Gentner and Forbus develop the analogy hypothesis as a structural mapping. This means that several relations from the source scheme are mapped to the target scheme. This is explained by the classic example of the analogy "solar system - atomic structure", in which the relations "attracts", "revolves around" are common, the relations "more massive" are distributed, and "hotter" are not distributed. In our opinion, this is an inaccurate definition of how the analogy of schemes in thinking is realized. Gentner's hypothesis uses several relations, and it is not clear how they are found, by what principle they are selected. In the scheme we propose, only one relation is used.

Observing a new phenomenon as the structure of an atom, we define the relations "revolves around the center without visible connection". And this is precisely what determines the choice of the solar system analogy as the most common analogous scheme. What is the difference? Firstly, it is the relation "revolves around" with one condition "without visible connection", and not several separate relations. Secondly, "revolves around" - such relations are present in many processes, for example, a carousel, a water funnel, a centrifuge. But it is the solar system that is chosen, and only due to the condition "without visible connection", which discards other options. And thus the choice occurs on the basis of one of the schemes of thinking, which will be described below. This is the scheme "find a generalizing scheme and a specializing feature in order to create a new scheme".

We are preparing an article for the journal, where we will show how known ways of thinking such as metaphor, syllogism, ontology and generalization, recognition and learning are realized with the help of analogy.