¹ The controversy raged most furiously among German economists (particularly between Menger and Schmoller) during this period and there is some reason for thinking that English analytical economics survived partially by grace of the language barrier. At any rate, analytical economics in Europe did not experience the enthusiastic advance that characterizes the English-speaking development from Alfred Marshall to the present day.

² The principal contender for that honour was, of course, Wesley C. Mitchell. In his early work Mitchell had expressed many of the same criticisms of orthodox theory as Veblen, and had demonstrated as well, in essays like “The Backward Act of Spending Money,” an ability to drive his points home with a literary style and logical power which, if gentler than that of his former teacher, equalled it in effectiveness. But Mitchell did not become the institutional theorist that Veblen had been. His later work, in the National Bureau of Economic Research, was a self-denying descriptive empiricism. He tried to reconstruct economics on foundations different both from orthodox theory and that of Veblen; for he considered them equally in error as unrealistic abstractions. The repudiation of both by the empirical economics he was trying to build was clearly stated in Mitchell's own words: “… if anything were needed to convince me that the standard procedure of orthodox economics could meet no scientific tests, it was that Veblen got nothing more certain by his dazzling performances with another set of premises.” Letter from W. C. Mitchell to J. M. Clark, published in S. A. Rice (ed.), Methods in Social Science (Committee on Scientific Method in the Social Sciences of the Social Science Research Council), p. 677.

³ I refer to the articles and notes in the American Economic Review over the past four years which began with R. A. Lester's article, “Shortcomings of Marginal Analysis for Wage-Employment Problems” (March, 1946) and F. Machlup's reply, “Marginal Analysis and Empirical Research” (September, 1946). See also Eiteman's, W. monograph, Price Determination: Business Practice versus Economic Theory (Bureau of Business Research, University of Michigan, Report No. 16, University of Michigan Press, 1949).Google Scholar

⁴ There can be little more reason than this for the introductory statement of a book for beginning students of economics: “Economics … is a science, one of the branches of that great systematic study of the world we live in which we call Science with a capital S.” Hicks, J. R. and Hart, A. G., The Social Framework of the American Economy (New York, 1945).Google Scholar

⁵ It is difficult to resist the conclusion that this is sometimes the primary motivation for the use of the term. In discussing papers on “Fiscal Policy in Prosperity and Depression,” presented at the sixtieth annual meeting of the American Economic Association in Chicago, 1947, Walter E. Spahr remarked: “… as additional reasons why there are inadequate grounds in science for general acceptance by scientists of … compensatory fiscal policy—much, if not most, of the recent and current advocacy of what are alleged to be compensatory fiscal policies seems not to have grown out of careful examination of available evidence, but appears, rather, to be the result of some internal cogitation or mental gymnastics out of which emerge theories that probably have little value insofar as the standards of science are concerned.” As an example of the neglected “evidence” which the “scientist” would consider, Spahr notes: “… some of the important and unhappy results of pump-priming during the years 1934-39 seem to be neglected, or ignored, or minimized. I refer, for example, to the persistent heavy unemployment, the mere trickle of capital into new enterprises, the low and generally-persistent decline in the velocity of bank deposits.” American Economic Association, Papers and Proceedings, May, 1948, pp. 407–8.Google Scholar See also, Hutchison, T. A., The Significance and Basic Postulates of Economic Theory (London, 1938).Google Scholar In the introductory chapter of this book the author discusses the “scientific” approach and lays to with great energy on the matter of “pseudo-science.” The opprobrious term however is freely applied to any who disagree with his own severely limited and somewhat naive positivism. Hutchison, in fact, simply dismisses such “pseudo-scientists” from his audience entirely (p. 13), and proceeds in the comfortable assurance that none remain to dispute with him on a definition which equates “science” with bare empiric fact, and categorizes all the rest as mere “philosophy.”

⁶ It is necessary to speak of “reality” in inverted commas because the interpretation of what is reality is itself an essential element of any epistemology. Such questions, although fundamental, cannot of course be fully discussed here. Suffice it to say that the view of “reality” which is implicit to this paper is that while a completely objective “reality” may exist, we cannot have any other knowledge of it than that which results from the human mind working upon sense data. As Whitehead has somewhere suggested, the real world is a deduced thing. As far as purposive action in everyday life is concerned, we find it necessary to operate with what the psychologists have described as an “assumptive world,” rather than the “real world” even in Whitehead's sense.

⁷ It is only fair to point out that in the hands of some economic positivists, theory plays a much larger role than this. Lionel Robbins, for example, makes his plea for economics as a positive science almost entirely on the basis that the content of economics is non-normative. See, An Essay on the Nature and Significance of Economic Science (London, 1937).Google Scholar

⁸ The mathematical nature of the theoretical probability distribution probably accounts for this inconsistency. As Lewis remarks: “… it is typical of empirical theorists that they begin with the stem demand that we must restrict ourselves to objective empirical facts, but very soon will be discovered talking pure a priori mathematics, though without any awareness that they have changed the subject.” Lewis, C. I., An Analysis of Knowledge and Valuation (LaSalle, Illinois, 1946), p. 283.Google Scholar

⁹ Ricardo, David, Economic Essays, ed. by Conner, (London, 1923), pp. 63 f.Google Scholar

¹⁰ The compilation of the various kinds of national income data is probably the clearest example of this in current economic research. While pioneering efforts in this field were made as early as 1915, it was the development of the Keynesian conceptual framework during the 1930's which made possible the highly integrated system of national income statistics we now possess.

¹¹ See, for example, T. C. Koopmans, “Measurement without Theory” (a review of A. F. Burns and W. C. Mitchell, Measuring Business Cycles), Review of Economic Statistics, August, 1947. See also comments by Vining, and Koopman's, reply under the heading of “Methodological Issues in Quantitative Economics,” Review of Economic Statists, 05, 1949.Google Scholar

¹² Pepper, S. C., World Hypotheses, a Study of Evidence (Berkeley and Los Angeles, 1942)Google Scholar, passim.

¹³ Ibid., p. 321.

¹⁴ Koopmans, T. C., “Review,” Review of Economic Statistics, 08, 1947.Google Scholar

¹⁵ The view is implicit in this paper that, although their basic epistemologies may be the same, the experimental methods of the physical sciences are inapplicable to the social sciences. A strong argument against this point of view is contained in Popper, K. R., “The Poverty of Historicism,” Economica, 1944 and 1945 Google Scholar; see especially Part II, August, 1944. A full exposition of what Professor Popper conceives to be the dangers of the “historicist” view (which he claims is the necessary alternative to the “piecemeal experimentation” methods of the physical sciences) is contained in his The Open Society and Its Enemies (2 vols., London, 1945).Google Scholar

¹⁶ Investigation of Motor Vehicle Accidents, memorandum of instructions to investigation officers under Privy Council Order 80/1045 (amended as P.C. 20/3570, Sept. 3, 1947). This memorandum requires the investigator to collect a prodigious amount of information without one reference to the degree of seriousness of the accident. In fact it begins with the statement that the instructions refer to “Every automobile accident involving an officer or servant of the Crown …” (my italics).

¹⁷ P. 173. The almost synonymous use of “discussion” and “description” in this passage also betrays a fundamental misconception of the nature of economic analysis.

¹⁸ Strachey, Lytton, Eminent Victorians (London, 1918).Google Scholar

¹⁹ It is by this criterion of adequacy that doubt is frequently expressed concerning the Keynesian analysis of the unemployment problem. The theory is couched in aggregative terms and while it is considered valid as an analysis of economic processes, the level of generalization may be too high for the purposes of policy implementation. Policies must necessarily be directed at more specific variables than the Keynesian aggregates, and until the aggregates are analytically broken down into those elements concomitant with the actual policy possibilities, the Keynesian theory will remain of doubtful adequacy in terms of the problem which it originally set out to meet. A great deal of the theoretical and empirical work presently being done by Keynesians is devoted to breaking down these aggregates. This effort is of very great importance to economic theory as a whole because it suggests the possibility of constructing a bridge between those two great bodies of existing general theory, the theory of price and the theory of national income.

²⁰ It is interesting to note that in the largest field of applied or practical economic research, that of government, the reluctance to recognize the subjective and judgment character of conclusions is perhaps the greatest. The studious elimination of the first personal pronoun from economic memoranda prepared in government departments may be not only a curious anachronism but one with serious consequences for the future of economics and its relation to public policy.

²¹ “A thorough-going empiricism is inconsistent with science itself. The essence of science, the understanding as distinct from the mere photographic reception of concrete phenomena, is theory and the essence of theory is analytical abstraction. Whatever its dangers, there is no other way.” Parsons, T., “Sociological Elements in Economic Thought,” Part II, Quarterly Journal of Economics, 08, 1935.Google Scholar (Parsons' italics.)

²² Emphasis on this point was one of the most important contributions of the early pragmatists to epistemology. Indeed, to some, it was the unique feature of the new theory. Henri Bergson, for example, in his introduction to the French translation of William James's Pragmalism, said: “On pourrait, ce me semble, résumer tout l'essentiel de la conception pragmatiste de la vérité dans une formule telle que celle ce: Tandis que pour les autres doctrines une vérité nouvelle est une découverte, pour le pragmatisme c'est une invention.” Le Pragmatisme (Paris, 1911), p. 11.Google Scholar

²³ Marxian theory, in the hands of some of its exponents, resists attack by this device. By insisting that the observer is, intellectually, the product of his own environment, no more need be done than assert a flaw in that background to “prove” that an observation is not “objective” or is “dialectically false.”

²⁴ Even in the physical sciences, the development of fundamental laws must necessarily be the result of mentally idealized “experiments”; for example, the discovery of the law of inertia in classical mechanics by the imagination of a body moving without friction or external influences and therefore moving in a straight line and without change of velocity.

²⁵ “In science the first criterion is the pragmatic one of success—in explanation.” Parsons, T., “Sociological Elements in Economic Thought,” Part II, Quarterly Journal of Economics, 08, 1935.Google Scholar

²⁶ When the physicist admits of a difference between “appearance” and “reality,” and that knowledge of the ultimate reality will never exist, the epistemology of that science becomes that put forward in this paper. For example: “Science is not just a collection of laws, a catalogue of unrelated facts. It is a creation of the human mind, with its freely invented ideas and concepts. Physical theories try to form a picture of reality and to establish its connection with the wide world of sense impressions. Thus the only justification for our mental structures is whether and in what way our theories form such a link.” Einstein, A. and Infeld, L., The Evolution of Physics (New York, 1938), p. 310.Google Scholar (My italics.) Many similar statements of the nature of physical theory are to be found throughout this book.

²⁷ I have here irtterpreted the physicist's problem of the indeterminancy of the behaviour of the single atom or molecule as a case of the lack of a known law, not as the lack of a governing rule in actuality. This interpretation may of course be the result of an unsophisticated understanding and if so, then the differences between the physical and the social sciences may be smaller than I have supposed them to be in this passage. But if that be the case it is not because the social sciences have realized their long aspiration to become as physics, but because the latter has pushed the boundaries of its knowledge to a point where its problems have become similar to those which social scientists have always had to meet. One interesting explanation of the current use of statistical statements in these areas of physics is that it is not necessary for the data to be more precise than this for any of the purposes which the physicist has at present in mind. This, of course, assumes a pragmatic criterion of adequacy which, fundamentally, is not different from that employed in the social sciences.

²⁸ The crucial character of these alternative approaches to social and economic theory has been very clearly demonstfated by Parsons, T., The Structure of Social Action (New York, 1937).Google Scholar

²⁹ It is the element of purpose which is the significant feature of the vitalist view of human activity. This is the essence of what is involved in the refusal to regard living organisms as nothing more than highly complex mechanical actions. It will readily be seen that the vitalist principle of purposiveness is essential to the epistemological argument put forward in this chapter and to the pragmatic philosophy in general. In fact, pragmatism in its simplest terms may be regarded as nothing more than an elaboration and development of the vitalist view of the phenomenon of life. The vitalist versus the mechanical view of human and social phenomena is one of the oldest debates in the history of intellectual endeavour. For some of the principal features of the controversy, see F. Kaufmann, Methodology of the Social Sciences (London, 1944), chap. VIII.

³⁰ The necessity for patterning the study of social phenomena on the methods of physical science was one of the primary principles of the work of Auguste Comte, founder of modern positivism and father of the discipline of sociology. See, for example, Hutton, H. D. (ed.), Comte: Early Essays on Social Philosophy (London, 1911).Google Scholar This aim has continued to characterize positivistic social science but it has not been supported by a continuing interest in epistemology. References to the more recent epistemological discussions are almost totally absent from the current literature of economics. Even works which deal directly with the question of method, such as the large collection of essays sponsored by the Social Science Research Council (Methods in Social Science, 1931) contain only passing references to epistemology, and the great bulk of these seem to be the result, not of direct study of that subject, but of second-hand derivations from the physical sciences.

³¹ For example, “Practical businessmen have always ridiculed descriptions of business activities that make use of … marginal analysis.” W. Eiteman, Price Determination, p. v. Eiteman's argument that marginal analysis is invalid because business men do not think in such terms is a case of what William James called the “psychologist's fallacy”—“the assumption that the mental fact studied must be conscious of itself as the psychologist is conscious of it.” James, W., Principles of Psychology (London, 1910), vol. I, p. 197.Google Scholar

³² For example, “The only significance of the equilibrium concept for realistic price theory is that it offers a basis for the prediction of the direction of change when equilibrium is not established. Long before a static equilibrium has actually been established, some dynamic change in the fundamental factors will ordinarily occur which will make quantitative changes in the conditions of equilibrium. The ordinary economic situation is one of disequilibrium moving in the direction of equilibrium rather than of realized equilibrium.” Viner, J., “Cost Curves and Supply Curves,” Zeitschrift für Nationalökonomie, vol. III, 1932, p. 29.Google Scholar

³³ J. M. Keynes made frequent references in his writings to the nature of theory as a tool of cognition and to the implications of that view which I have tried to set forth in the past few pages. For example: “The theory of economics does not furnish a body of settled conclusions immediately applicable to policy. It is a method rather than a doctrine, an apparatus of the mind, a technique of thinking which helps its possessor to draw direct conclusions.” (Introduction to the “Cambridge Economic Handbooks” series.) “The object of an analysis is, not to provide a machine, or method of blind manipulation, which will furnish an infallible answer, but to provide ourselves with an organized and orderly method of thinking out particular problems.” General Theory of Employment, Interest and Money (London, 1936), p. 297.Google Scholar

³⁴ Bridgman, P. W., The Logic of Modern Physics (New York, 1937), p. 5.Google Scholar (Bridgman's italics.) For an expression of the same idea in a science closer to economics, see Pratt, C., The Logic of Modern Psychology (New York, 1939).Google Scholar

³⁵ Hansen, A., “The Robertsonian and Swedish Systems of Period Analysis,” Review of Economics and Statistics, 02, 1950.Google Scholar

³⁶ Langer, S., Philosophy in a New Key, (Mentor Books) p. 67.Google Scholar

³⁷ Peirce's first expression of the concept was in a now-famous article in Popular Science Monthly (1878) entitled “How to Make Our Ideas Clear.” The ideas there expressed have however been variously interpreted, the operationists of the Bridgman school as well as modern pragmatists both claiming Peirce's authority for their views.

³⁸ William James referred to the “functional value” of concepts in putting forward this viewpoint. See, for example, “Percept and Concept—The Import of Concepts,” in his Some Problems of Philosophy (London, 1911).Google Scholar John Dewey has phrased the principle clearly in the following passage: “What is meant by calling facts operational? Upon the negative side what is meant is that they are not self-sufficient and complete in themselves. They are selected and described, as we have seen, for a purpose, namely statement of the problem involved in such a way that its material both indicates a meaning relevant to resolution of the difficulty and serves to test its worth and validity. In regulated inquiry facts are selected and arranged with the express intent of fulfilling this office. They are not merely results of operations of observation which are executed with the aid of bodily organs and auxiliary instruments of art, but they are the particular facts and kinds of facts that will link up with one another in the definite ways that are required to produce a definite end. Those not found to connect with others in furtherance of this end are dropped and others are sought for. Being functional, they are necessarily operational. Their function is to serve as evidence and their evidential quality is judged on the basis of their capacity to form an ordered whole in response to operations prescribed by the ideas they occasion and support. If “the facts of the case” were final and complete in themselves, if they did not have a special operative force in resolution of the problematic situation, they could not serve as evidence.” Logic, the Theory of Inquiry (New York, 1938), p. 113.Google Scholar

³⁹ Wicksteed, P. H., The Commonsense of Political Economy, revised edition (London, 1933), vol. I.Google Scholar

⁴⁰ One must recognize the danger that the coherence theory of meaning, when too strictly interpreted, would clamp all empirical knowledge into a predetermined mould and refuse recognition to any empirical facts which did not fit. Such a procedure, though not unknown in the history of science, is of course, nonsensical. Rendering the reality intelligible does not mean squeezing it into such a mould; it means the construction of a theoretical structure which permits the reality to be intellectually grasped, is itself a coherent structure, and squares with facts which are derived in such a way that there is some reasonable degree of freedom in their emergence. In practice this is largely a question of recognizing at what level the facts are clamped by the theoretical structure, and at what level freedom of emergence exists. Thus, for example, the statistics of national income do not prove the proposition that National Income is equal to National Expenditure; for no such freedom exists at that level of the data. The portions of income spent on the various expenditure categories can however be treated as empirical facts possessing freedom of emergence (within the limitation exercised by the definitions which are employed). Because of the fundamental nature of theoretical constructions, certain facts will always emerge which do not fit and their existence is a constant reminder that theoretical explanations are never perfect or complete. It is however a rather limited view of science which insists upon complete accord between theoretical explanations and facts.

⁴¹ Russell, B., The Problems of Philosophy (Home University Library, 1912), pp. 191 ff.Google Scholar

⁴² Lewis, C. I., An Analysis of Knowledge and Valuation, pp. 256–7.Google Scholar

⁴³ Dillard, D., “Methodology in Modern Economic Thought,” American Economic Review, 12, 1944.Google Scholar