Re: ¿Naro cel ei nau cepoa sia? ['naru,gil enQ,gibua'Za]

At 12:30 AM +0100 1/16/03, Christian Thalmann wrote:
>--- In conlang@yahoogroups.com, Dirk Elzinga <Dirk_Elzinga@B...> wrote:
>> Christian:
>>
>> I like your efforts to include "non-trivial" phonology in your
>conlang; it has a vaguely Numic feel to it.
>
>Thanks.
>
>
>
>> 1.  The transcription scheme you provide seems needlessly
>> complex; I would suggest that you stick to X-SAMPA (or your
>> favorite ASCII-IPA scheme) for a discussion of the phonology.
>
>Initially, I was going to write down every word in native,
>phonemic *and* phonetic transcription.  Then I noticed that
>the phonemic version was obsolete, since the native version
>*is* phonemic.
>
>Discussing the phonetic realizations of the phonemes (the
>large table) in an X-SAMPA notation alone would definitely
>be possible, but then I'd have to add a chapter to match up
>the native script with the X-SAMPA phonemic notation, which
>IMHO is a bit clumsy, seeing how phonemic the native script
>is.  The main factor of non-triviality in the native trans-
>cription is the use of |e o| for /i u/ and |i u| for /j w/.
Those mappings are not the only ones that are confusing. You haven't yet given a
compelling reason for transcribing [E] as /ai/, [Q] as /au/, [e] as /ei/, [y]
as both /eu/ and /ui/, and [o] as /ou/. In each case, it seems that the
phonetic quality represented by the X-SAMPA symbol would do fine as a
transcription. Are there alternations which indicate that the glide is in fact
present? If there are, you don't discuss them -- you do say that if a glide
follows a vowel within its syllable, it merges with the vowel in the ways
you've indicated. But there is no evidence that such a glide is present
underlyingly. As it stands, it reminds me of Particle Phonology or Dependency
Phonology, both of which use the symbols |i,u,a| as atoms to build up more
complex vowel segments. But these are *theories* of phonological structure, and
as such aren't appropriate this early in a *descriptive* enterprise.

Adding a chapter (or a section of a chapter) to explain the mapping of a phonemic
transcription (in X-SAMPA) to the transliteration you're using wouldn't really
be clumsy; plenty of published reference grammars do something similar. It
would also give you opportunity to talk about the native script, which would
also be interesting.

>As for the sandhi, those rules are already described in terms
>of phonemes rather than letters: "When one word ends on a
>vowel and the following one begins with a cluster (-V C-) or
>vice versa (-C V-), then...".
Well, "V" and "C" are not really phonemes, but I take your point.

> > 2.  Be very clear and specific about the kinds of
> > alternations that segments enter into. I found reference to
> > Palatalization before you explained what triggers it or what
> > the effects were.
>
>Noted.  I'll change that.
>
>
>
> > Also, in your chart of clusters, you organize them as [initial
>> - medial - final] without explaining the significance of these
>> terms. It may be possible for the reader to make a guess at what
>> you mean, but the reader's guess could be wrong.
>
>Yes, it seems I'm assuming a certain level of linguistic
>knowledge from the reader...  but who else would put up with
>such a language?   ;-)
>
>I'll make that clearer.  What I meant was WORD-initial (-medial,
>final), with phrasal sandhi sometimes fusing two words, causing
>word-initial and word-final clusters to become medial.
When you say "word-medial", do you mean between vowels, or simply not at the
beginning or end of the word? The alternations you show are typical of
intervocalic and postnasal position, but there are other "word-medial"
environments besides. If the |t-d-t| triplet is to mean anything, you need to
be very specific about when /t/ is realized as [d]. The chart should only come
at the end of your discussion of consonantal alternations in the language,
where it can be an elegant means to summarize those alternations.

> > Only later in
>> the section on phrasal sandhi do you explain more (but still not
>> as much as I'd like to see!).
>
>The definition of a phrase will come in the chapter about words.
[snip]

I wasn't referring so much to the absence of a definition of phrase, as I was to
the paucity of examples of phrasal phonology. It'll come in time, I'm sure.

> >  3.  Nasalization: are there examples of tautomorphemic clusters
>> of N + C that are not pronounced at the same place of
>> articulation? If not, then Nasalization can't be considered a
>> strictly phrasal phenomenon, but the extention to the phrase of a
>> generalization which operates between any two syllables. The
>> other sandhi alternations seem to have this character as well.
>
>This is a bit above my linguistic understanding.  As far as I can
>tell, you're wondering whether the assimilation of nasals is really
>a phrasal rather than an ubiquitous phenomenon.  The answer is yes,
>it only happens within phrases.  Example:
>
>|Amma ien golom rae|  [,amma'jen ,Nulum'brai]
>(Mother wash) (nose girl)
>"The mother washes the girl's nose"
>
>The |ien| and |golom| are in different phrases, thus the nasals
>don't assimilate.
No, that's not exactly what I had in mind, though the example is interesting in
that it clearly demonstrates phrasal boundaries. What I was wondering was
whether there are examples of single words which have nasal-consonant clusters
which aren't pronounced at the same place of articulation.

Dirk
--
Dirk Elzinga                                               Dirk_Elzinga@byu.edu

"It is important not to let one's aesthetics interfere with the appreciation of
fact." - Stephen Anderson