Neurons and Knowledge About the Environment • 67the nature of the environment, and perception by showing that there are neurons that
are tuned to respond best to things that occur regularly in the environment. We can
understand why this is important by considering the problem of designing a machine
that can perceive.DESIGNING A PERCEIVING MACHINE
Imagine that you are given the assignment of designing a computer-based system that
could scan a room and determine its layout. Luckily, you have at your disposal a power-
ful computer, an expert computer programmer, and an array of high-technology sensing
devices.
One approach to this problem would be to have the sensors scan the environ-
ment, determining the patterns of light and dark within a room, and have the com-
puter analyze this information to determine the layout of the room. But since we
know that it helps to have some knowledge of the environment, it would make sense
to design your computer program to be able to recognize elements that frequently
appear inside rooms. One of the fi rst things to do would be to be sure the program
was designed to pick up verticals and horizontals. These are features that are usu-
ally found in rooms; they are associated with the borders between the walls, ceilings,
and the fl oor. It would also make sense to program the computer to be able to sense
fl at surfaces, such as fl oors, ceilings, and walls. In other words, your computer-based
seeing system would operate more effi ciently if it were programmed to be especially
sensitive to features that occur frequently in rooms. This principle for designing a
perceiving machine is the same principle used by the “computer” for the human “per-
ceiving machine”—the brain.THE HUMAN “PERCEIVING MACHINE”
One of the basic operating principles of the human brain is that it contains some
neurons that respond best to things that occur regularly in the environment. When
we described physical regularities in the environment, we mentioned that horizontals
and verticals are common features of the environment, and behavioral experiments
have shown that people are more sensitive to these orientations than to other orienta-
tions that are not as common (the oblique effect, see page 63). It is not a coincidence,
therefore, that when researchers have recorded the activity of single neurons in the
visual cortex of monkeys and ferrets, they have found more neurons that respond best
to horizontals and verticals compared to neurons that respond best to other orienta-
tions, such as slants (Coppola et al., 1998; DeValois et al., 1992). There is evidence
from brain scanning experiments that this occurs in humans as well (Furmanski &
Engel, 2000).
Why are there more neurons that respond to horizontals and verticals? One pos-
sible answer is that through the process of evolution the brain has evolved to respond
best to situations or stimuli that are commonly found in the environment. According
to the theory of natural selection, genetically based characteristics that enhance an
animal’s ability to survive, and therefore reproduce, will be passed on to future gen-
erations. A person whose visual system contains neurons that fi re to important things
in the environment (such as verticals and horizontals, which would occur frequently
in the forest, for example) will be more likely to survive and pass on his or her
characteristics than will a person whose visual system does not contain these special-
ized neurons. Through this evolutionary process, the visual system may have been
shaped to contain neurons that respond to things that are found frequently in the
environment.
Although there is no question that perceptual functioning has been shaped by evo-
lution, it is diffi cult to prove whether a particular capacity is, in fact, “built in” by
evolution or acquired by learning (Kanwisher, 2003). There is, however, a great deal of
evidence that learning can shape the response properties of neurons through a process
called experience-dependent plasticity.Copyright 2011 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part.