Re: CHAT: minimum phonemes, was Re: vrindo

Proto-Indo-European at some stage seems to have had a very restrictive
vowel inventory -- perhaps as small as /e o/ (in which case vowel
contrast would have been primarily a matter of roundness?  PIE did
have rounded velar stops as well, of several varieties).  But then,
consonants (r, l, m, n, y, w, maybe laryngeals?) could step in and
become syllabic at times, which means the inventory was really much
larger for practical purposes.  It's just that the syllabized
consonants alternated morphologically (or according to suprasegmental
features with accent?) with ordinary consonants, while /e/ and /o/ did
not.

Or something like that.  Try getting a straight answer out of a
Proto-indo-europeanist!


+ Ed Heil ---------------------- edheil@postmark.net +
|    "What matter that you understood no word!       |
|    Doubtless I spoke or sang what I had heard      |
|           In broken sentences."  --Yeats           |
+----------------------------------------------------+

Kristian Jensen wrote:

> J. Barefoot wrote:
> >That reminds me... Does anyone know what the lower bound for number of
> >phonemes is for a natural language?
>
> Nik Taylor already mentioned Rotokas has the lowest number known.
>
> Let me just add a bit of theory.
>
> For vowels, it is said that the lowest contrast of major vowel
> features will always be in vowel height. Even if a language was
> only two vowels, these vowels are more likely to be /a/ and /@/
> rather than /i/ and /u/. Of course, a language with only one
> vowel, contrasts in major vowel features are ruled out
> altogether.
>
> As for consonants, a language will always have obstruents. Look
> at Rotokas for instance, from what Nik has posted, all but one
> are obstruents. Also, the contrastive POAs are likely to be
> labial vs coronal vs dorsal rather than anything else. Again,
> look at Rotokas.
>
> -kristian- 8)
>