The Category P Features, Projections, Interpretation

177
.
5.2.2.1 Op-movement: The syntactic operation which turns an argumental CP
into a predicative one (e.g. a relative CP, CP in object gap constructions) is the null
operator (Op)-movement familiar from Chomsky 1977, Chomsky 1986, Browning
1987, Rothstein 2001, among others. For instance, in the relative clause (8a) and the
English TC (8b), the Op generated in object position moves to spec-CP and binds its
trace, creating an operator-variable chain rendering the CP predicative, a CP with an
open position:
(8) a. The book [ CP Op i that [ IP Dan bought t i ]] is interesting
b. The book is easy [ CP Op i [ IP PRO to read t i ]]
Let us consider whether a similar syntactic operation is plausible for the lenominal
sequence in object gap constructions in Hebrew.
The le-nominal sequence can be analyzed either as an NP (with le- affix), or as a
PP. 6 If it is an NP, the Op base generated as the internal argument of the N, can move
only to spec-NP, as shown in (9):
(9) NP
Op N’
N
t
It is rather obvious that the Op-movement in (9) is completely illicit, as the
specifier of an NP is not an operator position, i.e. it is an A- rather than an A’-
position. An A-position is a position where a theta-role can be assigned (Chomsky
1981). An A’- position is a landing site of an operator and is disassociated from thetaassignment.
7 By hypothesis, then, specifiers and complements of lexical projections
(among them NPs) are A-positions. Indeed, it has been argued by various authors
(Ritter 1988, Szabolcsi 1992, Siloni 1994, 1997 and references cited therein), that the
6 See 5.2.1, where based on the obligatory indefiniteness of the nominal, it is assumed that the nominal
is not a DP.
7 The distinction between the two positions was redefined in Chomsky (1993) in terms of L-
relatedness. The definition does not change the status of a specifier position of a lexical projection.