A tactile screening-off problem for naïve realism

This paper develops a veridical-first screening-off problem for naïve realism using simple cases from touch. Unlike traditional, hallucination-based arguments, the problem arises entirely within ordinary tactile perception and realistic haptic simulation, and it turns on the multiple realisability of a single tactile experiential kind by distinct worldly conditions. A wooden ball, a metal ball, and a haptic device can all produce the same ‘smooth sphere’ experience. I argue that once we fix the shared pattern of local tactile interaction and the subject’s neural responses that realise this experiential kind, the particular external object that happens to produce it becomes explanatorily redundant with respect to phenomenal character. This challenges naïve realism’s claim that worldly objects play a non-derivative constitutive role in good-case experience. I close by considering four naïve realist responses and explaining why they do not resolve the tactile screening-off problem.