Clark L. Hull

Introduction

Clark Leonard Hull (24 May 1884 - 10 May 1952) was an American psychologist who set out to make the study of learning and motivation into an exact science. He wanted psychology to be deductive and quantitative in the way he believed classical mechanics was: a small number of axioms, stated as equations, from which the behaviour of an organism could be derived and predicted. For roughly two decades, from the late 1930s until the cognitive revolution of the 1960s, the resulting system was the dominant framework for animal learning research in the United States and the most ambitious theory of behaviour that psychology had yet produced.

Hull was born in Akron, New York, and died in New Haven, Connecticut. He spent his early career at the University of Wisconsin and the rest of it at Yale University, where he held the Sterling Professorship. He is remembered for drive reduction theory, for the hypothetico-deductive method he imposed on experimental psychology, for the formula relating habit strength and drive to the probability of a response, and, less flatteringly, for a theoretical apparatus that grew so elaborate that it collapsed under its own weight. He also founded the modern experimental study of hypnosis, and he trained or collaborated with a generation of psychologists who carried behaviourism into social learning, motivation and communication research.

Early Life and Education

Hull was born in a log house near Akron, New York. His family moved to a farm in Michigan when he was three or four, and he grew up doing farm work and attending a one-room schoolhouse of twenty to thirty pupils. He excelled at mathematics and struggled with grammar. He qualified as a teacher at seventeen and taught in a school much like the one he had attended before entering Alma Academy, where a bout of typhoid fever interrupted his studies and left him with a lifelong difficulty in remembering names.

He went on to Alma College intending to become a mining engineer, studying mathematics, physics and chemistry. After two years he took a job as an apprentice mining engineer in Hibbing, Minnesota, and two months into it contracted poliomyelitis, which permanently paralysed one leg and left him dependent on a brace and a cane. The illness ended his engineering career and redirected him. During a year of convalescence he read William James's The Principles of Psychology, decided on psychology instead, and taught for two further years before marrying Bertha Iutzi in 1911.

He entered the University of Michigan in middle age relative to his peers and took his bachelor's degree in 1913. He then moved to the University of Wisconsin as an assistant to Joseph Jastrow, completed his master's degree in 1915 and his doctorate in 1918 with a dissertation on the quantitative evolution of concepts. He taught at Wisconsin from 1916 until 1929, becoming a full professor there, and even designed and built a mechanical device for calculating correlation coefficients, an early piece of laboratory automation.

The biography explains a great deal about the theory. Hull came to psychology late, by way of engineering and mathematics, having lost one career to illness; he was a man who built a calculating machine because he needed one, and who thought about the mind in the idiom of mechanics because that was the idiom he had been trained in. His students described a personality of immense industry and little warmth, and the system he built has something of that quality: exhaustive, impersonal, and indifferent to whether anyone found it attractive.

Academic and Professional Career

At Wisconsin, Hull worked on aptitude testing and vocational guidance, publishing Aptitude Testing in 1928 and developing an interest in the quantitative treatment of individual differences. In 1929 he was recruited to Yale by its president, James Rowland Angell, to join the newly founded Institute of Human Relations, an ambitious Rockefeller-funded attempt to bring psychology, sociology, anthropology and medicine under one empirical roof. Hull remained at Yale for the rest of his life.

His first decade at Yale was largely given to hypnosis. He taught an introductory course for premedical students, developed an experimental programme on suggestion, and published Hypnosis and Suggestibility in 1933, the first large-scale statistical and experimental treatment of the topic. The book established that hypnosis is not a form of sleep, confirmed that phenomena such as hypnotic anaesthesia and posthypnotic amnesia are real, and deflated a good deal of stage-hypnotist folklore about extraordinary enhancements of sensation or cognition.

From the 1930s onward he turned to the larger project of a general behaviour theory. In a 1930 paper he argued that psychology is a genuine natural science whose primary laws could be expressed in a small number of equations, from which the behaviour of individual organisms and of groups could be derived. Working with collaborators including Neal E. Miller, John Dollard, O. H. Mowrer and Carl I. Hovland, he produced Mathematico-Deductive Theory of Rote Learning in 1940 and Principles of Behavior in 1943, the book that made him the most influential learning theorist of his era. He received the Warren Medal of the Society of Experimental Psychologists in 1945. He suffered a serious heart attack around 1948 and continued to revise his system until his death in 1952.

Major Contributions to Psychology

Drive reduction theory

Hull's central claim was that behaviour is motivated by biological need and learned by the reduction of the drives that need creates. Deprivation produces a need; the need activates a drive; the drive energises behaviour; and behaviour that reduces the drive is strengthened, so that the same response becomes more likely in the same situation. Learning, on this account, is the strengthening of a stimulus-response bond, or habit, and it occurs only when a response is followed by drive reduction.

The theory distinguished primary drives, which are innate and biological, from secondary or acquired drives, which are learned through association with the reduction of primary ones. Hull grounded the whole scheme in an evolutionary and homeostatic picture inherited from Charles Darwin and Walter Cannon: organisms act to restore internal equilibrium, and what is reinforced is what restores it. He drew on Edward Thorndike's law of effect, on Ivan Pavlov's conditioning and on John B. Watson's behaviourism, but went beyond all three in insisting that the intervening variables had to be defined quantitatively and anchored to measurable operations.

The formal system

Hull's best-known equation expressed reaction potential as a multiplicative function of habit strength, drive, stimulus-intensity dynamism and incentive motivation, later expanded with inhibitory terms and an oscillation factor. The multiplicative form carried a substantive claim: if either habit strength or drive fell to zero, the behaviour would not occur at all. Motivation did not create learning; it amplified habits already formed.

The method mattered as much as the algebra. Hull insisted on a hypothetico-deductive procedure: state postulates precisely, derive theorems, test them experimentally, and revise in the light of the outcome. Whatever one thinks of the specific theory, the standard of rigour set a benchmark for experimental psychology and trained a generation to state their variables precisely enough to be falsified.

The goal gradient

Among Hull's more durable empirical findings is the goal gradient hypothesis: organisms exert more effort as they approach a goal. Rats in a maze accelerate in the final stretch, for instance. Hull documented and formalised the effect in the 1930s, and the same pattern has since been observed in human task completion and in consumer behaviour, where progress toward a reward increases the effort invested in reaching it.

The human half of that claim was not demonstrated until sixty years later. Kivetz, Urminsky and Zheng reopened the question in the Journal of Marketing Research (2006, 43(1):39), showing in a field experiment that customers in a cafe loyalty programme bought coffee more often the closer they were to a free one, and that the effect is driven by perceived rather than actual distance: customers given a twelve-stamp card with two stamps already punched completed the ten purchases they still faced faster than customers given an ordinary ten-stamp card, a median of ten days against fifteen. Faster acceleration toward the first reward also predicted quicker re-engagement afterward. The finding has since become standard material in the design of loyalty programmes, progress indicators and gamified interfaces — the clearest case of a Hullian result outliving the theory it was derived from.

Major Publications

Aptitude Testing (1928)
Hypnosis and Suggestibility: An Experimental Approach (1933)
Mathematico-Deductive Theory of Rote Learning (1940), with Carl I. Hovland and others
Principles of Behavior: An Introduction to Behavior Theory (1943)
Essentials of Behavior (1951)
A Behavior System: An Introduction to Behavior Theory Concerning the Individual Organism (1952)

Impact and Legacy

Hull was one of the most frequently cited psychologists of the 1940s and 1950s, and his students carried his framework into neighbouring fields. Neal Miller and John Dollard extended drive theory to social learning and imitation, arguing that acquired drives could motivate behaviour without any link to tissue need. Kenneth Spence, his closest collaborator and most important critic, revised the system into what became known as the Hull-Spence theory, replacing Hull's multiplicative combination of drive and incentive with an additive one and arguing that reinforcement affects performance rather than learning itself. Miller, Dollard, Mowrer, Hovland and Spence all went on to substantial independent careers. In the 2002 Haggbloom survey of the most eminent psychologists of the twentieth century, published in Review of General Psychology, Hull was ranked twenty-first.

The criticisms eventually overwhelmed the system, and they are worth stating plainly. First, the drive-stimuli that were supposed to give behaviour its direction proved extremely difficult to identify, and it is not clear that they are necessary to direct behaviour under ordinary circumstances. Second, the deprivation evidence was not as clean as the theory required: whether an animal becomes more active when deprived depends heavily on species and on the apparatus used to measure activity, with effects visible in a running wheel absent in a stabilimeter cage. Third, and most damaging, learning demonstrably occurs in circumstances that preclude any reduction of drive. Tolman and Honzik's latent-learning studies showed that rats exploring a maze without reward learn its layout as thoroughly as rewarded rats and simply do not display the learning until a reason to do so appears.

Fourth, the reinforcement mechanism itself was challenged from several directions at once. Sheffield and others argued that what reinforces is the consummatory response rather than drive reduction; Harlow and others showed that monkeys will work for the opportunity to explore or manipulate with no deprivation at all; and the Olds and Milner work on intracranial self-stimulation, published just two years after Hull's death, showed that animals will work prodigiously for brain stimulation that reduces no drive whatsoever. Fifth, the theory became baroque: as exceptions accumulated, Hull added postulates and parameters, so that critics charged the system with being unfalsifiable by accretion. B. F. Skinner's far more parsimonious descriptive approach, which dispensed with intervening variables entirely, proved easier to use and eventually displaced it.

A sixth objection is the one that most undermines the formal apparatus from inside. Hull's intervening variables were to be anchored to operations, but in practice several of them were anchored to the very behaviour they were supposed to explain. Habit strength is inferred from performance, drive from activity level, and reaction potential from the probability of the response; when the prediction fails, one can always say that habit strength was lower than assumed, or that an inhibitory term was at work. That is the structural reason the system could absorb exceptions indefinitely, and it is the reason "unfalsifiable by accretion" was a description of the method rather than an insult.

His work on hypnosis has a different and more mixed afterlife. Hypnosis and Suggestibility established that the topic could be studied experimentally and cleared away a good deal of folklore, and the finding that hypnosis is not sleep is now standard. But Hull's own position was unusually sceptical about the phenomena, and later research has not sustained all of his negative conclusions: there is reasonable evidence that hypnotic suggestion can produce meaningful analgesia and can modulate perception and, under some conditions, memory — though the last of these remains genuinely contested, and his warning against credulous reports of hypnotic feats is still worth heeding.

Hull's defenders, and later historians, have argued that the criticism is partly unfair. His insistence on explicit definitions and testable derivations raised the methodological standard of the field, and the specific concepts he introduced, habit strength, incentive motivation, generalised drive, remain in the vocabulary of motivation research even where the formal system has been abandoned. The goal gradient survives intact. What failed was the ambition to capture all of behaviour in one closed set of equations, and that failure is itself one of the more instructive episodes in the history of the discipline — a demonstration that precision is not the same as correctness, and that a theory can be simultaneously the most rigorous of its era and entirely wrong.

Drive reduction has also had an afterlife Hull could not have anticipated, in computational neuroscience. Keramati and Gutkin's 2014 paper in eLife (3:e04811) proposed "homeostatic reinforcement learning" as a formal reconstruction of the drive-reduction idea, and stated explicitly which of the classic objections it was meant to answer: that drive reduction cannot explain anticipatory responding, in which an animal voluntarily increases rather than decreases its deviation from a set point; that it cannot explain how secondary reinforcers such as money or a light predicting food acquire motivational value when they reduce no drive; and that intravenous feeding, which reduces drive directly, is not rewarding. The reconstruction solves all three by making reward a function of predicted drive reduction computed from orosensory cues rather than of actual physiological deficit — which is to say that what survived is the homeostatic framing, while the mechanism was replaced. The framework has continued to develop: a 2024 study in Neural Networks scaled homeostatic reinforcement learning to deep reinforcement learning and to high-dimensional visual input, with simulated agents foraging and thermoregulating on that basis.

Learn More

Neal Miller — who extended drive theory to social learning and imitation
Kenneth Spence — his closest collaborator and most important critic
Carl Hovland — co-author of the rote learning work, later a founder of communication research
O. H. Mowrer — his student, who carried two-factor theory into the study of anxiety
B. F. Skinner — whose parsimonious alternative displaced the Hullian system

References

Hull, C. L. (1943). Principles of Behavior. Appleton-Century.