Monday, July 28, 2014

John Quiggin on Apriorism in Austrian Economics

John Quiggin has an excellent post here on the apriorist method in Austrian economics:
John Quiggin, “Austrian Economics and Flat Earth Geography,” John Quiggin, July 28th, 2014.

Also posted here:
John Quiggin, “Austrian Economics and Flat Earth Geography,” Out of the Crooked Timber, July 27, 2014.
One minor point here is that John Quiggin says that apriorist praxeology was “also endorsed, in a more qualified fashion, by Hayek.”

Actually, Hayek rejected Mises’ apriorism and adopted a – more or less – Popperian method for economics.

We know this because Hayek said so explicitly in a letter to Terence W. Hutchison dated 15 May, 1983. In this, Hayek stated:
“I had never accepted Mises’ a priorism. .... Certainly 1936 was the time when I first saw my distinctive approach in full clarity – but at the time I felt it that I was merely at last able to say clearly what I had always believed – and to explain gently to Mises why I could not ACCEPT HIS A PRIORISM”. (quoted in Caldwell 2009: 323–324).
In fact, in 1937 Hayek had published an article called “Economics and Knowledge” where he criticised Mises’ apriorism; Hayek also appears to have moved closer to Popperian ideas on methodology in later years:
“I became one of the early readers [sc. of Karl Popper’s Logik der Forschung, 1934]. It had just come out a few weeks before …. And to me it was so satisfactory because it confirmed this certain view I had already formed due to an experience very similar to Karl Popper’s. Karl Popper is four or five years my junior; so we did not belong to the same academic generation. But our environment in which we formed our ideas was very much the same. It was very largely dominated by discussion, on the one hand, with Marxists and, on the other hand, with Freudians. Both these groups had one very irritating attribute: they insisted that their theories were, in principle, irrefutable. Their system was so built up that there was no possibility – I remember particularly one occasion when I suddenly began to see how ridiculous it all was when I was arguing with Freudians, and they explained, ‘Oh, well, this is due to the death instinct.’ And I said, ‘But this can’t be due to the [death instinct].’ ‘Oh, then this is due to the life instinct.’ … Well, if you have these two alternatives, of course there’s no way of checking whether the theory is true or not. And that led me, already, to the understanding of what became Popper’s main systematic point: that the test of empirical science was that it could be refuted, and that any system which claimed that it was irrefutable was by definition not scientific. I was not a trained philosopher; I didn’t elaborate this. It was sufficient for me to have recognized this, but when I found this thing explicitly argued and justified in Popper, I just accepted the Popperian philosophy for spelling out what I had always felt. Ever since, I have been moving with Popper.” (Nobel Prize-Winning Economist: Friedrich A. von Hayek, pp. 18–19).
So Hayek lumped in apriorism in economics with Marxism and Freudianism!

It is also clear that Mises disliked Popper and Popper’s approach to epistemology in his seminal book Logik der Forschung (1934; later published in English as The Logic of Scientific Discovery, 1959):
“Popper was familiar with the early [sc. socialist] calculation debate – Polanyi’s seminar discussed it – but not much taken by it. He knew of Mises and his circle, but it is unlikely that he read Mises closely. He strongly disliked subjectivism and libertarianism. He ‘first met Mises early in 1935 in Vienna, owing to his interest in my first book. . . . Both he and I were aware of a strong opposition between our views in the field of the theory of knowledge and methodology. Mises saw me as a dangerous opponent.’” (Hacohen 2000: 478).
So it could not be more clear: Mises saw Popper “as a dangerous opponent”; but Hayek, rejecting apriorism, essentially endorsed Popper’s views on epistemology and method.

Nevertheless, Quiggin’s point that Misesian apriorist praxeology as a method for economics is untenable is right.

Mises needed synthetic a priori knowledge and a type of Kantian epistemology to justify his praxeology, but the arguments for synthetic a priori knowledge are unconvincing and must be rejected. Misesian praxeology does not yield universally and necessarily true empirical statements about economic reality.

Furthermore, not even the human action axiom can be known a priori: it is clearly a synthetic a posteriori proposition.

In addition, the very idea that Mises in Human Action succeeded in deducing all his theories by deductive logic is manifestly untrue, and Misesians and Rothbardians have never answered the challenge of George J. Schuller to set out Human Action in a formal symbolic form in which all axioms, premises, and deductions are shown formally and proven.

Even modern Kantians have rejected Mises’ attempt to ground his praxeology on Kant’s epistemology.

There has also arisen amongst modern Austrians a feeble and ignorant belief that Mises was not really using a synthetic a priori epistemology. This is simply untrue, as I have shown here and here, and even if it were true and praxeology were simply analytic a priori, then it would follow logically that praxeology cannot give necessary truth about the real world.

Further Reading
“Limits of the Human Action Axiom,” February 28, 2011.

“Hayek on Mises’ Apriorism,” May 23, 2011.

“What is the Epistemological Status of Praxeology and the Action Axiom?,” July 27, 2013.

“Barrotta’s Kantian Critique of Mises’s Epistemology,” July 28, 2013.

“David Friedman versus Robert Murphy,” August 4, 2013.

“Mises Fails Philosophy of Mathematics 101,” August 30, 2013.

“Bob Murphy All At Sea on Geometry and Economic Epistemology,” August 31, 2013.

“Mises’s Non Sequitur on synthetic a priori Knowledge,” September 2, 2013.

“Hoppe’s Caricature of Empiricism,” September 10, 2013.

“Hoppe on Euclidean Geometry,” September 11, 2013.

“Robert Murphy gets Mises’s Epistemology Wrong,” September 13, 2013.

“Hoppe on Euclidean Geometry, Part 2,” September 14, 2013.

“Mises on Kant and Praxeology,” September 15, 2013.

“Mises was Confused about the Analytic–Synthetic Distinction,” September 15, 2013.

“Schuller’s Challenge to Misesian Apriorists has never been answered,” December 7, 2013.

“Mises versus Ayer on Analytic Propositions and a priori Reasoning,” March 16, 2014.

“David Gordon on Praxeology and the Austrian Method: A Critique,” March 13, 2014.

“Why Mises’s Praxeological Theories are not Necessarily True of the Real World,” March 15, 2014.

“Mises and Empiricism,” April 17, 2014.

“Why Should we reject the Existence of Synthetic a priori Knowledge?,” May 23, 2014.

Caldwell, B. 2009. “A Skirmish in the Popper Wars: Hutchison versus Caldwell on Hayek, Popper, Mises, and methodology,” Journal of Economic Methodology 16.3: 315–324.

Hacohen, Malachi Haim. 2000. Karl Popper, The Formative Years, 1902–1945: Politics and Philosophy in Interwar Vienna. Cambridge University Press, Cambridge, UK and New York.

Hayek, F. A. 1937. “Economics and Knowledge,” Economica n.s. 4.13: 33–54.

Nobel Prize-Winning Economist: Friedrich A. von Hayek. Interviewed by Earlene Graver, Axel Leijonhufvud, Leo Rosten, Jack High, James Buchanan, Robert Bork, Thomas Hazlett, Armen A. Alchian, Robert Chitester, Regents of the University of California, 1983.


  1. I think that Mises probably felt that, in order to bring about the world he hoped for---in which all humans actually behave as the rational calculators described by economic theory--he had to pretend that that world already existed, and that he was describing it.

  2. Lars P Syll had post up on Austrian methodology today
    i noticed.

  3. The parallel postulate is an axiom, and the real world is more easily described by non-Euclidean geometry. This is not to say that Euclidean geometry doesn't apply to the real world.

    To say that only Euclidean or non-Euclidean geometry applies to the real world are both metaphysical statements. There is no evidence that denies Euclidean geometry from applying to the real world, because nothing says its IMPOSSIBLE for you to describe space and light as Euclidean.

    In fact, no evidence can prove or disprove whether Euclidean geometry applies to the real world or not. It would be silly to attempt this. Its an arbitrary designation made for modeling.

    However, to say that the mind perceives things as Euclidean is correct. You cannot visualize any non-Euclidean geometry, without visualizing it as Euclidean geometry. That is mental capacity of our mind, and is an inductive observation.

    I can prove this by asking you to draw an instance where the parallel postulate is invalid. Since its not possible for you to draw this, then its sufficiently proven.

    Schopenhauer had similar arguments shown here:

    So when people say the universe is Euclidean, it is the same as saying the mental capacity of the human mind only perceives Euclidean space. Objectively, the universe is not necessarily Euclidean, non-Euclidean, or any geometry at all. Any geometry is a mental analytical conception.

    This is where the confusion arises. Mises never used the term "axiom" because the basis of praxeology is formed of statements about the nature of how humans perform actions, which is inductive.

    Everything derived from this is apriori, and is based on the analysis of the causal factors of human actions.

    This isn't like geometry because these are not constructions for the purposes of modeling reality. They are observations about reality itself. They are inductions from introspection that are validated by literally asking any person.

    1. "In fact, no evidence can prove or disprove whether Euclidean geometry applies to the real world or not."

      Of course there is evidence that can.

      There is evidence that inductively proves that non-Euclidean geometry is the correct theory of spacetime in the universe.

    2. "So when people say the universe is Euclidean, it is the same as saying the mental capacity of the human mind only perceives Euclidean space.

      This is NOT what many rationalist apriorists have said down through the ages: many said it was necessarily and universally true of real space.

    3. "Mises never used the term "axiom" because the basis of praxeology is formed of statements about the nature of how humans perform actions, which is inductive.

      Everything derived from this is apriori,"

      So you think that the basis of praxeology is inductive, but at the same time a priori?

      Are you aware how stupid and ignorant this statement is?

    4. "There is evidence that inductively proves that non-Euclidean geometry is the correct theory of spacetime in the universe."

      This statement has no meaning. No geometry has physical existence, it only has conceptual existence. Its applicable to physical objects as descriptions of these objects. You can't "prove" geometry anymore than you can "prove" a triangle's angles add up to 180 degrees by measuring them. They are only proven by being derived from their axioms.

      Also, I only meant to say that if a person is not explicit as to their meaning, then it can be construed that the human mind perceives the physical world this way. I did not mean to say that every person holds this opinion.

    5. (1) "Its applicable to physical objects as descriptions of these objects. "

      That is what I meant, you fool: non-Euclidean geometry is the correct/right theory/model of spacetime in the universe.

      (2) It appears that you are using a narrow definition of "prove" in the sense of "prove" with apodictic certainty/truth in a shoddy fallacy of equivocation.

      When I say "prove" I mean **inductively prove with a degree of probability**: just as we can inductively prove that the heliocentric model is the right theory of how our solar system is actually structured with a high degree of probability.

  4. "Are you aware how stupid and ignorant this statement is?"

    Once its causes are inductively grasped, then applying them is analytic a priori. The inductions themselves are not a priori, except for whatever a priori nature is in induction itself.

    For example, any applications of these inductive concepts to explain economic historic events are a priori, because you are analyzing these events with PRIOR concepts.

    1. Oh, so your poorly written comment is supposed to mean that

      (1) you are saying that the human action axiom and all other axioms of praxeology are known a posteriori?

      (2) But when used as a premises in deductive arguments, the deductive argument per se, if valid and sound, is known as true a priori?

      Correct, but doesn't vindicate the synthetic a priori epistemology of Mises.

      In any case whenever a premise is synthetic a posteriori, you need empirical evidence and inductive argument to prove that this premise per se it is correct, so Mises is not freed from empiricism at all.