Dreaming

2. Dreaming and scepticism

Descartes used the dream argument in the Meditations (1641) in order to shake our belief in the reliability of the senses (see Descartes §4). Is it possible to describe our waking life as a dream? If so, how could we distinguish dreaming from perceiving? Of course, nothing depends specifically on dreams in the dream argument for the possibility of mistrust in perception; hallucinatory experiences would do as well. Moreover, perception is normally accompanied by reflexive awareness, whereas dreaming is normally not, and thus dreaming is not the best candidate for the sceptical argument, unless it be lucid dreaming. But these differences are hardly necessary differences, and the epistemological question of establishing how dreaming may be distinguished from perceiving remains open. (1) One might try a phenomenological answer, pointing out that our waking experience is distinguished by the fact that we believe it to begin upon awakening, and to end upon falling asleep. But – the sceptic will answer – nothing excludes that we dream that we both wake up and fall asleep. (2) Descartes proposed that clarity and distinctness of contents is a sign of veracity, but one has to admit that even the clearest and most distinct perceptions might turn out to be delusive. (3) Descartes, in the Sixth Meditation (1641), suggested that some form of the coherence criterion would allow, in some cases, a distinction between dreams and non-dream experiences; Leibniz, however, thought the knowledge provided through the application of this criterion not certain but only probable. But this criterion cannot exclude a certain set of coherent experiential reports being about an especially coherent great dream; as Hobbes remarked in his Objections to Descartes (see Descartes [1641] 1969), we might be dreaming of an all-encompassing coherent connection between experiences. It thus seems that the distinction between dreaming and waking is epistemologically not available.

An attack on scepticism, grounded in an argument against the received view that dreams are experiences, which is itself based on a very strong and controversial version of the principle of verification, has been put forward by Norman Malcolm (1959). Malcolm links (criterially) the occurrence of a dream with the (possibility of an) utterance of a dream report. Indeed, according to Malcolm, the only acceptable criterion for the occurrence of any experience or state is that one be able to provide a report thereof, a position which clearly relies on an extreme verificationist conception of meaning and truth (see Criteria §1). Malcolm’s main argument is that: (1) if one dreams, then one is asleep; (2) as there is no possibility of applying any criterion during sleep and the occurrence of a dream during sleep cannot therefore be verified, the question of whether a dream occurred during one’s sleep is meaningless; (3) for the same reason the question of whether a certain experience took place during one’s sleep is also meaningless. The conclusion is (4) that dreams cannot contain (or be identical with) experiences occurrring during sleep.

This is of some consequence for scepticism: if dreams are not experiences occurring during sleep, then the sceptic’s question of how we can tell dream experience from waking experience does not even arise. Moreover, the debate over the principle of coherence would be revealed as senseless, for it would presuppose that we can make sense of comparing contents and finding out that some do not cohere with most others (thereby discovering that we are not dreaming); this is, by Malcolm’s lights, absurd, for to compare is to judge, and no judgments occur during sleep.