On the Phasehood and Non-Phasehood of Cp

Abstract

In this paper, we argue that the combination of an Agree feature and a Tense feature on CP determines the (non-)phasehood of CP. We propose that the presence of both features makes CP a phase while the absence of one or both of the two features makes it a non-phase. It follows from this proposal that finite complements are CP phases while control complements, raising complements, and Accusative with Infinitives are non-phase CPs. It is then predicted that operations can apply across the CP of the latter types. We will show that this prediction is borne out.*

On the Nature of English Gerunds

“…Kanno (2008), which is based on the Feature Inheritance analysis, regards constructions that do not show a tense morphology as [−Tense]. Even though his framework is different from the present analysis, the property of the Acc-gerund falls into line with his notion of [−Tense].…”

“…6 The Acc-gerund has now turned out to be a complete CP structure with somewhat defective T inside, and the next task is to assess whether C under discussion forms a phase. Although C is basically assumed to be a phase head, here I follow the following suggestion by Kanno (2008) on the identification of the phasehood. (13) The phasehood of CPs is determined by the combination of an Agree feature and a Tense feature; if a CP possesses both Agree and Tense features, it is a phase; if a CP does not have one or both of the two features, it is not a phase.…”

Shared Labels and Partial Labeling

“…12 To establish a multiple Agree relation in ( 26), CP should be a non-phase (Kanno (2008)). An alternative is to assume that (obligatory) control complements do not have the CP projection (Bošković (1997), Grano (2015)).…”