This dissertation investigates the case markers ka and lul in Korean from an empirical perspective. Korean is a well-known nominative-accusative language in which the subject and object are marked by the post-nominal particles ka and lul respectively. As has long been observed, the subject and object are not always overtly marked by these particles, particularly in informal spoken Korean. In fact, the presence or absence of these particles has instigated much debate as to whether or not they are case markers, and this is still disputable. Despite many studies examining the structural properties of the case markers, there are very few studies investigating them from an empirical perspective. The case markers cannot be explained solely from structural perspectives, and discourse-based observations should also be considered in the discussion of the structural properties of the case markers. I propose two discourse-pragmatic factors, namely processing load and informational prominence, for their analysis.

The earlier part of the dissertation introduces several issues that are related to the so-called case markers ka and lul, and presents the issues that are pertaining to this dissertation, e.g., when the case markers do and do not overtly occur in a clause. The later part of the dissertation discusses the data and the factors used for the analysis of the case markers in Korean. Throughout the dissertation, the observations made for the case markers in Korean were compared to those in Japanese in an attempt to delineate the differences between the two languages.

In this dissertation, approximately five hours of informal two-party conversations were used for the investigation of the presence and absence of the case markers in Korean. I proposed several individual factors that are related to either processing load or informational prominence or to both. The two discourse-pragmatic factors that are used for the discussion of the case markers well capture the presence and absence of such markers, i.e ka and lul tend to overtly occur in a clause when the subject and object would represent processing load, informational prominence or both, and they tend not to overtly occur in a clause when neither of the two factors is represented by the subject and object.

The approach I take for the discussion of the case markers in Korean is original, and effectively describes their presence and absence. Along with other studies of the case markers from an empirical perspective, the present study contributes empirical evidence depicting cross-linguistic differences between Korean and Japanese case markers, which still need to be further delineated, and sheds light on the discussion of the structural properties of the case markers in Korean from an empirical perspective.