[Nicholas Humphrey:] What is it like to be ourselves? How can a piece of matter which is a human being be the basis for the experience each one of us recognizes as what it's like to be us? How can a human body and a human brain also be a human mind?
I've put forward several different answers in the last few years, and I'm no longer happy with the earlier ones. I was interested in introspection, and our intuitive knowledge of inner states of mind. I developed a theory of "reflexive consciousness," and I thought, basically, that reflexive consciousness is all that counts. You either have introspective knowledge of your own states of mind or you're not conscious at all. If this were right, consciousness would be a very high-level faculty. It might be something that has evolved only in the great apes and human beings. I suggested that it has evolved specifically to enable people to read their own and other people's minds — and so to become better "natural psychologists." This idea went down well with many people. When I published Consciousness Regained and The Inner Eye, colleagues like Richard Dawkins were full of enthusiasm: "I think Humphrey's got it! At last we have an answer to the big question: how human consciousness evolved."
I thought I'd done a good job on it. But there were problems. One of the consequences of this particular view of consciousness — consciousness identified with introspection and self reflection — was that it meant one had to exclude from the club of conscious beings a whole lot of animals, babies, and other more primitive organisms, which don't have this level of self reflection. The more I tried to persuade myself that, say, a rabbit in pain or a baby crying for its mother can't be conscious because they don't possess the ability to introspect, the more I got dissatisfied. I couldn't sell this idea even to myself, let alone to my nonphilosophical friends.
Surely consciousness can exist at a much lower level, exist unreflected on, just as the experience of raw being: as primitive sensations of light, cold, smell, taste, touch, pain; as the is- ness, the present tense of sensory experience, which doesn't require any further analysis or introspective awareness to be there for us but is just a state of existence. Surely that is what it's like to be me, or what it's like to be a dog, or what it's like to be a baby. That's what it's like to be conscious.
I call it the "thick moment" of consciousness. What matters is that I feel myself alive now, living in the present moment. What matters is at this moment I'm aware of sounds arriving at my ears, sight at my eyes, sensations at my skin. They're defining what it's like to be me. The sensations they arouse have quality. And it's this quality that is the central fact of consciousness.
This is where philosophers stumble. This is where the mind/body problem bites. How is it that anything going on inside a human body or inside a human brain could have a quality? How could the physiological activity underlying sensation have the conscious feel it does, how could it belong to a "sensory modality" — that is, be visual or auditory or tactile or olfactory?
I found myself thinking, Enough of this stuff about higher- level thought processes and the capacity for introspection! Enough, in a way, of all the recent advances in cognitive psychology, and AI! Enough of propositional discourse, and second-order beliefs, and so on! They're interesting problems in their own right, but the problem that people ought to be addressing is the problem of sensation.
At Cambridge, I did a degree in physiology and psychology. I was fortunate enough to have Larry Weiskrantz, an American psychologist, as my Ph.D. supervisor. At the time I was starting on my Ph.D., Weiskrantz had an experiment under way on the effects of visual-cortex lesions in monkeys. He was trying to confirm and extend the findings of the Chicago-based psychologist Heinrich Kluver that destruction of the visual cortex produces almost total blindness.
This wasn't my research project, and perhaps I should have left it alone. But Weiskrantz went away to a conference, and I got to spend some time with two blind monkeys in his lab. I hung out with them day after day, sitting with them, playing with them, interacting with them, trying to work out what was going on. I wanted to know whether they were as blind as people thought they were. It became clear after a couple of days that they were not so blind at all. When I moved my hand in front of their faces, they would follow it with their eyes. And soon enough I was able to persuade them not just to look at where my hand was but to reach out and take a piece of apple from it. The fact that they couldn't do it when the lights were down made it clear they were doing it by vision. When Weiskrantz came back from the conference, I told him I'd taught his blind monkeys to see.
I began working intensively with the monkeys, and within a few months I had brought them to the stage where they could reach out and grasp any small moving object. Weiskrantz and I quickly published a paper in Nature titled "Vision in a Monkey without Striate Cortex." A lot of people were amazed. It went absolutely against the standard view.
I went on working with one of the monkeys — she was called Helen — for seven years. She became a friend, a pet. I'd take her for walks around the countryside. By the end of her training, she'd progressed so far that in many ways she was just like a normal monkey. She could run around a room, avoiding obstacles, searching for nuts or currants on the floor. She had 3-D vision. She could reach out and catch a passing fly. Here was a monkey with no visual cortex, missing the apparatus required to see, and yet she was, in many respects, indistinguishable from a normal seeing monkey.
Weiskrantz made a remarkable new finding. Following up on what I discovered with Helen, he began looking for a similar capacity in human patients with damage to the visual cortex — patients who were supposed to be completely blind in the affected area of the visual field. It turned out that even though they believed themselves to be blind, they were in reality quite capable of using visual information. They could "guess" where a light was located, or what shape an object was, and get it right almost every time. They had a kind of unconscious vision, without any of the usual visual sensations, without anything that told them that light was arriving at their eyes. Weiskrantz called it blindsight. It's become a celebrated phenomenon, widely discussed by philosophers and cognitive scientists.
My own work took a different direction. I'd become uneasy about working with monkeys with brain lesions. Even though it was fascinating and exciting, I didn't want to go on doing experiments that were so hurtful to the animals. I decided to change tack entirely and study monkeys' esthetic preferences. If you give a monkey a choice of environment, what will it prefer? What do they like to look at and listen to? I soon discovered that monkeys have very strong color preferences. They like the blue/green end of the spectrum; they dislike the yellow/orange/red end. Their reactions are intense, much stronger than those of human beings. A red room really upsets them. Blue calms them down.
I'd hoped to find evidence of monkeys' showing preferences for beautiful shapes and forms and sequences, and so on. But there was very little in the way of these more sophisticated preferences: monkeys show no special liking for balance or harmony in visual forms; no interest in Mondrians or Picassos. They don't like any sort of music; in fact they always prefer silence. It's not that they have poor esthetic taste; they seem to have no interesting tastes at all.
The upshot was that I ended up writing theoretical essays about esthetics in humans rather than experimental papers about esthetics in monkeys. I wrote a paper called "The Illusion of Beauty," about the evolution of the sense of beauty. There was a lot of media interest in it. It was broadcast and won the Glaxo Writer's Prize. But I have to say that it was entirely speculative, not based on any solid evidence.
During this time at Cambridge, I ran a lab and I did a lot more experimental work, in areas such as concept formation and time perception. But something changed in me. I began to realize that while I loved doing experiments, I really wanted to get on with theorizing. So I thought I might as well leave the experimenting to other people. There were lots of good experimentalists around.
I made a decision not to go on working in a lab. I gave up my position at Cambridge. I wanted to concentrate on theoretical writing. I wrote Consciousness Regained, which came out in 1983. I got involved in movie stuff. I became distracted. There were different horizons around.
It was in 1987 that I went to work with Dan Dennett. It was a strange move. I had been a director of research at Cambridge; at Dan's Center for Cognitive Studies, at Tufts, I was a research assistant. I didn't feel entirely good about it, but it turned out to be a wonderful time. Over the next couple of years, Dan and I both started working on new books, Dan on Consciousness Explained while I wrote A History of the Mind.
Dan and I thought alike about a lot of things when I first went to work with him. We both had much the same view about what constitutes the central problem of consciousness — the problems of intentionality, self-reflection, and all that. But after spending some time there, I found that interacting with Dan made me realize how much was missing from his picture of the mind — and from mine. Of course, Dan is a much cleverer philosopher than I am, and, as a matter of fact, knows a good deal more cognitive psychology than I do. Maybe I needed to see how my own earlier ideas looked when Dan helped me express them more rigorously, to see what was wrong. What was wrong was that they left out the phenomenology. We were left with consciousness that didn't feel like anything.
I decided that the big problem I had to work on was the nature of conscious sensation. Dan and I used to drive up to his place in Blue Hill, Maine, weekend after weekend, furiously arguing about whether there's such a thing as the sensation of red, or the feeling of pain, or the taste of cheese. Dan would say, "Look, I hear what you're saying, but I simply don't have any reference point for it. Your raw sensations, if they exist, leave nothing behind. They might as well never have occurred." I'd say, "Yes, Dan, I know, but they just are. I'm having them now. I'm living these things." For Dan, if there's nothing left after the sensation has passed — nothing in the way of a text, which says something like, "Memo to self: have just had a sensation" — then it didn't happen.
For Dan, the basic constituents of consciousness are ideas, judgments, propositions, and so on. His problem is to explain how people come to have the particular thoughts they do: how they make decisions, pull things out of memory, construct verbal reports, and so on. To caricature it, his picture of the mind is of a kind of cerebral office, with memos, faxes, and phone calls flying around and competing for the attention of the frantic office staff. The final output of all this information processing is a "conscious text," expressed in words or their equivalent — what Dan calls the heterophenomenological text, corresponding to the stream of consciousness.
But for me, now, the basic constituents of consciousness are raw feelings or sensations. And my problem is to explain how people come to experience these sensations as such: how the "activity of sensing" results in sensations having their qualitative character, immediacy, present-tenseness, sense of belonging to the self, and so forth. My picture of the mind is more that of a cerebral cinema organ, with the organist creating music to match the mood of the film being played at the surface of the body. There doesn't even have to be a final output — at least, certainly not any kind of text — since the experience of consciousness consists essentially in the ongoing activity of playing the organ.
What it boils down to is whether or not you accept an instrumentalist criterion for "meaning": whether you say that only if something is instrumental in producing something else does it have any significance or value. This approach is closely allied to positivism and behaviorism. It also ties in with the Protestant political ethic, where everything is valued only in terms of its effect on the next generation. I suppose this ethic is part of Dan's cultural background. At any rate, it colors Dan's view of the mind. The meaning and value of a mental event consists in what can be made of it later.
Although it's hard to argue against this idea, it seems to belie the reality of our experience, the immediacy and presentness of sensations. For Dan, consciousness doesn't occur until the mind has made up a story and reported back. Consciousness is the story. I say that consciousness is the immediate reaction to the stimulus at the body surface. I make conscious sensations equivalent to an action, to an act of engagement with the stimulus.
The analogy I like to use is from art history. Until the French impressionists came along, most paintings were concerned with how a situation is developing in time: where things have come from and where they're going. It took Monet to value the present moment for itself. To say, "This is Rouen cathedral as I am experiencing it now; this is what hits my face as I look at it." The clock on Rouen cathedral in his paintings doesn't even have a hand on it. There's no time dimension here, no before and after, just a now. Monet grasped this moment, and celebrated it just for what it is, producing a thick painting, full of pigment, to represent a thick moment of his subjective experience, with no antecedents and no consequences. It's the same with the thick moment of sensation, the time we live in. Stand on a street corner in New York and look at the people passing by: the amazing thing is that they're living in the present.
The focus of almost all contemporary research in AI and cognitive science is on explaining thinking rather than feeling. It's been remarkably successful in its own terms. We already have thinking machines. We'll have better ones — fourth-generation, fifth-generation thinking machines. But we're not going to say, "Wow! This is something we could never have imagined." If, however, someone could devise a feeling machine, a machine that had conscious sensations, then we'd say "Wow!" But no one's working on that problem. IBM isn't interested in feeling machines.
Even if we did set out to design feeling machines, we'd probably be unable to design machines that had conscious feelings anything like ours. The reason is that so much depends on the particular biological history, the particular route, by which we got where we are today. Our sensations have what I call skeuomorphic features — features that derive from ancestral ways of doing things and which no longer have any relevance or payoff in today's world but nonetheless supply richness and quality. Think of the analogy with architecture. Many modern buildings contain features that derive from the way Greek or Roman temples were built but have nothing to do with the buildings' functions.
Here's a concise description of my model of reality: I'm me. I'm living an embodied existence, in the thick moment of the conscious present. I'm trying to work out why.
There's a poem I like by e.e. cummings:
since feeling is first who pays any attention to the syntax of things will never wholly kiss you;
Excerpted from The Third Culture: Beyond the Scientific Revolution by John Brockman (Simon & Schuster, 1995) . Copyright © 1995 by John Brockman. All rights reserved.