Approximation of anticipatory systems
Bertil Ekdahl · International Conference on Information Systems · 1998
In Systems Science and Computer Science, it is still customary to consider systems from a syntactically point of view. It is argued that, unlike causal systems, anticipatory systems should be studied from a linguistic point of view. Since language is complementarity in the sense that it consist of a description part and an interpretation part, language has to be considered holistically, i.e. language can not be fragmented into its part. Ibis has the consequence that anticipatory systems should be divided into description based and model-based where model is taken in its semantical meaning. The complementarity view now implies that model-based anticipatory systems, like planning creatures, is not describable and at first glance this seems to be a serious limitation. However, it can effectively be overcome by the productivity of Turing machines, meaning that every acting model-based anticipatory system can, with arbitrary accuracy, be approximated by a description-based anticipatory system. It has the consequence that there will be no limitation in understanding and building systems although we will never be able to construct a system which contain its own interpretation (model).