Hull, Clark L. (1884-1952)
HULL, CLARK L. (1884-1952)
Clark Leonard Hull was born in Akron, New York, on May 24, 1884, and died in New Haven, Connecticut, on May 10, 1952. He earned his bachelor's degree from the University of Michigan in 1913, his master's degree in 1915, and a Ph.D. in experimental psychology in 1918 from the University of Wisconsin. His graduate work was done primarily under the direction of Joseph Jastrow, Daniel Starch, and Vivian A. C. Henmon.
Hull recorded some of his earliest career plans in his Idea Books. "It seems," he wrote, "that the greatest need in the science at present is to create an experimental and a scientific knowledge of higher mental processes" (Ammons, 1962, p. 814). He planned to become the "supreme authority" in the psychology of abstraction, concept formation, and, possibly, reasoning—to "both know the literature and create the literature on the subject." His doctoral dissertation, Quantitative Aspects of the Evolution of Concepts, was published in Psychological Monographs (1920).
Hull was afflicted with a variety of health problems, and his relatively late entry into the field was another threat to his long-range goal. In 1930, at the age of forty-six, he wrote, "Sometimes I have been depressed and discouraged in my hope to achieve a major contribution to the theory of knowledge by the fact of my age. Recently, however, the examination of the ages at which several of the great critics have produced their best works has shown that I have by no means reason to be depressed" (Ammons, 1962, p. 836). The list included English philosopher Thomas Hobbes, sixty-three; Dutch philosopher Baruch Spinoza, forty-five; German philosopher and mathematician Gottfried Wilhelm von Leibnitz, sixty-eight; English philosopher John Locke, fifty-eight; and German philosopher Immanuel Kant, fifty-seven. (He could have added that Russian scientist Ivan Pavlov was fifty-six years old when he first proposed the principle of the conditioned reflex.)
Hull was elected to the American Academy of Arts and Sciences in 1935, to the National Academy of Sciences in 1936, and to the presidency of the American Psychological Association 1935. He received the Warren Medal of the Society of Experimental Psychologists in 1945. A measure of his influence is that during the decade spanning 1941 to 1950 approximately 40 percent of all experimental articles published in the Journal of Experimental Psychology and the Journal of Comparative and Physiological Psychology included references to his work (Spence, 1952). Today Hull is best remembered for one theoretical publication on learning theory, Principles of Behavior (1943). However, a case has been made that the twenty-one theoretical articles he published in the Psychological Review between 1929 and 1950, the earliest of which introduced the Pavlovian construct the fractional anticipatory goal response (rG-sG), represent at least as important a contribution to learning theory. (For an extensive account of Hull's work and influence, see Amsel and Rashotte, 1984.)
After receiving his doctorate, Hull stayed in the psychology department at the University of Wisconsin, where he became director of the laboratory in 1925. During these years, his projects included the effects of tobacco smoking on mental and motor efficiency; the definitive books of the time in aptitude testing and hypnosis; and, as forerunners of the computer age, a logic machine that automatically computed correlations, and a robot that learned.
For Hull, Pavlov's conditioning identified the principle by which an event in one subsystem could come to influence functioning in remote parts of such machines, and Edward Thorndike's law of effect, the principle whereby originally random movements could be selected and be linked as responses to specific stimulus patterns (trial-and-error learning and the principle of the habit-family hierarchy). The design of a machine that learned, he thought, should allow for common reactions to stimuli with sensory similarity (generalization); for the learning of common reactions to stimuli with quite different sensory qualities (discrimination); and for the possibility of differential responding to minute differences in stimulus pattern (afferent neural interaction). These became central principles in his papers and in his later formal writings (Hull, 1943, 1952).
In 1929, Hull accepted the position of research professor of psychology at the Institute of Psychology (later the Institute of Human Relations) at Yale University, where he remained for the rest of his career. When he arrived at Yale, at the age of forty-five, he made this entry in his Idea Books, which seems prophetic in the light of the computer age:
Just as the correlation machine has been intimately associated with the testing program, so it appears that the design and construction of automatic physical machines will be intimately associated with my attempts to work out my program involving the higher mental processes. … I am pretty certain to be criticized and called a trifle insane at the very least. But whatever genius I have quite evidently lies in this direction. I can do no less than make the best of it—let the tendency, have free rein and go as far as it possibly can. … It may lead to real insight into the higher mental processes. … It may [on the other hand] possibly serve as the final reductio ad absurdum of a mechanistic psychology. If it does, well and good. But even if this should take place, it may at the same time result in such a development in psychic machines displaying an utterly new and different order of automaticity that mechanical engineering of automatic machines will be revolutionized to a degree similar to the introduction of steam engines and electricity. (Ammons, 1962, pp. 828-829)
As research professor at Yale, Hull had no formal teaching assignment, but he engaged in graduate instruction through a weekly seminar that attracted students and faculty of the institute for discussion of a variety of issues in behavior theory.
By Hull's own account, the Institute of Human Relations was a loose organization of behavioral scientists from various fields, mainly psychologists, sociologists, and cultural anthropologists. However, the conceptual framework for the unified contribution it provided was Hull's theoretical system, and it seems fairly clear, in retrospect, that this conceptual framework was an amalgam of influences from English naturalist Charles Darwin, Pavlov, and Austrian neurologist Sigmund Freud. Their influences, filtered through Hull's own, are evident in several important publications written by members of the institute, all of which featured stimulus-response analyses of complex learning. Books in this genre include Frustration and Aggression (Dollard et al., 1939), Social Learning and Imitation (Miller and Dollard, 1941), and Personality and Psychotherapy (Dollard and Miller, 1950).
Even before going to Yale, Hull had planned to prepare a magnum opus on the scientific (mechanistic) analysis of higher mental processes. The earliest titles he considered for his work, listed in his Idea Books in 1928 (Ammons, 1962, pp. 824-825), indicate this intended focus and the mechanistic emphasis: Mechanisms of Thought, Mechanisms of Mind, Mental Mechanisms, Mechanisms of Mental Life, Psychology from the Standpoint of a Mechanist.
Hull's plan was to invert the direction taken by the great philosophers—David Hume, Locke, Kant, and Hobbes—who had attempted to construct a theory of knowledge, thought, and reason on the basis of conscious experience but who, in Hull's eyes, had failed. He planned to attack the same problem using the opposite strategy. He wrote, "I shall start with action—habit—and proceed to deduce all the rest, including conscious experience, from action, i.e., habit" (Ammons, 1962, p. 837).
To carry out this task, Hull planned a three-volume work. The first volume was intended mainly to present a set of formal axioms that constituted a logical system from which hypotheses about mammalian adaptive behavior could be deduced. Completed in Hull's fifty-eighth year, this volume appeared in 1943 and was titled Principles of Behavior: An Introduction to Behavior Theory. The orientation of the second volume, based on the principles set down in the first, was toward specific instances of more complex adaptive behavior, such as maze learning and problem solving. Completed in Hull's sixty-eighth year, it appeared shortly after his death in 1952 and was titled A Behavior System: An Introduction to Behavior Theory Concerning the Individual Organism. The third volume, which Hull thought would be the most important, was to have applied the system to some elementary phenomena of social mammalian behavior. It was never completed.
The basic formal structure of the theorizing in Hull's Principles, as revised in Essentials of Behavior (1951), was an intervening variable approach borrowed from Edward Tolman (1938). It involved antecedent, manipulable environmental conditions (independent variables), consequent behavioral conditions (dependent variables), and a bridge of lower-and higher-level constructs connecting the two (the intervening variables). Hull's final version of the theory is summarized in the first chapter of A Behavior System as a set of postulates and corollaries, most of which are restated in mathematical form. Examples are Postulate II, relating the strength of an intervening variable, the "molar stimulus trace" S1, to the independent variables, the physical stimulus (S), as a power function of time since the beginning of the stimulus; and Postulates XIV, XV, and XVI, in which the momentary effective reaction potential S ER, which already takes into account the behavioral oscillation S OR and the threshold for responding SLR, is related to four response measures—response latency, response amplitude, response frequency, and resistance to experimental extinction—by negatively accelerated functions in the first, third, and fourth cases, and by a linear function in the second. This is the nature of Hull's formal theorizing.
These later, more formal portions of Hull's theorizing were not taken up after his death except by Kenneth Spence (1954, 1956) and a few of Spence's students (Grice, 1968; Logan, 1960). One reason was the advent of the cognitive revolution in psychology, which was for the most part a reaction against B. F. Skinner's behaviorism, but also against Hull's stimulus-response associationism, his hypotheticodeductive version of it in particular (Amsel, 1989). Many thought that such a formal Newtonian treatment was both too general in scope and premature. However, more recent mathematical models of associative learning, greatly influenced by Hull but more restricted in explanatory scope (Rescorla and Wagner, 1972), remain influential. Still very important in learning theory is the general approach, taken from Hull's earlier work, of employing hypothetical constructs derived from Pavlovian conditioning in the explanation of instrumental behavior. Hull (1931) regarded this work as stimulus-response analyses of "goal attraction" and "directing ideas," and called the mechanism the fractional anticipatory goal response (rG-sG). An extensive treatment of such a mechanism in appetitive learning occurs in frustration theory (Amsel, 1962).
The similarity of Hull's theorizing in the 1930s and 1940s to recent computer-generated models of learning and memory has been noted by Hintzman (1992) in an article with the subtitle "Was the Cognitive Revolution a Mistake?" Hintzman, himself a product of the cognitive revolution, asserts that "Key ideas of the cognitive revolution, including cognitive organization and the computer metaphor, have been largely abandoned; and basic concepts from the era of behaviorism and functionalism, such as association, inhibition, similarity, unconscious learning, and transfer, have reemerged." He points particularly to the similarity of "Hullian theories of fifty years ago" to connectionist and production-system models of more recent cognitivists. Much the same can be said about the approaches to theory based on neural networks that often depend on what is called the Hebbian synapse. The proposition here is that the efficiency of cell A in firing cell B is increased as a function of a growth process that takes place at the synapse between cell A and cell B. As a principle of association, this is not greatly removed from Hull's reinforcement postulate. In short, it appears that a part of Hull's lasting contribution will be to modern theories of the neurobiology of learning and memory, an ironic outcome considering the criticisms Hull suffered for what was called his "neurologizing."
See also:HEBB, DONALD
Ammons, R. B. (1962). Psychology of the scientist: IV. Passages from the Idea Books of Clark L. Hull. Perceptual and Motor Skills 15, 807-882.
Amsel, A. (1962). Frustrative nonreward in partial reinforcement and discrimination learning: Some recent history and theoretical extension. Psychological Review 69, 306-328.
—— (1989). Behaviorism, neobehaviorism, and cognitivism in learning theory. Hillsdale, NJ: Erlbaum.
Dollard, J., and Miller, N. E. (1950). Personality and psychotherapy. New York: McGraw-Hill.
Grice, G. R. (1968). Stimulus intensity in response evocation. Psychological Review 75, 359-373.
Hintzman, D. L. (1992). Twenty-five years of learning and memory: Was the cognitive revolution a mistake? In D. E. Meyer and S. Kornblum, eds., Attention and performance XIV. Hillsdale, NJ: Erlbaum.
Hull, C. L. (1920). Quantitative aspects of the evolution of concepts. Psychological monographs 28, whole no. 123.
—— (1931). Goal attraction and directing ideas conceived as habit phenomena. Psychological Review 38, 487-506.
—— (1943). Principles of behavior: An introduction to behavior theory. New York: Appleton-Century-Crofts.
—— (1951). Essentials of behavior. New Haven: Yale University Press.
—— (1952). A behavior system: An introduction to behavior theory concerning the individual organism. New Haven: Yale University Press.
Logan, F. A. (1960). Incentive. New Haven: Yale University Press.
—— (1979). Hybrid theory of operant conditioning. Psychological Review 86, 507-541.
Miller, N. E., and Dollard, J. (1941). Social learning and imitation. New Haven: Yale University Press.
Rescorla, R. A., and Wagner, A. R. (1972). A theory of Pavlovian conditioning: Variations in the effectiveness of reinforcement and nonreinforcement. In A. H. Black and W. F. Prokasy, eds., Classical conditioning II: Current research and theory. New York: Appleton-Century-Crofts.
Spence, K. W. (1952). Clark Leonard Hull: 1884-1952. American Journal of Psychology 65, 639-646.
—— (1954). The relation of response latency and speed to the intervening variables and N in S-R theory. Psychological Review 61, 209-216.
—— (1956). Behavior theory and conditioning. New Haven: Yale University Press.
Tolman, E. C. (1938). The determiners of behavior at a choice point. Psychological Review 45, 1-41.