Case as Agree Marker

This paper argues in line with recent work by Pesetsky and Torrego that given the operation Agree there is no need for the concept Abstract Case in Narrow Syntax. Morphological case exists in many languages, and is here, following a suggestion by Halldór Sigurðsson, seen as the result of an instruction to PF how to translate syntactic structure into morphology. I will call this instruction CAM, short for “Case as Agree Marker”. In my paper I will explore how this instruction is formulated for Modern Icelandic, using an implementation of the Minimalist feature driven framework based on two features, tense (τ) interpretable in T and v, and a combined feature φ for person, gender and number, interpretable in DP. I will suggest a formulation of CAM that accounts for the distinction nominative – accusative in modern Icelandic, including quirky accusatives and other cases of “non-Burzionian” accusatives. Although CAM is not formulated for dative case, the distribution of many types of datives in Icelandic is indirectly accounted for by CAM as well.