The cognitive penetrability of perception brings some new problems to the discussion of perceptual justification in epistemology. In the above case, if the subjects were cognitively penetrated to see an entirely grey banana as yellowish-grey, did this experience give them the same amount of justification for believing that the banana was yellowish-grey as an ordinary, non-penetrated yellowish-grey experience would normally give? Many philosophers maintain that the penetrated experience has less justificatory power, although it remains hotly debated why cognitive penetration makes the experience epistemically downgraded.
In my article “Cognitive Penetration: Inference or Fabrication?” I critically examine a prominent approach to the epistemology of cognitive penetration, according to which some cognitively penetrated experiences result from bad inferences. One version of inferentialism takes the relevant inferences as between two sequential/simultaneous experiences (McGrath 2013), whereas another version allows there to be inferences from subpersonal mental states to personal-level experiences (Siegel 2017). I argue that the former theory fails to account for the banana case because the relevant inference would be between two color experiences, but evidence does not support the occurrence of such a transition. Moreover, the second theory, when combined with Bayesian theories of perception, implausibly implies that a large amount of our perceptual experiences lack justificatory power.
My alternative approach to the epistemology of cognitive penetration first offers more empirical evidence for a psychological mechanism, according to which cognitive penetration can occur through imagining-perception interaction (Macpherson 2012). One set of evidence comes from the cross-modal effects of sensory imaginings on perceptions, and another body of evidence is from neuroimaging. If such a mechanism of cognitive penetration is plausible, then what was involved in the banana case might be: the subjects’ background cognition that bananas are normally yellow initiated an imaginative process that would give rise to a yellow-banana experience, and this interacted with the perceptual process that would lead to a grey-banana experience, resulting in the subjects’ experiencing the banana as yellowish-grey. I also suggest that the priming effects of sensory imaginings on perceptions could be incorporated into the mechanism.
I offer an epistemological theory of cognitive penetration that draws inspiration from the epistemology of imagining. In particular, I propose that when a mental state results from a personal-level mental process, it needs a good evidential basis in order to have justificatory power; however, when a mental state results from a merely subpersonal-level mental process, it does not need a good evidential basis in order to have justificatory power. An experience is “fabricated” when it results from a personal-level mental process, but lacks a good evidential basis. Suppose that you hope that a blizzard will arrive, and your hope causes you to imagine seeing snow when looking out of the window. Your imagining is a fabricated experience, and fails to give you justification for believing that it is snowing. Fabricationism, I argue, explains the epistemic downgrade of cognitively penetrated experiences in various cases satisfactorily, and also leaves room for epistemically innocent/good cognitive penetration.