82 Anna Bondaruk
be the case in Polish, where the two pronouns in equatives linked by to (cf. (3) and
(26)) can differ in the person feature. What is more, they can also differ in the number
feature, as confirmed by (35), in which the first DP is singular, whereas the second is
plural:
(35) Ja to (jestem) oni.^20
I.nom cop am t h e y.nom
‘I am them.’
Therefore, we would have to diverge from Pereltsvaig’s line of analysis and suggest that
identity interpretation is possible in (34), even though the two DPs do not have an
identical set of φ-features (cf. Citko 2011: 185 for a similar assumption). This seems to
be justified, as even in Russian the two DPs can differ in gender, as confirmed by the
following sentence from Pereltsvaig (2007: 54), where the first DP is feminine and the
conjoined second DP comprises two masculine DPs:
(36) Valentina Ivanova _ sekretar’ gorkoma,
Valentina Ivanova.f secretary.m city_committee
i vernyj tovariš po partii.
and loyal.m comrade.m at party
‘Valentina Ivanova is the secretary of the city committee and a loyal party
comrade.’
One way to proceed is to assume, following Citko (2011: 185), that the two items in the
symmetrical relation must be identical in their categorial features only. The category
identity attested in equatives, however, may not represent an independent syntactic
constraint, but may rather follow from the semantics of equatives (as suggested by
one of the reviewers) and hence may belong to the realm of semantics, not syntax.
However, in Polish to-equatives, just like in other types of copular clauses with to, the
requirement that to link only identical categories is syntactic in nature, as it holds inde-
pendently of whether a given copular clause is predicational (as in (9) above), speci-
ficational (as in (10a)) or equative (as in (2)) and is thus independent of semantics.
The second problem with Pereltsvaig’s account concerns the absence of Agree in
equative sentences in Russian. Let us recall that in Pereltsvaig’s analysis, neither of
the two DPs in equatives enters into Agree with T. The fact that in Polish equatives,
the verb agrees in φ-features with the DP that moves to [Spec, TP] (see examples (3),
(4), (5) and (19)), clearly indicates that Agree must have taken place between these
two elements. However, no Multiple Agree is predicted to apply in to-equatives, since,- Although sentence (35) is syntactically well-formed, it sounds odd for pragmatic reasons,
as pointed out by one of the reviewers.