Labeling Defective T*

Abstract

This paper proposes a new analysis of so‐called ‘defective T’ (a postulate originally discussed in Chomsky 2000, 2001) and seeks a label‐based theoretical explanation of phenomena associated with the distribution of subjects in infinitival clauses. I argue that we do not need to postulate ‘defective T’ as an element of the lexicon. Following Chomsky’s (2013, 2015) labeling approach, which investigates only finite clauses, I argue here that the subject of infinitive‐to is pair‐merged structure <T, C> and the associated predicate forms an {XP, YP} structure, which is unlabelable. To attain labeling, the subject position ([spec, TP]) in the infinitival clause moves out and the label of the resulting structure is the amalgam <T, C>, which has no visible Spec (violating EPP). As a consequence, we eliminate the Merge‐over‐Move principle.

On the properties of phase heads in raising and passive clauses: DP movement and Transfer options

“… 11 As noted by a reviewer, Sugimoto (2021) (based on ideas put forward by Epstein et al 2016), also argues that C is present in infinitival clauses, but external pair-Merge of C and T makes C invisible. Sugimoto's approach, however, entails that in both raising and ECM clauses movement to the embedded Spec-TP is optional (Sugimoto 2021: 157). As discussed, C&F-S's observations regarding VP ellipsis suggest that this is not the case (see (1) above).…”