A Treatise of Human Nature / Being an Attempt to Introduce the Experimental Method Into Moral Subjects; and Dialogues Concerning Natural Religion

by David Hume

[1] p. 481 [Book I, part IV., sec. II.]

Locke’s theory of sensation disappears. Physiology won’t answer the question that Locke asked.

198. He thus logically cuts off his psychology from the support which, according to popular conceptions, its primary truths derive from physiology. We have already noticed how with Locke metaphysic begs defence of physic; [1] how, having undertaken to answer by the impossible method of self-observation the question as to what consciousness is to itself at its beginning, he in fact tells us what it is to the natural philosopher, who accounts for the production of sensation by the impact of matter ‘on the outward parts, continued to the brain.’ To those, of course, who hold that the only possible theory of knowledge and of the human spirit is physical, it must seem that this was his greatest merit; that, an unmeaning question having been asked, it was the best thing to give an answer which indeed is no answer to the question, but has some elementary truth of its own. According to them, though he may have been wrong in supposing consciousness to be to itself what the physiologist explains it to be—since any supposition at all about it except as a phenomenon, to which certain other phenomena are invariably antecedent, is at best superfluous—he was not wrong in taking the physiological explanation to be the true and sufficient one. To such persons we can but respectfully point out that they have not come in sight of the problem which Locke and his followers, on however false a method, sought to solve; that, however certain may be the correlation between the brain and thought, in the sense that the individual would be incapable of the processes of thought unless he had brain and nerves of a particular sort, yet it is equally certain that every theory of the correlation must presuppose a knowledge of the processes, and leave that knowledge exactly where it was before; that thus their science, valuable like every other science within its own department, takes for granted just what metaphysic, as a theory of knowledge, seeks to explain. When the origin, for instance, of the conception of body or of that of an organic structure is in question, it is in the strictest sense preposterous to be told that body makes the conception of body, and that unless the brain were organic to thought I should not now be thinking. ‘The brain is organic to thought;’ here is a proposition involving conceptions within conceptions—a whole hierarchy of ideas. How am I enabled to re-think these in order, to make my way from the simpler to the more complex, by any iteration or demonstration of the proposition, which no one disputes, or by the most precise examination of the details of the organic structure itself?

[1] See above, paragraph 17.

Those who think it will don’t understand the question.

199. The quarrel of the physiologist with the metaphysician is, in fact, due to an ignorantia elenchi on the part of the former, for which the behaviour of English ‘metaphysicians,’ in attempting to assimilate their own procedure to that of the natural philosophers, and thus to win the popular acceptance which these alone can fairly look for, has afforded too much excuse. The question really at issue is not between two co-ordinate sciences, as if a theory of the human body were claiming also to be a theory of the human soul, and the theory of the soul were resisting the aggression. The question is, whether the conceptions which all the departmental sciences alike presuppose shall have an account given of them or no. For dispensing with such an account altogether (life being short) there is much to be said, if only men would or could dispense with it; but the physiologist, when he claims that his science should supersede metaphysic, is not dispensing with it, but rendering it in a preposterous way. He accounts for the formal conceptions in question, in other words for thought as it is common to all the sciences, as sequent upon the antecedent facts which his science ascertains—the facts of the animal organisation. But these conceptions—the relations of cause and effect, &c.—are necessary to constitute the facts. They are not an ex post facto interpretation of them, but an interpretation without which there would be no ascertainable facts at all. To account for them, therefore, as the result of the facts is to proceed as a geologist would do, who should treat the present conformation of the earth as the result of a certain series of past events, and yet, in describing these, should assume the present conformation as a determining element in each.

Hume’s psychology will not answer it either.

200. ‘Empirical psychology,’ however, claims to have a way of its own for explaining thought, distinct from that of the physiologist, but yet founded on observation, though it is admitted that the observation takes place under difficulties. Its method consists in a history of consciousness, as a series of events or successive states observed in the individual by himself. By tracing such a chain of de facto sequence it undertakes to account for the elements common to all knowledge. Its first concern, then, must be, as we have previously put it, to ascertain what consciousness is to itself at its beginning. No one with Berkeley before him, and accepting Berkeley’s negative results, could answer this question in Locke’s simple way by making the primitive consciousness report itself as an effect of the operation of body. To do so is to transfer a later and highly complex form of consciousness, whose growth has to be traced, into the earlier and simple form from which the growth is supposed to begin. This, upon the supposition that the process of consciousness by which conceptions are formed is a series of psychical events—a supposition on which the whole method of empirical psychology rests—is in principle the same false procedure as that which we have imagined in the case of a geologist above. But the question is whether, by any procedure not open to this condemnation, the theory could seem to do what it professes to do—explain thought or ‘cognition by means of conceptions’ as something which happens in sequence upon previous psychical events. Does it not, however stated, carry with it an implication of the supposed later state in the earlier, and is it not solely in virtue of this implication that it seems to be able to trace the genesis of the later? No one has pursued it with stricter promises, or made a fairer show of being faithful to them, than Hume. He will begin with simple feeling, as first experienced by the individual—unqualified by complex conceptions, physical or metaphysical, of matter or of mind—and trace the process by which it generates the ‘ideas of philosophical relation.’ If it can be shown, as we believe it can be, that, even when thus pursued, its semblance of success is due to the fact that, by interpreting the earliest consciousness in terms of the latest, it puts the latter in place of the former, some suspicion may perhaps be created that a natural history of self-consciousness, and of the conceptions by which it makes the world its own, is impossible, since such a history must be of events, and self-consciousness is not reducible to a series of events; being already at its beginning formally, or potentially, or implicitly all that it becomes actually or explicitly in developed knowledge.

It only seems to do so by assuming the ‘fiction’ it has to account for; by assuming that impression represents a real world.

201. If Hume were consistent in allowing no other determination to the impression than that of its having the maximum of vivacity, or to other modes of consciousness than the several degrees of their removal from this maximum, he would certainly have avoided the difficulties which attend Locke’s use of the metaphor of impression, while at the same time he would have missed the convenience, involved in this use, of being able to represent the primitive consciousness as already a recognition of a thing impressing it, and thus an ‘idea of a quality of body.’ But at the outset he remarks that ‘the examination of our sensations’ (i.e. our impressions of sensation) ‘belongs more to anatomists and natural philosophers than to moral,’ and that for that reason he shall begin not with them but with ideas. [1] Now this virtually means that he will begin, indeed, with the feelings he finds in himself, but with these as determined by the notion that they are results of something else, of which the nature is not for the present explained. Thus, while he does not, like Locke, identify our earliest consciousness with a rough and ready physical theory of its cause, he gains the advantage of this identification in the mind of his reader, who from sensation, thus apparently defined, transfers a definiteness to the ideas and secondary impressions as derived from it, though in the sequel the theory turns out, if possible at all, to be at best a remote result of custom and association. We shall see this more clearly if we look back to the general account of impressions and ideas quoted above. ‘An impression first strikes upon the senses and makes us perceive pleasure or pain, of which a copy is taken by the mind,’ called an idea. Now if we set aside the notion of a body making impact upon a sensuous, and through it upon a mental, tablet, pleasure or pain is the impression, which, again, is as much or as little in the mind as the idea. Thus the statement might be re-written as follows:—‘Pleasure or pain makes the mind perceive pleasure or pain, of which a copy is taken by the mind.’ This, of course, is nonsense; but between this nonsense and the plausibility of the statement as it stands, the difference depends on the double distinction understood in the latter—the distinction (a) between the producing cause of the impression and the impression produced; and (b) between the impression as produced on the senses, and the idea as preserved by the mind. This passage, as we shall see, is only a sample of many of the same sort. Throughout, however explicitly Hume may give warning that the difference between impression and idea is only one of liveliness, however little he may scruple in the sequel to reduce body and mind alike to the succession of feelings, his system gains the benefit of the contrary assumption which the uncritical reader is ready to make for him. As often as the question returns whether a phrase, purporting to express an ‘abstract conception,’ expresses any actual idea or no, his test is, ‘Point out the impression from which the idea, if there be any, is derived’—a test which has clearly no significance if the impression is merely the idea itself at a livelier stage (for a person, claiming to have the idea, would merely have to say that he had never known it more lively, and that, therefore, it was itself an impression, and the force of the test would be gone), but which seems so satisfactory because the impression is regarded as the direct effect of outward things, and thus as having a prerogative of reality over any perception to which the mind contributes anything of its own. By availing himself alternately of this popular conception of the impression of sensation and of his own account of it, he gains a double means of suppressing any claim of thought to originate. Every idea, by being supposed in a more lively state, can be represented as derived from an impression, and thus (according to the popular notion) as an effect of something which, whatever it is, is not thought. If thereupon it is pointed out that this outward something is a form of substance which, according to Hume’s own showing, is a fiction of thought, there is an easy refuge open in the reply that ‘impression’ is only meant to express a lively feeling, not any dependence upon matter of which we know nothing.