Skip to main content Skip to Search Box
Summary Article: Suppes, Patrick from Biographical Dictionary of Psychology

Born: 1922, Tulsa, USA Nat: American Ints: Educational, evaluation and measurement, experimental, philosophical and theoretical psychology Educ: BS University of Chicago, 1943; PhD Columbia University, 1950 Appts & awards: Professor of Psychology, Stanford University, 1950-; Distinguished Scientific Contribution Award, American Academy of Sciences; President, APA; Member, American Academy of Arts and Sciences; Hon. D, University of Nijmegen, 1979, University of Paris V, 1982


Principal publications
  • 1969 Stimulus/response theory of finite automata. Journal of Mathematical Psychology, 6, 327-55.
  • 1971-90 Foundations of Measurement (3 vols). Academic Press (with D. H. Krantz; R. D. Luce; A. Tversky).
  • 1984 Probabilistic Metaphysics. Blackwell.
  • 1989 Current directions in mathematical learning theory. In Degreef, E.Roskam, E. E. Mathematical Psychology in Progress. Springer-Verlag.
  • 1991 Language for Humans and Robots. Blackwell.
  • 1993 Models and Methods in the Philosophy of Science: Selected Essays (Synthese Library, 226). Kluwer Academic.
  • Further Reading
  • Savage, C. W.Ehrlich, P. (1992) Philosophical and Foundational Issues in Measurement Theory. Erlbaum.
  • Wright, G.Ayton, P. (1994) Subjective Probability. Wiley.

  • Patrick Suppes was born in Tulsa, Oklahoma, in 1922, educated in Chicago and Columbia, and has been at Stanford University since 1950, holding a full professorship since 1959. He has made many substantial contributions to psychology at the philosophical, experimental and applied levels, and was also founder and chief executive of a related commercial organization, the Computer Curriculum Corporation, 1967-90. Suppes's most notable characteristic is a mastery of mathematics and physical science far beyond the reach of most psychologists and philosophers, combined with a detached and objective awareness of how these sciences resemble psychology and how they differ. He is one of the few psychologists to have taken the idea of psychology as a science seriously, not merely as a pious aspiration to be expressed in the early pages of chapter 1 and then forgotten.

    Suppes's active life covers the whole period of the birth and development of computers and he has been at the forefront throughout, being particularly concerned with computer applications in teaching and learning at all levels, including arithmetic and second-language learning in children, and learning at university level. His most recent publications summarize his life's work, and references to numerous earlier papers that were influential in their time may be found in them. The most substantial of these (in every sense of the word) is his three-volume Foundations of Measurement (1971-90). The argument in these volumes is technical and accessible only to specialists, but its influence has filtered down to some extent to workers at the coal face. His Probabilistic Metaphysics (1984) presents the views on science (including psychology) which permeate his writings. His basic premise is that ‘randomness and probability are real phenomena, and therefore are not to be accounted for by our ignorance of true causes’. Certainty of knowledge is unachievable. Logic (at the level of complexity at which it covers arithmetic) and mathematics (including arithmetic, geometry and set theory) are incomplete; so, therefore, must be the empirical sciences that depend on logic and mathematics. At the level of quantum theory there are ‘no hidden variables’. At the level of behaviour theory, even if we know the initial probability of a response, the parameters that determine how fast behaviour will change under different reinforcement schedules, and which reinforcement schedule was in operation, it remains the case that the response on any given trial is not uniquely determined. The sciences are in fact diverging; there is no bounded, fixed scientific theory towards which they converge. Suppes holds that though mind is absolutely dependent on brain it cannot be reduced to brain, simply because the neural hardware employed may differ from individual to individual. Elsewhere he writes: ‘the network of neurons in a human brain may constitute … a computationally irreducible system. The implications of this conjecture for having theories of learning or performance as detailed as one might ask for are quite pessimistic’. It follows that, in psychology, prediction may frequently be impossible; but prediction is impossible in many relatively simple physical systems, or works (as in psychology) only under conditions of strict experimental control. Biological systems, of which the system controlling human behaviour is one, are in fact more predictable than many physical systems. Suppes speculates: ‘Perhaps in the next century we will come to think of psychology as being a better predictive science than physics.’ His chapter ‘Current directions in mathematical learning theory’ (1989) considers the state of the art, but also develops his own (1969) work on stimulus/response theory which he has on occasion referred to as his most significant achievement.

    Suppes has made major contributions to the theory and experimental psychology of language. The meaning of a word, phrase, utterance, etc., is, in his view, a procedure or collection of procedures; it is private and probabilistic for each individual. The public meaning may be thought to correspond to the temperature of a gas, and the private procedures of each individual to the motions of the particles that make up the gas. Language for Humans and Robots (1991) is a collection of his articles on the subject from the last twenty years. There are four main themes. (1) He introduces the concept of ‘congruence of meaning’ (analogous to congruence of geometric shape). Every utterance asserts an inbuilt hierarchy of propositions, but propositions are only identical when the utterances from which they are abstracted are congruent. Suppes is critical of the tendency of philosophers to regard the proposition as a given. (2) He is also critical of the tendency to emphasize grammar and syntax in the study of children's speech, rather than semantics; and of the neglect of propositional attitudes in children. (3) He advocates a model—theoretic semantics without quantifiers or variables, having only constants — giving sets and relations and operations thereon. He shows how such a semantics may be obtained from his ‘procedural’ semantics. (4) He investigates methods of teaching robots to obey verbal instructions, and criticizes the relative neglect of ‘learning’ in artificial intelligence and computer science.

    Models and Methods in the Philosophy of Science (1993) is a collection of Suppes's contributions to that subject, containing much matter of interest that is not directly relevant to psychology. Having regard to the depth and breadth of his contributions, Suppes is one of the few psychologists who can seriously be compared with major figures in the older-established scientific disciplines. He has achieved very high repute in his profession but not the degree of influence that ought to have been his, perhaps because of the intrinsic difficulty of much of his work.

    N. E. Wetherick
    © 1997, 2002 Routledge

    Related Credo Articles

    Full text Article Preface
    Elsevier's Dictionary of Psychological Theories

    When the famous German-born American social psychologist Kurt Lewin suggested that nothing is as important as a good theory, he most likely was...

    Full text Article Science
    Blackwell Companions to Philosophy: A Companion to Cognitive Science

    The cognitive science of science studies the cognitive processes involved in carrying out science: How do scientists reason? How do scientists develo

    Full text Article Laws of Nature
    Encyclopedia of Philosophy & the Social Sciences

    After centuries of neglect, an intense interest in scientific laws was generated in the mid 20th century, principally by the seminal work on scienti

    See more from Credo