginning with ΣΠΙ, ΣΠΗ, and ΣΠΕΙ are treated together, as are words beginning with TE and TAI:

https://books.google.ca/books?id=0qBEAAAAMAAJ&printsec=frontcover&source=gbs_ge_summary_r&cad=0#v=onepage&q&f=false

According to Didymus Chalcenterus (1st century BC) the words λητουργεῖν and λιτουργεῖν differ in that “the [first] is (written) with an η, and the [second] with an ι” (quoted by Ammonius in his book Περὶ ὁμοίων καὶ διαφόρων λέξεων). If the η in λητουργεῖν and the ι in λιτουργεῖν were pronounced differently, Didymus’ gloss on the spelling would be redundant, like saying “Tell with an e, but till with an i.” What he is saying sounds more like “Cell with a c, but sell with an s.”

Now, even though Herodian and Didymus are not exactly classical authors, their testimony is of great value. The Hellenistic Κοινὴ they spoke was a direct offshoot of 4th-century-BC Attic. They read, compiled, discussed, and commentated upon classical authors, many of whom we don’t even have access to today. They had a much greater sense of continuity to the Classical period than we do. Furthermore, their age was characterized by a relatively high literacy rate, a standardized language, and dense communication networks between regions of the broader Greek-speaking world. So I highly doubt that there was room for a radical sound shift between the 4th century BC and their times. The shift must have happened before (cf Plato’s comment in the Cratylus above).

Therefore, seeing that H, EI = I and AI = E at the beginning of our period of investigation, at its end, and in between, I think we can be fairly confident that the sounds were homophonous throughout.



Argument 2. In and of itself, the Thucydides episode is not conclusive. However, there is good evidence to suggest homophony. Firstly, it is clear that the oracle was transmitted orally, since if it was written there would have been no room for confusion to begin with. Now, how does phonetic confusion arise?

Clearly, there must be some similarity between the words. One cannot say λοιμός and some other guy hears συμφορά. Let’s say I say the sentence “The truth is said.” It’s possible that someone might mishear this for “The truth is dead.” Even though “said” and “dead” are spelled differently, phonetically they differ in only one sound, namely s/d. How do λοιμός and λιμός differ? The only difference is the vowel (OI vs I). So that’s where the confusion had to occur.

We know from contemporary inscriptions that OI did in fact alternate with I, e.g MIΡΑΙ for MOIΡAI and ΠΕΡΙAΛΙΦΗΝ for ΠΕΡΙΑΛΟΙΦΗΝ. Furthermore, before labials (like the μ in λοιμός) the vowel I seems to have exhibited rounding. For example, μόλιβος is written with an ι in Homer but with an υ in Attic (μόλυβδος). The same applies to βίβλος, which was also written βύβλος. This phenomenon is confirmed in the Hellenistic papyri, where ἡμεῖς frequently alternates with ὑμεῖς (Geoffrey Horrocks, Greek: a History of the Language and its Speakers, 2010, 2nd edition, p. 118 ).

The exact same phenomenon occurs in Latin. There is a series of words whose orthography varies between an i and a u before a labial, e.g. documentum/docimentum, lacruma/lacrima, maxumus/maximus, lubet/libet. According to the Latin grammarians, this sound, termed sonus medius, had an intermediate value between u and i analogous to the Greek ypsilon (υ). Τhe sound was deemed important enough for Emperor Claudian to even devise a separate character to represent it: Ⱶ

Thus, if we assume that λοιμός and λιμός were both pronounced [lymos], they fit seamlessly into this larger pattern.


Argument 3. Robert Beekes, in his Etymological Dictionary of Greek (2010) derives both κοινός and ξυνός from a hypothetical Indo-European form *kom-ió. This derivation requires several assumptions (metathesis of i, m → n, o → u). The other option, which I am proposing, is that they were simply spelling variants in Ionic. Given that we know that οι and υ historically did coalesce, this proposal does not require any linguistic acrobatics. It only asks us to assume that oι had become [y] already in the Archaic Age, at least regionally.

(for the epenthetic σ in ξυνός, cf. μικρός/σμικρός)

67

Hi guys. So I’ve compared the musical renditions of the songs to the extant musical scores we have and as far as I can tell, the lengthened vowel in the long position actually seems to be a feature of the original notation. You can see this clearly for example in Pindar’s Pythian Ode, the epitaph of Seikilos, and Mesomedes’ hymns:

https://en.wikipedia.org/wiki/Mesomedes

And that’s what puzzles me. A phrase like ὅς μέγα and ἤθελον are both dactyls. The ο in ὃς is long even though there’s a word break after it. So it would have to be lengthened in anticipation of the following consonant. This seems unnatural to me. I’m not being dogmatic about this or anything, just stating my impression.

Also, no one addressed the typological and moraic problems I raised above and why A, I, and Y can be both long and short. I have never heard of a language with variable vowel-length tied to specific vowel-sounds.

For those who are interested, here are the links Joel mentioned:

https://www.youtube.com/watch?v=tkPWehao2kk

https://www.youtube.com/playlist?list=PLj-o4itdS1TYj8UpLfF1L_kOX0r-a0WDt

This article has a very interesting discussion on meter and rhythm I thought I would share:

https://www.jstor.org/stable/745886?seq=1

68

Evangelos, you should read carefully mwh’s post above. You clearly do not understand the difference between the length of vowels and the length of syllables.

Other than that, there are many other things that you have muddled.

For one, Hellenistic papyri start from the 3rd century BC and most are much more recent than that. I don’t think anyone denies that in them, “ι for η and ει, and ε for αι are extremely frequent”. However, that hardly provides evidence for how these vowels were pronounced in Athens several centuries earlier.

The Ionian alphabet, also called the Euclidean alphabet, from which the modern Greek alphabet ultimately derives, was officially adopted in Athens in 403 BC, although it was used unofficially to some degree already somewhat earlier. Before that, they used the Attic alphabet. In this alphabet, the letter Η represented the consonant /h/ and there were no letters for the vowels η ει ου ω. Instead, the sounds ε η ει were all three represented by Ε, and ο ου ω were represented by Ο. Basically, this meant that the old Attic alphabet didn’t distinguish between short and long vowels in writing, for ANY VOWEL, although these sounds were different in speech (even if you’re going to deny it). It doesn’t make any kind of sense that H and ΕΙ alternate with I in 6th century BC in Attic inscriptions, because H was used for the consonant /h/ and EI for long closed /e/ didn’t exist yet (or for /i/, according to your reasoning?).

"According to our oldest sources, the Phoenician Alphabet was introduced into Greece in 1,400 BC. "
Where did you get that? The general view is that the Phoenician alphabet was adapted for Greek around 800 BC, and the earliest inscriptions in alphabetic Greek writing come from about then. In 1400 BC and for a couple of centuries more Greek was written with a syllabic script called Linear B. A very similar syllabic script called the Cypriot syllabary was used in Cyprus for writing Greek until the 4th century. (You should google those!)

I do agree with you that reading e.g. the New Testament with reconstructed classical pronunciation (“Erasmian” you’d say) is inconsistent, and modern pronunciation might actually be closer. But wouldn’t it be surprising if the language hadn’t changed significantly in the preceding 500+ years?

69

You have said that those advocating reconstructed pronunciation can’t do it consistently. I think it’s true to a point, because everyone is limited by their native phonology. In most major modern European languages, the vowel system especially is very different from ancient Greek. For that reason, rendering vowel length can be quite difficult for a native French or English, for example. My native language is Finnish, which distinguishes short and long vowels, so that particular part is easy for me, but I suppose I have other blind spots.

This said, it can be done. If you listen to Ioannis Stratakis delivering the beginning of the Odyssey, it’s extremely credible (for me at least), remarkably consistent and very beautiful. We might of course argue about minor points (should he really pronounce those ϝ’s?), but there would be no point in that. Since Stratakis’ native modern Greek doesn’t have phonemic distinction between short and long vowels, there are a very few slips there, but I doubt anyone in this thread (except me with my native Finnish :smiley:) can pick them.

https://www.youtube.com/watch?v=UdyXlUmD3v4

70

Some examples:

I for H (pre-Euclidean E)

HEΡΜΙΣ (600-575 BC)
ΕΥΦΙΒΟΣ (530-500 BC)
ΑΓΑΠΙ (end 6th c. BC)
ΔΙΜΟΣΘΕΝΙΣ (5th c. BC)
ΑΘΙΝΑ (5th c. BC)

I for EI (pre-Euclidean Ε or EI)

ΠΙΣΙΣΤΡΑΤΟΣ (dated 7th c./550 BC)
KΛΙΤΙΑΣ (c. 570 BC)
ΚΛΙΤΟΜΕΝΕΣ (550-525 BC)
ΘΑΛΙΑ (530-500 BC)
ΙΝΑΙ (end 6th c. BC)

EI for I

TEIMAΡΧΟΣ (580-570 BC)

You can also refer to my comment to Chad above. The division between Classical and Hellenistic periods in terms of language is largely artificial. There is barely a 40-year window between the “end” of the Classical period and the beginning of the reign of Ptolemy Philadelphos. To put things into perspective, that means that it’s possible that some of the translators of the Septuagint were old enough to have been educated in Athens when Aristotle was living. Did the language really change so much in one generation?

Τhe introduction of the Phoenician alphabet is ascribed to Cadmus, hence why the ancients called the alphabet the Kαδμήϊα or Φοινίκια γράμματα. Herodotus dates Cadmus to 1,000 years before his time, which would put him in around 1,450 BC, which is also the usual date for the founding of Thebes. Herodotus also claims that he saw a Greek inscription in Thebes from the time of King Laius (13th century BC). There is also an Elean inscription from the 7th century BC which mentions an “anciently written law” (IGA 111):

[τ]ὰ ζίκαια κὰτ τὸ γράφος τἀρχαῖον εἴε κα.

https://archive.org/details/inscriptionesgra00deut/page/38/mode/2up

71

Out of curiosity, is there such a thing as long-by-position in Finnish, and if so, is the vowel lengthened? From what I’ve read Finnish stress is fixed to the first syllable, with a secondary stress the placement of which is determined by vowel length and syllable structure. If this is true, this is a very different prosody from Greek, where accent placement is not predictable and you can have long strings of syllables with only one stress. Also, isn’t it true that Finnish songs (e.g. the Kalevala) are all accentual based? I appreciate your imput :slight_smile:

72

The man has the voice (and the beard) of a god. It’s just too bad he ruins the consonants.

73

• I don’t know about Aristotle saying that ὁδός and ὀδός sound the same (hmmm, and even then, would “same” etc mean identical?), but I know that he wrote:
οὐ γὰρ ὁ αὐτός λόγος γίνεται,
διαιρούμενος, εἴπερ μὴ <ὡς> καὶ τό ὄρος, [καὶ] ὅρος τῇ
προσῳδίᾳ λεχθέν, σημαίνει ἕτερον. ἀλλ᾿ ἐν μὲν τοῖς γεγραμμένοις
τό αὐτό <τό> ὄνομα, ὅταν ἐκ τῶν αὐτῶν στοιχείων γεγραμμένον ἦ
καὶ ὡσαύτως (κἀκεῖ δ᾿ ἤδη παράσημα ποιοῦνται), τὰ δὲ φθεγ-
γόμενα οὐ ταὐτά. )
which, if there’s no “magic” involved, rather states exactly the opposite (but not about ὁδός, but ὅρος). Except if one doesn’t count prosody as part of language. In which text/place is the ὁδός-ὀδός thing?

• The (in)famous passage in Thucydides shows that the [οι - ῑ] don’t sound identical. If that was the case he would have written γράψουσιν and not ᾄσονται, since the written form of the word would have been the only way to differentiate the identically pronounced words. If that would be the case there wouldn’t be any reason to argue at all… So Thucydides wouldn’t have mentioned this case at all, or would have used appropriate words and not insist in using words that accidentally have to do with sound only. The 1st syllable of both words is long, the song was one that the old people present *remembered as been sung long ago (by others seemingly) and the old guys were not sure of “oi” or “i”.
Can one otherwise imagine a ridiculous scene of people saying/shouting the same word to each other?
Thucydides didn’t describe his contemporaries’ sounds but the relativity of meaning according to circumstances.

74

You forget that we do the exact same thing in English.

Let’s say I say "That person has nice [dʒiːnz]"

Without the context, you cannot tell if I am talking about a pair of pants (jeans) or about one’s genetic constitution (genes). You might imagine someone misunderstanding the sentence and being corrected by the speaker: “No, not jeans, genes!”

Or how about the following:

“I am going to the [sɛləɹ]”

Am I going to get a bottle of wine from the cellar or am I going to visit a salesman (seller)?

“The hikers discovered a [biːt͡ʃ]”

Did the hikers happen upon a sandy stretch of land by the water (beach) or a tree (beech)? Again, the context will elucidate things. Meaning is indeed relative to context, just like Thucydides said.


Minuscule letters and the consistent marking of accents were only introduced in the 9th century AD. Before that time, texts were written in unaccented uncial characters. So Aristotle didn’t write ὄρος καὶ ὅρος but ΟΡΟΣ ΚΑΙ ΟΡΟΣ. So it’s possible that what he actually intended was ὄρος (mountain) and ὀρός (whey), which would indeed differ τῇ προσῳδία λεχθέν. A bit further up in the same book, when Aristotle uses the word προσωδία, he uses it in reference to accents:

Παρὰ δὲ τὴν προσωδίαν ἐν μὲν τοῖς ἄνευ γραφῆς διαλεκτικοῖς οὐ ῥάδιον ποιῆσαι λόγον, ἐν δὲ τοῖς γεγραμμένοις καὶ ποιήμασι μᾶλλον, οἷον καὶ τὸν Ὅμηρον ἔνιοι διορθοῦνται πρὸς τοὺς ἐλέγχοντας ὡς ἀτόπως εἰρηκότα ‘τὸ μὲν οὗ καταπύθεται ὄμβρῳʼ. Λύουσι γὰρ αὐτὸ τῇ προσωδίᾳ, λέγοντες τὸ ου ὀξύτερον. καὶ τὸ περὶ τὸ ἐνύπνιον τοῦ Ἀγαμέμνονος, ὅτι οὐκ αὐτὸς ὁ Ζεὺς εἶπεν ‘δίδομεν δέ οἱ εὖχος ἀρέσθαι’, ἀλλὰ τῷ ἐνυπνίῳ ἐνετέλλετο διδόναι. (Sophistici Elenchi 166b 1-9)

Κατὰ δὲ προσῳδίαν (δεῖ διαλύειν), ὥσπερ Ἱππίας ἔλυεν ὁ Θάσιος τὸ ‘δίδομεν δέ οἱʼ καὶ ‘τὸ μὲν οὗ καταπύθεται ὄμβρῳʼ. (Poetics 1461a 21-23)

Aristotle is discussing two lines ascribed to Homer (the first one now considered lost or spurious). In the first case, the debate centred on whether Zeus had said δίδομεν δέ οἰ εὖχος ἀρέσθαι (we grant that victory be given to him [Agamemnon]) – with the accent on δι – or whether he said διδόμεν δέ οἰ εὖχος ἀρέσθαι (to order [Dream] to grant victory to him) – with the accent on δό – διδόμεν being the infinitive form.

Τhe second example is from the Iliad 23.328. The line, given in capitals, would be TO MEN OY KATAΠΥΘΕΤΑΙ ΟΜΒΡΩ. The question was whether one should read τὸ μὲν οὗ καταπύθεται ὄμβρῳ (where it rots away in the rain) or τὸ μὲν οὐ καταπύθεται ὄμβρῳ (which rots not away in the rain). The first reading (οὗ) takes a rough breathing but the second reading (οὐ) has a smooth breathing. You would think that this would be something Aristotle would mention. However, he completely disregards it. According to him, the difference between the words is one of accent: the second ου is pronounced more acutely (ὀξύτερον). Interestingly, in all of the Cratylus, Plato never once mentions the rough breathing either.

The passage I cited in my earlier post is from the Rhetoric to Alexander (1435b 18-24). Some people have suggested that the author of the work is actually Anaximenes of Lapsacus. That doesn’t really affect my argument, as Anaximenes and Aristotle were exact contemporaries. For a defence of the Aristotelian authorship, though, you can consult:

https://books.google.ca/books?id=JrTWngEACAAJ&pg=PR6#v=onepage&q&f=false

Here is the complete passage in Greek:

Τὸ δὲ τὰ ἀμφίβολα διαφεύγειν τοιόνδε ἐστίν· ἔνια τῶν ὀνομάτων ταὐτὰ ἐπὶ πλείοσι πράγμασι κεῖται, οἷον ὀδὸς τῶν θυρῶν καὶ ὁδὸς ἣν βαδίζουσιν. Δεῖ δ᾿ ἐπὶ τοῖς τοιούτοις τὸ ἴδιον ἀεὶ συμπαραλαμβάνειν. καὶ σαφῶς μὲν ἐν τοῖς ὀνόμασιν, ἂν ταῦτα ποιῶμεν, διαλεξόμεθα, εἰς δύο δ᾿ ἑρμηνεύσομεν διὰ τῆς προτέρας μεθόδου.

You cannot get any clearer than ἔνια τῶν ὀνομάτων ταὐτὰ ἐπὶ πλείοσι πράγμασι κεῖται, “we find that some words which are the same (ταὐτά) can refer to many things.”

If the words were pronounced differently, why are they ἀμφίβολα? Why would Aristotle feel the need to specify the ὀδὸς τῶν θυρῶν and the ὁδὸς ἣν βαδίζουσιν if the words were pronounced hodos and odos and could easily be distinguished by ear? It would be absolutely redundant, like saying hair, the stuff on your head, and air, the thing you breathe, or hedge, the thing in your garden, and edge, the thing you cut with. In fact, there would be no reason for this passage to exist at all.

We know that the Aeolic dialect had dispensed with the rough breathing already in Archaic times, as had the Ionic dialect by Herodotus’ time (which is why the Ionians were free to recycle the old grapheme H). Aristotle’s above two comments on οὗ and ὁδὸς suggest that by the 4th century BC Attic had lost it too.

75

It might be useless or even risky to revive a thread that is old and that at times has been somewhat ill-tempered, but I’d like to share the point of view and questions of a beginner or semi-beginner (I’ve just started reading lesson 11 of the Italian Ἀθήναζε) who has not read any of the books and articles referred to in the above messages. Having read the thread, where one side supports modern Greek pronunciation and the other a more archaic pronunciation, and other points of view supporting koine pronunciation (Lucian Pronunciation of Ancient Greek), I’m now utterly confused when I’m reading aloud as I think all three approaches are valid.

But the question I’m asking myself is : what’s my purpose when I read or speak ancient Greek ? Is it to be scientifically accurate ? No. An approximation will suffice. Is it to be culturally linked to modern Greece ? No, and I don’t intend to learn modern Greek. Is it so I can communicate with other ancient Greek speakers ? Not sure, probably not. (I do try to learn to speak Latin, though, because there’s a somewhat thriving modern Latin culture, and it’s not only extremely useful for language learning but also fun, to be part of it. I don’t think – I might be wrong – there is any such thing for ancient Greek.)

The point is that reading out loud (and possibly speaking) should help me learn the language, and specifically learn to understand the written language, and that’s all.

So, the new question is : what is the most useful pronunciation to learn the language ? Obviously, it should be a pronunciation that is as similar as possible to the written language, i.e. a form of Erasmian. Even if it’s scientifically proven that such a pronunciation had not been heard for centuries before our era, it will still help me understand how accents are placed (if long and short syllables are not distinguished, how would I know ? Correct me if I’m wrong), how to distinguish many morphemes, help me retain words and forms without the added burden of a discrepancy between pronunciation and writing, etc.

But maybe the answer is not that simple : modern or koine pronunciation could also be useful for learning if there’s plenty of interesting spoken material using those pronunciations. It so happens that such material exists for koine.

So I’m still hesitating.

I would be interested to know whether audio material with a modern Greek pronunciation but geared towards foreigners who don’t know modern Greek exist.

If someone can point me to such resources, I would be offered the invaluable opportunity to hesitate even more. Thanks in advance for that.

76

I’m always amazed at how passionately people can be invested in this topic. Simply this, use the pronunciation system which best helps you to learn and understand the language. There are good reasons to get a sense of historic pronunciation and the diachronic development of the language, and learning modern pronunciation helps you get around in Athens a lot easier, but the bottom line is what’s going to help you the most at the beginning stages. I participated in a “Koine conversation class” some time ago in which the majority of the participants used Buth’s reconstructed pronunciation. I did not, but something that probably sounded like Erasmean to them (though it wasn’t, more a modified Attic pronuncation). We had no problem understanding one another. In ancient times, different dialects were in play, and even during the Koine period someone from Corinth probably didn’t sound like someone who learned his Greek in Rome (regional differences). So relax and enjoy. I mean, regardless of pronunciation, it’s Greek, so what’s not to like? :slight_smile:

77

Counterpoint: poetry

78

Laurentius - I tend to go with modern Greek for the consonants (except Beta) and Attic reconstruction for the vowels myself, but as Barry says, use whatever works for you. In the end I don’t think it really matters, as long as it aids in learning the language!

As for jeidsath’s counterpoint - with poetry, he’s probably right - you’ll ultimately need a fully reconstructed pronunciation for the rhythm and meter to make sense, but frankly, by the time you are reading poetry and worrying about such things, your Greek will be quite well advanced enough to cope with knowing more than one pronunciation, and really it’s not a very big difference anyway.

79

Thanks to all for the sound advice!

80

I need to express my gratitude to Evangelos96, for his valuable contribution, on the subject of ancient Greeks’ pronunciation.
The only thing I have, as of now, to contribute, is, the recommendation of a book, thoroughly suppressed under the Erasmian’s descendance Vox Graeca by W.Sidney Allen, which has become sort of the standard guide for ancient Greek pronunciation; this is: «Βάσανος των περί της ελληνικής προφοράς ερασμικών αποδείξεων» (“Ordeal (probably in the sense of, “piece of work”) of the erasmian evidence about greek pronunciation”), by Παπαδημητρακόπουλος, Θεόδωρος.
Here is a link to download the document in Pdf format: https://anemi.lib.uoc.gr/metadata/5/8/6/metadata-87bd6543a927718e80eeea467060e661_1238058488.tkl
I have to add; the book is written in Katharevousa (Καθαρεύουσα), a form of Greek language, between Ancient and Demotic, and published in 1889, counting 774 pages.
Aim of the work is, to examine the Erasmian school of thought on its major evidence, at the time of writing, while presenting counter-evidence, to prove, when (according to the work, most of the times) and why the previous is wrong, and, finally, to prove, that pronunciation in ancient Greece, while it undertook major changes, in the interval between Homer and Hesiodus, and the Classical Period, from then on, its canon remained more or less unchanged.