Re: a grammar sketch...

Me (Daniel):

> > > I've got two words for ya: pivotless languages! Langs with no
> > > sentence alignment or grammatical relations whatsoever. There
> > > is just no way of telling if in the sentence 'Dog bites man'
> > > it is the man or the dog that is bitten.
Jörg:

> > So how does such a language *work* then?  Granted, what is subject and
> > what is object can often be reconstructed from the context, but there
> > are always cases of doubt, and the matter is utterly clumsy anyway.
czHANg:

> Sounds like either Lisu or a "Language Poetry" experiment ;)
Hehe. :) Let me quote Kibrik, again. He hasn't found any examples of
languages without syntactic relations, but he says:

"Flow-oriented languages should be similar to pivotless languages in
their neutrality to semantic relations. Their distinctive feature is
the prominence of flow concepts, one or more of which are the pivots
of NP marking -- which is not quite in accordance with the Hierarchy
of Dimensions, for which reason this type ought to be rare.
   Lisu, a Tibeto-Burman language from Thailand, claimed by Li &
Thompson (1976) for their topic-prominent type (based on Hope 1974),
is possibly flow-oriented. Sentences such as (12a-b) are equally
ambiguous as to which NPs are Agent and Patient, pointing to the role
neutrality of Lisu; the difference is merely on of topic, as marked by
clause-initial position and the topic particle _nya_.

  (12) a. lathyu  nya    ána  khùa
          people  TOPIC  dog  bite
          'People (topic) they bite dogs / dogs bite them.'

       b. ána  nya    lathyu  khùa
          dog  TOPIC  people  bite
          'Dogs (topic) they bite people / people bite them.'"

I also have some info on Awa Pit, which is a deixis-oriented language.
This is really cool too. If anyone's interested, I can post the info
I have on it.

> Another idea: an "anti-active" language.  Active intransitive verbs
> (such as "to laugh") treat their subjects like direct objects,
> while non-active verbs (e.g. "to fall") like transitive subjects:
>
> child-I stone-II throw
>
> child-II laugh
>
> stone-I fall
>
> (I and II are some kinds of cases, for which I haven't invented names
> yet; or use head marking instead.)
Hmm. I and II are fine names I think. As for the anti-active language,
it doesn't really make any sense semantically, which is the way I
define active languages. You might come up with a syntactic explanation,
but you're gonna have to ask Marcus about that. ;)

If you can find a way to explain why you treat the 'subject' of stative/
non-controlled/non-volitional/whatever predicates like transitive agents,
then it would be really cool. Perhaps it's possible to find a way to
make it fit semantics too.

Daniel