Introduction

The goal of this chapter is to provide a more complete basis for T's attraction properties by exploiting a neglected property of T, its Tense features. In Section 2, we will argue on cross-linguistic grounds that the properties triggering V-to-T movement and licensing null subjects should be kept distinct. In other words, we reject the idea, first proposed by Roberts (1985), that ‘strong’ agreement of some kind is always responsible for both. Instead, in Section 3, we suggest distinguishing T's tense (i.e. ‘verbal’ or V-related) properties from its agreement/ϕ (i.e. ‘nominal’ or D-related) properties and propose a correlation between richness of tense inflection and V-to-T movement on the one hand and, following numerous researchers since Rizzi (1982), between richness of agreement inflection and the availability of null subjects on the other (see Roberts (Chapter 1) and Holmberg (Chapter 2) for recent proposals in this respect). We then construct a new typology on the basis of this proposal. In Section 4, we first show how Modern English fits into this typology; in this connection, the auxiliary system is crucial, and we make some novel proposals in this regard (Section 4.1). Section 4.2 deals with relevant aspects of the historical development of this system, focusing on the development of systematic V-to-T movement from the loss of V2, and showing how this was a rather short-lived option, as predicted by the typology of Section 3.