Finiteness, inflection, and the syntax your morphology can afford
In this paper English and Norwegian are compared with respect to a range of syntactic constructions (negated clauses, negative inversion, polarity questions, declaratives, subjunctives, why-not- constructions and infinitives). The paper also discusses preterit-participle mix-ups in English and relates these to the behaviour of main verbs in the investigated constructions. It is argued that in both languages these constructions require either an explicitly encoded finiteness feature, or, for the second set of constructions, the explicitly encoded absence of such a feature. English auxiliaries and main verbs differ in their respective feature specifications as regards finiteness, hence main verbs are banned from many of the positions where auxiliaries appear, and vice versa. English main verbs are bereft of a productive morphologically expressed finiteness distinction (except for in a few, but frequent irregular main verbs distinguishing preterits from participles), whereas auxiliaries have and be exhibit full paradigms of finite and non-finite forms. The modals and auxiliary do, on the other hand, exist only as finite forms and are hence banned from positions where non-finiteness is explicitly required by a given syntactic operation or construction. Hence, the finiteness feature is unevenly distributed across verbs and auxiliaries in English, but evenly distributed in all verbal categories in Norwegian, where all main verbs, modals and auxiliaries have full paradigms of finite and non-finite forms. This state of events, it is claimed, gives rise to the observed parametric variation between the Norwegian and English verbal systems.