Dactylic Pruritus

[Bardo de Saldo]

Dear Textkitanici, I’m having some trouble digesting bits and pieces of information that I’ve gathered here and there concerning stress and pitch:

William Harris, in The Musical pitch accents in Greek says: “(...) Greek had no stresses (...)”

Benner, in his Selections from Homer’s Iliad says: “The first syllable of each foot is emphasized in oral reading. This stress of the voice is called ictus (...)”

Our own annis says somewhere in this forum (I quote from memory, and excuse me if I misunderstood): “Even if indeed the accents represent a change in pitch, that change is accompanied by stress.” (Side question: Do graves loose the pitch but save the stress?)

My conclusion is that, when reading dactylic hexameters, one would use the accents for pitch, disregard them for stress, stress every ictus (metrics overpowering individual words), and emphasize the stress where an accent falls on the ictus. Does that make sense?

[Eureka]

My conclusion is that, when reading dactylic hexameters, one would use the accents for pitch, disregard them for stress, stress every ictus (metrics overpowering individual words), and emphasize the stress where an accent falls on the ictus. Does that make sense?

The blue things are true, the red things are false.

(I think what Annis probably said is that a stress may have accompanied the pitch accent in common speech.)

[annis]

Benner, in his Selections from Homer’s Iliad says: “The first syllable of each foot is emphasized in oral reading. This stress of the voice is called ictus (...)”

This is an older practice. Well, I call it older. I suppose it is still taught in most schools. If you're going to go for a reconstructed pronunciation - pitch accent, attending carefully to vowel length - I would ignore Benner's advice here. In fact, I would advise tossing the whole "ictus" idea out the window.

Our own annis says somewhere in this forum (I quote from memory, and excuse me if I misunderstood): “Even if indeed the accents represent a change in pitch, that change is accompanied by stress.”

I don't think I would say that, but would say what Eureka said I would say.

And please, call me Will or William.

[Bardo de Saldo]

Thanks, aeodoi.

Will, if I toss the ictus out the window, where do I stress? The only alternatives that I can think of are stressing the accented vowels, or sounding flat. Could it be stressing based on phrase intonation rather than individual words? That would mean having to actually understand what you're saying!

[annis]

Will, if I toss the ictus out the window, where do I stress?

You don't.

Several languages (including one major European one) get along fine without any stress accent, or a greatly reduced one. The pitch accent is the word accent in Greek.

The only alternatives that I can think of are stressing the accented vowels, or sounding flat. Could it be stressing based on phrase intonation rather than individual words?[

If we're using the word "stress" here to indicate "volume," then I would say absolutely.

That would mean having to actually understand what you're saying!

Yep.

[Eureka]

Benner, in his Selections from Homer’s Iliad says: “The first syllable of each foot is emphasized in oral reading. This stress of the voice is called ictus (...)”

This is an older practice. Well, I call it older. I suppose it is still taught in most schools. If you're going to go for a reconstructed pronunciation - pitch accent, attending carefully to vowel length - I would ignore Benner's advice here. In fact, I would advise tossing the whole "ictus" idea out the window.

Really??? You mean ictus is all a modern scam?

[Bombichka]

Really??? You mean ictus is all a modern scam?

I think some Latin grammarians mentioned the ictus, otherwise it wouldn't even be called that (ictus is Latin for "stress")

I've read about it in my favourite Vox Graeca by W.S. Allen. There, Allen concludes that long (or rather "heavy") syllables had a greater stressability than the "light" ones, which implies they tended to be more emphasized while reciting poetry.

[Eureka]

Really??? You mean ictus is all a modern scam?

I think some Latin grammarians mentioned the ictus, otherwise it wouldn't even be called that (ictus is Latin for "stress")

I've read about it in my favourite Vox Graeca by W.S. Allen. There, Allen concludes that long (or rather "heavy") syllables had a greater stressability than the "light" ones, which implies they tended to be more emphasized while reciting poetry.

Then we must not ignore it, says I.

[Bombichka]

Really??? You mean ictus is all a modern scam?

I think some Latin grammarians mentioned the ictus, otherwise it wouldn't even be called that (ictus is Latin for "stress")

I've read about it in my favourite Vox Graeca by W.S. Allen. There, Allen concludes that long (or rather "heavy") syllables had a greater stressability than the "light" ones, which implies they tended to be more emphasized while reciting poetry.

Then we must not ignore it, says I.

I agree.

Nevertheless, the problem remains with steps like the spondee. It has two "heavy" syllables, i.e. two "stressable" syllables, of which we're used to put the stress only on the first one. Why?
Also, in iambic trimeters, it sometimes occurs that a step consists of three short syllables, and here too we tend to put the stress on one of them, although, according to Allen's theory, they all shound be "non-stressable".

I guess the verse rhythm has something to do not only with some inherent quality of the syllables themselves, but also with some sort of musical rhythmic measures superimposed on the syllables regardless on whether they sometimes happened to be of the "wrong" kind.

But, then again, when we speak of musical measures, we cannot dispense with the idea of stress/ictus. There's simply no other way.

[chad]

hi, i'm not really following this because i just follow the lengths and pitches in greek, and the lengths and pitch + stress in latin. i don't know about this ictus stuff. in greek, you definitely don't stress either long syllables or accented syllables unless your own judgment of the whole clause demands for added emphasis there, and that's just subjective, i usually don't try to stress at all unless i really understand a clause and have practiced it. in latin, i only stress the syllables with the word accent: you don't stress a syllable just because it's long. i think the ictus stuff is artificial...

[Bardo de Saldo]

I don't know about these things; I just know that the last time I appeared in front of an audience with a skull on my hand and proclaimed:
"to BE, or NOT to BE, that IS the QUESTion" in an impeccable iambic rhythm, I was laughed off stage.

It still hurts.

[annis]

i think the ictus stuff is artificial...

My objection to the arsis/thesis terminology is this: I don't march Homer, I recite it. It might be appropriate vocabulary for an anapestic run in Euripides for an entrance march, but I'm less convinced it's usefull for discussing Homer or non-choral verse.

[annis]

But, then again, when we speak of musical measures, we cannot dispense with the idea of stress/ictus. There's simply no other way.

Really? Plenty of musical traditions have fixed, rhythmical cycles without a strong pulse (stress).

[Bombichka]

But, then again, when we speak of musical measures, we cannot dispense with the idea of stress/ictus. There's simply no other way.

Really? Plenty of musical traditions have fixed, rhythmical cycles without a strong pulse (stress).

The question, then, is, should we number the Greek musical tradition amongst these, and why.

[annis]

Really? Plenty of musical traditions have fixed, rhythmical cycles without a strong pulse (stress).

The question, then, is, should we number the Greek musical tradition amongst these, and why.

Absolutely. I just don't want the possibility tossed out before consideration.

[Bombichka]

Really? Plenty of musical traditions have fixed, rhythmical cycles without a strong pulse (stress).

The question, then, is, should we number the Greek musical tradition amongst these, and why.

Absolutely. I just don't want the possibility tossed out before consideration.

you're right. but music theory isn't my strong side. So i can't take part in the consideration.

but here's one more point that just came to my mind as I was reading Theognis a while ago: syllables that are "long by position", like the first syllable in [face=spionic]ke/ntron[/face] were not really long, at least not as long as the first syllable in [face=spionic]skh=ptron[/face]. Allen quotes a passage from Dionysius of Halicarnassus where he speaks of the ancient musicians measuring the sounds in the following way:

a consonant: 1/2 mora ([face=spionic]xro/noj prw=toj[/face])
a short vowel: 1 mora
a long vowel: 2 morae

thus, from the point of vue of the [face=spionic]r(uqmikoi/[/face], [face=spionic]ke/ntron[/face] is relatively shorter than
[face=spionic]skh=ptron[/face].

but, from a prosodic point of view, they both count as "long" (Allen rightly prefers the Indian term "heavy") syllables in the verse.

now, if verse was based only on short-long syllable alternations, due to the considerable differences between all the kinds of syllables that were regarded as "long", this alternation couldn't have been so neat. so, shouldn't we suppose there must have been something else?

[annis]

but here's one more point that just came to my mind as I was reading Theognis a while ago:

Theognis? Did you read this on your own or for a class? I get the idea he isn't read much these days.

Allen quotes a passage from Dionysius of Halicarnassus where he speaks of the ancient musicians measuring the sounds in the following way:

a consonant: 1/2 mora ([face=spionic]xro/noj prw=toj[/face])
a short vowel: 1 mora
a long vowel: 2 morae

Already we're in a little trouble. In those languages which have long and short vowels, very few have a short:long ratio of 2:1. I believe 1.2:1 is the lowest and about 1.8:1 is quite common. So this mora system is already very likely a bit artificial.

thus, from the point of vue of the [face=spionic]r(uqmikoi/[/face], [face=spionic]ke/ntron[/face] is relatively shorter than
[face=spionic]skh=ptron[/face].

but, from a prosodic point of view, they both count as "long" (Allen rightly prefers the Indian term "heavy") syllables in the verse.

I prefer that terminology myself.

now, if verse was based only on short-long syllable alternations, due to the considerable differences between all the kinds of syllables that were regarded as "long", this alternation couldn't have been so neat. so, shouldn't we suppose there must have been something else?

No, because even if the phonetic reality isn't so neat, the poetic practice followed it. Some modern languages still arrange verse by syllable lengths, and some of those would consider a closed syllable with a long vowel longer than an open syllable with a long vowel (Urdu is one, I believe). So we have comparisons to make, and Greek does appear to consider heavy and extra-heavy syllables the same as far as the meter is concerned.

[Bombichka]

Theognis? Did you read this on your own or for a class? I get the idea he isn't read much these days.

I read it on my own. but, actually, I've had a class on some passages by him some time ago.

Already we're in a little trouble. In those languages which have long and short vowels, very few have a short:long ratio of 2:1. I believe 1.2:1 is the lowest and about 1.8:1 is quite common. So this mora system is already very likely a bit artificial.

Maybe you're right. But, still, the idea of "chronos protos" and the measurement proportions quoted by me are given by the ancient musicians. The whole "mora" thing is based on this, and the contemporary scholarly idea of "mora" is, in fact, a reflexion of the idea of "chronos protos", which is a notion of the native speakers of Ancient Greek. I've no reason not to trust them - even if they're a bit inaccurate from a modern phonological point of view.

No, because even if the phonetic reality isn't so neat, the poetic practice followed it. Some modern languages still arrange verse by syllable lengths, and some of those would consider a closed syllable with a long vowel longer than an open syllable with a long vowel (Urdu is one, I believe). So we have comparisons to make, and Greek does appear to consider heavy and extra-heavy syllables the same as far as the meter is concerned.

I'd like to know more about Indian meter. in fact, it was discussed in a book by the so-often-quoted W.S.Allen, as well as by a study of A.Meillet making a comparison of the Greek and the Vedic practice. But I'm not acquainted with those.

[Bardo de Saldo]

You have convinced me, William. You also agree with the best homeric performers that I know, Danek and Hagel ( http://www.oeaw.ac.at/kal/sh/ ). On their web page, past the bibliography, they talk about hot issues like enjambment and phraseology.

I also like the "flow to the caesura and ebb to the end of the line" metaphor for the hexameter better than the "marching to certain death" one.

[Eureka]

You have convinced me, William. You also agree with the best homeric performers that I know, Danek and Hagel ( http://www.oeaw.ac.at/kal/sh/ ). On their web page, past the bibliography, they talk about hot issues like enjambment and phraseology.

I also like the "flow to the caesura and ebb to the end of the line" metaphor for the hexameter better than the "marching to certain death" one.

Hmmm...

There's a few interesting things in that abstract:

"Thus, early Greek hexameter poetry is likely to have been sung to a fixed set of four notes, the melody governed by word accent and sentence intonation."

4 notes? That would make pitch modeling very easy.

"The performance of Ancient Greek verse, as heard today, which involves the so called ictus which overrides the word accents, has nothing in common with the ancient pronunciation."

OK then.

"With some training anyone who is able to read Homer can achieve to improvise the melody to any given Homeric text easily."



"The Greek aoidoi sang in unison with the accompaniment of the four-stringed phorminx, which implicates the use of only four notes for the melody, too."

Perhaps the other three strings on the kithara were only used between verses...


So, Bardo... You don't happen to understand German do you?

('Cause if you do, I need a fairly large favour: http://www.oeaw.ac.at/kal/sh/whb37.htm )

[Bardo de Saldo]

"So, Bardo... You don't happen to understand German do you?" (Eureka)

I was about to ask you the same question.

The charts and the music notations are self-evident, and this is what I think they do (I listen to the MP3 file): They start by playing the melody of the hexameter that they've chosen for that performance, and then they play and sing variations of that melody adapting it to each line's pitch wave (as marked by the accents). They hit a higher note at the main accented syllables and never drop it (that is, they play the same note) at the secondary ones. In between lines they play the last dactyl and spondee of their melody, with ocassional variations.

Can your ear tell if their highest note is an "octave" higher than the lowest one?

[Eureka]

I used Babelfish on that site, and one thing that's clear is that it is not a how-to guide. I doubt that they could give a full account of their method on that one page alone.

[Eureka]

Hmmm...



Just looking at the bottom line of diagram 2, it's hard to justify the first few notes on the second line being so low on diagram 3. Maybe Hagel is not the messiah.

(Just a note, I'm still trying to work out a system, it's just that the university is monopolising my time at the moment.)

[chad]

hi eureka, just quickly looking at those diagrams with the music staves for the 1st time, the first few notes on the bottom staff look right for a scale within the compass of the staff, as you have a run of graves, so from the first syllable to the acute in ai)gio/xoio is a steady rise, and if you position the accented syllable in ai)gio/xoio at the top of the staff, then the prior syllables must be approximately where they are. it's just in the diagram 2, hagel didn't account for anathesis after the grave su\n for some reason so the diagrams differ there.

[Eureka]

hi eureka, just quickly looking at those diagrams with the music staves for the 1st time, the first few notes on the bottom staff look right for a scale within the compass of the staff, as you have a run of graves, so from the first syllable to the acute in ai)gio/xoio is a steady rise, and if you position the accented syllable in ai)gio/xoio at the top of the staff, then the prior syllables must be approximately where they are. it's just in the diagram 2, hagel didn't account for anathesis after the grave su\n for some reason so the diagrams differ there.

Yes, I seem to remember that Devine and Stevens say that if there is a rise at the beginning of a sentence, then that rise is particularly steep, and so diagram 3 seems justifiable. It clearly disagrees with diagram 2 though, as you say, which is disconcerting.

What's also interesting is that they them seem to have a partial pitch reset at the caesura break (probably treating each caesura as a single minor phrase).

[Eureka]

I was thinking until recently that Hagel & Danek's method (i.e. first determining the overall curve as a smooth, nearly continuous path and then assigning each syllable to its closest note in the range you are using) was probably best. However that bottom stave indicates something very important, that often you must have a rise (or fall) regardless of how many notes you are using, or the minimum distance between the pitches.

That's a good argument, I think, for the basic premise of your method, Chad, whereby we in fact begin with the scale. After all, the contours, when singing, wouldn't necessarily have been as fluid as the contours when speaking.

[Bardo de Saldo]

When you break a line with a caesura, you end up with hemistiches. What are poets for?

I think that the speech wave charts and the music note charts are congruous.

A reason to start ΠΑΡ ΔΙΟΣ low would be the distance to the high note. The music is always upwards to the acute, and from there falls to the caesura, and repeat. It never skips a note and does a doopsie-doo (nice touch) at the circumflexions. If your first high note (acute) is far away, you either start low or play the same string all the way to the top.

The reason why they do that is in the abstract.

It sounds good and primitive, as it should. I think it is a good approximation, as they well call it. What I don't buy is the lack of chords.

[chad]

greek music didn't use chords. it was always based on simple melodies. in playing the stringed instruments it's conjectured that the player might have strummed several dampened strings, but they didn't use chords as we think of them. this is covered in west and other classic books on ancient greek music.

[Eureka]

Hmmm…

Judging by diagram 2, it would seem to imply that any anathesis is momentary. In other words, that it must later be counteracted so that subsequent words in the sentence are at the same pitch as they would be if there were no anathesis.

However, D&S 480: (Refiring to emphatic words.)
“After a focus obstruction on a proper name within a syntactic structure that normally corresponds to a downtrend, downtrend starts over from the higher pitch.”

In other words, the pitch of words following an emphatic are to be calculated via the pitch of the emphatic not the pitch of the word preceding the emphatic.


I haven’t found anything in D&S which refers to the pitch trend after words that have been lifted by graves, but I expect that it would be the same rule.

[Bardo de Saldo]

"greek music didn't use chords" (Chad)

If they played one note at a time, why didn't they use a flute? (Physical restrictions apart, if they were going to sing also... I guess they could play by fart.)

Isn't it the whole point of a string instrument that you can pluck more than one string at a time, and play all the strings in quick succesion with one move?

[chad]

Isn't it the whole point of a string instrument that you can pluck more than one string at a time, and play all the strings in quick succesion with one move?

no

[Eureka]

If I can ask you a technical question about your model, Chad: In working out the height of the drops that take place directly after acutes (or during circumflexes), did you use a simple ratio between the average falls in the Delphic Hymns (as documented in D&S) and the drops on your model?


Anyway, time to document some assumptions: Just like Danek & Hagel, let's assume that the only notes the bard will sing correspond to strings on his instrument. Therefore, he works the natural patterns of speech into a predefined set of notes. In this case, we have been using the seven notes of the kithara (its possible that different kitharai had different pitches).

D&S at several points state two things about grave accents; that they cause anathesis (nonlexical graves more so than lexical graves), but also that they cause the rise from one peak to the next to be more gradual. For our purposes, if we make our syllables rise by one tone each when approaching a peak when not influenced by a grave, then if a grave is involved not every syllable will be higher than the last (the choice being up to the bard).

So, Iliad line 1 could be: (Assuming that Πηληιαδεω is emphatic and that Αχιληος is not, this might be incorrect, but D&S had little evidence on emphatics to work with, and this seems appropriate to me.)

(Ignoring enjambment)


I don't know how high the grave and the emphatic should make the peak of Πηληιαδεω, but D&S possibly has the information necessary.)

If θεα takes an acute then it could be:



(Feel free to disagree at any point. )

[chad]

hi eureka, for the height of the drops (as well as for the height of the rises) i used relative intervals like this, it's not maths at all but just relative measurements:

* there are 3 types of drops, the smallest non-lexicals (= x), then compressed (i.e. catathetic) lexicals (=x + 1), then uncompressed lexicals (= x + 2).

* then the standard rise is a note-by-note rise (for relative purposes, = 1); then because the drop is greater than the rise, i put x=2 if that makes sense, i'm no mathematician

* finally for the absolute pitch height on an accent peak, anathetic highs are the highest (=peak), then the next highest relatively is a standard peak (=peak - 1), then a compressed peak (= peak - 2), then a further compressed peak (= peak - 3).

the huge flaw in this of course is that the relative units of the drops (=1) might not be the same at all as the relative units of the absolute peaks (=1 also). it's the simplest model i could adopt and it seemed to correspond to the musical fragments when i turned this into a model using the standard greek scale.

your point about the rise to the anathetic peak is definitely possible, it makes sense and 2 notes on the same pitch in extant greek music is not uncommon, although i can't see the pics in your post (and i still haven't been able to listen to your music files because of the media blocks on my pc). i wonder if we can get in touch with d&s about some of these points: you've already gone well beyond the limited stuff i've learnt about this

[Eureka]

Never mind that midi I sent you, it's different from those pictures. I'll make a new midi, when I'm confident of the system. (Can anybody else see them, by the way?) Is it something about your computer that sometimes pictures do not show? (Try refreshing a couple of times.)

I also think the basic size of your pitch falls sound appropriate. I'm surprised that, as far as I've read, there's no distinction between the falls after acutes and the falls during circumflexes.


However, I haven't read anywhere that emphatics have a greater pitch drop than non-emphatics. Or have I missed something?

[Bardo de Saldo]

I can see your music notation. It's missing the key. Are the actual notes irrelevant, and you're only interested in their relationship? (Maybe the key is not neccesary if you know the notes in a khitara.)

What's the purpose of your charts? I criticised the music you sent me assuming that it was meant for the ear, not for the brain (hope I didn't rub the wrong way).

[Eureka]

I can see your music notation. It's missing the key. Are the actual notes irrelevant, and you're only interested in their relationship? (Maybe the key is not neccesary if you know the notes in a khitara.)

I’m assuming that the kithara contained 7 notes that were separated by increments of a tone (Danek and Hagel seem to have made a similar assumption in having the 4 notes of the phorminx separated by 2 tone increments). The only other thing I know about them is that they were reasonably high in pitch, but not what the exact pitches are (they probably differed, and each kithara may have been set to its bard’s voice).

If you think they should/would use a different key or a particular set of notes, I’m open to suggestions.

What's the purpose of your charts?

Those two pictures were only posted to illustrate the proposed method of dealing with grave accents. (I've posted some of my thoughts on pitch modelling here in case someone wants to give a critique, or sees faults in the logic.) They are certainly not intended to be a definitive form of line 1 because of several unaddressed factors (so much to do, so little time).

I criticised the music you sent me assuming that it was meant for the ear, not for the brain.

It was only intended to guide a singer; it has no musical merit in itself. (I’m reworking it now, so don’t bother with that midi file.)

(hope I didn't rub the wrong way).

Not at all, criticism is good.