Pragmatism, Naturalism and Abstracta

Ross Pain (La Trobe University)

In 1980 Sellars and Quine exchanged a series of articles and letters on the problem of abstracta. For the most part, they were mutually sympathetic to each other’s broadly naturalistic goals. They differed fundamentally, however, over whether or not certain abstract entities (in particular, classes, sets and properties) are ontologically permissible. According to Quine’s methodological naturalism, any entities (abstracta included) may be admitted into one’s ontology if they are considered indispensable to a successful theory’s workability. Sellars, on the other hand, regards a requirement for abstract entities as a sign that a theory has strayed from a properly naturalistic framework. This disagreement is interesting in that both thinkers are neo-pragmatists inspired by the linguistic turn, and both are self-professed naturalists. In other words, the conflict points toward differences in the way the terms ‘pragmatism’ and ‘naturalism’ are employed as philosophical frameworks. I look in detail at the debate and draw some philosophical lessons from it. I argue that, in this area at least, Sellars has the better naturalistic credentials. In particular I claim that Huw Price’s distinction between object naturalism and subject naturalism sheds interesting light on the issues at the heart of the Sellars/Quine dispute, and points towards a new (pragmatic) way of resolving the harsh contrast between the manifest and scientific images.