--- On Wed, 12/28/11, neo gu <[log in to unmask]> wrote:

> > The issue in Talarian is that not all words that have
> inanimate
> > declension are actually inanimate by nature. Some are
> inanimate
> > only by accident of termination. Thus all the family
> nouns in -r, like
> > patar and matar, agent nouns in -r, etc. are all
> inanimate. Gosh --
> > how can an *agent* noun be *inanimate*!!??
> 
> How did -r nouns become inanimate in the first place, BTW?

A very ancient phenomenon indeed! The parent language had two classes,
animate and inanimate; Talarian has kept this distinction, eschewing
the proliferation of genders that the other Aryan languages have gone
in for. Back in the day, there were several subclasses of each. Some
animates terminated in -us or -as (what would eventually become fem. &
masc. in other languages), some in -@r (that would eventually become
various -r words, like "father", in other languages), some in -es; the
inanimates tended to terminate in -am, -um, -ar/-an and -em. Out of
all of these, -us, -am and -ar survived the "Great Vowel Collapse" of
some 18 centuries or so ago. This phenomenon led almost all vowels to
become either A or U. Long and short A remained; long O became long A;
short O became a kind of schwa; long and short E became long and short A
respectively; long U remained; short U also became a kind of schwa; 
long and short I remained. -us was already the declension of old
animate words, while -ar was already the declension of many key inanimates.
It scooped up some -am inanimates, while some seem to have become -as
animates. Eventually -us and -as become conflated into one declension,
and anything in -r become conflated into another.

Culturally speaking, the idea of animate beings and inanimate entities is
already very strongly ingrained, and as this period of time progressed,
each declension became highly associated with one or the other. Of
course, even in those days, you needed some way to say "Father spanked
his recalcitrant son"! Traditional solutions tended to place the object
(the son) in the subject position with a middle verb and the agent in the
instrumental: the son him spanked with the rod of his father.

The various other solutions appear as time progressed. The morphological
shifts are the newest solutions, probably appearing no more than a
century ago or so. It is possible that Talarian is suffering from some
areal peer pressure.

> > The Talarian solution is twofold: whenever such nouns
> don't need
> > to express their inherent animacy, they are simply
> declined as
> > inanimate nouns. But, when they need to express an
> agential role,
> > a new "transanimate" declension of some kind is used.
> 
> That's sort of what I'm thinking; treat them as if animate
> only when necessary.

This is certainly an easier solution -- less convoluted than bizarre
morphology or cumbersome circumlocutions!

> > In the case of the familial nouns, modified inanimate
> case forms are
> > used. So patar becomes patrâm (actually, obviously a
> kind of
> > accusative ending is being recycled here). Patrâm
> will thus be the
> > subject of active verbs; but patar will continue to be
> the subject of
> > stative verbs (and also the object of active verbs).
> In the case of
> > inanimate proper names, they are simply
> double-declined (that is,
> > animate case forms are simply tacked on to the end of
> the normally
> > declined inanimate case forms): Patarus (nom),
> Patanosáça (gen),
> > Patanam (acc). Agent nouns have followed protocol No.
> 4 and have
> > been reanalysed as animates: halomtaras = drinker, one
> who is
> > drinking.
> >
> >> For all of the above, modifiers take animate forms
> in the
> >> ergative and dative.
> >>
> >> 4. reclassify some of the inanimate nouns as
> animate,
> >> otherwise use the passive.
> >>
> >> 5. no change, always use the passive.
> >
> > This is also an option that is met with on many
> occasions.
> >
> >> I'd like to hear opinions. For reference, the
> current
> >> grammar is at
> >>
> >> http://qiihoskeh.conlang.org/cl/o/Naisek/NTOC.htm
> >
> >Hope this helps!
> >
> >Padraic
> 
> 
> I've taken a closer look at the Naisek texts and found:
> "from" + genitive phrase + passive is used in 2 relay
> translations
> instrumental + passive is used in 1 relay translation
> choice #3 above is used in 2 relay translations
> 
> A note I've discovered in the grammar suggests I should've
> used the 
> instrumental in all of the above texts. But in any case,
> that's a lot of 
> action done by inanimate objects!

Yeah. Quite possibly one reason why so many ways of making a grammatically
inanimate noun act like an animate one is simply the large number of
them! A large number of Rumelian and Teutonic cognates that are masc. or
fem. are inanimate in Talarian; it is even far more common to borrow a
masc./fem. foreign word into the inanimate declension than it is to
borrow it into the animate. For some reason, the -ar/-an declension has
simply become dominant in the language.

Might be interesting to speculate what would happen if *all* the animates
got gobbled up, except perhaps a few stubborn hold-outs. Quite possibly
the animate conjugations would disappear. Everything would be expressed
in the medio-passive-stative continuum; or else some exceptions could
still do the old instrumental of agency trick.

Padraic