William Edmund Hick

Introduction

William Edmund Hick (August 1, 1912 – December 20, 1974) was a British psychologist who came to experimental psychology from medicine and helped build two subjects as he went: the quantitative study of human information processing, and ergonomics. His name attaches to a single result — Hick's law, the finding that the time a person needs to choose between alternatives grows with the number of alternatives, and that it grows logarithmically rather than linearly.

That result is one of the few quantitative regularities in psychology that survived the cognitive revolution intact, and it is cited far outside the discipline, in interface design, aviation and control-room engineering. It is also routinely quoted in a stronger form than Hick himself established, and the gap between the two is worth knowing.

Early Life and Education

Hick was born in Scarborough, Yorkshire. He trained not in psychology but in medicine, taking the MB and BSc of the University of Durham in 1938 and later the MD of the same university in 1949. It is a common error to describe him as a Cambridge graduate: Cambridge came into his life only in 1944, and by then he was a qualified physician in his early thirties.

He joined the Royal Army Medical Corps in 1941 and served through the middle years of the war. Like a number of British scientists of his generation, he was then pulled into the applied work that the war had made urgent — the study of how operators perform under load, in cockpits, radar stations and control rooms — and he never went back to clinical medicine.

Academic and Professional Career

In 1944 Hick moved to Cambridge to join the Medical Research Council's Applied Psychology Unit at the Cambridge Psychological Laboratory. The Unit, under Kenneth Craik and then Frederick Bartlett, was the centre of British applied experimental psychology, and it was there that Hick did the work on choice reaction time that made his reputation.

He was appointed Reader by the University of Cambridge in 1953 and became a Fellow of St John's College. He was a founding member of the Experimental Psychology Group and served as its President in 1958, the year it reconstituted itself as the Experimental Psychology Society. He was also a founder member of the Ergonomics Society and a member of the Ratio Club, the informal cybernetics dining group whose members included Alan Turing, W. Ross Ashby, W. Grey Walter and W. A. H. Rushton.

Hick's applied side is visible throughout his less famous work. With J. A. V. Bates he wrote a 1949 report on The Human Operator of Control Mechanisms for the interdepartmental committee on servo-mechanisms, treating the person in a control loop as a component with measurable transfer characteristics — a framing that belongs as much to control engineering as to psychology. Membership of the Ratio Club placed him in the middle of British cybernetics, in a group that included Alan Turing, W. Ross Ashby, W. Grey Walter and W. A. H. Rushton, and the influence shows in his willingness to write about people as information channels with a bandwidth.

His 1948 paper "The discontinuous functioning of the visual cortex" is the clearest example. Arguing purely from reaction-time and flicker data, Hick proposed that visual processing is not continuous but quantised in time — that the cortex samples in discrete moments rather than tracking smoothly. It was a bold inference from behaviour to physiology made a decade before there was any physiological evidence for it, and although the specific claims have been superseded, the strategy — use timing data to constrain what the hardware must be doing — is the one that later made the Hick paradigm useful.

He also exemplified a pattern common to the Applied Psychology Unit: research done under the pressure of wartime operational problems, then converted into a general science of human performance. Together with his founding role in the Ergonomics Society, this is why Hick is claimed by two disciplines that do not otherwise speak to each other much.

Major Contributions to Psychology

On the rate of gain of information (1952)

The paper that carried his name appeared in the Quarterly Journal of Experimental Psychology (4, 11–26). Hick set subjects a choice-reaction task with varying numbers of equally probable alternatives and measured how long they took. The finding was that reaction time increases with the number of alternatives, and does so as the logarithm of the number, usually written:

RT = a + b · log₂(n + 1)

The "+1" is Hick's own touch: it represents the subject's uncertainty about whether to respond at all, not merely about which response to make, and it is why the formula works even for the single-alternative case where a plain log would go to zero.

Hick's way of putting the result was more ambitious than the equation. Borrowing from Shannon's information theory, he reported that the rate of gain of information was roughly constant with respect to time within a perceptual-motor act — a figure on the order of a few bits per second. In that framing the human operator is a channel of limited capacity, and the law is a statement about bandwidth rather than about arithmetic.

The Hick–Hyman law

The generalisation came a year later, and from someone else. Ray Hyman (1953) showed that the logarithm of the number of alternatives is only a special case: the right predictor is the amount of information transmitted, measured as entropy. When alternatives are not equally probable — which in real tasks they almost never are — reaction time tracks the entropy of the set, not the count. Most textbooks still print Hick's version; the version researchers actually use is the Hick–Hyman law.

Reaction time, intelligence and the Jensen box

Because the slope of the function differs between individuals, the Hick paradigm became a standard tool for studying the relation between speed of processing and intelligence. Arthur Jensen used an apparatus known in the literature as the Jensen box to measure choice reaction time and its slope across many alternatives, arguing that the intercept and slope carry different information about ability. The programme remains active, and it is the direct ancestor of the large population studies of reaction time and cognitive ability that followed.

Major Publications

  • Hick, W. E. (1948). The discontinuous functioning of the visual cortex. British Journal of Psychology, 39(4), 220–229.
  • Hick, W. E., & Bates, J. A. V. (1949). The Human Operator of Control Mechanisms. Inter-departmental Technical Committee on Servo Mechanisms.
  • Hick, W. E. (1952). On the rate of gain of information. Quarterly Journal of Experimental Psychology, 4(1), 11–26.

An obituary by A. T. Welford appeared in Ergonomics in 1975 (18(2), 251–252).

Impact and Legacy

Hick's law is one of the most frequently quoted results in applied psychology, and it has real force: adding alternatives does cost time, and the cost is not proportional to how many you add. That insight shaped control-panel layout, cockpit design and later menu design, and it remains the default starting point for anyone modelling choice.

The misapplications are at least as common as the applications, and the criticisms are specific.

The eponym is incomplete. The entropy formulation that makes the law usable with unequally probable alternatives is Hyman's, and the version in general scientific use is the Hick–Hyman law. Attributing the whole thing to Hick is an instance of Stigler's law of eponymy.

The law has documented exceptions. Longstreth, El-Zahhar and Alcorn (1985) reported a systematic exception: the logarithmic relation holds for the initiation of a response but not for its duration, which behaves quite differently as alternatives increase. The authors treated it as evidence that the law is a fact about decision, not about execution.

It depends on the mapping. Hick's law holds under a condition that is often forgotten — the stimulus–response mapping must be arbitrary or unfamiliar. When the mapping is spatially compatible, the work of Fitts and Seeger (1953) showed that compatibility effects can dominate the count of alternatives entirely; pressing the button where the light appeared can be nearly as fast with eight alternatives as with two. With extended practice the slope of the function can also flatten toward zero as the task automatises.

The "constant rate of gain of information" is not a constant of the nervous system. Hick's own framing invites the reading that there is a fixed bandwidth in bits per second, and that reading has not survived. The measured rate varies with modality, age, practice, stimulus quality and individual ability — which is precisely why the Hick paradigm became a tool for studying individual differences rather than a way of measuring a universal channel.

And it is over-applied in design. The familiar advice that Hick's law means "more choices always take longer" is true only for unfamiliar, equally likely, randomly mapped choices. In a menu where items are alphabetised, where the same layout is used every time, and where users know roughly where to look, search is often closer to linear in position than logarithmic in count, and adding an option that is never read costs nothing at all. Hick's law describes a particular laboratory situation extremely well; it is not a general law of interfaces.

Learn More

References

  • Hick, W. E. (1952). On the rate of gain of information. Quarterly Journal of Experimental Psychology, 4(1), 11-26.

Last updated: 2026-10-20

Menu
Loading