Re: Yes, another sketch for a new conlang! [very very long!]

Christophe:

I already love it: weird (mor)phonology, and classes (already tasty,
whatever the updates), and writing system...

I'd only suggest one tip to make the explanation more transparent, and the
system easier to evaluate.

I'm a bit confused by deep and surface phonemes being lumped together in
your description.

Do you have any objections against arranging some material in a more
generativist manner? I mean, first deep-level phonemes, then morphonological
operations (sandhi or whatever name you prefer), then surface-level
phonology as a system.

This way, everything can be checked separately.

For example, I'd expect that gaps and bans in the surface phonology would
form a logical and consequential system. For if they don't, there'll be a
tendency to fill in the gaps and violate the bans (in borrowings,
onomatopoeic words, new formations on analogy, etc). So surface forms will
appear in the language that cannot be generated from any deep sequence,
and you'll have to introduce special rules for re-analyzing such forms in
order to enable normal derivation from them.

The rules might be presented as a set of expressions like (A+)BC(+D) -> EF,
put in certain order. In a natlang the order would reflect the language's
history (but usually not in all details). It seems that the first rule in
your system would be about the syllabic/non-syllabic value of resonants - a
natural transition from deep phonemes' combinatorics to the other
operations.

Deep phonology is the opposite of the surface one. The set of deep phonemes
can be as asymmetrical as you wish, and their permitted sequences needn't
to be pronounceable at all. If later you choose to set up a historical
perspective for your language, you can always declare that some sounds
merged or dropped out without any consequences for the synchronic
description.

I hope all this doesn't sound too confused...

Happy conlanging,

Vasiliy Chernov
AKA Basilius
mailto:bc_@mail.ru
http://dmoz.org/profiles/basilius.html