Re: On Phonological Constraints: The Long Vowel Rule

On Sunday, January 23, 2005, at 04:27 , Nicolas Walker wrote:

> Dear All,
>
> Thank you for your input. I have learnt a lesson today - to be VERY
> careful with my words around such competent linguists! :)
Don't worry - it applies to us all. I've been reprimanded before now - in
the most friendly way, of course - for being rather lax in some
terminology or expressing something vaguely. It does help one concentrate
one's thoughts   :)

[snip]
> .... the rule itself is several years old and was part of a language I
> am attempting to somewhat revive.
[snip]
> .....However, in
> answer to such questions as 'why have the rule in the first place?', well,
> aja sula 'I wished it.'
Fair enough   :)

In that case, you might want to think in terms of morae: a long vowel =
two morae; short vowel = one mora.

Usually, I know, morae are applied to the syllabic nucleus (vowel) and the
coda. But the ancient Greek pitch accent was determined solely on vowel
length or vocalic morae.

There could be a rule, deriving from the rhythm of the spoken language,
that two contiguous syllables could consist of either two or three morae
only. That would preserve your rule and give it a reason that does not
conflict with your other rule: "the (regular) stress falling on the first
long vowel of a word."

> I still maintain that the rule is viable, and moreover, imparts a tone to
> the language which reflects my aesthetic of that 'period'.
OK - well, I have suggested a reason for the rule   :)

[snip]
> Anyway, if anyone else is GENUINELY interested in learning more about the
> rule, or would like some further details, please continue to post.
OK

> Otherwise, please dont cross-examine my turn of phrase. :(
Yes, but we may ask for the odd explanation now and again  :)

One thing I would appreciate more detail is the point Roger raised. He
suggested the language had only one long vowel in any morpheme. But you
imply that words may have more than one long vowel, tho not in two
contiguous syllables. So, I would like to know:
Is it only in polymorphemic words, where we can have more than one long
vowel in the same word? Or can this feature occur in the same morpheme?
Could, for example, a morpheme be: LV + C + SV + C + LV?

Can I return to something you wrote in the first email?
"The second part of the rule therefore proposes that should the
combination LV + C + LV occur (as a result of affixation,
declension or whatever), that the first vowel be shortened, causing the
sequence SV + C + SV. To give an example, u:r (=ear) would become uri:r in
the plural genitive."

In fact the example you give is u:r + i:r --> uri:r, and shows LV + C + LV
--> SV + C + LV.

Personally, I would want to keep the root morpheme unaltered, if possible,
  and have the affix change. One solution would be to have allomorphs of
such grammatical suffixes, one which is affixed to a root morpheme whose
final syllable contains a long vowel, and another affixed to a root
morpheme whose final syllable contains a short vowel. The simplest, of
course, would be just -i:r ~ -ir.

Ray
=======================================================
http://home.freeuk.com/ray.brown
ray.brown@freeuk.com
=======================================================
"If /ni/ can change into /A/, then practically anything
can change into anything"
          Yuen Ren Chao, 'Language and Symbolic Systems"

Reply