In the more than quarter century since Gettier published "Is Knowledge Justified True Belief?"'" the attempts to resolve the so-called Gettier problem have been numerous and disastrous. By this point the inductive evidence is almost irresistible that the next proposed solution will similarly fail. The cumulative effect of this is to make epistemology replace economics as the dismal science. Still hope springs eternal. In any case when someone as distinguished as Roderick Chisholm comes forward with a proposal and repeats it in successive publications one ought to take a serious look. First, however, we should remind ourselves of the structure of Gettier's original example.2 A person, S, has evidence, e, which would justify a belief in p sufficiently to make S know that p, at least if all the other conditions on knowledge were met. But in this case S does not believe p, but rather believes (p v q) where q is some other proposition for which S has no evidence whatever. As luck would have it p is false, but q is true, so (p v q) is true anyway. Because e is at least as strong evidence for (p v q) as it is for p, S's belief is justified as well as true.3 Many writers, including Chisholm, have fastened on the falseness of the justified disjunct. In the second edition of his Theory of Knowledge Chisholm's proposal was to say that there is something defective about the justification of p.4 This is because there was something defective about the evidence e, namely, it would sufficiently justify a false claim. Putting the matter this way, however, has the undesirable effect of denying that S knows e. For this reason Chisholm revised his definitions, publishing the result in