[Lingtyp] query: basic grammatical ground plans

[David Gil]

Dear all,

I am looking for an example of a language that has a construction that 
meets the following 4 conditions:

(1) The construction is of the form N A/V, i.e. a noun in construction 
with either an adjective or a verb — either linear order will do.

(2) The construction can be understood either predicatively, denoting a 
property or activity ("N is A/V"), or attributively, denoting a thing 
("N that's A/V").

(3) The construction is productive, i.e. it occurs across the board, and 
is not restricted to a particular configuration of grammatical features.

(4) The construction imposes restrictions (e.g. through case or 
agreement markers) on the semantic relationship between the two terms 
(e.g. specific thematic roles) — all while maintaining 
predicative/attributive vagueness or ambiguity as per (2).

Discussion:

Some European and other languages have constructions that meet 
conditions (1), (2), and (4), but are limited to particular 
configurations of features and thus fail condition (3).For example, in 
Hebrew, "ʕorvim šħorim" ('crow.PLM black.PLM'), with number-and-gender 
agreement, can be understood either attributively or predicatively, but 
the predicative interpretation is limited to a generic reading, and even 
there it is not the most natural way of conveying the meaning in question.

Colloquial Indonesian and my usual "ayam makan" ('chicken eat') example 
meets conditions (1), (2) and (3) but fails (4) — it is semantically 
vague with regard to most commonplace categories, including even 
thematic roles. Various optional grammatical morphemes can be added to 
"ayam makan" to narrow the range of semantic relationships between the 
two terms, but these tend also to eliminate the predicative/attributive 
vagueness or ambiguity in favour of either predicative or attributive 
readings — conditions (2) and (4) seem to be working against each other.

I have a vague recollection that I once saw an example from some 
Australian language that fit the bill, but I'll take examples from 
anywhere.Maybe if I stick my neck out and assert that no language can 
have a construction that satisfies the above 4 conditions then somebody 
will come up with the requisite counterexample.

Thanks,

David

-- 
David Gil

Department of Linguistic and Cultural Evolution
Max Planck Institute for the Science of Human History
Kahlaische Strasse 10, 07745 Jena, Germany

Email: gil at shh.mpg.de
Office Phone (Germany): +49-3641686834
Mobile Phone (Indonesia): +62-81281162816

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lingtyp/attachments/20190327/9660d46a/attachment.htm>