ABSTRACT
As late as the end of the eighteenth century, it was universally believed, by those who thought at all about such matters, that the norms of rationality were defined for all time by Aristotle’s logic, and were exemplified for all time by Euclid’s geometry. In the next 200 years, the traditional logic grew into modern symbolic logic, comprising such sub-disciplines as Boolean algebra, set theory and metermathematics, while Euclidean geometry was shown to be but one of several equally logical geometries, so that which described the real world was not to be settled by reason alone. In our own century, there has been another explosive growth in the conception of rationality. Until recent times, logic continued to focus on mathematical inference and proof. Rational thought about empirical matters now has become the subject matter of flourishing disciplines like statistics, theory of probability, theory of games, utility theory, decision theory and the like. Associated with these theories a new technology has sprung up, whose electronic components allow for an exploitation of the theoretical possibilities far beyond what can be done by a brain of flesh and blood – for certain sorts of tasks. If machines do not think, whatever it is that they do do puts our thinking to shame, given the same problems. More and more areas of rational decisions have been brought into the domain of such theory and practice, and its frontier is expanding rapidly. A brave new world of rationality is in the making. One of the functions of philosophy is to stand aside and say, “However . . .” The intent is neither to derogate what has been done nor to discourage further endeavor. It is to restore a sense of perspective, and by heightening an awareness of limits, to contribute to their expansion. The new technology of rational decision-making is significantly indebted to the design of weapons and weaponsystems, but this does not pre-determine its capacities. It has proven beyond dispute its usefulness in civilian life, especially in science and industry. With increasing urgency, the question arises of whether it can also enhance the rationality of decisions on broad issues of social policy and political confrontations. There are no serious difficulties in principle in processing the information contained in statutes, court-records and the like. Indeed, much of this has already been done. How is the outcome of the process to be used to increase the rationality of legislators, judges and juries? Computers will predict elections from early
returns, and even before the balloting. Can we use them to decide how to vote? That we cannot do everything does not imply that we can do nothing, but to moderate our expectations from the new theories and technologies of rationality is itself a component of rationality. Whether a specified belief or action is rational can no longer be deduced from the conceptual absolutes of tradition. Contemporary theories of rational decision-making also have their limitations. These limitations are not overcome simply by defining rationality as whatever fits the theories. Terms like “intuition,” “insight” and “creative imagination” are sometimes used as honorifics, and sometimes as devices to hide and protect irrationality. They may also be used to name thought-patterns that, properly understood, are quite rational. The logic we actually use cannot be identified by fiat with the logic we reconstruct in particular theories of rationality. What we call “logic” is sometimes counterposed to psychology, in which is localized the terra incognita of the irrational (“Here there be dragons”). What John Stuart Mill saw as a shortcoming of certain kinds of scientific theories can be seen also in certain kinds of reconstruction of such theories. He noted:
It is unphilosophical to construct a science out of a few of the agencies by which the phenomena are determined, and leave the rest to the routine of practice or the sagacity of conjecture. We either ought not to pretend to scientific forms, or we ought to study all the determining agencies equally, and endeavor, so far as it can be done, to include all of them within the pale of science; else we shall infallibly bestow a disproportionate attention upon those which our theory takes into account, while we misestimate the rest, and probably underrate their importance.1
