Corsican vowels

Question

In Corsican, some vowels are nasalized before a nasal consonant in the same syllable. What do these vowels have in common? Here are some examples:

'prin.tʃi.pe = prince

'fun.gu = mushroom

'ãn.ku = also

'kõn.tu = settlement

'põn.te = bridge'

fun.da.'mẽn.tu = foundation

I think that one similarity is the short pronounciation of the vowels due to nasalizing both vowel and consonant. But thats the only thing I can deduce and it is not a very scientifical approach or answer. Any hints would be appreciated.

Answer 1

The difficulty in seeing the solution derives from the theory of classifying groups of sounds. You can see that [i u] do not become nasal, and [a e o] do, but how do you encode that in a rule. One (bad) solution is to simply list the three letters that do nasalize, and forego any generalization of what the one class of sounds have in common. So you need a theory of how to describe sets of sounds. If you are trying to describe the sounds that undergo the rule using IPA terminology, you are doomed, because there is no IPA property that combines [e o a], especially given that the vowels in the middle are transcribed with [e o] and not [ɛ ɔ]. But there are other phonological theories for describing groups of sounds. One such theory, the SPE theory, has a small number pf physical properties plus values "+" and "-", where for example "voiced" is not a separate feature from "voiceless", instead there is a general property of voicing which may be present ([+voice]) or lacking ([-voice]).

So the answer depends on the theoretical framework that you are using for describing sounds and how groups of sounds cluster together a classes in phonological rules. Most phonological theories have a way to do this, but they vary wildly in how they do it.

Answer 2

Assuming that the data is correct and there's indeed good basis to say that /i, u/ don't become nasalised, then in the IPA vowel space what distinguishes these two vowels from all the rest is that they're high/close as opposed to mid or low/open. In generative / autosegmental phonology this is encoded as the vowel feature [+high].

In particle phonology, however, there's no single feature that distinguishes them, not even the two features I and U since they're also present in the vowels /e/ (AI) and /o/ (AU).

A phonetic explanation must be that:

1. the two high vowels /i/ and /u/ are the two shortest vowel sounds while nasalisation increases vowel duration, and
2. nasalisation pushes the vowels away from the corners and towards the middle of the vowel space, while what distinguishes the vowels /i/ and /u/ is their extreme corner peripherality.

The two in combination would suffice to disfavour phonemic (perceptually salient) nasalisation of these vowels.

I put in italics terms that you may benefit from googling.