Hume, David

Hume, David

(b. Edinburgh, Scotland, 26 April 1711; d. Edinburgh, 25 August 1776)

philosophy, economy, political theory, history.

His father, Joseph Home—David Hume preferred the phonetic spelling—was a country gentleman with a small estate, Ninewells, near Berwick-upon-Tweed. His mother, Catherine Falconer, was a daughter of Sir David Falconer, lord president of the Court of Session. Hume retained a lifelong admiration for the gentry, ascribing to them that “moderate scepticism” which he himself sought to foster. His father died young, in 1713, leaving Hume a small legacy on which he later could barely support himself.

Hume matriculated at the University of Edinburgh in 1723, but left three years later without taking a degree. Edinburgh was a center of Newtonian physics, and Hume most probably was taught its elements either by the mathematician James Gregory or by Newton’s popularizer, Colin Maclaurin. On the philosophical side, at Edinburgh there flourished a group of ardent Berkeley disciples. The religious atmosphere was a liberal Calvinism but at an early age, Hume told Boswell, he lost all belief in religion as a result of reading Locke and Samuel Clarke.

Following a family tradition, he set out to study law. He became convinced, however, at the age of eighteen, that he had made a great discovery which “opened up anew scene of thought,” and the determined to devote himself wholly to working out his new ideas.

There is considerable controversy about the nature of Hume’s “new scene of thought,” but there are good grounds for believing that it at least incorporated the idea of constructing a “science of man” by applying Newtonian methods of analysis to the workings of the mind. The further development of Hume’s ideas was delayed by the onset of an acute depression, which he tried to shake off by undertaking a career in business. In 1734 he abandoned business to go to France, taking up residence there at La Flèche, where Descartes had been educated. He had already taught himself French and had familiarized himself with such French sceptics as Pierre Bayle; in the extensive library at La Flèche he developed that intimate acquaintance with French philosophy which exerted so profound an influence upon him, uneasily coexisting with his Newtonianism.

Hume returned to England in 1737 with his Treatise of Human Nature completed. The first two books, “Of the Understanding” and “Of the Passions,” were anonymously published in 1739; the third book, “Of Morals,” was issued in 1740 with an important appendix containing his second thoughts. Hume was confident that the Treatise would create a sensation, but it was unenthusiastically received. In order to draw attention to its merits, he published what purported to be an anonymous review of the first two books as An Abstract of a Treatise of Human Nature (1740). As an advertising device, it failed, but the Abstract is a useful guide to Hume’s philosophical intentions, especially interesting for the stress it lays on his associationism. Concluding that the failure of the Treatise was a consequence of its length and complexity, Hume henceforth expressed his ideas more fashionably—in essays and dialogues.

In 1741, with a second volume in 1742, Hume published his Essays Moral and Political. It is often said that Hume abandoned philosophy for economics and politics in search of literary fame. But for Hume philosophy was “the science of man,” and economics, politics, history—understood as “philosophy teaching by examples”—formed for him part of it. He modified his literary style to meet the tastes of his age, but not his fundamental conception of the philosopher’s task. The first book of the Treatise had been intended as his theory of social inquiry, his “logic”, the second book as his moral psychology; and the third as his ethics. It was now time to pass on to the other social sciences.

His new prose style having proved successful, Hume made another attempt to present his logic to the public. This time it was in an abbreviated and popular form, no longer as a treatise but as Philosophical Essays Concerning Human Understanding (1748), renamed in 1758 An Enquiry Concerning Human Understanding. This was the work which, he told his critics in an advertisement first published in the posthumous edition of 1777, should “alone be regarded as containing his philosophical sentiments and principles,” his Treatise being, he explained, but a juvenile work.

Philosophers have been unwilling to take Hume at his word, for the Treatise contains a great deal of interesting philosophical analysis, especially of perception, which is not to be found in the Enquiry. But the Enquiry is in many ways the best introduction to Hume, especially in relating his philosophy to the history of scientific thought. It contains, too, a number of important essays—on miracles, on liberty and necessity, and on providence—which are not to be found in the Treatise.

Hume followed up the Philosophical Essays with his Enquiry Concerning the Principles of Morals (1751), an abbreviated and considerably modified version of book III of the Treatise. Although Hume thought it to be his best work, it has only recently received the detailed attention it deserves. At about the same time, Hume wrote the first draft of his Dialogues on Natural Religion, a potent criticism of the traditional arguments for the existence of God and especially of the argument from design. His friends warned Hume against publishing it; it appeared posthumously in 1779.

The Enquiry Concerning Human Understanding had excluded the sections on space, time, and geometry which formed part of the Treatise. Hume intended to write, he tells us in one of his letters, a separate work on “the metaphysical principles of geometry.” He prepared for inclusion in Four Dissertations (1757) an essay entitled “Some Considerations Previous to Geometry and Natural Science,” but the comments of Lord Stanhope, an able mathematician dissuaded him from publishing it. Hume’s talents, indeed, did not lie in that direction; the sections on space and time in the Treatise add little to what Berkeley had already argued. For very different reasons, he was also persuaded not to publish his essays “Of Suicide” and “Of the Immorality of the Soul”; these first appeared in an unauthorized French translation in 1770 and also in an unauthorized English edition in 1777 as Two Essays. He did include in the Dissertations, however, his “Natural History of Religion,” in which he sets out to show that classical mythologies are at once more reasonable and morally more enlightened than systematic Christian theology.

Knowing that he was about to die of cancer, Hume wrote in 1776 My Own Life, which was first published by his literary executor Adam Smith in 1777 and which is as much an apologia as an autobiography. He died after a long illness, bravely sustained. Hume was a man of exceptional personal qualities, nick-named in France “le bon David” and in Scotland “Saint David.” Adam Smith described him as “approaching as near to the ideal of a perfectly wise and virtuous man as human frailty will admit.”

Methodology. The subtitle of Hume’s Treatise describes it as “an Attempt to introduce the experimental Method of Reasoning into Moral Subjects.” Under “moral subjects” Hume includes logic, to which he assigns the task of explaining “the principles and operations of our reasoning faculty”; moral philosophy; political theory, which incorporates economics and history; and literary criticism. He sometimes wrote (as in the introduction to the Treatise ) as if he had fulfilled the common eighteenth-century ambition to be the Newton of human nature; as if, that is, he had constructed a science of man, paralleling physical science, by relating the elements of the mind in laws of association comparable to the laws of mechanics (Treatise, bk. I, pt. 1, sec. IV).

Hume’s important contributions to such moral subjects as economics and politics—he contributed nothing to and nowhere reveals any detailed knowledge of the physical sciences—did not depend on the use of a new method; he wrote as an intelligent and critical observer of the European scene, by no means as a methodological innovator. His approach is experimental only insofar as his explanations of social phenomena appeal to everyday human experience, rather than making use of such transcendental entities as “Providence.”

As for his positive methodology, that is dependent upon, and does not go far beyond, the “Rules of Reasoning in Philosophy” which Newton had laid down in the third book of his Principia mathematica. Hume himself wrote of his “rules by which to judge of causes and effects” (ibid., bk. I, pt. 3, sec. XV) that they are so obvious as scarcely to be worth the trouble of setting them out systematically. His importance lies not in his use or description of the experimental method, but quite elsewhere—in the doubts he raised about the rationality of the method.

His analysis of reasoning begins from a presumption universally accepted by his philosophical contemporaries, namely that what we are directly acquainted with are “perceptions in our mind,” as distinct from independently existing physical objects. Hume divides these perceptions into two classes, impressions and ideas. He counts as impressions not only sensations but any operations of the mind, including the passions, which are immediately apprehended. Ideas are “the faint images of impressions”; they are what men have before their mind when they think, as distinct from when they feel.

Since there are no ideas which do not derive from impressions, anybody who uses a word which purports to refer to an idea can properly be asked from what impression that idea derives. If the idea to which the word purports to refer does not derive from any impression, the word, Hume argues, must be meaningless (Abstract, p. 11). This is clearly the case, he tries to show, with such familiar metaphysical words as “substance” and “essence.” Hume’s analysis of perception thus provides him with a powerful polemical weapon to direct against all explanations that make use of concepts not derived from experience; explanations of this kind are, in his interpretation, mere word play.

Perceptions, whether impressions or ideas, occur in spatial and temporal sequences. Furthermore, very similar sequences of perceptions—“constant conjunctions”—regularly recur. Resemblance, spatiotemporal contiguity (in the Enquiry Concerning Human Understanding replaced by temporal priority), and constant conjunction are, according to Hume, “to us the cement of the universe” (ibid., p. 32). Men are able to progress from their perceptions to a belief in an orderly systematic world only by virtue of the fact that similar perceptions recur in particular ordered sequences.

Both science and common sense take it for granted, so Hume believes, that there are independently existing objects which are necessarily linked one with another (Treatise, bk. I, pt. 4, sec. II). Perceptions, on the other hand, depend upon the human mind for their existence and have no necessary connection with one another. Berkeley had rejected this contrast; perceptions and objects, he had argued, are identical, and science does no more than correlate perceptions. This analysis of scientific knowledge Hume dismisses, in spite of Berkeley’s protestations, as a form of absolute scepticism. Berkeley’s arguments, he says, if “they admit of no answer [yet] produce no conviction” (Enquiry, sec. XII, pt. 1). Although there are places in the Treatise (bk. I, pt. 2, sec. VI) where Hume writes as if he were a phenomenalist, he for the most part—particularly in the Enquiry (sec. XII, pt. 1)—takes it for granted that there are physical objects which give rise to perceptions in us. He does not seriously question, that is, the general world view constructed by Galileo, Boyle, Newton, and Locke: he asks, rather, what grounds we have for believing in its truth.

So long as science does no more than describe and compare perceptions no problem arises. Mathematics, according to Hume, is secure knowledge because it restricts itself to relating ideas one to another (Treatise, bk. I, pt. 3, sec. I). This is true, at least, of algebra and arithmetic; in the Treatise and the Abstract, although not in the Enquiry, Hume expresses some doubts about geometry. Nor is there any problem with what Hume calls “mental geography” so long as it confines itself to the “delineation of the distinct parts and powers of the mind” (Enquiry, sec. I).

In his more sceptical moods, admittedly, Hume does not allow even mathematics and “mental geography” to escape unscathed. Although the rules of mathematics are “infallible,” he says, the fact remains that mathematicians themselves are properly hesitant about the validity of their proofs and fully accept them only when their colleagues do so (Treatise, bk. I, pt. 4, sec. I); as for “mental geography,” that breaks down when it tries to give a satisfactory account of personal identity (appendix to Treatise, note to bk. I, pt. 3, sec. XIV). But to carry scepticism to the point of questioning the certainty of mathematics and “mental geography,” Hume suggests, is to carry it beyond the point at which it is humanly possible consistently to be a sceptic (Treatise, bk. I, pt. 4, sec. I).

The case is very different, Hume thinks, with what he calls matters of fact, assertions which go beyond perceptions by referring to independently existing, continuous objects and ascribing to them a necessary connection with other objects. Whenever the scientist makes a “matter-of-fact” assertion, according to Hume, he is relying upon some form of causal reasoning. Only causal reasoning can carry the mind beyond what it actually perceives to beliefs about what it has not perceived, for example, from beliefs about perceived smoke to beliefs about unperceived fire (ibid., bk. I, pt. 3, sec. II). Only if causal reasoning is rational, then, can science be securely grounded.

It cannot be demonstrated, Hume is confident, either that whatever happens has a cause or that a particular occurrence is the cause of a particular effect. (Hume counts as demonstrative only those arguments which prove that it is logically impossible for the conclusion to be false.) Metaphysicians who profess to demonstrate that every event has a cause always beg the question. Every perception, Hume tells us, is distinct and separate from every other perception. There can be no contradiction, then, in supposing that a perception, that is, without a cause (ibid., bk. I, pt. 3, sec. III).

For the very same reason it is impossible, according to Hume, to demonstrate that a particular effect has a particular cause. Since perceptions are distinct and separable there is nothing in any perception, taken by itself and prior to any further experience, which logically presupposes the existence of any other perception (ibid., bk. I, pt. 3, sec. VI). Our everyday experience confirms this philosophical conclusion. Prior to experience we have no way of telling how anything will behave, that fire, for example, will burn rather than thicken the human skin. Neither the effect itself, as Descartes thought, nor a power to produce the effect, as was widely presumed, is implicit in the cause; if it were, the scientist should be able simply by examining an object to discover what effects it will have, and this is impossible.

Only experience, then, enable the scientist to determine that a particular cause will have a particular effect. But experience tells him only that in the past certain similar perceptions A ₁, A ₂, A ₃,... have been constantly conjoined with certain other similar perceptions B ₁, B ₂, B ₃,... When the scientist holds that A is the cause of B, however, he ordinarily thinks of himself as being committed to something much stronger than this: that A is necessarily connected with B. Yet he has had no experience of necessary connection, as distinct from mere conjunction. Nor is there any general principle which would enable him to move from “B has, always in the past been produced by A ” to “B is necessarily produced by A.” It is quite easy to imagine a change in the course of nature such that A and B will no longer be constantly conjoined one with another; this is by no means a logical impossibility. Hence, Hume concludes, it is impossible to demonstrate that B cannot occur without A ’s having occurred. Anybody who perceives the conjunction may be led to believe that A and B are necessarily connected, but this “being led” is a psychological fact, not a logical necessity. It is not that there is a valid inference from constant conjunction to necessary connection; the belief that A is necessarily connected with B is reducible to the fact that we habitually suppose that A must have happened when B is perceived and expect B whenever A is perceived (ibid., bk. I, pt. 3, sec. XIV).

To understand scientific inference, then, we must turn to mental geography and the analysis of our mental habits, not to formal logic. The belief in any matter of fact has only two sources: the existence of a particular relationship between perceptions—constant conjunction—and the tendency of the mind to react in a certain way to constant conjunctions. That is why Hume is prepared to assert that the science of man is the fundamental science on which all other science rests; only with the help of mental geography can we explain why we hold our empirical beliefs.

If we ask, however, exactly what mental geography tells us about nondemonstrative inference, Hume’s answer is by no means clear or consistent. Sometimes he says that reason (that is, empirical reasoning) is “nothing but a wonderful and unintelligible instinct in our soul” which leads us to move from past experience to expectations about the future (ibid., bk. I, pt. 3, sec. XVI). This has led some commentators to assert that Hume is a naturalist who, in the manner of Pope’s Essay on Man, bids us rely on instinct rather than reason for our fundamental beliefs. At other times, however, the responsibility for causal inferences is assigned by Hume to the imagination.

Just how sceptical is Hume’s analysis of empirical inference? That, too, is a point on which he vacillates. On the one hand, he is anxious to dispute the claims of transcendental metaphysicians and theologians that they possess rationally grounded beliefs. With his eye on such opponents, he argues that it is quite absurd to go in search of remote causes for the Universe when we cannot even give a satisfactory reason for believing that a stone will fall or that the sun will rise tomorrow (Enquiry, sec. XII, pt. 3). A belief, he says, is nothing but an unusually vivid idea; to believe that the sun will rise tomorrow is simply to have a vivid idea that it will do so. This doctrine, too, is useful against those who argue that the moral sciences are intrinsically inferior to the physical sciences because they rest upon feeling; every form of science, Hume can reply, does so (Treatise, bk. I, pt. 3, sec. VIII).

On the other hand, Hume is equally anxious to destroy fanaticism and superstition. He can scarcely deny, however, that the superstitious and the fanatical have vivid ideas. He sometimes suggests, therefore, that a belief is rational provided only that it can be traced back to a constant conjunction; hence the rational justification for believing that the sun will rise tomorrow, as opposed to the irrationality of superstitious beliefs. From this perspective Hume distinguishes between demonstrations, proofs, and probabilities. It is ridiculous, he says, to declare as only a probability that the sun will rise tomorrow or that all men are mortal (ibid., bk. I, pt. 3, sec. XI). Inferences from constant conjunction, he suggests, are properly describable as proofs, even though they clearly do not constitute demonstrations. But when conjunctions are irregular—A being only sometimes conjoined in our experience with B, and sometimes with something else—the proper inference is only to probabilities, since the probability of a conclusion depends upon the relative frequency of the conjunctions on which it is founded. The conclusions of the superstitious have a zero or minimal probability because they are contrary to our regular experience.

This attitude is most fully developed in Hume’s critical analysis of the belief in miracles (Enquiry, sec. X). Hume there begins by asserting that a wise man will always proportion his belief to the evidence. A miracle is by definition a violation of the laws of nature, that is, an event which is contrary to our regular experience. The evidence in its favor, as in the case of those miracles on which the historical religions rely, is that some witness or an oral tradition tells us that the miracle happened. We are entitled to accept this testimony only, Hume says, if it would involve a greater miracle, a more manifest divergence from all past experience, to suppose that the testimony is false. Since this condition is not satisfied in the case of any recorded miracle, he says, we cannot properly treat miraculous occurrences as probable, let alone as proved.

Hume sometimes expresses his theory of “proof” in a way that links it closely with the workings of the imagination. The imagination, he tells us, has certain regular, associative ways of working, most clearly manifested in the case of causal inference. These ways we must accept as reliable and rational; to reject them is to undermine the whole foundation of our thought and action. The imagination, however, does not always work in a regular way; it has irregular and erratic tendencies which lead men into superstition. Conclusions derived from these irregular workings ought, on the face of it, to be rejected by rational men (Treatise, bk. I, pt. 4, sec. IV). The problem is that there exist unquestionably true beliefs—the belief in the independent existence of physical objects and the belief in personal identity, for example—which cannot wholly be explained in terms of causal inference, but which depend on the operations of irregular propensities of the imagination. So it is impossible, after all, to adopt a policy of accepting only those beliefs which are founded on constant conjunction (ibid., bk, I, pt. 4, sec VII).

In the Treatise especially, these considerations sometimes lead Hume to a posture of absolute scepticism, rather that the “mitigated scepticism” he generally adopts. But no man can live as an absolute sceptic (Enquiry, sec. XII). Mitigated scepticism, as Hume sums it up in his Dialogues (pts. VIII adn IX), asserts simply that it is impossible to demonstrate any matter of fact and that the nature of our experience, not some a priori principle of rationality, determines what we find intelligible. Such a position is substantially that of expiricism. But it is a different matter if our fundamental beliefs turn out to rest on nothing more solid than a trick of the imagination. We have only one defense against this sceptical conclusion, Hume suggests. Nature has not left our beliefs entirely to our choice; we cannot help coming to conclusions any more than we can help breathing (Treatise, bk. I, pt. 4, sec. I). Mitigated scepticism is therefore useful, for it prevents us from wandering into the wilds of metaphysical speculation by impelling us to reflect on the limits of our knowledge of even everyday physical experience and relationships.

BIBLIOGRAPHY

I. Original Works. The classical edition, although an imperfect one, of Hume’s works is T. H. Green and T. H. Grose, eds., The Philosophical Works of David Hume, 4 vols. (London, 1875). This does not include J. M. Keynes and P. Sraffa, eds., An Abstract of a Treatise of Human Nature (Cambridge, 1938), or Ernest C. Mossner and J. V. Price, eds.,A Letter from a Gentleman to His Friend in Edinburgh (Edinburgh, 1967). Especially for their indexes, consult also L. A. Selby-Bigge’s eds. of A Treatise of Human Nature (Oxford, 1888) and Enquiries Concerning the Human Understanding and Concerning the Principles of Morals, 2nd ed. (Oxford, 1902). The best text of the Treatise is the Mossner ed. (London, 1969).

See also Norman Kemp Smith, ed., Dialogues Concerning Natural Religion, 2nd ed., with suppl. (London, 1947). For Hume’s general writings on religion see Richard Wollheim, compiler, Hume on Religion (London, 1963).

II. Secondary Literature. John Hill Burton, Life and Correspondence of David Hume, 2 vols. (Edinburgh, 1846; repr. New York, 1968), is still valuable. The best modern life is E. C. Mossner, The Life of David Hume (Austin, Texas, 1954; London, 1955), which includes The Life of David Hume, Esq., Written by Himself or, as entitled in the original MS, My Own Life. See also J. Y. T. Greig, ed., The Letter of David Hume 2 vols. (Oxford, 1932), and Raymond Klibansky and E. C. Mossner, eds., New Letters of David Hume (Oxford, 1954).

It is impossible to give a straightforward, systematic, noncontroversial presentation of Hume’s views. That is one of the principal themes of J. A. Passmore, Hume’s Intentions, 2nd ed., rev. (London-New York, 1968). The most thoroughgoing commentary is N. K. Smith, The Philosophy of David Hume (London-New York, 1941), and the most useful introduction is D. G. C. Macnabb,David Hume: His Theory of Knowledge and Morality (London, 1951).

See also Charles W. Hendel, Studies in the Philosophy of David Hume rev. ed. (Indianapolis, 1963), with an account of recent work on Hume; Antony Flew, Hume’s Philosophy of Belief (London, 1961; New York, 1962), which concentrates on the Enquiuries ; H. H. Price, Hume’s Theory of the External World (Oxford, 1940); and Farhang Zabeeh, Hume: Precursor of Modern Empiricism (The Hague, 1960).

John Passmore