Consciousness

The core of the dispute between functionalists and phenomenolo- gists seems, then, to reside in their views of consciousness. Whether or not there are mental states—like unconscious beliefs—that are not in consciousness, there surely are conscious mental states.

One possibility, which was proposed by the British philosopher Hugh Mellor (who happens to have been one of my own teachers), is to say that conscious states are the states of our own minds about which we currently have beliefs; they are the ones we are currently aware of. So, in particular, a conscious belief that it is raining will be present, on this account, when I believe that I currently believe that it is raining. Let's call a belief about your own current mental state a “second-order” belief. A conscious sensation (of redness, say) will occur when I have the belief that I am currently seeing red.

The functional role of these second-order states will be specified by saying that they are caused by first-order states—like seeing red or believing it's raining—and that they play a role in shaping our behavior, in particular, in relation to ourselves. For one central form of behavior that a belief about something—call it “A”—can produce is behavior aimed at affecting A. So one kind of behavior my beliefs about my own current states is likely to affect is behavior aimed at changing or maintaining my current state.

An obvious example is this.

On the other hand, if I am driving and a traffic light in front of me turns red, I can stop the car, as we say, “automatically”: my belief that the red light is there and my desire to obey the traffic laws can operate directly without my coming to believe I believe anything. So, on this sort of functionalist view, some behavior can occur with­out consciousness.

There is another obvious kind of behavior that will require con­sciousness: telling you what I think or desire. For here, I need to form beliefs about my own mental states and then desire to com­municate what I believe. Indeed, since, as we shall see in the chap­ter on language, communication is a matter of aiming to get people to believe things about your own beliefs, all communication will require second-order beliefs—beliefs about what I currently believe—and so will require consciousness.

The view that both going to find out what the time is and lin­guistic communication require consciousness is, I think, intuitively appealing, as is the view that we sometimes act on our beliefs with­out any conscious mediation. In fact, it seems reasonable to suppose that people can act not just without conscious mediation but when they are not conscious at all.

An account of consciousness of this generally functionalist kind is likely to produce some impatience in the phenomenologist. For the apparatus of second-order states—states that are produced by other current states and that shape the behavior of a system by changing, or maintaining, its own mental states—could obviously be produced in an android: as I have already pointed out, M certainly has the full range of behavior that your mother has, including answering ques­tions and going to see what the time is. Perhaps, the phenomenolo- gist could concede, the functionalists' account of consciousness cap­tures something about consciousness, just as their account of belief—with its role in shaping behavior—captures something about belief. But it leaves out entirely the phenomenological char­acter of consciousness—what it feels like to be your mother or me or anyone else with consciousness. And without that character what you have is just a very good fake.

We seem to have reached an impasse: a situation where argu­ments have run out and there is still no secure conclusion. Faced with an impasse such as this, it is often helpful to ask whether there is some assumption shared by both parties to the debate—what we call a shared presupposition—that needs to be examined. If there are good arguments for both sides and both sides can't be right, maybe it's because they're both wrong in some way we haven't noticed. One shared assumption in the debate so far is an assump­tion about philosophical method. It is that we can discover the essence of the mind or of consciousness by a purely conceptual inquiry. We have been proceeding by making arguments that are based on our understanding of key terms, such as “belief,” “behav­ior,” “feeling.” I have mentioned no experimental explorations of the nature of the mind by psychologists.

For one thing, it might matter whether the thought experiments were about things that could in fact actually happen. It is not at all obvious, for example, that there could in fact be a creature like M. (Perhaps the only sort of thing that could exactly reproduce your mother's behavior would have to be made pretty much, molecule for molecule, as your mother is. And then most of us would probably suppose that there was something that it was like to be her, so that she would meet both the functionalist and the phenomenological criteria for being mentally the same as your mother.) What signifi­cance should we attach to our response to being told that something might happen, when, in fact, it can't happen? Why should we assume, that is, that ways of thought that work well enough in a rough-and-ready way in ordinary life would work just as well in a very different world?

Another, more fundamental line of objection would be to ask why we take it for granted that we have such internal states as beliefs and sensations at all. We are normally inclined to take it as obvious that someone has beliefs when they act, or sensations when they open their eyes on a lighted world. But the fact that this is part of the package of regular commonsense assumptions doesn't guarantee that we are right. People used to think it was obvious that some peo­ple were witches and that there were ghosts. (As a matter of fact, as we shall see in the final chapter, there are still places where most people think something similar.) Perhaps the very fact that our ordi­nary ways of thinking can lead both to functionalism and to phe­nomenology suggests that those ways of thinking are muddled.

The contemporary American philosopher Stephen Stich has sug­gested that we may indeed have to do just this. He has examined a good deal of recent work in cognitive psychology, the branch of the subject that seeks to explain how we perceive, remember, rea­son, decide, and then act, by postulating internal processes very like those in a computer program. Stich argues that there is already a good deal of evidence from cognitive psychology that our folk psy­chological theory is just plain wrong. In fact, he thinks, it may even­tually turn out that there is simply nothing at all inside our heads that operates in the way that our folk psychology of belief and desire supposes. If that is true, then there would be no beliefs or desires! And then we should have to proceed, guided by cognitive psychol­ogy or neuroscience (or perhaps some new field of science), to try to understand the causes of behavior in terms of internal states quite unlike those we have gotten used to. That is why the subtitle of his book From Folk Psychology to Cognitive Science is The Case Against Belief.

One natural response to this possibility is to say that even if sci­ence does end up showing this, we would still want to continue with our folk psychological theory for everyday purposes. We would still, that is, want to treat other people as if they had beliefs and desires and the rest, even if our official position was that they didn't. Another American philosopher, Daniel Dennett, has given this strategy a name: he calls it “adopting the intentional stance” toward them. We adopt the intentional stance toward someone (or something) when we predict its behavior on the basis of what it would do if it had beliefs, desires, and intentions, while leaving open the possibility that it does not, in fact, have them.

Stich argues that Dennett's proposal is intellectually irresponsi­ble. What's the point of explaining the way people behave in terms of states they haven't got, once you develop a theory that explains how they behave in terms of states they have? But to this objection one might reply that there may be practical reasons why it is easier to use the folk psychological theory. Perhaps, for example, we are attached to this theory because it is programmed into us by evolu­tion, so that, just as certain visual illusions persist, even once we know they are illusions, we will continue to think spontaneously of people as having beliefs, even once we realize they do not. Or per­haps the states that the new cognitive psychological theory postu­lates are rather difficult to identify, so that only a psychologist with special instrumentation can find out exactly what they are. (There is something odd about discussing what we should believe if there aren't really any beliefs!) The rough-and-ready apparatus of folk psychology at least has the advantage that we can apply it pretty eas­ily on the basis of looking and listening without special equipment.

But there is a natural response to both Stich's proposal and Dennett's, a response that challenges a presupposition they seem to share. It is that both of them ignore the fact from which the phe- nomenologist starts: the fact that each of us knows very well in our own case that we have beliefs, desires, sensations, and so on. In response to Stich, one wants to say:

I grant that I might be wrong about how my mental states work, and about their causal relations. But I can't be wrong about whether I have mental states. They are, as Descartes rightly insisted, the one thing in the world I am most certain of. By “belief” I just mean something like the state I am in when I look at a vase and come to believe that it has a flower in it.

And to Dennett one might say:

I can imagine taking the intentional stance toward somebody else, exactly because I can imagine that someone else doesn't really have beliefs and desires but only appears to do so. That is just the problem of other minds. But it's a problem of other minds; just because I have direct experience of my own internal states, I can't imagine taking the intentional stance toward myself.

1.12