Reconsidering structural case in Finnish

This paper is a response to Kiparsky (2000), who convincingly argues that the complex case marking in Finnish can only be captured in a theory that makes a clear distinction between abstract case and morphological case, and is coupled with the OT Correspondence Theory of McCarthy and Prince (1995). Acknowledging nearly all of Kiparsky’s theoretical presuppositions, as well as his empirical insights (to which I have nothing to add), I nevertheless arrive at different results. My paper is based on the assumptions of Lexical Decomposition Grammar (LDG, Wunderlich 1997, Stiebels 2000, among others), which in turn owes much to Kiparsky’s fundamental idea of abstract case features that encode the hierarchy of theta-roles. 1. The specification of case The examples in (1a) (taken from Kiparsky 2000:[21]) show that all four structural cases can realize the object role of a bounded predicate such as tuo- ‘bring’. Accusative is restricted to pronouns, genitive to singular, and partitive to the unspecific reading. Hence, these three cases are object cases, to be specified as [+hr] ‘there is a higher role ’ (which corresponds to [−HR] ‘non-highest role ’ in Kiparsky’s system). The genitive is also blocked in the imperative construction, shown in (1b), as well as in the passive. In the context of negation, only partitive is possible on the object, as shown in (1c); the same is true for unbounded predicates such as etsi- ‘seek’