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

Hi!

Dirk Elzinga writes:
>...
> I think that a strict segregation of morphology and phonology is
> probably a mistake in lg creation, whatever your analytical
> predilections are. For me, most of the interest in morphology is in
> its interaction with phonological forms. ...
Very true.  I often failed when I tried to use a phonology 'module'
for a conlang.  As most of my previous conlangs were engelangs, I
tried to separate steps, for the sake of getting reusable 'modules',
one of them being a phonology module.  However, when designing a new
language, it turned out that aspects from the phonology would
influence morphology design quite automatically.  Actually, I could
not simply replace any phonology easily because it was always tight to
much to the grammar.

E.g. in Qþyn|gài, the consonants are used to represent semantic
categories, the atomic part of the lexicon from which roots are
derived (root = category + another consonant, stem = root + vowel).
If I replaced the phonology, it would have to adhere to certain
constraints, e.g. it must have exactly 40 consonants and it must have
CVCV stems.  At least if I don't want to change the grammar.  (Of
course, such constraints are a bit different from interactions you
might get in naturalistic conlangs.)

And in S11 (wordless lang), word and morpheme boundaries are shown by
consonant mutation, which would need a total rework when replacing the
phonology.  (In this case, it may even happen that I replace the
phonology, but it will require a lot of reworking and may even result
in a new conlang.)

And there's also interaction between morphology and syntax and also
morphosyntax or phonology and lexicon, which makes conlanging so much
fun. :-)

> > P.S. more details of Miapimoquitch will be welcome.
>
> I'm hoping to finish the morphology section of the grammar by the end
> of the month; I'll be sure to let the list know when it's done.
Oh yes, please!

**Henrik