Explaining our Actions: A Critique of Common-Sense Theorizing (CUP, 2025)
Peter Carruthers
4. Belief and knowledge
Chapter 7 discusses the other main component in common-sense belief-desire psychology: beliefs. (Chapter 8 – which discusses the question whether attitudes like belief, desire, and intention admit of degrees – takes up the alleged existence of graded beliefs, or credences. It shows that the data are better explained in terms of beliefs and judgments that embed analog-magnitude representations of epistemic likelihood. This option has been overlooked in the field because philosophers seem unaware of the extensive evidence of analog-magnitude representations of likelihood in almost all animals that have ever been studied. See, for example, Stephens & Krebs 1986; Balci et al. 2009; Gallistel et al. 2014. It is analog-magnitude representations of likelihood that combine with analog-magnitude representations of outcome-value to compute expected values and hence to determine choice.) Chapter 7 shows that semantic and episodic memories are distinct kinds of belief, with differing types of content, differing functional profiles, and realized in partially separate brain networks. And in the ongoing debate about the distinction between episodic memory and future imagining it sides with Robins (2022) in arguing that the processes involved are, indeed, distinct.
When philosophers discuss beliefs, however, they generally have in mind states that result when one considers the available evidence and then makes up one’s mind. A juror considering the question of the defendant’s guilt, for example, and reflecting on the strength of the evidence, might soon make up her mind that he is, indeed, guilty. I argue, however (following the work of Frankish 2004, 2007), that in many such cases the resulting state is not a belief at all, but rather an intention. (Frankish himself follows folk-psychology by continuing to describe such states as beliefs.) When one makes up one’s mind that P, one commits oneself to the truth of P. And that means committing to thinking, reasoning, and acting in the future on the assumption that P is true. There are then two distinct ways in which such commitments can issue in subsequent behavior. One is that when cued into activity in a relevant context, they issue in action directly in the way that intentions usually do. Intending to act on the assumption that P is true, and being reminded of the intention, one does just that. Another is that one might access an episodic memory of having made up one’s mind that P, thereby coming to believe that one believes it. And this might then issue in action motivated by desires for self-consistency. In the first case, the state in question has the wrong direction-of-fit to be a belief, and in the second case it has the wrong content – it isn’t a belief that P, but rather a belief that one believes that P.
The final section of the chapter discusses knowledge. Williamson (2000) famously argues that knowledge is an intrinsically-factive mental state (allegedly like seeing and remembering in this regard), and that it is not a variety of belief. Many others have followed him in making such claims. I argue, in contrast, that the claim not only finds no basis in the science, but that it is actually inconsistent with one of the core commitments of almost all cognitive science. But I should emphasize that what is at stake is not the concept of knowledge (though I have my doubts about claims made by Williamson and others that our concept of knowledge is prior to and independent of our concept of belief, too), but rather the nature of the mental state that the concept picks out.
Cognitive science has no use for a kind of factive state of seeing, nor for a factive state of remembering, either as explanans or explanandum. While it may be a constraint on explanation in both domains that the processes giving rise to visual states and episodic memories are generally reliable ones, cognitive science’s interest is in the processes and underlying mechanisms themselves, whether or not they issue in correct perceptions and memories. There is nothing here to suggest that there might be a natural mental kind constituted by factive perceiving or factive remembering.
Moreover, the claim that knowledge is a real mental-state kind in its own right is actually inconsistent with one of the core commitments of cognitive science, which is to the representational theory of mind. (This had been briefly elaborated and defended in Chapter 1 of the book.) All mental processes of the sort studied in psychology involve representations of various kinds that have correctness-conditions or satisfaction-conditions. The processes themselves involve computations over the representations in ways that generally issue in correctness or satisfaction. But the two components can vary independently in any given case. There is no place here for a type of mental state that already has correctness somehow built into it (as – allegedly – does knowledge). The actual correctness of a representation (as well as its production via a generally reliable process, and together with any other signature properties of knowledge), are extraneous, relational, properties of the representation (the belief, percept, or memory), not intrinsic to it.
The obvious moral: the metaphysics of mind cannot be approached a priori.
Balci, F., Freestone, D., & Gallistel, C.R. (2009). Risk assessment in man and mouse. Proceedings of the National Academy of Sciences, 106, 2459–2463.
Frankish, K. (2004). Mind and Supermind. Cambridge University Press.
Frankish, K. (2007). Deciding to believe again. Mind, 116, 523–547.
Gallistel, C.R., Krishan, M., Liu, Y., Miller, R., & Latham, P. (2014). The perception of probability. Psychological Review, 121, 96–123.
Robins, S. (2022). Episodic memory is not for the future. In A. Sant’Anna, C. McCarroll, & K. Michaelian (eds.), Current Controversies in the Philosophy of Memory. Routledge.
Stephens, D.W. & Krebs, J. (1986). Foraging Theory. Princeton University Press.
Williamson, T. (2000). Knowledge and its Limits. Oxford University Press.