For TUCSON II volume
September 29, 1996
Do you remember Woodstock? Now I know what it was like to be a cop at Woodstock. Even at Woodstock there were good cops and bad cops and I'd like to be a good cop--one of those who went around saying, "Smoke what you like and have a good time and just try not to hurt anybody; it'll all blow over after a while. It may take 20 years, but go and let a thousand desert flowers bloom." I'm happy that this meeting is happening because I think it's a great opportunity to expose to the bright desert air a great variety of ideas, most of which will turn out to be just wrong--but that's all right. That's the way we make progress on this topic.
My own view, as David Chalmers has suggested in his introduction, is skeptical. It's funny: all through my youth I considered myself a sort of radical but now I see that I'm actually a sort of conservative. I say we don't need any scientific revolutions. We have all the materials we need right at hand--just ordinary normal science. The same science that can explain immunology and metabolism and volcanoes can explain consciousness. And what I'm seeing here is what we might call the heartbreak of premature revolution.
Let me give you some of my reasons for thinking that. Consider the cartoonists' convention of the thought balloon or thought bubble. It's a fine metaphorical way of referring to somebody's stream of consciousness. It's vivid, and maybe even accurate and precise, but it's a metaphor for what's actually going on somewhere.--I would say in the brain, but others at this gathering might want to locate it elsewhere. Now the problem of consciousness, as I understand it, is the question: If a thought balloon is the metaphorical truth, what's the literal truth? What is actually happening of which a thought balloon is such a fine metaphorical representation? That is, for me, the problem of consciousness.
The temptation is to view conscious observation as a very special sort of transduction. Let's remind ourselves about "transduction." It's a nice term because it cuts nicely across the artificial and the natural. We have artificial transducers, such as photocells, and natural transducers, such as the rods and cones in your retina. They take information in one medium and, at a boundary surface, transduce it; the same information is sent on in some other physical medium. It might be by turning photons into sound or by turning photons into spikes of electro-chemical activity in a neuron's axon. There's a temptation to think that what consciousness is is a very, very special sort of transduction--the sort that we call observation.
Here we have a conscious observer seeing a red light and letting us know that he's seen it, by saying "Red light!" Some people think that consciousness is a fundamental division in nature, dividing the things that are conscious from the things that are unconscious. The things that are conscious or sentient engage in this very special sort of transduction. Of course, just saying "Red light!" under this condition doesn't guarantee that you're conscious. After all, we might have an experimental subject who just stands there all day saying "Red light, red light, red light, red light." The fact that after you flashed a red light in his eyes such a subject said "Red light!" would be no indication at all that he was conscious of the red light.
The implication is just as dubious if we think about other things that the fellow might do. Consider stepping about the brake pedal. Suppose a person is driving along the highway and a red light comes on. The foot descends on the brake pedal, but does that response give us evidence of the driver's consciousness? Jeffrey Gray was just talking about this sort of issue, in playing tennis. His view is clear: reacting appropriately to such a stimulus wouldn't show consciousness at all. After all, we could readily design a simple AI gadget--it wouldn't even be near the cutting edge of AI--which could respond to a bright red light by pressing on the brake pedal of an automobile. There certainly wouldn't have to be any conscious transduction. There wouldn't have to be any special observation going on in such a case.
That's the way we're inclined to think about this--that consciousness is a very, very special sort of transduction. Consider this diagram from Frisby (1979). It nicely illustrates one of the reasons that people convince themselves that there is some special sort of late transduction event occurring in the brain which just is consciousness.
The light is transduced right at the retinas and then it the information works its way back, via the lateral geniculate nucleus, and by the time the information gets back to the occipital cortex, in the area known as V1 (for visual area number 1), it seems that it's all distorted. Of course if somebody looked at your cortex while you were looking at this woman in the figure, they wouldn't see the colored image depicted, but the patterns of stimulation on the cortex would actually be approximately as shown here--distorted, inverted and twisted in these various ways. A natural reaction when we learn this fact is to say something along the lines of "Well, that's interesting but that's surely not how it seems to me when I look at the woman--so I guess the seeming to me must happen at some later point in the process, where how it seems to me can be restored or put together in some later transduction!"
The transduction at the retina, into neuronal impulses, has taken us, it seems, into an alien medium, not anything we recognize as the intimate medium we are familiar with. That activity in V1 is not in the Medium, you might say. It may be a medium of visual information in my brain, but it's not . . . moi. It's not the medium in which I experience consciousness. So the idea takes root that if the pattern of activity in V1 looks like that (and that's not what consciousness is like), there must be some later second transduction into the medium that is consciousness.
This is David van Essen's famous color coded diagram of the visual areas in the brain of the macaque monkey. The big area on the left is V1. Another famous one is MT, the motion-serving area. V4 is where color is the main topic. Other areas deal primarily with shape and location. All these different aspects of vision are parceled out to specialist areas in the brain. When one looks at a map like this one is inclined to wonder--at least many people are--if there is some area on this map (or some more covert area not yet drawn, deeper in the system) where "it all comes together" for consciousness. That would be where the soundtrack was put together with the color track, which was put together with the shape track, the location track, the motion track, the smell track--everything would be brought together into a single multi-media representation: the representation of one's conscious experience.
If you're tempted to think that way--and if you're not, I think you're a very rare individual--you're making a fundamental mistake. This is the mistake that I call Cartesian Materialism. This is the idea that there is a second transduction, somewhere in the brain (that's why it's a form of materialism). This is the idea that there is a privileged medium in the brain where and when the consciousness happens, where all of the various features of a conscious experience get "bound"--and then, most importantly, appreciated.
This is another of van Essen's diagrams. It shows some of the connectivity between the areas in the previous diagram, and if you look closely you will see that there aren't any arrowheads on the connecting lines. The reason is that they would be redundant, since there would have to be arrow heads on both ends of just about every connecting line. There are at least as many pathways going "down" to V1 and "down" to V2 from "higher" regions as there are coming "up" from the eyes. The very tempting idea that we can move up, up, up from the sense organs at the periphery to "the summit" or in, in, in to "the center" must be abandoned as soon as you get to V1. When you get to V1, you're already "home"! There's no more central headquarters towards which "input" moves in the brain.
There might have been. This is a point I must stress because sometimes people ask me whether my theory is an empirical theory or an a priori theory?. It's both. I grant that it might have been the case that there was a little guy sitting in a little house in the middle of your brain, looking at many screens, listening to many speakers, and pushing many buttons. There might have been what I call a Cartesian Theater in each of our heads! It is an empirical discovery that there is no such thing. If there were, however, then we'd simply have to start our theory all over with him: what happened inside his brain (or whatever occupied that functional role)? The empirical side of my theory, then, is quite uncontroversial--though not trivial, since empirical researchers often stumble over it: there is no Cartesian Theater in the human brain. The conceptual point of my theory is that at some point, you have to get rid of the little guy sitting in the control center, and as soon as you do, you have to change the assumptions of your theory in a rather dramatic way.
Here is the Cartesian Theater (in and old spoof that first asppeared in Life magazine some years ago). We can all laugh at this; we all know this is wrong. The tough question is: What do we replace it with? Exactly what"s wrong with this picture? It's not that the little people are wearing white coats or that they have arms and legs or that there's two of them. What's wrong with the picture is actually much subtler. We already know what's wrong with the picture but we havent come to grips with all the implications:
The work done by the homunculus in the Cartesian Theater must be distributed in both space and time within the brain.
I think there is wide-scale agreement on this. But there's also a lot of backsliding; understanding what all this means is no easy task, and I want to pause and comment first on the word, "work." David Chalmers might say:
Yes, the work done in the Cartesian theater--all that functional work--that's all distributed around in various parts of the brain just as you insist. Getting clear about all that work--those are the Easy Problems. But after we've figured out how all that work is done (and distributed it around in the brain) we won't have touched the interesting, Hard Problem--which is the play, the fun, the qualia, the subjective experience itself (which of course we want to distinguish from all that work.
But I want to point out that the work that must be distributed isn't just pattern recognition and discrimination and locomotion and the like. It's also generating feelings of disgust or delight; it's the appreciation of the scene. Getting the appropriate emotional reactions is just as much part of the work that has to be distributed around in the brain as controlling the stroke of your tennis racket. If we recognize that the work done in the Cartesian Theater has to be distributed in space and time, then we have an answer to Jeffrey Gray's problem about the tennis player. It is simply a mistake to suppose that first there's all this work done in the brain and then comes the consciousness --ohmigosh, too late to matter! (As if consciousness itself were something that happened too late to control your tennis stroke.) Consciousness itself is distributed in time, not just in space, and there is no good reason at all not to say that part of your conscious reaction to the tennis stroke starts within 10, 15, 20, 50, 100 milliseconds of the arrival of the visual information at your cortex. It is a mistake to suppose that you have to wait 100, 200, 500 milliseconds for something--a finished product--to be created in your consciousness, a building process which will take too much time, so that it will be too late for you to do anything about it. We fall in this error if we suppose that there is still a task to be performed in the Cartesian Theater--a task of conscious appreciation and decision in addition to the processes that have already begun to take effect thanks to the distributed activity in the brain.
The task of deciding when and how to hit the tennis ball has to be distributed in time. Don't make the mistake of withholding the honorific of consciousness until all the work is done. (This is a point that has come up in discussion with Benjamin Libet's results: all of his evidence for how long it takes to become conscious of something--what he calls rising time to consciousness--could just as well be taken as evidence for what we might call curing time in memory. There can only be arbitrary grounds for taking some point in that interval of several hundred milliseconds and declaring it to be the onset of consciousness. Consciousness doesn't have to happen at an instant; it is much better to think of it as distributed in both space and time.
This is one of Vesalius's wonderful anatomical drawings of the brain. Right in the middle, marked L, is the pineal gland, or the epiphysis. But for the moment consider it a picture of the globe, and let me use it to tell a rather different story to dramatize the point I want to make. This is the story of the war of 1812 and the battle of New Orleans. The war officially ended on Christmas Eve, 1814, at Ghent (marked G in Vesalius diagram!), when the British and American ambassadors signed a treaty. The news that the war was over traveled out from Ghent in all directions, at various rather slow speeds. I don't know exactly when the news arrived in London, but it was hours, maybe even a day or two, after the signing. It took weeks for the news to get across the Atlantic to New Orleans where in the meantime, British and American troops fought a battle in which several thousand British troops were killed. It was a needless battle in one sense, since the war was officially over, the truce had already been signed. But neither army knew that. It may have been even later when the people in Calcutta or Bombay or Moscow or Hong Kong learned about the signing of the truce.
Now ask yourself a weird question: When did the British Empire learn about the signing of the truce? You may protest that the British Empire isn't the sort of thing that could be said to learn anything. But that is not really true. In a certain important sense the British Empire was a person, a legally responsible agent, a framer and executer of intentional actions, a promise-maker, a contract-signer. The Battle of New Orleans was a legitimate, intended activity of an arm (almost literally) of the British Empire, fighting under the British flag. It was not some sort of renegade action. The signing of the truce was another official action of the British Empire, executed by one of its proper parts.
Now suppose historians could tell us the exact day or hour or minute when every official, proper part of the British Empire learned of the signing of the truce. It is hard to imagine that any further facts could be relevant to the question of when the British Empire "itself" learned. But then we can readily see that we can't date the onset of this recognition, the onset of consciousness by the British Empire, any more finely than by specify a period of about three weeks. When did the British Empire learn? It learned sometime between Christmas Eve 1814 and several weeks into January, 1815. There's just no fact of the matter about when, more precisely, the British Empire learned. Now you might be tempted to disagree, saying that what mattered was when the news reached London (or Whitehall, or the ears of George III). (L'Empire, c'est moi?) After all, wasn't London headquarters of the Empire? In this instance, that's particularly implausible, since when George III learned anything was hardly decisive! (Not so many years ago many of us were going around chanting "What did he know and when did he know it?"--but that was Ronald Reagan, and in that case, too, the question was of merely ceremonial interest. It didn't really matter much when either of these "rulers" learned things.)
The point, then, is this: unless you are prepared to identify some quite specific subregion of the brain as headquarters, the place where moi is, so that entry into that charmed precinct is what counts, you simply cannot, logically, make precise determinations about when consciousness of one item or another happens. (In the discussion of Jeffrey Gray's talk, a number of the questions brought out that his subicular comparator, while perhaps a very important crossroads in the brain, would have to function by sending out appeals to here and there. The system under discussion was not just the hippocampus and the subiculum. It is only in concert with other large and widespread parts of the brain, as Jeffrey eventually acknowledged, that the effects that matter could be produced. It follows from this that there it must be arbitrary to point to any one point in this process and saying that's where the consciousness happens--or more important for present purposes--that's when the consciousness happens. When various vehicles of information arrive at the hippocampus is an interesting fact, but it does not settle anything about the timing of consciousness.)
Here is the moral of my story so far:
Since you are nothing over and above the various subagencies and processes in your nervous system that compose you the following sort of question is always a trap. "Exactly when did I (as opposed to various parts of my brain) become informed, aware, conscious, of some event?"
It's easy enough to see some of the reasons why people are tempted by the hypothesis of a second transduction. In some regards, the computer revolution has made this mistake easier. We now have many systems that are media neutral. Consider the steering system of a modern ship, in which the helm is attached to the distant rudder not by pulleys and wires or cables or chains but just by a couple of little electric wires. Information is all that is carried by these wires: information about the position of the helm, and feedback from the rudder. The wires could readily be replaced by any medium that carried the same information--glass fibres, radio waves, you name it. Such familiar examples give us a clear understanding of how one can have information transmission in an essentially media neutral way between the Boss--the steerer, the governor, the cyberneticker-- and the rudder. This is a valuable and powerful idea. As long as the information is preserved, it doesn't matter what the medium is. But then when we think of information transmission in neurons, well. . . mere nerve impulses, it seems, just can't be the medium of consciousness! It doesn't seem to be enough. Somehow, it seems, we have to put the helmsman back in there. It sems that we have to have a boss or an appreciator, some more central agent to be the audience for all that information. Otherwise the nervous system seems to be a phone system with no subscribers. There's anybody home to answer the phone. There's a television cable network without any viewers. It certainly seems as if we need to posit an audience somewhere to appreciate all that information--to appreciate it in a second "transduction."
The alternative to this bad idea takes some getting used to. It is the idea that the network itself--by virtue of its structure and the powers of transformation that it has, and hence its capacity for controlling the body--could assume all the roles of the inner Boss and thus harbor consciousness. That idea at first seems preposterous to many people, I know. Both David Chalmers and Michael Lockwood remarked in their sessions today that although they acknowledge that there are people who maintain this view, they think it is simply a non-starter. That "the subjective point of view" can somehow be captured in the third person resources of the structure of this functional network strikes them as inconceivable. Not to me it isn't. When people declare to me that they cannot conceive of consciousness as simply the activity of such a functional network, I tell them to try harder.
A common reaction to this suggestion is frank bewilderment, expressed more or less as follows: "OK. Suppose all these strange competitive processes are going on in my brain, and suppose that, as you say, the conscious processes are simply those that win the competitions. How does that make them conscious? What happens next to them that makes it true that I know about them? For after all, it is my consciousness, as I know it from the first-person point of view, that needs explaining!" That question, I think, betrays a deep confusion, it presupposes that what you are is something else, some Cartesian res cogitans in addition to all this brain-and-body activity. But what you are, however, just is this organization of all the competitive activity between a host of competences that your body has developed. You "automatically" know about these things going on in your body, because if you didn't, it wouldn't be your body! The acts and events you can tell us about, and the reasons for them, are yours because you made them--and they made you. What you are is that agent whose life you can tell about. For me, then, the Hard Problem is getting people to see that once you've solved the Easy Problems, you've solved them all--except for my Hard Problem which I'm continuing to work on as you can see!
Frisby, 1979, Seeing: Illusion, Brain and Mind, Oxford: Oxford Univ. Press.
van Essen (to come)