On understanding non-canonical case marking

Abstract

Case marking has been the focus of many grammarians from as early as Panini through the present. Its morphological form makes it easy to identify and its basic function is easily understood: case marking serves to identify the core arguments in a clause. This is intuitive and requires little explanation, and most modern theories of grammar can readily accommodate this sort of ‘default’ case marking. However, there is a wealth of data on the non-standard usage of core case markers that is so often overlooked. Consider Icelandic, an accusative language that typically marks subjects with the nominative case and objects with the accusative case. Icelandic is well known for its ‘quirky case’, where case markers are used to mark arguments not typically marked by that particular marker as in (1) below. (1) a. Accusative subject Mig dreymdi í nótt. me.ACC dreamed in night ‘Ihad a dream last night.’ b. Dative subject Sumum leiðist að læra heima. some.DAT is.bored to study at.home ‘Some (people) find it boring to do homework.’ One possible reason why non-canonical case marking may be overlooked could be because it is assumed to be a quirk of the grammar, simply to be modeled as a lexical stipulation. However, closer examination of the non-standard uses of case marking – within a language and across languages – reveals a semantic commonality for a given non-canonical case marker in different uses. This paper examines types of non-canonical case marking and discusses how they should be modeled theoretically. Moreover, I address what we can tell about the usage of non-canonical case and what the usage of non-canonical case can tell us about the structure