On Tue, Dec 09, 2003 at 08:59:02PM -0600, Nik Taylor wrote:

> Amanda Babcock wrote:
> > Here is the phonology I settled on (which actually ends up sounding
> > pretty Slavic, which is even cooler; thanks, Nik, for the i->ja and
> > u->wa rules!  I never would have thought of that!)
>
> :-)  I just borrowed it from one of the few settled sound changes
> between Uatakassi and its descendant Ivetsian, altho in that case it was
> only *long* /i/ and /u/ that underwent the change.  I originally got the
> idea by simply reversing what happened in English's GVS, namely /i:/ ->
> /aj/ and /u:/ -> /aw/

Oh, cool!  Well, it's very neat, although if the language develops
further all those w's and j's are going to have to do something to
the preceding consonants and then fade away... they're quite a mouthful.
(But I have no plans to explore the future of this language's development;
Witicku and Vjatjackwa are quite enough for now :)

> >         . Every third vowel dropped (except those following h) after
> >           causing certain mutations upon the previous consonant.  If
> >           a vowel divisible by three followed h, h dropped.
>
> The "every third" rule sounds a bit odd

It's a consequence of the dactylic stressed pattern.  It's the minimally
stressed vowels which drop.

The whole thing started because of a bug I got in the brain months ago
when reading on the list about where Slavic languages' phonotactics come
from (so it's come full circle!).  That conversation got me started thinking
about consonant clusters formed by vowel deletion, and I thought of a
word "Apaterakasuto -> Apatrakasto -> Aptraksto", which stuck in my brain.
(Ap-anything turns out not to be initially permissible in Witicku, although
I'm exploring deleting initial h's that are not hj's or hw's in Vjatjackwa -
either just the unstressed ones or all initial h's.)

I realized after the fact that the more likely vowels to drop would be
the ones right *after* the stressed vowels rather than the ones right
before them, but too late now.

> > "Mjatwam" as a title with which to address married women.
>
> Derived from a contraction of _sjakwaPmjatwam_, "my mother"?

Yep :)  Or possibly of one of the other two-syllable posessive prefixes,
but that's the one that makes the most sense :)

> Interesting source!  Classical Uatakassi used _tinani_ "mother" as a
> title of respect for a woman who was a mother (in later periods it
> simply became a title for any woman above a certain age)

I'm considering making the most powerful people in a clan/tribe be
the Widows.  Apparently my people have a high death rate for prime-aged
men...  (I just realized that this could be (taken as) an effect rather
than just a cause :)

> >                 Proto-Witicku     Wítickú         Vjatjackwa
> >
> > Cat             mimi              mími            mjamja
> > a cat           mamimi            mámiw           *mOmju
> > the cat         kamimi            kámiw           *kOmju
>
> Did Vjatjackwa lose articles completely, or did it develop new articles
> to replace the older ones?  Or perhaps analogizing those starred forms
> to mOmjamja and kOmjamja?

Well, I actually haven't worked that out yet.  There's a lot of handwaving
between the grammar of Witicku and that of Vjatjackwa.  I suspect that in
the real world, when synthetic grammar becomes isolating, it's because of
sound changes that completely erase the grammatical marking, rather than
just obscuring it (at some point the grammar would fight back and
re-analogize to prevent the sound changes from making off with all meaning!),
but in this case I'm not wearing down the sounds quite that much, just to
the point that the words become incomprehensible and people give up :)
So actually, Vjatjackwa probably looks more like the result of a creole
of Witicku, although I'm not too interested in having that much social
upheaval in its history.  Maybe the younger generation just got really
lazy and started stringing together words any which way... young people
today with their crazy talk, I swear... :)

I'll decide about definiteness later.  They'll probably lose it and then
reinvent it the usual way.  I did decide the other day that they keep a
simple one-syllable pluralization prefix, with the result that there is
an alternation between the singular and plural forms of words (and this
is also where I got the idea for h's dropping):

        Proto-W         Witicku         Vjatjackwa
egg     'tamako*        'tamag          'tOmOg
eggs**  'hetama'ko      'hetam'ko       i'tOmku

* Points for recognizing this.
** Ok, so I'm using the vocative/name form to derive the Vjatjackwa
words, which is all well and good and in accordance with tradition,
except... would the Witicku ever have *put* a plural prefix on a
hyper-definite/abstract word?  Erm...

(Ooh, I know!  The language must have lost the definite/indefinite/name
distinction before they lost the stress-alternation/plural-prefix
mechanism.  Whew!)

> Very interesting product!

Thanks! :)

I promised everybody something on verbs the other day, but I don't have
it yet.  Funny how times of stress (final exams, etc) are when I have the
most inspiration coupled with the least time...

Anyway, the verb system of Witicku will feature:

        . inanimate objects routinely incorporated into verbs
        . subjects not part of the sentence, must have been introduced
          in previous dialog or via topic marker and only appear on the
          verb as pronouns
        . verb agreement affixes use an animacy/proximacy hierarchy,
          an inverse marker, and the fact that all verbs are either
          transitive or intransitive (never ditransitive), with
          transitives being constructed from intransitives via applicatives
        . 3p -> 3p agreement where the object cannot be incorporated
          relies on animacy/proximacy hierarchy to disambiguate
          . animacy trumps proximacy
          . proximacy works near > far > yonder
          . if the arguments to the verb are of equal animacy *and* in
            the same location, the more topical one is the agent

And I'm going to do animacy the Cree way (as seen in _Describing
Morphosyntax_), with 2p > 1p > 3p:proximate > 3p:obviate; though they
also mark the obviative word, which I'm not currently thinking of
doing.  The animacy hierarchy was important because I've decided that
if there is no agreement marker for the object, and the verb is
transitive, it defaults to being one lower than the subject.  So with
the 2nd person marker, a transitive verb means "You X me", perfect for
command form, especially if the verb is made with a dative applicative.
With the 1st person marker, a transitive verb means "I X he/she/they/it",
perfect for recounting one's exploits.  With a 3rd person marker, a
transitive verb means "They X it/them-yonder", also useful.

Lots of other optional bound morphemes will exist for clarifying
things, including one to say "somebody in this sentence is female",
reflexive, reciprocal, and one to make third persons act on
equally animate/proximate third persons instead of distant or
inanimate ones.

As for topicality and proximacy: a new actor can be introduced either
in relation to a current topic, in which case it is of lower topicality
than the current topic, or it can be put on the top of the topicality
stack via some sort of topic-introduction construct.  I've been calling
this the "Now this guy" construct in my head, because it would work like
this:

"Now this guy Bill, see, now this guy John, he hits him" means "John
hits Bill".

So there you go!  Ok, that wasn't as clearly laid out as I'd have liked,
but at least it's written down now.  Now I just need to settle on actual
morphemes and start building words... I've been through three other verb
paradigms (not so animacy-driven) that I didn't like.  It's time to just
grab some syllables and run with them.

Amanda