Digamma?

[Eureka]

Do modern editions of the Iliad contain any digammas at all?

Line 33 is definitely missing a digamma in [face=SPIonic]e1dveisen[/face], and, as a result, the line no longer scans. Is it known whether performances in 5th century Athens would have involved the digamma-sound in words such as this?

[annis]

Do modern editions of the Iliad contain any digammas at all?

The most recent book I have that prints the digamma in the body of the poem is a 1907 "American Book Company" edition. If I recall correctly, it's not entirely clear Homer would have pronounced the digamma any longer.

Line 33 is definitely missing a digamma in [face=SPIonic]e1dveisen[/face], and, as a result, the line no longer scans. Is it known whether performances in 5th century Athens would have involved the digamma-sound in words such as this?

Probably not. Based on hints from orthography, various things were tried to make the verses scan where possible. I'm used to seeing your example given as [face=spionic]e)/ddeisen[/face], or least listed so in the ap.crit. (West, somewhat boldly in my opinion, spells it so in the main body of his Iliad).

[Eureka]

Ah, cheers… [face=spionic]e)/ddeisen[/face], that’ll work.

If I recall correctly, it's not entirely clear Homer would have pronounced the digamma any longer.

If he didn’t, he’d have to have his own way of making the syllable long. Wouldn’t that result in it becoming an official “epic” form of that word?

Probably not. Based on hints from orthography, various things were tried to make the verses scan where possible.

Various things

[Eureka]

So, what do we do if the necessary digamma came at the beginning of the word?

[annis]

Various things

Respelling (as in the example above), shuffling words, changing words, introducing decorative and hiatus-breaking [face=spionic]d', t'[/face]. There might be more.

[annis]

:? So, what do we do if the necessary digamma came at the beginning of the word?

What do you mean?

[Eureka]

:? So, what do we do if the necessary digamma came at the beginning of the word?

What do you mean?

Let's say the words were [face=SPIonic]kako\n ve/rgon[/face], and so the digamma was required to make the second syllable long.

It seems somehow less kosher to double the nu in this case, making it [face=SPIonic]kako\n ne/rgon[/face].

[annis]

It seems somehow less kosher to double the nu in this case, making it [face=SPIonic]kako\n ne/rgon[/face].

Ah. I hope Chad has something to say about this. Because my mind really latches onto linguistic minutia, the Homeric dictionary that lives in my brain has most of the digammas. I pronounce them but I'm not sure others should follow my lead in this.

[Eureka]

the Homeric dictionary that lives in my brain has most of the digammas. I pronounce them but I'm not sure others should follow my lead in this.

That's not a bad practice, I think. After all, you could end up accidently writing down [face=SPIonic]ne/rgon[/face] somewhere else, but there's probably no chance of writing [face=SPIonic]ve/rgon[/face] and not noticing the mistake.

[chad]

hi, digamma is a tricky one. on top of the fact that it's not known whether it was pronounced or not (and therefore i don't pronounce it) it's not clearcut how it worked as a "consonant". e.g. with w(/j which had digamma, it makes position in most cases: e.g. 2.190:

[face=SPIonic]daimo/ni' ou)/ se e)/oike kako\n w(\j deidi/ssesqai[/face]

and many other cases. however it can allow elision: e.g. 5.78

[face=SPIonic]a)rhth\r e)te/tukto, qeo\j d' w(\j ti/eto dh/mw|[/face]

and in other cases as well.

my guess is that the greeks didn't pronounce it but were mentally aware of it. from what i've seen of inscriptions, they didn't rely on the spelling of words for scansion: they didn't need to see/write a digamma or another delta to know how to read the 1st syllable in e)/deisen in line 33 (nor do we), but that's just a guess. i'm guessing that a lot of the double-deltas and things like that were added by medieval copyists or modern editors.

[Eureka]

Hmmm... As well as those modern copyists, I wonder how much of the Iliad was Atticised by the Athenians themselves.


I have one further question, this time about the /w/ sound, not the letter digamma itself.

I know that upsilon can indicate a /w/ sound if it is part of a diphthong preceding another vowel, and also that a /w/ follows ου if ου precedes another vowel. Vox Graeca does not mention o or ω in its digamma section, but it does say on page 96 that /w/ can follow o as well.

My question is, does /w/ follow ω, and do any of ε, η, and α have /y/ following them too?

[chad]

hi, i remember that vox graeca discusses the /y/ think if that means consonantal i, i don't know that standard pronunciation/phonetic symbol thing.

vox graeca discusses this, if you have it it's in there somewhere, i remember he talks about )Axaioi/ and other words, re diphthongs ending in iota followed by vowels or something.

[cweb255]

IIRC, the consonantal 'y' sound duplicated certain letters, such as thalattos, but I can't remember where I read this from.

[cweb255]

Ok, I found it now, it's about Verbs with a suffix in the present stem. What happens is the k/g/ch + y > tt; p/b/ph + y > pt; t/d/th + y > z. Like gaurd (phulake) becomes phulatto but fut. phulaxo.

[Eureka]

hi, i remember that vox graeca discusses the /y/ think if that means consonantal i, i don't know that standard pronunciation/phonetic symbol thing.

I think, technically, the symbol should be /j/, but the Wox uses both.


Cweb, I have absolutely no idea what you're saying. k/g + 2x ≥ √cos(iy^2)

(OK, I see you bunched similar consonants together, but when do you add y?)

EDIT: Spelling

[cweb255]

Sorry, it was making nouns like phulake into verbs, which you get phulatto, this is the result of the lost /y/ sound in Greek.

[Eureka]

Sorry, it was making nouns like phulake into verbs, which you get phulatto, this is the result of the lost /y/ sound in Greek.

Oh, well I was asking about semivowels that did exist, not semivowels that had existed.

[cweb255]

Well, I was merely showing you where it was and that it still had an impact.

[Eureka]

Well, I was merely showing you where it was and that it still had an impact.

Hmmm... I thought I knew what you were saying, but now I know I don't.

[Bardo de Saldo]

"the line no longer scans" (Eureka)

The lines always scan, Eureka; Homer wrote them. There's many things that we don't know about his poetry, but we know its rhythm. If a verse starts with a dactyl, especially if it's a standard like ΩΣ ΕΦΑΤ, what follows is going to be the start of the second foot. As such, it is going to hit the ground. In the case of line A33, if you make de first E of édeisen long and keep going at the rhythm, you'll get to the end.

Why Homer thought that it was OK to start the second foot with édeisen, I don't know.

I just remembered this little poem by George Starbuck:

HIGH RENAISSANCE

"Nomine Domini
Theotocopoulos,
None of this prelates can
Manage your name.

Change it. Appeal to their
Hellenophilia.
Sign it 'El Greco.' I'll
Slap on a frame."

[chad]

hi bardo, there are many lines in homer which don't scan, due to homer's errors in the use of his bardic art most probably. homer sometimes repeats a formula from one line (where it works) in another line (where it doesn't). e.g. the familiar 1.26

mh/ se ge/ron koi/lh|sin e)gw\ para\ nhusi\ kixei/w

works because the next consonant makes position. but he uses the same formula in 24.569

mh/ se ge/ron ou)d' au)to\n e)ni\ klisi/h|sin e)a/sw

ou)d' didn't have digamma otherwise other lines wouldn't work, e.g. 1.93

ou)/ t' a)/r o(/ g' eu)xwlh=j e)pime/mfetai ou)d' e(kato/mbhj,

1.220

a)\y d' e)j kouleo\n w)=se me/ga ci/foj, ou)d' a)pi/qhse

i.e. digamma would prevent epic correption in 1.93 and would make position in 1.220.

Will has also mentioned in an article if I recall some scansion errors in certain lines which have a gap between the masculine and feminine caesurae. Milman Parry and other books on homeric formulae discuss other errors; I think Hainsworth covers it in his book "The Flexibility of the Homeric Formula", i can't remember.

[Bardo de Saldo]

The lines of the masters always scan, Mr. chad. If a bard can’t make an hexameter out of a line of Homer, he should blame his rhythm stick.

24.569 ΜΗ’ ΣΕ, ΓΕ’ΡΟΝ, ΟΥΔ ΑΥΤΟ’Ν ΕΝΙ’ ΚΛΙΣΙ’ΗΙΣΙΝ ΕΑ’ΣΩ

In this line, saying the omicron of geron as an omega doesn’t kill anyone. Shakesperean mega-contractions are much harder on the apprentice. I can’t scan Greek on the fly like I can do with Spanish, but my guess is that a Greek bard would come out of the first foot, see all those diphtongs after the omicron, and go for the obvious. When you are at the end of book 24, almost done, the audience antsy, you don’t think of digammas, you just make whatever comes after the first foot long.

1.93 ΟY’Τ Α’Ρ ‘Ο Γ ΕΥΧΩΛΗ¨Σ ΕΠΙΜΕ’ΜΦΕΤΑΙ ΟΥ’Θ ‘ΕΚΑΤΟ’ΜΒΗΣ,

In this line, you start the 4th foot with an epsilon followed by two consonants, good sign. The accent tells you that the ai is short to complete the foot, and the 5th starts with a diphtong. If you put a digamma before outh I don’t see how that would change anything. (That may be your point.)

1.220 Α’Ψ Δ ΕΣ ΚΟΥΛΕΟΝ Ω¨ΣΕ ΜΕ’ΓΑ ΞΙ’ΘΟΣ, ΟΥΔ ΑΠΙ’ΘΗΣΕ

Here, if you put a digamma in front of oud you do mess things up. So why put it?

My point is that the study of digammas is a noble pursuit, and it is curious how digammas explain some “errors”, but one thing is philology and another thing is poetry.

[chad]

hi bardo, this isn't the 1st time that my limited knowledge of metrics has been criticised; i've said all i can in the above post; i'm sure a more experienced person here could continue this discussion

[Bardo de Saldo]

I don't doubt about your knowledge, chad, and I know that you know how to scan the lines that "don't scan". It's just that don Bardo (that's my internet persona, you can call me henry if you prefer) jumps when someone puts his masters' abilities in question.

[Bardo de Saldo]

"If you put a digamma before outh I don’t see how that would change anything." (Me.)

We all know, chad, how bold ignorance can be, especially among young apprentices (I can't use youth as an excuse, I'm afraid). We also know that the hearts of the wise are forgiving, if only for the greater cause of knowledge.

If epimémthetai was followed by a word starting with a consonant, would it loose its accent?

[annis]

Horray for metrics!

"If you put a digamma before outh I don’t see how that would change anything." (Me.)

Depends on where the digamma came from.

Certain kinds of sounds, at the beginning of a syllable, could close the preceding short vowel, even though it would normally be thought of as a single consonant. Those are [face=spionic]l, r, m, n[/face] and sometimes digamma. Why? Because an s sound before those dropped out fairly late in the history of Greek. The aoidoi preserved the old scanning, however. There are some relics of this, like the variation between [face=spionic]mikro/j[/face] and [face=spionic]smikro/j[/face]. Once the s disappeared, later poets adopted the possibility of lengthening a short vowel before all words starting with [face=spionic]l, r, m, n[/face].

The cluster sw- evaprated completely from words, but left metrical traces. There are a few instances of -nw- disolving into an upsilon with a "long" short vowel before it.

[annis]

The lines of the masters always scan, Mr. chad. If a bard can’t make an hexameter out of a line of Homer, he should blame his rhythm stick.

Or the text - Homer comes to us by a very twisty route. It is not everywhere secure.

The history of the hexameter is one of steadily increasing regularity. Homer's practice (let's call them "licenses" rather than "errors") - distorting vowel lengths, lengthening syllables in the princeps[1] by surprise - was cause for comment even by other Greeks not too much later than him.

[1] princeps - the first long of the dactyl. I avoid the ictus, arsis, thesis, etc. terminology.

[chad]

If epimémthetai was followed by a word starting with a consonant, would it loose its accent?

hi bardo, further to what Will said, if it's followed by a consonant, it will be short (for accenting purposes) but long (for scansion). see para 547 of pharr:

547. Final ai and oi are counted short when determining the accent, except in the optative and in oi)/koi... These diphthongs are regularly long in metrical quantity, and must be so treated when reading the verse, although considered short when determining the accent.

and so while in line 1.93 (as cited above) it's short (because no consonant follows) (this is called epic correption):

ou)/ t' a)/r o(/ g' eu)xwlh=j e)pime/mfetai ou)d' e(kato/mbhj,

in e.g. 1.284 it's long because a consonant follows:

e(/rkoj )Axaioi=sin pe/letai pole/moio kakoi=o.

see how the accent doesn't become paroxytone however.