Monday, December 13, 2010

Supporting Mises' Methodology Using Hayek's Mental Model?

I have just come to realize that one of the main things Hayek does in The Sensory Order is provide the mental basis for Mises' a priori propositional methodology of deduction from the axiom of action. The entirety of the book up through the following quote describes the brain as a classificatory organ. Then, Hayek observes that, because of how the brain classifies and relates concepts,

It is probably no accident that the formation of classes and the relation between classes were first studied in the attempt to analyze the principles of conceptual reasoning. It should be clearn now that the same kind of relationship which in logic has been developed as the theory of classes and relations is immediately applicable to that physiological process of multiple grouping or classification which we have been examining. (6.48)

In other words, logic is the natural extension from how our brains do in fact work. One of course in turn can use logicl to further clarify and refine thinking. Now, if logic is in fact how we perceive the world and logic is how we think and categorize the world, then one should use logic to try to understand the world. This is Mises' argument for what is and should be the proper methodology of economics -- for which it seems Hayek is providing the mental foundation. All things within our perception, then, are properly understood using logic; if economic behavior is properly understood as emerging out of action, then one needs to use deductive logic.

If this is true, might Hayek have been more of a Misesian than most people realize? Or, perhaps, more than even he realized?

1 comment:

Julien Couvreur said...

ET Jaynes has an excellent book, Probability Theory: the logic of science, in which he argues that Bayesian probability is the formalization of our common sense reasoning, and proves that it an extension of logic (and the only possible extension, within some reasonable requirements).

Basically deductive logic is a special case of the more general inference process. Logic and scientific reasoning rest on the same foundation (Bayesian inference).

You can read the first 3 chapters for free online, which are the most interesting in my opinion. And I'm working on a summary on my blog.

Overall, it makes me think the chasm between logic and empirical knowledge is not necessary.

If you have a strong prior in a theory (or a priori knowledge), then deduction is possible. Otherwise, you need observations to evaluate various theories are.