Lord Keynes — Why Mises’s Praxeological Theories are not Necessarily True of the Real World

My commentary:

There are two approaches to necessary truth, one scientific and one dogmatic.

Both approaches use axiomatic theories that are deductive in that theorems follow from application of formation and transformation rules to axioms. If the rules are applied consistently, then the truth of theorem following from axioms and postulates is said to be analytic and syntactical from the logical standpoint and capable of being known before experience (a priori) epistemologically.

The difference between science and dogmatic ideology is that scientists proceed to generate testable hypotheses from the theorems they deduce that generate experimental protocol statements that can be disconfirmed through observation. 

This aspect of science is quite obviously contingent in that truth is contingent on confirmation through observation. That which a model represents lies outside the model and is independent of the model, so that the truth of descriptions is additive or "synthetic," and can only be known after experience (a posteriori). This requires observation.

Dogmatic ideology stops at the necessary truth of the deductive aspect of the theory and holds that confirmation is unnecessary since the necessary truth of axioms, since they are postulated, and theorems, since they are logically deduced, shows that it is impossible that they be false. Examination of the model is therefore sufficient epistemologically. This involves confusion of logic and epistemology.

This is false logic in that it confuses the model with that which is being modeled. Necessary truth applies only to models, and the relation of model to that which is modeled is contingent on that which lies beyond the model. That can only be known exogenously to the model, regardless of the logical necessity endogenous to the model.

This is important for understanding MMT, for instance, where accounting identities provide the necessary truth of a stock-flow consistent model but say nothing about the causality involved. That must be determined from by observation, e.g., of operations.

Social Democracy For The 21St Century: A Post Keynesian Perspective
Why Mises’s Praxeological Theories are not Necessarily True of the Real World
Lord Keynes