>> ## Under co-enunciation it's not so much that words are invisible, but
>> rather that they're fused with other words; e.g. de + le = du. I've also
>> argued for (what I think is) a much more radical departure from the
>> original WG principles by arguing for PRO, which isn't a matter of
>> coenunciation; and we've considered the possibility that other optional
>> valents might in fact be present as invisible lexical items, as in And's
>> recent example of "these" meaning "these trousers".
>
>I think that if you allow de + le = du then it is possible to generalize
>this to cover PRO and deletion. I don't accept PRO, but nonetheless
>it wd be possible to have a rule saying it coenounces with its
>neighbour. Likewise for the deletion: each subordinate of "these"
>coenounces with "these".
>
>I don't see any natural nonstipulative way to prevent coenunciation from
>being able to do this sort of stuff, but given that it can, I think it's
>a good idea to recognize coenunciation as the mechanism behind both
>portmanteau-ization and deletion.
Nik: isn't this a problem? Don't we want as much of our grammar to be as
unstipulative as possible? Are we to asume taht everywhere you can have
co-enunciation you do?
>
>> > I am interested in establishing what kind of consensus there is
>among
>> >WGians on the existence of invisible words, where you find them, what
>> >diagnostics there are for them and what constraints there are on their
>> >(non)appearance. One of the main reasons I'm interested is because of
>> >comparisons with school-of-MIT syntax. In school-of-MIT syntax, you know
>> >you've got an invisible word because it blocks contraction.
>> ## That used to be the claim but surely it's been overturned because of
>all
>> those people who do allow contraction across traces, e.g. "That's the
>> person I wanna come to my party".
>
>BTW, I don't think any of us here have advocated an invisible word
>analysis for traces.
Nik: no we haven't, which is fascinating bcs we use invisible words for
other things, and don't need them for traces which are the canonical
invisible word in MIT-syntax!
Nikolas Gisborne
http://www.hku.hk/english/staff/gisborne.htm
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|