We are continuing our investigations of incremental learning in several directions: (1) combining domain-independent aspects of COBWEB, STAGGER, ID4, and other closely-related systems, (2) exploring incremental learning's utility in application areas such as intelligent tutoring, and (3) analyzing incremental concept learning in terms of other theoretical frameworks such as belief revision. We elaborate specific projects below, listing journal articles to benchmark our progress when appropriate. We also expect conference (AAAI, IJCAI, Cognitive Science) publications to precede our journal submissions, but do not list them here.