Belmont Greenlee
Farley, a pioneering brain researcher and computer scientist who
helped develop the world’s first fully-transistorized computer
and, with his colleague, W.A. Clark, created the first computer
simulation of a neural network, died on February 28, at
Cathedral Village, a retirement home in Philadelphia, P.A. He
was 87.
His wife, Elizabeth
B. Farley, confirmed his death, due to complications from
Alzheimer’s.
Starting in 1954 at
M.I.T.’s Lincoln Laboratory, Dr. Farley and Dr. Clark conducted
research on neural networks using mathematical models developed
by Alan Turing. According to the Encyclopaedia Britannica,
“They succeeded in running the first artificial neural
network—albeit limited by computer memory to no more than 128
neurons. They were able to train their networks to recognize
simple patterns. In addition, they discovered that the random
destruction of 10 percent of the neurons did not affect the
network's performance—a feature that is reminiscent of the
brain's ability to tolerate limited damage.”
Dr. Farley was born
on December 29, 1920 in Cape Girardeau, M.O. His father, Dr.
Belmont M. Farley, a schoolteacher and principal, later worked
as the director of public relations for the National Education
Association in Washington, D.C. His mother, Hazel R. Greenlee,
was a schoolteacher.
Dr. Farley entered
the University of Maryland at the age of 16 and obtained his
B.A. in mathematics in 1941. His subsequent doctoral studies in
mathematics at M.I.T. were interrupted by World War II. He then
joined a team at M.I.T.’s Radiation Laboratory that developed an
improved version of radar to detect objects at lower altitudes
than previously possible. In 1943, the team flew to England to
field-test this improved radar and assist the R.A.F. in its
battle against the Luftwaffe. He vividly remembered watching
the clustered blips on radar oscilloscopes as German planes
swooped across the English Channel.
He fondly recalled the lighter side of
wartime England, too. He and colleagues would sometimes scour
the English countryside for fresh, black market eggs when they
grew sick of powdered eggs and spam. Sometimes, the English
tired of their American allies. Dr. Farley would laugh as he
told of hearing British colleagues complain, “You Americans are
oversexed, overpaid and over here!”
After the war, at
the suggestion of Ernest C. Pollard, a biophysics pioneer, Dr.
Farley switched to physics and obtained his Ph.D. at Yale
University in 1948. He started his professional career at Bell
Telephone Laboratories from 1948 to 1952 where he joined the
research team headed by Dr. William Shockley that developed the
transistor.
Returning to M.I.T.
in 1953, Dr. Farley worked with a group that developed a
magnetic core memory for the first fully transistorized
computer, known as the Whirlwind Project. Components of this
computer are now exhibited at the Smithsonian and the Digital
Computer Museum in Marlboro, M.A. His groundbreaking work with
Dr. Clark on neural networks then followed. Additional
research, in conjunction with several colleagues, led to
computer techniques for the study of electroencephalograms
published in 1962.
In 1964, Farley
joined the University of Pennsylvania as Associate Professor of
Biophysics and published additional research on neural networks,
the electrophysiology of seizures, and computer simulations of
neural network behavior.
“The brain is an analog device, not a
digital one,” he frequently contended. “Neurons create
‘meaning’ by their frequency of firing,” he argued, “and by
their patterns of connecting to other neurons in networks, and
the intensity of their electrical activity and connectivity.
Computers, on the other hand, encode meaning by an on/off
semantic system. Their smallest units of significance are
either 0 or 1,” he would explain. “This is a binary or digital
system, very unlike the brain.”
In 1969, Dr. Farley
became the second member of Temple University’s new Department
of Computer and Information Science. He retired in 1986.
During his career,
Dr. Farley joined numerous professional organizations, including
the American Mathematical Society, the New York Academy of
Sciences, the American Physiological Society, and the
Biophysical Society, of which he was a charter member. He
was also a member of the National Institutes of Health’s ad-hoc
study section on biomedical engineering.
In addition to his
wife of 55 years, Elizabeth of Philadelphia, P.A., Dr. Farley is
survived by his son Martin of Lumberton, N.C.; his son Malcolm
of Brooklyn, N.Y.; his sister Frances Dennis of Selbyville, D.E.
and his brother Thomas of Carlsbad, C.A., near San Diego. His
eldest son Jefferson died before him.
