Scepticism and dreaming

Abstract

In a recent, and influential, article, Crispin Wright maintains that a familiar form of scepticismwhich finds its core expression in Descartes’ dreaming argumentcan be defused (or, to use Wright’s own parlance, “imploded”), by showing how it employs self-defeating reasoning. I offer two fundamental reasons for rejecting Wright’s ‘implosion’ of scepticism. On the one hand, I argue that, even by Wright’s own lights, it is unclear whether there is a sceptical argument to implode in the first place. On the other, I claim that even on the supposition that Wright has indeed succeeded in setting-up such an argument, he nevertheless fails to follow-through with an adequate response. A diagnosis of the failure of Wright’s approach is then given in the context of the wider sceptical debate. 1. WRIGHT’S RECONSTRUCTION AND IMPLOSION OF THE SCEPTICAL ARGUMENT. Crispin Wright has proposed what he calls an ‘implosion’ of the familiar Cartesian form of dreaming-scepticism.1 As is well known, this argument essentially contends that I cannot know/justifiably believe that I am not now dreaming and thus, since dreaming excludes perception, I am not in a position to claim knowledge/justification for a wide class of perceptual beliefs. Wright does not characterise the argument in these terms however, since he believes that when interpreted with either ‘knowledge’ or justification’ as the epistemic operator the argument is ineffective.2 For when expressed in terms of knowledge it is always possible, he contends, to respond to the sceptic via a ‘Russellian retreat’, whereby knowledge of the contested propositions is conceded in favour of justification.3 And when represented in terms of justification the argument will be ineffectual since, he argues, we clearly do have sufficient justification for believing that we are not now dreaming. Nevertheless, he does think that there is a sceptical argument worth responding to here, one that turns upon an epistemic “notion which, like knowledge, is answerable to non-internalist standards 1 ‘Scepticism and Dreaming: Imploding the Demon’, Mind 397 (1991), 87-116. 2 Ibid., 90-3. 3 It is actually unclear what exactly is achieved by a ‘Russellian retreat’. Intuitively one might regard justification as being whatever, along with true belief, is sufficient for knowledge (with, perhaps, a codicil to Gettier). But neither Wright nor Russell have this epistemic notion in mind, since the very point of a Russellian retreat is to concede that, even if one’s beliefs were true, they still would not count as knowledge. (If this were not the case then there would be no need for a retreat. One could claim knowledge whilst allowing that there is a possible world in which one’s contingent beliefs are false, as long as, that is, one also had reason to believe that this world is not the actual world or a relevantly nearby world, and this much is surely contained in the intuitive notion of justification, so defined). The manœuvre thus gives us a very weak epistemic notion that may be exploitable by the sceptic after all.