The traditional conception of knowledge as justified true belief is refuted in two famous counterexamples by Edmund L. Gettier. Roderick M. Chisholm has attempted to rescue a version of the traditional conception by distinguishing between defective and nondefective kinds of justification, and redefining knowledge more specifically as nondefectively justified true belief. Chisholm's revised definition avoids Gettier's counterexamples, but goes too far in the opposite direction, imposing conditions that are too narrow and not jointly necessary for knowledge. Chisholm's definition excludes some (...) claims that intuitively constitute genuine knowledge1 by entailing that if a true belief is invalidated as knowledge when defectively justified by a total body of evidence that also makes evident at least one false proposition, then no knowledge whatsoever can be supported by the same evidence. An alternative analysis of knowledge is proposed, according to which the potential loophole between the state of affairs that justifies belief in a proposition, and the state of affairs that makes the proposition true, permitted by the traditional concept of knowledge and discovered by Gettier's counterexamples, is closed by redefining knowledge as semantically‐epistemically evidentially relevant justified true belief. (shrink)