Phantoms in the Brain: Probing the Mysteries of the Human Mind (15 page)

So far so good. Dr. Sanders and Dr. Weiskrantz then noticed something very odd. Drew was obviously blind in the left visual field, but if the experimenter put his hand in that region Drew reached out for it accurately!

The two researchers asked Drew to stare straight ahead and put movable markers on the wall to the left of where he was looking, and again he was able to point to the markers, although he insisted that he did not actually "see" them. They held up a stick, in either a vertical or a horizontal position, in his blind field and asked him to guess which way the stick was oriented. Drew had no problem with this task, although he said again that he could not see the stick. After one such long series of "guesses," when he made virtually no errors, he was asked, "Do you know how well you have done?"

"No," he replied, "I didn't—because I couldn't see anything; I couldn't see a darn thing."

"Can you say how you guessed—what it was that allowed you to say whether it was vertical or horizontal?"

"No, I could not because I did not see anything; I just don't know."

Finally, he was asked, "So you really did not know you were getting them right?"

"No," Drew replied, with an air of incredulity.

Dr. Weiskrantz and his colleagues gave this phenomenon an oxymo−ronic name—"Hindsight"—and went on to document it in other patients. The discovery is so surprising, however, that many people still don't accept that the phenomenon is possible.

Dr. Weiskrantz questioned Drew repeatedly about his "vision" in his blind left field, and most of the time Drew said that he saw nothing at all. If pressed, he might occasionally say that he had a "feeling" that a stimulus was approaching or receding or was "smooth" or "jagged." But Drew always stressed that he saw nothing in the sense of "seeing"; that he was typically guessing and that he was at a loss for words to describe any conscious perception. The researchers were convinced that Drew was a reliable and honest reporter, and when test objects fell near the cusp of his good visual field, he always said so promptly.

Without invoking extrasensory perception, how do you account for blindsight—a person's pointing to or correctly guessing the presence of an object that he cannot consciously perceive? Dr. Weiskrantz suggested that the paradox is resolved when you consider the division of labor between the two visual pathways that we considered earlier. In particular, even though Drew had lost his primary visual cortex—rendering him blind—his phylogenetically primitive "orienting" pathway was still intact, and perhaps it mediates blindsight.

In other words, the spot of light in the blind region—even though it fails to activate the newer pathway, which is damaged—gets transmitted through the superior colliculus to higher brain centers such as the parietal lobes, 57

guiding Drew's arm toward the "invisible" spot. This daring interpretation carries with it an extraordinary implication—that only the new pathway is capable of conscious awareness ("I see this"), whereas the old pathway can use visual input for all kinds of behavior, even though the person is completely unaware of what is going on. Does it follow, then, that consciousness is a special property of the evolutionarily more recent visual cortex pathway? If so, why does this pathway have privileged access to the mind? These are questions we'll consider in the last chapter.

What we have considered so far is the simple version of the perception story, but in fact the picture is a bit more complicated. It turns out that information in the "new" pathway—the one containing the primary visual cortex that purportedly leads to conscious experience (and is completely damaged in Drew)—once again diverges into two distinct streams. One is the "where" pathway, which terminates in the parietal lobe (on the sides of your brain above the ears); the other, sometimes called the "what" pathway, goes to the temporal lobe (underlying the temples). And it looks as though each of these two systems is also specialized for a distinct subset of visual functions.

Actually the term "where" pathway is a little misleading because this system specializes in not just

"where"—in assigning spatial location to objects—but in all aspects of spatial vision: the ability of organisms to walk around the world, negotiate uneven terrain and avoid bumping into objects or falling into black pits. It probably enables an animal to determine the direction of a moving target, to judge the distance of approaching or receding objects and to dodge a missile. If you are a primate, it helps you reach out and grab an object with your fingers and thumb. Indeed, the Canadian psychologist Mel Goodale has suggested that this system should really be called the "vision for action pathway" or the "how pathway" since it seems to be mainly concerned with visually guided movements. (From here on I will call it the "how" pathway.) Now you may scratch your head and say, My God, what's left? What remains is your ability to identify the object; hence the second pathway is called the "what" pathway. The fact that the majority of your thirty visual areas are in fact located in this system gives you some idea of its importance. Is this thing you are looking at a fox, a pear or a rose? Is this face an enemy, friend or mate? Is it Drew or Diane? What are the semantic and emotional attributes of this thing? Do I care about it? Am I afraid of it? Three researchers, Ed Rolls, Charlie Gross and David Per−rett, have found that if you put an electrode into a monkey's brain to monitor the activity of cells in this system, there is a particular region where you find so−called face cells—each neuron fires only in response to the photograph of a particular face. Thus one cell may respond to the dominant male in the monkey troop, another to the monkey's mate, another to the surrogate alpha male—that is, to the human experimenter. This does not mean that a single cell is somehow responsible for the complete process of recognizing faces; the recognition probably relies on

a network involving thousands of synapses. Nevertheless, face cells exist as a critical part of the network of cells involved in the recognition of faces and other objects. Once these cells are activated, their message is somehow relayed to higher areas in the temporal lobes concerned with "semantics"—all your memories and knowledge of that person. Where did we meet before? What is his name? When is the last time I saw this person? What was he doing? Added to this, finally, are all the emotions that the person's face evokes.

To illustrate further what these two streams—the what and how pathways—are doing in the brain, I'd like you to consider a thought experiment. In real life, people have strokes, head injuries or other brain accidents and may lose various chunks of the how and what streams. But nature is messy and rarely are losses confined exclusively to one stream and not the other. So let's assume that one day you wake up and your what pathway has been selectively obliterated (perhaps a malicious doctor entered in the night, knocked you out and removed both your temporal lobes). I'd venture to predict that when you woke up the entire world would look like a gallery of abstract sculpture, a Martian art gallery perhaps. No object you looked at would be recognizable or evoke emotions or associations with anything else. You'd "see" these objects, their boundaries and shapes, and you could reach out and grab them, trace them with your finger and catch one if I threw it at 58

you. In other words, your how pathway would be functional. But you'd have no inkling as to what these objects were. It's a moot point as to whether you'd be "conscious" of any of them, for one could argue that the term consciousness doesn't mean anything unless you recognize the emotional significance and semantic associations of what you are looking at.

Two scientists, Heinrich Klüver and Paul Bucy at the University of Chicago, have actually carried out an experiment like this on monkeys by surgically removing their temporal lobes containing the what pathway.

The animals can walk around and avoid bumping into cage walls—because their how pathway is intact—but if they are presented with a lit cigarette or razor blade, they will likely stuff it into their mouths and start chewing. Male monkeys will mount any other animal including chickens, cats or even human experimenters.

They are not hypersexual, just indiscriminate. They have great difficulty in knowing what prey is, what a mate is, what food is and in general what the significance of any object might be.

Are there any human patients who have similar deficits? On rare occasions a person will sustain widespread damage to both temporal lobes

and develop a cluster of symptoms similar to what we now call the Klüver−Bucy syndrome. Like the monkeys, they may put anything and everything into their mouths (much as babies do) and display indiscriminate sexual behavior, such as making lewd overtures to physicians or to patients in adjacent wheelchairs.

Such extremes of behavior have been known for a long time and lend credibility to the idea that there is a clear division of labor between these two systems—and that brings us back to Diane. Though her deficit is not quite so extreme, Diane also had dissociation between her what and how vision systems. She couldn't tell the difference between a horizontal and a vertical pencil or a slit because her what pathway had been selectively obliterated. But since her how pathway was still intact (as indeed was her evolutionarily older "orienting behavior" pathway), she was able to reach out and grab a pencil accurately or rotate a letter by the correct angle to post it into a slot that she could not see.

To make this distinction even more clear, Dr. Milner performed another ingenious experiment. After all, posting letters is a relatively easy, habitual act and he wanted to see how sophisticated the zombie's manipulative abilities really were. Placing two blocks of wood in front of Diane, a large and a small one, Dr.

Milner asked her which was bigger. He found, not surprisingly, that she performed at chance level. But when he asked her to reach out and grab the object, once again her arm went unerringly toward it with thumb and index finger moving apart by the exact distance appropriate for that object. All this was verified by videotaping the approaching arm and conducting a frame−by−frame analysis of the tape. Again, it was as though there were an unconscious "zombie" inside Diane carrying out complicated computations that allowed her to move her hand and fingers correctly, whether she was posting a letter or simply grabbing objects of different sizes. The "zombie" corresponded to the how pathway, which was still largely intact, and the

"person" corresponded to the what pathway, which was badly damaged. Diane can interact with the world spatially, but she is not consciously aware of the shapes, locations and sizes of most objects around her. She now lives in a country home, where she keeps a large herb garden, entertains friends and carries on an active, though protected, life.

But there's another twist to the tale, for even Diane's what pathway was not completely damaged. Although she couldn't recognize the shapes of objects—a line drawing of a banana would not look different from a drawing of a pumpkin—as I noted at the beginning of this chapter, she had no problem distinguishing colors or visual textures. She was

good at "stuff" rather than "things" and knew a banana from a yellow zucchini by their visual textures.The reason for this might be that even within the areas constituting the what pathway, there are finer subdivisions 59

concerned with color, texture and form, and the "color" and "texture" cells might be more resistant to carbon monoxide poisoning than the "form" cells. The evidence for the existence of such cells in the primate brain is still fiercely debated by physiologists, but the highly selective deficits and preserved abilities of Diane give us additional clues that exquisitely specialized regions of this sort do indeed exist in the human brain. If you're looking for evidence of modularity in the brain (and ammunition against the holist view), the visual areas are the best place to look.

Now let's go back to the thought experiment I mentioned earlier and turn it around. What might happen if the evil doctor removed your how pathway (the one that guides your actions) and left your what system intact?

You'd expect to see a person who couldn't get her bearings, who would have great difficulty looking toward objects of interest, reaching out and grabbing things or pointing to interesting targets in her visual field.

Something like this does happen in a curious disorder called Balint's syndrome, in which there is bilateral damage to the parietal lobes. In a kind of tunnel vision, the patient's eyes stay focused on any small object that happens to be in her foveal vision (the high−acuity region of the eye), but she completely ignores all other objects in the vicinity. If you ask her to point to a small target in her visual field, she'll very likely miss the mark by a wide margin—sometimes by a foot or more. But once she captures the target with her two foveas, she can recognize it effortlessly because her intact what pathway is engaged in full gear.

The discovery of multiple visual areas and the division of labor between the two pathways is a landmark achievement in neuroscience, but it barely begins to scratch the surface of the problem of understanding vision. If I toss a red ball at you, several far−flung visual areas in your brain are activated simultaneously, but what you see is a single unified picture of the ball. Does this unification come about because there is some later place in the brain where all this information is put together— what the philosopher Dan Dennett pejoratively calls a "Cartesian theatre"?8 Or are there connections between these areas so that their simultaneous activation leads directly to a sort of synchronized firing pattern that in turn creates perceptual unity? This question, the so−called binding

problem, is one of the many unsolved riddles in neuroscience. Indeed, the problem is so mysterious that there are philosophers who argue it is not even a legitimate scientific question. The problem arises, they argue, from peculiarities in our use of language or from logically flawed assumptions about the visual process.

Despite this reservation, the discovery of the how and what pathways and of multiple visual areas has generated a great deal of excitement, especially among young researchers entering the field.9 It's now possible not only to record the activity of individual cells but also to watch many of these areas light up in the living human brain as a person views a scene—whether it's something simple like a white square on a black background or something more complex like a smiling face. Furthermore, the existence of regions that are highly specialized for a specific task gives us an experimental lever for approaching the question posed at the beginning of this chapter: How does the activity of neurons give rise to perceptual experience? For instance, we now know that cones in the retina first send their outputs to clusters of color−sensitive cells in the primary visual cortex fancifully called blobs and thin stripes (in the adjacent area 18) and from there to V4 (recall the man who mistook his wife for a hat) and that the processing of color becomes increasingly sophisticated as you go along this sequence. Taking advantage of the sequence and of all this detailed anatomical knowledge, we can ask, How does this specific chain of events result in our experience of color? Or, recalling Ingrid, who was motion blind, we can ask, How does the circuitry in the middle temporal area enable us to see motion?