Document Type



Master of Arts



Date of Defense


Graduate Advisor

Jon McGinnis


John Brunero

Eric Wiland


Are non-attributive uses of ‘good’ meaningful? Judith Jarvis Thomson, in her book, Normativity, argues that such uses are not. I argue that Thomson is mistaken, but in a way which improves on her basic account of the evaluative and directive judgments central to normative thought. Relations among kinds, and not just relations between kinds and individuals, can be understood as ethically relevant. Such kind-kind relations can provide a semantic basis for non-attributive uses of good. Accounting for such non-attributive uses allows Thomson to address three significant problems with her account. First, this account allows Thomson to make a more nuanced and charitable discussion of ethical realism and anti-realism. Second, this account provides a place in Thomson’s framework for ethically relevant distinctions such as intrinsic/extrinsic goodness, and moral/non-moral evaluation. Finally, this account provides Thomson the basis for resolving issues in her account of directives. Thomson account of directives such as ‘ought’ is defective in that the ‘ought’ of a more general kind eliminates that of a less general kind in cases of conflicting oughts. I demonstrate how such an account generates counter intuitive results and in fact undermines other parts of Thomson’s analysis. The kind-kind relations used in the account of non-attributive uses of good also can be used by Thomson as the basis of a non-eliminative account of directives. Such relations account for why some ‘oughts’ have more weight than others.

OCLC Number