How (not) to underestimate unconscious perception

Recent work questions whether previously reported unconscious perceptual effects are genuinely unconscious, or due to weak conscious perception. Some philosophers and psychologists react by rejecting unconscious perception or by holding that it has been overestimated. I argue that the most significant attack on unconscious perception commits the criterion content fallacy: the fallacy of interpreting evidence that observers were conscious of something as evidence that they were conscious of task-relevant features. This fallacy is prevalent in consciousness research: if unconscious perception exists, scientists could routinely underestimate it. I conclude with methodological recommendations for moving the debate forward.