diachronic syllabificiation of ἐξακολουθῶ

Are you reading Homeric Greek? Whether you are a total beginner or an advanced Homerist, here you can meet kindred spirits. Besides Homer, use this board for all things early Greek poetry.
Post Reply
bpk
Textkit Neophyte
Posts: 36
Joined: Fri Jan 04, 2019 12:18 pm

diachronic syllabificiation of ἐξακολουθῶ

Post by bpk » Wed Mar 13, 2019 4:18 pm

I have a question about syllabification through the various periods of Greek.

I recall from Smyth that permissible consonant clusters are typically syllabified together as a complex onset: e.g., ἔθνος = ἔ-θνος; ἐχθές = ἐ-χθές.

I suspect that these became a bit more like εθ νος and εχ θες during the Koine period, but I can't prove that (at the moment).

Smyth also mentions that, despite this rule about clusters, there tends to be a separation at morpheme boundaries so that prefixes on verbs (συν, εκ, etc.) would be syllabified separately, even if they would have otherwise created a permissible cluster.

Long story short, I am thus asking, in light of the two rules above, how a word like ἐξακολουθῶ would have been syllabified in Attic Greek. Would it be εξ-α-κο-λου-θω, ε-ξα-κο-λου-θω or εk-sα-κο-λου-θω And, if you feel so inclined and feel we have evidence for it, how it would have been pronounced at later periods (if different).

Any help is appreciated!

Hylander
Textkit Zealot
Posts: 1794
Joined: Mon Aug 17, 2015 1:16 pm

Re: diachronic syllabificiation of ἐξακολουθῶ

Post by Hylander » Wed Mar 13, 2019 5:12 pm

My guess would be that ἐξ- would continue to be syllabified as a unit, given that it would be recognizable as a morpheme or maybe sub-morpheme if there is such a thing. But I think ε-ξα-κο-λου-θω would be ruled out as long at least in poetry, as Greek poetry remained quantitative, because that syllabification would leave ε- as an open syllable with a short vowel, and hence a short/light syllable.

User avatar
jeidsath
Administrator
Posts: 3251
Joined: Mon Dec 30, 2013 2:42 pm
Location: Γαλεήπολις, Οὐισκόνσιν

Re: diachronic syllabificiation of ἐξακολουθῶ

Post by jeidsath » Wed Mar 13, 2019 5:21 pm

because that syllabification would leave ε- as an open syllable with a short vowel, and hence a short/light syllable.
I don't think that this is the case.

ἀλλ’ ἄγε λῆγ’ ἔριδος, μηδὲ ξίφος ἕλκεο χειρί

My impression was the syllabification doesn't have anything to do with the meter, and we know what we know of greek-specific syllabification due to statements from the ancient grammarians.
Joel Eidsath -- jeidsath@gmail.com

Hylander
Textkit Zealot
Posts: 1794
Joined: Mon Aug 17, 2015 1:16 pm

Re: diachronic syllabificiation of ἐξακολουθῶ

Post by Hylander » Wed Mar 13, 2019 8:58 pm

Greek syllabification doesn’t respect word boundaries (like Ancient Greek writing). It’s articulated as a continuous stream of syllables within the larger unit (colon) like French.

West explains how the prosodic rules for poetry are based on this feature at the beginning of Greek Metre. A long/heavy syllable is a closed syllable or a syllable with a long vowel. A short/light syllable is an open syllable with a short vowel.

Which, when I think about it, runs counter to my original suggestion that εξ- would be syllabified as a unit.

mwh
Textkit Zealot
Posts: 3292
Joined: Fri Oct 18, 2013 2:34 am

Re: diachronic syllabificiation of ἐξακολουθῶ

Post by mwh » Wed Mar 13, 2019 10:04 pm

You can’t have meter without syllabification.
With ἐξακολουθῶ in prose or verse I think it’s safe to say the first syllable boundary would at all periods normally come within ξ, i.e. between k and s. Syllables like to start with a consonant if there's one available. The morphological integrity of ἐξ is irrelevant to the prosody, as e.g. ξίφο|ς ἕλκεο is enough to show (syllabified before ς, though in writing a line would be not be divided there).
With ἀλλ’ ἄγε λῆγ’ ἔριδος, μηδὲ ξίφος ἕλκεο χειρί the syllable boundaries are:
ἀλ|λἄ|γε|λῆ|γἔ|ρι|δος|μη|δεk|sί|φο|ςἕλ|κε|ο|χει|ρί
That’s the conventional linguistic doctrine, and I know of nothing to contradict it. I would ignore Smyth on this.

User avatar
Paul Derouda
Global Moderator
Posts: 2133
Joined: Fri Nov 05, 2010 9:39 pm

Re: diachronic syllabificiation of ἐξακολουθῶ

Post by Paul Derouda » Wed Mar 13, 2019 10:27 pm

I have no idea, just asking: might it be possible that Smyth is reflecting spelling conventions (i.e. ways syllables are separated in case of line break etc.)?

User avatar
jeidsath
Administrator
Posts: 3251
Joined: Mon Dec 30, 2013 2:42 pm
Location: Γαλεήπολις, Οὐισκόνσιν

Re: diachronic syllabificiation of ἐξακολουθῶ

Post by jeidsath » Wed Mar 13, 2019 11:07 pm

But to answer Hephaestion, you can have syllable without meter. In Herodian, whose discussion seems to motivate Smyth, the syllable distinction is not about verse. It's about what vowels and consonants are pronounced ἐν συλλήψει and which are pronounced ἐν διαστάσει.
Τὰ ἄφωνα πρὸ τῶν ἀμεταβόλων ἐν συλλήψει εἰσὶν ἤγουν ὁμοῦ
εἰσιν, καὶ οὐκ ἔστι χωρὶς τὸ ἄφωνον καὶ χωρὶς τὸ ἀμετάβολον οἷον (5)
ἀκμή, ἀτμός, ὄκνος, ἀγρός, ἔκλαιον, ἔθνησκον, ἔπλεον, ἁβρός· ἰδοὺ
ἐπὶ τούτων δύο σύμφωνα ὁμοῦ εἰσιν, ἐπειδὴ ἄφωνον πρὸ ἀμεταβόλου
ἐστίν. δεῖ προσθεῖναι χωρὶς τῶν ὄντων ἀπὸ τῆς ἐξ προθέσεως. ταῦτα
γὰρ ἐν διαστάσει ἔχουσι τὸ ἄφωνον μετὰ τῶν ἐπιφερομένων ἀμεταβό-
λων, ἤγουν ἐν ἄλλῃ συλλαβῇ τὸ ἄφωνον καὶ ἐν ἄλλῃ τὸ ἀμετάβολον (10)
οἷον ἐκλῦσαι, ἐκνευρίσαι, ἐκρεῦσαι, ἐκμάξαι.
The discussion goes on for some pages, but it's clear that he's talking about how to pronounce words (and perhaps how to write them at line divisions as Paul says), and not how to scan them in verse. That is, pronounce ἐκλῦσαι as ἐκ|λῦσαι, not ἐ|κλῦσαι. Pronounce ἔκλαιον as ἔ|κλαιον, not ἔκ|λαιον. There is quite a difference on the tongue/to the ear, and I'm sure that it would have interested Herodian.

West's discussion of syllabification in Greek Metre 8ff is a metrical one and is rather unrelated to Herodian/Smyth discussion. Allen follows the second in Accent and Rhythm, and discusses all of this in terms of a motor theory of speech (he mostly follows Stetson).

For example, ἐγώ σε λαβὼν and ὡς ἔλασεν both scan the same (in the relevant portion), but would likely have been pronounced with a different motor pattern.
Joel Eidsath -- jeidsath@gmail.com

mwh
Textkit Zealot
Posts: 3292
Joined: Fri Oct 18, 2013 2:34 am

Re: diachronic syllabificiation of ἐξακολουθῶ

Post by mwh » Thu Mar 14, 2019 1:25 am

I don’t know what you mean by “But.” Of course you can have syllable without meter. All prose has syllables. Syllabification in verse is a regularization of syllabification in ordinary speech, that’s all. Herodian is talking about consonant clusters which may result in either a light or a heavy syllable in verse, according to the placement of syllable boundary.

Incidentally, I find it hard to think of a context in which ἐγώ σε λαβὼν and ὡς ἔλασεν would scan the same.

User avatar
jeidsath
Administrator
Posts: 3251
Joined: Mon Dec 30, 2013 2:42 pm
Location: Γαλεήπολις, Οὐισκόνσιν

Re: diachronic syllabificiation of ἐξακολουθῶ

Post by jeidsath » Thu Mar 14, 2019 2:55 am

Β261 εἰ μὴ ἐγώ σε λαβὼν ἀπὸ μὲν φίλα εἵματα δύσω,
Ψ615 τέτρατος, ὡς ἔλασεν. πέμπτον δ’ ὑπελείπετ’ ἄεθλον,

Referring to the ΩΣΕΛΑ, of course, making up the second foot of each.
Spoiler
Show
Looking for συλλαβή in the TLG, I found this fragment of Euripides. The very last seems like a stretch.

τὰ τῆς γε λήθης φάρμακ᾿ ὀρθώσας μόνος,
ἄφωνα καὶ φωνοῦντα, συλλαβὰς τιθείς,
ἐξηῦρον ἀνθρώποισι γράμματ᾿ εἰδέναι,
ὥστ᾿ οὐ παρόντα ποντίας ὑπὲρ πλακὸς
τἀκεῖ κατ᾿ οἴκους πάντ᾿ ἐπίστασθαι καλῶς,
παισίν τε τὸν θνῄσκοντα χρημάτων μέτρον
γράψαντα λείπειν, τὸν λαβόντα δ᾿ εἰδέναι.
ἃ δ᾿ εἰς ἔριν πίπτουσιν ἀνθρώποις κακά,
δέλτος διαιρεῖ, κοὐκ ἐᾷ ψευδῆ λέγειν.
Joel Eidsath -- jeidsath@gmail.com

Hylander
Textkit Zealot
Posts: 1794
Joined: Mon Aug 17, 2015 1:16 pm

Re: diachronic syllabificiation of ἐξακολουθῶ

Post by Hylander » Thu Mar 14, 2019 3:09 pm

Allen makes the point raised by Paul: namely, that Herodian and other grammarians confuse speech with writing. The "motor theory" of Stetson, which Allen endorses with some reservation (p. 2), apparently pertains only to the articulation of the individual syllable, not to syllable division. Allen doesn't seem to address synapheia -- the continuous stream of syllables in the colon or period without regard to word-divisions, which would have been a feature of Greek speech, like French, and is reflected in Greek meter. Synapheia

Allen Vox Graeca (3rd ed.) pp. 105-6:

"The statement of these rules [of syllable division] by the grammarians are somewhat misleading, since they tend to confuse speech with writing and to incorporate rules which apply more properly to orthographic word-division (at the enδ of lines). In particular they have a rule that any group of consonants which can occur at the beginning of a word (as e.g. κτ in κτῆμα) is allotted in toto to the following syllable even when it occurs in the middle of a word--thus τί-κτω (cf. Herodian, ii, p. 393L); but this is contrary to the phonetic division in Greek, which is τίκ-τω, giving heavy quantity for the first syllable." [Footnotes omitted.]

User avatar
jeidsath
Administrator
Posts: 3251
Joined: Mon Dec 30, 2013 2:42 pm
Location: Γαλεήπολις, Οὐισκόνσιν

Re: diachronic syllabificiation of ἐξακολουθῶ

Post by jeidsath » Thu Mar 14, 2019 6:23 pm

Allen has a whole chapter on the subject in Accent and Rhythm (ch. 3) which further details his points in the Vox Graeca discussion.
the continuous stream of syllables in the colon or period without regard to word-divisions, which would have been a feature of Greek speech, like French, and is reflected in Greek meter
UPenn's Mark Liberman doesn't believe in the distinction between syllable-timed and stress-timed languages. [1] [2] [3].
but this is contrary to the phonetic division in Greek, which is τίκ-τω, giving heavy quantity for the first syllable.
I don't think that Allen is justified in this statement. He's claiming that the metrical division can only be on a phonetic division, but I find nothing meter-busting about τι-κτω. In the following lines, it's trivial to achieve the same meter pronouncing "τι-κτε" and "ιφι κταμενοι" without prolonging the ιοτα. I'll provide a recording if people say it's impossible.

Β628 Φυλεΐδης, ὃν τίκτε Διῒ φίλος ἱππότα Φυλεύς,
Γ375 ἥ οἱ ῥῆξεν ἱμάντα βοὸς ἶφι κταμένοιο·
Γ70 συμβάλετ’ ἀμφ’ Ἑλένῃ καὶ κτήμασι πᾶσι μάχεσθαι·

Ζ227 πολλοὶ μὲν γὰρ ἐμοὶ Τρῶες κλειτοί τ’ ἐπίκουροι
Ζ228 κτείνειν ὅν κε θεός γε πόρῃ καὶ ποσσὶ κιχείω,

We have ι before κτ on a long beat in every one of these lines (or pair of lines at the end). Allen's suggestion only applies to the first two, making "τικ-τω" and "ιφικ ταμενοιο". It does nothing to change κτ self-adhering after the diphthongs and line-ends on the others. If Allen is right, we doesn't this overprolong the diphthongs and line end?

I don't think that there is any reason to depart from Herodian on the pronunciation of τίκτω.
Joel Eidsath -- jeidsath@gmail.com

bpk
Textkit Neophyte
Posts: 36
Joined: Fri Jan 04, 2019 12:18 pm

Re: diachronic syllabificiation of ἐξακολουθῶ

Post by bpk » Fri Mar 15, 2019 12:05 pm

Thank you everyone for your responses. I had a feeling this question might get a complex set of answers.

It seems that the following two issues are central to the discussion:

1) Were the ancient grammarians confusing spelling with pronunciation?

2) Is there a distinct pattern of syllabification for poetic meter vs. speech, or does the syllabification in poetic meter reflect that of speech?

Joel, I'd be curious to hear your thoughts further on #2 (I know you've already touched on it to a degree). I expect that, to a degree, a variety of patterns can be implemented for the meter when necessary. Are there any places where the grammarians explicitly mention or imply a difference between poetry and speech (or even prose) in terms of pronunciation or syllabification?

bpk
Textkit Neophyte
Posts: 36
Joined: Fri Jan 04, 2019 12:18 pm

Re: diachronic syllabificiation of ἐξακολουθῶ

Post by bpk » Fri Mar 15, 2019 3:24 pm

Two other factors to consider:

In Modern Greek, I don't believe things are any longer syllabified in the way that Smyth presents them, if they ever were. This means that at least at some point in the history of Greek syllabification changed.

Also, what can we learn from doublets like χθες and εχθες? If χθες is primary, then might the additional ε at the beginning of the word in the biform have been motivated by a desire to resolve the initial cluster as a sort of epenthetic (prothetic)? In such a case, could we learn that, at least phonetically in the dialects/styles that tended to use εχθες over χθες, this word was syllabified as εχ-θες? There may be another explanation, of course, but I'd appreciate your all's thoughts on this one.

User avatar
jeidsath
Administrator
Posts: 3251
Joined: Mon Dec 30, 2013 2:42 pm
Location: Γαλεήπολις, Οὐισκόνσιν

Re: diachronic syllabificiation of ἐξακολουθῶ

Post by jeidsath » Sat Mar 16, 2019 10:37 pm

I'd be interested in the syllabification of modern Greek.

I believe that the following advice, attributed to Isocrates, makes it clear that the rhythms of Greek poetry were not artificial, but similar to normal speech patterns (in contrast, say, to Sanskrit chanting). Specifically, he talks about mixing Iambic or Trochaic rhythms in oration.
Spoiler
Show
(10) Ἐκ τῆς Ἰσοκράτους τέχνης διδασκόμεθα ποῖαι τῶν λέξεων
λέγονται καθαραί· τοσοῦτον γὰρ πεφρόντικε τῆς καθαρότητος τῶν
λέξεων ὁ ἀνὴρ, ὡς καὶ ἐν τῇ οἰκείᾳ τέχνῃ τοιάδε παραγγέλλειν
περὶ λέξεως· δεῖ δὲ <ἐν> τῇ μὲν λέξει τὰ φωνήεντα μὴ συμπίπτειν
(χωλὸν γὰρ τὸ τοιόνδε), μηδὲ τελευτᾶν καὶ ἄρχεσθαι ἀπὸ τῆς @1 (5)
αὐτῆς συλλαβῆς, οἷον εἰποῦσα σαφῆ, ἡλίκα καλὰ, ἔνθα
Θαλῆς· καὶ τοὺς συνδέσμους τοὺς αὐτοὺς μὴ σύνεγγυς τιθέναι
καὶ τὸν ἑπόμενον τῷ ἡγουμένῳ εὐθὺς ἀνταποδιδόναι· ὀνόματι δὲ
χρῆσθαι ἢ μεταφορᾷ μὴ σκληρᾷ, <ἀλλ’> ἢ τῷ καλλίστῳ ἢ τῷ
ἥκιστα πεποιημένῳ ἢ τῷ γνωριμωτάτῳ· ὅλως δὲ ὁ λόγος μὴ λόγος (10)
ἔστω (ξηρὸν γάρ) μηδὲ ἔμμετρος (καταφανὲς γάρ), ἀλλὰ μεμίχθω
παντὶ ῥυθμῷ, μάλιστα ἰαμβικῷ ἢ τροχαϊκῷ
... Διηγητέον δὲ
τὸ πρῶτον καὶ τὸ δεύτερον καὶ τὰ λοιπὰ ἑπομένως· καὶ μὴ πρὶν
ἀποτελέσαι τὸ πρῶτον ἐπ’ ἄλλο ἰέναι, εἶτα ἐπὶ τὸ πρῶτον ἐπα-
νιέναι ἀπὸ τοῦ τέλους· καὶ αἱ ἐπὶ μέρος δὲ διάνοιαι τελειούσθωσαν (15)
ἐφ’ ἑαυτὰς περιγραφόμεναι. (Maxime Planude, V, p. 469)
Joel Eidsath -- jeidsath@gmail.com

Hylander
Textkit Zealot
Posts: 1794
Joined: Mon Aug 17, 2015 1:16 pm

Re: diachronic syllabificiation of ἐξακολουθῶ

Post by Hylander » Sun Mar 17, 2019 2:05 pm

Planudes was a 13-14th c. CE Byzantine scholar at a 1600+ year remove from Isocrates.

However, prose rhythms in orators such as Demosthenes and prose writers such as Isocrates and Plato apparently follow the same prosodical principles as ancient Greek poetry. To me, this suggests that Allen is right about syllabification in spoken Greek (to the extent it could be perceived), and that the rule of syllabification stated by Herodian (and Smyth who is simply repeating what Herodian wrote) to the effect that any consonant cluster that can begin a word in Greek should be syllabified as a unit with the following vowel is an orthographic rule, not a rule based on Greek speech.

User avatar
jeidsath
Administrator
Posts: 3251
Joined: Mon Dec 30, 2013 2:42 pm
Location: Γαλεήπολις, Οὐισκόνσιν

Re: diachronic syllabificiation of ἐξακολουθῶ

Post by jeidsath » Mon Apr 01, 2019 6:36 pm

Sorry to take so long to reply, but I've been thinking about this a bit. I can think of several word-end phenomena that occur in ancient Greek:

1. Caesurae. These can occur on a Vowel | Consonant Consonant boundary, V|CC, without difficulty, which one might think would be precluded if the syllables were split along VC|C.

2. Line end rules (final cretic, etc.). These rules observe the word-end, and not the proposed syllable end.

3. Accent rules. Accent is not influenced by syllable weight at all, only vowel length. But accent is also not moved by elision, correption, etc.

4. Preposition-verb composition. This isn't really a word-end phenomenon, but the augment does fall between the preposition and the verb in non-primary tenses.

5. Mute-liquids combinations cannot scan as light when separated by a word boundary, but are otherwise doubtful.

6. Rho after a word boundary causes a preceding light vowel syllable to scan doubtful.

A syllabification rule as proposed, that simply follows the metrical boundaries, really only "explains" one thing: that VCC is scanned heavy, whether V is short or long, except for mute-liquid combinations. The justification for the rule, as I've seen it (example Probert "Short Guide" 52-53) is that on mute-liquids, the syllable can fall as "πα-τρι" (light-light) or "πατ-ρι" (heavy-light). And VCBC (where B = word-boundary) is always split on the word boundary VC|C, making that first vowel heavy.

However VBCC is doubtful, and can be split V|CC or VC|C -- on the word boundary or not.

The proposed syllable rule also gives no explanation for why words that begin with rho cause the preceding syllable to be doubtful. After all, that is just VBC.

The metrical syllabification system of West/Probert/etc., can only explain one thing. It sheds no light at all on the many other word-end phenomena that existed in Greek language and poetry. The ancients did not ever notice such a system.

On the other hand, the system does bear some slight resemblance to Vedic syllabification, as pointed out by West, but I don't think that is enough to overrule the detractions.

I think a simpler explanation for all of these phenomena is that the syllabification was more or less as the ancients had it, and that it does not really affect the meter of poetry, which is determined instead by the speed of motor unit activation in pronouncing a word with a vowel followed by two consonants.
Spoiler
Show
One of the main problems I have with the West system, though I can't justify it other than by a gut feeling of wrongness, is that the augment in a composed verb generally adheres to the preceding preposition, rather than the following verb.
Joel Eidsath -- jeidsath@gmail.com

mwh
Textkit Zealot
Posts: 3292
Joined: Fri Oct 18, 2013 2:34 am

Re: diachronic syllabificiation of ἐξακολουθῶ

Post by mwh » Mon Apr 01, 2019 11:16 pm

Joel, there are too many muddles here for me to wade through. Just one or two points that I think are both unarguable and fundamental.
#Syllabification in verse is a regularization of syllabification in ordinary speech (as I said before).
#In verse, unlike in prose, the syllabified unit is the metrical line, the stichos, which behaves internally as if it were “a continuous stream of syllables,” regardless of its actual phonetics and of its lexical make-up.
#A stichos is prosodically independent of its neighbors.
#Syllabification in verse is necessarily consistent with the metrical pattern, and the metrical pattern in turn helps determine the syllabification.
#Phonetic behavior and metrical behavior, while interrelated, need not coincide. (So West can talk of “metrical syllables,” as distinct from phonologically ordinary ones.)
Hopefully these points may at least help free you from your (forgive me) misguided preoccupation with word-end.

User avatar
jeidsath
Administrator
Posts: 3251
Joined: Mon Dec 30, 2013 2:42 pm
Location: Γαλεήπολις, Οὐισκόνσιν

Re: diachronic syllabificiation of ἐξακολουθῶ

Post by jeidsath » Sat May 11, 2019 1:34 pm

I plead guilty to having many misunderstandings and understanding all of this imperfectly. However, I'd like to link to a discussion at the beginning of West's review of Allen's Accent And Rhythm about their differences in the matter of syllabification, and his objections to Allen's motor theory explanation.

https://www.jstor.org/stable/27686376
Joel Eidsath -- jeidsath@gmail.com

Post Reply