On the relation between V-to-I and the structure of the inflectional paradigm

In this paper an account is provided of the trigger for V-to-I movement. Th is account explains why relatively rich inflection involves a distinct infl ectional node in the clausal structure whereas relatively poor inflection i s base-generated on the verb (Rohrbacher 1999), without having to assume th at there are two fundamentally different types of inflection. I will argue that there is not a direct causal connection between rich inflection and V- to-I but, rather, that the two are related because having V-to-I and not ha ving rich inflection are alternative ways of avoiding a violation of the sa me constraint, namely a general constraint that disfavours complex structur e below the word (X-0) level. This constraint is a ranked, violable constra int, as in the optimality-theoretic conception of grammar. Its interaction with two other relevant constraints, which concern economy of movement and realisation of the input in the output, will account for the fact that lang uages can vary in their tolerance level for the amount of inflection on unm oved verbs. The analysis thus explains why no single definition of richness can exactly divide the V-to-I languages from the non-V-to-I ones. It will also account for the English 'V-to-I for auxiliaries only' puzzle.