Most of the time, Latin doesn't allow two instances of the same vowel next to each other: forms like *mee (from meus) are replaced with alternatives like .

However, in I-stem second nouns, the combination can occur (for example, filiī "sons"). Do we know how this was pronounced? Was there a consonant separating the two vowels, or did they merge into an extra-long i, or did the quality difference between the two vowels keep them apart?

I'm most interested in Classical pronunciation, but anything pre-Ecclesiastic is relevant!

(Related: this question covers uu, and the accepted answer mentions that a vowel before another vowel seems to have become long-like in quality, but doesn't cover groups like where the second vowel is long by nature.)

  • I bet your answer is in verse... metrics!
    – Rafael
    Commented Aug 8, 2018 at 18:16

1 Answer 1


I don't think we can know for certain, but the fact that the spelling "filii" occurred makes it likely that the word was perceived as having three syllables, and was so treated in poetry. There are various possibilities of how it was actually pronounced:

1) A single extra-long vowel, [i::], in which case the spelling represents perception only.

2) It's possible there may have occurred some subphonemic glide between the vowels, such as [iji:], or (less likely in my opinion) a subphonemic glottal stop, [iʔi:].

3) It's possible there may have been some tonal difference between the two syllables.

4) There's also the fact that short /i/ is generally thought to have had a relatively open pronunciation, i.e., [ɪ], so the pronunciation [ɪi:] may have been the case. (Of course, [ɪ] may have been used in possibilities (2) and (3) above, too.)

I don't think it's possible to know which of these was the case, lacking the evidence of a trained phonetician in Roman times or a time machine. But, I think these are all possibilities.

  • 1
    There seems to be less certainty about the pronunciation of short /i/ as /ɪ/ when it comes before another vowel, though. There are words in some Romance languages that seem to show reflexes of Latin [ɪ] > [e] (e.g. French "voie", from Latin "via"), but if I remember correctly, there are also words that seem to show the reflex of Latin [i] (I know of an old article that talks about this, "Hiatus and Vocalic Quality in Classical and Vulgar Latin" by G. K. Meadows (1946), but I forget any specific examples)
    – Asteroides
    Commented Aug 8, 2018 at 22:02
  • Allen also talks about this issue; I give the relevant quote in my answer here: latin.stackexchange.com/a/1072/9
    – Asteroides
    Commented Aug 8, 2018 at 22:04
  • 1
    So, the take-home message (for everyone) is we don't know for sure - and most likely never will - and we shouldn't really worry about it.
    – Alex B.
    Commented Aug 13, 2018 at 2:20

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.