Braithwaite, Richard Bevan (1900–1990)

views updated

BRAITHWAITE, RICHARD BEVAN
(1900–1990)

Richard Bevan Braithwaite, an English philosopher, was educated at King's College, Cambridge, where he studied physics and mathematics before turning to philosophy. Braithwaite was Knightbridge Professor of Moral Philosophy at Cambridge University. He served as the president of the Mind Association (1946) and of the Aristotelian Society (1946–1947). In the philosophy of science he made significant contributions on the nature of scientific theories and explanation, theoretical terms, models, foundations of probability and statistics, the justification of induction, and teleological explanations. He also wrote on subjects in moral and religious philosophy.

Scientific Theories

Braithwaite defended the view that a scientific theory consists of a set of initial hypotheses, with empirically testable generalizations that follow deductively. To explain a generalization is to show that it is implied by higher level generalizations in the theory. Often, especially in the physical sciences, the initial postulates will contain so-called theoretical terms, such as electron or field, that refer to items not directly observable. To understand the meaning of such terms, as well as the logical structure of the theory, one must begin by considering the theory as a formal calculus; that is, as a set of uninterpreted formulas. A calculus designed to represent a specific theory will have to be interpreted, but not all at once and not completely: Meanings are directly given only to those formulas representing the lower order empirical generalizations, rather than to initial formulas containing theoretical terms. The latter are indirectly and partially interpreted by the former.

Braithwaite's major contribution here consisted in the detailed attention he devoted to the nature of the initial or "theoretical" postulates. He divided these postulates into Campbellian hypotheses, which contain only theoretical terms, and dictionary axioms, which relate theoretical terms to observational ones. The latter include identificatory axioms, which identify single observational terms with theoretical terms—for example, a color word with expressions referring to wavelengths of light. Braithwaite argued that the advantage of systems containing theoretical terms over those whose initial postulates are entirely observational is that the former can more readily be extended to new situations than can the latter. However, Braithwaite held there is no special advantage to Campbellian hypotheses, because, at least for certain systems, the same testable consequences can be derived from identificatory axioms.

Scientific models are to be construed as alternative interpretations of a theory's calculus where the theoretical concepts in the original theory (such as molecules) are interpreted as designating more familiar and intelligible items (such as billiard balls). Accordingly, the theory and the model are to be distinguished; and while a model is not essential, it can sometimes be of help in extending a theory and clarifying its concepts.

Probability and Induction

Braithwaite proposed a novel finite-frequency theory of probability. Consider the statement (P), "The probability of a child being born a boy is 0.51," and the observed data that among 1,000 children 503 are boys. Such a situation is to be understood by imagining 1,000 sets of children, each containing 100 children of whom 51 are boys, and a selection of 1 child from each of the 1,000 sets, of whom 503 are boys. Since P is logically consistent with any observed data, the problem is to decide when to reject P. For this purpose it is necessary to have a rule specifying that a probability statement is to be rejected if the observed relative frequency differs from the probability postulated by more than a specified amount. This amount is determined by extralogical considerations involving the purpose for which the hypothesis is to be used and the value attached to possible consequences of its adoption. Such a rejection rule, Braithwaite claimed, is what gives empirical meaning to probability statements considered as constituents of theoretical systems. But suppose there are alternative probability hypotheses not rejected by the evidence in accordance with this rule. How is one to choose among them? Here again considerations of value must be invoked, and Braithwaite outlined a "prudential policy" of choosing the probability hypothesis that maximizes the minimum mathematical expectation of value.

Braithwaite also provided an original defense of Charles Sanders Peirce's solution to the problem of justifying induction. The problem was formulated by Braithwaite as follows: What warrant does one have for adopting the policy of accepting a hypothesis on the basis of many positive instances (the policy of "induction by simple enumeration")? The proposed answer consists of the following argument (where π is the principle of induction by simple enumeration): The policy of using π has been effective in many instances in the past; therefore (using π as the rule of inference) π will continue to be effective. Such an argument was traditionally dismissed as viciously circular, and Braithwaite undertook to prove this charge unjustified. The argument can be deemed valid and hence free from circularity, he claimed, because it enables one to pass from a mere belief in the general effectiveness of using π as a rule of inference, with a reasonable belief in π's past effectiveness, to a reasonable belief in π's general effectiveness. It would be viciously circular only if one were required to have an initial reasonable belief in π's general effectiveness. Since this requirement is unnecessary, the argument is not invalidated.

Moral and Religious Philosophy

Many of the conclusions and techniques of the philosophy of science were applied by Braithwaite in areas of moral and religious philosophy. Thus, just as one can defend the adoption of a particular scientific hypothesis by appeal to an inductive policy, so one can justify a particular action, such as returning a book, by reference to a moral policy, such as promise-keeping. Both sorts of policies are in turn justified by reference to the ends they subserve. Braithwaite showed how the mathematical theory of games, which he invoked in his discussion of hypothesis selection, can also be used to shed light on such notions as prudence and justice in situations involving human choices and cooperation between individuals. Finally, just as a moral assertion is to be construed as an expression of an intention to act in accordance with a certain policy, so a religious assertion must be understood, according to Braithwaite, as a declaration of adherence to a system of moral principles governing "inner life" as well as external behavior. The major difference between religious and moral assertions is that the former, being associated with empirical narratives, have a propositional element lacking in the latter.

Bibliography

Black, Max. Review of Theory of Games as a Tool for the Moral Philosopher. Philosophical Review 66 (1957): 121–124.

Hirst, R. J. Review of Scientific Explanation. Philosophical Quarterly 4 (1954): 351–355.

Russell, L. J. Review of Scientific Explanation. Philosophy 20 (1954): 353–356.

works by braithwaite

"Propositions about Material Objects." Proceedings of the Aristotelian Society 38 (1937–1938): 269–290.

"Teleological Explanation." Proceedings of the Aristotelian Society 47 (1946–1947): 1–20.

"Moral Principles and Inductive Policies." Proceedings of the British Academy 36 (1950): 51–68.

Scientific Explanation: A Study of the Function of Theory, Probability, and Law in Science. Cambridge, U.K.: Cambridge University Press, 1953.

An Empiricist's View of the Nature of Religious Belief. Cambridge, U.K.: Cambridge University Press, 1955a.

Theory of Games as a Tool for the Moral Philosopher. Cambridge, U.K.: Cambridge University Press, 1955b.

"Probability and Induction." In British Philosophy in the Mid-Century: A Cambridge Symposium, edited by Cecil Alec Mace. London: Allen and Unwin, 1957.

"Axiomatizing a Scientific System by Axioms in the Form of Identifications." In The Axiomatic Method with Special Reference to Geometry and Physics, edited by Leon Henkin, Patrick Suppes, and Alfred Tarski, 429–442. Amsterdam: North-Holland, 1959.

"Models in Empirical Science." In Logic, Methodology, and Philosophy of Science, edited by Ernest Nagel, Patrick Suppes, and Alfred Tarski, 224–231. Stanford, CA: Stanford University Press, 1962.