Donald Clark takes another crack at responding to my position in our recent debate about the use of the lecture in learning. He specific responses are pretty light and easily dismissed. His main point, that "there comes a point when the evidence (surely a fundamental tenet in HE) must win out," should be addressed. Clark is under the impression that the studies he cites support the contention that lectures must be abandoned. The current study - Improved Learning in a Large-Enrollment Physics Class - is a case in point. The authors' description of the study - "We compared the amounts of learning achieved using two different instructional approaches..." - strikes me as naive. "Amounts of learning?" Now of course we are blocked by a subscription wall from viewing the actual study (and Clark does nothing more than wave his hand in its direction). But a study of this sort - three hours of traditional lecture vs. three hours of "instruction based on research in cognitive psychology and physics education" - proves nothing. The sample is too small, to localized, and too biased (having already been through a course of lectures). It is frankly shocking that such a study would ever be published. I agree with Clark that the evidence should win out. But it is essential to have some clue about what constitutes evidence, the conditions in which it should be obtained, how it is to be gathered, and the conclusions it supports.