Joe:
> I think that the OTHER part of the answer might be this. There's
> a tension between different levels of specificity: At the level
> of more specific instantiations, the token:
>
> - The thing is[tonic stress], is that CLAUSE.
>
> is the most entrenched among 2BE speakers (due to its frequency).
> On the ther hand, the extensions from this central usage are of
> course less prototypical, and probably promote the formation of
> a more abstract construction type (something roughly like):
>
> - NP BE(,) BE[3rd person simple] (that) CLAUSE.
>
> So, to make a long story short, right now the token level is
> generating prototypicality effects because the type level has
> yet to become fully entrenched.
Is this, then, an example of 'lexical conservatism' (Pinker) restricting
a syntactic pattern to certain *inflectional* categories? That would be
very exciting if it were so.
>
> Of course, there is a residual problem here: how on Earth can
> all of this be proved?
How indeed.
>
Jasp
|