Ancient Greek: how worried do I need to be about "long" and "short" accents?

Question

I've been learning for a couple of months now, from a rather ancient book, which blissfully ignores all questions of accents.

But I recently found out about them (with help from Luke Ranieri among others, and I actually bought Allen's Vox Graeca, and have read most of it!).

I find to my satisfaction that getting obsessed with accented Ancient Greek pronunciation is the ideal way to make a hard task far far harder.

But I've now found another source of potential anxiety: in addition to all the marks ("diacritics"?) you can unearth from the Greek Polytonic keyboard, I find that there is one for "short" (ᾰ) and one for "long" (ᾱ).

My main reference source for looking up words is Wiktionary. I just looked up the word (in my simplified book) λυρα. In Wiktionary the entry for this is λύρα (seemingly corroborated by all the referenced conventional dictionaries) but the Wiktionary definition uses "λῠρᾱ".

I don't seem to recall seeing many of these "short" and "long" diacritics in Wiktionary. Are they actually Ancient Greek in origin (I mean, added by Aristophanes of Byzantium, or others from those times) or much newer innovations? I assume that knowledge of the length of these vowels will actually be very important if, 10 years from now, I get to the point of being able to try to understand Greek metre. But I also just don't want to pronounce things wrongly.

I'm wondering why the word λύρα/λῠρᾱ might have been singled out for this treatment in Wiktionary, and how important it is to write them down (when making vocab lists, as I tend to do as I go along). As far as I can tell, you can't in fact combine these marks with the other diacritics using the Greek Polytonic keyboard, but I could be wrong about that.

I'm also suspecting that there may be no need for these. Isn't a υ on its own always short? And does a long "α" at the end of the nominative of λυρα in fact mark this out as a bit of an oddity? I saw the note in Wiktionary: "A Mediterranean Pre-Greek substrate technical loan. Indo-European etymologies should be rejected."

Edit: combining macron diacritic + acute diacritic

Fresh with the knowledge imparted by Draconis I did a bit of searching to see if there was any solution to this (essentially how to combine a macron and an acute using Greek in Word).

I found that when I looked up his example, λυσω (i.e. future of λυω), that although it's barely discernable, the squiggle above the υ of λυσω is neither a "long" nor an "acute", but a combination. You can copy this to a Word document. And having copied, can then recopy wherever it is needed!

λῡ́σω: how about that? (In fact here this looks really rubbish on my computer screen, but the rendering in both Word and LO Writer is good). This turns out to be the following Unicode combination:

U+1FE1 : GREEK SMALL LETTER UPSILON WITH MACRON
U+0301 : COMBINING ACUTE ACCENT {stress mark; Greek oxia, tonos}

Here's a Word macro which types that:

Sub MacronAcuteU()
    With Selection
        .InsertSymbol CharacterNumber:=8161, Font:="Times New Roman", Unicode:=True
        .InsertSymbol CharacterNumber:=769, Font:="Times New Roman", Unicode:=True
    End With
End Sub

A similar thing can be done for α (8113) and ι (8145).

And you can even add smooth/rough breathing:

Sub MacronAcuteUSmooth()
    With Selection
        .InsertSymbol CharacterNumber:=8161, Font:="Times New Roman", Unicode:=True
        .InsertSymbol CharacterNumber:=787, Font:="Times New Roman", Unicode:=True
        .InsertSymbol CharacterNumber:=769, Font:="Times New Roman", Unicode:=True
    End With
End Sub

Sub MacronAcuteURough()
    With Selection
        .InsertSymbol CharacterNumber:=8161, Font:="Times New Roman", Unicode:=True
        .InsertSymbol CharacterNumber:=788, Font:="Times New Roman", Unicode:=True
        .InsertSymbol CharacterNumber:=769, Font:="Times New Roman", Unicode:=True
    End With
End Sub

In my version of Word, and using this font at least, the result is not good, with the breathing mark and acute pushed to the right. I also tried in LO Writer and with various fonts. Is it possible that some Ancient Greek specialist font might actually render these properly?

Answer 1

I don't seem to recall seeing many of these "short" and "long" diacritics in Wiktionary. Are they actually Ancient Greek in origin (I mean, added by Aristophanes of Byzantium, or others from those times) or much newer innovations?

They're newer. Some ancient authors (though I don't think Aristophanes of Byzantium) used the macron and breve (ᾱ vs ᾰ) to indicate the weight of syllables when talking about poetic meters: a dactyl in poetry, for example, is – ◡ ◡. But the use to indicate specifically vowel length was not done in ancient times.

I assume that knowledge of the length of these vowels will actually be very important if, 10 years from now, I get the point of being able trying to understand Greek metre. But I also just don't want to pronounce things wrongly.

It's most important for meter, but also for pronunciation: the position of the accent, for example, depends on vowel length. Personally, I'm terrible at remembering it myself, so I look it up when I need it; somehow I'm much worse at remembering vowel quantity in Greek than in Latin.

I'm wondering why the word λύρα/λῠρᾱ might have been singled out for this treatment in Wiktionary, and how important it is to write them down (when making vocab lists, as I tend to do as I go along).

Wiktionary tries to do it for all Ancient Greek words. For example, here's θύρα. The inconsistency is due to multiple editors. Personally, I would write it down in the vocab list, but it's not as crucial as remembering the letters themselves—you need to know the letters to look up the accents and lengths in a dictionary after all!

I'm also suspecting that there may be no need for these. Isn't a υ on its own always short?

ε and ο are always short, while η, ω, and diphthongs (including "spurious diphthongs") are always long. The remaining vowel letters, α ι υ, can be either long or short—you'll notice these are the three that support macrons and breves on the Polytonic Greek keyboard. For a specific example, ypsilon is short in λύρα but long in λύσω.

And does a long "α" at the end of the nominative of λυρα in fact mark this out as a bit of an oddity?

Originally, the first declension feminine singular nominative ending was either ᾱ (in 90% of all words) or ᾰ (in a handful of special nouns like θάλαττα/θάλασσα). In Attic, though, this ᾱ shifted to η everywhere except after epsilon, iota, or rho.

In this case, it's after a rho, so the long alpha remained alpha, instead of becoming eta.

Answer 2

Be reassured. Greek is not Hebrew. Words written using the alphabet alone, without dots squiggles and wiggles, are perfectly memorable and perfectly comprehensible. Don’t be overly distracted by the ornaments. Don’t worry about pronouncing the words in a way that would make Aristotle laugh. Remember he didn’t use those things either.

As far as accents are concerned, there are very few places where a different accent means a different meaning (mostly, though not exclusively, monosyllables), and you can pick them up as you go along.

As far as long and short markings are concerned, Greek defaced with them is as distractingly illegible as Latin is when pockmarked with macrons. When you get on to verse you will find that the structure of hexameters means that there are practically no places where the verse doesn’t tell you whether a given α is short or long. (Choruses in tragedy are a different story of course!)

I’m not saying that you shouldn’t be interested in which endings are long and which short and why. I had no idea how fascinating it all was until I read Draconis’ answer and I look forward to pursuing it myself. But don’t overburden yourself. This is all meant to be a pleasure, not a duty!

Answer 3

From the fact that you refer to macrons and breves as "accents" and wonder if you can do without them, I get the sense that you don't really understand what vowel length is. In English, we have what are traditionally called "short" and "long" vowels, but in reality, we've long lost the distinction of vowel length; the "short" and "long" vowels are just completely different vowels now. Ancient Greek, on the other hand, did have a true length distinction, and the meaning of this distinction has to do with the rhythm of the language.

English has a very irregular rhythm. We lengthen stressed syllables while quickly brushing over unstressed syllables, and we speed up or slow down for emphasis.

But if you hear someone speaking, e.g., Japanese, you'll notice that it does not sound like this. It has a steady stressed-unstressed rhythm. This rhythm rarely seems to slow down for any reason, and even when it does, the rhythm usually stays steady at the slower tempo; it doesn't pause. They convey emphasis primarily through intonation, not by lengthening syllables.

This is because Japanese—like Ancient Greek—is pitch accented. Accent is marked by intonation, not stress. Stress does not carry any meaning in Japanese; instead, it just has a steady pulse of alternating stresses and non-stresses that starts at the beginning of the utterance and doesn't care (so to speak) about how it matches up with the syllables of the words. The same word might be stressed differently depending on where it happens to fall in an utterance, and the speaker is free to reset the alternation mid-utterance if they feel like it.

Ancient Greek might not have worked exactly the same as Japanese, but it was probably similar in having some kind of steady alternation of stresses and non-stresses that wasn't required to match up with a particular syllable of each word. They probably emphasized words via intonation and did not slow down or lengthen syllables for emphasis, at least not regularly the way we do in English.

And lastly, there's one more complication: In Ancient Greek, syllables did not match up one-for-one with rhythmic units. Short vowels did only take one beat, but long vowels took two beats. This is what the long-short distinction means. Closed syllables—those that end in a consonant—also took two beats, as well as syllables containing diphthongs.

Japanese actually has true long and short vowels too, by the way; I just didn't mention it for simplicity. Also, I've been using Japanese as an example because I know it well, but Lithuanian is probably a better example, since it's Indo-European on top of being pitch accented and having distinctive vowel length. I just didn't use it because I don't know much about it, so I can't talk as confidently about it. In any case, Japanese or Lithuanian is probably a good language to listen to to get an idea of what pitch accented languages with long and short vowels sound like.

So can you do without vowel length? Well, sure. It's a dead language; who cares how you pronounce it? But you evidently aren't the type to be satisfied with that, so I would say that yes, vowel length is absolutely worth memorizing—and pronouncing correctly.

Sorry if I just explained stuff you already know, but I can't imagine someone who cares enough about pronunciation to buy and read Vox Graeca wondering if he can do without vowel length, if he knows what vowel length is.

Anyways, remembering vowel lengths isn't as difficult as you might think. As Draconis said, of the seven vowels in Ancient Greek, α, ε, ι, ο, υ, η, and ω—technically nine, counting the "spurious diphthongs" ει and ου—only three of them have ambiguous vowel length, α, ι, and υ. Of those three, one of them only has distinctive length in specific contexts, after ρ, ε, or ι. If the vowel shows up in the word's ending, then you can immediately tell its length by what the ending is; e.g., the -ας in the first declension genitive singular or accusative plural is long, while the -ας in the third declension accusative plural is short. (... Except when the nominative singular is -εύς.) Within the stem, these vowels are short more often than not; the most common exception is when vowel contraction or compensatory lengthening is involved.