This paper addresses morphological causativization in Hindi/Urdu to argue that a distinction between temporally dependent and temporally independent subevents is what underlies the direct vs. indirect causation distinction in that language. I thus argue against analyses couched in terms of a lexical vs. syntactic distinction, and also against a syntactic embedding account of the two types of causative (i.e. `inner ' vs `outer' , or 1st causative vs. 2nd causative). The analysis is framed in a syntactic decompositional system, and uses as evidence the distributional facts about the attachment of the -aa and -vaa suffixes in Hindi/Urdu, as well as the interpretational possibilities of the instrumental case-marked adjunct in -se to argue for the more aspectual analysis involving subevents.