Abstract

This paper explains a flaw I find in Michael Martin's argument for the view that the phenomenal properties of a hallucination are its being indiscriminable from a certain sort of veridical perception. The argument relies, I argue, on the assumption that if a certain broad mental property sometimes has a certain narrow realizer, it never has any other narrow realizer. This assumption is false. Accordingly, the argumentation fails to rule out "positive multidisjunctivism": the fundamental kind of the veridical experience is a broad property, while some matching hallucinations are subjective and others are intentional.