Re: Types of Possession -- Tesäfköm: A Constructed Language (S11)

Hi!

Thomas Hart Chappell <tomhchappell@...> writes:
>...
> >1) Many languages divide nouns into two types; possessible and non-
> >possessible.  ...
> > [snip]
> >
> >2) Many languages divide nouns into two types; those on the one hand that,
> >inherently, must be possessed, and those on the other hand that need not,
> >inherently, be possessed.  ...
> > [snip]
> >
> >3) Many languages ... divide types of possession into alienable ...
> > versus inalienable ...
> >
> >4) Many languages use a combination of two of the above three ideas.  It is
> > logically possible, and I think there may exist some natlangs attesting,
> > all three at once -- I don't know, does anyone else?
> >
> > [snip]
>
> Henrik Theiling's conlang Tesäfköm (S11) does embody all three of these
> ideas.
Indeed! :-) Of course, I stole the idea from various books I read.  I
found it very interesting and designed my conlang according to it.

BTW, I love to introduce irregular structures that make it near
impossible to regularly 'fix' a defective system, e.g. the inalienable
construct state in S11 is irregular so that no regular absolute state
can be derived and used.  And in S7 and also in S11, I embedded the
mood/evidence markers so deeply in the grammar structure that no-one
can easily leave them out (e.g. due to sloppiness).  In S11 they
have a secondary function of marking the beginning of a sub-clause.

Thanks for looking to deeply into my conlang (despite my often
talkative and confusing web sites) -- it is nice to share ideas! :-)

>...
> 3) Henrik gives S11 two construct-states; the alienable one and the
> inalienable one.
>...
An upcoming website update will show that I removed one of the
construct states, but still, the distinction is there: alienable
possession now requires a completely different structure, namely a
relative clause.  After the recent discussions about the various types
of possessive I decided to throw out a concept so underspecified and
ambiguous as alienable possession and further found that instead of
introducing many different possessive markers (e.g. like gzb and
Ithkuil), I'd better use what I already have: a relative clause.  The
phrases get much longer by this, but so be it. :-)  E.g.

   man's house = EARLIER:  man house.<alienable-constr-state>
                 NOW:      MOOD-owned man-owns REL house
                 OR:       house-MOOD-owned man-owns RES

   MOOD= some mood/evidence marker (also indicates start of sub-clause)
   REL=  relative clause finalising particle (externally headed RC)
   RES=  resumptive pronoun (internally headed RC)


So instead of two or three morphemes (depending whether the construct
state is a morpheme of its own), yow now need six.

Anyway, inalienable possession is still present in exactly the way as
before:

   man's arm = man arm.<inalienable-constr-state>

And actually, this is currently the only binary relation the language
has.  Everything else that is naturally binary or n-ary, n > 2, has to
be expressed with a serial verb construction.  I'm not sure yet about
how to handle clause subordination, but we'll see.

Discussions about different types of inalienable possession are very
welcome, maybe I've overseen something and it is still ambiguous?  How
do, say gzb or Ithkuil or other conlangs handle inalienable possession
and what types are possibly distinguished?

**Henrik

Reply