V-to-I movement in the absence of morphological cues: Evidence from adult and child Northern Norwegian

Abstract

Several people have pointed out that there seems to be a close correlation between inflectional morphology and verb movement (see e.g. Kosmeijer 1986, Holmberg & Platzack 1988). The nature of this correlation has been claimed to go in both directions. Vikner (1994, 1995) and Rohrbacher (1999) have both suggested that the verb can only move to an inflectional head if the morphology is rich enough. Bobaljik (1995), Thráinsson (1996), and Bobaljik & Thráinsson (1998), on the other hand, argue that the correlation goes in the other direction, i.e. that rich inflection is a reflection of verb movement, rather than the cause for it. A correlation between morphology and verb movement has also been suggested in first language acquisition (Santelmann 1995 on Swedish, Clahsen et al. 1996 on German, Déprez & Pierce 1993, and Meisel 1994 on French). Several of these studies indicate that children use inflectional morphology as a cue for verb movement in the acquisition process, and that they employ verb movement as soon as they acquire verbal inflection. In this paper I will present new data from a dialect of Northern Norwegian which challenge the strong correlation between verb movement and inflectional morphology in both the adult language and in the acquisition of this dialect. More specifically, this dialect appears to have optional independent V-to-I movement despite the fact that the inflectional morphology is very poor. With respect to the acquisition of this dialect, preliminary data from one subject seem to indicate that children to some extent overgeneralise this verb movement pattern into constructions where adult speakers would not allow it.