Hume

‘Hume’, in S. Nadler (ed.), Blackwell Companion to Early-Modern Philosophy, Oxford: Blackwell, 2002, pp. 483–504 DAVID HUME HUME’S LEGACY When in April 1776 David Hume, aware of his impending demise, took a retrospective look at his life, he could confidently write in My Own Life that he had been a reasonably successful man. But even in the valedictory balance and stoicism of his brief autobiography he still recalled with a tinge of regret, if not of bitterness, the fate of his first published work, the Treatise of Human Nature: citing Alexander Pope’s Epilogue to the Satires, he wrote that it had fallen ‘dead-born from the press’. For a long time this assessment of the reception of the Treatise must have seemed right. It had first been published in three volumes in 1739 and 1740, and many of its not very numerous early readers and reviewers were uncomprehending and hostile. After this disappointment, Hume started publishing shorter and more readable works, essays on political and critical matters which, after being published in 1741-42, went through many recensions. He also recast in essay form what he felt could be salvaged from the Treatise in the Enquiry Concerning Human Understanding (1748) and then the Enquiry Concerning the Principles of Morals (1751)—and by the mid-1750s he was open in acknowledging that he found his unlucky first literary production positively irritating for the evident youthful eagerness inspiring it. The essays ‘The natural history of religion’, ‘Of the passions’, ‘Of tragedy’ and ‘Of the standard of taste’ were published in 1757 as Four Dissertations (the more provocative ones ‘Of suicide’ and ‘Of the immortality of the soul’, also at some point meant to appear in this collection, were first published much later, in French translation in 1770, and in English in 1777). The six volumes of his History of England (1754-62) gave him both a good income, and the literary fame he admits he had much wished for as a young man. His last major philosophical work, the Dialogues Concerning 1 Natural Religion, was to be published posthumously in 1779. By the time he was writing My Own Life Hume was known as the author of original and witty essays, as an important and elegant history writer, as a dangerous sceptic and possibly an atheist, and as a very nice man. Contemporary anecdotes about his amiability and kindness abound: as his friend James Caulfield wrote, ‘of all the philosophers of his sect, none … ever joined more real benevolence to its mischievous principles’; he was nicknamed the Socrates of Edinburgh; provided the model of the compassionate and humane sceptical philosopher in the sentimental novel-writer Henry MacKenzie’s ‘Story of La Roche’; indeed, his reputation for a virtuous character was only challenged by those who did not know him personally. In spite of Dr Johnson’s strongly expressed disapproval, his friend, admirer and biographer James Boswell knew and admired Hume too, and contemplated writing his biography. His portraits show that he was a big man, and there are stories about the deceitfully rustic and stupid look of his fat face. In the catalogue of the British Library he is “Hume, David, the Historian”. Contemporary responses to Hume and his philosophical work were, as I said, mostly unsympathetic; but their most striking overall feature is that they were very diverse, and very much at odds with each other. So Thomas Reid thought that Hume had reduced Locke’s empiricism and the whole ‘way of ideas’ to absurdity, while Kant claimed that Hume’s thought had woken him up from his dogmatic slumber. James Beattie and several others among his Scottish contemporaries found that Hume’s scepticism involved an intolerably irreligious, indeed atheistic stance, while Heinrich Hamann argued that it was in fact conducive to faith. The Hegelian Thomas Green regarded Hume’s philosophy as the Pyrrhonian outcome of Locke’s and Berkeley’s empiricism; among the agnostics, Leslie Stephen thought of him as a ‘systematic sceptic’ and Thomas Huxley made him, together with Kant, into a founding father of agnosticism. And so on. Today Hume’s Treatise is one of the great classics of Western philosophy, and Hume has become, in turn, one of the major figures in the canon. The diversity in the more recent responses to his philosophy is still impressive. Hume was one of the few recognised as ancestors by positivists, for what they regarded as his consistent empiricism and his antimetaphysical stance (Zabeeh, 1960); he was read as a founding father by the phenomenologists (Reinach, 1983; Salmon, 1929); and similarities have been found between the post2 modern approach and his antifoundationalism and irony (Parusnikova, 1993). Hume’s thought appears in the writings of Hans Reichenbach and Edmund Husserl, and there are books on Hume written by Alfred Ayer and Gilles Deleuze (Ayer, 1980; Deleuze, 1991). There is even a tradition comparing aspects of Humean philosophy with aspects of Buddhism (Stafford Betty, 1971). Hume’s presence in the genealogies of such disparate philosophical traditions must depend on some quality of ambiguity, or of richness, in his philosophical writings themselves: like all ‘classics’, they are endlessly stimulating and open to many different, even contrasting readings. Accordingly, in the scholarly literature there is, to date, hardly any issue in Hume’s writing on the interpretation of which one can find reasonably general agreement. It is evidently neither within the scope, nor among the aims of this essay to provide an assessment of the numerous interpretative and philosophical debates about Hume’s writings and thought. So in what follows I try to offer a balanced overview of some of the controversial points, but my attempt is also, inevitably, guided by my own bias in reading Hume’s work. In the Treatise and in the Enquiries Hume called his investigations a ‘science of human nature’. Indeed human nature may be taken as providing the unifying theme of his whole written work, from the abstruse analyses of the Treatise to the elegantly written Enquiries, from his political essays to his pieces of religious critique to his massive History. In his more explicitly philosophical writings, Hume begins with the treatment of traditional metaphysical and epistemological issues—the origin, nature and limits of our knowledge, cause and effect, the nature of our beliefs, the existence of external objects— and only moves on to the passions, to morals and to critical and political matters at a later stage. Given his emphasis on human nature, this means that he starts by treating the workings of the understanding. So the foundation of all discussion of human nature is an account of the operations of the mind. In this essay I follow his lead in this. After Norman Kemp Smith’s classic The Philosophy of David Hume (1941), many have agreed that Hume’s main interest is not metaphysical or epistemological, but moral. I share this conviction; indeed it is my belief that Hume’s philosophy is, first and foremost, a style of life. This does not necessarily mean that it is in his moral philosophy that we are to look for his most interesting or thought-provoking contributions. I therefore start with Hume’s account of perception, and proceed with his discussions of the idea of cause and effect 3 and of the inference from the past to the future, his scepticism and his views on the nature of philosophy. I then discuss his contributions to moral and political philosophy, and conclude with his critique of religion. SENSE IMPRESSIONS, PASSIONS AND IDEAS Our experience, according to Hume, begins with what he calls ‘impressions’: sense impressions and passions (or, in his terms, ‘impressions of reflexion’). I see the black marks on the white page of a book, and wish to read it. These impressions are subsequently duplicated in our mind. That is, they are copied into ideas: in this case, roughly speaking, the ideas of certain colours, shapes, textures etc. which compose the complex idea of the page of the book, and the idea of my wish to read it. All our basic experiences are, according to Hume, of this sort: sense impressions and passions come before thinking, are followed by it, and are duplicated in it (Treatise, p. 1). This is Hume’s so-called ‘copy principle’, which he presents as his ‘general maxim’ and the ‘first principle’ in his science of human nature (pp. 6, 7). Application of it will, in his view, rapidly expose all obscure and confused notions and pseudo-ideas. Ideas are copies of impressions; they are not quite so vivid and compelling—‘lively’ and ‘strong’—as impressions, but apart from this they are accurate duplications of impressions. The distinction between impressions and ideas is reflected in their different force and liveliness. This difference is crucial because when we are faced with obscure and confused ideas we can, on the basis of the copy principle, trace them back to their more vivid and clearer originals, thus easily sorting out all difficulties and, when necessary, exposing fictions which, due to our natural inclination to relish paradoxical and far-fetched notions, we had taken for genuine ideas (p. 33). As many readers have noticed, however, difference in strength and liveliness is neither an unambiguous, nor an entirely clear criterion for the distinction between impressions and ideas (Stroud, 1977, pp. 27-33). For a start, it is evident that these terms are not synonyms—what is Hume trying to convey by them? When using them to define belief, he realises that they are not satisfactory (Treatise, p. 105); but all he does about it is to acknowledge the problem and 4 to introduce a whole new list of terms, ‘firmness, or solidity, or force, or vivacity’ (p. 106, see also p. 629). Also, it is not difficult to imagine counterexamples, cases in which an idea may turn out to be more ‘lively’ than the impression from which it is copied. And while it is true that, as Hume points out, we all ‘readily perceive the difference betwixt feeling and thinking’ (p. 2), does this mean that one is therefore exonerated from accounting for it? These difficulties appear less serious if we keep in mind that the fundamental difference between these two kinds of perceptions is that impressions are the originals of which ideas are the copies. With this, however, problems are by no means at an end. Of course, in the case of more complicated or ‘complex’ ideas, things are less straightforward: for example, I cannot honestly say that my idea of Paris is an accurate copy of any particular impression or set of impressions (p. 3). But the basic principle remains: all ‘simple’ ideas are copies of ‘simple’ impressions (pp. 3-4). The only exception Hume is prepared to acknowledge is that of a continuous gradation of a quality—his example is the colour blue—which is only missing one particular grade: in his view, in this case our mind would be able to fill the gap, and produce an idea without copying it from a former impression (p. 6). The case of this ‘missing shade of blue’ is a notorious brain-teaser for students of Hume (Fogelin, 1992, pp. 70-80). Is Hume trying to suggest that all exceptions to the copy principle are similarly far-fetched? Is it really all that far-fetched? And in any case, does it not mean that the copy principle is to be regarded as an empirical generalisation, rather than an a priori principle? But if so, how is Hume entitled to use the copy principle to vet, so to speak, our ideas, as he seems enthusiastically and often rather aggressively inclined to do? (see for example the opening of the section ‘Of modes and substances’, Treatise, p. 15; also pp. 72-3). To answer the last question we should now turn to consider what use Hume makes of this principle. It then becomes apparent that all the ideas Hume sets out to investigate through the copy principle turn out to be, for a variety of reasons, recalcitrant to it. There is a whole range of these unobvious and difficult ideas: for example, there are the ideas of space and time and the ideas of empty space and changeless time, the idea of substance, the idea of our own self, the idea of virtue and the idea of the necessary connexion 5 between cause and effect. The use of the copy principle produces a taxonomy of these odd ideas. Some of them create serious problems, giving rise to philosophical puzzles which Hume openly admits he does not know how to solve, while others are indeed found to copy some impression, but not of the sort one would have expected. A few—in fact surprisingly few, given his stated intentions—are simply rejected as pseudo-ideas. Let us consider some examples. The investigation into the origin of the ideas of space and time is one of the first applications of the copy principle to be found in the Treatise. Objects of widespread metaphysical and natural-philosophical dispute in the age of Newtonianism, these ideas seem to be an ideal case for the application of the copy principle. And this is the way Hume presents the matter: he starts his discussion by saying that nothing could be more fortunate than to have such a tool as the copy principle to further our investigation into their nature (p. 33). The sight of the purple surface of my table, he goes on to say, is enough to give me the idea of space. Once we try to follow him and unpack the complex impression of the table’s surface to single out the individual impression at the origin of the idea of space, however, we run into a difficulty which he does not explicitly acknowledge: there is no impression of space separate from the individual impression of the coloured points; all we can find is the impressions of coloured points ‘disposed in a certain order’ or ‘manner’. The idea of space is, Hume says, copied from the overall ‘order’ or ‘manner of appearance’ of the impressions of the points (p. 34). Similarly, the idea of time is copied from the overall ‘manner of appearance’ of, say, five notes played on a flute, rather than from a distinct sixth impression (p. 36). In other words, there are no distinction impressions of space and time corresponding to the ideas. (Something similar also applies to the idea of existence, p. 66-8.) Hume does not, however, conclude that since they apparently fail to satisfy the copy principle the ideas of space and time are only pseudo-ideas. Rather, he seem to claim that they are reflections in thought of features of our complex impressions. By contrast, after explaining that there is no impression, either of the senses or of reflexion, at the origin of our putative ideas of empty space and changeless time, he shows that it is the way we talk of matter and motion which misleads us into thinking that we do have such ideas. In this, 6 as in so many cases, we mistake words for ideas. Together with the idea of substance, the ideas of empty space and changeless time are among the very few putative ideas which Hume suggests philosophers should be prepared to regard as mere fictions or pseudoideas. He does not, however, maintain that the corresponding vocabulary is to be discarded: we may well follow Newton and use the word ‘vacuum’ in accounting for the motion of bodies, if we are so inclined, provided we are aware of its lack of any psychological or epistemological, let alone metaphysical, underpinning, just as we may well decide to follow the Cartesians in avoiding it, and be aware of the consequent difficulties in accounting for motion (pp. 61-2 and 638-9). The idea of self is similar to the ideas of space and time: there neither sense impressions nor impressions of reflexion which are good candidates for the impression corresponding to the idea of self. For we cannot pin down any impression which is distinct and self-contained, and yet is common to every single one of our perceptions and identifies each and all of them as ours. Our mind is like a theatrical performance of our perceptions (p. 253); it is ‘a bundle or collection of different perceptions, which succeed each other with an inconceivable rapidity, and are in a perpetual flux and movement’ (p. 252). The identity of our self is therefore similar to the identity of a river (p. 258) or of a republic (p. 261), that is, an identity enduring through a constant change of component elements. And yet it is obvious that we do have a particularly vivid idea of our own self—so much so that later in the Treatise, in the discussion of passions in Book 2, Hume suggests this idea is rather like an impression (p. 17). THE IDEA OF CAUSE AND EFFECT So in the case of the ideas of space and time and of the self Hume’s application of the copy principle has given rise to a genuine philosophical puzzle: there is no doubt that we do possess such ideas, and they do seem to have some empirical basis; yet they cannot be regarded as copies of any distinct impression. We know that at least in the case of the idea of self Hume openly acknowledged that there was indeed an unsolved puzzle in his theory: he says so in the appendix published at the end of Book 3 of the Treatise (pp. 7 633-6). Ideas such as those of virtue and vice, and of ‘necessary connexion’ between cause and effect present yet another sort of anomaly. These ideas are indeed copies of corresponding impressions but, contrary to our expectations, these turn out to be not sense impressions, but impressions of reflexion. Consider the ideas of virtue and vice. These ideas, Hume maintains, are not copied from specific distinct qualities in a virtuous or vicious person or action, as one would imagine. Parricide is, Hume says, the worst of crimes; and yet we do not disapprove of a sapling which, by its steady growth, ends up destroying the parent tree, even though it reproduces parricide in every detail. Incest is a criminal action among humans, but we regard it as completely innocent among animals. The difference between these cases cannot have to do with humans being rational and hence in a position to know better. For reason can only uncover what is already there, without giving rise to anything new (p. 466-8). There is neither virtue nor vice in the objects themselves: vice and virtue are, like colours, tastes and the other secondary qualities, ‘not qualities in objects, but perceptions in the mind’ (p. 469). The ideas of virtue and vice are copies of our own feelings of approval or disapproval—in other words, it is our approval or disapproval that give rise to moral distinctions, and not vice versa. So when we imagine that the origin of the ideas of virtue and vice is something out there, in a quality of the virtuous or vicious person or action, we mistakenly project an internal impression onto the objects. The case of the ideas of virtue and vice is based on a more general feature of the mind, its ‘great propensity to spread itself on external objects’ (p. 167). Some interpreters have noticed that he idea of the necessary connexion between cause and effect is similar, in this respect, to the idea of virtue (Stroud, 1977, pp. 177-8, and Garrett, 1997, pp. 107 and 202-3; but cf. Kail, 2002). What is it that we call a causal sequence? When I see a cause followed by its effect, all I actually see is, Hume points out, an A followed by a B. There is nothing in what I see to give rise to the idea of an interaction between the two. Nor is there any impression of reflexion of this sort—there is no impression of reflexion even in a causal sequence involving my voluntary bodily motions. Yet if I keep observing more and more cases of the same causal sequence, after a certain number of repetitions I find that I have acquired the idea of a necessary 8 connexion between this cause and this effect. This new idea comes from the habitual association of perceptions, which has given rise to an expectation, when I have the sense impression of the cause, to conceive the idea of the effect and to transfer to this idea part of the strength and liveliness of the present sense impression of the cause. In this way I find that I am compelled to conceive the effect in a particularly strong and lively manner—in other words, I believe that it is going to occur. The feeling of expectation is an impression of reflexion which only arises as a consequence of repetition and habit and which, in turn, gives rise to the idea of necessary connexion. In short, the idea of necessary connexion derives from an habitual inference, rather than giving rise to it (Treatise, p. 88). We do perceive a ‘necessary connexion’ of sorts, after all; but it is a connexion between our sense impression of the cause and our strong and lively idea of the effect, and we mistake it for a connexion between the cause and the effect out there in the world. Hume’s discussion of the idea of cause and effect is one of the best known and most celebrated parts of his philosophy. It is important to remember that in his discussion he is not considering causation, but simply the ideas of cause and effect. He is not discussing the metaphysics of causation, but rather is offering a science-of-human-nature—that is, roughly speaking, partly epistemological, and partly psychological—account of the process which makes us conceive and expertly use causal notions and language. But it is also a matter of historical fact that his treatment of the ideas of cause and effect has given rise to an entirely new way of regarding the metaphysical problem of causation. This is hardly surprising. Before Hume, cause and effect were normally supposed to be somehow homogeneous with each other, with the cause being more ‘powerful’ or more ‘perfect’ than the effect. Now Hume is saying that, for all we know, ‘anything may produce anything’ (p. 173): the idea of cause and effect is not a priori, so we can stare at something as long as we like, but we will never be able to tell what sort of effect that particular something is going to have if we have no previous experience of anything like it. Now, if according to Hume there is no perceivable necessary connexion between cause and effect, it may well be that all there is to causation is a mere sequence of things taking place one after the other. But is it right that we have no grounds for referring to some causal ‘cement’ connecting things to each other ? And if so, how are we to define 9 causation so as to account for the ways in which we successfully employ rather complex and sophisticated causal notions? These and other related questions still constitute hot topics for philosophical discussion. Hume himself tried to answer by offering not one, but two definitions; with his discussion of our belief that, when we see the cause, the effect will follow; and by bringing to our attention a set of rules which guide causal judgments. Hume’s first definition of cause and effect is in terms of pairs of constantly conjoined objects or perceptions such that without the ‘cause’ the ‘effect’ would not take place; while the second definition says that cause and effect pairs are such that the present impression of the cause determines our mind to form a lively idea of the effect, that is, the expectation and the belief that it will occur (p. 170). Much has been written to show that these two definition are not very satisfactory, either individually or as a pair. For example, neither of them accounts for what actually binds an individual cause-and-effect pair, hence neither of them can account for a cause operating once only. Also, it is easy to imagine pairs of events which would count as causal according to the first, but not according to the second definition, and vice versa. Certainly Hume’s two definitions do not provide a useful basis for a metaphysics of causation. On the other hand it is also true that, as has been recently pointed out, they provide a consistent, indeed a very good science-of-human-nature account of the idea of cause and effect, if we take the first definition in the subjective sense—the constant conjunction is from the point of view of an observer— and the second in the absolute sense—the beliefs described are those of an ideal observer (Garrett, 1997, pp. 112-4). The two definitions are an expression of Hume’s conviction that our ideas of cause and effect do not involve any ontological assumptions about causation. In his view, they can be fully accounted for by a combination of feeling and practical rules. To start with feeling: as is clear from the second definition, our expectation that a cause will be followed by the effect usually accompanying it is a crucial component of our idea of cause and effect. Belief is indeed a major character of Hume’s discussion of the whole issue. The way he talks about belief is deliberately reminiscent of his discussion of the distinction between impressions and ideas: a believed idea is stronger and more lively 10 than a merely entertained one, just as an impression is stronger and more lively than the corresponding idea. So when we are in the presence of the sense impression of the cause, this impression lends part of its strength and liveliness to the idea of what is habitually associated with it, that is, to the idea of the effect (Treatise, p. 98). In this sense our guide in our causal judgments, that is, in all our judgments concerning matters of fact, is a feeling: it is a propensity of our mind—its propensity to associate ideas, to acquire habits, to give ideas an emotional colour, so to speak—that produces our recognition of some sequences as causal, and of some others as casual. Commenting on Hume’s role in the history of aesthetics, Terry Eagleton has attributed to him the ‘alarming claim’ that both cognition and morality are, in this sense, ‘aesthetic’ (Eagleton, 1990, p. 45): as Hume himself put it, ‘’tis not solely in poetry and music, we must follow our taste and sentiment, but likewise in philosophy’ (Treatise, p. 103). So our causal judgments turn out, in the end, to be based on a faculty which is far less steady and reliable than reason or the understanding in its operations: the imagination. This is where the rules help us out—our reasonings about causes and effects do not result simply from feeling and the imagination. For the imagination itself not only controls the associations of ideas guiding us to project from the past to the future, from the observed to the unobserved, and so on; it is also the faculty responsible for the most unwarranted flights of our fancy and for our tendency to be gullible when faced with exciting and entertaining fictions. But feelings and the imagination constitute only one side of causal judgments. The other side is a ‘logic’ constituted by a set of ‘rules by which to judge of causes and effects’. These include the spatial and temporal contiguity of cause and effect, the priority of the cause, and their constant conjunction; that the same cause always produces the same effect, and vice versa; that different causes produce similar effects in virtue of their similar features, and vice versa; and so on (pp. 173-4). Hume’s cavalier conclusion is that no other logic, no other presupposition is necessary for us to reason correctly: these rules are more than enough (p. 175). Hume really is very laid back here; but, as Baier pointed out, the presence of the scanty three pages devoted to this ‘logic’ seems to suggest that he is not only trying to demolish old prejudices about cause and effect, but also to suggest an alternative account (Baier, 1991, pp. 56-7). 11 PROBABILITY AND THE INFERENCE FROM PAST TO FUTURE Hume is clear that we can hardly overestimate the importance of the relation of cause and effect to our experience and to knowledge. For, he points out, by enabling the mind to pass from the impression of the cause to the idea of the effect, it allows us to associate something present to us to something absent. In this way it provides the foundation for our inferences from past to future, and more generally from what we are directly acquainted with to other matters—it is the basis of our inductive inferences (Treatise, pp. 73-4). Posing the problem of induction is widely regarded as another of Hume’s crucial and original contributions to philosophy. This is how he put it. Our reason can only ever produce either demonstrative arguments about relations of ideas, as in mathematics and in all deductive forms of reasoning, or probable arguments about matters of fact. Now, it is clear that there cannot be a demonstrative argument to the effect that the course of nature must be uniform: a change being conceivable, and hence possible, uniformity cannot be established a priori. On the other hand, all probable arguments are based on the presupposition that the course of nature is indeed uniform, therefore there can be no non-circular argument for uniformity. But as all arguments available to us are either demonstrative or probable, no argument can prove the uniformity of nature (Treatise, pp. 89-90; Enquiries, p. 30). Or in other words, inductive arguments cannot by definition be justified deductively, while any attempt at an inductive justification begs the question. So, since deduction and induction are the only modes of inference available to us, a justification of induction is impossible. What grounds then do we have to believe that the future will resemble the past, for example, that because the sun has risen daily in the past it will rise again tomorrow? Hume’s answer is naturalistic: repetition naturally gives rise to a habit, and equally naturally the habit brings about belief. So the belief results from the way the human mind works (Treatise, pp. 104-6; Enquiries, p. 36); human reason itself is not, after all, all that different from the instinct of animals—it ‘is nothing but a wonderful and unintelligible instinct in our souls’ (Treatise, p. 179). 12 Curiously, Hume never used the word ‘induction’ in what we commonly consider the typical Humean sense. His discussion of the (Humean) problem of induction belongs to a much wider discussion of what he calls ‘probability’. In the Treatise this discussion includes probabilities amounting to proofs (there is really no doubt that the sun will rise tomorrow, or that all men must die, p. 124), the balancing of probability in the presence of different degrees of uncertainty (as in throwing differently biased dice, p. 127), and various forms of ‘unphilosophical probability’ (as, for example, in the cases of prejudice, pp. 146-7), as well as some very interesting, if scattered remarks on the assessment of testimony (pp. 145-6, also pp. 83-4). Issues concerning probability and probable inference were very popular among contemporary readers. So it is no surprise that the lengthy discussions of this part of the Treatise, now on the whole neglected by philosophers, are the ones which were discussed reasonably favourably in some of the contemporary reviews (such as the one published in the Nouvelle Bibliothèque in July and September 1740). Hume had intended to devote the culminating section of this treatise-within-theTreatise on probable arguments to the assessment of the probability of reported miracles. His view of the matter, as it is presented in the essay ‘On miracles’ in the Enquiry concerning Human Understanding, is as follows: no human testimony can be strong enough to establish a miracle, except in cases where the falsity of the testimony would have to be regarded as the greater miracle (Enquiries, p. 91); no testimony for a miraculous event has, as a matter of fact, ever been in this position (p. 98). As he put it in the dead-pan conclusion to the essay, ‘the Christian Religion not only was at first attended with miracles, but even at this day cannot be believed by any reasonable person without one’: for a believer must be aware of a true miracle in his own person, allowing faith to subvert all the principles of his understanding and determining him ‘to believe what is most contrary to custom and experience’ (p. 101). Here Hume was intervening in an important and widespread contemporary debate. He was aware that his approach to the matter was likely to give offence—this is why he cut out all discussion of miracles from the Treatise. When his views were published in the first Enquiry, as he had anticipated, they did not fail to attract plenty of attention: ‘On miracles’ is one of the most notorious pieces of Humean philosophy, and among the responses it aroused there is 13 even one, Richard Whately’s Historic doubts relative to Napoleon Bonaparte (1849), which tried to show, on the basis of Hume’s criteria of assessment of evidence and testimony, that the French emperor had never existed. (MODERATE) SCEPTICISM The study of the understanding undertaken in Book 1 of the Treatise of Human Nature is concluded with a Part 4 entitled ‘Of the sceptical and other systems of philosophy’. Here Hume discusses a variety of issues, from the certainty of demonstrative knowledge and the mathematical sciences, to the distinct and continuous existence we attribute to external objects; from the ancient philosophers’ occult qualities, to the distinction proposed by modern philosophers between primary and secondary qualities; from the doctrine of the immateriality of the soul to the theory of the substantial self. In these discussions Hume’s writing presents major interpretative problems: in some parts oddly personal, in other cases bitingly sarcastic, always very ‘literary’; sometimes in perfect ironic balance, but more often ambiguous between mockery and candour, between spoof and serious philosophical argument, it is impossible to bracket off matters of style and rhetoric to attempt a purely philosophical reading of these pages. Unsurprisingly, for some of Hume’s readers the Enquiry concerning Human Understanding is a clear improvement on the Treatise on account, among other things, of his having all but abandoned most of these discussions. According to others, however, in spite of its interpretative difficulties Part 4 is to be regarded as one of Hume’s most important pieces of philosophy: for this is where we have a chance to uncover the crucial moral significance of his investigations of human nature and the understanding (Baier, 1991, ch. 1; Livingston, 1998). It is useful to begin with Hume’s discussion of our belief in the existence of external objects, in one of the most notoriously difficult sections of Part 4, the one ‘On the scepticism with regard to the senses’. Yet again, it is important to remember that Hume is not asking a metaphysical question. Our belief in the existence of external world is such an unbreacheable fact of our nature, that questioning it would not even make proper 14 sense (Treatise, p. 187). But it makes very good sense to investigate human nature: where does this unshakeable belief come from? Hume first describes the naïve or absentminded attitude of the ‘vulgar’, those who are not thinking of metaphysical matters at all—all of us, Hume says, including the deepest philosophers, belong to this group most of the time (pp. 205, 206). When we are getting on with the ordinary business of life, or when we are thinking, however deeply, about something else, we take it for granted that our perceptions and external objects are one and the same. Hume then follows this naïve consciousness through the doubts instilled by the first opening of metaphysical questions: reason tells me that my perceptions are fleeting and volatile, and yet I do not imagine the world to be annihilated and brought into being again at my every eyeblink. Objects must be stable. But then, how can perceptions and objects be one and the same? According to Hume, this contradiction is created by the conflict of two faculties of the mind: reason, which discovers the brokenness of perceptions, and the imagination, which, on the basis of the constancy and coherence of our fleeting impressions—the world looks very much the same after an eyeblink, after all, and it only changes, when it does, according to regular and familiar patterns—constructs the notion of continuous objects. The natural outcome of this stage is the formation of the false philosophical consciousness which Hume calls ‘the opinion of a double existence’ (p. 211). Perceptions and objects are divorced from each other: so that at the same time reason is granted, in the sphere of perceptions, the interruptions it has discovered, and the imagination, in the sphere of the so-called external objects, the continuity that it has itself constructed. In other words, this duplication of the world only arises out of the irresoluble contrast between the two equally strong suggestions of reason and the imagination, which combine rather than neutralising each other. The theory of the double existence is, in this sense, parasitic on our pre-philosophical conviction that our perceptions are the only objects and, at the same time, that objects still exist when we are not perceiving them: but then, far from dissolving the contradiction, it is no more than an expression of it (pp. 212-3). The resolution of this conflict can only take place at a new level, that of ‘true philosophy’, with the acknowledgement that the nature of our mind itself both makes the whole dialectic inevitable and provides the remedy: our inability to maintain the intellectual tension necessary for philosophical reflection means that, in spite of such intense arguing 15 and counterarguing, once we return from our closet to the ordinary business of life we rapidly snap back into the absent-minded naivety, or, as Hume calls it, into the ‘carelessness and in-attention’ of the vulgar and of the true philosopher alike (p. 218). The dialectical pattern of false starts, errors and failed attempts in evidence in this discussion is also to be found in the overall structure of Part 4, where, it has been suggested, Hume seems to be offering a genealogy of philosophical consciousness, or a ‘cavalcade’ of philosophical errors (Livingston, 1998). Insofar as it is to be grasped by limited and fallible human minds, even the certainty of deductive reason is open to Pyrrhonian doubts (section 1); as we have seen, the same applies to our senses, which, in spite of the compelling immediacy with which they present their data, cannot in fact give us access to the world (section 2). Ancient and modern philosophers alike are defeated in their attempts to make sense of the qualities of bodies (sections 3 and 4); and neither theologians nor philosophers are able to account for the unity of body and mind or for the nature of our self (sections 5 and 6). After such debacles, what are we to think of our own philosophising, of our own reason, indeed of ourselves? The (temporary) triumph of radical scepticism Hume voices at the start of the final section of Book 1 is due not to the impact of an argument or set of arguments, but to the melancholy, despair and solitude brought about by so much intense and apparently fruitless philosophical reflection. Hume represents this state of mind in such vivid terms as to suggest close comparison with the autobiographical testimony he left, in the form of a letter written in 1734 to a physician, of a crisis he suffered in his twenties (see for example Sitter, 1982, pp. 26-33). The answer to melancholy lies, yet again, in doing philosophy ‘in a careless manner’: a true philosopher is able to doubt his philosophical doubts as well as his philosophical convictions. This approach has suggested to some that Hume was not a ‘true philosopher’: that he was very clever, but superficial and not serious (Taylor, 1934, p. 365; Pritchard, 1950, p. 174). I think that the opposite is the case. When philosophising, Hume follows his passion for truth with the utmost keenness and concentration—and he certainly is good at that; his ‘careless manner’ means, in my view, that Hume takes philosophy seriously enough to realise that there is a time for a philosopher to abandon his abstruse speculations, to return to more ordinary, habitual 16 ways of thinking, and to rejoin human society and polite conversation (Treatise, p. 273). This is how the true philosopher goes about doing philosophy, and this is the essential feature of Hume’s ‘moderate’ or ‘mitigated’ scepticism: a combination of problemsolving eagerness in the first-order philosophising exercise and, at the metaphilosophical level, the wisdom of alternating the practice of philosophy with that of sociability and conversation (Baier, 1991, ch. 1; Livingston, 1998). Thus these investigations and their outcomes show that the main focus of Hume’s philosophy, even when he is investigating the operations of the understanding, is not metaphysical and epistemological, but moral. MORAL FEELINGS So far I have discussed aspects of Hume’s study of human understanding. The investigations that we have considered are introspective and solitary; and it has been observed that Book 1 of the Treatise is dominated by the impression of the lonely philosopher thinking hard in the silence of his chamber (Sitter, 1982, pp. 23-24). As we have just seen, Book 1 concludes with a surprisingly personal representation of the dangers of philosophical solitude and the recommendation that we harmonise philosophical solitude and conversable sociability. Hume’s moral philosophy complements this recommendation with a study of human nature in common life and conversation. When he wrote that in philosophy, as in poetry and music, we must rely on sentiment, Hume did not have in mind only the study of the operations of the understanding: of course, he was also thinking of moral philosophy. In presenting the uses of the copy principle I have already mentioned that, just like the idea of necessary connexion between cause and effect, the ideas of virtue and vice are also copies of impressions of reflexion, namely, in this case, of our feeling of approval (or disapproval) for actions which are agreeable or useful (or the opposite) to the agent or to others. This reconstruction of the origin of the ideas of virtue and vice is typical of Hume’s approach to morality, as well as embodying several characteristic tenets of his moral philosophy. Let us try to unpack them. 17 For a start, in this case as in the rest of his moral philosophy, the focus of Hume’s discussion seems to be the spectator of action rather than the moral agent: as has been noticed by some interpreters, Hume seems deliberately to leave out any assertion of duties or obligations and to avoid putting forward any general normative doctrine. He describes human nature without actively engaging with issues of right or wrong (see Mackie, 1980, pp. 5-6). Such moral rationalists as Samuel Clarke had maintained that moral duties and obligations have nothing to do with the consequences of actions, and derive from nature prior to, and independently of, the authority of either god or man. In their view right and wrong are completely objective, eternal and necessary, and directly available for rational knowledge. Hume’s descriptive approach is an expression of his more general opposition to this rationalism in favour of a naturalistic approach. ‘Reason is and ought only to be the slave of the passions’ (Treatise, p. 415) is one of his most famous slogans: the conflict between reason and the passions and the notion that virtue is the triumph of reason over the passions are, in his view, mere philosophical mistakes. A passion can only be in conflict with and contrasted by another passion, and so only passions can motivate us to act or to refrain from acting. So it is clear that Hume agrees with Francis Hutcheson that morality is based on a ‘moral sense’—it is ‘more properly felt than judg’d of’ (p. 470). Virtue is by definition amiable, and vice odious: moral assessments proceed from sentiment. Reason has nothing to offer to explain the origin of moral distinctions, for there is nothing in virtuous or vicious actions corresponding to our idea of virtue or vice. In a similar vein Hume also observes that the shift from ‘is’ and ‘is not’ to ‘ought’ and ‘ought not’ expresses a new relation which is neither observed nor explained (p. 469). It is also clear from Hume’s discussion of the origin of the ideas of virtue and vice that at the foundation of our moral assessments is our ability to value something which is advantageous to someone other than ourselves: we develop our feelings of approval or disapproval because we can feel for and with others. In other words, morality is underpinned by our capacity for sympathy. Sympathy, our natural tendency to be cheerful with the happy and to mourn with the miserable, was vividly described in the colourful variety of its social manifestations by Hume’s contemporary and friend Adam 18 Smith in his Theory of moral sentiments (1759). It is also of paramount importance within Hume’s moral philosophy, where it is the root of all virtue. We find other people very similar to ourselves, Hume says, and part of the extraordinary liveliness with which we always perceive our own self is transferred to the conception of the feelings and passions of others. Due to sympathy, feelings and passions tend to be infectious: the idea of another’s passion can easily be so enlivened by sympathy as to turn into that very passion (Treatise, p. 319). What we observe in others is, of course, no more than the behavioural side of passions. But this is enough to set us off: ‘as in strings equally wound up, the motion of one communicates itself to the rest’ (p. 576). When we observe in another’s demeanour the effects of a passion, our mind naturally moves to conceiving the cause—the passion itself—with such strength and liveliness as to give rise to the same passion in ourselves. Finally, sympathy is also naturally responsible for our approval of virtues which promote the general good of mankind (pp. 578-9). As in the case of belief, however, there is something erratic and arbitrary in the operation of sympathy. In particular, the transference of liveliness is made easier, the closer the relation between our own self and another person. This is why we sympathise more easily with those whose manners or personality more closely resemble ours, or with those who are more closely associated with us through blood or through acquaintance (pp. 317-8). So, as in the case of causal judgments, sentiment is not all there is to moral judgments; here too experience intervenes with a balancing act. It is true that the virtues of Marcus Brutus, about whom we know only from history books, may not inspire such lively feeling of affection as those of someone, however less impressive, with whom we are personally acquainted (Hume’s example here is that of a faithful servant). And yet we would not say that our esteem for the two must, or indeed that it does, vary in exactly the same manner as the liveliness of our feeling, for ‘reflexion’ and ‘general rules’ help us to correct the vagaries of our feeling and to steady our judgment (pp. 581-2, 631). I have noted above that Hume’s moral theory, like his theory of the understanding, is informed by his naturalistic approach. This naturalism does not entail that all dictates of our moral sense are to be regarded as entirely natural. There are important cases in which our moral sense gives rise to feelings of approval or disapproval as a consequence of an 19 ‘artifice or contrivance, which arises from the circumstances and necessities of mankind’ (p. 477). This is the case with women’s chastity and modesty and with men’s courage, which, Hume maintains, it is easy to show have no natural foundation, even though this does not at all mean that we could do without them (pp. 570-3). This is also the case with justice, which arises from the combination of our natural selfishness, or ‘confin’d generosity’, with the scarcity of resources, and is articulated through the three laws of the ‘stability of possessions’, of their ‘transference by consent’, and of the obligation to respect promises. Hume shows the artificiality of justice by arguing first that the merit of a virtuous action is in the virtuous motive, not in the outcome (p. 478), and then that no such virtuous motive independent of the regard for justice itself is to be found for just actions (p. 480). The effects of time on issues of justice and property also serves to show the artificiality of their foundation: a title which is clear and certain now will be obscure and doubtful fifty years hence, even though no other circumstances have changed; and long possession of something will give the possessor a title to it. Time alone cannot give rise to anything real; it can only affect sentiments and the imagination—so, like virtue and necessary connexion, property itself is not ‘any thing real in the objects, but is the offspring of the sentiments’ (p. 509). Yet another sign of the artificiality of justice is that while every single instance of a natural virtue promotes some good and is the object of a spectator’s natural feeling of approval, it is easy to imagine single instances of justice which might not result in any particular advantage to anyone, or which might even be of positive disadvantage to someone. The great advantage of justice and artificial virtues derives from the ‘concurrence of mankind, in a general scheme or system of action’ (p. 579). But, in spite of their artificiality, justice and its rules are essential for human society, and hence for the well-being of humans. The ‘state of nature’ is, Hume says, a mere philosophical fiction: the artificial but by no means arbitrary rules of justice must have been invented by mankind at the very beginning of social life. For such rules were surely necessary in order to compensate as promptly as possible for our natural but fatally antisocial selfishness and tendency to favour our kin in the face of the scarcity of material goods and the instability of their possession (pp. 487-8). The convention of abstaining from appropriating others’ possessions established society, itself so crucial for our well20 being and subsistence; and the conclusion of this brief piece of conjectural history is that the idea of justice must have immediately followed suit, together with the notions of property, right and obligation (pp. 490-1) (see Phillipson, 1989, on Hume as a historian). The situation is, of course, simpler in the case of what Hume regards as the natural virtues. As we have seen earlier, the qualities that we naturally regard as virtues are those which are useful or agreeable to the person himself or to others. As a young man, Hume used to measure his own behaviour and character according to the strict, unforgiving standards of John Bunyan’s The Whole Duty of Man: he said so himself, years later, commenting on the demoralising effect of such practice and adding that, apart from murder and theft, he used to find himself guilty of practically all possible vices. By the time he was writing his philosophical works, it is evident that he had moved to a very different, entirely secular frame of reference: in his catalogue, which he declared he had derived from his beloved Cicero (‘one of the finest gentlemen of his age’, Hume calls him in the Enquiries, p. 128), there are such useful qualities as benevolence, discretion, caution and enterprise, industry and frugality, presence of mind, quickness of conception and facility of expression, as well as such agreeable ones as cheerfulness, a generous pride, serenity, and wit, politeness and modesty. Similarly, his examples of natural virtue come, for the most part, from his extensive readings of classical historians, and include Sallust’s characters of Caesar and Cato: the first’s amiability produces love, he writes, and we would like to find his virtues in a friend, while the second’s sterner character is ‘awful’, and such as ‘we wou’d be ambitious of in ourselves’ (Treatise, p. 607). HUMAN NATURE AND RELIGIOUS BELIEFS On 7th July 1776 James Boswell, knowing that Hume was close to death, went to see him, and was shocked at how diminished he looked. Inevitably their conversation turned to death, the immortality of the soul and religion. As Boswell reports, Hume expressed himself as serenely and cheerfully as ever on the subject: ‘when he heard a man was religious, he concluded he was a rascal, though he had known some instances of very good men being religious’ (Dialogues, p. 76). Hume’s attitude to religion is well 21 summarised in this episode—and what he said to Boswell on that occasion about the possibility of an afterlife is indeed perfectly consistent with the views he expressed in the dissertation ‘Of the immortality of the soul’. Hume had quite a reputation among his contemporaries for his irreligious attitudes: apart from the well known bookish reactions mentioned above, such as Beattie’s, there are plenty of anecdotes about this. He wrote about religious matters more than about any other topic, and devoted some of his finest philosophical writing to religious critique. We have seen something of his approach to Christianity in talking about ‘On miracles’: that essay is a good illustration of Hume’s attitude to revelation, and it is not surprising that it gave rise to varied, and usually strong, reactions. For natural religion and rational theology the situation is even more complex. In the dissertation on ‘The natural history of religion’ Hume investigates the principles and causes of religious belief. Contrary to the views common at the time, Hume maintains that ‘’tis a matter of fact uncontestable, that about 1700 years ago all mankind were idolaters’. On the basis of unanimous historical testimony we must conclude that the primary religion of mankind cannot have been monotheistic (Four Dissertations, p. 3). He identifies the origin of religious beliefs not in a rational contemplation, in the Newtonian style, of the order of nature—for the curiosity and attention of such ‘barbarous, necessitous animals’ as primitive men would be aroused by monstrosity rather than regularity (p. 7)—but in ‘the incessant hopes and fears, which actuate the human life’, in the ‘various and contrary events’ of it (pp. 13-5), and in the dread of what future may hold (p. 94). Polytheistic religions do often promote one of their gods above the others (p. 45); and men tend to praise their gods in more and more exalted terms (think of the progressive glorification of the Virgin Mary before the Reformation and of St Nicholas in Russia, pp. 47-8), till they gradually reach the notion of an all-powerful being (p. 51). The rise of an omnipotent god is not, however, irreversible; rather, there is ‘a flux and reflux in the human mind’ from idolatry to monotheism and back again (p. 54). Nor is it clear that moving from polytheism to monotheism is a progress. Hume openly suggests that, on the contrary, a polytheistic popular religion has all sorts of advantages over monotheism: it is inherently more tolerant (pp. 58-64); since it does not appeal to reason, it does not first incorporate, and then inevitably end up destroying philosophy (pp. 69-71); and the stories on which 22 traditional mythological religions are based, however groundless, do not imply blatant absurdities and contradictions, as well as being well suited to impress men by affecting their imagination (pp. 92-3). The dissertation ends with a demolition of any argument to support the naturalness of religious belief on the basis of the universal consent of mankind. The belief in a perfect being creator of the world must be a natural one. And yet the religions with which we are acquainted are full of contradictions: between the sublimity and the capriciousness at once attributed to god, between the verbal protestations and the actual conduct of men, between their zeal and their hypocrisy, between the highest hopes in talk and the most dismal terrors in fact (pp. 114-5). And while it seems true that, as the proverb goes, ignorance is the mother of devotion, we must also acknowledge that only the most brutish of peoples are entirely devoid of religion (p. 116). Thus the whole natural history of religion seems to reveal an insoluble riddle, faced with which the only sensible strategy is, Hume suggests, suspension of judgment—or it would be, were it not too difficult for the frail reason of humans to sustain the effort to escape the ‘irresistible contagion of opinion’: Hume’s final suggestion is therefore that we ‘enlarge our view, and opposing one species of superstition to another, set them a quarrelling; while we ourselves, during their fury and contention, happily make their escape into the calm, tho’ obscure, regions of philosophy’ (p. 117). Hume’s natural-historical investigation starts with the acknowledgement of the foundation of religion in reason as non-problematic—‘the whole frame of nature bespeaks an intelligent author’ (p. 1)—an acknowledgement repeated over and over again and contrasted with the ways religious beliefs arise in the primitive and in the vulgar (for example pp. 6, 10-1, 35, 42, 112, 114, 115). By contrast the in Dialogues concerning Natural Religion, where he tackles the attempted arguments a priori and a posteriori to demonstrate the existence of God, Hume concludes that religious belief does not have any foundation in reason either. The Dialogues are devoted, for the most part, to an argument from experience, the socalled design argument, very popular and endlessly discussed in the eighteenth and nineteenth centuries. On the basis of the appearance of order and finality in nature, this 23 argument establishes an analogy between nature and human artefacts; then, from our knowledge of the human mind as the cause of artefacts, it infers a similar, but proportionally more powerful and perfect mind as the cause of the world. Modelled on Cicero’s dialogue ‘Of the nature of the gods’ (Mossner, 1977; Battersby, 1979), Hume’s Dialogues are staged as a series of conversations between three characters: Demea, who supports a form of a priori argument; Cleanthes, who defends the argument from design; and Philo, who attacks both kinds of arguments until the very end, where he appears to recant. This conversation is reported years later, from memory, by Cleanthes’ ward and pupil Pamphilus, who introduces his former master by comparing his ‘accurate philosophical turn’ to Demea’s ‘rigid inflexible orthodoxy’ and to Philo’s ‘careless scepticism’ (Dialogues, p. 128). It is again Pamphilus who, in conclusion, adjudicates the dispute by maintaining that, in his view, ‘Philo’s principles are more probable than Demea’s; but that those of Cleanthes approach still nearer to the truth’ (p. 228). Philo’s ‘careless scepticism’ allows him to argue ad hominem without qualms, thus defending at different times inconsistent positions, and freely joining forces now with Demea against Cleanthes, now with Cleanthes against Demea. The resulting interaction among the three characters is lively and complex; and the attack thus orchestrated against the design argument is regarded by many as fatal. Hume proceeds by undermining the inference from the similarity of the effects to the similarity of the cause by exposing the arbitrariness and feebleness of its basis. From a house we infer an architect or a builder because we know from experience that architects and builders are the means through which houses come into being; but we cannot have any such empirical knowledge about the universe, which is a case without parallel in our experience (p. 144). Also, if we suppose a spiritual cause of the material order, are we not bound to look for a cause of that cause, thus opening up an infinite regress (pp. 1602)? Anyway, all we can know of the god as designer of the universe is that, given the similarity between the universe and human artefacts, his power, wisdom and goodness must be proportionately greater than ours, so as enable him to produce the world as we know it—but this does not mean that we have any real reason to regard the designer as infinite and perfect (p. 166, p. 203). For all we can tease out of the similarities between the world and a human artefact, say a house or a clock, when god created the world he 24 might well have been juvenile, or incompetent, or practising, or he may have been senile, and be now dead, or indeed he may have been working in a team (pp. 167-8). Moreover, the world may be regarded as more similar to an animal or to a plant than to a house or a clock, with the consequence that its cause would then turn out to be not an infinitely or immensely intelligent, powerful and benevolent architect or clockmaker, but a blind process of generation or vegetation (p. 178). Again, for all we know matter itself may well contain a principle of order (p. 146, pp. 174-5); indeed, we cannot rule out even the old Epicurean hypothesis of an appearance of art and contrivance emerging from the motion of matter (p. 183). In any case, if we try to strengthen the analogy between god and man we border on anthropomorphism and are at risk of making god finite (p. 166, p. 203), and also of making him responsible for the shortcomings and ills of the world. If, to avoid anthropomorphism and to appease the problem of evil, one follows Demea and loosens the analogy, highlighting the impenetrability of god’s intentions and actions to our feeble understanding, one does not really explain evil away, while being left with a notion of god so vague and abstract as to be almost vacuous (p. 203). In the course of the discussion Cleanthes and Philo gang up to dispose of all priori arguments (they specifically consider the necessity of a first cause, proposed by Demea, p. 189) by repeating that matters of fact and existence are simply not demonstrable by a priori argument, so that the very expression ‘necessary existence’ must be meaningless (p. 190). The only point against which Philo again and again fails to argue is the appeal of the design argument to the imagination, put forward by Cleanthes on several occasions (p. 154, p. 163), an appeal whose acknowledgement marks the beginning of Philo’s peroration (p. 202). After reviewing the main conclusions of ‘The natural history of religion’ on the pernicious effects of false religions (Dialogues, pp. 219-22) and on the origin of religious beliefs in fear of the future (pp. 224-6), Philo’s famous conclusion is that we should certainly agree that ‘the cause or causes of order in the universe probably bear some remote analogy to human intelligence’ (p. 227)—a conclusion which is, however, sufficiently feeble to be easily conceded even by an atheist. Among Hume’s philosophical writings the Dialogues is probably the most difficult to interpret and the most controversial. As Hume himself wrote in a letter to Adam Smith, while revising the text he found that it could hardly be ‘more cautiously and artfully 25 written’. What exactly did he mean by this? Philo is the dominant figure in the text. His talk occupies more space than either Demea’s or Cleanthes’, he is provocative and fun, and he has the edge on either of them, and by far, in cleverness, inventiveness and wit. Pamphilus’ final verdict is against him; but the way Hume presents him—a ward and a pupil of Cleanthes’, and reporting from memory alone conversations long past—is clearly meant to suggest that he is not absolutely reliable. Also, it is true that Cleanthes and Demea on several occasions do express typical Humean views—such as the distinction between relations of ideas and matters of fact (Cleanthes, in p. 190), and the notion that the self is a bundle of perceptions and faculties (Demea, in p. 159). But in general Demea’s approach and often his words can be traced to Samuel Clarke’s Demonstration of the being and attributes of god, and Cleanthes’ to such works as George Cheyne’s Philosophical principles of religion: natural and revealed (1715), Samuel Butler’s Analogy of Religion (1736), and Colin MacLaurin’s An Account of Sir Isaac Newton’s philosophical discoveries (1748). By contrast, not only is Philo the most ‘Humean’ of the three in approach and doctrine, but also it is difficult to trace his position (and his words) to anyone else. So it is no surprise that he has often been regarded as Hume’s mouthpiece (for example by Mossner, 1977). Yet, there is enough ambiguity in the Dialogues to make many readers wonder. If Hume was to express his point of view directly via Philo, why did he choose to write a dialogue at all, rather than an essay or a treatise? Also, what are we to make of Philo’s several apparent inconsistencies, and especially of his notorious final recantation? Perhaps Hume was trying to hide his point of view in order to make the book more acceptable to his readers? We know for a fact that Hume did try to be as discreet as possible; but many readers feel that this cannot be the whole story. So for example Livingston identifies the connexion between philosophy and conversation as one of the crucial issues in the whole of Hume’s work, and observe that all Hume’s philosophical writings—including the Treatise, in spite of its systematic appearance—are, in fact, dialectical. In this sense, in the Dialogues Hume finally found the literary genre most suited to his philosophy (Livingston, 1984, p. 19). A reading of this kind makes it plausible that Philo is, after all, Hume’s mouthpiece. For his recantation would express the extent of what he and Cleanthes can genuinely share, and provide the idea of a 26 ‘philosophical’ religion, a true religion so different from the corrupt and distorted ones described in Philo’s rant at the end of the Dialogues and in ‘The natural history of religion’, that Hume himself would not, perhaps, be reluctant to accept it (Livingston, 1998, pp. 76-9. According to others, the Dialogues are entirely informed by irony; the religion which Philo and Hume are prepared to admit is an entirely naturalistic and humanistic one, without a god, and in the end coinciding with moderate scepticism itself. In this reading, Philo is Hume, but in a way so are Cleanthes and Demea too: for a sceptic is inevitably unstable, or, if he is to be honest and true to himself, even inconsistent in his thinking (Mossner, 1977, p. 5; Battersby, 1979, p. 250-1). Yet again, this suggestion does cast some light on the issue. But for many it is a fact that the main overall impression of the Dialogues is of ambiguity; and, given Hume’s usual command of his stylistic means, some find it difficult not to think that this is a deliberate effect of a literary strategy aimed at manipulating his readers (Christensen, 1987, p. 4). If so, what effect was Hume trying to achieve? To answer this question, it has been suggested that in the case of the Dialogues there are special links between the dialogue form and the message. Before Hume, dialogues about religion staged the movement from an initial multiplicity of views to the eventual consent and order. But in the Dialogues the harmonious agreement reached by Cleanthes and Philo at the end seems at least dubious; and, more importantly, Demea leaves in a huff well before the conclusion. In other words, Hume subverts the genre of the religious dialogue, appropriating it to stage the failure of Cleanthes, Philo and Demea to compose their disagreements and reach a final consent (Prince, 1996, ch. 5). Similarly, it has also been suggested that the author’s ‘artfulness’ consists in his deliberate disappearance from his text. The uncertainty thus induced in the readers as to the design and intentions of the author of the Dialogues reproduces and reinforces the indecision communicated to them by the meandering discussions of the three characters about the design and intentions of the author of the universe. The reason of this deliberate uninterpretability is that Hume is not interested in simply producing rational and consistent, if perhaps unpersuasive, arguments: he intends to affect the readers’ imagination and to arouse their feelings, to counteract the natural appeal of the argument from design (Dancy, 1995). 27 Be this as it may, while reading the text it is difficult not to find the affinities between Philo and Hume striking. Equally striking is, on the other hand, the rough treatment Philo frequently receives at the hands of Hume—the numerous times he is made to shut up, or to look embarrassed and silly, even when he would have perfectly good ways of standing his ground. Perhaps Pamphilus’ presentation of Philo’s attitude as a ‘careless scepticism’ is a clue to his true identity. As we have seen, ‘carelessness’, ‘inattention’ and ‘scepticism’ are the virtues of the true philosopher as Hume had described him, so many years before, in his first book; and we know that the mature Hume came to find the ‘positive air’ of the Treatise and the ‘ardour of youth’ inspiring it an embarrassment and an irritation. Philo may well be, seen through the eyes of the mature author of the Dialogues, the young and ardent author of the Treatise. 28 REFERENCES Ayer, A. (1980) Hume. Oxford: Oxford University Press. Baier, A. C. (1991). A Progress of sentiments. Reflections on Hume's Treatise. Cambridge, MA and London: Harvard University Press (the best work to date on Hume and moral philosophy). Battersby, C. (1979). ‘The Dialogues as original imitation. Cicero and the nature of Hume’s skepticism’, in D. F. Norton, N. Capaldi and W. L. Robinson (eds.). McGill Hume studies (pp. 239-52). San Diego: Austin Hill Press. Christensen, J. (1987). Practicing enlightenment. Hume and the formation of a literary career. Madison and London: University of Wisconsin Press. Deleuze, G. (1991) Empiricism and subjectivity. An essay on Hume’s theory of human nature. (C. D. Boundas, Trans.) New York and Oxford: Columbia University Press (Original work published 1953). Eagleton, T. (1990). The Ideology of aesthetics. London: Blackwell. Fogelin, R. J. (1992). Philosophical Interpretations. Oxford: Oxford University Press. Garrett, D. (1997). Cognition and commitment in Hume's philosophy. Oxford: Oxford University Press. Gaskin, J. C. A. (1978). Hume’s philosophy of religion. London: Macmillan. Hume, D. (1995). Four Dissertations. Bristol: Thoemmes Press. Cited in text as Four Dissertations. Kail, P.J. E. (2002). Projection and realism in Hume. Oxford: Oxford University Press. Kemp Smith, N. (1941). The Philosophy of David Hume. London: Macmillan and Co. Kemp Smith, N.(ed.). (1947). Hume's Dialogues concerning natural religion, 2nd ed. with supplement. Edinburgh and London: Thomas Nelson and Sons (1st ed. Oxford 1935). Cited in the text as Dialogues. Livingston, D. W. (1984). Hume's philosophy of common life. Chicago and London: University of Chicago Press . Livingston, D. W. (1998). Philosophical melancholy and delirium. Hume’s pathology of philosophy. Chicago and London: University of Chicago Press. Mackie, J. L. (1980). Hume’s moral theory. London and New York: Routledge and Kegan Paul. 29 Mossner, E. C. (1977). ‘Hume and the legacy of the Dialogues’. In G. P. Morice (ed.). David Hume. Bicentenary papers (pp. 1-22). Edinburgh: Edinburgh University Press. Parusnikova, Z. (1993). ‘Against the Spirit of foundations: postmodernism and David Hume’. Hume Studies, 19, 1-18. Phillipson, N. (1989). Hume. London:Weidenfeld and Nicholson. Prichard, H. A. (1950). Knowledge and perception. Oxford: Clarendon Press. Prince, M. (1996). Philosophical dialogue in the British enlightenment. Theology, aethestics and the novel. Cambridge: Cambridge University Press. Reinach, A. (1976), ‘Kant’s interpretation of Hume’s problem’ (J. N. Mohanty, Trans.). In K. R. Merrill and R. W. Shahan (eds.). David Hume many-sided genius (pp. 161188). Norman: University of Oklahoma Press.(Original article published 1908. Zeitschrift für Philosophie und philosophische Kritik, 141, 176-209). Salmon, C. V. (1929). ‘The central problem of David Hume's philosophy. an essay toward a phenomenological interpretation of the first book of the Treatise of human nature’. Jahrbuch für Philosophie und phänomenologische Forschung, 10, 299-449. Selby-Bigge, L. A. (1975) Hume's Enquiries concering the human understanding and concerning the principles of morals. Ed. with introduction, comparative table of contents, and analytical index by L. A. Selby-Bigge, 3rd edition, with text revised and notes by P. H. Nidditch. Oxford: Clarendon Press. Cited in the text as Enquiries. Selby-Bigge, L. A. (ed.). (1978) Hume's Treatise of human nature, ed. with analytical index by L. A. Selby-Bigge, 2nd edition with text revised and variant readings by P. H. Nidditch. Oxford: Clarendon Press. Cited in the text as Treatise. Sitter, J. E. (1982). Literary loneliness in mid-eighteenth century England. Ithaca and London: Cornell University Press. Stafford Betty, L. (1971). ‘The Buddhist-Humean parallels: postmortem’. Philosophy East and West, 21, 237-53. Stroud, B. (1977). Hume. London: Routledge and Kegan Paul. Taylor, A. E. (1934). ‘David Hume and the miraculous’. In Philosophical Studies (pp. 330-65). London: Macmillan. Zabeeh, F. (1960) Hume precursor of modern empiricism. An analysis of his opinions on meaning, metaphysics, logic and mathematics. The Hague: Martin Nijhoff. 30 FURTHER READING Beauchamp, T. L. and Rosenberg, A. (1981). Hume and the problem of causation. New York and Oxford: Oxford University Press. Beck, L. W. (1978). Essays on Kant and Hume. New Haven: Yale University Press. Burns, R. M. (1981). The Great debate on miracles from Joseph Glanville to David Hume. Lewisburg: Bucknell University Press. Church, R. W. (1968). Hume's theory of the understanding. London: George Allen and Unwin Ltd. (1st ed. 1935). Dancy, J. (1995). ‘”From here the athor is annihilated”: Reflections on philosophical aspects of the use of the dialogue form in Hume’s Dialogues concerning natural religion’. In T. Smiley (ed.). Philosophical Dialogues: Plato, Hume, Wittgenstein (pp. 29-60). Oxford: Oxford University Press. Fogelin, R. J. (1985). Hume’s skepticism in the ‘Treatise of human nature’. London: Routledge and Kegan Paul. Frasca-Spada, M. (1998). Space and the self in Hume’s ‘Treatise’. Cambridge: Cambridge University Press. Hall, R. (1978). Fifthy Years of Hume scholarship. A Bibliographical guide. Edinburgh: The University Press (yearly up-dates in Hume Studies). Harrison, J. (1976). Hume’s moral epistemology. Oxford: Clarendon Press. Harrison, J. (1981). Hume’s theory of justice. Oxford: Clarendon Press. Hurlbutt, R. H. (1965). Hume, Newton and the design argument. Lincoln: University of Nebraska Press. Jessop, T. E. (1966). A Bibliography of David Hume and of Scottish philosophy from Francis Hutcheson to Lord Balfour. New York: Russell and Russell. Jones, P. (1982). Hume's sentiments. Their Ciceronian and French context. Edimburgh: Edinburgh University Press. Keynes, J. M. and Sraffa, P. (eds.). (1938) An Abstract of a treatise of human nature, 1740. A pamphlet hitherto unknown of David Hume, reprinted with an introduction. Cambridge: Cambridge University Press. Maund, C. (1937). Hume’s theory of knowledge. London: Macmillan and Co. Miller, E. F. (ed.). (1987) David Hume, Essays moral, political, and literary, revised edition. Indianapolis: The Liberty Classics. 31 Mossner, E. C. (1954). The Life of David Hume. Edinburgh: Thomas Nelson and Sons. Passmore, J. (1980). Hume’s intentions. London: Duckworth. (Original work published 1952). Penelhum, T. (2000). Themes in Hume. The self, the will, religion. Oxford: Clarendon Press. Read, R. and Richman, K. A. (eds.). (2000) The New Hume debate. London: Routledge. Richetti, J. J. (1983). Philosophical writing. Locke, Berkeley, Hume. Cambridge, MA: Harvard University Press. Yolton, J. W. (1984). Perceptual acquaintance from Descartes to Reid. Oxford: Blackwell. 32