Differentiating by Differences in Knowledge
Steven James
University of Texas at Austin

I argue that perceiving, (perceptual) remembering, and (perceptual) imagining are best distinguished in terms of differences in the knowledge they respectively underwrite.

Historical discussions have typically focused on things made available via introspection and the general consensus is that such proposals are unsatisfactory, although there remains disagreement over precisely why they are unsatisfactory. I claim that part of the trouble stems from a tendency to run two kinds of issues together: (i) issues concerning self-knowledge of mental states; (ii) conceptual issues concerning the taxonomy of mental states. To keep things simple, I focus on two questions. (Q1) How does a subject S know that she is seeing, and not remembering, or imagining? (Q2) What are the fundamental differences between seeing, remembering, and imagining such that theorists can differentiate them?

Answers to Q1 naturally appeal to introspection and in this context something in the spirit of the historical approaches is plausible. In contrast, introspection-based answers to Q2 are wrongheaded. The problem is simple; although introspection may typically underwrite subjects’ knowledge that they are e.g. remembering and not seeing, it does not reveal fundamental differences between remembering and seeing. Consequently, it can lead to the misclassification of mental states and something else is needed.  

I propose that an answer to Q2 can be developed in terms of the knowledge that these states respectively underwrite. In short, perceiving underwrites knowledge of a present perceptual environment; remembering underwrites knowledge of past perceptual environments; and imagining underwrites knowledge of possible perceptual environments. Unlike the introspection-based proposals, this proposal captures difficult cases and provides a framework for further research.

Here is an outline of the paper in more detail:

In §1 I introduce the topic by way of example and distinguish the two key questions.

In §2 I briefly consider three plausible kinds of introspection-based answers to Q1 cf. (i) Hume 1740; (ii) Russell 1913, 1921, Byrne 2010; and (iii) Wittgenstein 1953, Urmson 1967, McGinn 2004).

In §3 I discuss three cases that demonstrate the limitations of these proposals. (i) In cases of de re hallucination (cf. Penfield 1958; Sacks 1970; Dennett 1991), the proposals mistake something previously perceived (i.e. arguably remembered) for something presently perceived; (ii) In cases of the “Perky Effect” (cf. Perky 1910; Byrne 2007), the proposals mistake something perceived for something imagined; and finally (iii) in some everyday cases (cf. Martin and Deutscher 1966), the proposals mistake something remembered for something imagined. Something else is needed if we are to answer Q2.

In §4 I develop a proposal that classifies these mental states in terms of the knowledge they underwrite and I show that the proposal satisfactorily handles the aforementioned cases.

In §5 I defend the proposal against two kinds of objections: (i) the knowledge criterion is too demanding (cf. Bernecker 2010; Byrne 2007, 2010; Gendler 2011); (ii) the knowledge criterion is too thin. And finally, I briefly discuss directions for further research.