Contribution to Book Symposium on Joëlle Proust’s The Philosophy of Metacognition Unwitting Self-Awareness?

A central question in the philosophy of metacognition is whether the standard tests of metacognition are, in fact, tests of metacognition. For an animal to answer questions only when it knows the answer, and to “opt out ” of answering otherwise, does it need to metacognize? It depends, obviously, on what you think metacognizing requires. If, like Josef Perner (2012) and Peter Carruthers (2011), you hold that metacognizing requires the use of mental state concepts in thoughts about your own mind, then—for reasons I will discuss—the answer may be no. But if, like Joëlle Proust (2013), you think that metacognition can occur in a non-propositional format, without the use of mental state concepts, and without one’s representing one’s own mental states, then the answer is not so clear. But neither, in that case, is it very clear what one means by “metacognition. ” Or so one might object. In response, Proust points to metacognitive know-how—what she terms ‘procedural metacognition. ’ On her view, humans, and a handful of other species, are able to monitor and evaluate their own cognitive states and dispositions, and to fruitfully use such self-evaluations to guide their behavior, without knowing that they have minds at all. This monitoring and evaluation (typically of confidence levels) is accomplished through one’s sensitivity to nonconceptual, nonpropositional representations she calls “noetic feelings. ” While noetic feelings serve in the monitoring and control o