Summary In the last decade analytical philosophy of science has been considered by many people as a descriptive activity. In part I of this paper we show that philosophy of science has been designed as normative logical analysis by Reichenbach and Carnap before world war II. Thus the identification analytical = descriptive is historically unjustified. In part II we discuss three tasks of analytical philosophy of science, the logical reconstruction of concepts, theories, and methods. While the first is mainly descriptive, (...) the third mainly normative, the second may be both, depending on the intended goal. (shrink)

The determination of the meaning of theoretical terms by the axioms of theories as meaning postulates and the merely fictitious character of a basic observational language leads to Feyerabends problem of the incommensurability of physical theories. Different theories are actually compared by physicists as well. They might have a common sub-language as the language of macroscopic physics in atomic physics. Furthermore shared metalaws define an equivalence relation which identifies terms of different theories and enables physicists to talk about them in (...) a common language. (shrink)

The determination of the meaning of theoretical terms by the axioms of theories as meaning postulates and the merely fictitious character of a basic observational language leads to Feyerabends problem of the incommensurability of physical theories. Different theories are actually compared by physicists as well. They might have a common sublanguage as the language of macroscopic physics in atomic physics. Furthermore shared metalaws (e. g. invariance principles) define an equivalence relation which identifies terms of different theories and enables physicists to (...) talk about them in a common language. (shrink)