Measurement: A Systems Approach

Abstract

It has become popular in philosophical literature to refer to a certain formulation of the problem of induction as ‘Hume’s problem’, with the inevitable questioning whether Hume actually had such a problem as the formulator proposes. I should characterize the central theme of TEI as a concern with ‘Kant’s problem’, with the inevitable admission that Kant himself never faced the problem as I formulated it. ‘Kant’s problem’ is stated exactly in the middle of the text:

... we now seem forced to the following paradoxical conclusions:

• 1. We must agree with Kant that the construction of an intelligible world out of the immediacies of sensation demands a certain a priori equipment, i.e., demands the assumption of certain principles not derivable from the elementary facts of experience.

• 2. But we must disagree with Kant that a priori laws are known intuitively, and that their verification is independent of observation.

(This paper is a review of an earlier work of mine, Theory of Experimental Inference (TEI), on its twenty-first birthday, in order to assess its promises.)