In this paper, mainly based on the data with the constituent XP-man-i ('XP-only-F'), I discuss some distributional and functional properties of the markers -ka in Korean to suggest that the marker -ka that is stacked on the dative marker -eykey ('to') or on particles like -man ('only') (which is generally called stacked -ka) is not a case marker, but a focus marker with a special function, and that unstacked -ka may also function as a focus marker in certain contexts. I also discuss further data with the marker -lul to suggest the following: First, stacked -lul is a focus marker and unstacked -lul may function as a focus marker in certain contexts, like unstacked -ka. Second, there are two types of focus markers, which differ in their morphological realization patterns. Based on the discussions, I conclude that case duplication or case stacking is not allowed in Korean.