🚀 go-pugleaf

Mark Line <markline@henson.cc.wwu.edu> wrote:
 > > Huh?  Center-embedding produces intuitively *unacceptable* sentences 
 > > pretty fast, as your example shows.  Its unacceptability is a datum.
 > > Nonetheless, it is as grammatical as all get-out.  What's been 
 > > "fabricated" here?
 >
 > How do you know it's "grammatical" if it's "unacceptable"? Unless of
 > course you create a grammar by fiat and then state, wholesale, that
 > sentences which fit that grammar are "grammatical" _by definition_,
 > whether they are "acceptable" or not. That, however, wouldn't buy you
 > anything, since you then have no way of empirically falsifying your
 > grammar.

Slow down.  First, you've presented a false dilemma.  The mere fact
that we distinguish acceptability from grammaticality doesn't mean 
that the grammar is a matter of fiat.  The relation of acceptability 
to grammaticality is itself a theoretical matter, as it ought to be.
The issue is complicated, involving among other things the whole
competence/performance distinction.
   Second, the mere fact that we distinguish acceptability from 
grammaticality doesn't mean the grammar isn't empirically falsi-
fiable.  If acceptability diverges from grammaticality and there
is no explanation, there are gonna be problems.  In the center-
embedding case, there's a natural and obvious explanation: it's
hard to process the damn things, so despite their grammaticality,
they're objectionable (in a wider sense than just *grammatically*
objectionable).
==================================================================
Michael Hand                      (409)845-5660, fax (409)845-0458
Philosophy Dept, Texas A&M Univ, College Station TX 77843-4237 USA
hand@tam2000.tamu.edu                               mhand@tamu.edu
==================================================================