Abstract
Goodman's Grue Paradox may be intransigent as a version of the problem of induction, but may be resolved within the more limited context of confirmation theory in which the task is to explicate the basic notion of evidential relevance. Although the green and grue hypotheses are equivalently confirmed if we follow Goodman's use of the Hempelian instance confirmation relation, there are asymmetries than can be exploited if we adopt an "ontic" confirmation theory that uses a causal notion of evidential relevance. I sort out a variety of interpretive confusions about the intended content of the definition of grue and show how the causal approach resolves each in a way that is not paradoxical.
