Popper and Hayek on Reason and Tradition

Two Types of Rationalism

After the publication of The Road to Serfdom and Open Society, Hayek and Popper discuss rationalism on numerous occasions—in their publications, in their correspondence, and, presumably, in personal conversation. This was a sequel to their wartime correspondence, in which one of the arguments was whether it was possible to intervene in society. Hayek deals with rationalism, for instance, in “Individualism: True and False” (1945), The Constitution of Liberty (1960), and “Kinds of Rationalism” (1964). He mainly criticizes (particularly in “Individualism: True and False”) what he considers to be the wrong kind of rationalism, which he associates with Descartes, the French Encyclopaedists, Rousseau, and the physiocrats. In Logik der Forschung (1934), Popper had developed his philosophy of science in which rationality consists in provisionally accepting the hypotheses that have best stood up to severe criticism. This has later been generalized into critical rationalism. Popper had extensively discussed the birth of the tradition of rationalism in ancient Greece in his Open Society, but he had never written about the possible conflict between reason and tradition. Yet in 1948 Popper gave a lecture with the title “Towards a Rational Theory of Tradition” in which he does just that, in addition to, among other things, linking a discussion of rationalism with that of the method of social science. Hayek had done so in the articles that he published during the 1940s, which were published as The Counter Revolution of Science in 1955. In the light of Popper’s previous publications, however, the discussion of a possible opposition between reason and tradition and the link between reason and the method of the social sciences make the article rather unusual, and for a long time I have wondered why Popper wrote it. ⁶ The puzzle began to make sense once I started to realize that it is a criticism of Hayek’s ideas on the role of tradition.

What pleads against my thesis, however, is that Popper never mentions Hayek. Instead, he directs his criticism at an author that Hayek often approvingly refers to, Edmund Burke. As I have observed, there is no doubt that Popper and Hayek closely interacted with and admired each other. Their mutual esteem is confirmed by Popper’s dedicating Conjectures and Refutations (1963) to Hayek and by Hayek’s dedicating Studies in Philosophy, Politics and Economics (1967b) to Popper. That is no convincing evidence that Popper (1948) is a criticism of Hayek. It is, however, part of a complex of arguments that make my thesis plausible, as I hope to show.

Hayek identifies rationalism with the Cartesian tradition, in which he includes Saint Simon and Comte. He attributes to them the idea that society can be shaped consciously according to our wishes, an attitude that he refers to as constructivism, or the engineering approach to society. Hayek bases his criticism on the argument that individual actions have unintended social consequences. So rationalism is a negative label for Hayek; the correct social philosophy is antirationalism. This is argued, for instance, in “Individualism: True and False” (Hayek 1945), which I have already mentioned. Instead of the constructivist-rationalist approach to society, Hayek emphasizes the importance of

the traditions and conventions that evolve in a free society and which, without being enforceable, establish flexible but normally observed rules that make the behavior of other people predictable in a high degree. The willingness to submit to such rules, not merely so long as one understands the reason for them but so long as one has no reasons to the contrary, is an essential condition for the gradual evolution and improvement of rules of social intercourse; and the readiness ordinarily to submit to the products of a social process which nobody has designed and the reasons for which nobody may understand is also an indispensable condition if it is to be possible to dispense with compulsion. (23)

Here Hayek depicts rationalism and tradition as mutually incompatible. In “Towards a Rational Theory of Tradition,” Popper develops the contrary argument that they are complementary. My hypothesis is that Popper’s article (1948) is a criticism of Hayek, which, moreover, summarizes the discussions they had in correspondence and face-to-face. This is what he says about “the problem of tradition” (it is a problem for rationalists):

The anti-rationalists in the field of politics, social theory, and so on, usually suggest that this problem cannot be tackled by any kind of rational theory. Their attitude is to accept tradition as something just given. You have to take it; you cannot rationalize it; it plays an important role in society, and you can only understand its significance and accept it. (120)

Even though Popper refers to Burke, it is not difficult to recognize Hayek in this description. On the next page, he writes,

I feel that there does exist an anti-rationalist reaction of a serious kind and among very clever men, and that it is connected with this particular problem. Quite a number of outstanding thinkers have developed the problem of tradition into a big stick with which to beat rationalism. (121)

Popper mentions Michael Oakeshott, but I suggest that, once again, his criticism is (at least in part) aimed at his friend Hayek, whom he deeply admired (“clever man,” “outstanding thinker” ⁷ ) but perhaps did not want to criticize in public (I say more about this in the conclusion).

Popper (1948) proposes a more balanced view of tradition than the antirationalism advocated by Hayek, Burke, and Oakeshott. Instead of the uncritical acceptance of tradition, he proposes another possibility: “a critical attitude, which may result either in acceptance or rejection, or perhaps in a compromise” (122). Immediately applying what he has said to the tradition of rationalism itself, Popper introduces two arguments as examples—but I argue in the conclusion that there is perhaps more to them.

I should say that in our invaluable rationalist tradition (which rationalists so often accept too uncritically) there are quite a few points which we ought to challenge. A part of the rationalist tradition is, for example, the metaphysical idea of determinism. . . . Another element in the rationalist tradition which we should question is the idea of observationalism. (122-23)

Let us return to the main line of Popper’s argument. He refers to the “conspiracy theory of society,” which he had introduced in Open Society and which he says is held by many rationalists: the idea that social phenomena are the intended results of individual objectives and actions. He argues—exactly as Hayek does in Counter Revolution of Science—that this theory is unsustainable because individual actions almost always have unintended effects. But he introduces an element that signals a divergence from Hayek’s ideas: “The characteristic problems of the social sciences arise only out of our wish to know the unintended consequences, and more especially the unwanted consequences” (Popper 1948, 124). Now, of course Hayek, too, recognizes the possibility of unwanted consequences, but he argues that we ought to accept them. If we do not, we risk destroying the spontaneously evolved social order, which contains the accumulated experience of the past. For Hayek, rationality consists in accepting tradition. I call this position conservative rationalism. ⁸

Popper then draws a parallel between traditions and scientific theories, saying that both serve “to bring some order into the chaos in which we live so as to make it rationally predictable” (131). But the analogy goes further: theories and “traditions have the important double function of not only creating a certain order or something like a social structure, but also giving us something upon which we can operate; something that we can criticize and change. The point is decisive for us, as rationalists and as social reformers.” (131; my emphasis). Popper’s openness to the possibility of changing social institutions is consistent with what he observes about the objectives of social science. Their task is not only to explain unintended consequences but also to discover the functions of social institutions. This functionalism is combined with the theme of social control:

The working of institutions, as of fortresses, depends ultimately upon the persons who man them; and the best that can be done by way of institutional control is to give a superior chance to those persons (if there are any) who intend to use the institutions for their “proper” social purpose. (133-34)

Popper’s functionalist approach and his idea that we can discover what the proper objectives of social institutions are ⁹ draw the attention to a fundamental problem in Hayek’s social philosophy. It consists of the duplicity between what we as social theoreticians may know about society and do with that knowledge, on one hand, and what the individuals who constitute society—presumably including the very same theoreticians—know and how they act on that knowledge, on the other.

“Kinds of Rationalism” of 1964 is one of many publications in which Hayek contrasts Cartesian constructivist rationalism with what he no longer labels antirationalism, as he had done in “Individualism: True and False,” but more prudently circumscribes as the type of rationalism that recognizes the limits to the powers of human reason. He writes:

Among contemporary philosophers it is particularly Professor Karl R. Popper who has provided important new philosophical foundations for this strand of thought. He has coined for it the name “critical rationalism.” . . . It seems to me the best term for describing the general position which I regard as the most reasonable one.

So, Hayek says that his own position is identical with Popper’s. A closer reading of the article, however, raises the suspicion that he was responding to some of the criticisms of Popper that I have just discussed.

Hayek elaborates the consequences of the fact that society is complex. This is the objective counterpart of the limited character of our subjective knowledge. He says that he has discovered that classical liberalism is

an instance of a general method of indirectly creating an order ¹⁰ in situations where the phenomena are far too complex to allow us the creation of an order by separately putting each element in its appropriate place. It is a sort of order over the particular manifestation of which we have little control, because the rules which determine it determine only its abstract character, while the detail depends on the particular circumstances known only to its individual members. It is therefore an order which we cannot improve upon but only disturb by attempting to change by deliberate arrangement any one part of it. The only way in which we can effectively improve it is by improving the abstract rules which guide individuals. (Hayek 1964, 92; my emphasis throughout).

What Hayek does here is to apply the results of his theory of cultural evolution, which he had developed it in “Rules, Perception and Intelligibility” (1963b) and “Notes on the Evolution of Systems of Rules of Conduct” (1967a). Part of the complexity that we face in social affairs is due to the complementary character of the (mostly implicit) rules of behavior that constitute a particular culture. This implies that we cannot change one rule without having to change others. I have discussed Hayek’s theory of cultural evolution elsewhere (Birner 2009), including Popper’s influence on it and the influence that it—and, particularly, the idea of group selection—exerted on his own theory of evolution (specifically, his theory on the evolution and selective power of ecological niches). I briefly come back to this in the conclusion.

I would like to draw the attention to some difficulties that are inherent in Hayek’s idea that we can improve a social order only by improving its abstract rules.

First, this seems to me to be inconsistent with Hayek’s identification of his own position with Popper’s piecemeal engineering. Popper understands by piecemeal engineering not the changing of abstract rules but small concrete changes that act as social experiments, such as “the introduction of a new kind of life-insurance, of a new kind of taxation, of a new penal reform. . . . Even a man who opens a new shop, or who reserves a ticket for the theatre, is carrying out a kind of social experiment on a small scale” (Open Society, 1:162). ¹¹

Second, Hayek’s advocacy of the changing of abstract rules of behavior seems more coherent with the large-scale social engineering that he has always rejected than with a more modest approach to social change that is respectful of spontaneously grown institutions.

Third, to change the abstract rules, we must first know what they are. But if individuals know only the concrete circumstances of their immediate environment, as Hayek has maintained time and again, how can they know the abstract—and largely tacit and implicit—rules of behavior? Their discovery would be the task of the social scientist. But what makes social scientists less bound than others to knowledge of the concrete circumstances of their environment? The analysis of this last problem has to take into account Hayek’s philosophy of science, which is discussed in the last paragraph. I would here like to return to Hayek’s position with respect to rationalism.

Popper, Hayek, Hume

Both Popper and Hayek count David Hume among their intellectual ancestors. Hayek adopts, it seems, the whole of Hume’s thought, from his epistemology to his monetary theory and his legal and political philosophy (cp. Hayek 1963a). Popper adopts Hume’s logical criticism of induction, but he criticizes his conclusion that induction, though logically impossible, is psychologically necessary. According to Popper (1972, 95), this “led Hume, one of the most reasonable thinkers of all time, to give up rationalism and look at man not as endowed with reason but as a product of blind habit.” Popper argues that Hume was right in thinking that there is no solution to the logical problem of induction. But he was wrong to think that induction is psychologically necessary. From this paradox, as Popper calls it (95), Hume draws the wrong conclusion that man is irrational and a product of “blind habit” (95).

The solution of the paradox is that not only do we reason rationally, and therefore contrary to the principle of induction. We do not act upon repetition or “habit,” but upon the best tested of our theories which, we have seen, are the ones for which we have good rational reasons; not of course good reasons for believing them to be true, but for believing them to be the best available from the point of view of a search for truth or verisimilitude. (95)

But he warns against a false conclusion: “This solution of Hume’s paradox does not, of course, say that we are thoroughly rational creatures. It only says that there is no conflict between rationality and practical action” (95). Popper had expressed the same thought in a letter to Hayek (11 November 1958): ¹² “And as to Hume, his irrationalism was based, exclusively, upon his correct finding that induction is impossible—exactly as Polanyi’s anti-rationalism. But this perfectly correct finding has no anti-rationalist consequences—except if you are (as Hume and Polanyi) a disappointed inductivist.” ¹³ This falls little short of suggesting that Hayek drew the same conclusion.

Hume’s thought contains the key to the differences between Hayek’s and Popper’s research programs. ¹⁴ The main elements of Hume’s philosophy are empiricism, skepticism, conservatism, and the logical critique of induction. Popper adopts Hume’s logical critique of induction and his skepticism, or a skeptical version of empiricism (observations serve only to test and not to justify theories), which are the corner stones of his critical rationalism. In Hayek’s thought, however, an observationalist variant of Humean empiricism is combined with conservatism. To show this, I must return to a discussion of Hayek’s very first work on the philosophy of mind and to his philosophy of science, which are closely related.

Hayek as a Radical Thinker

Right from its origins, Hayek’s “grand research programme” ¹⁵ is characterized by a feature of his thought that has been little noticed: its radicalism. Once Hayek has defined an area of research, or, as I prefer to call it, one of his partial research programs, he carries it out rigorously to its ultimate logical consequences. ¹⁶ Other examples are his radical antisocialism, which goes so far as to reject all nonliberal political ideologies, including social democracy, and his radical rejection of social justice as a concept deprived of meaning. ¹⁷ But the first instance of Hayek’s radicalism is his theory of mind, in which he derives consequences from Schlick’s epistemology that go beyond Schlick’s own. All its main features are contained in the unpublished manuscript of 1920 “Beiträge zur Theorie der Entwicklung des Bewusstseins,” which he elaborated into The Sensory Order (1952). ¹⁸ In both the manuscript and the book, Hayek draws the epistemological conclusions that he thinks follow from his theory of mind, which is why a discussion of this theory is in order here. In what follows, I quote the passages in Sensory Order that Hayek took over from the manuscript.

Radical Empiricism

Though Hayek adopted Schlick’s empiricism, he felt that it failed to solve the problem posed by Kant—namely, that we need a preexisting mental framework that makes consciousness and knowledge about the world possible:

There is . . . a part of our knowledge which, although it is the result of experience, cannot be controlled by experience, because it constitutes the ordering principle of that universe by which we distinguish the different kinds of objects of which it consists and to which our statements refer. (The Sensory Order, 8.18)

But rather than criticizing empiricism from the outside, Hayek says that he has led it to its logical conclusion:

In so far as we have been led into opposition to some of the theses traditionally associated with empiricism, we have been led to their rejection not from an opposite point of view, but on the contrary, by a more consistent and radical [sic] application of its basic idea. Precisely because all our knowledge, including the initial order of our different sensory experiences of the world, is due to experience, it must contain elements which cannot be contradicted by experience. (The Sensory Order, 8.27) ¹⁹

Radical Epistemological Structuralism, Naturalism, and Functionalism

Hayek’s theory of mind leads him to a radical epistemological structuralism, naturalism, and functionalism. To see this, let us examine what Hayek says about scientific explanation in Counter Revolution of Science, which was directly influenced by his work in the philosophy of mind. According to Hayek, “to explain” means to reclassify our sense impressions until we are left with an abstract system of relationships without any sensory qualities. We are left with “pure structure” without content, for the description of which we need mathematics. The process of scientific explanation as Hayek describes it is the same as the process by which the human mind forms classifications of the world (Counter Revolution of Science, 20). Moreover, the functions of the mind could be realized in any physical substrate that has the characteristics of a classifier system and is not necessarily restricted to the brain’s physical multilayered neural networks. ²⁰

He hints at how we can use the theories of physics to make predictions: “their significance is due to the fact that we possess rules, a ‘key,’ which enables us to translate them into statements about perceptible phenomena” (Counter Revolution of Science, 21). Presumably, Hayek has in mind the correspondence rules that were discussed in philosophy of science texts of the 1960s. ²¹ What Hayek writes about explanation is an elaboration of some of the ideas with which he concluded his manuscript of 1920 and which found their way into the last chapter of The Sensory Order: when the mind has completed its work, we are left with a system of definitions or tautologies. Since tautologies are necessarily true, we must accept them. That leads Hayek to another radical conclusion (one that I have already mentioned).

Radical Realism

Hayek seems to argue that the tautological nature of our “finalized” knowledge of the physical world makes its acceptance compulsory:

Its concern [i.e., of “Science”] is not what men think about the world and how they consequently behave, but what they ought to think. The concepts which men actually employ, the way in which they see nature, is to the scientist necessarily a provisional affair and his task is to change that picture, to change the concepts in use so as to be able to make more definite and more certain our statements about the new classes of events. (Counter Revolution of Science, 22; emphasis added)

Apparently, “Science” will produce the only valid classification possible, which therefore ought to be accepted. As I have argued above, this applies only to the explanation (or, rather, classification) of the physical order; the explanation of the social order, or the domain of the “moral sciences”, must always remain incomplete to us. That is because our minds not only explain but also constitute this order; human minds are part of both the explanantia and of the explananda:

Until Science has literally completed its work and not left the slightest unexplained residue in man’s intellectual processes, the facts of our mind must remain not only data to be explained but also data on which the explanation of human action guided by those mental phenomena must be based. (Counter Revolution of Science, 24)

Given the self-referential character of the social sciences described above, the argument is to be considered to be a reductio ad absurdum.

Kant’s Question and the Theory of Mind

I have argued that Hayek and Popper have different stances vis-à-vis rationalism. This can also be put in terms of their different positions vis-à-vis Hume. But even more fundamental are the different answers they give to Kant’s question about the possibility of knowledge. This contrast can be traced back to the origins of their intellectual careers. The very first scientific interest of both was the explanation of the human mind. In the early 1920s, Hayek is a rigorous follower of what became the ideals of the Vienna Circle, expounded in Schlick’s Allgemeine Erkenntnislehre. He puts on his research agenda the development of a theory that explains the mind (or consciousness) exclusively in terms of the laws of physics. ²² Popper, on the contrary, rejects Schlick’s physicalism and follows the psychology of the Würzburger Schule. Hayek and Popper’s ways part right here. Nevertheless, their ideas—or, rather, their research programs—develop in strikingly parallel ways. They publish their ideas on the free or open society and on the ideologies that threaten its existence at the same time. They elaborate their methodologies of social science in the same period. They elaborate their ideas on evolution in the same years. Part of the parallels can be explained by the fact that Hayek and Popper interacted with each other ever since they became colleagues at the London School of Economics in 1946. But even before that date, they discovered that they had common intellectual interests: both worked on problems related to the functioning of society and the correct method to deal with them. ²³

Hayek’s solution to Kant’s formulation of the problem of the possibility of knowledge is worked out in his neural-network theory of the mind, which he himself considers to be a radical elaboration of empiricism and which respects the methodological principle of the Vienna Circle that everything should be explained in terms of the scientific laws governing the physical world. Popper (1952), however, criticizes and corrects Kant’s question before answering it:

His question, we now know, or believe we know, should have been: “How are successful conjectures possible?” And our answer, in the spirit of the Copernican Revolution, might, I suggest, be something like this: “Because, as you said, we are not passive receptors of sense data, but active organisms. Because we react to our environment not always merely instinctively, but sometimes consciously and freely. Because we not only invent myths, stories and theories, . . . but try them out and see whether they work and how they work.” (95)

The 1920 manuscript on the theory of mind that forms the core of The Sensory Order contains a number of ideas that influenced the development of Hayek’s research program. They suggested ²⁴ to Hayek a series of models, theories, or images of the nature of knowledge, of the economy, and of society as a whole that are “mind-like” or “(neural) network-like.” In the case of his philosophy of science there is a direct link. It derives from Hayek’s neural-network theory of the brain, which describes the physical processes (Popper would say, the world-1 processes) underlying human consciousness. We as subjective individuals acquire knowledge about the world through a process of reclassification of our sense impressions that, if allowed to pursue its full course, would lead to a definitive system of certain knowledge, consisting of nothing but a set of tautological definitions, a purely structural image of the world without empirical content—as far as the physical order is concerned. Hayek’s suggestive image of the economy is that of a decentralized structure of individual units—each of which is equipped with a brain—that have limited knowledge of their immediate environment. The image is rather similar to that of the structure of the brain itself: a network of neural networks. Later, Hayek generalizes this image of the economy into a model of society as a whole.

Conclusion

Hayek’s epistemology shows a number of striking differences with Popper’s. For Hayek, knowledge is a system of classifications, whereas for Popper, it is more: knowledge is a set of descriptive theories that are also explanatory. In Hayek’s epistemology, knowledge is tautological or definitional, whereas in Popper’s, it is contingent. According to Hayek, certain knowledge is in principle attainable, whereas Popper rejects the idea of certain knowledge. For Hayek, knowledge is subjective, a world-2 phenomenon that is produced by the world-1 physical processes of the brain, whereas for Popper, knowledge arises out of the interaction of our subjective, world-2 activities with the objective world-3 contents of thoughts, ideas, and theories that obey regularities that are independent from us. For Hayek, knowledge is something that “happens to us” through physical brain processes, whereas for Popper, knowledge is a product of our active interventions and conscious efforts to discover it. Hayek’s epistemology and methodology are based directly on his theory of the mind and hence may be called naturalistic, whereas Popper maintains a strict separation between context of discovery and matters of fact, on one hand, and context of justification and questions of logic, on the other. Hayek’s epistemology has a strong deterministic flavor, whereas Popper rejects determinism. As we have seen above, Popper is particularly critical of the uncritical acceptance by many in the rationalist tradition of determinism and observationalism.

These fundamentally different epistemological ideas are the source of all the further differences that arise in the course of the development of Hayek’s and Popper’s research programs. They carry over into their ideas on the place of social science in society. As to social institutions, Hayek advocates the conservative principle “if it ain’t broke, don’t fix it.” Popper, on the contrary, is of the conviction that we are not only “auf der Suche nach einer besseren Welt” (in search of a better world) ²⁵ but also capable of, and responsible for, improving it. Hayek is opposed to individual irrational man, with his limited knowledge interfering with the social processes that produce “Reason” (the argument of “Individualism: True and False”). He advocates a conservative rationalism. Popper is more optimistic about the power of individual reason, which, if wielded judiciously and in a piecemeal way, allows us to learn from our mistakes and construct a better society. This is the crux of Popper’s critical rationalism. The concept of the individual who, guided by ideas, actively intervenes in his environment has also found its way into Popper’s theory of the evolution and selective powers of ecological niches in nature and culture. This is very different from Hayek’s idea that “the brain is an organ enabling us to absorb, but not to design culture” (Hayek 1978, 8). ²⁶

That leaves me with a last issue. It concerns the remarkable fact that, despite their blatant and fundamental differences, Hayek and Popper hardly ever criticize each other explicitly in their publications. ²⁷ This would deserve a separate article, but let me just indicate a couple of factors that may explain this. That Popper refrained from criticizing Hayek, except in veiled terms, in his publications has no doubt to do with Popper’s profound sense of gratitude for his friend: Hayek had helped him to escape from Austria before the war; Hayek had provided him the position at the London School of Economics; and Hayek had been instrumental in getting The Poverty and Open Society published. ²⁸ Ever since the publication The Road to Serfdom and Open Society and Poverty, Hayek and Popper felt that they had to join forces against a common enemy. As far as they were aware of their differences, they did not judge it opportune to discuss them in public. Furthermore, despite a number of fundamental philosophical differences that separated them, Hayek and Popper had stimulating effects on each other’s thought. Popper and Hayek did not always understand their own ideas, their implications, and the incompatibilities with the ideas of the other. The instances in which their interaction produced negative results is a perfect illustration of Popper’s idea that nobody is an authority on his own thought. Popper was probably more critical of Hayek—particularly of his theory of mind ²⁹ —than the other way around. That does not change the fact that, despite some negative mutual influences and internal inconsistencies, their research programs have produced some of the most important ideas of the twentieth century.