An analysis of Korean case marking proposes an explanation for several aspects of marking, especially stacking and spreading, from a lexical perspective. The explanation has advantages over previous theories in that (1) it explains the morphology of Korean case marking without reference to several levels of syntactic derivation, including intermediate levels for which there is little morphological evidence; (2) it explains case spreading in terms of a cross-linguistically recurrent trend in which grammatical relations may be shared rather than passed on, avoiding the apparent contradiction a feature-passing analysis introduces in some sentences; and (3) it allows differences between two dialects to be described in terms of a slight difference in strategies for morphological representation of case. (MSE)