ὃν οἱ θεοὶ φιλοῦσιν, ἀποθνῄσκει νέος.
Is the meter incorrect for this famous quote? Why is there a broken anapest in the fourth foot? Shouldn’t it be:
ὃν οἱ θεοὶ φιλοῦσ’ ἀποθνῄσκει νέος.
This puts a quasi-caesura at the end of the third foot and seems to fix the difficulty. However I must be missing something important because I can’t find the second version anywhere.
From the Loeb website:
The line is cited by Stobaeus, Eclogae 4. 52b. 27 (περὶ ζωῆς), with the full heading Μενάνδρου Δὶς Ἐξαπατῶντος. Without the play-title but with the author’s name it is cited by [Plutarch], Consol. ad Apoll. 119e; Clement of Alexandria, Stromata 6. 2 (p. 436. 10 Stählin); a scholiast on Homer, Odyssey 15. 246; and Eustathius, commenting on the same line of Homer, 1781. 2. It appears also in the collection of monostichs ascribed to Menander, 583 Jäkel.
ΣΥΡΟΣ
ὃν οἱ θεοὶ φιλοῦσιν, ἀποθνῄσκει νέος.Cf. Plautus, Bacchides 816–17 (Chrysalus) quem di diligunt / adulescens moritur, dum ualet, sentit, sapit.
Actually, I see now. θν is a mute + nasal, so it’s a tribrach in the fourth foot…though θνησκω never seems to behave like this in Homer. The θν always seems to be treated as a double-consonant.
Yes it’s a perfectly ordinary resolution after 4th-foot caesura. Light syllables often stay light before θν- and other plosive+nasal(or liquid) combinations in Attic and later Greek. E.g. Soph.OT623 (Oed. to Creon) ἥκιστα· θνῄσκειν, ου φυγεῖν σε βούλομαι.
After your post, I read through Allen’s Correptio Attica article, as well as a few other discussions of Correptio Attica in older books. I had seen the Allen article before but must have only skimmed it, because it was mostly news to me. The takeaway seems to be that plosive (mute) + liquid/nasal combinations (except for voiced plosives + liquids) are pronounced as single consonants in Attic, but much more rarely in Homer. Comedy reflects this closely, with Epic reflecting it not as much.
So it seems that there was a pronunciation difference between Homer and Attic, and one that was important to the meter.
This stumbled my scanning of the Menander line for most of the evening last night, but I think I’ve mostly got it now. I’ve included a line from the Iliad for comparison: https://www.youtube.com/watch?v=NjHom-lOftM
I suppose it doesn’t matter for most classicists, since they all tend to use the fricative pronunciation for the aspirates χ,θ,φ. A fricative (σ was the only one) + liquid/nasal always made the previous position strong in Attic Greek.
Surely it does matter (to a certain extent) for most classicists, who regardless of how they pronounce theta need to know that in Attic the second syllable of e.g. αποθνησκ- will not necessarily be long—though it could be, if it suited the meter.
So your “takeaway” is valid only in the most general terms. West’s Greek Metre gives more precise data. Much good info on Attic in Andreas Willi’s Languages (sic) of Aristophanes.
I feel a little silly asking such an elementary question, but I am baffled by a comment I stumbled upon in Reading Greek (2nd ed.) They quote the Greek saying and then remark "The Roman comic poet Plautus translates this: quem di diligunt adolescens moritur. Byron mistranslated this line as ‘Whom the gods love die young.’: why is that wrong? What would it be in Greek? For the life of my I cannot find the error. Hold on! I think I have it. It is “die” plural. It should be “dies” singular. So would the Greek be: οὓς οἱ θεοὶ φιλοῦσιν ἀποθνῄσκουσιν νεοί? I am a long way from exploring the subtleties of meter.
Yes, that would be right (minus the moveable ν), though not metrical.
Replying to this old comment from mwh:
Surely it does matter (to a certain extent) for most classicists, who regardless of how they pronounce theta need to know that in Attic the second syllable of e.g. αποθνησκ- will not necessarily be long—though it could be, if it suited the meter.
Well, it couldn’t be lengthened in tragedy (which is λεκτικόν, says Aristotle).
Hi Joel, plosive + liquid within a word can make a syllable heavy (not syllable-releasing) in tragedy, as Michael says.
Are you basing your conclusion on Attic correption generally? Tragedy has its own conventions. See for instance page 40 of my old notes for iambic trimeters in Greek tragedy:
http://aoidoi.org/articles/GrkIambicComp-23-Apr-06.pdf
See also West who explains how plosive + liquid treatment differs across authors, dialects, genres etc. (Greek metre, pp. 16–17):
In early poetry this treatment of plosive + liquid [i.e. as syllable-releasing] is normally found only at the beginning of a word [> footnote: Even this is avoided in the iambi of Archilochus and Semonides and in ‘normal’ Lesbian lyric, with the doubtful exception of Acl. 332. 1. In Stesichorus, however, it is the rule.> ] or in words which the verse would not otherwise admit. Theognis, Solon, and Pindar show a greater freedom, and in Attic speech short syllables regularly remained short before these consonant combinations (‘Attic correption’): > in tragedy they are often lengthened within the word> , but in comic dialogue only where tragedy is being parodied or the style is elevated. …
A very helpful intro to metre is by David Butterfield, on the Antigone site. Check out the first and fourth bullets under N.B. on page 4 of the lecture notes here, making the same point:
https://antigonejournal.com/wp-content/uploads/2021/05/Metre-I-II.pdf
Much more often than not, you’ll find the syllable light in trimeters, but heavy syllables do occur, as noted in the sources above.
Incidentally, for resolution of the 8th position in the trimeter (which kick-started this thread) check out the statistics I compiled on resolutions of the different trimeter positions in Aeschylus and Sophocles on page 53 of my old notes:
http://aoidoi.org/articles/GrkIambicComp-23-Apr-06.pdf
(I should mention that these notes are over 15 years old; I had only been studying Greek and Latin a few years before that.)
Cheers, Chad
I am not talking about plosive+liquid in general. I am talking about θν specifically. θν does not ever make syllables heavy in tragedy. The Stobaeus fragment there needs a fix (as is usually the case for his quotations).
EDIT: Chad’s post originally referenced “παισίν τ’ ἀποθνῄσκοντα χρημάτων μέτρον”
In general, it is interesting that these doubtful consonant combinations in tragedy much more often follow a short than not. They never make augment or reduplication long. Lengthening the preceding syllable seems to have been the license.
Hi Joel, yes, your final sentence of your last post is right. It’s consistent with my post, Michael’s and the sources I linked above, e.g. the fourth bullet on page 4 of David Butterfield’s notes:
By contrast, in Attic dialogue a short vowel is > very often not lengthened > by a following mute-cumliquid; when this does occur, it is almost always within words and > extremely rare > between them. This preference for retaining short syllables is known as Attic correption.
Yes, originally I had posted that Euripidean fragment you mention above (which shows θν making the syllable heavy in 4th position in the trimeter) but took it out as I wanted to focus on the rule and not any particular example.
Cheers, Chad
Chad, before continuing in this vein, wouldn’t it be fair to say that you just do not have the information required to evaluate the rest of my post? Or, in fact, to evaluate the statement about -θν- in tragedy that got you to jump in on this thread?
Again, I repeat:
The second syllable of “αποθνησκ” could never be long in tragedy, as -θν- never causes lengthening of a short syllable in tragic meter.
These consonant combinations in tragedy much more often follow a short than not. The ratios (anceps excluded), for example, for ὑβρίζειν is 6.50:1, for ὕβρις 6.67:1, νεκρός 6.58:1, πότμος 4.33:1.
These consonant combinations in tragedy never make augment or reduplication long.
Anyone who wants to argue with the above (Chad), rather than arguing with me, could instead take it up with the spirits of H. Darnley Naylor, “Doubtful Syllables in Iambic Senarii”, and T. G. Tucker, “On a Point of Metre in Greek Tragedy.”
Seems to me that -θν- would be more likely to be treated as a syllable onset in a transparent compound such as απο-θνηισκει. Just a thought.
By the way, how can we be sure that θν never makes position in tragic meter when so little tragedy has survived? We have a bigger chunk of Euripides, but not much Aeschylus and Sophocles . . . and there were other tragedians, too.
Yes, as an obvious compound, it would also be unlikely to be lengthened. There are a few examples of compounds seeing lengthening, which you can see in the Tucker article. But it’s very rare.
Could -θν- have caused lengthening from time to time, and we just don’t have enough examples of it? Sure. But with ἔθνος and ὀθνεῖος both used without lengthening, you at least have a statistical bound on the probability. (~11% is my very quick and dirty estimate, simply given the fact that with 67% of words with doubtful syllables, you can find at least one example. So with two words, the chance of no lengthenings occurring by sheer chance is 0.33 * 0.33 ~ 0.11. But this would need to be modified for frequency.)
Regardless, the only reason that I brought up the -θν- and tragedy point, was that the classicist could be usefully served with a phonetic theory that goes a bit beyond “doubtful syllables can be used to make meter”. Tucker gives a couple of bad emendation examples where it could help:
Electra 629: οὐδεὶς παρῆν Ἀργεῖος, ὀθνεία δὲ χείρ
Agam 1664: σώφρονος γνώμης δ᾿ ἁμαρτεῖν τὸν κρατοῦντα <θ᾿ ὑβρίσαι>
He also suggests, plausibly, that the more frequent use of the lengthening license in the mock-tragedy of comedy would have been recognizable to the ear of the average Athenian, and contributed to the absurd effect.
On the metrical point at issue the net obviously has to be cast wider than θν, and I’d refer to somewhat less antiquated authorities than Tucker, e.g. Dale on Eur.Alc.542, Barrett on Hipp.760.
Besides, all this talk of specifically tragic practice is a diversion from the fact that the line in question is from New Comedy, not even paratragic.
It would be great if these statistical tables showed up in newer stuff, but if they’ve been replicated, I don’t know where. I suspect that they haven’t been, as D&S cite them as their authority. Judging by the Greek Metre statements and bibliography, West may only have only second-hand information on it. We’ll have to deal with old numbers, I think. Hopefully they have not gone bad…although every now and then I read a linear algebra paper from this time period, and I haven’t come across any number rot yet.
Archilochus on the old paper problem:
οὔ τις αἰδοῖος μετ᾽ ἀστῶν οὐδὲ περίφημος θανὼν
γίνεται· χάριν δὲ μᾶλλον τοῦ ζοοῦ διώκομεν
οἱ ζοοί, κάκιστα δ᾽ αἰεὶ τῷ θανόντι γίνεται.
But this discussion is not really occasioned by the line in the topic, but by the later comment (in this old 2016, certainly rotten, thread, started by me when I knew much less Greek, and making an unlikely claim about fricatives) that classicists don’t really need a phonetic basis for their understanding of doubtful syllables, and that it’s good enough to simply know that they can make meter when needed.
Hi, there are a few universal statements in this thread which I believe should be nuanced.
A. That θν can never make a syllable heavy in tragedy. This should be nuanced for two reasons.
Firstly, there is a counterexample. It’s cited in the thread already: Eur. Frag. 578.6 παισίν τ’ ἀποθνῄσκοντα χρημάτων μέτρον. Devine & Stephens take this as a valid example of θν making position (1994 p. 33).
Secondly, many occurrences fall in anceps position, and we have no way of knowing whether they are heavy or light in any case. At least three factors are at play.
Any supposed universal statement as to the weight of these syllables is therefore a weak induction at best. We just can’t tell in anceps, for instance.
B. That reduplication followed by mute plus liquid is never heavy. This was irrelevant to the original thread discussion and so I did not engage with it. This universal affirmation should be nuanced for two reasons.
Firstly, there are counterexamples. I cited one above from Sophocles: Soph. El. 366 πάντων ἀρίστου παῖδα κεκλῆσθαι, καλοῦ. These are rare (as I noted in my original post), but go against any universal rule.
Second, many occurrences fall in anceps position, and so the same issues I mentioned in A above apply here.
C. Same statement for B but on augment. This was irrelevant to the original thread discussion and so I did not engage with it. This universal affirmation should be nuanced due to the anceps point I mention above in A. I wouldn’t expect to see an augment making position, nor would I hazard a universal rule for the reasons outlined above. We can only go off the evidence we have.
For phonetic analysis (which is not yet present in the thread as far as I can tell), I’d direct readers to the leading analysis which I use: Devine & Stephens 1994 who consider the treatment of mute plus liquid (with a focus on tragedy) from a cross-language perspective, considering metrical evidence, sonority, morphology, orthographic evidence and linguistic evidence at e.g. pages 31–43 and elsewhere in the work (e.g. from page 59 on). All the examples from my posts come from this leading source (and the principles of mute plus liquid are supplemented from West and others in my first post).
As for the patterns, I believe we are all saying close to the same thing; I just wouldn’t universalise for the reasons stated above.
Cheers, Chad
Thank you Chad. A much better post, and to the point. Though an “oops” would have been appreciated.
The appeal to D&S as an authority for ἀποθνῄσκειν is not convincing. They are wrong to cite it. Both that line and the next are fixed up by editors, of course. But more importantly, ἀποθνῄσκειν is not a word ever used in tragedy. It is in constant use in Menander and Aristophanes (though with a short second syllable, interestingly).
I am aware of the (extremely rare) reduplication exceptions. I suspect that confusion with poetically-reduplicated κικλήσκω is likely to be at play with κεκλῆσθαι. But even given an exception, it’s actually a far stronger rule than any rule about θν, without exceptions. You are using the wrong framework to talk about existence of exceptions or universality for a phonetic rule. This isn’t axiomatic mathematics. Percentage of exceptions and regularity is what matters.
ἀπότροποι is a fossilized prefix. (I could give more examples in this vein.)
D&S, whom I’ve already referred to in this thread, are useful to have around, but are a minefield in the details. Beyond things like this ἀποθνῄσκειν issue, their chapter on accent and musical notation is fundamentally ruined by using an incorrect statistical tool for small datasets.
Hi Joel, I understand from the sentence in your post above “Percentage of exceptions and regularity is what matters.” that you agree that the universal statements in the earlier posts need to be nuanced.
That brings your position in line with my posts and the sources I quoted in my posts (my position is the same throughout). Agreement being reached on this, we can conclude. (If you wish to say that we are not in agreement and I and the leading sources on the matter which I quoted are wrong in some way which I cannot yet see from your posts, that’s fine too; let’s still conclude.) As to the other matters in the thread, I will leave those to others.
I try to leave others the last word on this forum even where I disagree (readers can read the threads and make their own assessments) and so if you’d like to add anything to wrap this up, please feel free.
Cheers, Chad
This is a revisionist statement of the course of the argument. If a reader looks upthread, he’ll see nothing of the kind. My statement about -θν- was a specific point about the non-universalness of the regular rule. Chad missed that, and mistakenly and didactically jumped in with a universal rule (that was what in fact needed nuance) to disprove the exception, “…plosive + liquid within a word can make a syllable heavy…”.
My own hope had been to have a conversation about the well-known sonority sequencing constraint, and how it might tell us something about how syllable pronunciation may have changed between the Epic and Classical period, and what sort of consonant cluster pronunciations might be adopted by classicists to allow them to better get the poetry of both periods by ear and tongue, rather than a set of secondary scanning rules that the ancients had no need for.