88 Anna Bondaruk
namely the DP that moves provides the ‘bare’ small clause with a label and becomes a
closer goal for Agree with T and hence determines subject verb agreement.
Having outlined the syntactic derivation for Polish true equatives advocated here,
let us now turn to their semantics. Since equatives establish the identity between the two
referential DPs around the copula, their semantics must be different from that of predi-
cational clauses, in which only the pre-copular element is referential, while the post-
copular one is non-referential. Following Williams (1983), Partee (1987) posits a single
be for all types of copular clauses, and argues that the identity interpretation found in
equatives cannot be in any way linked with the meaning of the copula, but rather results
from the type shifting operation that turns a referential expression into a non-referential
one. In Partee (1987, 1998) the type shifting operation involves an IDENT operator,
whose function is to turn an 〈e〉 type expression into an 〈e, t〉 type one. After type shift-
ing has applied, the second DP in an equative such as (38) above is understood as the
property of being (identical to) ty ‘you’, which is then predicated of the other DP. Con-
sequently, the IDENT type-shift turns equatives into predicationals in semantics, and
removes the anomaly that equatives seem to give rise to, resulting from the fact that
they apparently combine two referential expressions.^29 Under Partee’s approach, equa-
tion is regarded as a special type of predication, and the only difference between these
two types is the presence of the IDENT operator in the former and its lack in the latter.
The final question to be addressed here is how the analysis offered here fares with
respect to the Russian equatives, which have served a starting point for our account of
Polish data. It seems that the analysis advanced in the paper is fully applicable also to
Russian equatives. The only difference between the two languages that emerges from
the discussion presented above lies in the categorial status of the word ėto/to. Whereas
to in Polish is a functional head, ėto seems to represent a maximal projection (cf. Geist
2008 , and Reeve 2010 in Section 4 above). A different approach to the status of ėto
is taken by Błaszczak and Geist (2000), who treat ėto as a head of its own projection
which they call ĖtoP. If we follow Błaszczak and Geist’s proposal concerning the cat-
egorial status of ėto, then the only modification that we would have to make in the
structure such as (38) and (41) is to posit an additional projection below TP and above
vP, namely ĖtoP. Consequently, a Russian true equative such as (43) below would have
the structure as in (44):
(43) Ciceron ėto byl Tullij. (Geist 2008: 87)
Cicero.nom cop was Tully.nom
‘Cicero was Tully.’- Partee (1998) mentions other type shifting operators. One involves the Iota operator
that shifts a predicate into an entity, i.e. the 〈e,t〉 type into the 〈e〉 one. Another one is the Pred
operator of Chierchia (1984) that maps an individual onto its property correlate.