Re: IPA/SAMP question

> Date:         Wed, 14 Nov 2001 08:01:52 -0500
> From: Matthew Kehrt <matrix14@...>
>
> [Eviendadhail] does, actually, distiguish between [e], [e_o] and
> [E]. At the moment, it's sort of a problem. I'm thinking of having
> [e_o] be nasilized to help distinguish. On the other hand, there are
> very few cases in which it is necessary to tell the difference,
> except in determining aspect in a very few verbs.
Yes, if Eviendadhail is a human language, I think it's a problem to
have contrasting vowels that close together. It's not uncommon for a
natlang to have an unconditioned four way height contrast in the front
vowels, but then it's usually /i e E a/. I don't think I've heard
about a five way phonemic contrast, though there are languages where a
greater number of vowels heights occur in different contexts (short
vs. long, for instance).

I've only found one short sample of your language, but I can see both
<i> and <a> in there --- otherwise you could perhaps have lowered the
/E/ to [{] to leave room for the extra phoneme (as [E]).

But instead, there are several ways of moving the extra phoneme off
the pure unrounded front vowel column. Nasalization (as you mention
yourself), centralization, rounding, creaky or breathy voicing... you
could even use advanced/retracted tongue root, though I admit that I
don't know how it sounds myself.

Lars Mathiesen (U of Copenhagen CS Dep) <thorinn@...> (Humour NOT marked)

Replies