PART IV
Naive Realism, Sensation, and Apperception
9
Sensation and Apperception
THE REFLECTIONS about the topic of perception and its relations to conception that follow are the product of more than seventeen years of thinking about John McDowell’s great book Mind and World. If I find myself forced to disagree with him at certain points (Amicus Plato, sed magis amica veritas), that does not alter the fact that his book was a pathbreaking one in the highest sense of the term. For that reason I shall begin the present chapter by reviewing some its arguments, and explain my own view as a correction of his.
John McDowell’s Reasons for Thinking That Experience Is Conceptualized
I once wrote,
What McDowell means by saying that our conceptual powers are “drawn on” in experience, albeit “passively,” is not anything mysterious, nor is this to be construed as psychological speculation of some kind; it is articulated by the work that this idea has to do: to show how experience involves “openness to how things anyway are.”
If we put aside temporarily McDowell’s difficult (and fascinating) discussion of experiences with “inner accusatives” (e.g., experiences of pain, or of after-images) and confine our attention to experiences of how it is “out there,” what McDowell is saying is that such experiences, when they are experiences in what McDowell calls “the demanding sense” (when they function in the justification of belief), are intrinsically about the outer world, and the possibility of having them depends on the possession of the relevant world-involving concepts.
They are not inner signs with a magical connection to the outer world, but takings in of how it is (in the best case), or how it seems to be (in more problematic cases), with the outer world.
I confess that I no longer understand how I could have thought that McDowell’s claim that the possibility of having perceptual experiences “depends on the possession of the relevant … concepts” can be anything but “mysterious.”
However, McDowell means exactly what he says, as is shown by other things he says. For example, in Mind and World, he writes, “No subject could be understood as having experiences of color except against a background of understanding that makes it possible for judgments endorsing such experiences to fit into her view of the world.”
And he makes clear that this is supposed to apply to inner experiences such as the experience of “seeing red” produced by a blow on the head, or even the judgment that I have a pain. According to this view, experiences must be conceptually articulated. What I want to explore is how McDowell arrived at this metaphysical position! What follows is a very brief account.
McDowell Believes the Space of Reasons Cannot Be Reduced to Facts about the Causation of Our Beliefs
McDowell is not satisfied with a merely reliabilist account of justification and other epistemological notions. And I agree with him. Such an account, if offered in a reductionist spirit, presupposes many notions that the reductionist naturalist is not entitled to, for instance, intentional notions such as reference and truth, not to mention the use of counterfactuals, often accompanied by talk of “possible worlds” a là David Lewis, as well as being open to a number of counterexamples.
McDowell must be understood as seeking to produce an account of experience compatible with the idea that perceptual experience justifies beliefs about the layout of the world around one and doesn’t merely causally “trigger” true beliefs a high percentage of the time. This is the first thing one needs to know about Mind and World.
This idea, that perceptual experiences justify accepting and rejecting beliefs about the world and do not merely trigger noises and subvocalizations is what McDowell calls “minimal empiricism.” McDowell clearly identifies this claim with the claim that “impressions” can do this, and I will come back to this identification later.
James Conant has distinguished between two varieties of skepticism, which he calls “Cartesian skepticism” and “Kantian skepticism.” Cartesian skepticism, in Conant’s sense, is skepticism about the possibility of knowledge of things and events “outside” the mind. Kantian skepticism is a puzzle about the very possibility that one’s thoughts, whether seemingly about an external world or even about one’s own sense impressions, can have content at all. Cartesian Skepticism assumes our thoughts about the world are genuine thoughts, that is, that they are true or false, and only worries about whether we can ever really know that any of them are true. Kantian skepticism threatens to undercut even Cartesian skepticism. This poses a problem for empiricism, in that it is not clear how merely saying that our judgments come from experience explains how any of them have content, as opposed to being merely what Rorty called “marks and noises.” “Come from” needs to mean something more than “are caused by,” McDowell tells us, if experience is to be a “tribunal” before which our judgments are to stand. The second thing we need to know to understand McDowell is that he is concerned with Kantian, not Humean, skepticism in Mind and World.
McDowell’s Way into the Problem
McDowell’s way into the problem is via an interesting interpretation of Wittgenstein’s Private Language Argument, which he reads in the light of Wittgenstein’s remarks on the supposition that words acquire meaning by “ostensive definition.”
Wittgenstein’s point can be illustrated thus: Russell (as of 1912) thought that we acquire our basic concepts by “acquaintance by introspection” (in effect, by a sort of inner pointing to our private impressions). But concepts are general, as Russell well knew. So to get the concept red (as applied to sense data), because I obviously can’t direct my attention to all my red sense data, including the future ones, I have to “abstract” the appropriate quality. In other words, I need a private ostensive definition. But all ostensive definition (e.g., holding up a glass and saying “glass” to teach someone the concept) presupposes that the pupil possess at least implicit knowledge of the relevant sortal concept (e.g., “implement” as opposed to “material”). So ostensive definition can’t be the way we acquire all our concepts. That is, pointing to “bare presences” can’t explain how language and minds “hook on to the world.” (In addition to Wittgenstein, McDowell attributes this insight to Sellars and Davidson.)
A way out that McDowell doesn’t discuss would be to say that some concepts are innate. (Think of Quine’s “similarity spaces.”) But innate similarity spaces in the behaviorist sense are just innate patterns of response, dispositions to be caused to respond (say, with the noise “red”), and if that’s all we have, the corresponding judgments are merely conditioned responses to make certain noises, and we lose the fact that our concepts are concepts. And if this behaviorist story isn’t what the nativist has in mind, then it isn’t clear what an “innate concept” is. (Chomsky, today’s best-known nativist, rejects the question of how words refer as too unscientific to discuss.)
I agree with McDowell, Davidson, and Sellars that appealing to “bare presences” can’t provide an answer to the question as to how concepts and experiences are connected, or, in McDowell’s terms, how experiences can rationally constrain beliefs. That confrontation with bare presences can do this is what Sellars meant by “The Myth of the Given,” and what Davidson meant by the idea of “content” in the phrase “dualism of scheme and content,” according to McDowell. (Davidson should have said that what he rejected was the dualism of conceptual scheme and Given, on McDowell’s interpretations.) These are fascinating and plausible interpretations, needless to say.
Davidson
Davidson—who evidently despaired of finding a rational linkage between concepts and experiences (or impressions—we need to keep an eye on McDowell’s identification of experiences with impressions!) held (reviving an idea of Neurath’s) that justification begins with beliefs (which, for Neurath, were just sentences) and not with experiences. But how sentences about experiences can be justified by other sentences is just as problematic! (The coherence of the whole system does the work? How do we know it isn’t “spinning in the void?”) And if observation sentences (Neurath called them “protocols”) aren’t justified by anything, how can they in turn justify anything? Davidson responds to the problem that bothers McDowell by giving up minimal empiricism. But that can’t be right, McDowell feels. Nor can the answer be to just recoil back to the Given. We saw that that didn’t work. So what way out is left?
McDowell’s Way Out
McDowell’s way out is to say that perceptual experiences aren’t just the products of our sense organs. According to McDowell our conceptual capacities are “in play” in experience. Impressions “already possess conceptual content.” Since both beliefs and “impressions” are conceptually articulated, there is supposedly no obstacle to the idea that impressions can rationally justify, and not merely “trigger,” beliefs. Thus, McDowell claims that we can “dismount from the seesaw” that threatens to keep us oscillating between the Myth of the Given and an equally untenable “coherentism.”
It seems to me that McDowell, far from giving a philosophical worry peace by “exorcizing the question,” has worked himself into an unbelievable metaphysical position! What I find unbelievable is not the claim that some of our experiences are conceptualized (in some sense of “conceptualized”), nor the claim that conceptualized experiences are epistemologically fundamental, but the claim that all experiences, indeed all sensations, involve and presuppose our conceptual powers. Surely, the reader is going to want to know, how exactly is it supposed to be the case that my conceptual capacities are “in operation?” A prima facie difficulty is that any given experience has an enormous number of different aspects. Moreover, as McDowell recognizes, “demonstrative expressions” have to be used to describe many of those aspects; it is not the case that one always has the appropriate concept before having an experience that falls under that concept. McDowell (as of Mind and World) was clearly wedded to the idea that one could not have a particular sensation (“impression”) if one did not have the concepts under which that sensation falls. But how could one’s conceptual abilities be “in play” when one has a sensation (say, a particular color-sensation) if one didn’t previously have that concept?
McDowell’s view is clearly in flux here. In Mind and World, the answer to this question was that one forms the demonstrative concept and one has the sensation at the same time. In Having the World in View, however, the view McDowell defends is that to have a sensation it suffices th...