Boghossian on externalism and privileged access

Abstract

1. There's been much recent discussion about whether content externalism is compatible with the sort of knowledge we have of our thought contents.1 According to externalism, some of a subject's thought contents are partly individuated in terms of her environment. However, it's been argued that externalism is incompatible with the following plausible claim about the nature of our self-knowledge: that a subject can know what her own thought contents are just by reflection and without undertaking an empirical investigation (let's call this claim 'privileged access'). Much of the discussion has focused on reductio arguments of the form first proposed by McKinsey (McKinsey 1991).2 He argued that the assumptions of externalism and privileged access generate an absurd conclusion, namely that a subject can have a priori knowledge of the nature of her environment. Thus, he concluded that externalism and privileged access are incompatible. The sort of reductio argument McKinsey proposed could in principle be applied to any of the different varieties of externalist theses.3 However, in considering whether the argument is successful, it's tempting to simplify matters by considering just one specific externalist thesis. After all, it would be significant and interesting to show that one externalist position is incompatible with privileged access, and one could afterwards consider whether the argument could be extended to other varieties of externalism. A clear example of someone who uses this strategy is Boghossian. In Boghossian 1997, he focuses on the issue of whether externalism generated by Putnamian Twin Earth thought experiments concerning natural kind terms is incompatible with privileged access. He makes it clear that he's leaving other varieties of externalism aside for consideration at a future date. In particular, he explicitly sets aside Burgian externalism, saying that he expects that the same argument which he applies to Putnamian externalism could be later applied to Burge's position (Boghossian 1997: 163, n.3). Although Boghossian's strategy might seem to be a sensible way of approaching the issue, and one that in no way affects the outcome of the argument, I will argue that this is not so. For, by deferring consideration of Burge until later, I think that Boghossian effectively rules out of consideration a number of objections to his argument that arise from Burgian