Author: Falk, Yehuda N. Date: Dec 30 1998 Title: Unmarked Case Email: msyfalk@mscc.huji.ac.il Remarks: This is a draft; comments eagerly awaited Abstract: There is good evidence for two conflicting approaches to nominative/absolutive nominals. Under one approach, despite their superficial appearance, they have the features of ergative or accusative Case. This approach is supported, inter alia, by agreement phenomena. Under the other approach, nominative/absolutive nominals are unmarked for Case. This is a very traditional view, and is supported by agreement and other syntactic phenomena that seem to distinguish unmarked Case from marked Case. The only way to reconcile these apparently contradictory analyses is a two-level approach to Case, in which "Case specification" and "Case realization" are separated: the former being the specification of features on a nominal argument by a head; the latter being the functional category K. In LFG, this means that Case realization is c-structure Case and Case specification is the f-structure feature CASE.