Abstract
This paper argues that there is a non-standard but theoretically important notion of “veridicality”, on which perception is only veridical if it does not scramble the objective physical structure of the environment. I argue that non-veridicality in this sense is compatible with veridicality in more familiar senses, and motivate the importance of the notion. For example, I think a certain kind of realism about the scientific enterprise (that it can uncover nature’s natural structure by inference from the manifest image), assumes that perception is veridical in this sense. I think the best reconstruction of Hoffman, Singh and Prakah’s “Interface Theory” is as the view that perception is non-veridical in this non-standard sense – a view that I think is reasonably understood as a kind of transcendental idealism, because it makes the objective structure of the world unknowable to us. They offer debunking arguments against perceptual veridicality (in this special sense). I respond to these arguments, and sketch a realist alternative.