A recent critique of a priorism

Natsuka Tokumaru is a doctoral student at Kyoto University in Japan. At the 2007 Rethinking Popper conference in Prague she delivered a paper titled “Popper’s analysis of the probems of induction and demarcation and Mises’ justification of the theoretical social sciences”. The proceedings of the conference have been published by Springer.[My comments are in square brackets.]

Tokumaru considered the arguments advanced by von Mises to assess whether they are adequate to solve the problems of induction and demarcation which Popper regarded as the two fundamental problems in epistemology.

[I have reservations about the demarcation problem, it may have been Bartley who suggested that we need to 'demarcate' between better and worse theories, in terms of their capacity to solve problems and stand up to various forms of criticism. Attempts to establish criteria for the 'scientific' status of theories and methods have run often run into the sands of technicality (in the hands of the inductivists) and into anti-scientism (in the hands of others such historicists and Austrians). It is interesting to note, as a historical aside, that during the 1920s and 1930s Popper was employed as a school teacher and worked on the fundamental problems of epistemology in the evening, while nearby von Mises ran the Austrian economy during the day and went home to relax and work on epistemological problems in economics. His book appeared in 1933, followed by Popper's book in 1934/5.]

Tokumaru takes up the Misian claim that the theoretical social sciences, called praxeology, the sciences of human action, are justified as an a priori discipline, presupposing the “category of human action”. She identified for possible formulations of this category.

1. An observational statement or a statement describing experiences from introspection.

2. A proposition about the basic ontological form of the social universe, describing its essential characteristics.

3.  A definition, adopted as a convention.

4. A methodological principle, of the kind required by methodological individualism.

She concluded that under no definition or interpretation can Mises justify the social sciences as a piori disciplines.

Tokumaru  noted that von Mises used Kantian terminology but the tenor of his thought is rather like Descartian rationalism ( the clear and distinct intuition of  ideas) and so “it is quite unclear what kind of  of aprioism Mises defends, and on what grounds.”   Because the action axiom “all humans act to achieve aims” asserts something about the world, it may be regarded as synthetic, but if  is synthetic and also a priori, how is it to be justified on logical, methodological and epistemological grounds?

Tokumaru first examines von Mises’ perception of the epistemology of the natural sciences to ascertain his stance on induction and demarcation. She then examines his position on the human sciences.

Mises’ epistemology and methodological views of the natural sciences

He shared Popper’s view that empirical science is defined by strict universality and empiricism [but of course Popper's empiricism does not take the positivist form and his position can just as well be called "modified rationalism/intellectualism" or "fallible apriorism".] Mises was quite happy to accept inductivism and positivism in the natural sciences “The modern natural sciences owe their success to the method of observation and experiment. There is no doubt that empiricism and pragmatism are right as far as they merely describe the procedures of the natural sciences.” (HA 32)

For Mises, inductive inferences are applicable to the natural sciences because of the regularity of nature.

All human knowledge concerning the universe presupposes and rests upon the cognition of the regularity in the succession and concatenation of observable events…Inductive infrence is conclusion from premises that invariably include the fundamental proposition of regularity. HA 21-22

Being aware of Hume’ arguments he suggests that there  is an induction principle to justify inductive inferences; in methodological terms the key factor is is the repeatability of observations, assuming that they have been “correctly observed” (HA 22, Mises’ italics). But this is like the principle of induction which Bertrand Russell and others thought might have to be used to justify the logic of inductive  inferences. It just shifts the argument back a step (to justify the principle). Likewise in this case, the principle of correct and repeated observations needs to be justified itself, and so on.

What is a Priori?

Tokumaru notes the distinction between the psychological a priori (innate expectations and capacities) and a priori validity.  [Nobody is likely to deny the existence of innate capacitites, that is the simple truth, so far as we know.]  However Tokumara notes that innate expectations, for example the infant’s expectation of  food and nurturing can be unfulfilled. And a priori validity is a claim about the the truth status of propositions but the validity of  such claimes can only be decided in the light of experience. 

She points out that Mises sometimes refers to psychological a prioris in the sense of evolved endowments which fit us to our environment. This is the territory of evolutionary epistemology [and again, nobody will deny the existence of evolved endowments] but Mises wants to make major claims about the validity of a priori theories of  praxeology, while at the same time distancing these claims from the psychlogical a priori.

Mises’ Epistemological and Methodological Position in the Theoretical Social Sciences

At this point Tokumaru returns to the four different interpretations of the caterory of action, listed above. She finds that all four can be found to some extent in Mises’ writings. 

1. The category of action is an observational statement or a statement describing experiences from introspection

She suggests that Mises could have used singular observation statements (about the way that people evaluate situations and then act) as an empirical base to postulate social regularities by generalization. He wrote:

The starting point of praxeology is a self-evident truth, the cognition of action, that is, the cognition of  the fact that there is such a thing as consciously aiming at ends (Ultimate Foundations, 1962, 5-6)

So he might have used a princple of induction (like the one he relied on in his account of the natural sciences) to make the leap from (correct) observations of individual instances of human action, to the universal statement of the action axiom.

The logical steps run:

First, premises in the form of singular statements in the form ‘At time t and location k there is a human who acts rationally to achieve an aim’.

Second, the fundamental proposition of regularity (principle of induction) ‘What is correctly observed in one case must also be observed in all other cases offering the same conditions.’

Conclusion (a strictly universal statement) ‘All humans act rationally to achieve aims’.

[Acute observers will note that this conclusion is identical to Popper's much discussed and problematic RP, the Rationality Principle, in his theory of Situational Analysis].

Tokumara points out that this approach makes the axiom hostage to the inductive principle and hence not valid a priori.

[Of course Misians in their capacity as methodological dualists would not regard this as a valid criticism because they insist that the inductive principles of the natural sciences do not apply in this domain, but this is a logical exercise by Tokumaru to explore various ways of interpreting the axiom and its basis.]

[Against the tide of  dualism, some Austrians adhere to "causal realism" which can be found in Menger (Salerno, in The Review of Austrian Economics, 2010) . I don't  have the citations at finger tips but sometimes the Austrians speak of  'exact laws', which are universal (valid for all times and places) and explain economic phenomena such that wherever certain "initial conditions" apply, certain effects will follow.  This is precisely the mode of explanation in the natural sciences, though in complex (realistic) situations the laws will predict tendencies rather than precise outcomes due to the multiplicity of factors at work and the impossibility of getting a full account of the initial condidions. More precision can be obtained in stable and isolated systems like the solar system and experimental models.]

2. The category of action is a proposition about the basic ontological form of the social universe, describing its essential characteristics.

In this formulation “Praxeological reality is not the physical universe, but man’s conscious reaction to the given state of this universe” (HA, 1949,  p 92).

And “We do not want to discover a new method, but only to characterize correctly the method that is actually used in economics” (1933, p 18)

The deduction reads as follows:

Premise 1. Social knowledge exists (social knowledge is possible).

Premise 2. If social knowledge exists, humans exist who act rationally.

Conclusion. Humans exist who act rationally (there exists human action).

This line of argument depends on the validity of Premise 1 in all circumstances, meaning that it has to assume the status of a sythetic a priori, which begs the question of its justification and leads to another in finite regress statements, each claiming to be a priori valid.

3.  The category of action is a definition, adopted as a convention.

Indeed, Mises stated that “Action is, by definition, always rational” (1933, p. 3)

[On this account the theorems of economics would be tautologies on a par with maths and logic, with all the content contained in the inititial definitions. Clearly  Mises thought along this line some  of the time, especially when he compared monetary theory to geometry so everything is contained in the concept. Such is supposed to be the case with the axiom of action, but Smith and others have explained that when you get to details (cases or situations) and unpack the concept you end up with a plethora of items that are not contained in a simple definition of action, indeed it gets to look like a very complex and sophisticated Popperian Situational Analysis].

Tokumaru’s critique of this approach is very complex and contains some symbolic logic which makes my head hurt. I am not certain what this adds to the really simple objection that the logical products of definitions are hostage to the truth status of the definition, assuming it is relating to the natural or human world and is not an abstract creation. But Tokumaru points out that Mises insists that the theories are not only a priori true but also “…convey exact and precise knowledge of real things” (ibidem!).

On Tokumaru’s account Mises resorted to pragmatic or instrumental arguments to explain the possibilities of theories which are definitions and also pertain to experiences in a helpful manner. [The pragmatic approach has some similarity to the proposal that we test theories for their capacity to solve problems (especially to provide explanations) and to stand up to tests of various kinds.] 

Tokumaru concludes this section saying that if Mises took this option he would have to settle for usefullness instead of a priori truth as the criterion for acceptance of economic theorems.

4. The category of action is a methodological principle, of the kind required by methodological individualism.

This is entirely consistent with Mises’ aims, his principles and his practice. However methodological principles are normally regarded as conventions, to be accepted or rejected on the basis of their utility, that is to generate true descriptions and explanations of phenomena. Questions of truth and falsehood can arise, for example methodological individualism may be supported by refuting claims about group minds or the spirit of the age, however methodological principles cannot be claimed to be a priori truths.

Tokumaru concludes that none of the four formulations of the category of action justify claims to a priori truth.

This entry was posted in Uncategorized. Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *


− 3 = one

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>