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] Pp. 376 & 388. [Book I, part III., secs. II. and VI.]

Not seeing this, Hume has to explain inference to latter system as something forced upon us by habit.

330. From the avowal of such paradoxical results, Hume only saved himself by reverting, as in the passage before us, to the popular view—to the distinction between two ‘systems of reality,’ one perceived, the other inferred; one ‘the object of the senses and memory,’ the other ‘of the judgment.’ He sees that if the educated man erased from his knowledge upon us by of the world all ‘facts’ but those for which he has ‘the evidence of his senses and memory,’ his world would be unpeopled; but he has not the key to the true identity between the two systems. Not recognizing the inference already involved in a fact of sense or memory, he does not see that it is only a further articulation of this inference which gives the fact of judgment; that as the simplest fact for which we have the ‘evidence of sense’ is already not a feeling but an explanation of a feeling, which connects it by relations, that are not feelings, with an unfelt universe, so inferred causes and effects are explanations of these explanations, by which they are connected as mutually determinant in the one world whose presence the simplest fact, the most primary explanation of feeling, supposes no less than the most complete. Not seeing this, what is he to make of the system of merely inferred realities? He will represent the relation of cause and effect, which connects it with the ‘system of memory,’ as a habit derived from the constant de facto sequence of this or that ‘inferred’ upon this or that remembered idea. The mind, ‘feeling’ the unchangeableness of this habit, regards the idea, which in virtue of it follows upon the impression of memory, as equally real with that impression. In this he finds an answer to the two questions which he himself raises: (a) ‘Why is it that we draw no inference from one object to another, except they be connected by the relation of cause and effect;’ or (which is the same, since inference to an object implies the ascription of reality to it), ‘Why is this relation requisite to persuade us of any real existence?’ and (b), ‘How is it that the relations of resemblance and contiguity have not the same effect?’ The answer to the first is, that we do not ascribe reality to an idea recalled by an impression, unless we find that, owing to its customary sequence upon the impression, we cannot help passing from the one to the other. The answer to the second corresponds. The contiguity of an idea to an impression, if it has been repeated often enough and without any ‘arbitrary’ action on our part, is the relation of cause and effect, and thus does ‘persuade us of real existence.’ A ‘feigned’ contiguity, on the other hand, because we are conscious that it is ‘of our mere good-will and pleasure’ that we give the idea that relation to the impression, can produce no belief. ‘There is no reason why, upon the return of the same impression, we should be determined to place the same object in the same relation to it’. [1] In like manner we must suppose (though this is not so clearly stated) that when an impression—such as the sight of a picture—calls up a resembling idea (that of the man depicted) with much vivacity, it does not ‘persuade us of his real existence’ because we are conscious that it is by the ‘mere good-will and pleasure’ of some one that the likeness has been produced.

[1] P. 409. [Book I, part III., sec. IX.]

But if so, ‘system of judgment’ must consist of feelings constantly experienced;

331. Now this account has the fault of being inconsistent with Hume’s primary doctrine, inasmuch as it makes the real an object of thought in distinction from feeling, without the merit of explaining the extension of knowledge beyond the objects of sense and memory. It turns upon a conception of the real, as the unchangeable, which the succession of feelings, in endless variety, neither is nor could suggest. It implies that not in themselves, but as representing such an unchangeable, are the feelings which ‘return on us whether we will or no,’ regarded as real. The peculiar sequence of one idea on another, which is supposed to constitute the relation of cause and effect, is not, according to this description of it, a sequence of feelings simply; it is a sequence reflected on, found to be unchangeable, and thus to entitle the sequent idea to the prerogative of reality previously awarded (but only by the admission as real of the ‘fiction’ of distinct continued existence) to the system of memory. But while the identification of the real with feeling is thus in effect abandoned, in saving the appearance of retaining it, Hume makes his explanation of the ‘system of judgment’ futile for its purpose. He saves the appearance by intimating that the relation of cause and effect, by which the inferred idea is connected with the idea of memory and derives reality from it, is only the repeated sequence of the one idea upon the other, of the less lively feelings upon the more lively, or a habit that results from such repetition. But if the sequence of the inferred idea upon the other must have been so often repeated in order to the existence of the relation which renders the inference possible, the inferred idea can be no new one, but must itself be an idea of memory, and the question, how any one’s knowledge comes to extend beyond the range of his memory, remains unanswered.

… which only differ from remembered feelings inasmuch as their liveliness has faded. But how can it have faded, if they have been constantly repeated?

332. What Hume himself seems to mean us to understand is, that the inferred idea is one of imagination, as distinct from memory; and that the characteristic of the relation of cause and effect is that through it ideas of imagination acquire the reality that would otherwise be confined to impressions of sense and memory. But, according to him, ideas of imagination only differ from those of memory in respect of their less liveliness, and of the freedom with which we can combine ideas in imagination that have not been given together as impressions. [1] Now the latter difference is in this case out of the question. A compound idea of imagination, in which simple ideas are put together that have never been felt together, can clearly never be connected with an impression of sense or memory by a relation derived from constant experience of the sequence of one upon the other, and specially opposed to the creations of ‘caprice’. [2] We are left, then, to the supposition that the inferred idea, as idea of imagination, is one originally given as an impression of sense, but of which the liveliness has faded and requires to be revived by association in the way of cause and effect with one that has retained the liveliness proper to an idea of memory. Then the question recurs, how the restoration of its liveliness by association with an impression, on which it must have been constantly sequent in order that the association may be possible, is compatible with the fact that its liveliness has faded. And however this question may be dealt with, if the relation of cause and effect is merely custom, the extension of knowledge by means of it remains unaccounted for; the breach between the expectation of the recurrence of familiar feelings and inductive science remains unfilled; Locke’s ‘suspicion’ that ‘a science of nature is impossible,’ instead of being overcome, is elaborated into a system.

[1] Part I., sec. 3; cf. note on p. 416 [Book I, part III., sec. IX.].

[2] P. 409. [Book I, part III., sec. IX.]