Abstract

Rules of inference can be interpreted as rules of (purposive) behavior; in such a case the behaptor consists in accepting a certain statement, called conclusion. The justification of a rule of inference with respect to a given end consists in showing that it is the most efficient method of realizing that end; the meaning of the word “efficient”, and the character of the end, should, of course, be made clear.The article is an attempt at reconstructing a part of the theory of mathematical statistics (the so-called parametric inference) along those lines. It is assumed that the end to be realized by following rules of statistical inference is, broadly speaking, a cognitive one