accents in Sappho 44.12-13

Greek Learning Greek
1

Dear all,

I am currently part of a reading club. A few days ago, we read Sappho 44 and I stumbled upon a problem regarding accents. The lines in question run as follows:

φάμα δ’ ἦλθε κατὰ πτόλιν εὐρύχορον φίλοις.
αὔτικ’ Ἰλίαδαι σατίναι[ς] ὐπ’ ἐυτρόχοις
ἆγον αἰμιόνοις, …

My question is about the bold parts. Both adjectives start with ευ-, but the former has the accent on the second letter and the latter on the first letter. I was hoping someone on this forum could explain why this is. Our (i.e. members of the club) first guess was that in some cases, the υ didn’t ‘count’ as a vowel, but as a consonant in Sapphic dialect, hence the υ doesn’t get accentuated in some cases. I myself thought it was a typo..

Any thought is greatly appreciated.

Kind regards,
GJCaesar

2

It’s not a typo.

Your question relates to breathings, not accents, of course.

In the first line, the basically “dactylic” meter requires that ευ- in εὐρύχορον be treated as a diphthong, giving the word the metrical shape _ υ υ _.

In the second line, ευ- in ἐυτρόχοις must be two syllables–not a diphthong. That’s why the breathing is placed over the ε, not the υ. Some editions would use a diaeresis to make this clear, printing ἐϋτρόχοις, but this would be redundant. -δαι σατίναι[ς] ὐπ’ ἐυτρόχοις has the metrical shape _ υ υ _ υ υ _ | υ χ.

It’s not unusual in Homeric hexameters for ευ- to be treated as either one or two syllables depending on metrical considerations, and this poem (the marriage of Hector and Andromache) is almost certainly echoing epic language and meter.

Both of these lines are the same meter. The first two syllables are the “Aeolic basis” and can be either long or short (x x). Then we have _ υ υ _ υ υ _ υ υ _. Finally, υ x. So the whole verse looks like this:

x x | _ υ υ _ υ υ _ υ υ _ | υ x

The most common Aeolic pattern is the glyconic: x x | _ υ υ _ | υ x. Sappho’s meter in this poem can be viewed as an expansion of the central element of the glyconic _ υ υ _ (a “choriamb”) by inserting “dactyl”-shaped elements. (However, it’s wrong to think of this element as dactylic–that’s why I put it in quotes. It’s actually a succession of longs and shorts, with no division into dactylic “feet”. It could just as well, or as wrongly, be described as expansion by appending anapests.) Instead of expanding by inserting “dactyls” (or appending “anapests”), the basic glyconic shape can be expanded adding “choriambs”: elements shaped like the central element of the glyconic: _ υ υ _, to get verses shaped like this:

x x | _ υ υ _ (_ υ υ ) ( υ υ _) | υ x

(Horace uses a Latinized version of this meter–the “Greater Asclepiadean”–in the famous poem that begins: Tu ne quaesieris, Leuconoe, quem mihi quem tibi and contains the phrase carpe diem.)

In contrast to the hexameter, Aeolic meters are isosyllabic (i.e., each verse in a particular metrical pattern has the same number of syllables, with no substitution of two shorts for a long). However, there are obvious similarities of Aeolic meters like this one to the Homeric hexameter, and some think that the hexameter emerged from “Aeolic” meters like this, though of course there can be no certainty on that point.

Here, I suspect, it’s the other way around–Sappho has consciously chosen to use an Aeolic meter that has a similarity to the hexameter. It’s difficult not to think that the pathos of this poem has some relationship to the tender and moving scene between Hector and Andromache in the Iliad: was Sappho inspired by that scene . . . or was the passage in the Iliad inspired by Sappho’s poem?

3

It’s generally thought that the Aeolic type of meter, with isosyllabism, represents the most archaic type of Greek meter. It’s well known that there are parallels to the glyconic in the anuṣṭubh, a meter used in many of the units of the Rg Veda, and it’s thought that this reflects a common descent of Greek and Indic meters from Indo-European oral poetry. (I disclaim any knowledge of Sanskrit.)

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

The puzzle is where the anisosyllabic hexameter, which allows substitution of two shorts for a long in the second half of each foot except the last, came from. At one time it was thought that the hexameter wasn’t of Indo-European origin–that it was adapted from a non-Greek meter. However, I think the hexameter could be derived from an isosyllabic Aeolic type of meter something like this:

x x | _ υ υ _ υ υ _ υ υ _ υ υ _ | χ

The first element would be an Aeolic basis, with two anceps syllables. The second element would be an expanded choriamb, similar to the second element of the Sapphic verse discussed in the preceding post. The ending would be a “catalectic” ending, with just one anceps syllable (or, more accurately, brevis in longo, a syllable that is protracted at the end of a line) instead of υ χ. In fact, catalectic verses are common in Aeolic meters. The catalectic glyconic, x x | _ υ υ _ | x, is known as a “pherecratean.” So the verse pattern set out above is an essentially a pherecratean expanded by “dactyls” or “anapests” if you prefer, in the same way that the Sapphic verse can be viewed as an expanded glyconic.

So how did anisosyllabism come from? The long central element, _ υ υ _ υ υ _ υ υ _ υ υ _ , would be very difficult to adhere to strictly, given the “composition in performance” nature of epic. (The mature hexameter is already hard enough to compose even with writing.) So substitutions of a long for two shorts could have begun to be allowed. Perhaps, too, contraction of two adjacent vowels contributed to this, as existing formulas could no longer fit the metrics of an isosyllabic verse.

And once the central part of the verse began to be felt as a succession of elements shaped _ υ υ, perhaps that pattern was generalized back to the Aeolic basis.

That’s all pure speculation, of course, but it’s one conceivable path (among others) from an isosyllabic, Aeolic type of verse to the hexameter.

4

I like your theory. But what do you think about the idea maintained by West (and maybe others) that the original elements of the hexameter are (– υ υ – υ υ –), (υ υ – υ υ – υ υ – –), (– υ υ – υ υ – υ), and (υ – υ υ – υ υ – –), i.e. the individual parts that come before and after the caesura. I don’t suppose your idea (with some modifications) is incompatible with this theory?

5

I’m very skeptical of those sorts of explanations for metrical evolution. That line of argument and evidence couldn’t explain, for example, why the Japanese chant “Nippon” instead of “Nihon” at soccer matches. And that’s a very simple problem compared to the sort of thing that we’re talking about here.

Was ἐϋτρόχοις pronounced any differently from εὐτρόχοις I wonder? The idea that it was two full monophthongs with zero glide strikes me as unlikely. So was the glide merely protracted (ie., eyu-)? If you’re already pronouncing diphthongized ευ at 1.8x the length of the short vowels, an extension might have been inaudible.

I have a paper somewhere about Finnish that discusses the length equivalence between long vowels and shorts followed by two consonants. It’s apparently a matter of debate for Finns whether long vowels followed by two consonants are special. Certain native speakers claim so, but it can’t be discovered by timing the syllables. It’s possible that εὐτρόχοις/ἐϋτρόχοις is such a difference.

6

what do you think about the idea maintained by West (and maybe others) that the original elements of the hexameter are (– υ υ – υ υ –), (υ υ – υ υ – υ υ – –), (– υ υ – υ υ – υ), and (υ – υ υ – υ υ – –), i.e. the individual parts that come before and after the caesura.

The hemiepes (_ υ υ _ υ υ _), which is also the second half of an elegiac “pentameter”, looks like an expanded choriamb. This would relate the hexameter and elegy to various meters used by Archilochus in epodes.

I don’t know which specific theory is right, and no one else does either (although of course West’s ideas are to be taken very seriously). But my main point is that there seems to be a relationship between the hexameter and the expanded choriambs that form the core of Aeolic meters.

7

Was ἐϋτρόχοις pronounced any differently from εὐτρόχοις I wonder?

The same alternation between ευ- as diphthong and as adjacent short monophthongs (the second occurring in a long/heavy/closed syllable) can be found in the Homeric poems. There must have been some difference–the difference is metrically guaranteed by the Homeric poems and by Sappho’s meter, too.

I’m very skeptical of those sorts of explanations for metrical evolution.

There is a large family of Aeolic meters in both the fragments of Sappho and Alcaeus and choral poetry (e.g., dramatic choruses). If you look at the family as a whole, the patterns on which these meters are constructed becomes apparent. They’re built around the choriamb, which can be expanded by either dactyls/anapests or choriambs. They usually have an Aeolic basis and an ending, which can be expanded to three or truncated to one syllable. My point is that the hexameter could conceivably have emerged from this type of meter.

That line of argument and evidence couldn’t explain, for example, why the Japanese chant “Nippon” instead of “Nihon” at soccer matches

Not sure I follow or see how this is relevant.

8

Here are the relevant metrical phenomena that I know of in Greek. Of course, I might be missing something:

  1. Diphthongs are (absent exceptions like this) always metrically equivalent to long vowels and heavy syllables.

  2. Sequences of two adjacent short vowels are generally only equivalent to a long vowel / heavy syllable / diphthong in a resolvable section of the meter.

  3. Neither long vowels nor diphthongs are metrically prolonged by sequences of double consonants.

  4. We sometimes — rarely — find a normally diphthongized word (but never a long vowel or heavy syllable) in a non-resolvable section of the meter. Grammarians have historically used a diaeresis to indicate the separation.

  5. Allen suggests that coincident short vowels generally were likely to have been pronounced with a glide.

  6. The difference between a glide and a diphthong can be squidgy.

This ἐϋτρόχοις example is a good one because we’ve got the heavy syllable. So one element of the difference between εὐτρόχοις and ἐϋτρόχοις must be length. But what would that length difference have looked like? West argues for a 1.6-1.8:1 difference between biceps and princeps (others argue for 2:1). So we’re talking about a very short elongation. 1.6-1.8 → 2.0. That sort of difference would fit well into general metrical fudge factors.

“Nippon! Nippon! Nippon!” (heavy-heavy heavy-heavy heavy-heavy) sounds right but “Nihon Nihon Nihon” (light-heavy light-heavy light-heavy) doesn’t. Look it up an Forvo and try saying it for yourself. The right/wrong seems very basic – you don’t even need to know Japanese. For some reason this seems related to the reason that Haiku (quantitative syllabic poetry, like Greek) is written in a 5/7/5 meter and read aloud in an 8/8/8 meter.

We can’t really discuss the evolution of Greek syllabic poetry without knowing what “sounds right” and what doesn’t. In fact, I would be surprised if – in true detail – it would have looked much like evolution. The similarities between different meters would, I’d guess, have less relation to genetic progress than to patterns that “fit” the language.

9

Meter proves that it was (as Hylander pointed out). εὐ diphthong has to sound like a heavy syllable, ἐϋ in ἐϋτρόχοις has to sound like a light syllable followed by a heavy one. Otherwise it would not “sound right.”

Of course, I might be missing something.

Well, yes.

10

West’s 1973 articles and Nagy’s 1974 book were both successful in relating the dactylic hexameter to Aeolic versification, which is closer to its IE origins than is the Greek epic tradition. Nagy postulated a proto-hexameter in the form of an expanded pherecratic,
x x _ υ υ _ υ υ _ υ υ _ υ υ _ x.
This is precisely what Hylander is now proposing,

Nagy found an example of this proto-hexameter in a pair of lines of Alcaeus (368 LP), which run:
κελομαι τινα τον χαριεντα Μενωνα καλεσσαι,
αι χρη συμποσιας ετ’ ονασιν εμοιγε γενεσθαι
(“I bid someone invite the delightful Menon,
if Ι’m still to get enjoyment from a drinking party.”)

It seems to me, however, that these don’t really work as potential hexameters, because the epic hexameter’s metrical behavior is fundamentally different. (Only one of the lines has a third-foot caesura, and both of them break between the two shorts of the potential fourth foot, which the hexameter rarely does.)
So I prefer to see the hexameter as the outcome of a fusion of two shorter verses (generalized versions of common 7/8/9-syllable IE verses), resulting in the caesura that is the hallmark of the hexameter.

None of these ideas do much to explain spondaic substitution, the great innovation from an IE perspective,* but the factors Hylander mentions (difficulty of sustaining double-shorts over a line of such length, phonological developments such as contraction) presumably had something to do with it.

  • The great linguist Antoine Meillet had established the IE origin of Greek meters in his ground-breaking book of 1923, but had found himself unable to accommodate the hexameter. He was forced to posit an “Achaean” origin for it.
    —Incidentally, it was Meillet who pointed to the formulaic nature of Homeric composition and set Milman Parry on his path.
11

Either hiatus (which is how we write it) or an almost imperceptibly lengthened glide would answer that. It’s the difference between how we prolong “advancéd” here:

And Death’s pale flag is not advancéd 'there.

Versus how we prolong:

O come O come Emmanuel

Which did the Greeks do in this case? Allen suggests a lot of glides. Despite my argument for the glide above, I think that he may be wrong. I was hoping to have an actual discussion of why we think it must be hiatus.

Διαίρεσις δέ ἐστι συλλαβῆς μακρᾶς ἀνάλυσις ἢ διάστασις εἰς δύο συλλαβάς. τὸ δὲ πάθος Ἰωνικόν ἐστι καὶ Θετταλικόν, τὸ δὲ πλέον Ἰωνικόν ἐστιν, οἷον κοῖλον κόϊλον, οἰωνῶν ὀϊωνῶν, παῖς πάϊς, Διομήδους Διομήδεος, βέλους βέλεος, (εἰ μὴ ἀνάπαλιν τὸ μὲν Διομήδους καὶ βέλους συνῄρηται, τὸ δὲ Διομήδεος καὶ βέλεος ὁλοκληρότατά εἰσιν, ἐπεὶ αἱ εἰς οῡς λήγουσαι γενικαὶ ἀπὸ περιττοσυλλάβων συναιροῦνται)· καὶ πάλιν πυλῶν πυλέων.

Τρύφωνος Περὶ παθῶν

Yes, I was thinking of Nagy’s arguments as well – though I haven’t read the West articles because they aren’t online. It was Luigi E. Rossi who critiqued what he called “genetic” arguments about metrical evolution of Homeric Hexameter.

12

mwh: One explanation for the strong caesuras in the hexameter might be, again, the “composition in performance” origin of epic verse–the need for a pause in a long line in a long song for the singer to catch his breath and collect his thoughts for the second part of the verse. This might not be so important in a shorter, occasional poem like that of Alcaeus, whether that was composed orally or with writing.

On the other hand, my recollection is that in reading the Homeric poems I’ve occasionally stumbled over instances where otherwise short/light syllables have been lengthened before a 3rd-foot masculine caesura–could this license be a survival of a fusion of two shorter verses?

Of course this is all speculation, but the point is that the hexameter, however and whenever it evolved (I think West traced it back to the middle of the 2d millenium), is not, as Meillet thought, alien to the family of Greek meters that go back to Indo-European.

Incidentally, in my senior year as an undergraduate I took a seminar on the Iliad with Greg Nagy–that would have been in the spring of 1968, and I read his book many years ago. So I was exposed to the about the derivation of the hexameter from an expanded pherecratic earlier at some point, but in my effort to explain the meter of the Sappho fragment yesterday the idea came back to me.

jeidsath: ευ is derived from εσυ, so perhaps when it was articulated as two syllables, the two vowels were separated by a slight aspiration rather than a glide. I don’t think a glide would be required to get from ε to υ, just a rounding of the lips, with no change in the position of the tongue–certainly not a raising of the tongue that would produce a j-glide.

13

Thank you all for the vivid discussion about my question! Your comments have kind of solved my problem. @Hylander I know it’s about breathing, but I used the term ‘accent’ in its widest meaning.

I’m afraid that my knowledge about metre in general falls short to be able to discuss it here with you. I might have to dig into it at some point, but I am currently too busy with other stuff!

14

“Diacrirical mark” or just “diacritical” would be better than “accent” to refer to a breathing mark. “Accent” is refers exclusively to the tone accents, acute, circumflex and grave. But your question was a good one.

If you’re reading Sappho, you should definitely try to understand the meters. The Aeolic meters aren’t as hard as the hexameter: the verses are generally shorter, and there’s no substitution of two shorts for a long. You should at least familiarize yourself with the Sapphic stanza, as in φαινεται μοι κηνοσ ισοσ θεοισι and ποικιλοθρον’ αθανατ’ Αφροδιτα.

Meter is much more important than breathings or accents, especially since accents were uniformly recessive in Aeolic, and Aeolic was already psilotic–the aspiration, i.e., rough breathing, characteristic of Attic had been lost in Aeolic, leaving only smooth breathings.

15

I am familiar with the Sapphic stanza, both in Greek (your example) and Latin (ille mi par esse deo videtur etc.), but the terminology especially can be quite puzzling. I am a ‘fan’ of metre, in the sense that I always try to understand and appreciate what metre is and what it does to our understanding of the text. I never understood why people skip choral lyrics in tragedy, just because they are hard and written in notoriously difficult metre!