GROUPLANG: optional features and case

Nik Taylor wrote:

>Carlos Thompson wrote:
>> I've suggested we have a compulsory gender/class dictintion with other
>> optional "genders" or markers which would be needed for agreement
>> differenciation or extention of the meaning.  I've post sex being one of
>> them, deixis or another referencial mark as another one.  Number would be an
>> optional mark also.  Politeness would be also optional.
>
>I don't know about all these "optional" features.  I don't know of any
>language where such a thing is optional.  And politeness couldn't be
>optional.  By having no indication of politeness level, that in itself
>would say something, just as in English, not saying "please" says
>something.  And optional number just seems odd.  If we're going to have
>a number system, we oughta make it mandatory.  Please, let's not have an
>obviously artificial language.  We're not re-inventing NGL here.  (I
>intend no disrespect to NGL, my point was that NGL is obviously
>artificial, and unnatural; I'm hoping to have a natural-looking lang)
>
I agree with you in that I want a nat-looking lang too. But I have to say
that my particular way to achieve that wouldn't be avoiding features that
are not present in natlangs and/or seem odd. (Of course too much oddity
is not good either.)

My opinion on optional features:

Politeness can't be optional; either you're specifically polite, or not.
But maybe Carlos meant that we should have a politeness affix and a
(zero affix) informal marker. A polite pronoun could be the plain form
+ a politeness inflection (maybe thought as a polite word that merged
with the pronoun).

Optional number doesn't seem odd to me. If you leave the word uninflected,
then it doesn't imply anything; or else you specifically say it's singular,
or plural. I think I *have* read about optional number distinction in
several natlangs.

Gender should be mandatory, as well as tense for predicates when it's
not present (or a "narrative" tense).

>As for the cases, I really don't like the proposal.  Is it set in stone
>yet?  I'd like to propose another system.  This isn't all-or-nothing, of
>course
I don't see the current case proposal as "set in stone" exactly. But
we've been discussing it for a looong while already. I don't love it,
but I think it's fine.


>1.  Absolutive
>2.  Ergative
>    These might be in a mixed system, thus more accurately
>    absolutive-nominative
In the earliest attempts of this, I posted a mixed system...


>3.  Genetive
>4.  Dative
>5.  Instrumental
>6.  Role (I don't know the term)
>7.  Comitative (with; co-agent)
>8.  Aversive (avoidance; he avoided *the fire*, he fears *fire*)
>    Of course, this could be handled by other cases, perhaps local cases
>9.  non-Comitative (I don't remember the term; without)
>10. Associative (non-possessive genetive, e.g., *his* illness)
>11. Partative
>12. Vocative
>13. Comparative (he is older *than me*)
>14. Various local cases - I'm not particularly attached to any specific
>    system, but the more cases, the better, as far as I'm concerned.
>
>Perhaps think about multiple case-marking?
Multiple case-marking is OK to me.
"The more cases the better" is not. Case is not very comfortable
sometimes. The current proposal is just over the line dividing
"interestingly complex" from "innecessarily intricate" (at least
for me). I see you have a special feeling for cases :) (I saw your
W. pages).

As for your cases, I could live with some of them, especially
associative and instrumental, to replace attributive and causative
case in certain ambiguous situations. I prefer to leave local cases to
postpositions.


--Pablo Flores