Re: Hadwan stress system renewed

On Sat, 28 Jul 2001, Muke Tever wrote:

> Well, the talk about stress inspired me to study, and I found an actually
> consistent way of accenting Hadwan words.  (Before I was going by ear and
> partly influenced by the vocalization of the proto-root.  I only had to
> change a couple of words to fit.)
I love it when this kind of thing happens -- regularity
springing up from intuition.

> Stress on verbs is morphological: in the nonpast the first syllable of the
> ending is stressed and in the past tense the last syllable of the stem is
> stressed.   So:
>
>         bá-mai     /BQ:"mAI/   "I say/am saying"
>         rour-eñ    /"rUwrEN/   "I made room for smth."
>         gigwi-mizh /gI"gwImIZ/ "We have lived"
Sounds reasonable. I also like morphologically determined stress
and accent systems; PIE, Salish ...

> Stress on everything else can fall only on the first three syllables of a
> word.  The heaviest (and, in a tie, the rightmost) of these syllables
> receives the stress.  So:
I don't think I've ever seen a three syllable window at the left
edge of a word, but I don't know why it shouldn't be possible.

> Coda sonorants don't count towards syllable weight...
Now this is weird. If there were a distinction among the
moraicity (weight-bearing-ness) of consonants, you'd expect
sonorants to be moraic but obstruents to be non-moraic. But hey;
if it works for you, go with it.

> Some words still have idiosyncratic stress, however.  This usually is
> because Hadwan dropped initial unstressed /e/ and /i/ from its parent but
> retained the original stress:
>
>         jormos  /jUr"mUs/  "collection"
>
> looks normal but
>
>         jormosho /jUr"mUSU/ "(gen.)"
>
> appears where /jUrmU"SU/ would be expected, because the stress is
> "remembered" from the earlier *iGor"moSo.  (You might say it's also
> 'remembered' on regular <jormos>, but in that case the outcome would be the
> same either way anyway, so it doesn't matter.)
>
> [A northern dialect developed a consistent nonverb stress on the second
> syllable.  This dialect still is affected by the previous rule, and accents
> the first syllable of words like <jormos>.]
Lexical exceptions are always nice to have; IMO, they lend an
air of "authenticity" to an artlang. (They're just annoying in
auxlangs; cf Interlingua. Which reminds me. Last year I picked
up the Interlingua Grammar and Dictionary and tried working out
an analysis of Interlingua stress. Turns out to be every bit as
complex as Spanish, and in pretty much the same ways. I'll have
to polish up my notes and post it here sometime.)

> I don't know what to do with borrowed words, however.  Should a word like
> "kétos" be stressed normally (on the final syllable) or as in the original
> Greek (on the primary syllable)?  Might it depend, maybe, on how 'nativized'
> the word is?  Hmm..
Yes. One of my very favorite examples in historical linguistics
is the multiple borrowing into English of the French word
'gentil'. From it we get gentle, gentile, genteel, and jaunty
(in that order). This illustrates the nativization process
pretty well, I think.

> I'm still not sure about (or whether) secondary stress, either.
>
> Any comments?
If you have morphologically determined main stress, but
phonologically determined secondary (perhaps via mora counting),
that would be fun ...

> How much of this is actually impossible?
Your intuition has already led you to some regularity. Trust it
some more.

Dirk

--
Dirk Elzinga                          dirk.elzinga@m.cc.utah.edu

"The strong craving for a simple formula
has been the undoing of linguists."               - Edward Sapir

Replies