Abstract
It is one thing to have phenomenal states and another thing to think about phenomenal states. Thinking about phenomenal states gives us knowledge that we have them and knowledge of what they are like. But how do we think about phenomenal states? These days, the most popular answer is that we use phenomenal concepts. Phenomenal concepts are presumed to be concepts that represent phenomenal states in a special, intrinsically phenomenal, way. The special nature of phenomenal concepts is said to be important for defending materialism against epistemic arguments for dualism. In this paper I present an account of phenomenal knowledge that does not depend on phenomenal concepts. In fact, I argue that we have no phenomenal concepts. Instead my account appeals to mental pointing, a process that I explain in terms of phenomenal demonstratives. Phenomenal demonstratives are sometimes referred to as concepts in the literature, but I suggest that this is a mistake. I also present a theory of phenomenal demonstratives that equates them with attentional control structures in working memory. In a concluding section I describe how this theory can be used to defuse the knowledge argument for dualism. That is only a subsidiary goal, and my response to the knowledge argument echoes others in the literature. I think the project of developing a substantive, empirically informed theory of phenomenal knowledge has interest independent of debates about mental ontology. That is my central focus. Thinking about phenomenal knowledge can shed light on the relationship between consciousness, attention and memory. This paper has a philosophical agenda and an empirical agenda. Those who reject my philosophical claims about the nonexistence of phenomenal concepts, the conditions..