Abstract

In this paper I lay down some groundwork for a pragmatist scientific realism, which will be fully consonant with actual scientific practices. Scientific realism demands that our best scientific theories should give really true descriptions of the world. Truth here is usually conceived in terms of a ‘correspondence’ between theory and reality – ultimate, metaphysical, and mind-independent reality. However, this idea is useless in practice because such reality is inaccessible to us, and the alleged correspondence only makes sense as a metaphor based on actual representational activities, in which both the ‘model’ and the ‘target’ are accessible. This metaphor appears to make literal sense only because we take part in the illusion of the ‘ready-made world’, according to which reality, independently of any conceptions we impose on it, already has well-defined parts and relations between the parts. Abandoning the illusory metaphor, I propose that realists should accept pragmatism in relation to ‘primary truth’, which does not consist in agreement with other things that we already know to be true. Primary truth is based on the ‘operational coherence’ of activities that we engage in; if some coherent activities rely on a certain proposition, then that proposition is true within the domain of those activities. Once we have some primarily true propositions, then correspondence to them defines the secondary truth of other propositions. However, the picture I propose is not a foundationalist one in the traditional sense: a given proposition may be true in a primary or a secondary way, or even both. ‘Truth happens to an idea’ (William James), and the manner of that happening depends on the contingent contexts of truth-making and truth-finding activities.