In response to:

Discovering Karl Popper from the May 2, 1974 issue

To the Editors:

Singer’s essay on Karl Popper (NYR, May 2) confirms my impression that Popper’s sole contribution (and Singer suggests no other) to the philosophy of science is the argument Popper published in 1944-1945 that the laws of science should be under- stood to rest not on the “principle of verification” but on the “principle of falsifiability,” i.e., that a truth of science cannot be based on the numerous observations that seem to confirm it but rather its acceptability as law must rest on the claim that inferential predictions within the range of present conception have been attempted and have so far failed to refute that law. From this it follows that in order to qualify as “scientific” a statement must be so formulated that it can be refuted through the test of such prediction, i.e., that through a chain of logical deductions, it can be made to refer to a universe of specific events so that it can be definitively refuted and rejected if it fails to predict any one of these.

This argument makes good sense; certainly it shaped my own thinking when I worked with Popper some thirty years ago. Still, that as his lifelong contribution it can be taken as the basis for proclaiming Popper to be the world’s greatest living philosopher of science, reflects not well on the quality of the philosophy of science.

Singer bills the same principle of falsifiability as an attack on logical positivism. It seems to me rather an eddy in the current of positivist thought. Popper did not propose his principle simply as a preferred criterion in evaluating scientific truth. He postulated it rather as the key to the actual, historical development of science, and in this, his conception of the process of scientific development is entirely in accord with, indeed gives a new logical force to, the positivist image of a unilinear unfolding of scientific truth through the ceaseless and steady application of observation and logic, now with hypotheses and laws discarded at the instant they encounter the limit of their predictive powers to be (somehow)’ replaced by more powerful hypotheses and more encompassing laws, until these in turn are refuted through a prediction that fails, hence discarded, replaced, and so on and on, forever, amen.

Singer brings Thomas Kuhn into the picture through a brief mention of Kuhn’s contribution to a symposium The Philosophy of Karl Popper, wherein Kuhn challenges the internal consistency and coherence of Popper’s principle of falsifiability. While Kuhn’s argument is impeccable, some will read it as philosophical nit picking. What Singer fails to say and what must be understood is that Thomas Kuhn in his Structure of Scientific Revolutions heralds a conception of science and of its development that is fundamentally opposed to Popper’s and to the whole positivist approach. Of the Popper-Positivist image of scientific law replacing scientific law through a relentless march of test and refutation in an unbroken and unilinear point by point evolution of the great edifice of truth, Kuhn would hold that it doesn’t and has never happened that way at all; that the sciences operate within “paradigmatic” structure consisting of a set of related generalizations and theories, attitudes and expectations, conventional assumptions and experimental models. Such structures are tough, resistant to change and exist so to speak as islands in a sea of anomalies, i.e., of the contradictory and inexplicable. When there is a significant (revolutionary) advance in science, it is through the discontinuous shattering and displacement of a total paradigm under a complex of pressures that neither Kuhn or anyone else has yet explained. One cannot accept Kuhn’s (or, say, Michel Foucault’s) conceptualization of the development of science and at the same time give credence to Popper’s. In Kuhnsian terms, Popper’s contribution is a last elaboration upon a positivist paradigm ready for discard.

The force of Popper’s argument against Hegel, Marx, or Freud is that their theories are not formulated so that they can be refuted through a specific inferential prediction. Nor is it Popper’s position that because these theories cannot be subjected to the kind of testing that is supposed to be the rule in physics, that they might be accepted as credible on other grounds. Rather he would seem to hold that if they cannot be accepted as science they must be dismissed as nonsense.

Popper was installed at the London School of Economics under the patronage of Lionel Robbins. Robbins, then Professor of Economics and the great power broker at the School, was and is a ferocious ideologue of laissez faire liberalism, of the sort we associate here with Milton Friedman at the University of Chicago. It would seem that Robbins brought Popper and Von Hayek to LSE as intellectual commandos for an assault against the left. It is understandable perhaps, though shameful nevertheless, that even after Karl Popper had acquired such great prestige as to be independent of the goodwill of his patron, evidently he never dared to make it clear that if, under the principle of falsifiability, the theories of Marx and Freud are unscientific and should be discarded, that under the same criterion, the theories of his patron are likewise unscientific and should be discarded. Even more so than that of Marx, the classical and neoclassical general equilibrium theories of economics (of Ricardo, of Mill, of Marshall, of Hicks, et al., and indeed Keynes’s General Theory as well) are all so formulated as to be removed from the possibility of refutation through any specific prediction.

Robert Solo

Michigan State University

East Lansing, Michigan

Peter Singer replies:

I did not describe the principle of falsifiability itself as an attack on logical positivism. Popper did criticize the positivists’ verification principle, on the ground that it was itself unverifiable. This was an objection to the positivists because their principle was supposed to distinguish meaningful utterances from meaningless ones. Popper’s principle of falsifiability was not supposed to do this job at all; it was intended to demarcate the area of scientific inquiry from other, unscientific but perhaps still meaningful concerns. I do agree, however, that there is a sense in which Popper can himself be seen as a continuation of some aspects of positivist thinking—though these aspects are not especially characteristic of positivism itself, rather of a very widespread attitude to science.

Has Kuhn challenged this whole attitude to science? In some ways, he has, but he has himself been considerably influenced by Popper, an influence he has acknowledged in print. The difficulty in assessing Kuhn’s work is that it is not clear to what extent he is describing what has in fact happened in science, and to what extent he is setting limits to what is possible for science. If he is only doing the former, the incompatibility with Popper is not so marked, since they are interested in different issues; but if the latter, then I do not think Kuhn has successfully made out his case. In any case, there has now been so much written on this topic that it can hardly be discussed briefly. (See especially I. Lakatos and A. Musgrave [eds.], Criticism and the Growth of Knowledge, Cambridge University Press, London and Boston, 1970).

It is precisely because Popper does not use his principle of falsifiability as a principle distinguishing sense from nonsense (an interpretation he has rejected in print several times) that he is not committed to holding that if the theories of Hegel, Marx, or Freud cannot be tested they must be dismissed as nonsense.

As for the present quality of philosophy of science, and Popper’s obligation to attack Robbins and other economists, readers may draw their own conclusions.

This Issue

August 8, 1974