Self-Knowledge for Humans

The phrase “for humans” in Quassim Cassam's title signals a central concern of the book. We are, he emphasizes, Homo sapiens, not members of the fictitious tribe Homo philosophicus that figures in much of the recent philosophical literature on self-knowledge, namely, in the work of those he calls “Rationalists.” Rationalism embraces the “Transparency Method” (TM) to account for our knowledge of our propositional attitudes. He quotes a well-known passage from Gareth Evans about how answering the question whether one believes there will be a third world war requires attending to the evidence we would attend to if asked whether there will be a third world war, and he cites Richard Moran, who quotes the Evans passage with approval, as a prominent advocate of TM. A formulation of TM he quotes from David Finkelstein tells us that “the question of whether I believe that P is, for me, transparent to the question of what I ought rationally to believe—i.e. to the question whether the reasons require me to believe that P. I can answer the former question by answering the latter” (4). The method can be applied to attitudes other than belief, for example, to questions about what one wants or fears.TM will provide a good route to self-knowledge if one is Homo philosophicus—that is, if one's attitudes are always what they ought rationally to be. Cassam discusses at some length the “disparity” between Homo philosophicus and Homo sapiens, the latter being ourselves as we really are. He thinks that we depart in various ways from the requirements for being Homo philosophicus, and that this is not always a bad thing. We are given to “fast thinking” that involves heuristics and shortcuts, and while this sometimes leads us astray, it by and large serves us better than the “slow thinking” that would go with being Homo philosophicus. Here he cites a good deal of psychological literature.Cassam notes in his preface, and elsewhere in the book, that nonphilosophers whose interest is excited by the idea of a philosophical investigation of self-knowledge are disappointed when they learn what most recent philosophical discussion of self-knowledge is really about. In the Rationalist literature, what he calls “substantial” self-knowledge—knowledge of one's character, one's values, one's emotions, and so forth—is mostly ignored, with attention focused instead on mundane or “trivial” self-knowledge, such as (his example) one's knowledge that one believes that one is wearing socks.TM is of little or no help in providing substantial self-knowledge; considering what I ought rationally to value, or what character I ought rationally to have, is not a way of discovering what my values and character actually are. TM also cannot provide self-knowledge of sensations. And in general, it faces a “matching problem”—there is for all kinds of mental states the possibility of a mismatch between one's actual attitude and what one's attitude ought to be, which raises the question of how one can determine what one's attitude is by determining what it ought to be. There is also the “Substitution Problem” for Rationalism; TM represents us as substituting what is often a more difficult question (“Do the reasons require me to believe that P?”) for an easier question (“Do I believe that P?”).Cassam considers various ways of combining an appeal to TM with acceptance of the “disparity,” that is, the fact that we are not Homo philosophicus. One is what he calls “Psychological Rationalism.” This sees only a small gap between Homo philosophicus and Homo sapiens and holds that it is a condition of our having such mental states as beliefs and desires that these are generally what they ought rationally to be. It offers a “transcendental argument” for the “Similarity Thesis,” that is, the view that we are similar enough to Homo philosophicus for Rationalism about self-knowledge to be a viable account of how we know about our beliefs and the like. This invokes an “interpretationalist” account of what it is to have such mental states, an account he finds in Dennett, Davidson, and McDowell. He doesn't outright reject interpretationalism but argues that it is far from giving Rationalism the support it needs. Another response to the disparity is what he calls “Normative Rationalism.” This interprets Rationalism “not as a theory about what human beings are actually like but about how humans ought to be or how they are supposed to be” (75). He distinguishes two different versions of this view but argues that neither provides an account that saves Rationalism.Cassam does not in the end deny that TM has some role to play in accounting for our self-knowledge. But the role it plays is a very modest one. While it is tempting to look for a single source that will account for all of our intentional self-knowledge, in fact “there just is no magic bullet; there are multiple sources of intentional self-knowledge, depending on the type and content of the attitude known. Among the sources of some intentional self-knowledge is TM, but TM only accounts for a narrow range of intentional self-knowledge, and even in the best case the knowledge it accounts for doesn't come out as based on no evidence” (120–21). He says here that the knowledge provided by TM is based on evidence, and he repeatedly says that, contrary to what the proponents of Rationalism suppose, the self-knowledge provided by TM is not immediate and noninferential. A central claim of the book is that all self-knowledge is inferential. In the case of the knowledge provided by TM, the knowledge is inferred from the rationality assumption that takes one from what one's attitude ought to be to what it is.I will come back to this later. But first let's look at what he has to say about the sources of self-knowledge other than TM. One view he discusses and rejects is the “object perception model”—the view that we know our mental states by a kind of perception. One objection to this he considers is the “neo-Humean argument.” This argues that the content of a thought is determined by its causal or relational properties, and that we cannot ascertain an item's causal or relational properties perceptually. Although he expresses some sympathy for this, he rejects it on the grounds that the perception involves inference and that we can know an object's causal or relational properties perceptually. But he says that when you are introspectively aware that your socks are stripy, “this isn't a matter of singling this belief out and distinguishing it from other beliefs you are also introspectively aware of. Propositional attitudes aren't ‘objects’ waiting to be ‘singled out’ on the basis of introspectively available information about their relational and non-relational properties” (131). And it isn't clear what the “intrinsic, nonrelational properties” of a belief might be. He also rejects D. M. Armstrong's version of the object perception model, which he thinks “makes introspection out to be fundamentally no different, epistemologically or phenomenologically, from clairvoyance” (34).Cassam's own account of self-knowledge is inferentialism; such knowledge is always grounded on evidence. He stresses that this is not the view, mistakenly attributed by some to Ryle, that self-knowledge is grounded on behavioral evidence. To some extent, it is so grounded, but what distinguishes our knowledge of our own mental states from our knowledge of the mental states of others is that it is based in part, sometimes entirely, on “internal promptings”—occurrent attitudes and feelings. Our access to the latter is “interpretive”—what enables you to make sense of them is “your background knowledge of your current circumstances, current or recent behavior, and current or recent mental life” (165). He avoids foundationalism and vicious regress by holding that “self-knowledge is holistic rather than linear, and the circularity its holism implies is genuine but not vicious” (169).What Cassam says about the way self-knowledge can be inferential is plausible and well argued. But his application of this to what he calls “trivial” self-knowledge, for example, my knowledge that I believe I am wearing socks, seems to me questionable. As is apparent from “Moore's Paradox,” there is an incoherence involved in combining an affirmation of P with a denial that one believes that P, and this suggests that the belief that P brings with it the belief that one believes that P. Cassam rejects the view that some of the states of which we have self-knowledge are “self-intimating.” Yet it does seem that the belief that one is wearing socks necessarily brings with it the belief that one has that belief. One doesn't infer that one has the latter belief from the fact that one is rational and that one's reasons support that belief. The passage from Gareth Evans quoted by him and Moran needn't be read as claiming that we make such an inference; Evans is merely saying that attending to the outward phenomena that are evidence for or against the claim that there will be a third world war can serve to establish one's belief about the matter, and that this brings with it the belief that one has that belief. The “transparency” just consists in the fact that the second-order belief is implicit in the first-order belief it is about. At one point (160), Cassam supposes that the belief that one believes that one is wearing socks is inferred from the judgment that one is wearing socks. But later (170), he suggests that this cannot be right because knowing that you judge that P is already knowing that you believe that P. I think we are left with no plausible account of how self-knowledge of such beliefs can be inferential.Even if I am right about this, much of what Cassam says in criticism of current views about self-knowledge is interesting and challenging. The book is an important contribution to discussion on this topic

Epistemology · Homo sapiens · Skepticism · Sociology · Tribe · Anthropology · Epistemology, Ethics, and Metaphysics · Philosophical Ethics and Theory · Philosophy · Philosophy and Theoretical Science · Psychology