Abstract

The central claim of this paper is the following. There are two different types of notions applied in science, and in fact in any discourse whatsoever. Some of the notions are veristic, i.e. they refer to some specific fairly well defined entities (individual objects, species, relations, systems, regularities, etc.). But there are also heuristic notions that serve the users to call attention to the fact that a specific situation can be accounted for, or even less - there is a chance that it can be accounted for, in a certain way known already from some earlier application of similar ideas. On numerous occasions we argue by analogy. Now, to argue by analogy means to treat the situation accounted for as „similar” to some other already familiar situations. The similarity in question can amount to some fairly explicitly stated expectations, which go beyond the available evidence. The heuristic notions can be characterized as the notions in terms of which such expectations are stated. Since the right interpretation of those expectations should be defined anew any time when an argument by analogy is applied, also the concepts they involve share this characteristic. Thus while the veristic concepts have some fixed references, the reference of a heuristic concept should be defined anew every time when the concept is used. A theory which involves any heuristic notions is called heuristic. A theory is said to be veristic if all the descriptive notions it involves are veristic. The notion of truth can be in a straightforward way applied to veristic theories only