And Rosta wrote:
>
> #> > > #1. Play a trick.
> #> > > #
> #> > > #2. a. A magician's trick.
> #> > > #   b. A dirty trick.
> #> > > #
> #> > > #3. a. ?? Play a magician's trick. magic
> #> > > #   b. Play a dirty trick.
>
> #Isn't complementation located at the crossroads between grammar and
> #the lexicon?
>
> Yes, though I largely agree with WG's denial of the grammar/lexicon
> distinction. [I won't here go into which bits I do & don't agree with; I'm
> sure we must have discussed it on the list in the past.]
>
> But still, I don't get what your example is supposed to show.

With this example I just merely meant to show that low-level stuff
are constructions too. They are productive, but not unconstrained.

I think that you'd agree that idiosyncratic complementation cannot be
done solely at the level of syntax proper. Otherwise you are just
doing symbol manipulation and you get interminable amounts of junk.
My view of constructions is that they make use of form-meaning
pairings, and they are structured at various (many?) levels
specificity.

As you say, there is nothing to stop an abstract grammar from
generating "play a magic trick," but my point is that grammar
doesn't stop at the highest level of generalization. Instead,
it stretches all the way down to the lexicon.

Joe
_________________________________________________________________
Home page:        http://lingua.fil.ub.es/~hilferty/homepage.html
__________________ http://www.ub.es/filoan/hilferty/homepage.html