God Soul Mind Brain

between vision and language, and are also expert at computing mental states? Areas TE and STP (see Diagram 6-3) are prime candidates.

    Social neuroscience began in the 1960s, when Gross and his colleagues studied neurons in area TE of monkeys. (Charlie Gross was my graduate advisor and is my longtime mentor and friend. I cut my teeth scientifically in the lab where social neuroscience began.) Area TE lies at the pinnacle, the highest hierarchical level, in the visual stream that processes object shape and color. Gross and colleagues discovered two especially interesting types of neurons: hand neurons and face neurons.

    A hand neuron fires off signals when the monkey looks at a hand or a picture of a hand. If the monkey looks at a picture of a leg or of an alligator, or if he closes his eyes, that neuron withholds signals. When the monkey opens his eyes and looks at a hand, the neuron begins to deliver signals again. The signals are sent on to other parts of the brain, in a sense telling the rest of the brain, “There’s a hand in view!”

    If the monkey looks at an obvious picture of a hand, then the hand neuron fires off signals at a high rate, shouting, “It’s a hand!” as often as fifty times a second. If the picture is ambiguous, the hand neuron fluctuates in its decision. It shouts out its signal less often. If the monkey is looking at a block of wood instead of a hand, the hand neuron shouts out its signal perhaps once every two or three seconds, hundreds of times less frequently than with an actual hand.

    Face neurons are similarly choosey. They become active when the monkey looks at a face, whether a monkey face, a human face, a profile, a frontal view, an actual person in front of the monkey, a photograph, or a line drawing. A face neuron might become more active in some of those conditions than others—a real face usually drives a bigger response than a mere drawing—but the preference for faces is consistent. Other, non-face images will generally cause only a weak response in the cell.

    Hands and faces appear to be special. There are no banana cells, or bird cells, or tree cells. No other category of object has its own dedicated class of neurons, at least that anyone has found thus far in monkeys. A generic machinery seems to be sufficient for processing and recognizing the vast majority of visual images, whereas hands and faces are evidently so important that they require their own, dedicated population of neurons. (The scientist Kanwisher and her colleagues found the corresponding region in the human brain that emphasizes faces.)

    It is easy to understand why faces require a specialized machinery. Not only are faces extremely important for social communication, but they are also extremely similar to each other. One face is almost identical to the next. Suppose you had a hundred heads of lettuce in front of you, laid out on the ground. Would you be able to recognize them individually so that, at a glance, within a fraction of a second, you would know without doubt that you were looking at Lot 56, #3, the lettuce head, instead of any of the other ninety-nine? If you think heads of lettuce are much too similar for any chance of success, consider just how different they really are in size, shape, rottenness, the exact pattern of the outer leaves—there is more visual variety in a hundred heads of lettuce than in a hundred faces. But we can recognize faces, even thousands of them, with so little effort that we take the skill for granted. We almost certainly owe our astonishing ability to the presence of a dedicated population of face cells whose job it is to perform this task.

    It is less obvious why monkeys and humans need hand cells. One possibility is that hand cells have a similar social importance. We certainly look at each other’s hands in order to judge each other’s gestures and intentions. It may also be that we primates spend so much time manipulating objects, looking at our own hands, that our visual systems have become hand experts simply through overexposure.

Biological motion

    After discovering hand and face cells, Gross and colleagues began to study an area of the cortex next to TE that they named STP. (The name stands for the Superior Temporal Polysensory area, referring to its location in the brain and its properties. It is shown on Diagram 6-3.) Here, they found a bewildering range of cells. Some were face cells, just as in TE. Some were sensitive to motion, especially