Mieszko Tałasiewicz

In the paper the author tries to show that semantico-categorial analysis in the traditional Ajdukiewiczian application as a tool of evaluating semantic completeness of complex expressions is ineffective. There is, however, another application, to which the analysis is much better fitted. Namely, it allows to precisify the actual uses of natural language - in all those cases in which on the one hand precision is a superior value, and on the other hand, the formalisation is not possible and one has to stay within natural language. Obviously, in order to achieve this aim many particular problems, sometimes not at all trivial, have to be solved. Amongst the most important the author sees determining a binding canon of the rules of analysis and determining pragmatic criteria of the admissibility of paraphrases and of filling in ellipses

The starting point of my considerations is the idea that the criteria of rationality should be distinguished from the essence of rationality (in linguistic formulation: from a non-criterional connotative definition) and that the question of the universality of rationality arises separately for the criteria and for the essence. I notice that a similar distinction between the essence (or a definition) and the criteria applies to the notion of truth. The thesis which constitutes ny standpoint is a compromising one: the criteria (the sets of criteria) of rationality change in time and in space, the essence of rationality is universal and unchangeable. Therefore, the principal challenge for the comprehensional conception of rationality, which is to back up my standpoint, is a formulation of a good, non-criterional connotative definition of the rationality of science. My definition is the following: „Science is rational” means „the consecutive stages of its development allow us to better understand (in an identificatory sense) the world”