Long Syllable

Latin Learning Latin
1

I am a real stickler about pronouncing Latin (and Ancient Greek) as correctly as I can. Just the right quality and length of vowels, just the right elision, just the right pronunciation of gn, bs/bt, s (retracted), z, r, final m (without turning it into a velar nasal), etc.

However, I have come across two different ways of determining which syllables with short vowels followed by consonants are considered “long by position”.

I traditionally learned it this way (as the Bantam New College Latin & English Dictionary: Third Edition states):

A syllable is long . . . if it contains a short vowel followed by x, z, or any two consonants except a mute (b, d, g, p, t, c) follwed by l or r: sax•um, Ma•zentius, mit•tō, cur•sor (such a syllable is said to be > long by position> , but the vowel is pronounced > short

.

This is similar to what the Dickinson College Commentaries say:

A syllable containing a short vowel followed by two consonants (except a mute before l or r) or by a double consonant (x, z) is said to be > long by position> , but the vowel is pronounced > short> .

Henle’s Grammar also rings similar, though with an interesting addition:

A syllable is long if it contains a vowel that is long by nature or a vowel that is followed by two consonants other than a mute (c, g, p, b, t, d) > or f > followed by a liquid (r, l)

So Henle says pretty much the same thing, except that he adds that any short vowel followed by -fr and -fl is still short. It is this convention that I have followed in all my many hours of reading Latin.


However,
A first hint of a challenge to this system came in the Companion book to Familia Rōmāna, which says the following about this topic (in a footnote!):

If a combination of letters could be used to begin a word (like the > sp > in > hi spa ni a> ), those letters are
kept together and go with the following vowel.

This would mean that in a word like asper, the first syllable would be short instead of long! It would be pronounced a•sper instead of as•per! However, this can’t be the case (at least in poetry). Look at the following line of the Aeneid:

ōstia, dīves opum studiīsque asperrima bellī > - > Aeneid > 1.14

According to the meter of dactylic hexameter, the first syllable of asperrima must be long. It must be pronounced as•perrima and NOT a•sperrima.

Okay, so I stick with the first convention. However, I just came across this passage from GILDERSLEEVES itself which leads me to really question my convention:

Any combination of consonants that can begin a word (including > mn> , under Greek influence) belongs to the following vowel ; in other combinations the first consonant belongs to the preceding vowel : > a•sper> , > rough> ; > fau•stus> , > lucky> ; > li•bri> , > books> ; > a•nmis> , > river> .

So according to Gildersleeve, the first syllable of asper IS actually short!

Which method is correct?

2

Sorry, I accidentally posted this twice. The website started glitching and I couldn’t figure out what was happening.

3

From what I’ve seen, the first syllable of “asper” is treated as long by all the ancient poets (Lucretius, Vergil, Ovid, Horace, Tibullus etc.), and it appears quite often in poetry, so we can be very sure it is, in fact, long. The same goes for “amnis”. This makes me wonder if Gildersleeve isn’t talking about syllable division as something separate from syllable length because I cannot imagine that he got this wrong. But I’ll wait what others have to say.

4

Oh my goodness you are right! Yes, that part I quoted is in the section about Syllable Division (and that footnote I quoted from the Companion to Familia Rōmāna is also talking about syllable division). Later on Gildersleeve talks about syllable quantity and does make that mute + liquid distinction.

Interesting. So I must be confusing syllable division from syllable quantity. I thought that syllable quantity was based off of syllable division. According to Gildersleeve, asper would be divided as a•sper but the first syllable would still be long. So is syllable quantity not directly linked to syllable division?

In addition, I’d be curious to know everyone’s thoughts on Henle’s addition of “fr/fl” to the mix. Would you all consider the word Africa with the first syllable long or short? According to Henle, the first syllable is short. According to everyone else, the first syllable is long. Virgil (as in the Aeneid, line 86) seems to consider it long.

Okay, so I guess Gildersleeve had it right, as did the dictionary and the Dickinson commentaries. And Henle was wrong to add in “fr/fl”?

5

The syllable division method for working out quantity on paper is just a school thing.

fr:

Corripit Aeneas extemplo avidusque refringit

6

The thing about “Africa” is that the initial “a” is long (cf. the related word “Afer”, where the quantity of the vowel can be seen), so the syllable is long anyway.

Do you know “Musisque Deoque”? (https://mizar.unive.it/mqdq/public/ricerca/avanzata) It is an excellent tool for all things metrical because you can look up how certain words are used by the poets. You can search for “refr*”, for example, and find many verses like the one jeidsath cited.

7

Oh, right. Thanks for reminding me. It does make sense that short syllables followed by fr/fl would be short. I wonder why Gildersleeve and the others don’t include it?

8

I don’t know about the others, but my Gildersleeve & Lodge has the rule at 704.

https://archive.org/details/gildersleeveslat00gilduoft/page/446/mode/2up

9

A possible point of confusion with G&L.

G&L section 6 (page 3) groups the fricative, f, in with the stops, calling the group the mutes.

With this definition in mind, section 704 (rule III) is taken to apply to both stops and the fricative, f.


Allen, Vox Latina, covers this issue in a footnote on page 90.

“The situation with f + liquid is not entirely clear.”
Here, Allen is contrasting the rule for stops (aka plosives) + liquid with f + liquid. That is, he is not grouping f in with the stops (as in G&L).

He gives a few examples where the vowel preceding the f is short:

  • in compounds (where there is a morpheme boundary), such a re-fringo;
  • in certain personal names, Ru-fras and Sa-froni;
  • in one example of a non-compound and non-personal name, va-frae.

Allen does not give a general rule for f + liquid, however, so he is not exactly agreeing with G&L on this.

10

Thank you so much. This helped resolve the question. I didn’t notice that Gildersleeve includes “f” with the mutes. I guess he refrains from calling the group “stops” because “f” is a fricative in contrast to a plosive. But it’s interesting to note that this (in light of Allen’s opinion) actually is a point of contention amongst linguistics. For me personally, I think that a syllable containing a short vowel plus fr/fl should be short (unless there is a specific instance in poetry when it has to be long). It sounds the most natural to me.

11

You also asked about s + plosive, as in asper.

I’m looking at one of Allen’s earlier books, Accent and Rhythm, 1973, in which he discusses this.

On page 137, he states that this regularly results in a preceding heavy syllable, e.g. pes.tis, constituting “arrest + release.”

That is, the s arrests the preceding syllable, ensuring that it is heavy, regardless of vowel length.


He then continues on page 137 to discuss the case of plosive (and f) + liquid.

For plosives, he states that the following syllable is treated as a “complex release,” that is, the preceding syllable is not arrested, and therefore is light (assuming that the vowel is short).

Examples are tene.brae, pa.tris, and po.plus.


But in considering f + liquid, Allen says that it is a case of “arrest + release.” The f arrests the preceding syllable (as does an s), ensuring that it is heavy. This is contradictory to your feeling of what sounds natural.


So we see that in Allen’s earlier work, he is more convinced that f + liquid should be treated as “arrest + release” than he seems to be in Vox Latina, where he says that things are “not entirely clear.”


As for the underlying linguistic theory on this, Allen mentions that “difference in aperture” accounts for f arresting the preceding syllable, contrasted with a plosive not arresting the preceding syllable.

“…the difference in aperture between f and a liquid is less than that between a plosive and a liquid…”

12

That makes sense.

That also makes sense. I feel like whether the syllable is long or short is a mixed back, but it seems to be more short than long, from what I’ve seen. Here are some quotations from the Aeneid and On the Nature of Things:

Short
Silvifragis vexat flabris: ita perfurit acri - On the Nature of Things 1.275
Accipitres atque ossifragae mergique marinis - On the Nature of Things 5.1079
Latonaeque genus duplex Ianumque bifrontem - Aeneid 12.198
Verberat hic idem, cum ducitur atque reflatur - On the Nature of Things 4.938
Impavidos, illam tereti ceruice reflexam - Aeneid 8.633 (also in 10.535, 632 and in 11.622)
Cum refluit campis et iam se condidit alueo. - Aeneid 9.32

Long
Hic natura suis refrenat viribus auctum - On the Nature of Things 2.1121 (also in 4.1085, 6.531)
Refrigescit enim cunctando plaga per auras - On the Nature of Things 4.703
Inde ubi iam morbi reflexit causa reditque - On the Nature of Things 3.502

13

Thanks for putting together this good set of examples.

All nine of your examples involve compound words so we might expect them to fall into Allen’s example of re-fringõ, that is complex release of the fr or fl, and the preceding syllable light.

The first six do fall into that class.

The last three examples, still being compound words, have arrest + release, with the preceding arrested syllable (ref) being heavy.

hīc.nā|tū.ra.su|īs.ref|rē.nat|vī.ri.bu|sauc.tum

ref.rī|gēs.ci.te|nim …

in.de.u|bi_iam.mor|bī.ref| …

I think I can conclude from this set of examples, comprising only compound words, that the expected complex release is going to be the most likely, but metrical constraints may shift the result to arrest + release.

Another conclusion seems to be that compound words are going to be the most likely to involve the fr/fl issue.