Pretend you are a member of an audience watching several people dribbling and passing a basketball among themselves. Your job is to count the number of times each player makes a pass to another person during a 60-second period. You find you need to concentrate, because the ball is flying so quickly. Then, someone dressed in a gorilla suit ambles across the floor (right). He walks through the players, turns to face the viewers, thumps his chest and leaves. Astonishingly, as Daniel J. Simons, now at the University of Illinois, and Christopher F. Chabris of Harvard University learned when they conducted this study, 50 percent of people fail to notice the gorilla.
We think of our eyes as video cameras that make a flawless recording of the world around us, but this demonstration shows how little information we actually take in at a glance.
The gorilla experiment is the culmination of a long line of related studies on attention and vision that began more than three decades ago by, among many researchers, Ulric Neisser of Cornell University, Ronald A. Rensink of the University of British Columbia, Anne Treisman of Princeton University, Harold Pashler of the University of California at San Diego and Donald M. MacKay of Keele University in England.
Researchers refer to the gorilla effect as "inattentional blindness" or "change blindness," which in turn is part of a more general principle at work in our visual system. Our brain is constantly trying to construct meaningful narratives from what we see. Things that do not quite fit the script or that are not relevant to a particular task occupying our interest are wiped wholesale from consciousness. (Whether such deleted information is nonetheless processed unconsciously has yet to be investigated.) A simple example of how the brain's running narrative can interfere with perception is the children's game "spot the difference" (below). The two images are similar enough that the brain assumes they must be identical; it takes minutes of careful inspection to locate the disparities.
The value of having an underlying brain "story" becomes clear when you consider how jumbled sensory inputs can be. As you survey the room around you, the image on your retina is jumping rapidly as various parts of the scene excite different bits of retina. Yet the world appears stable. Researchers once believed that the experience of having an unbroken view was entirely created by the brain sending a copy of the eye movement command signals originating in the frontal lobes to the visual centers. The visual areas were thought to be "tipped off" ahead of time that the jumping image on the retina was caused by eyes moving and not by the world moving.
But an effect you can demonstrate for yourself at home shows that this cannot be the entire reason. (Jonathan Miller, an opera director in London, and one of us [Ramachandran] independently observed the effect in the early 1990s.) Turn a television set upside down. Gently! Better yet, flip the TV's image optically with a prism. Alternatively, you can turn the TV sound off and then stand slightly to the side of the set, looking at the screen with your peripheral vision. Put the TV on any channel and watch what happens. You will see sudden, jarring changes and visual jolts. Next, gaze at the broadcast with the TV right side up, viewing it straight on and with the sound at normal volume. Now the cuts and pans of the camera flow smoothly and seamlessly into one another--in fact, you do not even notice them. Even when the scene switches, say, from one talking head to the other as they alternate in conversation, you do not see a head transforming or morphing from one to the other as your mind alternates between each of the two speakers. Instead you experience your vantage point shifting.
Have you noticed any gorillas walking by? How can you be sure that none did?