What Makes Economics a Human Science?

What Makes Economics a Human Science?

Zoöpraxeology and the Abstractive Foundation of Catallactics

In Human Action (1966), Ludwig Von Mises stresses both the purely formal character of praxeological logic as a description of human action, and the unique character of goal-directed action by which man is set apart from animals. To the extent the former is true, however, the latter becomes less unique, and we may usefully interpret animal behavior using the praxeological method. This suggests in turn that the study of the market order must be founded elsewhere than in the axiom of purposive action. This foundation is more properly the uniquely human capacity for abstraction, which Mises failed to distinguish from purposefulness due to his overly strict dichotomy between the methods of praxeology and Verstehen.

First of all, it will be useful to begin from Roger Koppl’s (2002) reinterpretation of Mises’ methodology through Schutz. Rather than Mises’ strict dichotomy between “exact” praxeology and the empirical and contingent Verstehen – between theory and history as he liked to put it – Koppl instead interprets the difference as the use of ideal types of varying anonymity. Ideal types become less anonymous as we fill in substantiative content – for example from “human” to “American male” to “local postman” to “Dave Johnson”. The formal and exact laws of praxeology are the most highly anonymous ideal types, applying to people qua people. Importantly, using any of these types is equivalent with an act of Verstehen.

Second, in order to clarify what exactly Verstehen is, we must establish its basis in congruent mental structures under what Koppl calls closed reflexivity (i.e. I know that you have the same structure as I do, I know that you know that I do, and so on). This congruence allows us to impute meaning that is not contained merely in the objective facts of the situation, and may consist in structures inborn or acquired through experience. Acquired congruent structure allows use of less anonymous (more substantiative) ideal types in dealing with a neighbor than, say, someone who speaks a different language and grew up in a different culture. The fact that Hayek (1952), in his account of the cognitive basis of Verstehen, used as his working hypothesis a mass of initially undifferentiated neurons indicates that he was primarily concerned with the use of these semi-anonymous ideal types acquired through experience, though he does concede off-handedly the possibility of structures embodying the “experience of the species”.

Advances in linguistics in the decades since the publication of The Sensory Order have, however, vindicated Mises’ use of highly anonymous ideal types by appeal to such biologically ingrained structures. In particular, the biological structure of Universal Grammar is the prototype for all acts of Verstehen. The specific problem in that field is language acquisition given the fact that no amount of linguistic input could suffice to infer the rules of a language unless the child already knew what to listen for (Chomsky 1965). This is known as the “poverty of the stimulus” and is a specific example of the more general inability to infer meaning from the objective facts of a situation.1 Thus, in overcoming this problem, language acquisition is the first and primary act of Verstehen. Innate mental structures of this sort, shared across the entire species, are the basis of the highly anonymous ideal types of which praxeology consists. This is true regardless of the “objective” validity of those types, and for this reason it would be more apt to call Mises’ epistemology “necessarian” in the Kantian sense than absolutist. He argues as much in a section entitled “A Hypothesis about the origin of A Priori Categories” (1962, p. 17):

In stressing the fact that the logical structure of the human mind is common to all specimens of the species Homo sapiens, we do not want to assert that this human mind as we know it is the only or the best possible mental tool that could be devised or that has ever been and will ever be called into existence. . . . Perhaps there will once somewhere live beings who will look upon us with the same condescension as we look upon amoebae. But scientific thinking cannot indulge in such imagery. It is bound to limit itself to what is accessible to the human mind as it is.

Chomsky’s work, by vindicating the Kantian approach, shows just how far ahead of its time Mises’ epistemology was. Nevertheless, the analogy to language also highlights the problem with what for most people is the main stumbling block in reading Mises: he equivocates on two senses in which something may be said to be “apodictically certain”. The rules of English grammar, for example, are known to me in a sense a priori. I know, along with every other native speaker of English, that there are three different past tense markers for weak verbs ([-t], [-d], and [-əd]). Despite the fact that most of these native speakers could not tell you the rule that determines which form to use, they know the rule in the sense that they can execute it effortlessly and with perfect accuracy, and immediately identify when someone else has used it wrongly.2

Likewise with the praxeological types. It may be true that I know, a priori, certain things about other people that help me to impute meaning to their behavior. But it is this knowledge, and not the ex post articulation, that we can call “apodictically certain”. Reconstructions of grammatical rules have been in error, and in principle there is no reason a reconstruction of the categories of action might not be as well. After all, Mises’ is a minority position, which is an uncomfortable place to be for a doctrine alleging self-evident and perspicuous truths.

Thus, Koppl’s reformulation of Mises’ method as a gradient of anonymity, rather than a dichotomy between apodictically certain praxeology and empirically contingent Verstehen, suggests the question: to what extent might the ideal types of praxeology be sufficiently anonymous (that is, grounded in mental structures sufficiently primordial) to extend backward along the phylogenetic tree, to other primates, mammals, or vertebrates?

Mises himself dismisses the question:

There is nothing in between a being driven exclusively by instincts and physiological impulses and a being that chooses ends and the means for the attainment of these ends. (1962, p. 8)

We interpret animal behavior on the assumption that the animal yields to the impulse which prevails at the moment. As we observe that the animal feeds, cohabits, and attacks other animals or men, we speak of its instincts of nourishment, of reproduction, and of aggression. We assume that such instincts are innate and peremptorily ask for satisfaction. But is different with man. Man is not a being who cannot help yielding to the impulse that most urgently asks for satisfaction. Man is a being capable of subduing his instincts, emotions, and impulses; he can rationalize his behavior. (1966, p. 16)

Yet, he says of human instinct immediately before the latter quote,

[R]ationalism, praxeology, and economics do not deal with the ultimate springs and goals of action, but with the means applied for the attainment of an end sought. However unfathomable the depths may be from which an impulse or instinct emerges, the means which man chooses for its satisfaction are determined by a rational consideration of expense and success. He who acts under an emotional impulse also acts.

Being driven by instinct is a fact about ends, and does not impair the formal validity of the praxeological types. A distinction is therefore warranted between the capacity for abstraction – which is common among and exclusive to human beings – and the capacity for purposive action, which would appear to be more or less common to all beings with a central nervous system.

Mises’ own strict dichotomy between praxeology and Verstehen does not allow for this latter distinction, both sides of which he identifies with “rationality”. Nevertheless, we see him crediting Freud as the first to attribute rationality to the mentally ill within the context of their own ends (Mises 1944). This example is intended to illustrate that however nonsensical we judge the mentally ill man’s ends to be – however incongruous the illness has rendered his mental structure with our own – there is nevertheless enough congruity that we are bound to call action in pursuit of those ends purposive. The fact that animals are driven primarily by instinct, rather than by reason, does not alter the fact that the congruity of the means-ends structure of action extends much further up the phylogenetic tree than the singular quality of abstract reasoning. We may suppose a specialized structure in the brains of migratory birds gives rise to the impulse to fly south in response to some stimulus indicating the onset of winter. The bird cannot explain what drives him – but neither can humans explain very often what drives them. The fact that the evolutionary significance of migration is altogether different from the significance the bird itself attaches to its migratory act is no different in principle from the fact that many human norms have a cultural-evolutionary significance entirely different from that attributed by their practitioners (e.g. Leeson 2012). And though we cannot know the ineffable quality of being driven southward by the change of season, we also cannot know the quality of having been raised in, say, a Confucian household, which is to say nothing more than that Verstehen is limited by the congruence of mental structures, whether inborn or acquired, and however so congruent the two may or may not be. The ideal type of purposive action, as distinct from the capacity for abstraction or language (which are likely the same capacity), appears to be sufficiently anonymous that we can legitimately call the bird’s migration “purposeful”.

The capacity for abstraction, however, is the foundation of man’s ability to subdue his instincts, and thus of all activity usually regarded as “economic” – certainly all calculating activity. In other words, humans are characterized by a uniquely low time preference among animals, though of course “instinct” is merely the word for an animal impulse that arises from a mental structure sufficiently non-congruent as to be inscrutable to Verstehen (Mises 1966, p. 27). If Rational Choice is essentially a translation of the praxeological types into integral space, then economic imperialism could perhaps conquer zoology similarly to how it conquered criminology (Becker 1968), modeling animal behavior as a Robinson Crusoe with a discount rate close to 1.

It is possible that the praxeological types only result in a useful research program when applied to subjects with a sufficiently low time preference. Perhaps zoology has, with its own methods, exhausted all the insights that might be gained this way. Even so, the cognitive foundation of this methodological point indicates that the widely shared ideal type of purposive action is insufficient grounds for a science of the uniquely human extended order: it must be grounded instead in the uniquely human ideal types of language and abstraction.


  1. The Duhem-Quine Thesis identifies the same problem in the context of the scientific method: the objective facts of the situation are, on their own, never sufficient to decisively confirm or reject a theory.
  2. The rule, for what it’s worth, depends on the last phoneme of the verb stem: [-əd] follows [t] or [d], [-d] follows other voiced consonants or vowels, and [-t] follows devoiced consonants.


EpistemologyLanguageLinguisticsMethodologyPraxeologyF.A. HayekGary BeckerLudwig Von MisesNoam ChomskyRoger Koppl


Facebook Twitter Reddit Threads

Leave a Reply

More Content