George Miller: Cognitive Revolutionist and Integer Persecutee

George Armitage Miller (1920–2012) was a pioneer of the cognitive revolution that restored mental phenomena to the field of psychology. Although best remembered for applying mathematical and computational theories to human behavior, Miller also advocated for psychology to focus on solving societal needs. This article overviews his life, career, and contributions to the field of psychology.

1920s and 1930s

Miller grew up in West Virginia amid the Great Depression. He was raised by Christian Scientists, taught to distrust medical science. At George Washington University, Miller was appalled to encounter a psychology textbook depicting the brain: “I was…trained to avoid materia medica, and I could recognize the devil when I saw him” (1989).

In 1938, Miller transferred to the University of Alabama, studying history and speech. One day, he accompanied his drama club friend Katherine “Kitty” James to a psychology seminar at Professor Donald Ramsdell’s home. Miller admitted that “it was not…holistic neurology that attracted me. I was there because I was interested in Katherine James; she was interested in psychology” (1989). Yet, he became intrigued by the psychology conversation and its “complex arguments over points I would never have thought to question” (Miller, 1989).

After earning his bachelor’s degree, Miller stayed at the University of Alabama as a graduate student in speech. He earned his master’s degree but turned down a fellowship because he could not afford to relocate. Fortunately, Ramsdell offered Miller a position teaching psychology. It was an easy decision: “Katherine James had become Kitty Miller, our first child had appeared, the Depression was still a reality, and I needed the job” (Miller, 1989). Thus, Miller began teaching 16 weekly discussion sections for Ramsdell’s introductory psychology course.

1940s

Optimal Design of Jamming Signals

In 1942, Miller went to summer school at Harvard and later enrolled as a doctoral student in psychology. His speech background proved useful as a research assistant in Stanley Stevens’ Psycho-Acoustic Laboratory. During World War II, Miller conducted research on how to jam signals to render the enemy’s radio communications unintelligible. He later recalled:

My generation saw the war against Hitler as a war of good against evil; any able-bodied young man could stomach the shame of civilian clothes only from an inner conviction that what he was doing instead would contribute even more to ultimate victory…The problem was clear: wars are noisy; armies must communicate. (Miller, 1977)

Miller (1947) studied the intensity, frequency, and temporal pattern of masking stimuli. He discovered that the acoustic properties annoying in everyday life — such as high pitches and unpredictable cadences — were ineffective for jamming signals. Miller persuaded the Army Signal Corps that “it was not the sound of a Stuka dive bomber coming right at you that reversed peristalsis and caused loss of sphincter control. I conducted irrefutable demonstrations that optimal masking signals are better than the best nuisance signals” (1989).

The war defined Miller’s doctoral experience. He felt more military researcher than psychologist and found that “as the faculty we had come to study under became increasingly involved in the war, formal instruction shrank to nothing” (Miller, 1989). His lab work isolated him from Harvard’s social and clinical psychologists. After the war, the polarized psychology department split in two. Miller kept his psychophysical approach but networked with the psychologists in the new social relations department, claiming that “if ever there was a social behavior, it was language” (Crowther-Heyck, 2017).

Information Theory

Miller earned his PhD in 1946, submitting his military research as a dissertation. Staying at Harvard as a speech researcher, he soon read Claude Shannon’s “A Mathematical Theory of Communication.” Shannon’s (1948) information processing theory covered signals, noise, channel capacity, and information measurement. He defined the binary unit, or bit, as the smallest unit of information with only two alternatives. Shannon also argued that the probability of each event in a sequence depends only on the state attained in the previous event.

Miller considered Shannon’s article a revelation. First, it explained his wartime data’s inverse relationship between intelligibility of a speech signal and number of possible alternative signals. Second, information theory could help explain human behavior. By conceptualizing information as anything that reduces behavioral uncertainty, Miller and Frederick Frick (1949) created an index to measure the degree of unpredictability in an organism’s behavior. They predicted that over time, with learning and conditioning, an organism’s actions would become less unpredictable. They labeled their research as behavioral statistics, or behavioristics, because “behavior seems to be the observable datum of psychology. We have modified the noun by the word statistical because behavior has that kind of complexity that seems to preclude, at least for the present, a completely deterministic account” (Miller & Frick, 1949).

1950s

At the time, behaviorism had a hold on psychology. Behaviorists believed that human thought processes could not be accessed, and therefore psychology must focus on observable human and animal behavior. Miller joked that psychology had “first lost its soul and then its mind” (Bruner, 1983). In the 1950s, behaviorism was challenged by cognitivism. Miller argued that if psychology were to succeed, “mentalistic concepts would have to integrate and explain the behavioral data” (2003). Initially apprehensive, Miller was emboldened by psychologists Jerome Bruner and Noam Chomsky to drop his behaviorist pretense. Miller recalled that cognition was “considered a naught word…but Jerry Bruner proudly nailed it to the door that [Harvard] Dean McGeorge Bundy gave us, and…we set about making it respectable” (1977).

Language and Communication

Miller’s Language and Communication (1951) helped establish psycholinguistics, which studies the cognitive processes of how people learn, use, and invent languages. Using information theory, Miller argued that words cannot be broken down devoid of context. If every action is contextualized by the action before and after it, then each word in a sentence depends on the word before and after it. Miller (1951) also addressed idiosyncrasies between cultures in language use, developmental linguistics, and the social aspects of language.

Massachusetts Institute of Technology (MIT)

In 1951, Miller was recruited to MIT to help establish a Human Resources Research Laboratory. MIT had broad departments and a task-based, rather than discipline-based, orientation. Miller enjoyed the teamwork, collaborating with engineers, physicists, mathematicians, and linguists on human factors (Crowther-Heyck, 2017). He also organized biweekly “Pretzel Twist” meetings to discuss current psychological topics (Hérbert, 2006).

Miller continued researching speech perception, including how context affects the correct perception of a word (Miller et al., 1951; Miller et al., 1954). He also found that word recognition and discrimination improve with free recall practice (Bruner et al., 1955). He identified the phonetic features critical to listener recognition of consonants spoken over voice communication systems with frequency distortion and random masking noise (Miller & Nicely, 1955). His resulting matrix for spoken consonants is still used today.

Return to Harvard

In 1955, Miller returned to Harvard in a tenured role. Harvard’s psychology department was again polarized, this time between Stevens’ psychophysics and B.F. Skinner’s behaviorism. Stevens, who recommended someone else for the role, was not interested in Miller’s research or budding cognitive orientation. Miller instead collaborated with department chair E.B. Newman to obtain grants. Miller also better defined his area of study as the psychology of communication.

The Magical Number Seven, Plus or Minus Two

Preparing a lecture for the Eastern Psychological Association, Miller sought to link his two research projects: applying information theory to absolute judgments and studying how recoding could extend the span of short-term memory. He decided both had numerical similarity. Miller presented experimental findings that people can retain about seven — plus or minus two — chunks of information in short-term memory. His research showed that people could retain about seven numbers, six letters, or five words, correlating to the increased bits of each information type. Miller speculated that short-term memory limits followed an evolutionary advantage to remember “a little information about a lot of things” (1956).

Miller’s lecture was published in 1956 in the Psychological Review. His opening lines have been widely cited, even in obituaries, as a testament to his wit:

My problem is that I have been persecuted by an integer. For seven years this number has followed me around…Either there really is something unusual about the number or else I am suffering from delusions of persecution. (Miller, 1956)

The paper was an important contribution to the cognitive revolution, appearing “at a time when many psychologists were looking for new ways to think about their science” (Miller, 1989). By applying information theory to human short-term memory, Miller revealed that the brain could be studied as an information processor with systems and mathematical rules. Current models of cognition still argue that knowledge is organized into hierarchical networks of nested chunks.

Symposium on Information Theory

Miller (2003) cited a 1956 symposium on information theory, held at MIT, as the birth of the cognitive revolution. Miller presented the limits of short-term memory; Herbert Simon and Alex Newell presented their logic machine; IBM tested Donald Hebb’s theory of cell assemblies; Chomsky presented transformational generative grammar. Miller left the symposium convinced that “experimental psychology, theoretical linguistics, and the computer simulation of cognitive processes were all pieces from a larger whole” and there would be a “progressive elaboration and coordination of their shared concerns” (2003).

Miller and Chomsky

Miller wanted to estimate the amount of information per word in conversational speech but was unable to extend information theory beyond Shannon’s analysis of letter sequences in written texts. Fortunately, he met Chomsky at the MIT symposium. Chomsky persuaded Miller that Shannon’s probabilistic processes would not explain natural language. Information theory was based on probabilities, which was not easily compatible with cognitivism. Rather, grammatical rules are “mentalistic hypotheses about the cognitive processes responsible for the verbal behaviors we observe” (2003). With Chomsky’s help, Miller began using syntactic theory to examine the cognitive processes responsible for structuring human language.

In 1957, the Social Science Research Council held a summer workshop at Stanford University to help social scientists think mathematically. Miller and Chomsky led a seminar on mathematical theories of communication, even moving their families in together on campus. They continued developing theorems about grammar and parsing that ultimately established mathematical linguistics, which studies computational problems for interpreting syntax.

1960s

Cognitivism soon outpaced behaviorism. Miller’s definition of psychology evolved from the science of behavior to the science of mental life, proclaiming that “human minds exist and it is our job as psychologists to study them” (1962). The American Psychological Association (APA) gave Miller an Award for Distinguished Scientific Contributions in 1963, demonstrating that cognitivism was taking hold in mainstream psychology.

Plans and Structure of Behavior

Miller spent a year at Stanford’s Center for Advanced Study in Behavioral Sciences. He collaborated with Eugene Galanter and Karl Pribram on mathematical models and cognitive simulations. Their 1960 book, Plans and Structure of Behavior, explored how a robot could be programmed to plan and act. Miller et al. (1960) argued that a feedback loop, rather than a stimulus-response, was the building block of behavior. Although they later decided that their feedback loop was too general, it was an important step in combining artificial intelligence research with psychology to explain human behavior. It became a manifesto for the cognitive revolution and still influences theories of problem solving and information processing.

Center for Cognitive Studies

Returning from Stanford, Miller was dissatisfied with Harvard’s narrow, polarized conception of psychology. He decided that “either Harvard would have to let me create something resembling the interactive excitement of the Stanford Center or else I was going to leave” (Miller, 1989). He confided in Bruner, and in 1960 they co-founded the Center for Cognitive Studies at Harvard. Students and research fellows experimented with sentence memory, demonstrating top-down processing in language comprehension by showing that people better recognize words in grammatical and sensible sentences. Miller deemed the Center a success: “the bright young graduates grew up to become important psychologists unafraid of words like mind and expectation and perception and memory” (Miller, 2003).

A Full Plate

In 1964, Miller became chairman of the Harvard psychology department. It was no easy task. Ousted chairman Newman was upset; Stevens and Skinner did not want to teach. Nevertheless, Miller attended official meetings as chairman, co-directed the Center, and lobbied weekly in Washington, D.C. to fund the Center. In 1967, an exhausted Miller decided “to hell with it” and went to Rockefeller University where “they had professors and each professor was a department unto himself” (Hérbet, 2006). He stayed involved in national science policy with the APA, National Institutes of Health, National Academy of Sciences, and President’s Scientific Advisory Committee.

APA Presidential Address

In Miller’s presidential address at the APA convention, he advocated to “give psychology away” to the public rather than develop specialized, elite knowledge (Miller, 1969). He argued that psychologists are citizens first and scientists second, obligated to share practical value with society. Miller said, “I can imagine nothing…more relevant to human welfare…than to discover how best to give psychology away” (1969). Just before Miller’s speech, the Black Students Psychological Association had stormed the stage to share their grievances as Black psychologists and American citizens (Pickren & Rutherford, 2010). In retrospect, their grievances reflected Miller’s message that psychology had fallen short of its societal obligations.

Miller (1969) also proclaimed that psychology had fallen short of its goal to understand, predict, and control mental and behavioral phenomena. It had provided society with nothing as tangibly impactful as gunpowder or the steam engine. Miller (1970) later criticized psychotechnologies such as polygraphs and personality tests. He demanded to assess the “reliability, social relevance, safety, [and] informed consent” of these tools before introducing them to society (Miller, 1970). He was also adamant that any released psychotechnologies should be accessible to all groups, not just privileged subgroups.

1970s

Developmental Psycholinguistics

Visiting the Institute for Advanced Study in Princeton, Miller collaborated with Philip Johnson-Laird on mental representation of word meanings in long-term memory. In Language and Perception, Miller and Johnson-Laird argued that words and percepts provide “avenues into a conceptual realm that is itself the central concern of cognitive psychology” (1976).

Back at Rockefeller, Miller set up a developmental psycholinguistics lab to research how children acquire word meanings. Miller (1977) coined the phrase “spontaneous apprentices” to describe how children acquire language from adults, beyond what adults consciously teach them. He suggested scientists should also be childlike apprentices of nature to gain knowledge.

Two Foundations

In 1976, the Sloan Foundation created a Special Program in Cognitive Science to link the study of brain and mind. Miller proposed that many disciplines should be involved, including psychology, linguistics, neuroscience, computer science, anthropology, and philosophy. He felt each discipline “had inherited a particular way of looking at cognition and…recognize[d] that the solution to some of its problems depended crucially on the solution of problems traditionally allocated to other disciplines” (Miller, 1989). He became part of an interdisciplinary committee to develop appropriate next steps for a field of cognitive science. The Sloan Foundation provided grants to universities to promote interdisciplinary communication.

In 1986, the McDonnell Foundation wanted to establish a research program in cognitive neuroscience. Per Miller’s recommendation, the foundation funded study panels on memory, emotion, higher cognitive skills, action systems, and attention. These five panels, composed of neuroscientists and cognitive psychologists, created the research program guidelines. Miller became chair of the new McDonnell-Pew Program in Cognitive Neuroscience.

1980s

State of Psychology

As the new editor of the Psychological Bulletin, Miller criticized past issues as “strangely parochial” for catering to small, niche groups (1980). He implored the journal to do better:

Do authors really believe that no one outside psychology is concerned with psychological questions? Or that literal facts can be recorded objectively, untouched by human bias? Or that important psychological developments are irrelevant to the culture and society in which they occur? (Miller, 1980)

Furthermore, Miller feared that psychophysiology was drifting away from psychology, and called for article submissions from experts in neuropsychology, psychopharmacology, cognitive neuroscience, and “all the rapidly twigging branches of the brain sciences” (1980).

In 1985, Miller contributed a chapter to A Century of Psychology as Science. He argued that the constitutive problem of psychology is the “explanation of the capacity for conscious experience” (Miller, 1985). He argued that psychology had been pulled in two directions — biological and social — that could be reconciled through the study of immediate experience, or consciousness. Little progress had been made because “we have been too willing to change the subject, too easily distracted by problems for which we were not yet ready, too uncertain what a science of consciousness would look like” (1985). Optimistically, Miller shared that the “revival of work in cognition, spurred by the advent of intelligent machines and a better understanding of the structure of intelligent systems, seems to me a sign of a new awakening” (1985).

WordNet

Miller joined Princeton University in 1979 as the James S. McDonnell Distinguished University Professor of Psychology. In 1986, he co-founded the Princeton Cognitive Science Laboratory and began developing WordNet, an electronic reference database of English words. WordNet groups nouns, verbs, adjectives, and adverbs into cognitive synonyms based on their meaning. Miller wanted to create precise definitions of words to help computers avoid confusion with words with similar sounds or meanings. He also wanted a way to test psycholinguistic theories of how humans use and understand words. WordNet is still in use; it has been recreated with other languages and incorporated into many systems for search engines, language translation, and more. It is an example of “giving psychology away” to help society.

1990s to 2010s

In 1990, Miller became professor emeritus and senior research psychologist at Princeton. He received many awards and recognition, including the APA Life Achievement Award, APA Award for Lifetime Contributions to Psychology, and National Medal of Science.

Miller’s wife Kitty died in 1996. Although she was the one who introduced Miller to psychology, little is known about her life beyond a credit for contributing to WordNet (Miller & Fellbaum, 2007). Miller stayed involved in research, spent more time playing golf, and remarried in 2008. He died at home in 2012 of complications from pneumonia and dementia.

Obituaries sang Miller’s praises. The Los Angeles Times called him an “iconoclastic psychologist who played a crucial role in shifting his field from the study of behaviors to the direct examination of thought processes” (Maugh, 2012). The New York Times penned that Miller “set off an explosion of new thinking about thinking” (Vitello, 2012). Psychologist Steven Pinker wrote that Miller “fomented the cognitive revolution, invented psycholinguistics and cognitive psychology, imported powerful ideas from the theories of information, communication, grammar, semantics, and artificial intelligence, and left us a sparkling oeuvre that proves that a rigorous scientist needn’t write in soggy prose” (Pinker, 2013).

Legacy

George Miller has a great legacy as a founder of the cognitive revolution. But diving into his curriculum vitae reveals something more. He was provocative and highly reflexive about psychology and his role in it. With only four degrees of separation from founding father Wilhelm Wundt, Miller was taught that psychology is biological, the experimental method is the key to scientific objectivity, and that “pandering to public interest” would “weaken and eventually destroy the scientific core that holds the field together” (Miller, 1985).

Miller came to think differently. He was frustrated by the polarity between biological, micro-level psychology and social, macro-level psychology. Miller called psychology an “intellectual zoo,” sharing that “I have discovered several other psychologies, all claiming proprietary rights to the label, all competing for disciples, and each contemptuous of the others” (1985). As Miller broadened his epistemological stance, he quipped that “reality is vastly overrated. It is merely the point of origin from which everything interesting departs” (1989). A singular, objective reality was neither obtainable nor interesting to him.

Miller saw a return to the mind as a recalibration of the field to study consciousness and immediate experience. He championed the role of language in the field, believing it made internal psychological phenomena observable and measurable (APS, 2012). Despite all of his criticisms, he envisioned a bright future for the field of psychology:

The original dream of a unified science that would discover the representational and computational capacities of the human mind and their structural and functional realization in the human brain still has an appeal that I cannot resist. (Miller, 2003)

Miller was clearly more than a pioneer of cognitive psychology (and cognitive science, and cognitive neuroscience). He pushed psychology to realign on its goals of understanding human behavior through consciousness and immediate experience. And above all, he measured psychology’s progress by how well it helped society.

References

APS. (2012, September 26). Remembering George A. Miller. Association for Psychological Science — APS. Retrieved June 22, 2022, from https://www.psychologicalscience.org/observer/remembering-george-a-miller

Bick, C., & Rabinovich, M.I. (2009). Dynamical origin of the effective storage capacity in the brain’s working memory. Physical Review Letters, 103(21), 218101.

Bruner, J. (1983). In Search of Mind. New York: Harper & Row.

Bruner, J.S., Miller, G.A., & Zimmerman, C. (1955). Discriminative skill and discriminative matching in perceptual recognition. Journal of Experimental Psychology, 49(3), 187–192.

Crowther-Heyck, H. (1999). George A. Miller, language, and the computer metaphor of mind. History of Psychology, 2(1), 37–64 .

Hébert, R. (2006). The Miller’s tale: A genealogy of the father of the cognitive revolution. APS Observer, 19(6).

Maugh, T. H. (2012, August 6). George A. Miller dies at 92; pioneer of cognitive psychology. Los Angeles Times. Retrieved May 19, 2022, from https://www.latimes.com/local/obituaries/la-me-george-miller-20120806-story.html

Miller, G.A. (1947). The masking of speech. Psychological Bulletin, 44(2), 105–129.

Miller, G.A. (1962). Some psychological studies of grammar. American Psychologist, 17(11), 748–762.

Miller, G.A. (1969). Psychology as a means of promoting human welfare. American Psychologist, 24, 1063–1075.

Miller, G. A. (1970). Assessment of psychotechnology. American Psychologist, 25(11), 991–1001.

Miller, G.A. (1977). Spontaneous apprentices: Children and language. New York: Seabury Press.

Miller, G. A. (Ed.). (1980). Aspirations of the Future Editor: An Editorial. Psychological Bulletin, 87(1), 213–215.

Miller, G.A.. (2003). The Cognitive Revolution: A Historical Perspective. Trends in Cognitive Sciences, 7(3), 141–144.

Miller, G. A. (1985). The constitutive problem of psychology. In S. Koch & D. E. Leary (Eds.), A century of psychology as science (pp. 40–59). American Psychological Association.

Miller, G. A., Bruner, J. S., & Postman, L. (1954). Familiarity of letter sequences and tachistoscopic identification. Journal of General Psychology, 50, 129–139.

Miller, G.A., & Fellbaum, C.D. (2007). WordNet then and now. Language Resources and Evaluation, 41, 209–214.

Miller, G. A., & Nicely, P.A. (1955). An analysis of perceptual confusions among some English consonants. Journal of the Acoustical Society of America, 27, 338–352.

Miller, G.A., & Frick, F.C. (1949). Statistical Behavioristics and Sequences of Responses. Psychological Review, 56(6), 311–324.

Miller, G.A., Heise, G.A., & Lichten, W. (1951). The intelligibility of speech as a function of the context of the test materials. Journal of Experimental Psychology, 41(5), 329–335.

Miller, G.A., & Johnson-Laird, P.N. (1976). Language and Perception. Belknap Press.

Pickren, W., & Rutherford, A. (2010). A History of Modern Psychology in Context. Hoboken, NJ: John Wiley & Sons.

Pinker, S. (2013). George A. Miller (1920–2012). American Psychologist, 68(6), 467–468.

Raymond, W. D., & Healy, A. F. (2017). Breaking into the mind: George A. Miller’s early work in the American Journal of Psychology. The American Journal of Psychology, 130(3), 269–282.

Simon, S. (2007, July 7). The magical qualities of the number 7. NPR. Retrieved May 24, 2022, from https://www.npr.org/templates/story/story.php?storyId=11803762

Vitello, P. (2012, August 1). George A. Miller, a Pioneer in Cognitive Psychology, Is Dead at 92. New York Times. Retrieved May 19, 2022 from https://www.nytimes.com/2012/08/02/us/george-a-miller-cognitive-psychology-pioneer-dies-at-92.html