History of Western Philosophy - Bertrand Russell [373]
Recollection is one of a whole genus of effects which are more or less peculiar to the phenomena that we naturally call 'mental'. These effects are connected with habit. A burnt child fears the fire; a burnt poker does not. The physiologist, however, deals with habit and kindred matters as a characteristic of nervous tissue, and has no need to depart from a physicalist interpretation. In physicalist language, we can say that an occurrence is 'perceived' if it has effects of certain kinds; in this sense we might almost say that a watercourse 'perceives' the rain by which it is deepened, and that a river valley is a 'memory' of former downpours. Habit and memory when described in physicalist terms, are not wholly absent in dead matter; the difference, in this respect, between living and dead matter, is only one of degree.
In this view, to say that an event is 'perceived' is to say that it has effects of certain kinds, and there is no reason, either logical or empirical, for supposing that all events have effects of these kinds.
Theory of knowledge suggests a different standpoint. We start, here, not from finished science, but from whatever knowledge is the ground for our belief in science. This is what Berkeley is doing. Here it is not necessary, in advance, to define a 'percept'. The method, in outline, is as follows. We collect the propositions that we feel we know without inference, and we find that most of these have to do with dated particular events. These events we define as 'percepts'. Percepts, therefore, are those events that we know without inference; or at least, to allow for memory, such events were at some time percepts. We are then faced with the question: Can we, from our own percepts, infer any other events? Here four positions are possible, of which the first three are forms of idealism.
(1) We may deny totally the validity of all inferences from my present percepts and memories to other events. This view must be taken by anyone who confines inference to deduction. Any event, and any group of events, is logically capable of standing alone, and therefore no group of events affords demonstrative proof of the existence of other events. If, therefore, we confine inference to deduction, the known world is confined to those events in our own biography that we perceive—or have perceived, if memory is admitted.
(2) The second position, which is solipsism as ordinarily understood, allows some inference from my percepts, but only to other events in my own biography. Take, for example, the view that, at any moment in waking life, there are sensible objects that we do not notice. We see many things without saying to ourselves that we see them; at least, so it seems. Keeping the eyes fixed in an environment in which we perceive no movement, we can notice various things in succession, and we feel persuaded that they were visible before we noticed them; but before we noticed them they were not data for theory of knowledge. This degree of inference from what we observe is made unreflectingly by everybody, even by those who most wish to avoid an undue extension of our knowledge beyond experience.
(3) The third position—which seems to be held, for instance, by Eddington—is that it is possible to make inferences to other events analogous to those in our own experience, and that, therefore, we have a right to believe that there are, for instance, colours seen by other people but not by ourselves, toothaches felt by other people, pleasures enjoyed and pains endured by other people, and so on, but that we have no right to infer events experienced by no one and not forming part of any 'mind'. This view may be defended on the ground that all inference to events which lie outside my observation is by analogy, and that events which no one experiences are not sufficiently analogous to my data to warrant analogical inferences.
(4) The fourth position is that of