It has been questioned whether it is at all possible to give the term “probability” one uniform and consistent interpretation not merely in mathematics, statistics, evaluation of evidence, testing of hypotheses but also in any situation where we want to express a degree of belief. The question itself of assigning a probability to our belief in a scientific theory has raised important questions of an epistemological as well as mathematical nature.

This paper is a tentative step toward formalizing in symbolic logic some of the central assumptions of Clark L. Hull's theory of behavior. There will be noticeable deviations from Hull's own terminology and form of statement, but this will be largely in the interest of greater logical simplicity. The notation of symbolic logic will not be used, but the behavioral axioms will be so stated as to admit of easy translation into such notation. Such easy translatability seems to be a prerequisite if theorems are to be derivable with genuine deductive rigor.

Professor Feigl's admirably lucid and concise appraisal of the major considerations which are still significant in the controversy between realism and phenomenalism includes the important reminder that the problem at hand should be viewed as concerning, not the truth or falsity of two conflicting hypotheses, but rather the comparative adequacy of two alternative proposals for the rational reconstruction of scientific knowledge. Feigl advocates the approach of semantic realism in preference to a phenomenalistic type of reconstruction on the grounds that in his critical examination, the latter has been “found wanting,” whereas semantic realism—and only semantic realism—can provide a satisfactory systematic account of the meaning and the experiential foundation of scientific hypotheses. While I fully agree with Feigl's general appraisal of the issues involved, I have doubts as to the superiority of the proposed realistic interpretation, and I should like to state these here in outline.

In mathematics we have demonstrably insoluble problems, one example being that of trisecting an arbitrary angle in elementary geometry. Every now and then, we encounter engineers and others who offer solutions of the insoluble and make some stir. To those who feel convinced of the demonstration of insolubility, these claimed solutions do not seem to deserve any serious considerations. In fact, such solutions have long since ceased to attract attention from mathematicians.

The situation in philosophy seems different. Here we do not possess demonstrations as clear and distinct as in mathematics. For example, it has been a fashion among philosophers for two centuries or so to answer Hume's scepticism concerning induction. Although, presumably, there never has been any answer or solution which is as widely accepted among philosophers as Hume's sceptical argument, yet philosophers continue to enjoy elaborating arguments for and against Hume's scepticism. And people who feel convinced by Hume's resaoning often find it difficult to understand why philosophers persist in trying to contradict Hume flatly instead of seeking for some more devious way out.

A special merit of Professor Feigl's stimulating essay lies in the clarity with which he draws the issue between phenomenalistic and realistic interpretations of science as one concerned entirely with the evaluation of their relative adequacy as analytic reconstructions of empirical knowledge. Responsible and fruitful discussion of these interpretations cannot therefore be a debate over the truth or falsity of scientific theories. For as he carefully notes, the differences between these alternative analyses are not “pragmatic differences” of the ordinary sort, since no differences in regard to any observable consequences that could weigh for or against a scientific theory can be associated with them. Feigl's own rejection of phenomenalism and his espousal of semantic realism must in consequence also be judged in terms of the standard he himself has recognized as alone relevant in this discussion.

The contention of the above comment, as I understand it, is that there is, analogously to the problem of trisecting an arbitrary angle in mathematics, a sound demonstration, along the lines employed by Hume and Russell, of skeptical conclusions concerning our inductive knowledge of the future, and that hence one is mistaken in imputing to that argument, as I have done, a logical slip arising from a confusion in the use of ‘future’ and other similar words. I am indebted to Mr. Wang for the stimulation, and to the Editor of Philosophy of Science for the opportunity, to try to make my points clearer.

The proponents of these skeptical arguments maintain that, apart from some supporting postulates which we must make in inductions about the future, we have no evidence concerning the future of any given moment. If we restrict ourselves to the evidence we have from the present and the past, and do not bolster this evidence with some Principle of Induction, like that of Hume to the effect that the future will be like the past, “we have no theoretical assurance,” as Mr. Wang puts it, “that the practice of guiding our activities by what we learn from ... [the past] will lead to success on the whole and in the long run.”

The position that science is a technique establishing the means to achieve any stipulated end has now fanned out and been defended by social scientists as well as by natural scientists. It is the thesis of this paper that the bifurcation of science and morality derives from the social status of both science and scientists today, and involves, wittingly or unwittingly, an uncritical acceptance of dominant social values. Science is thus not non-moral, as is claimed, but rather appropriates conventional morality. Certain conditions have brought about this role for science and scientists.

In Human Knowledge, Its Scope and Limits Bertrand Russell recognizes and attempts to deal with what is one of the central questions of philosophy since Descartes, “The relation between individual experience and the general body of scientific knowledge.” He states that his purpose is “to discover the minimum principles required to justify scientific inferences.” The search for such principles arises from his belief that “data are private and individual.” He has only scorn for those who, “finding these problems distasteful” have “tried to deny that these problems exist.”

The literature of science speaks of many things in the heavens and on earth that neither the man in the street nor the scientist in his laboratory, can directly see with his eyes or touch with his fingers. Scientific hypotheses refer to atoms and genes, cosmic rays and electrostatic charges. No one can deny the importance of these hypothetical constructions; without them we could neither understand nor manage the world which we directly experience. They simplify the system of laws that relate observable variables to one another. They suggest further research and enable us to predict what will be found if certain operations are carried out. The sporadic and disjointed collection of observed objects and events, and the occurrence of the unusual ones which we call illusions, are explained by correlating them with a hypothetical physical or physiological state of affairs that has order and spatio-temporal continuity.

“We are in the position of a little child,” said Einstein in a press interview twenty years ago, “entering a huge library whose walls are covered to the ceiling with books in many different tongues. … The child does not understand the languages in which they are written. He notes a definite plan in the arrangement of the books, a mysterious order which he does not comprehend, but only dimly suspects.”

There has recently been considerable controversy in the literature over the concept “disposition to respond” as used by Charles Morris in his criterion by which it is to be determined whether or not something functions as a sign. It is often assumed that this concept serves as a means for broadening a strictly behavioral criterion through the introduction of something which is a characteristic of sign situations when behavior is not. But dispositions have no practical value in this respect, for Morris's criterion can not be used to obtain sign knowledge about situations not giving rise to overt behavior. However, dispositions do have a practical function in Morris's criterion, a function which has been overlooked in these discussions.

In this rejoinder to the critical comments elicited by my essay “Existential Hypotheses,” I propose to deal first with the challenge coming from the avowedly different philosophical outlook of Professor Churchman. My other critics, Professors Frank, Hempel, Nagel and Ramsperger, on the whole, share my basic conception of the tasks of philosophy of science and epistemology, even if they dissent in one important respect or another from the special solution I suggested. But since I discern even in Professor Nagel's remarks (and possibly also between the lines of Professor Ramsperger's comments) a pragmatist or instrumentalist strain akin to the major contentions of Professor Churchman, it will be well to begin with a defense and further clarification of my underlying point of view. Only after this restatement of my platform will I undertake to defend semantic realism against the specific criticisms advanced by the last four authors.