Re: Attributive Nominal Forms and Syntax in a lang experiment

Josh Roth wrote:

> >(I'm not sure here who posted the data...)
> > myezai 'bear'
> > myeze 'bear-attr'
> > kun 'eat'
> > ta 'him,he'
> >
> > shi ni            myeze     ta kun (ta)  (1)
> > exist this-attr   bear-attr he eat (he)
> >
> > only means:
> > 'this is the bear that he ate'
> >
> > shi ni myeze kun ta  (2)
> >
> > means both:
> > 'this is the bear that he ate'
> > 'this is the bear that ate him'
>
> How come the first sentence has only one interpretation and the second has
> two? Since the language seems to be SVO, I'd expect the first to mean
"this
> is the bear he ate" only, and the second "this is the bear that ate him"
> only (like in English).
>
That would be my feeling too-- in (1) the resumptive pronoun (the first
_ta_) would seem to indicate a change of subject, so it's
"this is the bear(x) [that] he(y) ate it(x)"

whereas the lack of the subj.pronoun in (2) ought to show no change of
subject--
"this is the bear(x) [that] (x) ate him(y)"

Kash would be similarly ambiguous--

tala oso re yan yanahan
this (is) bear REL 3s(acc-anim) 3s-eat--

could have either reading since yan and ya- could refer to either party.  So
by convention, when both subj. and obj. are animate, the use of the subject
pfx. indicates change of subj. in the rel.clause, and the above sentence
means only "This is the bear that he ate"; whereas--

tala oso re yan nahan = this is the bear that ate him.

With a neuter object, the obj.pronoun is not necessary and the use of the
subj.pfx. is optional though preferred (since you're still introducing a new
subject)--

talu yemo re yanahan
this (is) fruit REL 3s-eat
this is the fruit (that) he ate

(for hypercorrect "talu yemo re yu yanahan"
this (is) fruit REL it/neut-acc 3s-ate)

Oso, of course, is < Spanish, since there are no bears on Cindu.  There
might well be a bear-like animal, but it will probably be called _pruwañ_
['pruaN] or _paroñ_ ['paroN], stolen from Indonesian beruang or Javanese
barong.