Re: storage vs computation

Nik Taylor wrote:

> Ed Heil wrote:
> >
> > Good point.  The 'Construction Grammar' theorists such as Charles
> > Fillmore have suggested that grammar is *NOTHING BUT* sets of pre-made
> > constructions, which are interlinked and fitted inside the other.
>
> I have a hard time accepting that, just as I doubt theories that state
> that language is ALL rule-based.  I suspect the truth is closer to the
> middle, a combination of premade constructions and rules.
I suspect that the Construction Grammarians would formulate what you
would call "rules" as extremely generalized constructions; e.g. the
rule "prepositions come before their objects" would be formulated as a
construction [ [PREPOSITION] [OBJECT] ] which is stored exactly like
any other lexical item.

Calling them constructions rather than rules would be justified on
larger, more theoretical grounds, but the differences in description
in a simple instance like this would be purely a formality.

+ Ed Heil ---------------------- edheil@postmark.net +
|  Waitin' for th' dinner bell t' do the bell thing  |
|        Dinner bell dinner bell ding  --TMBGs       |
+----------------------------------------------------+