HEBB, DONALD (1904-1985)

Donald Olding Hebb was born July 22, 1904, in the small community of Chester Basin on the Atlantic coast of Nova Scotia, Canada. Both his parents were physicians; his two brothers followed in their footsteps, and his sister Catherine became a well-known neurophysiologist after receiving her medical degree. Donald, however, was the family's nonconformist: he wanted to be a novelist, and in 1925 he graduated from Dalhousie University with a degree in English. He spent the next two years teaching and traveling. Hebb never completed his novel, but he was an excellent writer (his classical work on the neurological bases of behavior, The Organization of Behavior, 1949, is eminently readable), and he taught numerous students how to communicate clearly.

Early Career

In 1927, having become interested in the work of Sigmund Freud, Hebb enrolled as a part-time graduate student in psychology at McGill University. He supported himself by teaching and soon was appointed principal of a local elementary school. He experimented with teaching methods in his school, and at McGill he developed an interest in the "nature-nurture" question.

Around this time, newly married, Hebb almost decided to make a career in education, but illness and tragedy diverted his course. He was immobilized by tuberculosis of the hip for more than a year, and during his illness he wrote a theoretical master's thesis, "Conditioned and Unconditioned Reflexes and Inhibition," which his university examiners rated cum laude. In this thesis he speculates about a phenomenon that later became known as the Hebb synapse, stating that "the discharge of one neuron into another is increased by the discharge of the second neuron." Hebb further proposes that: "An excited neuron tends to decrease its discharge to inactive neurons," an idea that he apparently considered too audacious to include in the famous neurophysiological postulate that appears in his 1949 book, The Organization of Behavior. Boris Babkin, a McGill physiologist and former student of Pavlov, read Hebb's thesis, and when Hebb was mobile again, Babkin encouraged him to embark on a conditioning experiment. This type of research soon palled, however, and after his wife was killed in a motor accident, Hebb decided to leave Montreal. He was offered an assistantship at Yale University, but Babkin recommended that he apply to study with Karl Lashley in Chicago. Hebb took his advice, a crucial step in his career, and was accepted.

Karl Lashley as Mentor and Colleague

At the University of Chicago, Hebb came under the influence of L. L. Thurstone, C. Judson Herrick, Nathaniel Kleitman, Wolfgang Köhler, and, of course, Lashley himself. Lashley believed that the goal of psychology was to discover how the brain determines behavior, so the questions he asked were different from those arising at the time in mainstream behaviorism.

After a year Lashley moved to Harvard University and Hebb went with him, completing his experiments on the effect of dark-rearing on rat visual perception. After receiving his doctorate, Hebb remained at Harvard for a year and then received a fellowship to study Wilder Penfield's patients at the newly established Montreal Neurological Institute. A major project was to determine the effect of frontal-lobe removal on measures of intelligence. Hebb found that extensive damage to that area had a negligible effect on standard intelligence test scores, which pleased Penfield and posed an enduring problem for Hebb. Another finding was that large right-temporal-lobe ablations impaired visual perception.

At the end of his two-year fellowship, Hebb went to Queen's University, in Kingston, Ontario. There he lectured and experimented with rats to follow up his work at the Neurological Institute. He devised a rat intelligence test (Hebb and Williams, 1946) that he considered analogous to the tests used on human patients and, on the basis of his experiments, concluded that early experience had a significant effect on intelligence. Hebb developed this theme and used it to explain the lack of effect of large frontal-lobe lesions in adults (Hebb, 1942).

In 1942 Lashley was appointed director of the Yerkes Laboratories of Primate Biology in Orange Park, Florida, and he invited Hebb to join the staff. The opportunity to work with Lashley again easily outweighed Hebb's doubts as to the suitability of chimpanzees as experimental subjects, and he seized the opportunity. Lashley intended to develop tests of learning and problem solving while Hebb worked on tests of temperament and emotionality. The effects of brain lesions on these tests would then be studied. In fact no operations were performed until after Hebb left five years later, largely because Lashley's experience with rats did not adequately prepare him for a battle of wits with chimpanzees. In the meantime Hebb made some interesting discoveries about the causes of fear, and he accumulated a fund of chimpanzee anecdotes that enlivened his writing and lectures for the next forty years. But it was the time he spent pondering the neurophysiology of thought and intelligence, and discussing his ideas with Lashley and other members of the staff, that Hebb valued most. In 1944 a paper by Lorente de Nó on neuronal loops and recurrent circuits came to Hebb's attention and restructured his conception of the brain. Strangely, although many psychological phenomena, such as attention, cried out for explanations in terms of feedback from central to peripheral brain areas, and neurophysiologists had known about recurrent circuits for some time, the idea of one-way sensory-to-motor paths remained deeply ingrained among psychologists. The only feedback route envisaged was via the effects of responses on sensory input.

One of the most important questions raised by Lashley in his monograph Brain Mechanisms and Intelligence (1929) was why the connections established during learning could not be localized by brain lesions. Another topic that must have been discussed at length by Lashley's circle was stimulus generalization. Why, for example, is learned recognition not disturbed by a change in the size of a visual stimulus? Hebb thought that Lorente de Nó's neural loops would provide answers to many long-standing problems concerning the nature of the representation of objects in the brain.

A Neural Model of the "Idea"

The theory that finally emerged (Hebb, 1949) was, in keeping with the views of the day, strongly empiricist. Hebb postulated that initially random connections between cortical cells become organized by sensory input into densely interconnected groups that he called cell assemblies. Hebb's answer to Lashley's localization problem was that the cells of each assembly are dispersed over a large area of the cortex so that enough interconnected cells survive all but the largest lesions, ensuring that objects continue to be represented. Hebb explained generalization by stipulating that during the initial investigations of an object, many different sensory patterns contribute inputs that are incorporated into, or closely linked to, a single cell assembly representing the object. The dense intrinsic connections ensure that the whole assembly fires when only partial input is presented.

Cell assemblies that are active successively acquire connections with each other, which explains expectancy and the association of ideas. In effect, Hebb developed a neural model of the "idea," rescuing it from the obloquy of mentalism under which it had languished for almost half a century. Many psychologists were chafing under the constraints and dogmas of behaviorism, and they received Hebb's liberating ideas with enthusiasm. Hebb presented psychologists with an alternative to the mind-body dichotomy that had forced behaviorists to outlaw many important psychological phenomena.

The establishment of cell assemblies and the connections between them requires neural plasticity. The mechanism Hebb proposed was that if an axon terminal and its postsynaptic neuron were active at about the same time, the effectiveness of the synapse would increase. This model of synaptic change is now known as the Hebb synapse, and in the years since his death it may be that Hebb has come to be viewed as a neurophysiologist. If so, this is a pity; Hebb's attempt to explain how the brain represents the outside world, in which the Hebb synapse played a role, was a more original contribution than the Hebb synapse itself and, by illustrating the power of neurological models, was of more far-reaching influence.

Although Hebb made no attempt to test a computer or mathematical model of his conception of the cell assembly, a not very successful attempt to do so was made by Rochester, Holland, Haibt, and Duda (1956) at IBM shortly after The Organization of Behavior was published. The cell-assembly idea has influenced all subsequent neural-network models of learning, from the perceptron (Rosenblatt, 1962) to the parallel distributed processing models of Rumelhart and McClelland (1986). Although Hebb's neural network was a paper and pencil model that he showed little inclination to computerize, its pioneering quality should make Hebb nerve nets would be a more fitting memorial to his name than Hebb synapses.

Later Career

Hebb returned to the McGill department of psychology in 1947 and became chairman a year later. He studied the effects of rearing environment in rats and dogs, obtaining results that were influential in shaping programs for providing more stimulating surroundings for disadvantaged infants. Later he measured effects of prolonged sensory deprivation on human subjects and probed the cell assembly hypothesis more directly by studying the patterns of breakdown of perception at reduced contrast, as when an image is held fixed on the retina.

Hebb was elected president of the Canadian Psychological Association in 1952 and of the American Psychological Association in 1960. In 1966 he was made a fellow of the Royal Society, and in 1979 a foreign associate of the National Academy of Sciences (USA). He was chancellor of McGill University from 1970 to 1974. Hebb retired to his birthplace in 1977 and died there in August 1985.

See also:PARALLEL DISTRIBUTED PROCESSING MODELS OF MEMORY

Bibliography

Hebb, D. O. (1932). Conditioned and unconditioned reflexes and inhibition. Master's thesis, McGill University.

—— (1942). The effect of early and late brain injury upon test scores, and the nature of normal adult intelligence. Proceedings of the American Philosophical Society 85, 275-292.

—— (1949). The organization of behavior. New York: Wiley.

—— (1980). D. O. Hebb. In G. Lindzey, ed, A history of psychology in autobiography, Vol. 7, pp. 273-303. San Francisco: Freeman.

Hebb, D. O., and Williams, K. (1946). A method of rating animal intelligence. Journal of General Psychology 34, 59-65.

Orbach, J. (1998). The neuropsychological theories of Lashley and Hebb. Lanham, MD: University Press of America.

Rochester, N., Holland, J. H., Haibt, L. H., and Duda, W. L. (1956). Tests on a cell assembly theory of action of the brain, using a large digital computer. IRE Transactions. Information Theory IT-2, 80-93.

Rosenblatt, F. (1962). Principles of neurodynamics: Perceptrons and the theory of brain mechanisms. Washington, DC: Spartan Books.

Rumelhart, D. E., and McClelland, J. L. (1986). Parallel distributed processing, Vol. 1: Foundations. Cambridge, MA: MIT Press.

Peter M.Milner