So I’m thinking about pitch accents and melodies. In the third syllable of the Song of Seikilos we have an eta with a circumflex, but it only gets one note! According to Hagel, the pitch should rise between a third and a fifth and since this is a circumflex presumably fall back. So it seems the melody would have to have three notes for that one syllable! Had the pitch accent fallen by the wayside by then? They say it was written between 200BC and 100AD.
More generally, does anybody know how songs in languages with pitch accents work? Actually, let me back up. Do other pitch languages work like Greek where there is a rise during the syllable? Or is there just a higher pitch relative to the other syllables? If they work like Greek, then do composers have to put two notes for every stressed syllable? It seems so. How would the singer know how much to raise the pitch otherwise?
I just finished reading Sydney Allen’s Vox Graeca, where that song gets some treatment (p. 118 ff.). It says according to Dionysios of Halicarnassus, “melody is subordinate to the requirements of the music”, but in the case of the epitaph of Seikilos or Sicilus there’s a “tendency to agreement between the music and what we believe to have been the melodic patterns of speech.”
Looking at the musical score, there seems to me to be some correlation with the melody and the pitch accent - look especially at the two other circumflexes. I guess we have tendencies, not absolute correspondences here.
I’ve studied Swedish (it’s compulsory in Finland), and I’ve never noticed there was any correlation between the melody of Swedish speech and Swedish music. But only Swedish Swedish has the pitch accent; the marginal dialect of Finnish Swedish (like 200 000 native speakers), the one we are taught at school, doesn’t have it. So maybe I’m not a judge.
And since I’m totally unmusical, I guess I’m not a judge anyway…
It says according to Dionysios of Halicarnassus, “melody is subordinate to the requirements of the music…”
On an even more basic level than the pitches, every version of the song that I have heard places the accent (wrongly) thus
τὸ τέλος ὁ χρόνος ἀπαίτει instead of τὸ τέλος ὁ χρόνος ἀπαιτεῖ.
presumably because the latter would not line up with the tune.
We do this in English songs, of course, where the tune will trump the way we normally stress words.
We sing “My country 'TIS of thee..,” whereas I think we would say in normal speech “My country, 'tis of THEE..”
It’s just an uninformed guess, but I have always believed that in actual Ancient Greek speech, the rules of stress (to say nothing of the rules of pitch) would have been routinley modified for a variety of reasons. Even more so, I assume, when singing.
Lynyrd Skynyrd wrote:
The POLice said, ‘we can’t drink in the bar;’ what a shame…
Analogies from English (or any other language) are dangerous, but I’m pretty sure that rules of stress are routinely violated in singing and speech in Modern Greek as well.
Isn’t it rather one note on “o” and two notes with falling pitch on “lws”? That would be in line with the melody of speech as reconstructed: a rising pitch on the acute and a falling pitch on the syllable following it.
And yes, Dionysios of Halicarnassus said the first quote and the second is endorsed by Allen.
If I understood Allen correctly, he was saying that basically in poetry the heavy syllables probably were stressed in a way not unlike stressed syllables in a language with stress accent. That is, you sort of have the natural Greek pitch accent and a stress accent of poetry working at the same time. Probably these aren’t the terms he’s using but that was his point I think.
E.g. μῆνιν ἄειδε θεὰ Πηληϊάδεω Ἀχιλῆος
Bolded syllables should pronounced with some stress.
I count 10 syllables in all in the song of Seikilos having either the acute or circumflex accent, and all except the first two seem to behave logically:
with an acute, the syllable has a rising tone while the next syllable has a falling tone
with a circumflex, the syllable has first a rising and then a falling tone
With the third syllable, the eta with circumflex, the fact that the next syllable has an acute might be an interfering factor.
With the first syllable, the acute o of hoson, the fact that it’s the beginning of the song might the reason for the exception.
Just my guesses, somebody who understands music tell me if this makes sense.
Correction: the last syllable in ἀπαιτεῖ doesn’t seem to match exactly this rule, no rising tone, but it still has a falling pitch in the end. ἀπαίτει wouldn’t match the tune at all.
Allen dates the epitaph of Seikilos to “no earlier than 2 c. BC and probably 1 c. AD”. According to him, there’s no convincing evidence of a change from melodic to a stress accent before late 2nd - early 3rd century AD. (p. 130)
So according to Allen, the pitch accent was still alive and well…
Ah! My apologies, I didn’t read all the way through and, in my mind, it was a 1st century AD piece a transitory period when it comes to accents (and many other non linguistic things obviously )
as to why the first word in the score goes up rather than down, this could be because it’s at the beginning of the melody, see devine and stephens pg 166 which talks about this particular case: http://books.google.fr/books?id=poau61NlcP4C&pg=PA166
according to what people have worked out on the pitch accent, a circumflex didn’t get 3 pitches (a pitch, then a pitch a 3rd to a 5th higher, then a lower pitch), just the latter 2, like the other circumflexes in the musical score linked above.
pster it’s definitely worth working through the evidence on pitch in devine and stephens, in the book that i linked to above. i tried to summarise that stuff here: http://iliad.envy.nu/GreekPitchModel.pdf.
if i had spare time in this world i would dig into how this worked with the enharmonic mode, to be able to hear how homer may have sounded in the enharmonic mode, which seems to be the most likely from what i read in west on ancient grk music… when i put some time into this when i was younger it seemed to me that the pitch drop from an acute or the 1st part of a circumflex couldn’t occur within a πυκνόν and so – to get an adequate pitch drop – you’d need to drop down from the note above the πυκνόν into the πυκνόν, or from the πυκνόν down into the next tetrachord, but working out the details of this is something that i hope someone else with time will eventually do, as for me that ship has sailed
Thanks all. I’m glad that the tension between language pitch and music pitch has been explored by others. And it seems there is a decent body of knowledge about how this tension plays out for living languages in musical settings.
I’m a bit pissed because I don’t have easy access to either the Allen or the Devine and Stephens, although they both look excellent. Anybody have some spare copies they want to barter?
Can we be careful to distinguish a rising or falling pitch on a syllable and a higher or lower pitch relative to an adjacent syllable? It will cause endless confusion otherwise.
There are a number of authors that seem to get a fair bit of respect when accent matters come up. I’m thinking of West, Hagel, Allen, Devine and Stephens, etc. Are there any serious disagreements among them? So for example, I am just learning for the first time that the circumflex is just a falling pitch. Do all the authors agree about that? (I learned from Mastronarde and he says rise and fall on the syllable.)
@ Chad. I can’t get that pdf to open. Do I have to sign up?
The most interesting thing I saw while looking at the Allen through Amazon was Mountford’s point that (assuming that the strophe and the anti-strophe had the same melody) the language pitch had to submit to the musical pitch since the strophic correspondence doesn’t extend to identity of accentuation (p. 118).
Is anybody brave enough to summarize the consensus for how to pronounce acute and circumflex accents clearly distinguishing between rises and falls in pitch on a syllable and higher and lower pitches on a syllable relative to surrounding syllables?
All I gathering so far is that an acute must be followed by a syllable at a lower pitch and that a circumflex is made up of two tones with the second one lower. This seems to leave open the possibility that the accented syllables are no higher in pitch than preceeding syllables!
Oh, won’t somebody rid me of these troublesome questions!
I can’t find the Hagel passage where he talks about the pitch difference being between a third and a fifth. Now I’m starting to wonder whether it was even him.
Hagel says on his website:
“Individual accents produce smaller deviations from the overall melody (as in the extant musical documents). Each (major) accent may be realized by a rise to the accented syllable, but must be followed by a melodic fall. In the case of circumflex syllables, the post-accentual fall may be realized on the second part of the accented syllable, resulting in a two-note ‘melisma’. The gravis accent forces the melody to rise without any downtrend up until the next accented syllable.”
But on another thread in this forum, William Annis writes:
“Only on a long vowel or diphthong would the acute represent what sounds like a rising pitch. And the circumflex was very probably just a falling pitch, not a rise then fall. The grave was merely the absence of a pitch accent — definitely not a falling contour in any interpretation of the evidence.”
So Hagel seems to have the grave doing upwards work!
And Annis seems to rule out a higher pitch on an acutely accented short syllable relative to the preceeding syllable, while Hagel seems to leave that open.
And when we consider the acutes on ultimas, isn’t the idea that the only requirement is that the next syllable be at a lower pitch rather odd? After all, that is going to be a new sentence. Why accent ultimas at all with acutes if that is the case?
pitch rises syll by syll through grave-accented words to the next acute/circumflex in the following word. see my reconstructed iliad 1 doc for e.g.s, on this old site http://iliad.envy.nu/
Thanks Chad. I spent a while looking at the Pitch Model doc by mistake! That one is a bit harder to understand! Hehe. For Illiad 1, I’m curious why you didn’t use two notes for circumflexes. I would have thought that would been an important part of such an undertaking.
I’m still trying to hunt down the Hagel place where he talks about a third to a fifth. One thing that is bothering me is that those are rather large intervals, especially a fifth. I’ll have to see if I can find something on Swedish. But the musician in me is bothered by the range that is implied by all of this. If a circumflex spans a fifth (or even a third), then it seems likely a sentence will frequently span about an octave. But most songs span about an octave. In other words, it just seems like too much.
hi pster, i did show both notes of the circumflexes. see e.g. the 1st syll of the 1st verse. the ~ marks the second pitch of the syllable - i use a ~ throughout the doc to show the pitch of the 2nd beat of a long syll. cheers, chad
)