Re: head-marking

on 8/12/03 8:19 PM, JS Bangs at jaspax@U.WASHINGTON.EDU wrote:

> Nik Taylor sikyal:
>
>> I don't know if there are any languages where the preposition would have to
>> agree with its object, but if there are, I'd suspect that they'd also be
>> ergative, and that the prepositions in question are derived from verbs.
>>
> What about Italian and French? Yes, I know that you'll tell me that those are
> just the articles phonetically mashed in with the preposition, but by modern
> times this relationship is pretty opaque and can arguably be called pure
> agreement.
I wouldn't call it agreement. If it were just a prep. with an agreement
morpheme, then the problem is that there's an article missing. For example,
'nella cucina' would consist of just a prep. and a noun. So you'd need a
rule that a noun with a prep. can't take an article (*'nella la cucina') ...
unless it's indefinite, that is ('in una cucina'). But then the prep.
doesn't agree! I think it's much easier to call it an article phonetically
mashed with a preposition. More specifically, I guess, cliticization, or
portmanteau, depending on the case.

Put another way, the preposition/article combo actually contains the
information (definiteness) that an article should give, in addition to the
semantic info of the preposition and gender/number agreement. So it seems
that the article is still there in some sense. You could still say it's
really definiteness agreement, but that doesn't happen in any other part of
Italian or French, and it would still be strange that the definite article
can't be expressed after a prep., and that neither definiteness nor n/g
agreement happens with indefinites. That explanation just makes things
unnecessarily complex.

Well you did say arguably. :-)

Josh Roth

Reply