V's, W's, and reconstructed pronunciations

[Marius Magnus]

In another thread:

Oh geez Marius, don't get into the idiots who tell you Latin v's are pronounced as w's...that's professorial stupidity.
I mean...say "veni, vidi, vici" (whether with the Italian ch sound or not) with the v's pronounced that way, and it'll sound like baby talk.

Well, a few things:

First, there is no single "pronunciation of Latin" at all. The language changed, in pronunciation and grammar, over the several centuries spanned by our extant documents. The so-called "Ecclesiastical Pronunciation" reflects the pronunciation in Italy toward the very end of the Empire, which is why it is essentially the same pronunciation as modern Italian. The diphthongs ae and oe merged to e early on, and afterward C and G were palatalized before soft vowels; and the semivowels W and Y gradually shifted to the consonants V and J.

In the earlier Classical period, however, V and I were most certainly pronounced as semivowels. There is plenty of evidence for this; notably Roman spelling (why would they have used the same letter to represent two different sounds? They didn't have any historical reasons to do so as we do today; it was their own alphabet, after all), and also in borrowings into other languages (the English wine and vine, for example, are both borrowed from the same Latin word, but at different points in time).

And in pre-Classical times, pronunciation was still different. The letter F (borrowed from the Greek digamma) originally represented our sound W, while in the places where V would later show up, proto-Latin had GW. A pre-Classical sound shift brought W to F, and GW to W. There is linguistic evidence of this as well, notably in PIE cognates:

Proto-Latin *gwigwo vs. English quick (as in, "the quick and the dead")

Also, there are parallel processes all across the linguistic spectrum of W becoming V in some dialects, remaining W in others (this being one of the consonant shifts which separated the Anglo-Frisian branch from the main Germanic).


Second, what is actually silly-sounding about "Wehnee, weedee, weekee" (they are all long vowels, are they not?) in the first place? It certainly doesn't make any sense to judge the sound of the phrase by modern English aesthetic values. Remember that these words had "immediate meaning" to the Romans; they were not mere sounds, but came with their own array of emotional associations. Say it like you're actually serious about it, and believe me, it can sound compelling.

There is nothing inherently "silly" about W sounds anyway...we have plenty of "heavy", "respectable"-sounding short W-words in English as well, such as "wealth", "worth", "well", "war", "weight", etc.

Also it is worth noting that English, so I am told, sounds rather ugly to speakers of most other languages, even the poetic phrases that we find beautiful.

 

[QMF]

I didn't think the w sound was like that, it's the "way" "wee" "wee" sound. I dunno, it just sounds silly.
This is a good thread topic though.

 

[Andy]

Another hint of this may be found in Spanish. For instance the words ciudadano (see-ü-da-da-nou) (citizen) and civil (see-bill) (civilian, adj.), if I spelled this out the Roman way:

CIVDADANO
CIVIL

Most certainly they're words with common origins. Hence, hinting at the u sound of v in Latin. Civitas was probably pronounced Kee-uee-täs.

I agree with Marius that one should not judge with today's standards yesterday's worth.

 

[Marius Magnus]

Spanish etymology is pretty interesting, also. Whereas in Italy, the Latin combinations ci, ce became chi, che, in the Iberian peninsula they became tsi, tse. These were their values by the 16th century, when the Conquest of the Americas was taking place.

Later, the now-regionally-divided language further diverged in pronunciation, with the Spanish in the Americas going from /ts/ -> /s/, and the Spanish of Spain going /ts/ -> /th/. Which is the reason today that Castilian has a "lisp" and Latin American Spanish does not.

It is also interesting that in the same time period (16th c.), the initial F was not yet lost in Spanish either, and so

facer vs. hacer (to make, to do)
fijo vs. hijo (son)
fermosa vs. hermosa (beautiful)
fierro vs. hierro (iron)
Fernando vs. Hernando (Ferdinand, which is a Germanic name)

In the Old Spanish, the Latin cognates are much more readily apparent.

In fijo, the transformation from the Latin -li- to -j- (pronounced /kh/) is consistent in other forms also: mujer, mejor, etc.

 

[QMF]

Civitas was probably pronounced Kee-uee-täs

It's very difficult to phonetically illustrate the difference between the long i and the short i when they are both pronounced as "ee". The long is essentially the same sound said for twice as long...but it's still difficult.

 

[Marius Magnus]

It's the difference between the vowel sounds in seat vs. seed.

The difficulty for English speakers is producing long and short vowels independently of the surrounding consonants, since in English, vowel length is an allophonic variation determined by context: generally, vowels are short before unvoiced consonants, and long before voiced ones.

 

[Cato]

It's the difference between the vowel sounds in seat vs. seed.

The difficulty for English speakers is producing long and short vowels independently of the surrounding consonants, since in English, vowel length is an allophonic variation determined by context: generally, vowels are short before unvoiced consonants, and long before voiced ones.

This is an excellent point Marius, and one I may borrow. I am certainly not a phonetics expert, but would another example in English be "let" and "led", where "led" can be pronounced for a longer time than "let"?

My thinking here is that this might offer insight into the importance of vowel quantity to the Roman ear. Obviously the rules would be different, but examples in English may make the idea less far-fetched for new students of Latin poetry...

 

[Marius Magnus]

I am certainly not a phonetics expert, but would another example in English be "let" and "led", where "led" can be pronounced for a longer time than "let"?

Yes; in fact, I have yet to come up with any counterexamples, especially in final syllables:

phase vs. face
had vs. hat
tried vs. trite
ledger vs. lecher
noise vs. Noyce
house (verb) vs. house (noun)

Note that it applies even to diphthongs. Actually, it applies even where there is an intervening liquid or nasal:

planned vs. plant
send vs. sent
cold vs. colt
chord vs. court

The last may present some disagreement, though, because in non-rhotic varieties of English (such as British or Australian), the r makes the previous vowel long. I'm not sure if that makes the vowel in chord doubly long or not; there do exist languages with three (and even four!) gradations of quantity.

So if you want to teach students to pronounce long vowels properly, one approach would be to ask them to interchange the final consonants of such pairs of words while maintaining the original vowel quantity.

However, a harder trick will be teaching them to pronounce long vowels in unstressed syllables, which I think never happens in English.

My thinking here is that this might offer insight into the importance of vowel quantity to the Roman ear. Obviously the rules would be different, but examples in English may make the idea less far-fetched for new students of Latin poetry...

It might even help to compose some poems in English that use the Latin poetic devices based on syllable length. The rules for determining syllable length are slightly different in English than in Latin, however; merely ending in a consonant is not enough to make a syllable long as a whole, because "let" is definitely short...so perhaps a syllable has to end in a voiced consonant, or two consonants, or something like that.

Or did I just repeat what you said? Anyway, I keep telling myself I'm going to write some English poetry according to the Latin rules, but I haven't got around to it yet (I have to decide just what the rules fo English syllable length are first...).

 

[Marius Magnus]

Incidentally, modern Italian has allophonic vowel quantity also, with a very simple rule: open syllables have a long vowel, and closed syllables have a short vowel. Therefore something like fettuccine all'Alfredo with macrons expressed would be:

fettuccīnē all'Alfrēdō

Modern Spanish, on the other hand, has no vowel length at all: all vowels are short (and there are also no geminate consonants, except rr).

 

[Andy]

Modern Spanish, on the other hand, has no vowel length at all: all vowels are short.

I would disagree most respectfully with that (that is the most polite way of saying - I am most likely wrong )

But notice the difference between canto as in el canto (the singing) and calor (heat).

In the following, understand ^ as a macron.

El Cânto v. Calor.