The section on stress in Vox Graeca

[chirpis]

Can anyone give a better explanation of the rules of the hypothetical stress accent described in the last section of the last chapter of Vox Graeca, or give some examples? I find it rather difficult to follow. I think jeidsath also mentioned that there are typos in this section. Is that right?

[chirpis]

There's a chapter about it in the last edition of Vox Graeca, although some printing errors on the formulas make it hard to parse. Accent and Rhythm has a longer discussion. Martin L. West has a critique of the theory it in his review, which can be found on JSTOR.

Here's the quote I'm referring to. Could you elaborate on the printing errors? Also, what is the title of West's review?

[jeidsath]

Vox Graeca pg. 135

Show

In the above, the third "not derivable" formula is derivable, and the second "derivable" formula is not. At least that's how it seems to me. Allen has a text-only version of these rules on page 333 of Accent and Rhythm.

The article that I mentioned is:

Reviewed Work: Accent and Rhythm. Prosodic Features of Latin and Greek. A Study in Theory and Reconstruction by W. Sidney Allen
Review by: M. L. West
Gnomon
48. Bd., H. 1 (Feb., 1976), pp. 1-8

https://www.jstor.org/stable/27686376

I've done some stress marking of words in a few metrical texts, by Allen's rules, and I now think that if there is a tendency like the one Allen talks about, it's a weak one.

[David Beckham]

Very interesting topic.

Anyone who attributes to W S Allen the idea that there was any sort of stress accent in ancient Greek is mistaken - West, by the way, does this on p.5 of his review. He did hold that some syllables were stressed in ancient Greek speech but this is not the same thing: the last syllable of a word could (according to Allen) be either stressed or unstressed depending on its context. If I remember rightly (no doubt I should have checked this before writing) accentuation is so defined that it cannot behave like this; the accented syllable of a word has to remain, at least for the most part (admittedly enclitics constitute an exception to this), what it is regardless of context. This would be the reason chapter 16b of Accent and Rhythm is entitled "Non-accentual Stress".

I see West at one point claimed that Allen came close to maintaining that "the stress varies to suit the metre". I cannot see that this is fair. The only way his position comes at all close to being like this is that he held that a verse-end or a caesura could cause the stress on the immediately preceding word to be pushed back by one syllable, just as in prose a sentence-end could have the same effect. I think Allen even replied to this criticism in one of the later editions of Accent and Rhythm.

I would love to have an in-depth debate about this but no time right at the minute.

[mwh]

The practical problem is that hardly any native speaker of English is capable of divorcing pitch from stress. And if we could hear an ancient Greek speaking—which we can’t—only a highly trained phonetician would be able to tell just what was going on. We fool ourselves if we think we can reproduce ancient Greek speech. Not to mention that as with any language pronunciation varied enormously over time and over place.

If you haven’t yet studied Devine and Stephens, I’d say that should be the next step.

[smitterle]

I believe it is very interesting to compare with other accentual systems. The Basque language seems to be using several different systems depending on region http://www.ehu.eus/ojs/index.php/ASJU/a ... /7813/7007

When I learned it it was very interesting to notice the lack of a fixed stress and still a correct and a wrong way to use tone depending on the functional context.

Interesting to notice, too, that Norwegian and Swedish are very much alike (mutually intelligible) although Swedish does have a tone accent while the other hasn't.

I guess the Greek pitch accent (that I'm very eager to get right) is focusing the Attic way.

Today I'll finally get Devine and Stephens. Hoping to find some answers.

[David Beckham]

The practical problem is that hardly any native speaker of English is capable of divorcing pitch from stress. And if we could hear an ancient Greek speaking—which we can’t—only a highly trained phonetician would be able to tell just what was going on. We fool ourselves if we think we can reproduce ancient Greek speech. Not to mention that as with any language pronunciation varied enormously over time and over place.

If you haven’t yet studied Devine and Stephens, I’d say that should be the next step.

Here you mention both a theoretical problem - how can we know what ancient Greek sounded like? - and a practical problem: even given some recommended pronunciation, is it psychologically and physiologically possible for us to reproduce it? I think the two problems should be separated: even if one problem is insoluble, that does not necessarily mean they both are.

I am no expert on the theoretical problem, so will not try to comment on it here. I did once read a few pages of D&S and concluded that I certainly do not have the time, and probably not the expertise, to read the whole book.

Regarding the practical problem of separating pitch from stress, I am not as pessimistic as you are. I think I have heard a few recordings on the web made by people who were managing to do exactly that. Anyway, if you resign yourself to being unable to separate pitch from stress, how should you read ancient Greek poetry? I can see only two alternatives.

One is to stress the (pitch)-accented syllables. I presume this is what modern Greek speakers do, but to me it makes the poetry sound completely unmetrical - especially as we English speakers also have a tendency to lengthen stressed syllables.

The other is to ignore the Greek pitch-accent and follow your natural tendency to let the pitch rise on the syllables that you think should be stressed. If, following Allen, you think that these coincide by and large with the syllables bearing the ictus, the poetry will certainly come out sounding metrical - but also, in my opinion, very monotonous and boring. Besides, if you are going to ignore the Greek pitch-accent when reading poetry, should you not, in the interest of consistency, also ignore it when reading prose? So, while separating pitch from stress is certainly difficult, it is going to be the only aesthetically satisfying solution for me personally.

Allen thought there was a tendency for the mora following a high-pitched mora to carry some stress. I think, if you practise for long enough, this begins to feel almost natural.

[mwh]

What I do is substitute stress for pitch. That's to say, I put some stress—not more than I can help—on the accented syllables (ignoring graves), and let the intonation take care of itself. What’s paramount, at least in verse and non-koine prose, is a sense of the rhythm, which is primarily a function of the relation between light and heavy syllables. (So yes, modern Greek reading of Homer is a monstrosity.) I don’t believe in ictus.

I know this is not authentic, but it’s ersatz-authentic, so to speak, and for me is a good enough approximation, and acceptable in terms of auditory experience, as an authentic pronunciation could not be.

[jeidsath]

The meter comes from getting the vowel quantities right and practicing until you can read fast and smoothly. The accent works best, I've found, by reading from accent to accent rather than word to word. Just as a matter of mental focus that is, not pausing on the accents or anything. I try to use a pitch accent, but it's only slowly improving over the course of years.

[David Beckham]

It is about time someone answered Chirpis's original question. My nearest library does not seem to have Vox Graeca so I looked at chapter 16 of Accent and Rhythm instead.

It may be worth mentioning that Allen thought (p.311) that stress in ancient Greek was much lighter than in heavily stressed languages like English.

Allen's theory is simple enough so long as you disregard secondary stresses and pre-pausal words, i.e. words immediately preceding a sentence-end, a verse-end or a caesura. His hypothesis was that every word which was not pre-pausal would have a "primary stress" on the last heavy syllable, unless the last three syllables were all light, in which case the primary stress would be on the third last syllable if the fourth last syllable was light and on the fourth and second last syllables if the fourth last syllable was heavy (p.333).

So far, so simple. To understand the further refinements of his theory, you have to get your head around the concept of a "stress-matrix" which is generally, though not always (p.293), equal to either two light syllables or one heavy syllable. Most of the time, he talked about stressing stress-matrixes rather than simply syllables. Secondary stress is governed by the rule: starting from the primary stress in a non pre-pausal word and working backwards, every second stress-matrix is stressed. Pre-pausal stress is governed by the rule: starting from the primary stress in the last word before a pre-pausal word and working forwards, every second stress-matrix is again stressed (p.334).

When a stress-matrix can consist of a single light syllable - this is where things become seriously complicated.