Re: Euphonic phonology (Was: 'Nor' in the World's Languages)

On 8/8/06, Henrik Theiling <theiling@...> wrote:
> [snip]
>
> Anyway, do others also have such a hard time finding personally
> pleasing phonologies?  I find it awefully difficult.
Reading over Philip's reply to Henrik's question, I realized that
there are a number of ways to understand the word "phonology". As a
practicing linguist (it's what puts bread on the table), I tend to
think of the subject in different terms than most people around here
seem to do. So when I say that I find it difficult to create a
pleasing phonology, I do not mean that I find it difficult to come up
with an inventory of sounds that I like, or even to design suitable
rules for the realization of these sounds in context. Rather, the
rules themselves need to fit together in a consistent and pleasing way
that isn't always obvious until you try them out together.

Let me give you an example of what I mean. In Miapimoquitch, I have a
stress rule which stresses the first syllable of every word and every
odd numbered mora thereafter. (Proclitics are excluded from the stress
rule.) This means that a syllable which follows a heavy syllable is
also stressed, yielding a sequence of two stressed syllables in a row.
So in the word [táppùni] 'a rabbit', the first syllable is stressed by
the initial stress placement rule. Since the first syllable is heavy
and contains two moras, the third mora falls in the second syllable,
so it is stressed by the "law of alternating stresses".

In Miapimoquitch there is also a rule of intervocalic lenition. A
voiceless stop which immediately precedes a stressless vowel is
realized as a voiced fricative; for example, the word [píðɨ] 'see' is
underlyingly /pitɨ/. Since the second vowel is stressless, the /t/
preceding it is lenited.

These two rules interact, since lenition is dependent on stress
placement. Over the years I've tinkered with both rules. Lenition used
to apply to any intervocalic voiceless stop regardless of stress, and
stress placement used to avoid stress clashes. I liked the idea of
making lenition dependent on stress, but finally coming to grips with
where stress would fall has taken quite a while. Now Miapimoquitch
sounds "right", at least with respect to the interaction of stress and
lenition.

This is the kind of thing I thought of when Henrik posed his question;
I've been wrestling with this (and other phonological issues) for
quite a while now, and it is this part of the game which I find
particularly challenging (and rewarding).

Dirk

Replies