The proper thesis of connectionism and the ability to model

Summary form only given, as follows. It is noted that the problem concerning the relation between a symbolic representation of some object and the corresponding subsymbolic storing in a neural network can be seen as a sophisticated form of the old philosophical problem of interactionism. In order to solve the problem one may adopt the connectionist approach, according to which the connections between cognitive models and neural models (or between symbolic and subsymbolic representations) can be found. This view, which can be called the proper thesis of connectionism, involves also another point of view, namely the conceptual modeling approach. Describing the representation of any kind of entity in (artificial or natural) neural networks presupposes at least some kind of theory of reference, but also a frame for conceptual modeling within which the epistemological status of the modeling ability of the modeler is described. The author introduces a frame for conceptual modeling which fits this demand and the thesis of connectionism. A theory of reference which fits the possibility of existence of semantically active states of artificial neural networks has also been developed