Re: CHAT: vowel reduction (was Re: CHAT: A flat minor)

En riponse ` "Isaac A. Penzev" <isaacp@...>:

> In addition to the previous post:
> ObConlang: Does any of your conlangs have regular vowel reduction bound
> to stress shift, like in Russian [dOm] - [dV"ma] "house" - "houses",
> ["gOr@t] - [g@rV"da] "city" - "cities"; or (Old) Hebrew [kQ:"Bo:D] -
> [k@Bo:"Do:] "glory" - "his glory", [zQ:"qe:n] - *[z@q@"ne:] > [ziq"ne:]
> "elder" - "elders of"?
>
Maggel has vowel reduction bound to stress shift, but it's not regular (indeed,
Maggel and regularity don't cohabit very well ;))) ). It doesn't happen for all
processes that create a stress shift. Unfortunately I don't have any example of
stress shift to show. I used to have (with the non-singular augment |e(i)-|
used by neuter nouns to mark non-singular numbers), but I discovered later that
prefixes didn't provoke stress shifts (bringing the only exception to the
stress-on-first syllable rule for words containing neither long vowels nor
tense consonants) but were still pronounced as if they were stressed (hence
without vowel reduction).

Actually I see that I do have one example of vowel reduction due to stress
shifts. The word for 5 is (in its declarative form) |fomd| ['fO~]. Its
distributive form (meaning "5 by 5") is formed in this case by adding the
suffix |-eo| pronounced [e:]. Since it's a long vowel, it attracts the stress,
and the previous vowel gets reduced. Hence |fomdeo| is pronounced [f@~'de:].
There are other examples of this in the formation of distributive numbers, but
this is the only one which is really regular (other distributive numbers are
more irregularly derived from the declarative forms).

But vowel reduction is not regular in Maggel. It is an existing phenomenon, but
it doesn't always affect vowels (for instance, the digraph |oa| often
represents [a]. The difference with simply |a| is that the [a] represented by
|a| will undergo vowel reduction when it falls out of the stress, while the [a]
represented by |oa| won't. But it's only on of the rare cases when such a
distinction is marked in writing :)) . Another example of that is |ae| for
unreduceable [e]).

Hehe, I've been checking my notes, and discovered quite a few mistakes, in
which I forgot the stress shift! ;))) So I have other examples of it :))) .

For instance, the uncountable "sugar" is (in the plural definite form, used
when a specific amount of sugar is counted) |tyigrim| ['Dykr\@~m]. But its
plural indefinite form (used when it's not counted), written |tyisgr|, is not
pronounced *['Dyk*r\(=)] as I first wrote it down, but [DY'k*r\=] (* marks a
tense consonant, and the [r\=] is here mandatorily syllabic since it's
stressed), because when the vowel following a tense consonant is of the same
type (short, long, diphtongue or long diphtongue) as the vowel which should get
stressed according to the stress rules, it attracts the stress. And in the case
of this word, the previous vowel gets reduced from [y] to [Y]. Another example
is the plural indefinite of |aotod| ['aTV]: bag, written |aosted|, which is not
*['aT*@] as I first wrote it (ridiculous, since tense consonants also prevent
vocalic reduction of the following vowel) but [@'T*E] (same reason as the
previous example). Prefixes don't change stress position, whatever their
pronunciation, but I never said anything about suffixes, infixes and other
changes ;))) .

Well, it seems I'm gonna have a lot to correct in my notes. Maggel is actually
so Maggelish that even I underestimate it! ;)))

Christophe.

http://rainbow.conlang.free.fr

Take your life as a movie: do not let anybody else play the leading role.