Abstract

In speaking of empirical science as a self-correcting process one implies that a proposition accepted in accordance with the rules of procedure may have to be eliminated later according to these very rules. Taking this for granted one realizes that a particular empirical science, say physics, should be defined in terms of rules of method rather than as a system of propositions representing our knowledge at a given time. Obviously both the science of Galileo and Newton and the science of Einstein and Bohr are called ''physics'' and we do not regard this as a mere equivocation. It may, however, be objected that a definition of a science in terms of rules of method is incompatible with the view that a science consists of a set of propositions. To meet this objection I propose that we distinguish between the structure of a science as defined in terms of the rules of method and the corpus of this science at a given time, i.e., the set of propositions accepted at this time in accordance with the rules of method of this science. The relation between the structure of a science and the corpus of a science is similar to that obtaining between the structure of a legal order, e.g., American law, which is defined by the rules of the process of legislation in the broadest sense and the legal norms which are the corpus of law at a given time. It should then be noted that the assertion that a certain proposition belongs to a given science is incomplete without a time-index. The scientist changes the corpus of his science either by introducing new propositions to it or by eliminating propositions which were previously accepted and did not stand further control. The significance of whatever else he does in the course of his inquiry can be determined only in the light of this selection. There are two stages in this process of selection. The first is exclusively concerned with meanings as such. The scientist has to ascertain that a given aggregate of words represents a proposition, that this proposition contains only objective meanings, that it is not self-contradictory, that it belongs to the subject-matter of the particular science. The second phase of the selection, exhibited in scientific procedure in the strict sense, applies only to such propositions which have passed this preexamination. We shall call those rules of method in terms of which the first stage of selection is defined rules of the language of science and those rules in terms of which the second stage is defined rules of the procedure of science. In this paper I shall try to show how important it is to distinguish clearly between these two types of rules. I shall first deal with the rules of the language of science and confine myself to sciences dealing with the physical world.