Life's Potentials
89
One of the mathematicians, computer pioneer John von Neumann,
later elaborated the concept in great detail, but basically it's rather sim-
ple. In analog computers, changes in information are expressed by analo-
gous changes in the magnitude or polarity of a current. For example, if
the computer is to use and store the varying temperatures of a furnace,
the rise and fall in heat can be mimicked by a rise and fall in voltage.
Analog systems are slow and can handle only simple information, but
they can express subtle variations very well. Digital coding, on the other
hand, can transmit enormous amounts of data at high speed, but only if
the information can be reduced to yes-no, on-off bits—the digits 1 and
0. If the brain was such a hybrid computer, these early cyberneticists
reasoned, then analog coding could control the overall activity of large
groups of neurons by such actions as increasing or decreasing their sen-
sitivity to incoming messages. (A few years later neurologists did find
that some neurons were "tuned" to fire only if they received a certain
number of impulses.) The digital system would transfer sensory and
motor information, but the processing of that information—memory
and recall, thought, and so on—would be accomplished by the syn-
ergism of both methods. The voltage changes Burge found in response
to major alterations of consciousness seemed to fit within this frame-
work, and his observations were extended by the Harvard-MIT group
and others. Much of this work was done directly on the exposed brains
of animals and of human patients during surgery. When cooperative
patients elected to remain awake during such operations (the brain is
immune to pain), human sensations could often be correlated with elec-
trical data. Contributors to this endeavor included nearly all of the
greatest American neurophysiologists—Walter B. Cannon, Arturo Ro-
senblueth, Ralph Gerard, Gilbert Ling, Wilder Penfield, and others.
Measurements on the exposed brain quickly confirmed the existence of
potential voltages and also revealed possible currents of injury. When-
ever groups of nerve cells were actively conducting impulses, they also
produced a negative potential. Positive potentials appeared from injured
cells when the brain had been damaged; these potentials then expanded
outward to uninjured cells, suppressing their ability to send or receive
impulses. When experimenters applied small negative voltages to groups
of neurons, their sensitivity increased; that is, they would generate an
impulse in response to a weaker stimulus. Externally applied positive
potentials worked in the opposite way: They depressed nerve function,
making it harder to produce an impulse. Thus there did seem to be an
analog code, but how did it work? Did the potentials come from direct
currents generated by the nerve cells themselves, or did they merely