Turning
to (30) and (31), the theory would receive support if
the position in which PRO occurs can be regarded as
ungoverned, or if it can be maintained that actually
what is indeed there in the position of PRO is trace,
and PRO has been moved from its place of origin; the
kind of analysis proposed for the English construction
"she bought a book to give to John" (Chomsky;
1981b: 64). Now, PRO cannot be claimed to occur in an
ungoverned position, given the possibility of PRO occurring
in the complement position of the CM Ø , only
under which an analysis the ungrammaticality of (25)
- (29) could be accounted for. If one would rather treat
PRO in (30) and (31) as ungoverned, not only would the
above instances be problematic for such an approach,
(38) will also be so because in this construction there
is AGR in INFL, which is singular and feminine and agrees
with kitab with which PRO is coindexed. There
will be no straight forward account of why PRO cannot
be governed by INFL. The second alternative, the one
in which PRO moves to the COMP position leaving behind
its trace (37) is also not viable. |
