Abstract

The paper reconstructs the evolution process of scientific knowledge. The evolution theory has been applied hitherto exclusively to the famous reference problem. It the eye would be incapable seeing something really available it could not establish itself it the reality as such evolutional achievement. Contrary to this view the author states that the cognitive apparatus could survive not due to their achievements in the representations of the external world but rather due to their selfreproductive capabilities. By extrapolation of this view on the level of the epistemology it means that the knowledge itself selects that it can know on the base of that it already knows. The author suggests the principles of such cognitive evolution - the mechanisms of variety, selection and restabilization. The mechanism of variety concerns exclusively some particular operations (i.e. the communicative occurrences). Something innovative (unexpected, out of the ordinary) which has been recently created would occasionally be uttered, suggested, described, and probably printed under sole condition that it is apprehensible and writable. The selection is always based on some structures i.e. on the expectations of some reproductive use of some meaning affitudes. Only the structures can be marked out symbolically: applied to the science it means that they are marked as the true or the false ones. Finely, the stabilization level consists in the continuality of the autopoesis of the scientific communication.