What Mad Pursuit (21 page)

I soon found out that what I had learned amounted to very little. Apart from the fact that a lot of work had been done in neuroanatomy and neurophysiology since I had first glanced at these subjects, there were whole areas, such as psychophysics, of which I knew absolutely nothing. (Psychophysics is not some new California religion. It is an old term for that branch of psychology that deals with
measuring
the response of a person or animal to physical inputs, such as light, sound, touch, etc.).

Moreover, I found there was a new subject that called itself cognitive science. (It has been said, somewhat unkindly, that any subject that has “science” in its name is unlikely to be one.) Cognitive science was part of the rebellion against behaviorism. Behaviorists thought that one should study only the behavior of an animal and should not try to take account of, or make models of, any postulated
mental
processes inside the animal. Behaviorism became the dominant school in psychology in the earlier part of this century, especially in America.

Cognitive scientists, in opposition to the narrow views of behaviorists, think it important to make explicit models of mental processes, especially those of humans. Modern linguistics is an important part of cognitive science, since it does just that. There is no great enthusiasm, however, for looking into the actual brain itself. Many cognitive scientists tend to regard the brain as a “black box,” better left unopened. In fact, some people
define
cognitive science as studies that take no account of such things as nerve cells. In cognitive science the usual procedure is to isolate some psychological phenomenon, make a theoretical model of the postulated mental processes, and then test the model, by computer simulation, to make sure it works as its author thought it would. If it fits at least some of the psychological facts it is then thought to be a useful model. The fact that it is rather unlikely to be the correct one seems to disturb nobody.

I found all this most peculiar and still do. Basically it is the philosophical attitude of a functionalist, a person who believes that study of the functioning of a person or animal is all important and that it can be studied, by itself, in an abstract way without bothering about what sort of bits and pieces actually implement the functions under study. Such an attitude, I found, is widespread among psychologists. Some even go so far as to deny that knowing exactly what goes on inside the head would ever tell us anything useful at all about psychology. They are apt to bang their fists on the table in support of such statements.

When pressed as to why they think in this way, they usually say that the whole bag of tricks is so fiendishly complicated that no good is likely to come from looking at it closely. The obvious answer to this is that if indeed it is so complicated, how do they ever hope to unscramble the way it operates by looking solely at its input and output, ignoring what goes on between? The only reply I have ever had to such a question is that it is essential to study organisms at higher levels and that the study of neurons,
by itself
(my italics), will never solve such problems. With this I entirely agree, but I cannot see that it justifies ignoring neurons altogether. It is not usually advantageous to have one hand tied behind one’s back when tackling a very difficult job.

My own prejudices are exactly the opposite of the functionalists’: “If you want to understand function, study structure,” I was supposed to have said in my molecular biology days. (I believe I was sailing at the time.) I think one should approach these problems at
all
levels, as was done in molecular biology. Classical genetics is, after all, a black-box subject. The important thing was to combine it with biochemistry. In nature hybrid species are usually sterile, but in science the reverse is often true. Hybrid subjects are often astonishingly fertile, whereas if a scientific discipline remains too pure it usually wilts.

In studying a complicated system, it is true that one cannot even see what the problems are unless one studies the higher levels of the system, but the
proof
of any theory about higher levels usually needs detailed data from lower levels if it is to be established beyond reasonable doubt. Moreover, exploratory data from the study of lower levels often suggests important ways of constructing new higher-level theories. In addition, useful information about lower-level components can often be obtained from studying them in simpler animals, which may be easier to work on. An example would be recent work on the mechanism of memory in invertebrates.

My first problem was to decide what sort of animal to concentrate on. Some of my fellow molecular biologists had opted for small, rather primitive animals. As mentioned, Sydney Brenner had selected a nematode. Seymour Benzer had chosen to study the behavioral genetics of the little fruit fly,
Drosophila
, partly because so much basic genetics had already been done on it.

I decided that my main
long-term
interest was in the problem of consciousness, though I realized that it would be foolish to start with this. Consciousness is most apparent in man—at least I know I am conscious and I have good reasons to suspect that you are too. Whether a fruit fly is conscious is an open question. There are, however, grave experimental handicaps to working on human beings, since so many experiments are impossible for ethical reasons. It seemed reasonable, therefore, to concentrate on animals close to man in evolution; that is, the mammals and in particular the primates—the monkeys and apes.

My next problem was to choose some particular aspect of the mammalian brain. As I knew very little I decided to make the obvious choice and concentrate on the visual system. Man is a very visual animal (as are monkeys), and much work had already been done on many aspects of vision.

How can one study vision in man by working on monkeys? The obvious approach is to do what one can on man, and, in parallel, study the same system in a monkey or other mammal. In work on perception, it is now becoming standard practice to use arguments from detailed psychophysical studies on man (plus rather cruder psychophysical studies on a monkey) combined with all the neuroanatomical and neurophysiological knowledge available on the relevant part of a monkey’s brain. Occasionally other data from man can be used, such as evoked potentials (a type of brain wave), or various rather expensive scans, but as yet these have a much lower resolution, either in space and/or time, and thus usually give us much less information.

This is why, to someone like myself, the visual system is attractive since, as far as we can tell, a macaque monkey sees in much the same way as we do. There are, of course, few subjects more important to us than language, since it is one of the main differences between man and all lower animals. Unfortunately, for this very reason, there is no suitable animal for such studies. This is why I believe that modern linguistics, sophisticated though it is, will run into a brick wall unless much more can be found out about what happens inside our heads when we talk, listen to speech, and read. If language is anything like as complex as vision (which seems more than likely), the chance of unscrambling the way it really works without this extra knowledge seems to me to be rather small. Linguists, not surprisingly, usually find this argument unacceptable.

I also decided that, at least at first, I would not attempt to do experiments. Apart from the fact that, technically, they are often very difficult, I thought I could contribute more from a theoretical viewpoint. It seemed to me that I might perform a useful function by studying the problem of vision from as many points of view as possible. I hoped that I might help to build bridges between the various scientific disciplines, all of which studied the brain from one point of view or another. I had rather little expectation of producing any radically new theoretical ideas at such an advanced age, but I thought I might interact fruitfully with younger scientists. In any case I expected that the subject would prove endlessly interesting and that at my time of life I had a right to do things for my own amusement, provided I could make an occasional useful contribution.

Having decided that I could learn about the mammalian visual system, my next problem was to select which aspect to study first. I had never had a medical education, so my knowledge of neuroanatomy was almost zero. I decided to tackle that first, as I expected it to be the dullest part of the subject. It would be as well, I thought, to get it out of the way before going on to other, more interesting, topics.

To my surprise I soon discovered that there had been a minor revolution in dry-as-dust neuroanatomy. Thanks mainly to the introduction of various rather simple biochemical techniques, it was now possible to discover how the various regions of the brain were connected together. Moreover, the techniques were not only powerful but considerably more reliable than most of the older methods. Unfortunately most of them cannot be used on humans (one cannot, at the end of an experiment, “sacrifice” the graduate student who has been acting as the subject, as one can do with animals, for obvious ethical reasons). We thus have the curious situation that more is known about neural connections in the brain of the macaque monkey than about those in the human brain. In fact, we shall soon know so much about the broad pattern of connections in the macaque, and about the location in the brain of various chemical transmitters and the receptors for them, that the only way to cope with all this new information will be to store it in computers, in such a way that it can be displayed in some vivid graphic form for easy comprehension.

I first started by reading experimental papers and reviews. I found it was not difficult to approach experimentalists provided one was genuinely interested in what they were doing and had first made some effort to discover from their publications what they were up to. In this way I made many new friends, far too many to list here. I was lucky in finding in La Jolla several people interested in vision or in theory. A group at the Psychology Department at the University of California, San Diego (UCSD), under the leadership of Bob Boyton, studied mainly the psychophysics of vision. Other psychophysicists I got to know were Don MacLeod and V. S. (“Rama”) Ramachandran when he came to San Diego from Irvine. I also interacted with another group in the same department, led then by David Rumelhart and Jay McClelland, that did theoretical work. After a while the department appointed me an adjunct professor of psychology, in spite of my very flimsy knowledge of the subject.

In 1980 Max Cowan came to the Salk, setting up a large group of neuroscientists there. Some of these people, such as Richard Andersen (now at M.I.T.) and Simon LeVay do experimental work on the visual system. Although Max left in 1986, the Salk still has a strong interest in neuroscience and has recently recruited Tom Albright, an experimentalist from Princeton.

Another blessing was the arrival, in 1984, of the Canadian philosophers Paul and Pat Churchland, to take up chairs in the philosophy department at UCSD. It is unusual to find philosophers who are even remotely concerned about the brain, so it is a great help to have the advice of two people who do take a keen interest in it. Both had written very well on reductionism (a dirty word to some, especially to those who regard me as an archreductionist). More recently Pat has written a large book, called
Neurophilosophy
, published by the Bradford Book section of the M.I.T. Press, setting out the philosophical, theoretical, and experimental aspects of their new point of view. Its subtitle is “Towards a Unified Science of the Mind-Brain.”

Ramachandran and Gordon Shaw (a physicist at U.C. Irvine) were the co-founders of the Helmholtz Club, named after the nineteenth-century German physicist who pioneered the scientific study of perception. The members meet about once a month, starting with lunch and ending with dinner. In between we have talks by two speakers, on topics mostly connected with the visual system. This schedule allows plenty of time for discussion. The meetings are held at Irvine, which is midway between Los Angeles and San Diego, so that members and guests from the other university campuses can attend.

This is not the place for me to attempt to outline what we now know about the visual system—that would take at least another whole book—let alone the rest of the brain. I will restrict myself to rather general comments. In the first place, it is not obvious to most people why we need to study vision. Since we see so clearly, without any apparent effort, what is the problem? It comes as a surprise to learn that in order to construct our vivid mental representation of the outside world, the brain has to engage in many complex activities (sometimes called computations) of which one is almost completely unaware.

We succumb all too easily to the Fallacy of the Homunculus—that somewhere attached to our brain there is a little man who is watching everything that is going on. Most neuroscientists don’t believe this (Sir John Eccles is an exception) and think that our picture of the world and of ourselves is due solely to neurons firing and other chemical or electrochemical processes inside one’s body. Exactly how these activities give us our vivid picture of the world and of ourselves and also allow us to act is what we want to discover.

The main function of the visual system is to build a representation inside our head of objects in the world outside us. It has to do this from the complex signals reaching the retinas of our eyes. Though these signals have much information implicit in them, the brain needs to process this information to obtain
explicit
representations of what interests it. Thus the photoreceptors in our eyes respond to the wavelength of the impinging light coming from an object. But what the brain is mainly interested in is the
reflectivity
(the color) of an object, and it can extract this information even under quite different conditions of illumination of that object.