Impossible Models

Abstract

It is a widespread belief among all kinds of people, including scientists that reality, though difficult to capture, can nevertheless finally be approximated by making models better and better. The meaningfulness of the wordings “approximated” and “better” in this context, however, may be doubted. To be sure, there is a real world supplying us with a great variety of inputs. Also, we are more or less able to act within the real world, applying thereby guidelines stemming from such diverse sources as everyday experience and full fledged science. Living in the real world and making models of parts of it, however, does not mean that reality is approximated by the models used. The latter would require the notion of a distance between reality and models which makes little sense, I think. (For an opposite view see [Niiniluoto, 1987] . For drawing my attention to Niiniluoto’s work as well as for other valuable remarks I would like to thank Stephan Hartmann.)