Dick:
> Jasper:
> >It strikes me that I missed an opportunity to say the following on the
> >strand ('constructions?') started by And last year:
> >
> >An important consequence of (or precondition for) saying that WG 'has
> >constructions' is that its lexicon and grammar are not separated. The
> >only difference between a word and a word class is generality. Similarly
> >for the statement that RELY takes the complement ON and the statement
> >that verbs can take noun subjects.
> ## I agree. Also WG is a static constraint-based model, without procedures
> - that's another important plank in the construction-grammar platform which
> distinguishes it at least from Chomsky.
TG is formulated in procedural terms because that it Chomsky's predilection,
but I think he's in a minority in having that predilection, and plenty of
his followers would want to insist that the model is not inherently
procedural and that it could be reformulated declaratively with no change
to the model. I share this view, on the whole, and I think of TG in
declarative terms. All that the procedural formulation does is put in
place an unnecessary barrier to comprehension and a potential source of
confusion and misunderstanding.
--And.
|