II. THE PROBLEM OF THE PRIMARY LITERATURE, PART 1: THE APPARENT INCONSISTENCY OF MISES’S WRITINGS ON ECONOMIC METHODOLOGY

The chief issue at stake in the literature on Mises’s apriorism concerns not the extent of his apriorism—i.e., not how many or how few of the fundamental propositions of economic theory he thought to be a priori—but the epistemological status of the so-called action axiom that constitutes the theoretical foundation of praxeology (Scheall Reference Scheall2017b), Mises’s general science of human action, the “best-developed part of” which is economics (Mises [1949] Reference Mises1998, p. 237). The action axiom states that human action is purposeful behavior (Mises [1949] Reference Mises1998, p. 11). Mises made several claims about the epistemology of human action that have struck many readers over the decades as extreme and dogmatic, if not unscientific.

I have described two of these extreme epistemological claims elsewhere as the “Reason without Experience” and “Greater Certainty” theses (Scheall Reference Scheall2017b). According to the first, our knowledge of human action is due entirely to the faculty of reason, i.e., to pure reason, intuition, introspection, or reflection on “inner experience,” entirely unaided by prior sensory experience. According to the Reason without Experience thesis, our engagement with the world of “outer experience” is impotent with respect to this knowledge; the external world neither leads us to nor can undermine our knowledge of human action.

According to the second thesis, the Greater Certainty thesis, by virtue of being introspectable, intuitable, etc., our knowledge of human action is more secure, more certain, than knowledge acquired through experience of the external world. Indeed, according to Mises, our knowledge of the purposefulness of human action is absolutely secure or “apodictically certain.” Theorems validly deduced from the action axiom “are not only perfectly certain and incontestable, like the correct mathematical theorems[, t]hey refer, moreover, with the full rigidity of their apodictic certainty and incontestability to the reality of action as it appears in life and history. Praxeology conveys exact and precise knowledge of real things” (Mises [1949] Reference Mises1998, p. 39; emphasis added).Footnote ⁷

Putting these theses together, we get the outlines of a rationalist epistemology according to which it is possible to possess apodictically certain and incontestable knowledge about “real things”—that is, about human action in the external world of outer experience (“in life and history”) without any contact with, or input from, this same world. It is not necessary to excavate far below the surface of Mises’s methodological writings to find what seems a fairly radical rationalist epistemology.

This being said, however, there are other places in relevant texts where Mises made subtle gestures toward more moderate epistemological stances. It is here that the problem of finding an internally consistent reading of Mises’s methodological writings is manifest, for these elusive hints of more moderate epistemological attitudes do not sit comfortably with his strong statements of the Reason without Experience and Greater Certainty theses.

Mises ([1949] Reference Mises1998, pp. 35, 85–86; Reference Mises1962, pp. 14–16) made comments that seem to imply that human knowledge might be subject to evolutionary processes over historical time. Yet, he never explained how his seemingly rationalistic justification of the apriority of the action axiom fit within this sketch of a proto-evolutionary epistemology. If human knowledge evolved over time, it is not immediately obvious how the part of it that concerns human action “in life and history” could have been exempted from the effects of experience.

I do not mean to suggest that a consistent story is impossible to tell here, merely that combining the Reason without Experience and Greater Certainty theses in an evolutionary epistemology is not obviously consistent and, therefore, requires some explanation. Indeed, given the creative ways that intelligent thinkers often find of unifying their seemingly inconsistent ideas, the problem is more that Mises provided no guidance for determining which of the many ways his claims might be made consistent he intended.

Or, again, consider the fact that Mises made strong assertions of human epistemic fallibility (Mises [1949] Reference Mises1998, p. 68), yet never explained why his fallibilism stopped short of the action axiom. Anyone who thinks his knowledge of human action is apodictically certain is manifestly not a fallibilist about this knowledge.

Similarly, Mises seemed favorably inclined to a sort of pragmatism (Mises [1957] Reference Mises2005, p. 248; Reference Mises1962, p. 96) but rejected a pragmatic interpretation of the apriority of the action axiom—it was not, according to Mises ([1949] Reference Mises1998, pp. 39–40, 85; Reference Mises1962, pp. 17–18), merely an assumption made for reasons of scientific convention, ease and convenience, or practical utility.

Thus, the evidence from the primary literature indicates that Mises said some things that suggest a fairly radical rationalism and some other things that undercut to some degree the extremeness of this rationalism.Footnote ⁸ However, these things do not—or, at least, do not obviously—all sit comfortably together. The combination of claims for a priori “exact and precise knowledge of real things” due to pure reason with evolutionary or pragmatist epistemology is somewhat unusual in the history of ideas. Most epistemological fallibilists are fallibilists about all of human knowledge. Limiting his fallibilism, as Mises did, to everything but the action axiom is a non-standard epistemological stance. Some elucidation is in order. Unfortunately, Mises never explained how his radically rationalist statements were, in his mind, consistent with his more moderate statements.

Internal consistency might seem to be the least we can reasonably expect from a theory of knowledge, but I want to suggest the possibility that Mises’s epistemology fails to meet even a weaker criterion: taken in its entirety, Mises’s epistemological commentary is so unclear and, relatedly, as we will see below, the personal circumstances that informed it so uncertain that, ultimately, we do not know what his theory of knowledge was, and cannot tell even whether it was internally consistent or not. Mises’s epistemology fails to meet even a criterion of minimal clarity required to discern its internal consistency.

III. THE PROBLEM OF THE SECONDARY LITERATURE: RECAPITULATING THE INCONSISTENCY OF MISES’S WRITINGS ON ECONOMIC METHODOLOGY

The most prominent line of interpretation in the secondary literature reads the epistemology at the core of Mises’s apriorism as a kind of extreme rationalism. Such a reading can be found in Terence Hutchison (Reference Hutchison1938, Reference Hutchison1981), Murray Rothbard (Reference Rothbard1957), Paul Samuelson (Reference Samuelson1964), Mark Blaug (Reference Blaug1980), Gillis Maclean (Reference Maclean1980), Bruce Caldwell (Reference Caldwell1982, Reference Caldwell1984), Hans-Herman Hoppe (Reference Hoppe1995), Maurice Lagueux (Reference Lagueux, Davis, Hands and Maki1998), Wade Hands (Reference Hands2001), Dimitris Milonakis and Ben Fine (Reference Milonakis and Fine2009), Eugen Schulak and Herbert Unterköfler (Reference Schulak and Unterköfler2011), and Scott Scheall (Reference Scheall2015). Most, but not all, of these interpretations are negative: Rothbard, Hoppe, and Maclean defend, rather than criticize, the apparent extremeness of Mises’s apriorism and, in varying respects and to differing degrees, the Kantian epistemology upon which it seems to have been founded.Footnote ⁹

It is, of course, precisely those strains of moderation mentioned in the previous section—Mises’s gestures toward evolutionary epistemology, fallibilism, and pragmatism—that interpreters take up who wish to save him from accusations of extreme rationalism. Mises’s apriorism has been variously interpreted as assimilable in some degree to Popperian falsificationism (see Champion [manuscript]; Di Iorio Reference Di Iorio2008); to Imre Lakatos’s (Reference Lakatos1968) methodology of scientific research programs (see Machlup Reference Machlup1955; Rizzo Reference Rizzo and Kirzner1983; Zanotti and Cachanosky Reference Zanotti and Cachanosky2015); and to W. V. O. Quine’s (Reference Quine1951) epistemological holism (see Leeson and Boettke Reference Leeson and Boettke2006). Alexander Linsbichler (Reference Linsbichler2017) reads Mises as almost a conventionalist.

Unfortunately for these interpreters of Mises as a merely moderate apriorist, as noted above, he explicitly undercut several of these readings elsewhere in his writings. Those moderate readings that are not explicitly undercut are implicitly undone by extreme statements like the Reason without Experience and Greater Certainty theses.

I have argued that moderate readings of Mises’s apriorism can ultimately be sustained only if the most prominent criticisms that have been set against it in the secondary literature are simply ignored (Scheall Reference Scheall2017b). Interpreters of Mises’s apriorism as moderate do not so much confront and counter these criticisms, or provide a moderate reading of his extreme statements of the Reason without Experience and Greater Certainty theses, as pretend they do not exist. Nevertheless, those furtive hints of a more moderate Misesian epistemology are present in his writings. It is no more legitimate to ignore them simply because they might seem to be dominated by his more rationalistic comments than it is for the interpreters of Mises as moderate to ignore the latter.

We are obligated by the principle of interpretative charity to assume that Mises had some clarifying idea in mind that unified his seemingly inconsistent epistemological statements. However, Mises did not provide this explanation himself. The task of unifying his apriorism was left to his descendants in the Austrian tradition. The variety of conflicting interpretations adds to, rather than diminishes, the confusion surrounding the primary literature. Anyone who sought to understand Mises’s methodology and who found frustration in the master’s own writings would find no solace, and probably only additional frustration, in the commentary upon these writings. In this regard, the secondary literature on Mises’s methodology decidedly fails to fulfill its function of clarifying the primary material.

The evidence from the primary literature suggests the possibility of internal inconsistency. The fact that the secondary literature exhibits such varied and mutually inconsistent interpretations, each supported to varying degrees by reference to Mises’s own texts (or, as we will see below, to certain of his extra-textual remarks), reinforces the impression of the possible internal inconsistency of his epistemology without bringing us closer to a definitive determination. It is of course perfectly normal to find tension among different interpretations in any body of secondary literature, but Mises’s epistemology offers an especially egregious example. Tension in the secondary literature is normal; it is not normal to find in the secondary literature of a coherent thinker mutually exclusive interpretations each more or less equally supported—and undermined—by the relevant texts (or by some extra-textual remark).

We are left in the awkward position of granting to Mises the charitable notion that some unifying conception underlies his seemingly inconsistent assertions, even though the clues we possess regarding this explanation are themselves inconsistent, leading in multiple directions at once. This is another way of saying that, if there is an internally consistent reading that unifies the whole of Mises’s methodologically relevant writings—i.e., an interpretation that does not just emphasize this or that aspect of his writings while ignoring other aspects that undercut the interpretation—we do not know what this interpretation might be.

IV. THE PROBLEM OF THE PRIMARY LITERATURE, PART 2: MAKING SENSE OF MISES WITHIN HIS MILIEU

If it could be shown that Mises’s epistemology was an artifact of his milieu, there might be hope for understanding it by considering the circumstances from which it emerged. This is not possible in Mises’s case. It is difficult to recognize his methodological arguments as artifacts of the historical and geographical context in which they were propagated. The two leading families of interpretations in the secondary literature both treat Mises’s apriorism as radically anachronistic, albeit in opposite senses.

In its extreme interpretation, Mises’s apriorism is, as F. A. Hayek (Reference Hayek1978, p. 137) put it, a relic of “almost eighteenth-century rationalism.” Mises seemed to have missed much of what the nineteenth century added to the history of philosophy and science, at least outside of economics. Mises penned his methodological writings at around the same time and almost literally across the University of Vienna quadrangle from the room in which the members of the Vienna Circle of Logical Positivism were rewriting the philosophy of science. Yet, Mises’s writings on economic methodology reflect a near complete ignorance, or innocence, of the scientific and philosophical problems that led the Viennese Positivists (and many other philosophers of the era) to deny the possibility of synthetic a priori knowledge and to reconceptualize the epistemological significance of analyticity.

This is not the place to recount this history in detail. Suffice it to say that such was the impact of Immanuel Kant’s ideas that the status of a priori knowledge, analytic or synthetic, became one of the central intellectual problems of the nineteenth and early twentieth centuries, occupying philosophers, mathematicians, logicians, and physical scientists. Over the course of the sesquicentenary that separated Kant from Mises, various claims that Kant had made for the apriority of particular domains of knowledge were successively whittled down to effectively nothing by the time Mises was writing. First, in the 1830s, the mathematician Bernard Bolzano showed the inadequacy of Kant’s arguments for the synthetic a priori nature of arithmetic (Coffa Reference Coffa1991, pp. 22–40). Around the same time, several scholars—the Hungarian János Bolyai and the Russian Nikolai Ivanovich Lobachevsky (Reference Lobachevsky1829–1830), and, a bit later, Bernhard Riemann (Reference Riemann1854) —established the possibility of non-Euclidean geometries. Not long after, that giant of nineteenth-century science Hermann von Helmholtz showed that these possibilities call into considerable doubt Kant’s claim for the apriority of the Euclidean nature of our spatial perceptions (Coffa Reference Coffa1991, pp. 47–60). Einstein’s writings on relativity in the first decades of the twentieth century called such claims into further doubt (Coffa Reference Coffa1991, pp. 317–320; also see Reichenbach [1920] Reference Reichenbach and Reichenbach1965).

Mises’s methodological writings do not engage with these developments. There are passages in his writings that indicate he was either not privy to or did not appreciate the significance of many recent (and often local) developments in contemporary philosophy and natural science. It is not that Mises-the-rationalist argued for the possibility of (synthetic) a priori knowledge against the many arguments that had been advanced against it in the previous century and a half; he simply ignored (or was unaware of) them.

Mises ([1949] Reference Mises1998, pp. 72–91) argued against the possibility of multiple logics long after polylogism was a well-established empirical fact. This ignorance or negligence is particularly odd, given the contributions of his Viennese colleagues Karl Menger ([1932] Reference Menger and Mulder1979) and Kurt Gödel (Reference Gödel1930, Reference Gödel1931, Reference Gödel1932) to the relevant literature. Mises knew Menger, in particular, very well. In addition to being a professor of mathematics at the University of Vienna and a member of the Vienna Circle, Menger was the son of the founder of the Austrian School of economics, was active in Viennese economic circles throughout the 1920s and early 1930s (Mises left Vienna in 1934; Menger, in 1937), made scholarly contributions to the school, and, in one of these (Menger [1936] Reference Menger and Mulder1979), even warned Mises that his loose approach to deductive logic failed to meet modern standards of mathematical logic. Menger said, “Mises wrote me that he learned a great deal from the paper” (Menger [1936] Reference Menger and Mulder1979, p. 279). Despite this, Mises ([1949] Reference Mises1998, pp. 72–91) seemed to think that his rejection of racial- and class-based relativizations of logic sufficed to counter polylogism. However, this is an obvious non sequitur: though it was perhaps an admirable stance to defend in the wake of Nazism and Stalinism, it does not follow from the fact that logics are not relative to race or class that they cannot be relative to anything at all.

Recall that, in apparent ignorance or innocence of the dubiousness of making such an assertion in 1933, Mises’s statement of the Reason without Experience thesis includes the claim that “[w]e conceive all this [i.e., the action axiom] from within, just as we conceive logical and mathematical truths, a priori, without reference to any experience” (Mises [1933] Reference Mises2003, pp. 13–14). That Mises saw no problem with trying to justify the (synthetic) apriority of the action axiom by referring to the apriority of logic and mathematics, and saw no need to explain the latter claim in light of relevant historical circumstances—i.e., Bolzano, Menger, and others—does not support the notion that he was familiar with recent developments in these fields. Similarly, as late as his last methodological work, 1962’s Ultimate Foundation of Economic Science, Mises (Reference Mises1962, pp. 12–14) argued for Euclidean geometry as an example of the Kantian synthetic a priori, a position undermined by Helmholtz’s work in the nineteenth century (Caldwell Reference Caldwell1984, p. 368, makes a similar point).Footnote ¹⁰ Mises ([1949] Reference Mises1998, p. 35) claimed that “[m]an acquired … the logical structure of his mind in the course of his evolution from an amoeba to his present state” right after asserting that the “character of the logical structure of the human mind” was “essential and necessary” (Mises [1949] Reference Mises1998, p. 34), a conjunction of propositions that suggests confusion, either about the highly contingent nature of evolutionary processes or the multiple meanings of necessity by then common in the philosophical, logical, and scientific literatures.

On the other hand, in its moderate interpretation as a kind of evolutionary epistemology, Lakatosianism, or Quineanism, Mises’s apriorism is a foreshadow of philosophical ideas that emerged later as potential resolutions of the very same debates over the possibility of a priori knowledge, of which, if the extreme interpretation is sound, he was seemingly either ignorant or unaware. The German-language literature on evolutionary epistemology is typically dated from Konrad Lorenz’s ([1941] Reference Lorenz and Ruse2010) “Kant’s Doctrine of the A Priori in the Light of Contemporary Biology,” but there is no evidence that Mises was familiar with this work (in any case, his conception of a priori knowledge ignores Lorenz’s arguments).Footnote ¹¹ The field of evolutionary epistemology really burgeoned with Karl Popper’s (Reference Popper1962) Conjectures and Refutations and the work of Donald T. Campbell in the 1960s. Other moderate readings try to make Mises’s methodology out to be a foreshadow of proto-Lakatosian conventionalism (Machlup Reference Machlup1955; Rizzo Reference Rizzo and Kirzner1983; Zanotti and Cachanosky Reference Zanotti and Cachanosky2015), or proto-Quinean holism (Leeson and Boettke Reference Leeson and Boettke2006), several decades before Lakatos (Reference Lakatos1968) or Quine (Reference Quine1951) made their seminal contributions to philosophy of science and epistemology. Such readings implicitly attribute remarkable prescience to Mises as a prophet of philosophical change.

The difficulty of understanding Mises’s methodology as an artifact of its time and place is complicated by the fact that the personal circumstances from which it emerged are largely obscured. It is not obvious from his writings which contemporary philosophers and scientists he had read, or with whom he might have argued, who could have encouraged the development of his seemingly unique epistemology. The rediscovery in the 1990s of much of Mises’s Vienna library, long thought lost, in a Soviet state archive has not brought clarity to this issue.

It is difficult to square the Mises whose epistemology seems to reveal an almost eighteenth-century rationalist with the Mises who both gestured at the possibility of an evolutionary epistemology and defended (albeit in a limited way) fallibilism and pragmatism. More to the present point, it is hard to understand why Mises would have advanced his seemingly extreme statements of the Reason without Experience and Greater Certainty theses at the time and place that he did: Vienna in the early 1930s.Footnote ¹² It is difficult to square the notion, implied in the latter anachronism, that Mises was ahead of his time in matters epistemological with some of the seemingly naive things he said about the same fields from which mid-century conventionalism would eventually emerge. Mises did not seem to appreciate the significance of the problems in mathematics, geometry, logic, epistemology, physics, etc. that led many scholars in (and philosophers of) these fields to modern conventionalist philosophy of science.Footnote ¹³

In short, Mises’s epistemological arguments are so incongruous with his intellectual environment and our grasp on his philosophical understanding is so weak that we cannot rely on knowledge of relevant intellectual circumstances either historically or geographically local to its development to help us understand Mises’s epistemology. Whether one reads his methodology as extreme or moderate, Mises’s apriorism is an anachronism, out of time and place.

Given the inconsistencies in both the primary and secondary literatures, and the impossibility of using knowledge of his historical or personal circumstances to reconstruct Mises’s understanding of epistemology, whatever he really thought about the theory of knowledge and philosophy of science is mostly lost to us. This thesis is further strengthened by the possibility, considered in the next section, that Mises’s methodological pronouncements may not have truly expressed his actual beliefs about economic methodology.

V. THE PROBLEM OF THE HISTORICAL MISES: A RHETORICAL SHELL GAME?

Unfortunately, the problem of the primary literature cannot be taken to be identical with the problem of what the historical Mises believed about epistemology. It is possible that Mises’s writings do not accurately reflect his actual beliefs about pertinent issues.

What are we to make of Israel Kirzner’s claim (Reference Kirzner2001; quoted in Leeson and Boettke Reference Leeson and Boettke2006, p. 248n2) that Mises “told him [Kirzner] that the action axiom was derived from ‘experience’”? Assuming the veracity of Kirzner’s testimony, what are we to make of Mises’s epistemology? If the historical Mises actually believed that knowledge of human action is due to experience, it is far from obvious how to interpret those several passages in his methodological writings in which he explicitly denies precisely this. Did Mises come later in life to change his mind about Reason without Experience and Greater Certainty in favor of some kind of empiricism about the action axiom? Were his statements of Reason without Experience and Greater Certainty mere obfuscations designed to conceal (for whatever reason) the empiricism he always accepted and eventually revealed to Kirzner? If so, why did he not make clear that he had come to reject his earlier rationalistic pronouncements? Or was he merely shining on Kirzner about his seemingly newfound empiricism? Was Mises just confused about various epistemological notions? Did he fail to recognize the tensions in his epistemological statements? Did he use the word “experience” differently in unique contexts such that what seems a contradiction is in fact not?

I do not know the answers to these questions. Indeed, I do not even know which are the right questions to ask. (And, I submit, neither do you.)

Hayek—who, of course, knew his “mentor” very well—occasionally suggested that Mises exaggerated his own methodological views. In reviewing Mises’s ([1933] Reference Mises2003) Epistemological Problems of Economics, Hayek ([1964] Reference Hayek, Klein and Kresge1992, p. 148; emphasis added) claimed, without providing an explanation, that Mises’s “emphasis on the a priori character of theory sometimes gives the impression of a more extreme position than the author in fact holds.” In his introduction to Mises’s (Reference Mises and Sennholz1978) Notes and Recollections, Hayek ([1988] Reference Hayek, Klein and Kresge1992, p. 158; emphasis added) asserted, again without much explanation, that “considering the kind of battle that he had to lead, I also understand that he was driven to certain exaggerations, like that of the a priori character of economic theory, where I could not follow him.”Footnote ¹⁴

Combined with the problem of the primary literature, the possibility that Mises was in fact an empiricist and the methodology seemingly expressed in his writings merely some kind of rhetorical shell game would seem to force the conclusion that “[u]ltimately, we simply do not know what the historical Ludwig von Mises believed about epistemology” (Scheall Reference Scheall2017a, p. 114).

Again, more carefully, I am not denying that there might be a reading that makes everything that Mises wrote (and/or believed) about epistemology both internally consistent and contextually respectable relative to the historical and geographical circumstances in which it was propagated. I am saying that it is unlikely, if such a reading exists, we will ever both know what it is and discover anything approaching univocal grounds for attributing it to Mises.

If this is right, then any interpreter who moves from his own reading of Mises’s apriorism to the claim that this reading represents Mises’s intentions has traveled a bridge too far. If the evidence from the primary and secondary literatures is so confused that we do not know what Mises actually believed about epistemology, then attributing some position to him, be it extreme, moderate, or in-between, is only so much wishful thinking on the part of the interpreter.Footnote ¹⁵ Such interpretations may well, for all we know, say more about the methodological disposition of the given interpreter than they do about Mises. Mises’s apriorism seems to be less a coherent position than a cipher upon which generations of Austrian sympathizers and antagonists have written their methodological fever dreams.