on this assumption, they explain the complementary distribution of the mor-
pheme ore and V-movement to C. If their analysis is applied to Chomsky’s
(2013) analysis, the following derivations are carried out (although Car stens
and Shoaff present other derivations based on a different movement system.)^2
(6) VSO
[CP Crel [TP V+v+T [vP [DP NP]...
→ [CP V+v+T+Crel [TP <V+v+T> [vP [DP N P]...
(7) Ore SVO
[CP Crel [TP [DP ore NP] V+v+T [vP...
→ [CP ore+Crel [TP [DP <ore> NP] V+v+T [vP...In (6), C minimally searches the V+v+T amalgam before the subject DP
moves to T-Spec. (as schematized in (2)). As a result, the V amalgam moves
to C. In (7), on the other hand, the subject DP is moved to T-Spec before C
minimally searches. Under minimal search, C searches and fi nds D within the
subject DP and D-to-C movement takes place. (as schematized in (3)). That
is, both of these orders generate convergent outputs and the two derivations
are equally optimal/effi cient, computationally, applying all and only the (universal
syntactic) operations necessitated by convergence.
4 A note on INTRA and INTER I-language variation
Notice that variation between I-languages and variation among “constructions
in a given I-language” must both satisfy the same constraints; UG and third
factor under experience. The theory therefore predicts, in principle, a cor-
respondence between (A) the variation displayed by different I-languages and
(B) the variation displayed within a particular I-language. For example Obata
and Epstein (2011, 2012) argue that the enduring oddity of English tough-
cons tructions is ‘nothing mor e’ than a type of lexical item inducing the same
kind of (computationally effi cient) object agreement as is seen characteristically
in Kilega wh-movement (Carstens 2005, 2008). That is, English tough-
constructions are “just Kilega”. The pre diction that in tra-language construc-
tional (lexical) variation comports with inter-I-language variation is predicted
by the theory – all such systems are constrained by UG and third factor, with
third factor leaving certain aspects of derivations underspecifi ed, thereby
allowing variants triggered by externalized outputs (generated by variant
systems, allowed by third-factor underspecifi cation, generating variable exter-
nalized outputs.... etc.!) By hypothesis, the orders in (2) and (3) both
apply “within English”, thereby generating variant outputs in English subject
wh-movement cases:
(8) a. Which man visited New York?
b. Which man did visit New York?
(with emphatic in tonation of did; Koopman 1983)
134 Miki Obata and Samuel Epstein