Abstract

In his influential criticism of scientific realism, Bas van Fraassen assumes that this doctrine is incompatible with empiricism, according to which the sole ultimate basis of knowledge is experience. This claim has been generally accepted in the contemporary literature in philosophy of science. Thus, the very distinction between scientific realism and empiricism is often forgotten, the term 'empiricism' being now widely used to designate a range of anti-realist positions, such as van Fraassen's "construct ive empiricism". In this paper it is argued, first, that empiricism, in the traditional and proper sense of the word, is a thesis about the problem of the foundations of knowledge, and should therefore be clearly distinguished from anti-realism, which concerns the issue of the extension of knowledge. It is then conceded that the main arguments for scientific realism do indeed require that extra-empirical characteristics of scientific theories, such as simplicity and explanatory power, should be ascribed epistemic weight. Although this point lends support to van Fraassen's claim, it is indicated here that his constructive empiricism is threatened by the same kind of epistemological objections which he raises against his opponents. Like some other scientific anti-realists, van Fraassen avowedly embraces realism concerning ordinary material objects; but it is not clear that this form of anti-realism remains tenable when explanatory power, simplicity, etc. are regarded as merely pragmatic, non-epistemic virtues.