Dreaming

Abstract

The counter-intuitive features of Malcolm's analysis are reviewed. An attempt is made to avoid these difficulties by allowing the admission of criteria for dreaming other than Malcolm's. This leads to a consideration of Malcolm's use of the term "criterion" and reference is made to Malcolm's interpretation of Wittgenstein on the subject. It is concluded that language is not criterion-governed in the way that Malcolm (but not Wittgenstein) thinks. An analysis of dreaming is then attempted on the basis of all the types of evidence we use for the occurrence of a dream, the approach being to determine what dreaming must be like if these can all be evidence for dreaming. It is argued that our concept of dreaming is of a datable occurrence, but it is admitted that experimental evidence could arise which would incline us away from this view. This approach tends to represent dreaming as an unobservable, and to progress further, one needs to take account of our concept of dreaming as an experience during sleep. Malcolm's arguments against this are rejected. The implications of treating dreaming as an experience are reviewed, and it is shown how this view avoids the difficulties which beset Malcolm's analysis. It is argued that Malcolm's treatment of dreaming as a basis for scepticism, dependent on his view that one cannot make judgements in sleep, fails as there is reason to reject this view. However one of his arguments against the coherence principle is found pertinent, and an attempt is made to refurbish the coherence principle in the light of this. This leads us to a further consideration of scepticism, in which it is maintained that the problem of dream-scepticism, if pushed to its limits, loses its specific reference to dreaming.