REPLY TO BREWERMy inferentialist analysis of hijacked experiences allows that these experiences can be conclusions of inference. If experiences are conclusions of inference in these cases, what kinds of mental states can figure as the inputs to the inferences?In Brewer's (2019) initial gloss on these cases, the input to such inference is always another experience, and that experience always has a baseline amount of epistemic power. He then asks how this gloss would apply to memory color, where one ends up experiencing a banana as yellowish.The answer is that it wouldn't because the gloss is too limited. The inputs to inferences that culminate in experiences can be informational states of the perceptual system as well as cognitive states such as beliefs, suspicions fear (more exactly, its representational component, as well as such components of other emotions). Inferences to experience can occur wholly within the perceptual system. 1 In my analysis of the case of memory color, the ultimate status of the experience depends on the structure of the underlying processing. 2 The conclusion is false, but this fact alone does not worsen the inference. An inference to a false conclusion could be epistemically good. It will be a poor inference, I claim, when information indicating that the banana is grey is not given its due weight, or the prior assumption that bananas are yellow is given too much weight, or both. Here, there is no other experience that figures in the inference. A fortiori there is no positively charged experience that figures as an input to the inference. So premise E1 in Brewer's reconstruction is not part of my analysis. There is no greyish appearance, either in the actual experiments (as they're typically interpreted) or in the cases I discuss. 3 It strikes me as incoherent to say that a single experience at a time both presents the same visible parts of a banana as two different hues at once, greyish and yellowish. If Brewer thinks this may be my view, I wonder why he did not criticize it on these grounds.In line with the mistaken assumption that perceptual inputs are always experiences that retain baseline epistemic power, Brewer takes me to be committed to a "baseline experiential input immune from influence" by prior assumptions, and then says this commitment creates instability. But there is no such commitment. Part of what makes information processing perceptual is that it takes in information from the environment. The importance and interest of the phenomena that motivate the Rationality of Perception thesis stem from the fact that experiences can occur far downstream of such inputs. Phenomenologically, experiences are passive and receptive, but informationally, they can in principle be heavily shaped by one's prior outlook. When that happens, any epistemology of perception that Philosophical Issues. 2019;29:403-410.wileyonlinelibrary.com/journal/phis