A renewed interest in the old problem of the relationship between science and metaphysics has been fuelled by the ongoing debate between naturalistic metaphysicians and non-naturalistic metaphysicians. However, I maintain that this debate is missing the mark because it is focused on the problem of the credibility of a metaphysics that is not ‘scientific’, instead of focusing on the presence of metaphysics in science. In order to show that metaphysics pervades all stages of scientific inquiry, and after analysing the distinction (...) between presuppositions and assumptions, I address the complex problem of the relation of metaphysics to truth and to experience. I advocate that there is an indirect relation of metaphysics to experience and that it is possible to choose between rival metaphysical theories. But metaphysics, according to my view, is not present in science merely as a background of presuppositions and assumptions. It is present at every step of the scientific inquiry and also in a later moment: the interpretation of the findings of science and the elaboration of unifying theories. Keywords: science, metaphysics, presuppositions, truth, experience, unification. (shrink)

info:eu-repo/semantics/publishedVersion.

info:eu-repo/semantics/publishedVersion.

In this paper I intend to thoroughly analyse Karl Popper’s relation to metaphysics. I start with his first writings, where he states the differences between science, pseudoscience and metaphysics. I then describe how his thoughts on the subject evolved to culminate in his reflection on metaphysical research programmes and the need for a revival of natural philosophy. A major concern is Popper’s famous testability criterion to set apart science from non-science. I point at the problems of the conception of metaphysics (...) as non-testable theories and, in order to avoid these problems, I propose to retain nothing but the traditional conception of metaphysics as the general theories about the nature of the world. This leads me to the conclusion that science is not only an empirical task but also, and in a very important sense, a speculative one. (shrink)

info:eu-repo/semantics/publishedVersion.