Pronunciation of Gnaeus

Pacifica

grammaticissima
Staff member
Someone more expert in phonetics than I am will probably answer your question, but my, did you make me chuckle.

What I do know is that a word like gnatus became natus, so it would seem to suggest that that gn sound in the beginning of a word could be problematic for native speakers as well. Now, was it always resolved to a simple n even where it went on being spelled as gn, or was a full g restored? Or did something else happen? I'm not sure.
 

Godmy

Sīmia Illūstris
My general recommendation with the reconstructed pronunciation(s) is: don't rely on models/recordings and on imitation! All of them may always be slightly impefect + you may not always hear the thing correctly even if the speaker articulates it corretly, because the sound may be too alien to your ears => then the model is useless for you nonetheless.


1) learn phonetic notation and understand basics of phonetics

2) learn how to articulate both random vowels and random consonants just based on a phonetical description PRIOR to hearing them. Especially for vowels this is important, because there may be theoretically an infinite amount of vowels when it comes to quality and it is important that wherever I point my finger on the mouth/tongue phonetic vowel map, you will be able to pronounce it no matter that it hasn't been done before you in your living memory.

If you manage to do that, the quirks of the restored Greek and Latin pronunciations won't startle you even if you don't get it 100% correctly. And then, yes, being able to hear to various other [modern] languages with different phonologies will give you an inspiration here and there how to adjust your ancient language pronunciation... (but one should be really careful while doing this, not just blindly copy some modern language, that would be most likely erroneous).
 
Last edited:

Anbrutal Russicus

Active Member
Initial G in GN was certainly silent already by the time of Plautus. There's some probability that it could have been retained if a vowel preceded, especially in some stable combinations - in this case it sounded as in the middle of the word (eg. agnōmen), namely as you describe, [ŋ(g).n]. In the Classical pronunciation it's a simple [n].
I have no idea how our normally well-informed Godmy came up with the pronunciation above, which is [ɲ]. This is how it's pronounced in most if not all Romance pronunciations of Latin including the Roman Ecclesiastical, but this sound first appeared post-Classically (probably imported from Oscan), was always geminate and never existed at the start of the word until /gn/ was introduced to the language(s) again from Germanic in the Middle Ages.
(So, do we have the German barbarians to thank for gnocchi?..)
 
Last edited:

Anbrutal Russicus

Active Member
One other thing worth addressing is the syllabification /na.e̯us/. There's just two attestations of this word in poetry, both in the awesome all-verse grammar of Terentianus Maurus, which kind of hints that others might have felt ill at ease with its syllabification. Maurus' heavy first syllable means for him it contained a dipthong. This necessarily means that the second syllable starts with a glide: Latin bans sticking a single consonant to the left syllable, forcing the syllabification of PRAEOPTARE as /pra.e̯op.tā.re/. The name most probably is the same word as naevus 'mole', and in this case that glide would be the unspelt /w/ - it's common in Latin orthography to not spell double VV, eg FLA(V)VS, I(V)VENIS, POS(V)VIT. Of course it's conceivable that the African Maurus had simply never heard the name in person, or it had become conflated with the word for 'mole', or this was the same archaizing phenomenon as in /Dī.ā.na/, /dī.vus/ for /de.us/, both from *dejw-.

Or the glide could have been /e̯~j/ - /nae.jus/. In that case the difference from the spelling GAIVS /gā.i.(j)us/ would be that the I made the latter trisyllabic - tho you'd still have to know it wasn't read like MAIVS /maj.jus/, which it eventually was. I've often thought about the problem of these two names' spelling and I'm still on the fence about how they were actually pronounced.
 
Last edited:

Godmy

Sīmia Illūstris
You, are mistaken, @Anbrutal Russicus: it wasn't [ɲ] but [ŋ] which is a difference. A velar nasal [stop], not a palatal nasal. See Sidney Allen, Vox Latina, page 23 (I hope you aren't event commenting on that issue without prior reading the book or a similar resource).

That is why I said one should get good training in phonetics and THEN study the restored pronunciation instead of relying on recordings/models because they won't be able to hear alien sounds to them correctly anyway even when they were articulated correctly. You heard it wrong. What you meant (what is used in Romance languages and the ecclesiastic pronunciation) is not [ŋ] which is what I pronounced (and it's also in certain positions present in my native language) but indeed [ɲ] which is also present strongly in my native language and IT IS NOT what I said. I could never mistake those two sounds as they would constitute a minimal pair in my own language! (and I did not) If you wish, we can take the audio apart or I can make a recording where I can say both sounds so you may hear how they differ.

That's what happens when one doesn't study the academic sources for the restored pronunciation or isn't a skilled phonetician.

This has nothing to do with the palatal nasal, this has nothing to do with Romance pronunciations! (goddammit)

Maurus' heavy first syllable means for him it contained a dipthong
What I said could be (in some instances) taken as a diphthong as there is often an initial release and in the same time it is what Sidney suggest might have been happening [ŋn]. But it's not certain.
 
Last edited:

Pacifica

grammaticissima
Staff member
I can't speak to the pronunciation of Latin gn, whether your recording is right or not, but in any case I can confirm that you didn't say [ɲ]. [ɲ] exists in my language as well and what you say in the recording definitely isn't it (though I guess it's kind of close — kind of close, but not it).
 

Pacifica

grammaticissima
Staff member
Words like trépigner, oignon, enseigner... Or at least that's what Wiktionary says. I'm not actually sure how that differs from /nj/. It sounds pretty much the same to me.
 
Last edited:

Bitmap

Civis Illustris
Words like trépigner, oignon, enseigner... Or at least that's what Wiktionary says. I'm not actually sure how that differs from /nj/. It sounds pretty much the same to me.
To me too.
 

Pacifica

grammaticissima
Staff member
Si ita est, quinam sonus sit [ɲ] nescio, nec de eo superius loqui debui.
 

Anbrutal Russicus

Active Member
That's what happens when one doesn't study the academic sources for the restored pronunciation or isn't a skilled phonetician.
Totally incorrect on both counts, I'm afraid ^_^ I'm one of the least suitable targets around to direct this kind of comments at. Probably half the PDFs on my PC are academic sources on phonetics, and two thirds of these have something to do with Latin.

What you're saying might not be an exact [ɲ], but it's certainly palatal and not a [ŋ]. Your language has no /ŋ/ as a phoneme so you cannot claim any native language advantage in producing it. What you produce is not Allen's [ŋn] - it's one sound, and there's no tongue-teeth contact required to make an [n]. That one sound is not [ŋ] either because, again, it's clearly palatal (as a Russian I know palatal when I hear it), so [ɲ] is as good a description as any. Palatalised velar nasal? I don't think there's an IPA for that - in fact linguists struggle to pinpoint the articulation of Romance palatals, especially of the lateral (Italian gli).

Here's how a single initial [ŋ] sounds (make sure to check several languages). It's plain to hear that these lack the palatal articulation in your recording.

I've already mentioned an example word where a word-initial [ɲ] can be heard - Italian gnocchi. Here's an example before /a/: Lombard & Venetian gnaro. And here it is in Polish. Will anybody claim that Godmy's recording is not closer to that than to the Southeast Asian/Amerindian sound?

More important for this discussion is that the initial g- had disappeared from ordinary speech by the classical period, as even Allen notes. You didn't say /gnaeus/ any more than you said /gnātus/ or /gnōtus/. That's not to imply that nobody did so, but the impression one may get that GNAEVS was spelt with a G more often than the other two words, therefore it was pronounced with it more often, is a false one. It was normally spelt CN, and when one wanted to look archaic while spelling it out, they'd spell CNAEVS with a C. I don't imagine you'd recommend a spelling pronunciation for that.
What I said could be (in some instances) taken as a diphthong as there is often an initial release and in the same time it is what Sidney suggest might have been happening [ŋn].
What you're saying makes no sense to me. Vowels have no release, consonants do. What can be initial about a dipthong I don't have a clue either. You seem to be talking about the vowels -ae- and the consonants gn- in the same sentence at the same time and I can't tell where one starts and the other begins. Any way, your pronunciation very obviously has no dipthong, the first syllable is light/open and ends in /a/.
 
Last edited:
Anbrutal Russicus dixit:
Your language has no /ŋ/ as a phoneme so you cannot claim any native language advantage in producing it.
I think you're maybe basing that assumption on your native language? It may not be /ŋ/ in R банка, but it is in C banka, at least according to my comparative Slavic grammar & the wiktionary entry, attested in the recording.

I will agree, though, that his pronunciation doesn't accord with A&G.
 
Someone more expert in phonetics than I am will probably answer your question, but my, did you make me chuckle.

What I do know is that a word like gnatus became natus, so it would seem to suggest that that gn sound in the beginning of a word could be problematic for native speakers as well. Now, was it always resolved to a simple n even where it went on being spelled as gn, or was a full g restored? Or did something else happen? I'm not sure.
Probably problematic less in that it might have been difficult to pronounce (I myself find it quite easy and ordinary to pronounce 'gn' as in 'ignite', regardless of its position) and more in that it might not have been distinctly heard, and thus not properly reproduced by younger generations who might have been largely illiterate.
 

Anbrutal Russicus

Active Member
I think you're maybe basing that assumption on your native language? It may not be /ŋ/ in R банка, but it is in C banka, at least according to my comparative Slavic grammar & the wiktionary entry, attested in the recording.
The key words there are "as a phoneme" as opposed to as a(n) (allo-)phone, and the key characters are the slashes // as opposed to square brackets []. Czech does have [ŋ] as an allophone of the phoneme /n/ before velars, and Russian indeed doesn't, where my explanation is that it wouldn't fit into the soft/hard opposition that Russian has, and also by the fact that Russian had historically inherited no /nK/ clusters (all were nasalised) and as a result no velar assimilation rule. Altho the latter is true for all the Slavic languages, so I guess Russian is special in not having developed/borrowed it despite being surrounded by languages that have it.

Anyhow, you wouldn't manage to make an average Czech native to pronounce a [ŋ] at the beginning of a word for the reason stated above.
 
Ah, правильно, правильно. My mistake.
 

Anbrutal Russicus

Active Member
Probably problematic less in that it might have been difficult to pronounce (I myself find it quite easy and ordinary to pronounce 'gn' as in 'ignite', regardless of its position) and more in that it might not have been distinctly heard, and thus not properly reproduced by younger generations who might have been largely illiterate.
There's no conceptual distinction between these two scenarios. You don't hear sounds that you haven't acquired as phonemes, and you find them difficult to pronounce. As a rule, no native finds the sounds of their own speech difficult to hear or to pronounce, or if they do, this is their way of saying "sometimes we don't actually say these sounds even though we know they're there", a description of a phonetic rule. If a child doesn't reproduce some feature of their parents' speech, this means they don't hear it, and this is an instance of language change. Literacy doesn't really come into play until well after the primary language acquisition stage (somewhere till 3 years of age); even modern education is incapable of reverting language change, particularly when it comes to phonology, by for example teaching all the English-speaking children to pronounce the "k" in "knit". If you try, your best success will be to get half the students to force a "kuh-nit" out of themselves when reading under the threat of bad grades. Language acquisition defies intrusive correction after-the-fact.
 
Last edited:
Top