The Tell-Tale Brain: A Neuroscientist's Quest for What Makes Us Human

Perhaps we can test the idea of isolated brain regions more directly. One approach would be to use functional brain imaging such as fMRI, which you may recall measures magnetic fields in the brain produced by changes in blood flow while the subject is doing something or looking at something. My ideas about isolation, along with Allan Snyder’s ideas, predict that, when you look at cartoon sketches or doodles of faces, you should get a higher activation of the face area than of areas dealing with color, topography, or depth. Alternatively, when you look at a color photo of a face, you should see the opposite: a decrement in the relative response to the face. This experiment has not been done.

Peekaboo, or Perceptual Problem Solving

The next aesthetic law superficially resembles isolation but is really quite different. It’s the fact that you can sometimes make something more attractive by making it less visible. I call it the “peekaboo principle.” For example, a picture of a nude woman seen behind a shower curtain or wearing diaphanous, skimpy clothes—an image that men would say approvingly “leaves something to the imagination”—can be much more alluring than a pinup of the same nude woman. Similarly, disheveled tresses that conceal half a face can be enchanting. But why is this so?

After all, if I am correct in saying that art involves hyperactivation of visual and emotional areas, a fully visible naked woman should be more attractive. If you are a heterosexual man, you would expect an unimpeded view of her breasts and genitalia to excite your visual centers more effectively than her partially concealed private parts. Yet often the opposite is true. Similarly, many women will find images of hot and sexy but partially clad men to be more attractive than fully naked men.

We prefer this sort of concealment because we are hardwired to love solving puzzles, and perception is more like puzzle solving than most people realize. Remember the Dalmation dog? Whenever we successfully solve a puzzle, we get rewarded with a zap of pleasure that is not all that different from the “Aha!” of solving a crossword puzzle or scientific problem. The act of searching for a solution to a problem—whether purely intellectual, like a crossword or logic puzzle, or purely visual, like “Where’s Waldo?”—is pleasing even before the solution is found. It’s fortunate that your brain’s visual centers are wired up to your limbic reward mechanisms. Otherwise, when you try to figure out how to convince the girl you like to sneak off into the bushes with you (working out a social puzzle) or chase that elusive prey or mate through the underbrush in dense fog (solving a fast-changing series of sensorimotor puzzles), you might give up too easily!

So, you like partial concealment and you like solving puzzles. To understand the peekaboo law you need to know more about vision. When you look at a simple visual scene, your brain is constantly resolving ambiguities, testing hypotheses, searching for patterns, and comparing current information with memories and expectations.

One naïve view of vision, perpetuated mainly by computer scientists, is that it involves a serial hierarchical processing of the image. Raw data comes in as picture elements, or pixels, in the retina and gets handed up through a succession of visual areas, like a bucket brigade, undergoing more and more sophisticated analysis at each stage, culminating in the eventual recognition of the object. This model of vision ignores the massive feedback projections that each higher visual area sends back to lower areas. These back projections are so massive that it’s misleading to speak of a hierarchy. My hunch is that at each stage in processing, a partial hypothesis, or best-fit guess, is generated about the incoming data and then sent back to lower areas to impose a small bias on subsequent processing. Several such best fits may compete for dominance, but eventually, through such bootstrapping, or successive iterations, the final perceptual solution emerges. It’s as though vision works top down rather than bottom up.

Indeed, the line between perceiving and hallucinating is not as crisp as we like to think. In a sense, when we look at the world, we are hallucinating all the time. One could almost regard perception as the act of choosing the one hallucination that best fits the incoming data, which is often fragmentary and fleeting. Both hallucinations and real perceptions emerge from the same set of processes. The crucial difference is that when we are perceiving, the stability of external objects and events helps anchor them. When we hallucinate, as when we dream or float in a sensory deprivation tank, objects and events wander off in any direction.