From an understanding of this first step on the sensory pathway, researchers have edged up to analyzing how messages about a sensory stimulus travel through the brain to the cerebral cortex and how these messages are coded.
They know that nearly all sensory signals go first to a relay station in the thalamus, a central structure in the brain (named after the Greek word for "couch" because the cerebral hemispheres seem to rest comfortably on it). The messages then travel to primary sensory areas in the cortex (a different area for each sense). There they are modified and sent on to "higher" regions of the brain. Somewhere along the way, the brain figures out what the messages mean.
Many factors enter into this interpretation, including what signals are coming in from other parts of the brain, prior learning, overall goals, and general state of arousal.
Going in the opposite direction, signals from a sensory area may help other parts of the brain maintain arousal, form an image of where the body is in space, or regulate movement.
These interactions are so complex that focusing on the activity of single neuronsor even single pathwaysis clearly not enough. Researchers are now asking what the central nervous system does with all the information it gets from its various pathways.
In more authoritarian times, scientists believed that the brain had a strictly hierarchical organization. Each relay station was supposed to send increasingly complex information to a higher level until it reached the very top, where everything would somehow be put together. But now "we are witnessing a paradigm shift," says Terrence Sejnowski, an HHMI investigator who directs the Computational Neurobiology Laboratory at the Salk Institute in La Jolla, California.
Instead of viewing the cortex as a rigid machine, scientists see it as "a dynamic pattern-processor and categorizer" that recognizes which categories go together with a particular stimulus, as best it can, every step of the way, according to Sejnowski. "There is no 'grandmother cell' at the top that responds specifically to an image of Grandma," he emphasizes. "We recognize a face by how its features are put together in relation to one another."
Sejnowski, a leader in the new field of computational neuroscience, studies neural networks in which the interaction of many neurons produces surprisingly complex behavior. He recently designed a computer model of how such a network might learn to "see" the three-dimensional shape of objects just from their shading, without any other information about where the light came from. After being "trained" by being shown many examples of shaded shapes, the network made its own generalizations and found a way to determine the objects' curvature.
Vision and the other senses evolved "to help animals solve vital problemsfor example, knowing where to flee," says Sejnowski. Large populations of sensory neurons shift and work together in the brain to make this possible. They enable us to see the world in a unified way. They link up with the motor systems that control our actions.
These neurons produce an output "that is more than the sum of its parts," Sejnowski says. Just how they do it is a question for the next century.
< Previous | Top of page
Each of the five senses activates a separate area of the cerebral cortex, the sheet of neurons that makes up the outer layer of the brain's hemispheres. This brain is a computer reconstruction...