Theory of Ideas

Hume accepts the Cartesian doctrine of the distinct idea - conceivability subject only to the principle of contradiction - as both the unit of reasoning and the criterion of truth. But the doctrine of the distinct idea means that every noncontradictory idea expresses an a priori logical possibility. And the speculative freedom of the imagination to conceive opposites without contradiction makes it impossible to demonstrate any matter of fact or existence. This argument leads to a distinction between relations of ideas (demonstrations which are true a priori) and matters of fact (the opposite of which is distinctly conceivable). And this distinction excludes from the domain of rational determination every factual event, future contingent proposition, and causal relation. For Hume, since truth is posterior to fact, the ideas of reason only express what the mind thinks about reality.

Distinct ideas, or imaginative concepts, are pure antinomies apart from experience as every factual proposition is equally valid a priori. But Hume does acknowledge that such propositions are not equally meaningful either to thought or action. On the level of ideas, Hume offers a conceptual correlative to the exemption of sensation as a form of cognition by his recognition that the meaning of ideas is more important than their truth. What separates meaningful propositions from mere concepts is the subjective impression of belief.

Belief, or the vivacity with which the mind conceives certain ideas and associations, results from the reciprocal relationship between experience and imagination. The cumulative experience of the past and present - for example, the relational factors of constancy, conjunction, and resemblance - gives a bias to the imagination. But it is man's imaginative anticipations of the future that give meaning to his experience. Neither the relational elements of experience nor the propensive function of the imagination, from the viewpoint of the criterion of truth, possesses the slightest rational justification. Hence the interplay between the criterion of truth and the logic of the imagination explains both Hume's skepticism and his conception of sensation and intellection.

The most celebrated example of this argument is Hume's analysis of the causal relation. Every statement which points beyond what is immediately available to the senses and memory rests on an assumption and/or extension of the cause and effect relation. Let us examine two cases: I see lightning and hear thunder; I see a rabbit and then a fox. The question is why I am right in concluding that lightning causes thunder but wrong in believing that rabbits cause foxes. Experience, in both instances, reveals an A that is followed by B, and repeated experiences show that A is always followed by B. While the constant conjunction of A and B might eliminate the rabbit-fox hypothesis, it is of no help in explaining causality because there are all sorts of objects, such as tables and chairs, which are similarly conjoined but not supposed to be causally related. Thus experience reveals only that constant conjunction and priority are sufficient but not necessary conditions for establishing a causal connection. And it is necessity, understood as that which cannot be otherwise than it is, which makes a relation causal in the propositional form of "If A then B must appear and if no A then no B."

But if necessary connection explains causality, what explains necessity? Experience yields only a particular instance and tells us nothing about the past or the future. Nor is there any necessity discoverable in repeated experiences. That the sun will rise tomorrow because it has in the past is an assumption that the past necessarily causes the future which is, of course, the connection that is to be demonstrated. If experience cannot account for necessity, then reason fares no better. I can always imagine the opposite of any matter of fact without contradiction. If someone tells me that Caesar died of old age or that thunder is uncaused or that the sun will not rise tomorrow, I will not believe him, but there is nothing logically incorrect about such statements since for every probability "there exists an equal and opposite possibility." Thus there is no justifiable knowledge of causal connections in nature, although this is not a denial that there are real causes. Man's supposed knowledge results from repeated associations of A and B to the point where the imagination makes its customary transition from one object to its usual attendant, that is, "an object followed by another, and whose appearance always conveys the thought to that other."

Because of his skeptical attitude toward the truths of reason Hume attempted to ground his moral theory on the bedrock of feeling - "Reason is, and ought only to be, the slave of the passions." In this, Hume followed the "moral sense" school and, especially, the thought of Francis Hutcheson. The notion that virtue and vice are to be derived ultimately from impressions of approbation and blame or pleasure and pain shows that Hume anticipated Jeremy Bentham's utilitarianism, a debt which the latter acknowledged. Although Hume considered himself to be primarily a moralist, this doctrine is the least original part of his philosophical writings.