Advances in the Syntax of DPs - Structure, agreement, and case

78 Anna Bondaruk

(30) TP

DP T′

vP

v

ja

ty

T

EqP

jestem DP Eq′

Eq DP

ø

to

The main problem with the structure in (30) concerns the question of how to deter-

mine that T in sentences such as (29) attracts only one DP, and never enters into Mul-

tiple Agree. Since it is Multiple Agree that underlies the PCC-effect in our account,

the application of Multiple Agree in (30) would predict that the structure should be

unacceptable as the person features of the two DPs do not match (cf. (24) above). The

fact that sentence (29) is perfectly grammatical indicates that no PCC-effect ever arises

in this case and consequently, no Multiple Agree takes place in it. To block Multiple

Agree in (30), we could perhaps claim that EqP is a phase and, therefore, the second

DP is never attracted by T, being trapped in the complement of the phase, by the Phase

Impenetrability Condition (Chomsky 2001).

Consequently, if we adopt the asymmetrical structure above, we will be left with

no explanation for why the PCC-effect does not show up in equatives, but it does in

predicational to być clauses, unless we assume that PredP is not a phase, whereas EqP

is. This claim, however, seems to run counter to the conviction that PredP is actually a

phase (cf. den Dikken 2006, for instance, who uses RP, instead of PredP).^16 One alter-

native that might be adopted to eliminate the possibility of Multiple Agree in (30) is to

claim that T in equatives is not a multiple probe, but actually probes only once. This

assumption, however, lacks any empirical support, therefore it must be abandoned.

All in all, Reeve’s proposal for equatives fails to explain why both DPs, found

within the same domain of EqP, cannot undergo Agree with T and why they do not

give rise to the PCC-effect, in contradistinction to predicational to być clauses. The

very fact that Reeve postulates a new functional projection is not innocent, either.

16. The analysis advocated here also runs against the claim that PredP is a phase, as Multiple
    Agree can target both the specifier and the complement of the Pred head in to być predicational
    clauses (cf. Section 3.2)