Re: deeply embedded VSO nightmare
From: | William Annis <annis@...> |
Date: | Tuesday, October 23, 2001, 13:11 |
>From: Matthew Pearson <Matthew.Pearson@...>
>
>Many verb-initial natlangs have flexible word order, allowing
>subjects and objects to permute, at least in special cases. The most
>obvious solution to your dilemma that I can see is to allow subjects
>to follow objects if they're especially 'heavy'. Thus:
>
> Lero aldoven na tath daipoth fidial
> "The woman walking quickly is going home"
Possibly. When I first designed the language, I gave it a
pronounced aversion to separating the V and the S. Perhaps I could
weaken that a bit. My tendency now would be to shift 'aldoven' out to
the front if it goes with 'lero.'
>There shouldn't be any problem with this, since I notice you have
>case marking in your language.
Yep. In theory I could do almost anything to disambiguate,
but I feel that if I'm going to insist that the main word order is
VSO, and the modifiers follow the thingie modified, then there had
better be a good reason to go messing about with the word order.
>This kind of operation is called 'Heavy-NP Shift', and many
So, I read this last night, and thought about it for a while.
Strangely, "heavy-NP shift" turned into "heavy-NP *lift*" so I had
this mental image of linguists hanging out in Gold's with other
gym-bunnies talking about heavy-NP lifting and their favorite protein
drinks.
>However, if the direct object is especially 'heavy', it can shift to
>a sentence-final position, as shown below.
This is interesting. This, and some of the other responses,
gives a very good start on this problem.
--
William Annis - System Administrator - Biomedical Computing Group
"When men are inhuman, take care not to feel towards them as they do
towards other humans." Marcus Aurelius VII.65