Hume on human understanding

David Hume on human understanding Anne Jaap Jacobson David Hume’s A Treatise of Human Nature1 was published before he was 30 years old. It is often said to be the greatest philosophical work written in English. Bold and ambitious, it is designed by its author to be a significant step in the construction of a science of human nature. In his subtitle for the Treatise, Hume tells us that he will use the experimental method to develop this science. We learn elsewhere that he thinks of the experimental method in contrast to another method: we can only expect success, by following the experimental method, and deducing general maxims from a comparison of particular instances. The other scientific method, where a general abstract principle is first established, and is afterwards branched out into a variety of inferences and conclusions, may be more perfect in itself, but suits less the imperfection of human nature, and is a common source of illusion and mistake in this as well as in other subjects.2 Whether science is indeed best pursued by the experimental method or by starting with ‘clear and self-evident principles’3 was a matter being debated in Great Britain in the eighteenth century. Hume would have been aware of the debate while he was a student at the University of Edinburgh, during which time he appears to have conceived of the project issuing in the Treatise. Hume’s declared intention to use the experimental method means, in effect, he is siding with Newton, as opposed to, among others, the continental philosophers Descartes and Leibniz, and the English philosopher, Clarke. From Newton, and discussions of Newton, Hume derived not just a conception of method, but also a conception of success. Hume hopes to advance us towards an understanding of human nature that rivals in its systematicity the systematicity of Newton’s mechanics. Locke and Berkeley also had a strong influence on Hume’s philosophy. For example, Hume’s theory of ideas is explicitly an amendment of Locke’s. Hume’s exposition of his theory of ideas is somewhat swift, an indication that he saw himself as largely modifying familiar claims. And Berkeley’s philosophy shows up in several Humean theses, including Hume’s attack on abstract ideas and his account of causation in terms of constant conjunction, as we will see below.4 In addition, during his study at Edinburgh University, Hume clearly paid serious attention to other philosophers. Outstanding among these were those of the moral sense school of philosophy, particularly Hutcheson. The moral sense school of philosophy challenges the picture of human beings as at their best when functioning rationally, where rational functioning is a matter of arguing from truth to truths. In its place, Hutcheson et al. emphasize the role of the passionate side of human nature in our acquisition of some of our most important attitudes and beliefs. At the same time, there is a decidedly continental influence on Hume’s philosophy. The Treatise appeared after Hume had spent several years studying and writing in France; part of his time was spent at La Flèche, the school where Descartes was educated. Hume clearly absorbed work by Descartes and by his followers, most obviously Malebranche. Hume certainly rejects some of what he read in these sources. For example, Hume’s conception of method is at times described in explicit opposition to Descartes’.5 An additional example is Hume’s discussion of personal identity which clearly attacks Descartes’ account of the self. And Hume singles out some of Malebranche’s views for explicit rebuttal.6 None the less, Malebranche’s philosophy also has a constant conjunction view of causality among material objects which is almost certainly the immediate ancestor of Berkeley’s and Hume’s; this is a matter to which we will return.7 And Descartes has an account of natural belief, which is picked up by and elaborated by Malebranche and which will have reinforced the Hutchesonian influence on Hume.8 Despite the fact that Hume’s philosophy is significantly influenced by his cultural context, it presents us with a systematic working out of a radical idea. The radical idea, greatly extending the claims of Hutcheson, is that in the most important aspects of human life—over a vast range of phenomena—we are and must be creatures ruled by the nonrational in our nature. Our principle source in our following discussion will be Hume’s Treatise. However, the Treatise undeservedly had a very poor reception (see the concluding section below) and Hume reworked and rewrote some of its most important parts. Those directly concerned with the understanding received their final articulation in An Enquiry Concerning Human Understanding. The title ‘the first Enquiry’ is often used to distinguish it from his An Enquiry Concerning the Principles of Morals. The first Enquiry will be our second major source. THE THEORY OF IDEAS Hume’s science of human nature starts with a theory of the understanding and the theory of the understanding starts with his theory of ideas. The theory of ideas is, to use Hume’s terminology, a theory of perceptions. Perceptions include sensations, passions and emotions. The shock of cold as one falls into a icy pond is an example of a sensation. A person craving tobacco or someone angry at a rude comment will also have perceptions. Such perceptions, which Hume thinks of as particularly vivid and forceful, are impressions. There is another class of perceptions. In addition to feeling cold, we can think about being cold. We can make plans about the best way to survive in a very cold place. Or we can remember the cold we felt on some occasion before. Or we can imagine being cold in the future. In such cases, Hume maintains, we have faint images of impressions. These faint images are called ideas. Impressions come from either sensation or reflexion, the latter including passions, desires and emotions. Hume singles out two particular sources of ideas: memory and imagination. The imagination plays a very important role in Hume’s philosophy, though its introduction in the theory of ideas does not really prepare us for this. Hume starts simply by remarking that the ideas of the imagination are less lively and strong than those of the memory and that the imagination, unlike the memory, is at liberty to transpose and change its ideas. Hume’s distinction between impressions and ideas is an explicit amendment of Locke ‘s theory of ideas, which does not attempt a corresponding distinction. Hume does not tell us much at all about how to draw the distinction or decide a problem case, though he thinks that, in a few cases, we can have ideas nearly as vivid as impressions or impressions nearly as faint as ideas. None the less, he thinks the distinction is in general quite obvious and thus it is not Very necessary to employ many words in explaining [it]’.9 Having introduced impressions and ideas, Hume gives us a distinction which applies to both categories. Both impressions and ideas can be simple or complex. Hume tells us that ‘simple perceptions or impressions and ideas are such as admit of no distinction nor separation’.10 As first examples Hume gives the colour, taste and smell of an apple. Colours are prominent in his discussions of simple ideas. For example, the idea of red and the idea of a particular shade of bluefigure in his discussion in the Treatise.11 The simple-complex distinction Hume employs is not actually entirely clear. We will see this below when we discuss abstract ideas. For now we need to note that his simple-complex distinction allows Hume both to attempt to explain the creative powers of the human mind and to hold at the same time that in some sense all our ideas are derived from impressions. In the formation of our impressions, the human mind seems to be entirely passive; on the level of impressions there is no hint in Hume of the later, Kantian account of the mind making nature-as-weexperience- it.12 But Hume is well aware of at least some of the creative powers of the human mind: To form monsters, and join incongruous shapes and appearances, costs the imagination no more trouble than to conceive the most natural and familiar objects…. What never was seen, or heard of, may yet be conceived; nor is anything beyond the power of thought, except what implies an absolute contradiction.13 Hume explains our ability to form complex ideas which are not directly derived from impressions as the ability to compound, transpose, augment or diminish the materials of experience.14 He tells us that we can analyse our thoughts or ideas, however compounded or sublime, into simple ideas; each complex idea is composed of simple ideas which are and must be derived from impressions.15 Thus, simple ideas are basic ingredients for the creations of the imagination, among other things. We should want to understand, not just the major conclusions of Hume’s science of human nature, but also the way he thinks a science should argue. Does Hume follow an experimental method? Hume gives us two arguments for the thesis that ideas depend on impressions. The first argument is a model causal argument. Simple ideas are constantly conjoined with corresponding simple impressions and vice versa. Such constant conjunctions prove16 a causal dependence; and the direction of dependence is clear from our experience. One gives someone an idea of something by furnishing the opportunities for gathering an impression and not vice versa; hence, it is the impressions that are the causes. The second argument is based on ‘the plain and convincing phenomenon’ that people who lack a particular impression also lack the corresponding, resembling idea.17 These arguments reflect a use of the experimental method as Hume conceives of it; they are based on observation and experience, as opposed to supposedly undeniable first principles. Hume does qualify his thesis that all ideas are derived from impressions. The qualification concerns ‘the missing shade of blue’. Suppose that one has had experience of all the shades of blue except one. Suppose further that a chart of all the shades of blue (except for the one not experienced) going from the lightest to the darkest is placed before one. Would it not be possible to spot the lack and have an idea of what is lacking? Interestingly, Hume answers in the affirmative and remarks that it is such a strange case that it should not lead him to alter his general maxim about the dependence of ideas on impressions. Possibly Hume thinks that merely imaginary cases have no bearing on the basic laws governing human cognition.18 Whatever the correct adjudication here might be, it remains the case that Hume thinks there are some universal principles and that he can tell us what some of them are. Hume’s principles of association play a particularly crucial role. They are the source of much of the mind’s creativity and they are the source of much in our ordinary beliefs. It is Hume’s thesis that there are regular patterns to our thought and the principles of association give us his formulation of the patterns. There are three such principles: Resemblance, Contiguity and Causation. Given an impression or idea, our imagination naturally ‘runs to’ ideas resembling it. For example, one sees a horse and thinks of the horse one’s neighbour owns. Similarly, the imagination also ‘run[s] along the parts of space and time in conceiving its objects’; that is, an impression or idea of some object naturally leads us to ideas of spatially or temporally related objects.19 For example, someone sees a picture of the Pope and thinks of the Vatican, which does not at all resemble the Pope, but which is usually fairly near him. By far the most important of the principles of association is cause and effect. ‘[T]here is no relation, which produces a stronger connexion in the fancy, and makes one idea more readily recall another, than the relation of cause and effect betwixt their objects’.20 Thus, someone sees a vase being pushed off a table and then has an idea of the vase shattered into bits. Underlying the creativity of the human mind is a systematic ability we have. If one has an idea of a orange circle and an idea of a brown equilateral triangle, one therewith becomes able to have the idea of an orange equilateral triangle. It seems Hume wants to explain this ability in terms of, in this case, our combining a simple idea of orange with an idea of an equilateral triangle. But what Hume says in his discussion of abstract ideas makes the use of the notion of ‘simple idea’ in such a context problematic. We will now turn to Hume’s account of abstract ideas and then return to the question of how to understand the ‘simple’ of ‘simple ideas’. In his discussion of abstract ideas, Hume is attempting to account for the fact that we seem capable of uttering and understanding sentences which have an indefinite or even potentially infinite number of implications.21 For example, (a) ‘All human beings are mortal’ and (b) ‘X is a human being’ imply ‘X is mortal’ for every fill-in for ‘X’. Thus, ‘Socrates is a human being and Elizabeth I is a human being’, conjoined with (a), imply ‘Socrates is mortal and Elizabeth I is mortal’. And so on for any fill-in for ‘X’ in ‘X is a human being’. How do we manage to have such large thoughts? One answer is that we have abstract ideas; for example, an abstract conception of human being which applies equally to Socrates and Elizabeth I. Hume opposes such an account and instead locates generality in our language. Hume aims to explain how we can have and use such terms. We only have specific ideas and so ideas attached to each general term are specific. But, Hume notes, we are also able to notice the similarity among, for example, various red objects that are correctly called ‘red’ or various objects that are correctly called ‘globes’. In addition, when we use general terms, our imagination brings to mind lots of specific ideas of resembling objects, ideas of other red objects or ideas of other globes. The imagination here operates ‘by custom’. Finally, if we make a mistake when using a general term, we are also so constituted that we are able to correct it. Should one think, ‘No persons are snubnosed’, a counter-example will occur to one, if there are counter-examples one can be aware of. Note that this explanation of our use of general terms invokes the imagination at important points; below we will see other tasks the imagination performs. In his arguments, Hume maintains that we cannot separate out features in the way proponents of abstract ideas appear to have thought we can. We cannot literally conceive, Hume thinks, of some abstract triangle which contains or models the triangularity every triangle (equilateral, isosceles, large, small, etc.) somehow is supposed to have. As he says, ‘’tis impossible to form an idea of an object, that is possest of quantity and quality, and yet is possest of no precise degree of either’.22 Hume asserts, for example, ‘When a globe of white marble is presented, we receive only the impression of a white colour dispos’d in a certain form, nor are we able to separate and distinguish the colour from the form’.23 But this cannot be Hume’s final word; if it were, then he would be denying himself access to the systematic ability underlying creativity which he is clearly aware we do have. To refer to a previous example: If one’s idea of orange cannot be separated from an idea of circularity, then it is not a simple idea. This is so because, by definition, simple ideas are the ones that can be so separated. But then one is not going to be able to form an idea of that orange in a different shape. And now the theory leaves as highly questionable the thesis that if one has an idea of an orange circle and an idea of a brown equilateral triangle, one therewith becomes able to have the idea of an orange equilateral triangle.24 Not only do we lose our systematic ability to recombine such ideas, but, further, what Hume is saying seems to involve a staggering error, another sign that we may well not be interpreting him correctly. It is false that the colour of a particular circle one sees cannot be separated from that shape and size or that the colour of a particular globe cannot be separated from that globe. One could cut up the circle and rearrange the pieces; one could break off tile-like pieces from the globe and rearrange them. Further, Hume holds a thesis which entails, and could be used to explain, such possibilities.25 According to Hume, extended surfaces are made up of physical points which are neither divisible nor extended; they are non-extended points. Any other extension, on Hume’s account, can be regarded as decomposable into its points or atoms. Such an atomistic view makes it possible for the points to be rearranged. Further, according to Hume, our visual field is similarly constituted by such points. The thesis that our visual fields are so constituted makes the rearrangement particularly accessible to our imaginations. What this suggests is that when Hume says we cannot separate the colour from the form, he means that we cannot get the colour alone, without any shape. This is quite different from saying that we cannot get the colour without the particular shape it has at some one time. It may be a mistake to push Hume further on the problem of reconciling a fruitful account of simple ideas with his denial of abstract ideas. For one thing, the area in which he is working is enormously difficult and it is arguable that every extant theory in the area is full of problems. More immediately concerned with our project is the fact that Hume has not pushed himself to a final explanation, as we can see from the fact that his account of abstract ideas simply invokes without explanation our ability to spot similarities. We have looked at some of the central ingredients of Hume’s theory of ideas. But what is the theory a theory of, what is it supposed to be explaining? Of course, one thing the theory is telling us is what sensing and thinking are. The first is the having of impressions; the second the having of ideas. Hence, among other things, the theory is Hume’s first move in building a theory of our perception of an external world. Hearing, seeing and feeling just are having impressions, as far as what goes on inside of us is concerned. But the theory is to tell us more than this. Thus the theory will deliver an account of belief, as we will shortly see. That is, the theory will tell us the difference between, for example, imagining it is cold outside and believing that it is. In addition, the theory yields quite directly a theory of meaning.26 ‘If you tell me, that any person is in love, I easily understand your meaning, and form a just conception of his situation’, where understanding the meaning and forming a just conception consist in, it is clear from the context, having ideas.27 A major condition on meaningfulness which Hume proposes, one of his most important critical tools, follows from Hume’s theses that all complex ideas are composed of simple ideas, that all simple ideas are derived from impressions and that meanings require ideas: When we entertain, therefore, any suspicion that a philosophical term is employed without any meaning or idea (as is but too frequent), we need but enquire, from what impression is that supposed idea derived? And if it be impossible to assign any, this will serve to confirm our suspicion.28 There are strict limits on the impressions we can have and these limits transfer into conditions on what ideas we can have and so on what we can intelligibly say. There is a further aspect to the theory of ideas which needs emphasizing. From the start, Hume’s examples of impressions give a prominent place to the impressions of reflection, our passions and emotions. This feature reflects the importance in Hume’s philosophy of the non-rational aspect of human nature. In what follows we will be looking at Hume’s descriptions of the mechanisms underlying many of our most important beliefs. As we will see, the roles of imagination and custom or instinct are very central. At the same time, some of Hume’s discussion may make us wonder about whether, in Hume’s philosophy, our ordinary beliefs are really all that likely to be true. Hence, it is very important to realize that Hume does not recommend that we abandon all these beliefs. What he in general is recommending will be discussed principally in the final section below. We will next look at (i) Knowledge and Causation, (ii) The Existence of the External World and Personal Identity, and (iii) The Question of Humean Scepticism. KNOWLEDGE AND CAUSATION In the Treatise, Hume distinguishes between knowledge, proofs and probabilities.29 Hume reserves the title ‘knowledge’ for those beliefs which cannot be false; for example, the belief that two plus two is four. The thesis that knowledge is restricted to what cannot be false is largely present in Locke and quite clearly derives from Descartes. Proofs belong to a species of belief commonly regarded as particularly well grounded and exceeding our merely probable beliefs. We could see ‘knowledge’ and ‘proofs’ as technical terms which together cover a significant amount of what we would ordinarily count as knowledge. The distinction between knowledge and proofs stems from Hume’s difficult theory of relations; similar material, in a more accessible form, can be found in the Enquiry. In the Enquiry, the corresponding material is introduced with a distinction between relations of ideas and matters of fact.30 Relations of ideas, like knowledge in the Treatise, cannot be false; our discovery of them depends, Hume says, merely on the operation of thought. That 3×5=one half of 30, a relation of ideas, is something of which we can be wholly certain, however the course of nature unfolds. Matters of fact, on the other hand, can be false in the sense that they are not necessarily true. To use Hume’s example, however sure we are that the sun will rise tomorrow, it is possible that our belief is false. We can always imagine what it would be like for such a belief to be false, Hume maintains. Hence, though our evidence for it can be great, it cannot measure up to that we can have for relations of ideas. Those matter of fact beliefs for which our evidence is great include the ‘proofs’ of the Treatise. While Hume does allow that we do sometimes have proofs and that our evidence for matter of fact beliefs can be very great, he is typically seen as a sceptic about such beliefs. That is, he is typically interpreted as doubting or denying that we have any knowledge in such areas even when we use ‘knowledge’ in a much weaker sense than Hume uses it. He is so interpreted because he maintains that there cannot be any good arguments for our matter of fact beliefs. More precisely, he maintains this about our matter of fact beliefs which go beyond our beliefs about our present sensory environment, and our memories of such environments. Hume thinks that our matter of fact beliefs which are about things we do not and have not observed cannot be supported by any good arguments.31 In both the Treatise and the Enquiry, Hume’s arguments for the claim that a host of our supposedly very probable matter of fact beliefs cannot be founded on argumentation begin with a discussion of necessary connexions and what we can know about them.32 Hume’s central claim in the Enquiry is that our matter of fact beliefs about the unobserved require connexions: ‘And here it is constantly supposed that there is a connexion between the present fact and that which is inferred from it. Were there nothing to bind them together, the inference would be entirely precarious’.33 If As are merely followed by and not connected with Bs, then the occurrence of an A places no restrictions on the occurrence of alternatives to B. If smoking is not connected to lung cancer, then the mere fact that someone is a smoker does not make lung cancer any more probable than not. If, as far as what we observe goes, there are no connections, then, as far as what we observe goes, B is no more probable than any alternative to B. But, Hume adds in, we do not observe any necessary connections. Thus, as far as what we do observe goes, for each cause there are a vast number of possible results each of which is, in an important sense, as possible as any other. Suppose, then, one has observed an A followed by a B, where as far as anything one observed goes, there was a vast array of alternativesto B each of which was as likely as B. It would certainly be rash—and irrational—to expect a B before one occurs. Suppose, again, that one observes the conjunction repeating, though as far as observation goes, each time a B occurred, it had a host of equally likely alternatives. Let us add in (as Hume does) the fact that even our best reasoning by itself, and in operating on our observations, also failed to reveal any connections. Under such suppositions, can we really have strong reasons for thinking that a B will occur, given that an A has occurred? Remember, if there are no connections, then there are a host of alternatives to a B, each of which is as likely as a B. Just because of this, the prospects of one’s getting a good argument securing the conclusion that a B will occur are very dim. Hume does consider whether we might shore up an argument going from present and past observations to a conclusion about the unobserved by adding in a general belief about uniformity, namely, that the future will resemble the past (or the unobserved resembles the observed). The answer is that we do not really improve the argument. If there are no connections, the claim that the future will resemble the past is at least as doubtful as is the idea that a B will occur when an A next occurs. Hence, an argument for ‘The future will resemble the past’ has the same sort of problem that is had by an argument for the claim that a B will occur when an A next occurs. If ‘The future will resemble the past’ is needed to shore up an argument concluding that B will occur, it is every bit as much needed for any argument to show the general conclusion that the future will resemble the past. But, clearly, with the more general conclusion we have reached a position that cannot be shored up by ‘The future will resemble the past’, because that would make our argument circular. We may worry that Hume is unnecessarily restricting what he is willing to count as a good argument. Even if we concede that we cannot construct a good argument directly from ‘As and Bs have been conjoined in the past’ to ‘If an A occurs tomorrow, then a B will occur’, might not there be some sort of legitimate argument from conjunctions to connections which would solve the problem Hume has located? Hume has an argument against such an objection. The argument is an attack on the legitimacy of the notion of necessary connection. And his conclusion is that in so far as necessary connections are supposed to be more than constant conjunctions, they are in the observer, not the observed. Before we look at this attack, we need to consider Hume’s positive account of how we do acquire beliefs about the unobserved. How, then, does Hume think we manage to have any beliefs about the unobserved? Hume argues that our judgements are founded on our observations of constant conjunctions. Our belief that eating some piece of bread will nourish us depends on our having observed in the past a regular conjunction between eating bread and being nourished. When we have observed a series of conjunctions between two types of objects, we pronounce the one a cause of the other. And given an observation of the cause, we expect the effect. The details of this transition from the experience of a cause to an expectation of the effect are very important. Such transitions draw on the imagination and custom or instinct. It just is a basic fact about the way our minds operate that our observations of a series of conjunctions among As and Bs will lead us to expect a B when we perceive an A. Further, there will be a transfer of vivacity from our impression of A to our idea of B. When vivacity is transferred to the idea of B, it counts as a belief, for beliefs just are more vivacious ideas in such a relation to a present impression.34 The difference between merely raising the possibility that it is cold outside and believing it is cold outside is the fact that in the latter case, the idea has much more vivacity. ‘Thus’, Hume concludes, ‘all probable reasoning is nothing but a species of sensation’.35 It is a matter of how you feel. How, then, does Hume argue that any connection which is more than constant conjunction is in ourselves, not in the objects we pronounce connected? The argument rests on the theory of ideas. If we have an idea, we must have had the relevant impression (s).36 The only possible relevant impression sources for an idea of necessary connection are (i) our experience of constant conjunctions and (ii) our inferential reactions to such experiences. Thus, any content to our idea of necessary connection which goes beyond constant conjunctions comes from our inferential reactions. Hume concludes, ‘Either we have no idea of necessity, or necessity is nothing but that determination of the thought to pass from causes to effects and from effects to causes, according to their experienc’d union.’37 Hume accordingly gives two definitions of cause; the first gives us ‘cause’ in terms of constant conjunction, the second in terms of our inferences.38 Just as other philosophers, prominent among them Locke, have argued that colours-asseen are really in the mind and merely projected onto the world, Hume holds that necessary connections are really just in the mind. ’Tis a common observation, that the mind has a great propensity to spread itself on external objects, and to conjoin with them any internal impressions, which they occasion, and which always make their appearance at the same time that these objects discover themselves to the senses… Mean while ’tis sufficient to observe, that the same propensity is the reason, why we suppose necessity and power to lie in the objects we consider, not in our mind, that considers them; notwithstanding it is not possible for us to form the most distant idea of that quality, when it is not taken for the determination of the mind, to pass from the idea of an object to that of its usual attendant.39 But, Hume holds, it is literally unintelligible to ascribe our inferential reactions to the things regarding which we are inferring. Hence, it is literally unintelligible to ascribe necessary connections to items we are saying are causally related. But when…we make the terms of power and efficacy signify something, of which we have a clear idea, and which is incompatible with those objects, to which we apply it, obscurity and error begin then to take place, and we are led astray by a false philosophy. This is the case, when we transfer the determination of the thought to external objects, and suppose any real intelligible connexion betwixt them.40 To think of one thing as necessitating another is to think of one thing as a proof of another; that is, to think of one as a premise and the other as a conclusion.41 And this is not, strictly speaking, something of which we can make sense. From what we have just seen, one might well expect Hume to be a thorough sceptic about our beliefs about the unobserved and doubt or deny any statements which go beyond what we have observed. However, Hume’s philosophy is full of causal generalizations which go way beyond what Hume could have observed. Hume in fact gives us rules by which we are to distinguish between good and bad causal judgements. Further, he does not take any doubts about our ability to know the unobserved to compel a silence about, for example, how human minds he has not observed do in fact work. Moreover, Hume uses his discussion of causation to propose a solution, in a frankly nonsceptical way, to the problem of freewill. In discussing freewill Hume gives us the first thorough articulation of a compatibilist position; that is, a position which says we can be both causally determined and morally responsible. Hume’s arguments for compatibilism draw heavily on his two definitions of causation and Hume presents us as having a vast array of non-problematic beliefs about various constant conjunctions which go way beyond what we could have observed. Accordingly, we cannot simply assert that Hume was sceptical about all such beliefs. None the less, Hume was acutely aware of the sceptical possibilities in his treatment of causation: The sceptic…justly insists…that nothing leads us to this inference [regarding the unobserved] but custom or a certain instinct of our nature; which it is indeed difficult to resist, but which, like other instincts, may be fallacious and deceitful. While the sceptic insists upon these topics, he shows his force, or rather, indeed, his own and our weakness; and seems, for the time at least, to destroy all assurance and conviction.42 At the same time, Hume does not conclude this discussion by endorsing the scepticism. Rather, he maintains that the scepticism ought to be rejected, because no durable good can come from it. (Notice, however, that he does not attack the truth of the sceptic’s premises or the power of the sceptical arguments.) As we have found, though Hume raises very serious, sceptical questions about the content and justification of many of our ordinary causal statements, he does make some causal statements himself. We might well be tempted to see this as just inconsistent or, perhaps worse, very sloppy. But Hume is much too good a philosopher for either epithet to explain adequately a wide-spread feature of his thought. Rather, what he is dealing with is a general and profound philosophical problem. This is a topic we will return to in the final section below. We have looked principally at the relation between Hume’s account of causation and his assessment of our beliefs about the unobserved. This is the emphasis in the Enquiry. There is another aspect of Hume’s discussion, part of which is placed more in the foreground in the Treatise, which has been mentioned briefly and which we should look at in a little more detail. Many seventeenth-and eighteenth-century philosophers had, we might say, a problem about causation. Many had reasons for believing that there are three domains in which causation operates or appears to operate: (1) God causes events in the world; (2) we cause our actions and (3) events in the world cause each other. It is not easy to see how these three tasks can all be performed. For example, if God causes physical events and physical events cause each other (which already may look to be too much), there seems to be little or no causation left for us to effect. And if there is nothing left which requires our causing, how can we be responsible for any of our actions? Hume’s discussion of causation intersects with this larger problem of causation at several points. As we have seen, Hume uses his account of causation to argue that we can be both causally determined and morally responsible. And Hume attacks the seeming obviousness of some of the principles which create the problem of causation. For example, he argues that the principle that every event has a cause, which underlies many proofs for the existence of God, is not self-evidently true. We may feel, as some seventeenth- and eighteenth-century philo-sophers did, that causation by both God and worldly events is too much. Malebranche, and Berkeley in part under his influence, both did and both saw themselves as removing or downgrading causation among worldly events. What we think of as causation among worldly events is really just a matter of regularity or constant conjunction which is arranged by God, they each maintained. Genuine causation, which they thought of as embodying a kind of necessity, was reserved for another realm. It looks to be the case that Hume is in part reacting to just this picture. For Hume actively argues for the view that all there is to genuine causation among material objects is constant conjunction; what he adds is the denial that there are special necessitating connections anywhere. BODIES AND SELVES Hume tells us at the beginning of his discussion of body in the Treatise that ‘We may well ask, What causes induce us to believe in the existence of body? but ‘tis in vain to ask, Whether there be body or not? That is a point, which we must take for granted in all our reasonings’.43 And Hume says that as a consequence he will restrict his attention to the question of what causes us to believe in bodies which are external to and independent of us, and which may continue to exist when we cease to perceive them. Hume’s readers of this discussion have reason to wonder whether he is being disingenuous. For it very much looks as though Hume gives us an account which fails to reveal our belief in body as anything we would be willing to describe ourselves as in fact taking for granted. None the less, Hume is serious. And what we will come to see is that what we ‘take for granted’ may show the efficacy, in Hume’s philosophy, of the imagination, even when it is at odds with reason. There are strong similarities between Hume’s discussion of our belief in body in the Treatise and the Enquiry. In each of them, Hume considers two accounts of what the belief in body is, what its content amounts to. The first is the belief held by the vulgar (‘that is, all of us, at one time or other’).44 Their belief in continued and distinct bodies is really a belief about their perceptions. Hume tells us, ‘[T]he vulgar confound perceptions and objects, and attribute a distinct continu’d existence to the very things they feel or see’.45 He maintains, ‘The very image, which is present to the senses, is with us the real body’.46 This is a very striking attribution and we need to ask why Hume thinks we do so confound. In part the attribution is the result of Hume’s general thesis that ideas and so meanings are determined by impressions. But we can alsosee two further components behind Hume’s claim that we do so believe. In part he is representing us as direct realists; we believe that what we see—for example, a chair or a table—is immediately47 present to us. Many philosophers would agree with Hume that the vulgar are such direct realists. Additional to this is Hume’s account of how we come to believe this. It is really this last part which has the implication that our direct realism consists in beliefs about perceptions. According to Hume, the belief held by the vulgar is not the product of reasoning or argumentation; the vulgar do not reach the belief by ‘reasoning beyond’ their perceptions. As a consequence, the vulgar must be understood as having beliefs not based on their perceptions, but instead as in some sense about their perceptions. We will later look more at the details of this account. At the same time, our belief that tables and chairs are directly present to us is unstable, at least for someone constructing a science of the mind. [T]he slightest philosophy…teaches us, that nothing can ever be present to the mind but an image or perception, and that the senses are only the inlets, through which these images are conveyed, without being able to produce any immediate intercourse between the mind and the object. The table, which we see, seems to diminish, as we remove farther from it: but the real table, which exists independent of us, suffers no alteration: it was, therefore, nothing but its image, which was present to the mind.48 Thus our direct realism does not withstand the slightest scrutiny. It involves a fiction which Hume appears to count as literally false.49 Before we look more carefully at the details, we should consider Hume’s second account of what the belief in body amounts to. Because the belief of the vulgar is so clearly problematic, philosophers tend to replace direct realism with a theory about two distinct existences. The philosophers’ view is that there are internal images, directly present to us, which are caused by and resemble bodies which are external to us. This thesis of “a double existence’50 is what, on Hume’s second account, a belief in bodies involves. The philosophers’ conception of body is hardly unproblematic. The universal opinion of us all—that is, the belief of the vulgar—is the product of primary instincts. It is something we find difficult to resist. But once we see its problems, and attempt to replace it, we end up unable to support the new view. We cannot simply maintain that, like our beliefs about the unobserved, the thesis of a double existence is merely a matter of custom and instinct. For, as we have seen, custom and instinct do not lead us to the philosophers’ view; rather, they lead to the vulgar view. Neither, however, will we be able to produce a convincing argument for the philosophers’ view. For we should argue for such a view as we would for any causal thesis. That is, we should start to observe whether the relevant constant conjunctions obtain. ‘But here experience is, and must be entirely silent’.51 All we have present to us, Hume maintains, is one sort of conjunct, the images. We cannot reach around our perceptions to see if they are constantly conjoined with resembling objects. Hence, we have no good argument for the causal thesis that our perceptions are caused by resembling objects. When we take the existence of body for granted, then, we either believe something false or assent to something for which we cannot have a good argument. Further, in addition to the fact that we cannot successfully argue for the philosophers’ view, that view seems plausible only because of our original assent to the vulgar view. If we had not started with a natural tendency to think we do perceive something which is independent of us, we never would have been tempted to construct the double existence view.52 Consequently, if it is vain to ask whether there is body, if that is something we must take for granted, it is not because the belief in body is luminously true. Rather, we must take it for granted the way a ball whose movement is unimpeded must move when struck. That is, our belief in body is causally fixed. There are some aspects of Hume’s description of how vulgar belief in bodies is causally fixed which particularly merit close attention. Hume argues that the belief in an external body which continues to exist even when not observed is not the product of the senses or of reason; instead it is the product of the imagination. What sets the imagination off is the fact that our perceptions have a coherence and constancy. Our perceptions cohere in that they appear according to regular patterns; the sight of a hand knocking on a door is regularly conjoined with the sound of a knock. Further, this coherence is much greater and more uniform if we suppose the directly perceived objects to have also continued existence when not perceived. And ‘as the mind is once in the train of observing an uniformity among objects, it naturally continues, till it renders the uniformity as compleat as possible’.53 Thus we naturally ‘read’ our experience as experience with continuing objects. Even more influential is the constancy. Constancy consists in the great similarities among perceptions which leads us to judge that some object is being re-experienced. Hume notes, ‘[We find] that the perception of the sun or ocean, for instance, returns upon us after an absence or annihilation with like parts and in a like order, as at its first appearance’.54 In such cases, we tend to think of the perceptions as really the same individual from one time to the next. Hume tells us, ‘[W]e arenot apt to regard these interrupted perceptions as different, (which they really are) but on the contrary consider them as individually the same, upon account of their resemblance.’55 We have such a tendency because their similarity means the imagination passes from one to the other with great ease and in such a case, surveying the different perceptions feels very like a surveying of a single object. As a consequence, we confound a succession with an identity and attribute sameness to every succession of such related objects.56 The claim that interrupted perceptions amount, none the less, to the same thing does ask that we account for how they can be both one thing and interrupted. We respond to this by ‘feigning a continu’d being, which may fill those intervals, and preserve a perfect and entire identity to our perceptions’.57 We in effect create the fiction of a continued existence: ‘This propension to bestow an identity on our resembling perceptions, produces the fiction of a continu’d existence;…that fiction, as well as the identity, is really false, as is acknowledg’d by all philosophers’.58 As we have seen, the philosophers respond to such a view, by attempting to replace it with a thesis of double existence. But the philosophers’ view is simply the monstrous offspring of two contrary principles: the imagination’s propensity to ascribe an identity to distinct perceptions and reflection’s insistence that ‘our resembling perceptions are interrupted in their existence and different from each other’.59 The imagination’s propensity to feign an identity is strong enough that we will succumb to it. However, when we return to Hume’s critical perspective we may well want to say, with him, that I feel myself at present of a quite contrary sentiment, and am more inclin’d to repose no faith at all in my senses, or rather, imagination, than to place in it such an implicit confidence. I cannot conceive how such trivial qualities of the fancy, conducted by such false suppositions, can ever lead to any solid and rational system.60 There are several points at which one might object to what Hume is saying. For example, we might insist that he is wrong in describing our starting-point. We might insist that we start, not with our internal impressions, but with our shareable observations of a public world. Or we might question Hume’s reasons for maintaining that we have no good reason for believing the philosophers’ thesis of a double existence, with impressions on the one hand and resembling objects causing them on the other. His reason here is that we cannot verify this causal thesis the way we need to be able to; we cannot reach around our impressions to check on the existence of the resembling causes. One might object that Hume has placed his standards of proof too high. We often do, as a matter of fact, make inferences about classes of causes which we cannot observe. Such inferences are inferences about the best sort of explanation a phenomenon has. An example of such inferring was the inference to the existence of a gene in advance of our even being able to specify what its chemical composition is.61 Whatever the best philosophical position here is, it is important in understanding Hume that we realize that he has other attacks on body, on the coherence of a supposition that there is a publicly accessible world of material objects, which support his view. Hume argues that our thought regarding material objects is subject to further infections by imagination. Thus, Hume maintains, the imagination is apt to feign something unknown and invisible to make sense of our conviction that bodies are more than momentary and unstable clusters of qualities. This is ‘substance, or original and first matter’.62 Such fictions are formalized in ‘the antient philosophy’,63 but their effects are also present in ordinary thought. In addition, Hume agrees with, and adds to, Berkeley’s attack on the primary/secondary quality distinction. Hume takes learned opinion, ‘modern philosophy’, to assert that ‘colours, sounds, tastes, smells, heat and cold…[are] nothing but impressions in the mind, deriv’d from the operation of external objects’.64 On this modern view, primary qualities—extension and solidity in their different mixtures and modifications—are the only properties really possessed by external, material objects. But, Hume objects, if secondary qualities are really in the mind, so also must primary qualities be. To suppose otherwise is to suppose the possibility of a kind of separating which, in his discussion of abstract ideas, Hume has argued is impossible.65 Hume concludes: Bereave matter of all its intelligible qualities, both primary and secondary, you in a manner annihilate it, and leave only a certain unknown, inexplicable something, as the cause of our perceptions; a notion so imperfect, that no sceptic will think it worth while to contend against it.66 At several points in his discussion of bodies, some of which we have not explicitly considered, Hume appeals to the activity of the imagination. In the Treatise he contrasts the principles of the imagination which give rise to our reasonings concerning cause and effect with those principles which give us our belief in an external and independent world. The former are the more reputable; the latter are trivial.67 (This contrast appears to be rejected in the Enquiry.)68 As we will see, the less reputable principles of the imagination are also at work in our construction of our concept of ourselves. Hume wants to place his discussion of the self in the context of a discussion of bodies. But the perspective in this discussion of bodieswill be quite different from the discussion seen above. In discussing personal identity, Hume takes it for granted that there are planets, mountains, plants and animals. And he examines how we think about the identity of such things in order to extract some general principles to enlighten our discussion of the self. We do think of plants and animals (and planets and mountains) as continuing to exist through a series of losses, or increments, of parts. Suppose someone removes a thorn from a rose bush. The rose bush is still there, we may feel very inclined to say, even though it now has one less thorn. Hume disagrees. He maintains that, strictly speaking, any mass of matter is the same mass only if there is no addition or subtraction of any matter.69 Why do we think otherwise? The answer is one we have seen before: as long as the change is small, the passage of thought (the imagination) from the earlier to the later objects is so smooth and easy that we take the case to be one of identity. Hume maintains that there are other describable general reasons why we disregard change and pronounce identity. For example, we do not take gradual and insensible change to interrupt identity. Another factor: we can even combine identity and a large change if the parts function toward some common end or purpose which is not destroyed by the changes. For example, an extensively repaired ship or remodeled house may be allowed to be the same ship or house. In addition, when the parts interact in promoting the common end, we can tolerate vast changes in matter. Thus, nearly all of the matter of a large tree will be different from that in a sapling planted, let us suppose, thirty years ago. This need not impede a judgement that the older tree is the tree planted thirty years ago. Finally, if the nature of the object is to be changeable, we may disregard change. Thus we speak of the same part of a river even when the river is rapidly flowing. If Hume is right, then our ordinary talk of material objects is replete with error. None of the seemingly familiar objects in our present environment can have really existed for very long, despite our great inclination to believe otherwise. Given we make these errors, it is going to seem less odd to think we are making a similar error about ourselves. And Hume is going to maintain that we are wrong in the way we think about ourselves. Understanding what errors Hume thinks we make is made difficult by the fact that Hume appears to reject his initial discussion. In an appendix to the Treatise,70 he seems to take it back, apparently telling us his two main principles are not consistent. Thus the account Hume gives us in the body of Treatise is presented with a confidence we need to regard as provisional. Hume’s discussion in ‘Of personal identity’ begins with an attack on Descartes, among others: There are some philosophers, who imagine we are every moment intimately conscious of what we call our SELF; that we feel its existence and its continuance in existence; and are certain, beyond the evidence of a demonstration, both of its perfect identity and simplicity…. [N]or is there any thing, of which we can be certain, if we doubt of this.71 Hume maintains that this description cannot be the description of a genuine, contentful idea of the self. Such a self is supposed to exist at every moment of our lives, yet we are without any awareness of such a continuant. It is, further, something supposedly additional to our perceptions, something which has the perceptions. Not only do we lack any impression of such a continuant, all we do have are our changing impressions. Nothing in our experience gives content to an idea of, or provides evidence for, such a continuant. Having attacked the idea that Cartesian self-reflection supports a view of the self as that to which all our perceptions are referred, Hume attempts, first, to explain why we have thought the Cartesian view of the self is so plausible, and, second, to tell us what the truth is. We find the Cartesian view plausible because of a tendency we have seen Hume discuss above.72 As we have seen in the case of body, when the passage of the imagination is very smooth as it surveys a diversity of perceptions or (what is for Hume) a series of trees, we have a tendency to pronounce the case to be one of identity. Similarly in the case of the mind, the easy transition of the imagination in its survey of our perceptions leads us to think of the self in terms of identity, and not diversity. In both such cases we ‘substitute the notion of identity, instead of that of related objects’.73 Nor do we stop with the self at ‘boldly asserting] that these different related objects are in effect the same, however interrupted and variable’.74 Rather, In order to justify to ourselves this absurdity, we often feign some new and unintelligible principle, that connects the objects together, and prevents their interruption or variation. Thus we feign the continu’d existence of the perceptions of our senses, to remove the interruption; and run into the notion of a soul, and self, and substance, to disguise the variation.75 As what we have just seen suggests, Hume thinks that our tendency to pronounce identity when we really have diversity does not merely result in a verbal flourish. Rather, we take on a commitment to a further kind of being. Thus, our propensity to declare identity when there is really diversity is compounded by a tendency of ours to think there is something more which is really unchanging and identical. We do this with our creation of the fiction of the self or soul. But even when our reaction to diversity is not as extreme as inventing a further thing to be there, we are still apt to ‘imagine something unknown and mysterious, connecting the parts, beside their relation’.76 In the case of ourselves, as in the case of ordinary material objects, certain relations among the diverse elements leads us to pronounce identity. With ourselves, the diverse elements are perceptions and the relations are similarity and causation.77 It is largely memory which accounts for the similarity among perceptions; and this similarity does promote an easy passage of the mind from one cluster of perceptions to another. Memory supplies us with copies of earlier perceptions and, as copies, they will be similar. In addition, our perceptions are causally interrelated, as again memory helps us to see. Impressions cause ideas which in turn can cause other impressions and ideas. The self is, then, an evolving cluster of perceptions related by similarity and causality. In his famous misgivings in an appendix, Hume reviews his conception of the self and shows it to be anchored in two theses which he cannot renounce: ‘that all our distinct perceptions are distinct existences, and that the mind never perceives any real connexion among distinct existences’78 Hume appears to say that these principles positively block any satisfactory account of the unity of the self, or at least that he cannot find a way to unite these principles with a needed, better account. The implication is that his earlier account of the self is defective and that he cannot do any better. His reaction to the resulting philosophical problem is to declare himself a moderate sceptic, who avoids any dogmatic conclusion and instead pleads the problem is too difficult for him to understand. That Hume declares himself a moderate sceptic is significant, as we will see. A CONCLUSION REGARDING SCEPTICISM Hume’s work is informed by the radical idea that in the most important aspects of human life—over a vast range of phenomena—we are and must be creatures ruled by custom and instinct, where this contrasts to being continuously guided by rational argumentation. To describe a set of beliefs, attitudes and actions as irrational in this way is not necessarily to denigrate them. In fact, one can easily and consistently hold that some patterns in our lives are very beneficial and even efficient in getting us true beliefs without thinking of them as consisting in rational arguments. Hume’s work has, however, often been regarded as negative and destructive. While Hume tells us that the remarkable outpouring of his own youth, A Treatise of Human Nature, ‘fell dead-born from the Press, without reaching such distinction as even to excite a Murmur among the Zealots’,79 this comment looks like wishful thinking. The radical and ambitious Treatise in fact cost Hume dearly. As Mossner, his biographer, remarks: the Treatise was sufficiently alive in 1745 to lose for Hume the Professorship of Ethics and Pneumatical Philosophy at Edinburgh…. And after a quiescent period of more than a decade, the ‘Murmur among the Zealots’ began to rise in the 1750s, reaching something like a roar in the 1770s.80 The idea that Hume’s work is full of perverse errors has persisted down to our own time. The latest such view of Hume is Antony Flew’s. Flew refers to what he terms ‘the strangeness, the often self-frustrating perversity’81 of some of Hume’s positions. He takes philosophical discussion of Hume to proceed best by reversing the arguments Hume gives for ‘outrageous conclusions’82 and turning them into disproofs of the premises. Is Hume’s work in the end perversely destructive? Or is he, like Kames and Turnbull,83 other followers of Hutcheson, confident that our ideas unfold in harmony with the world they seem so obviously to represent? One way to decide this might be to see how much in our ordinary beliefs Hume wants us to discard. The extent to which Hume thinks we do have satisfactory knowledge of the world around us is currently a matter of very considerable debate among Hume scholars. While some scholars view Hume’s work as essentially sceptical,84 others see Hume as principally a constructive philosopher85 or as sharply limited in his scepticism.86 Indeed, it is only recently that the philosophical community has begun to see that Hume’s philosophy has many constructive features. In addition, the sort of naturalism Hume espouses—his picture of us as continuous with the brutes, acting and believing instinctively—seems to be increasingly confirmed by recent work in cognitive science. Hence, scholars are still in the process of re-examining the rich and subtle arguments Hume has given us. But at the same time, this ferment means that the community of Hume scholars has not reached a consensus on much at all. There is an ongoing, sometimes heated debate about how to understand Hume’s arguments. I am going to suggest a moderate, safe and easy answer to the question of whether Hume is a sceptic; then I am going to complicate things a bit. Before we approach the answer, however, we need to bring together some of what we have seen of Hume’s philosophy. We have looked at some areas regarding which we make knowledge claims (in a less restricted sense of ‘knowledge’ than we saw Hume use in theTreatise). That is, we do claim to have some knowledge about some things which we have not observed and about causal relations in which one thing makes something else occur. And we do claim to have some knowledge about our external environment. Further, we do think we can often tell quite easily that, for example, the coat we wore yesterday is the same as the one we have on today. Finally, most of us feel quite certain that our selves are substantial things which have mental states and are not just clusters of such states. Taking all these areas together, we should ask whether Hume is a sceptic regarding them. The safest and easiest answer says that Hume is simply a moderate sceptic. The moderate sceptic eschews excessive or Pyrrhonic scepticism, which, according to Hume,87 does maintain that we should suspend most or all of our beliefs in the areas we are discussing. A moderate sceptic also urges us to bear in mind our own and others’ fallibility, to undertake our investigations cautiously, with the awareness that we are not going to get the answers we want to grand questions about the ultimate truth. In favour of this moderate-sceptic interpretation is the fact that Hume does clearly reject excessive scepticism and does say that his investigations do reveal limits on our ability to know. Indeed, we might say that Hume has a healthy scepticism regarding our claims to knowledge, where this means that he stresses that our knowledge is not boundless and that we are far from infallible. Finally, we might add that Hume assigns positive merit to the beliefs we have because of instinct and custom, to our natural beliefs. Assigning such merit is not, however, a matter of seeing that the more extreme sceptic’s arguments are based on false premises or are argued illicitly. Rather, the moderate sceptic offers us a practical solution to a problem we tend to think of as a purely theoretical problem. The moderate sceptic addresses issues about how to form our beliefs without continuing to attempt to defeat the claim that the beliefs may not be true. There is a complication. One thing that Hume is also arguing is that ordinary thought proceeds on metaphysical assumptions that are philosophically indefensible. Our spreading necessity on nature, our fiction of external, resembling objects, and our fiction of a simple, continuing self are all metaphysically questionable. The slightest philosophy reveals that much in the beliefs of the vulgar cannot be justified philosophically. Whatever else one can say about the beliefs, we have no good reason to think them true. In addition, Hume’s investigations have shown that the modern philosopher has failed to give us acceptable alternatives to the beliefs of the ordinary person. Thus, taking Hume to be merely a moderate sceptic does not dispel all the tension in his work. Moreover, if we picture Hume as a moderate sceptic who is prepared to concede that truth may residewith the extreme sceptic, then the confidence with which the Treatise opens and the enthusiasm with which Hume continues his investigations in Books II and III become quite puzzling. It is hard to believe that Hume really is prepared to concede that most or all of what he says in Books II and III may be false. A genuinely adequate model for the place of scepticism in Hume’s work may have to put even more thoroughly into question the idea that a great philosopher must attempt to give us the final, true answer on the questions addressed. Such a model could pick out several different personas and perspectives in Hume’s work, among them that of the vulgar, the modern philosopher, the extreme sceptic, the moderate sceptic and the scientist of the mind, the last being the persona who continues with Books II and III of the Treatise. The scientist of the mind, we could say, turns the moderate sceptic’s practical solution of ignoring the excessive sceptic into a theoretical verdict against that sceptic’s claim to have the truth.88 Somewhat similarly, we might question whether we need or can get the final truth about how best to understand Hume. Late twentieth-century post-modernism has put in question in many areas the insistence that there be one right answer. If we keep such a view in mind as we read Hume, we may see a philosopher who is even more radical than has been thought. NOTES 1 [6.5]. Page references to the Treatise are to this edition. 2 [6.4], 174. Page references to An Enquiry concerning Human Understanding and those to An Enquiry concerning the Principles of Morals are to this edition. The second Enquiry begins with p. 167. 3 [6.4], 150. 4 I am indebted here to conversations with John Yolton. See [6.32]. 5 See [6.4], section XII, part one. 6 [6.5], 159–60. 7 See [6.22], chs V and VI. 8 See [6.31], 221–30. 9 [6.5], 1. 10 [6.5], 2. 11 [6.5], 3, 6. 12 Though as we will see with the principles of association, there is considerable activity in the mind at other junctures. 13 [6.4], 18. 14 [6.4], 19. 15 Ibid. 16 Hume’s word, [6.5], 4. 17 [6.5], 5. 18 For a useful discussion of this issue, see [6.29], 33–5. 19 [6.5], 11. 20 Ibid. 21 See [6.5], 24, ‘they can become general in their representation, and contain an infinite number of ideas under them’. 22 [6.5], 20. 23 [6.5], 25. 24 One might see Hume as attempting to resurrect the thesis through distinctions of reason as he discusses them in [6.5], 24–5. There are, however, two reasons why this suggestion is faulty. First of all, the rhetoric Hume is employing is all wrong for a discussion of the fundamental ability which underlies all our creative cogitating. Second, the passage does not aim at explaining what is in question; namely, how we manage to recombine ideas of colors and shapes. 25 [6.5], 26–39. 26 As David Pears points out in [6.25], there is some controversy about whether Hume’s theory of ideas is principally a theory of meaning or principally a theory of evidence. Like Pears, I think it is concerned with both meaning and evidence. 27 [6.4], 17. 28 [6.4], 22. 29 Hume is not completely consistent in his use of some of the terms central to this discussion, especially the terms ‘reason’ and ‘reasoning’ which are sometimes restricted to fairly formal arguments, consisting of premises and conclusions, and sometimes used to encompass a wider class of episodes of thinking. In addition, the distinction between proofs and the merely probable is not made consistently. Thus at points Hume includes proofs under ‘probable reasoning’. 30 [6.4], 25–6. This distinction is an immediate ancestor of the now infamous analytic-synthetic distinction which has been the object of much scrutiny in the second half of the twentieth century. In [6.4] the distinction between proofs and probability is relegated to a footnote; see p. 57, n. 1. 31 My ascription to Hume of this negative assessment is a traditional interpretation; see [6.29]. Several commentators have recently argued against the interpretation. See [6.10] for a discussion of this disagreement. 32 The reader should know that the interpretation given below has a serious point of disagreement with the standard interpretation. On both my and the standard interpretation, Hume locates a problem in our beliefs about the unobserved. On the standard interpretation, the problem is that such beliefs rest on an unsupportable belief that the future will resemble the past. On my interpretation, the problem is fundamentally an ontological one about necessary connections. For a very good version of the standard interpretation, see [6.29], 42– 67. I have defended my interpretation in [6.21]. 33 [6.4] 26–7. 34 [6.5], 103. 35 Ibid. 36 The appeal I am making here to Hume’s theory of ideas is controversial, though it is also the conventional reading. It has been challenged recently by, among others, [6.31] and [6.28]. However, it has been ably defended in [6.30]. 37 [6.5], 166. 38 See [6.4], 76–7, and [6.5], 169–72. 39 [6.5], 167. 40 [6.5], 168. 41 [6.4], 76. My interpretation here is not the standard interpretation. The standard interpretation maintains that the determination is a kind of feeling. I have discussed the standard interpretation and argued that it is not fully accurate in [6.20]. 42 [6.4], 159. 43 [6.5], 187. 44 [6.5], 205. ‘Vulgar’ here means just ‘ordinary folk’. 45 [6.5], 193. 46 [6.5], 205. 47 Hume’s word on [6.5], 212. In believing that what we see is directly present, we believe that we know what we are seeing without having to argue from some sort of visual clues, 48 [6.4],152. 49 [6.5], 209. See my discussion below. 50 [6.5], 182 and 205. 51 [6.4], 153. 52 [6.5], 215. 53 [6.5], 198. 54 [6.5], 199. 55 [6.5], 199. 56 [6.5], 204. 57 [6.5], 208. 58 [6.5], 209. Of course, fictions do not have to be false, but Hume says that this one is. A work of fiction might tell a story which happens to be true. What makes something a fiction is, roughly, that there is little or no connection in the creator of the fiction between the existence of the fiction and its truth, if such there be. 59 [6.5], 215. 60 [6.5], 217. 61 It is immaterial to my point whether or not we want to say that we can now observe genes (with electron microscopes, for example); rather, the point is merely that we counted ourselves as knowing such genes existed before anyone was willing to say we observed them. BIBLIOGRAPHY Works by Hume 6.1 Philosophical Works, 4 vols, ed. T.Hill Green and T.Hodge Grose, London, Longman, Green, 1875. 6.2 The Letters of David Hume, ed. J.Y.T.Grieg, New York and London, Garland Publishing, Inc., 1983. 62 [6.5], 220. 63 [6.5], sect. III of part IV of book I. 64 [6.5], 226. 65 The separating in this sort of case would amount to separating colour from any shape. 66 [6.4], 155. 67 [6.5], 197 and 217. 68 [6.4], 159. 69 Hume’s reasoning here seems at least in part to be a kind of slippery slope argument: In many or most cases, we would agree that subtracting or adding a large mass would destroy identity. Given this, one might argue, small changes must destroy identity. Otherwise, a series of small changes will add up to a large change and we will have to say both that we do have identity (each small change preserved identity) and that we do not have identity (the large change occurred and, ex hypothesis, that destroys identity). 70 [6.5], 633–6. 71 [6.5], 251. 72 Jane McIntyre and Phillip Cummins have suggested to me in conversation that a very careful reading shows that Hume does not genuinely think that the vulgar—we—do have a Cartesian conception of the self. A more focused study would be needed to decide the issue. 73 [6.5] 254. 74 Ibid. 75 Ibid. 76 [6.5], 255. 77 Contiguity, Hume maintains, drops out. 78 [6.5], 636. 79 ‘Hume’s My Own Life’, quoted in [6.23], 612. 80 [6.23], 117. 81 [6.16], 3. 82 [6.16], 103. 83 See [6.24], 152–238. 84 See, in addition to Flew, [6.19], [6.24], [6.27]. See also [6.12]. 85 See especially [6.10]. Both [6.26] and [6.29] are very important as initiating the current revision in our view of Hume. 86 See [6.14], [6.28] and [6.31]. 87 Hume’s interpretation of Pyrrhonism is questionable; see [6.18]. 88 The view of Hume adumbrated here is developed further in my forthcoming ‘A New Model of the Place of Scepticism in Hume’s Philosophy’. An early version of this paper was presented at the Hume Society Conference in 1993. 6.3 New Letters of David Hume, ed. E.C.Mossner and R.Klibansky, Oxford, Oxford University Press, 1954. While the standard editions of Hume’s works are the above Green and Grose editions, references in this chapter are to the commonly used Nidditch editions: 6.4 An Enquiry Concerning Human Understanding and An Enquiry Concerning the Principles of Morals, ed. L.A.Selby-Bigge, 3rd rev. edn by P.H. Nidditch, Oxford, Oxford University Press, 1975. 6.5 A Treatise of Human Nature, ed. L.A.Selby-Bigge, 2nd rev. edn by P.H. Nidditch, Oxford, Oxford University Press, 1978. Other Historical Texts 6.6 The Philosophical Writings of Descartes, vols I & II, trans. J.Cottingham, R. Stoothoff, D.Murdoch, Cambridge, Cambridge University Press, 1985. 6.7 Locke, John, An Essay Concerning Human Understanding, ed. P.H.Nidditch, Oxford, Oxford University Press, 1975. 6.8 Malebranche, Nicolas The Search after Truth, Elucidations of the Search after Truth, Philosophical Commentary, trans. T.M.Lennon and P.J.Olscamp, Ohio, Ohio State University Press, 1980. Twentieth-century Texts 6.9 Ardal, P.S. Passion and Value in Hume’s Treatise, Edinburgh, Edinburgh University Press, 1989. 6.10 Baier, A. A Progress of Sentiments Reflections on Hume’s Treatise, Cambridge, Mass., Harvard University Press, 1991. 6.11 Box, M.A. The Suasive Art of David Hume, Princeton, Princeton University Press, 1990. 6.12 Bricke, J. Hume’s Philosophy of Mind, Princeton, Princeton University Press, 1980. 6.13 Chappell, V.C. (ed.), Hume A Collection of Critical Essays, London, MacMillan, 1966. 6.14 Craig, E. The Mind of God and the Works of Man, Oxford, Clarendon Press, 1987. 6.15 Danford, J.W. David Hume and the Problem of Reason Recovering the Human Sciences, Newhaven, Conn., Yale University Press, 1990. 6.16 Flew, A. David Hume: Philosopher of Moral Science, Oxford, Blackwell, 1986. 6.17——Hume’s Philosophy of Belief A Study of His First Inquiry, New York, The Humanities Press, 1961. 6.18 Frede, M. ‘The Skeptic’s Beliefs’, in Essays in Ancient Philosophy, Minneapolis, University of Minnesota Press, 1987, pp. 179–200. 6.19 Fogelin, R.J. Hume’s Skepticism in the Treatise of Human Nature, London, Routledge and Kegan Paul, 1985. 6.20 Jacobson, A.J. ‘Inductive Scepticism and Experimental Reasoning in Moral Subjects’, Hume Studies (Nov. 1989): 325–38. 6.21——‘The Problem of Induction: What is Hume’s Argument?’ Pacific Philosophical Quarterly (Sept/Dec. 1987): 265–84. 6.22 Loeb, L.E. From Descartes to Hume, Ithaca, Cornell University Press, 1988. 6.23 Mossner, E.C. The Life of David Hume, 2nd edn, Oxford, Oxford University Press, 1980. 6.24 Norton, D.F. David Hume Common-Sense Moralist, Sceptical Metaphysician, Princeton, Princeton University Press, 1982. 6.25 Pears, D. Hume’s System An Examination of the First Book of his Treatise, Oxford, Oxford University Press, 1990. 6.26 Smith, N.K. The Philosophy of David Hume, London, Macmillan & Co. Ltd., 1941. Repr. New York and London, Garland Publishing, Inc., 1983. 6.27 Stove, D.C. Probability and Hume’s Inductive Scepticism, Oxford, Oxford University Press, 1973. 6.28 Strawson, G. The Secret Connexion Causation, Realism, and David Hume, Oxford, Oxford University Press, 1989. 6.29 Stroud, B. Hume, London, Routledge and Kegan Paul, 1977. 6.30 Winkler, K.P. ‘The New Hume’, The Philosophical Review (October 1991): 541–79 6.31 Wright, J.P. The Sceptical Realism of David Hume, Minneapolis, University of Minnesota Press, 1983. 6.32 Yolton, J. Perceptual Acquaintance from Descartes to Reid, Oxford, Blackwell, 1984.