11-10-03 13:32, Ray wrote:
> Shouldn't we analyze a sentence's structure on the basis of how it
> meant and was used at that particular stage, rather than on the basis
> of how it used to mean in an earlier stage?
>
> Since a reanalysis had taken place and 'genotudne' carried the
> meaning of a main verb(as opposed to an auxiliary), why not claim it
> is a main verb in the sentence? The speaker who uttered(or wrote)
> that sentence probably treated it as a main verb.
>
> I know it is inflected as an adjective due to the presence of mete,
> but that's where I think the problem lies: The participle-form verb,
> formerly an adjective, now agrees with the object only.
Of course when the past participle was still felt to be part of the
possessive construction, it had to agree with the object in number and
case. The "true perfect" construction, also common in OE, involved an
uninflected past participle. Being "at a transitional stage" means that
a construction remains ambiguous; I suspect that the sentence in
question was still potentially interpretable as a "possessed object plus
complement" construction, especially because of the presence of a
possessive pronoun (<hiora>). The writer may have had used the inflected
participle mechanically, on the analogy of sentences like "They had
their house-ACC built-ACC", ignoring strict logic. It's true that you
can't eat a cake and _have_ it if <have> means literally 'possess, own',
but meaning is a flexible thing, and <have> has always expressed various
kinds of abstract relations. It's this semantic generalisation of <have>
that made the development of the perfect possible.
Piotr