Question formal status of trees in H&GPSG
carnie at U.Arizona.EDU
Sat Sep 6 14:55:02 EST 2003
> Do you mean that say, once an XP is merged, it should always be moved
> (manipulated, if talking in the spirit of this thread) as a unit, and the
> things being dominated should be treated as sth fossilized forever?
Partly correct: SOME XPs are fossilized. Once a Phase is created, you
can't mess with it. But this only applies to the internal structure of NP,
vP and CP (the nodes that define phases). More precisely, once you build
something on top of an NP, vP or CP, the phase is "closed". Before that
happens however, you can manipulate the internal structure.
Also there is one "escape hatch" (phase edges), which are accessible by
higher phases. Some of us don't believe in edges -- but that's a debate
that's sort of irrelevant.
More information about the HPSG-L