DATR is a formal language designed for the compact representation of lexical knowledge. In recent years, it has been successfully applied in a number of domains, most saliently as a means of generating natural-language morphology from minimal lexical information. We have begun work involving modifications to and diverse applications of DATR. We are currently using DATR to model the declensional systems of various Old and Middle Indic languages, investigating the hypothesis that the differences between the Old Indic declensional system and the divergent Middle Indic declensional systems are largely the effect of a tendency for declension-class membership to realign in such a way that a word's declension and its gender become mutually predictable. We are also using DATR to develop a comprehensive generative morphology for Yiddish. This will serve as the basis for two programs that will be made available as on-line Web resources: the first of these will return a complete morpholexical analysis for any Yiddish word entered as input; the second will check an input word's spelling. We have enhanced the semantics of DATR, and the enhanced version, KATR, is being implemented. We will begin incorporating it into our research on Middle Indic and Yiddish once it is ready.