Does Hungarian have a case system?

I argue that case markers in Hungarian are best thought of as ‘fused postpositions’. There is no need to set up a separate syntactic or morphological [Case] attribute as such. Rather, we just need a morphological principle stating that nominals (including pronouns) have a special form, the traditional case form. In this respect Hungarian is crucially different from languages such as Latin (which requires both a morphological and a syntactic [Case] feature) or Finnish (which requires at least a syntactic [Case] feature). I discuss certain typological issues arising from this analysis, arguing that when grammarians refer to Hungarian ‘cases’, they are really referring to a rather more general notion of ‘canonical grammatical function markers on dependents’.