EN
This paper deals with the problem of the so-called universal reasoning theory. First it presents two theories that claim to be universal in describing reasoning. Both are built on the algorithmic level what means that they have to specify the nature of mental symbols and the processes that operate upon them. The protagonists of the historically first theory, the mental rule theory, claim that reasoning requires building mental proofs for given arguments. Such proofs are syntactic in nature, and their steps resemble the use of rules known from courses on formal logic. The protagonists of the second theory say that reasoning consists of building mental models (hence the name: mental model theory) that correspond to different possibilities that may happen given the truth of the premises. Both theories seem to account for reasoning processes on the universal level, and yet this claim is too broad for both of them. As noted by Roberts (1993) both theories omit the problem of individual differences. However, the study of individual differences might shed a new light on the rules vs. model debate and provide important new data on the reasoning processes described on both algorithmic and higher levels (e.g. Anderson’s rational level).