Hi,
Where can I find about vowels becoming short in front of other sounds? It’s weird, but I can’t find what I’m looking for.
For example, take the 4th conjugation (theme + desinence):
1st. s. audi-ō
2nd. s. audī-s
3rd. s. audi-t
1st. pl. audī-mus
2nd. pl. audī-tis
3rd. pl. audi-u-nt
This /i/ is long whenever it is possible, but it isn’t in front of another vowel (-ō or -u-) and in front of that /t/ in 3rd. s. (but not in front of -tis in 2nd. pl.!).
Do you know the explanation for it and where to find about it? I’m missing something, because I haven’t found in Allen-Greenough or in a book about phonetic history!
Thank you very much!
Leonardo
Allen & Greenough sec. 10:
- Vowels are either long or short by nature, and are pronounced accordingly (§ 8 ).
a. A vowel before another vowel or h is short: as in vĭa, nĭhil.
http://www.perseus.tufts.edu/hopper/text?doc=AG+10&fromdoc=Perseus%3Atext%3A1999.04.0001
There are exceptions to this. Many Latinized Greek words are an exception (Greek has no restriction on the quantity of a vowel before another vowel); another exception is the genitive ending -ius on certain pronouns: e.g., nullius, illius, istius, where the -i- is long.
Can anyone think of other exceptions?
See also Allen & Greenough sec. 167:
- A long vowel is shortened before the personal endings -m (-r), -t, -nt (-ntur): as, ame-t (for older amē-t ), habe-t (for habē-t ), mone-nt , mone-ntur .
http://www.perseus.tufts.edu/hopper/text?doc=AG+167&fromdoc=Perseus%3Atext%3A1999.04.0001
We’ll remember the dictum Vocalis ante uocalem corripitur, which Hylander quoted in English above. Remember that in Latin an h is only graphic between vowels — it will have no phonetic reality but marking a syllable boundary (= glottal stop).
Hylander pondered exceptions to this rule. They will include cases where the preceding vowel is itself preceded by another vowel (at least when it’s an i), so rēs : *rēī > rĕī but diēs : diēī. But we also have the notable exception of fīō, fīēbam, fīam (and most of its conjugation [not fĭĕrī, fĭĕrem etc.]). Further examples of not shortening a vowel before a vowel will be Greek loan-words like Alexandrīa, Acadēmīa, and Thalēa, but loan-words are often a special case, not necessarily (at least straightaway) fully adapted to the phonology of the language they’re borrowed into.
The Latin phonology may be, as a subject, too vast to be treated in its entirety on Textkit, but we’ll all be glad to discuss it further on the OP’s cue, so do feel free to pose further questions, which we’ll answer if we can. You’ll definitely want to read the the treatment on phonology in Manu Leumann’s grammar (p. 43—254). Within that, pages 79—124 will discuss all the phenomena of vowels due to their place in a word (or sentence), including e.g. the change in quality (gen_i_tor but gen_e_trix [the latter having muta cum liquida]) and quantity (făctus < făcere but āctus < ăgere).
Fascinating! Thank you very much, Timothée and Hylander!