The Copy Principle ‘ocke argues in Essay Liv that there are no innate ideas; all ideas are instead derived from sensation or reflection. This denial of innate ideas—highly infh- ential in eighteenth-century British philosophy—helps to justify Locke's de facto rejection of the Cartesian intellect, and thereby also to justify his methodological ‘empiricism and commonsense skepticism toward a number of scholastic and ratio- nalist metaphysical doctrines concerning both the physical and the mental realms. ‘The Copy Principle, which isthe first of Hume's two most fundamental princi- ples, is a direct successor to this Lockean doctrine, as Hume himself acknowledges (See EHU §17n). As frst presented in the Treatise, the Copy Principle states: [All our simple ideas in their first appearance are deriv from simple impressions, which ate correspondent o them, and which they exactly represent. (THIN 4) ‘As introduced in the first Enquiry, it states: [All our ideas or more feeble perceptions are copies of our impressions or more lively ones. (EHU §13) This principle, as we saw in chapter 1, involves both a Resemblance Thesis (that «every simple idea has an exactly resembling simple impression) and a Causal The- sis (that every simple idea is at least partly caused by a simple impression).! Hume uses the principle to justify one of his prime methodological directives; namely, that in order “perfectly to understand any idea,” we must [trace] it up to its origin, and {examine} that primary impression, from which it arises” (THN 74-75). Even more important, the principle plays a crucial role in his arguments concerning such central topics as space, time, causation, substance, personal identity, and ‘morality. In each case tis used to argue that a certain supposed idea—of vacuum, of time without alteration, of necessary connection in nature, ofa subject of inher- ‘ence, of a unified and identical self, or of real moral relations existing in the objects alone, respectively —does not exist. Yet it has seemed to many commentators that Hume fails to support the Copy Principle with evidence sufficient to justify the manner in which he applies it in 42 Cognition and Commitment in Hume's Philosophy these central arguments. Indeed, it has often been concluded that, despite the ‘expectations aroused by his announced methodological empiricism, he effectively ‘treats the principle asa priori, o as necessary, or both. Thus, Antony Flew writes of Hume's procedure that it amounts to making such sentences as “all our ideas... re copies of our impressions «ambiguous: most ofthe time they are taken to express a contingent generaizar tion; but at Some moments of ersis he apparently construes them as embodying a nec- essay proposition. Such manoeuvres have the effeet of making it look as if the imma nity to falsification of a necessary truth had been gloriously combined with the substantial assertiveness of a contingent generalization” However, no sooner does Hume state the principle than he himself explicitly admits an exception, or counterexample, to it, involving the idea of a “missing shade of blue”: Thete is however one contradictory phanomenon, which may prove, that "tis not will readily be allow’, that the several dstnet ideas of colours, which enter by the «xe, or those of sounds, which are convey by the hearing, are realy diferent fom cach other, tho! at the same time resembling. Now if this be tue of diferent colour, it ‘must be no les 0 ofthe diferent shades ofthe same colour, that each of them pro- duces a distinc idea, independent ofthe rst. For if his shoud be deny’, ts possible, by the continual gradation of shades, to run a colour insensibly into what is most remote from it; and if you will not allow any ofthe means to be different, you cannot without absurdity deny the extremes to be the same. Suppose therefore a person to hhave enjoyed his sight for tity years, and to have become perfectly well acquaint with colours ofall kinds, excepting one particular shade of blue, for instance, which never has been his fortune to meet with. Let all the different shades ofthat colo except that singe one, be plac before him, descending gradually from the deepes 10 the lightest; "ts plain, that he will perceive a blank, where that shade is wanting, and will be sensible, that there is a greater distance in that place betwit the contiguous colours than in any other. Now I sk whether tis posible for him, fron ination, to suppl this deficiency, and raise up to himself the idea ofthat particulat shade, tho it ad never been conveyed to him by his senses? I belive there ate few but willbe of opinion that he ean; and this may serve a a proof, thatthe simple ideas are not alvays derived ftom the corespondent impression; tho the instance isso partic lar and singular, that's saree worth our observing, and does not merit that for it «lone we should alter our general maxim. (THN 5-6 and FHU §16) But this is not all. After discussing the missing shade of blue, he seems almost immediately to go on to allow several other counterexamples without any apparent qualms, Immediately afer reiterating his endorsement of using the Copy Principle (THN 33), he asserts that our idea of time “is not deriv'd from a particular impres- sion mixd up with others, and plainly distinguishable ftom them” (THN 36). In light of the immediate context, in which he has been emphasizing the parallel between the ideas of space and time, itis evident that he would affirm a similar doctrine about the idea of space.’ And yet immediately after seeming to grant that the ideas of space and time are not copied from impressions, he concludes his dis- cussion of the ideas of space and time by denying that there is any idea either of‘The Copy Principle 8 “vacuum” (i.e,, empty space) or of “time without a changeable existence” —pre- cisely on the grounds that there are no impressions from which these latter ideas can be copied (THN 65). He then concludes Treatise Lii, a mere page later, by ‘maintaining that our “idea of existence is not deri'd from any particular impres- sion” (THN 66) All ofthis suggests to many readers that Hume applies the Copy Principle both arbitrarily and unfairly—adopting it uncritically, insisting on it when it suits his purposes, and then allowing exceptions to it whenever he wishes. In this chapter, I try to show that there is a consistent and principled approach behind these seemingly careless and inconsistent features of Hume's adoption and use of the Copy Principle. First {argue that he does not, at any point, treat the principle as either a priori or necessary, and in the process, I will outline his own argument for the principle. Second, I consider Hume's admission of the missing shade of blue as an exception to the Copy Principle and argue that it does not sig- nificantly undermine his subsequent uses of the principle. | then argue that con- trary to appearances, the ideas of time, space, and existence are not exceptions to the Copy Principle. Finally, | argue that Hume is perfectly consistent in using the Copy Principle to deny the existence of any idea of either “vacuum” or “time with- ‘out a changeable existence,” even though he admits the existence of the ideas of space and time. ‘The Status and Grounds of the Copy Principle ‘The Intended Status of the Copy Principle As his erties are well aware, there are several strong reasons to believe that Hume does not intend to treat, and does not believe he has treated, the Copy Principle as either a priori or logically necessary. One of the mast obvious of these reasons les in Hume’s explicit nomological empiricism. He repeatedly maintains that we can- not have any a priori knowledge of causal conditions (sce, for example, THN 86, EHU §23); indeed, most of his uses of the term ‘a prior’ oceur in the course of making or reiterating this very point. Yet in the course of explaining and defending the Copy Principle, he explicitly identifies it asa claim about a causal condition for the occurrence of ideas. From this characterization and his nomological empiri cism, it follows immediately that the Copy Principle cannot be a priori—so imme- diately, indeed, that Hume could hardly miss the point‘ Furthermore, his Concciv- ability Criterion of Possibility entails that the Copy Principle cannot be a necessary truth if its denial is conceivable, but he is equally insistent that the denial of any ‘causal claim is conceivable. Thus, it follows unmistakably, from claims that Hume not only accepts but emphasizes, both that the Copy Principle cannot be a priori and that it eannot be necessary. ‘A second piece of evidence that Hume does not intend the Copy Principle to be either a priori or necessary is, of course, his own admission that the idea of the “missing shade of blue” would constitute a counterexample to the Copy Principle. He could hardly fail to notice that the discovery of even one empirical counterex- ample would be incompatible with the universality implicit in treating the prinei-“4 (Cognition and Commitment in Hume's Philosophy ple asa priori or necessary. Nor is that counterexample later forgoten or rejected; on the contrary, he reproduces his entire discussion of it verbatim in the first Enquiry. ‘The Argument for the Copy Principle Outlined A third piece of evidence that Hume does not intend the Copy Principle to be a priori or necessary is the character of the argument he provides for it. That argu- ‘ment, as it occurs in the Treatise (THN 4-5), may be outlined as follows: 1. {E}very simple impression is attended with a cortespondent idea, and every simple idea with a corespondent impression. 2. Such a constant conjunction, in such an infinite numberof instances, cn never aise fom chance. 3. [There is ether a] dependence of the [simple] impressions upon the [simple] ideas, ‘or ofthe [simple] ideas upon the [simple] impressions (fom | and 2) 44 Togivea child an idea of scarlet or orange, of sweet or bitter, present the object, or in other words, convey to him these impressions; but proceed not so absurdly as to ‘endeavor to produce the impressions by exciting the ideas 5. Our ideas upon thet appearance produce not their correspondent impressions, nor do we perceive any colour, o fel any sensation merely upon thinking of them [but] any impression either of the mind or body is constantly followed by an idea, \which resembles i and is only different in the degrees of force and liveliness 6. In their fist appearance ... the simple impressions always take the precedence of their cortespondent ideas, but never appear in the contrary order. (fom 4 and 5) 7.[Wihereever by any accident the faculties, which give rise to any impressions, are ‘obstructed in their operations, as when one is born blind or deaf; not only the iimpeesions are los, but abo thee correspondent ideas. 8, [The correspondent ideas ar lst likewise] where the organs of sensation... have never been put in action to produce a particular impresion. [For example, we ean not form a jus idea ofthe taste ofa pineapple without actually tating it] 9. (All our simple ideas in thei fist appearance are deriv from simple impresions, Which ate corespondent to them, and which they exactly represent. (Rom 3, 6,7, ond 8) The frst premise, (1), is simply the Resemblance Thesis conjoined with the par allel claim that for every simple impression there oceurs a resembling simple idea > ‘We have already seen Hume's grounds for the Resemblance Thesis in chapter I: an appeal to introspective experience together with a challenge to produce a coun- terexample * As I argued in chapter 1, he is confident that no damaging counterex- ‘amples to the Resemblance Thesis can be produced because (i) he has confidence in Locke's and his own accounts of the origins of problematic concepts; (ii) he believes that philosophers who claim to employ counterexamples to the thesis derive implausible and conflicting results from them; and (ii) he is confident of the ability of his own theory of abstract ideas to explain fully the crucial phenome non of generality in thought. His argument forthe further causal claim —that sim- ple ideas are derived from the corresponding simple impressions—appeals first to (1's implication that there isa constant correlation, and hence a prima facie causal‘The Copy Principle 45 relation, between simple impressions and resembling simple ideas (2-3). It then provides two kinds of empirical evidence to establish the nature ofthe causal rela- tion, The frst kind of evidence lies in the order in which impressions and ideas appear, in those cases where both do occur (4-6), The second kind of evidence li in our inability to have ideas where the corresponding impressions have not occurred (7-8). The argument as a whole thus appears to be based entirely on ‘empirical, contingent premises. Reasons to Interpret the Principle as A Priori and Necessary Although Hume evidently does not intend to treat the Copy Pri priori or necessary, two related grounds might be offered forthe claim that he nev- ertheless does so. Firs, it may be thought that the empirical evidence he actually presents forthe premises of his argument is inadequate or nonexistent. Bennett, for example, writes boldly that “if we take the [copy] principle at face value, asa theory about the preconditions for having unlively ‘perceptions’ or quasi-sensory states,” then “we cannot bring evidence to bear upon it” (Bennett, 1971, pp. 226-227; see also Flew, 1986, pp. 22-23). In support ofthis claim, he continues: Cleatly, Hume will not bow to any fool or knave who claims to have a counterexam- ple, any congenitally blind man who says “Ihave an idea of purple” To “produce” an idea one must not merely say but show that one has it; and Hume is confident that his challengers will fail n this larger task, eg that a congenitally blind man who says “I have an idea of purple” won't be abe to give us reasons for believing him. But the blind man might well satisfy us that he isnot Ivin, and then Hume’ only resort would be to say that the bind man did nat know what “purple” means. This, 1 sugges, isthe source of his confidence: he is sure that the congenitally blind man would not be able to “produce” an idea of purple because he would not beable to sat isfy us that he knew what “purple” means. Now, what of the people whose ideas Hume counts as positive evidence fortis the 179? He hat not asked them what ideas they have, and even if he dd, why should he believe thei answers? He must say: “Well, they clearly understand the word ‘purple, {and thats good enough for me.” Ifhe doesnot say this, then itis perfectly obscure how hhe can have any postive evidence fr his theory ab applied to anyone but himself. Ifhe does sy it, then anyone counts as having an idea of purple ifhe understands “purple” ‘ora synonym oft in some other language. (Bennet, 1971, p. 227; se also MacNabb, 1951, pp. 27-29) (On this basis, Bennett argues that although Hume offers the Copy Principle asa claim about the relation between impressions and ideas, it is better understood as the thesis that experience is a precondition for “understanding,” where the latter is construed as consisting in certain linguistic abilities. Although Bennett grants that this latter thesis has some considerable empirical warrant in its own right, he also suggests that is relation to certain “analytic” (and hence presumably a priori and necessary) truths about the connection between meaningfulness and “empirical cashabilty” helps in some way to explain Hume's adoption and/or use of the Copy Principle. And although Flew granis that some empirical evidence does support the Copy Principle itself—construed as a claim about Humean mental entities—6 Cognition and Commitment in Hume's Philoeophy he also holds that Hume's confidence in itis due largely to a groping understand- ing ofa related necessary truth of the kind that Bennett describes. ‘A second reason sometimes offered for the claim that Hume treats the Copy Principle as a priori or necessary is the appearance that Hume is unwilling to take seriously the possibility that the putative ideas he rejects on the basis of the Copy Principle are instead counterexamples to it—as he ought to do ifthe prineiple were merely based on experience. Thus Flew, for example, writes (just after the remark cited at the beginning ofthis chapter: But this—not to put too fine a point on it—is outrageous Its all very wel to support such a psychological generalization {the Copy Prineiple] by citing the kind of evidence which Hume does cite, and then to challenge all comers to produce a counter-exar ple. But it simply will not do at all to turn the generalization thus supported into the supposedly sure foundation of a method of challenge: dismissing anything which might be proffered as a counterexample as being, on that ground alone, necessarily disered- ited. (Flew, 1986, pp. 20-21) ‘These two grounds, although distinet, are related. For the weaker the empirical grounds for the principle itself, the more it becomes incumbent on Hume to take putative counterexamples seriously—and the more telling his failure to do so will ‘eas evidence for the view that he treats the principles asa priori and/or necessary. Hume's Empirical Evidence Does Hume offer adequate empirical support for his premises? Certainly there would often be strategic difficulties in demonstrating to other people that one was having a simple idea that was not preceded by a corresponding impression. For ‘example, a person who, for whatever reason, never has and never will have an impression of purple obviously cannot say without selEcontradiction, “Do you see that purple object over there? The impression that you and I both have when we Took at that object corresponds to an idea for which I have not had the preceding. impression.” But this does not mean that we can never obtain any information con- ‘cerning the mental images (ideas) of the sensorily deprived or the relation of those images to their remaining impressions or sense experiences. Itis a fact, for example, that the blind and the deaf do not report mental images—that i, Humean “ideas"—that are unrelated to any simpler elements pre- viously experienced in sensation or feeling, Ifthe blind or deaf did report having such images, of course, we might not be able to determine precisely what images, they were experiencing, but would there be any reason to doubt that they were cor rect in reporting the occurrence of some such images? Just as one need not know the omithologist’ name for a bied in order to ceport that one is presently seeing a picture of some bied that one has never encountered before, so one need not know the name of a kind of image in order to report that one is presently having an image unlike any previous impression, ‘The fact that the blind and deaf can and do report aspects of their mental lives but do not report such images is surely some evi dence that they do not have them. Furthermore, sensory deprivation is not always permanent. This is obvious inthe‘The Copy Principle "7 ceases where sensory deprivation is due to simple lack of opportunity, but it may also be the case where the organic damage of the congenitally blind or deaf person is reptited or otherwise overcome. IF such persons frequently reported that they now hhad for the frst time sensations or feelings of which they had previously had only images, Hume would certainly conclude that the Copy Principle was false. That such persons do not make such reports is surely some evidence in favor of that prin- ciple. Nor are we utterly without evidence conceming the experiences of persons whose sensory exposure is normal. If they frequently reported having simple images unrelated to their sense impressions or feelings, there need not be any reason to doubt the truth of their claims. In fact, however, they generally do not make such claims, and that, once again, is some evidence for the Copy Principle ‘To treat all ofthis as evidence for the Copy Principle is, of course, to assume that the verbal and other behavior of other persons—so far as it goes—can be good evidence about the content oftheir mental images. But this assumption will be granted by everyone who is not a complete skeptic about the content of other minds. Hume himself may not have formulated an adequate specific refutation of skepticism about other minds, but in this he resembles neatly every other philosopher who claims to have information about other minds. Furthermore, in order to treat others’ behavior as evidence, there is no need to translate Hume's claims about ideas into claims about the linguistic ability to use terms (such as ‘purple’) for specific kinds of images. Indeed, even committed logical behaviorists (that is, conceptual reduetionists about the mental states of them- selves and others) need not make such a translation. They will, of course, propose that claims about the mental states of others are equivalent in meaning to claims about behavior, but they will still have their own (behaviorist) analyses of claims about mental images, and those analyses will likely remain largely distinct in ‘meaning from their analyses of claims about the understanding of particular words. Thus, there is no need even for them to interpret Hume’s Copy Principle as a disguised claim that understanding, construed as a behavioral linguistic capacity, requires experience. ‘The Copy Principle itself isthe relatively straightforward empirical claim that the presented content of those mental representations that are less “lively” than (Humean) impressions is copied from the experienced content of these impres- sions. However, Hume also accepts his predecessor’ and contemporaries’ view that tunderstanding—construed as a mental act—requires such representations and depends on their presented content. From the Copy Principle and this additional premise, it follows for Hume that understanding causally requires experience ince he would no doubt agree that, in general, complex behavioral linguistic capacities causally require understanding, he would likely agree, too, that complex behavioral linguistic capacities causally require experience. But these further claims and consequences would be distinct from the Copy Principle itself. Indeed, rather than offering a principle on which “evidence cannot be brought to beat,” Hume himself brings (in addition to the positive evidence he cites) some evidence to bear against it, in the form of the missing shade of blue. Hume does not dig in his heels and refuse to grant thatthe person claiming to have the idea of this shade can make any successful reference to it. Ifanything, Hume's standards8 (Cognition and Commitment in Hume's Philosophy for admitting contrary evidence are a bit too lax: he cheerfully accepts the ant pated results ofa merely hypothetical experiment with color samples. Itmust be granted that Hume's confidence in the Copy Principle is not based on. extensive empirical research directed toward that very question: he has not, for ‘example, sent out questionnaires. Rather, he thinks he has observed enough about human nature to know that, if counterexamples to the Copy Principle were at all ‘common—among the blind and deaf, children, or others—then he would have heard about them. Moreover, the publication of the Treatise and especially the first Enguiry (with its specific challenge to produce counterexamples) itself constitutes a sort of ex post facto questionnaire: if counterexamples are at all common, then these discussions should help to elicit reports of them. Hume thinks that both his ‘own experience and his knowledge of what other people do and do not report pro- vide extensive support for the Copy Principle. In the face ofthis support, an ocea- sional dissenter may be questioned with respect to his or her sincerity or accuracy, just as rare reported failures to produce otherwise widely replicable experimental results may be questioned. But the existence of a large number of dissenters whose dissent could not be convincingly explained as the result of natural errors would certainly call the principle itself into question. Barring universal skepticism about other minds, and given the general framework of representational categories he has adopted, the evidence Hume presents is reasonable, relevant evidence for the Copy Principle construed as an empirical generalization about the relation between two classes of perceptions Applications of the Copy Principle Even if Hume presents some genuine empirical evidence for the Copy Principe, hhowever, it remains possible that his critical use ofthe principle is more rigid than. that evidence warrants. Does he actually apply it as if it were either a priori or nec- essary? He does not imply such a status for the Copy Principle when he uses it to estab- lish his methodological directive to trace problematic ideas back to the impressions from which they are derived. He argues only that fainter than impressions, and thereby liable to obscurity and confusion” (THN 73), whereas “the examination of the impression bestows a like cleamess on the idea” (THN 75; see also THN 33). Because impressions are stronger and firmer than ideas, itis naturally useful to refer to them in problematic cases. The Copy Princi- ple serves to justify this methodological directive by asserting that there will be such ‘an impression to be found. When he makes the same point in the first Enquiry, Hume adds that we may also use the methodological directive in asessing the sta- tus of terms that we suspect have no corresponding idea: When we entertain, therefore, any suspicion that a philosophical term is employed without any meaning or idea (si is but to fequent) we need but enquire, fom ‘what impresion is that supposed idea derived? And iit be impossible to assign any, this will serve to confi our suspicion. By bringing ideas into x clear light, we may rex Sonably hope to remove all dispute which may ase, cancering their nature and ea iy. (HU §17)‘The Copy Principle 9 But again, there is no need to interpret Hume as maintaining that its either a pr- ori or necessary that every simple idea has a corresponding simple impression. He need only maintain that we have found this tobe the case, thereby raising a reason- able expectation that the search for an original impression for a problematic idea will shed light (due to the greater clarity and vivacity of impressions) on whether the idea really exists and, ifit does, on its nature. In fact, when faced with an initial inability to find an impression from which our idea of necessary connection is derived, he quite explicitly declines to treat the Copy Principle as either a prior or necessary Shall the despair of success make me assert, that lam hete possest of an idea, which is not preceded by any similar impression? This wou'd be too strong a proof of levity and inconstaney; since the contraty prineiple has been already so firmly establish’, as to admit of no farther doubt; at last, till we have more fully examin‘d the present die cally. (THN 77) Even when Hume does argue against the existence of certain ideas by appealing to the Copy Principle and to the observed absence of any corresponding impres- sion, the contexts of these instances provide no reason to suppose that the principle itself is intended to be a priori or necessary. For when he argues against the exis tence ofa certain (putative) idea, he never argues merely that we do not find such a corresponding impression in experience; he also always argues that no impression could possibly satisfy the requirements we implicitly demand for such a perception. In the case of “vacuum” and “time without a changeable existence” (as we shall see shorily), he argues that any such impression would really have to be a nonimpres- sion, since it would be an impression of nothing at all Inthe case of necessary con- nection (as we shall see again in chapter 5), he argues that a perception of a neces- sary connection in nature would “amount to a demonstration, and wou'd imply the absolute impossibility for the one object not to follow, orto be conceived not to fol- low upon the other” (THN 161-162)—a demonstration that he argues we can already see to be impossible. Similarly in the case of the metaphysicians’ concep- tion of mental or physical substance (considered in chapter 8), the perception of such a thing would be the perception of something that bestowed simplicity at a given time and identity through time on what we can already see to be intrinsically ‘complex and plural rather than simple and identical (THN 219-21; THN 252). In Book Ill of the Treatise (as we shall see in chapter 9), Hume does not argue merely that we do not actually find any distinctive moral relations (or other moral tatters of fact) so long as we confine our attention to the objects or actions them- selves. He also argues (THN 464-466) that no relation or matter of fact pertaining conly to the objects or actions themselves could be guaranteed to exert at least some motivating or approbative force on any being who perceived is existence—as the perception of vice and virtue must. In each of these cases, admitting a counterexample to the Copy Principle would mean not merely violating the Resemblance Thess but violating it in such a way as toallow nonimagistc ideas that could not, even in principle, resemble impressions. It would thus require the admission of an entitely distinct representational faculty, and hence a very serious modification in the cognitive psychology that Hume50 (Cognition and Commitment in Hume's Philosophy thinks he finds otherwise well supported by experience. That is why he thinks it ‘would be “too strong a proof of levity and inconstancy” on his part to admit coun- terexamples to the Copy Principle too hastily in such cases. Hume intends the Copy Principle to be, and consistently teats it as, a well-con- firmed empirical generalization. And indeed, the evidence for it, within the context ‘of Hume’s cognitive psychology, is reasonably strong. In view ofthe greater strength and firmness of impressions, the Copy Principle naturally gives rise to a method- ‘ological cirective—namely, to seek the impressions from which problematic puta- tive ideas are derived. It also creates a strong inductive presumption that where no prior impressions can be found, no idea will be found either. Alleged Counterexamples to the Copy Principle ‘The Missing Shade of Blue If the Copy Principle is indeed an empirical generalization, doesn’t Hume's admis- sion about the missing shade of blue undercut to a considerable extent the critical tuse he makes of the principle? He himself allows that, should anyone produce a counterexample to the Copy Principle, “it will then be incumbent on us, if we would maintain our doctrine, to produce the impression, or lively perception, which corresponds to it” (EHU §14). Sympathetic critics have sometimes suggested that Hume could easily have avoided treating the missing shade of blue as a counterexample to the Copy Principle by treating all shades of color instead as complex ideas, in one of the ways suggested by color theory. Thus, the missing shade of blue might be treated as a complex com- posed of experienced hues of varying intensities, or as a complex composed of an experienced hue, an experienced degree of saturation, and an experienced degree of brightness. This expedient is notin fact available to Hume, for teasons I explain at the end of chapter 3. Instead, Hume cheerfully grants thatthe missing shade is counterexample, and in concluding his discussions of it, he offers a reason why it joes not merit an alteration in “our general maxim’: it s because the “instance” isso ‘particular and singular” (THN 6 and EHU §16).In one sense, actually the instance is not singular. Analogous counterexamples might easly be constructed around a ‘missing tone,” a “missing degree of heat,” and so on. But in another sense, all such ‘counterexamples are “particular and singular"—that is, they have quite particular kinds of explanations. Hume has at least two quite specific explanations ready to hand cof how an idea ofthe missing shade of blue can be obtained without the eorrespond- ing simple impression. Both explanations depend essentially on his doctrine of nat- tural resemblances among simple perceptions. ‘Some resembling complex perceptions resemble each other in virtue of one containing, as a part, some simpler perception that is qualitatively identical to a part of the other. An idea of a lion and an idea of a chimera, for example, may resemble each other in virtue of involving the same kind of head. But for Hume, not all resemblance between perceptions is ofthis kind. In an Appendix note added to Treatise 1i.7, he writes:‘The Copy Principle 51 "Tis evident, that even different simple ideas may havea similarity to each other; nor is it necessary thatthe point or circumstance of resemblance shoud be distinct or sepa be from that in which they differ. Blue and green are different simple ideas, but ace ‘more resembling than blue and scale; tho’ their perfect simplicity excludes all possi- bility of separation or distinction. “Tis the same with particular sounds, and tastes and ‘smells. These admit of infinite resemblances upon the general appearance and com- parison, without having any common circumstance the same, (THIN 637) He then goes on to remark that any two simple perceptions must resemble each, ‘other at least in respect of their being simple, and yet by definition no two simple petveptions could have proper parts ofthe same kind in common Hume's view that perceptions can resemble each other without having any quali- {atively identical pars is often either ignored or regretted by etitics because it is thought to be inconsistent with his “nominalism.” But if this view is inconsistent with nominalism, then Hume is no nominalst. Far from wanting to explain or to explain away the fact that even simples can resemble one another in various respects, he regards the fact as (i) undeniable by anyone, as the case of “simplicity” itself shows, and (i) so fundamental that no further explanation is possible. in fact, if nominalism and realism are attempts to give some account of how instances of ‘even simple qualities can resemble each other, then Hume (like Locke) is neither a norinalist nor a realist Given the existence of natural resemblances among simple perceptions, Hume has available a plausible explanation for a subject’s ability to form a simple idea of the missing shade of blue in the absence of an exactly corresponding impression: the subject has instead a very large number of simple impressions that naturally resemble the missing impression very closely and are even arranged in such an order as positively to point, given the nature ofthe resemblances, to the content of the missing impression, The operation of the ming in using an attay of resembli shades to fill in the blank within an ordering of simple ideas—especially when itis an ordering of clements whose principle the mind understands (through the rele- ‘ant abstract ideas)—is arguably quite similar to the operation of the mind when it interpolates missing elements into other series that do not require the formation of a new simple idea. ‘The process is also rather similar to a phenomenon that Hume deseribes in Book Ml of the Treatise: Ideas may be compared to the extension and solidity of matter and impressions, espe cially reflective ones, to colours, tastes, smells, and other sensible qualities. Meas never admit ofa total union, but are endowed with a kind of impenetrablity by which they ‘exclude each other, and are capable of forming a eompound by their conjunction, not by their mixture. On the other hand, impressions and passions are susceptible of an entire union, and, like colours, may be blended so perfectly together, that each of them ‘may lose itself, and contribute only to vary that uniform impression which arises from the whole. Some of the mast curious phenomena ofthe human mind ate derived fiom this property ofthe passions (THIN 366) In these cases, two distinct emotions—each of which is a simple impression for Hume~are “blended” into a third emotion, potentially similar to cach. The “blending” involved is somewhat metaphorial, for the resulting emotion is itself a82 Cognition and Commitment in Hume's Philosophy simple impression, and hence is not a complex impression containing the fist two «as pars, Instead, it is a new impression that both replaces, and causally results from, two previous emotions. Let us now imagine a similar process by which ideas of a lighter and darker shade could be imaginatively “blended” so as to produce a sim- ple idea intermediate between them. This would differ from the blending of impressions only in the lesser force and vivacity of the perceptions on which it oper ated. And although Hume describes the blending process as one that applies to impressions rather than to ideas, he also characterizes it as one that applies specifi- «ally to colors. Hence, one might not be surprised to find the imagination having at least a limited capacity to “blend!” ideas of closely resembling colors.? Either—or both—of these explanations are readily available to Hume, Although the idea of the missing shade of blue need not—given very special circumstances of an ordering of similar ideas—be derived from an exactly corresponding impression, it still ultimately derived from a set of very closely resembling impressions. It thus constitutes a very near miss forthe Copy Principle. Moreover, Hume can have rea- sonable confidence that no similar explanation, in terms of closely resembling mpressions, would be available for such cases as vacuum, eventless time, necessary ‘causal connections, subjects of inherence, metaphysical selves, and real moral rela- tions. The instance of the missing shade of blue and its correlatves for other sense ‘modalities are in that way “particular and singular."!9 ‘The exception of the missing shade of blue may thus be admitted without fatally undermining either (i) the general usefulness of the methodological directive to search for the impressions from which an idea is derived, or (i) the conclusion that we should view with deep suspicion any alleged idea for which we cannot find an ‘original impression or set of impressions. The missing shade of blue itself requires at most a very slight and understandable amendment to Hume's general descrip- tion ofthe powers ofthe imagination. However, to admit an idea of vacuum, event- less time, necessary causal connection, substance, the self, or moral relations would, as we have seen, violate the Resemblance Thesis in a far more radical way, a way that would require the reintroduction of an entirely different representational faculty, a faculty of intellect that he has already rejected on several different grounds, Hume is well aware that, on his own principles, empirical investigation could always force him to admit the existence of counterexamples to the Copy Principle that would be more philosophically significant than the missing shade of blue. But the missing shade of blue, because of its special explanations and near- miss character, has little effect on Hume's philosophical uses ofthe principle Space, Time, and Existence Hume's available special explanations forthe ides shade of blue do not seem applicable to the ideas of “space,” “time,” and “existence,” however. ‘Thus, if he were to admit such ideas as these without allowing corresponding, impressions from which they are copied, he would be granting far more serious breaches of the Copy Principle. Infact, however, Hume does not say that there are 1no impressions of time, space, and existence from which the corresponding ideas are copied. Rather, he says that the idea of time (ike, presumably, that of space) is‘The Copy Principle 3 not copied from a “particular impression mix'd up with others, and plainly distin- guishable from them” (THN 36), and that the idea of existence is similarly not ‘copied from “a particular impression” (THN 66). In order to see how this distine- tion makes a difference, its essential to recall his theory of abstract ideas, set out in ‘Treatise Li.7 (and described in chapter 1). Let us apply that theory ist to the idea of pace. Hume explicitly notes that the ideas of space and time are abstract ideas (THN 34-35) and that they are composed of parts (THN 34, 39). On his view, in fact, all visual (ie., colored) and tactile perceptions are composed of minima sensibilia (minimum sensibles), and space is 2 manner in which two or more such minima sensibilia are ordered or arranged relative to one another. It follows from his gen- eral theory of abstract ideas that the abstract idea of space is some particular spa- tilly complex idea, consisting of a number of simple colored or tangible ideas, an idea that is associated with a general term and a disposition to call up a set of spa- tially complex ideas (the term’s revival set) when needed. Any complex idea that serves as the abstract idea of space will ether be copied directly from a correspond- ing complex impression of various colors or tactile qualities, or will be made up of simpler ideas that have been copied ditectly from impressions of colors or tactile qualities. Thus, although there is no separate impression of space, every spatially complex impression is an impression of space—and of various other things as well (elaborate on this use of the simple/complex distinetion, and on the way in which an impression can be an impression of more than one thing, in chapter 3.) Hume's theory thus contrasts with Locke's For according to Locke, experience presents us with a simple idea (that is, what Hume would call an impression) of space, numeri cally different from such other simple tepresentations as those of “extension,” “ig- ‘re,’ and “motion.” ‘An exactly parallel account applies also to Hume's idea of time. Time is another ‘manner in which two or more temporally minimal perceptions (which in this case need not be visual or tactile) are ordered or arranged relative to one another. One's abstract idea of time is thus a temporally complex idea, associated with a general term that revives a disposition to call up other temporally complex ideas. This tem porally complex idea is—either directly or in its simpler parts—copied from corre- sponding impressions.!! Once again, this contrasts with Locke's theory, according to which experience presents us with a unique simple idea (that is, what Hume would call an impression) of time oF “succession,” in addition to any other simple representations we might experience. Similar considerations apply also to the idea of existence, with one important dif- ference, To conceive of something and to conceive of it as existing are the same thing, according to Hume, so every impression whatever will at the same time be an impression of existence. The idea that serves as the abstract idea of existence, accordingly, may be either simple or complex. Once again, Hume's view contrasts with Locke's theory, according to which every experience presents us with a separ rate, and simple, idea (again, what Hume would call an impression) of existence, entirely distinct from all other simple representations.” ‘Thus, although there are no “separate and distinet” impressions of space, time, and existence for Hume, every idea of space, of time, and of existence is an idea4 Cognition and Commitment in Hume's Philosophy that has been copied from previous impressions. Moreover, each of these previous impressions has been an impression of (among other things) space, time, and exis- tence, respectively—that is, of spatially ordered things, temporally ordered things, and existing things. Empty Space and Changeless Time ‘We have now seen why Hume allows ideas of space, time, and existence. Why does he not also allow ideas of “vacuum” and “time without a changeable existence”? ‘The answer again depends largely on the theony’of abstract ideas, ‘The case of the vacuum differs from that of space in a crucial respect, There are ‘many impressions that are “impressions of space,” and hence many different ideas that ean serve asthe abstract idea of space. There is, however, no impression that is ‘an impression of a vacuum, and hence there is no idea that is eapable of serving as the abstract idea of a vacuum, For a “vacuum’ is supposed to be empty space, yet any spatial perception will be a complex impression whose parts ae al either visual or tangible—and visual and tangible perceptions represent thei locations a filled, not as empty. This is true whether the perception is an impression or an idea. Hence there is no idea of a vacuum and no impression from which such an idea can be derived. ‘This conclusion is much less philosophically pregnant than it might appear. I is ‘a claim about representations, made within a cognitive science of representations, and it has no negative consequences for those who deny that the universe is a plenum. Hume also writes: [If tbe askd, whether oF not the invisible and intangible distance be always full of body, oF of something that by an improvement of our organs might become visible or tangible, I must acknowledge, that I ind no very decisive arguments on ether side; tho’ Iam incln to the contrary opinion, as being mote suitable to vulgar and popular notions. Ifthe Newtonian philosophy be rightly understood, it will be found to mean no more. A vacuum is asserted: That is, bodies are said to be plac afer such a mane ner, as to receive bodies betwixt them, without impulsion or penetration. (THN 639) How can Hume say this, ifthe term ‘vacuum’ does not stand for any idea? Our mistaken belief that we have an idea ofa vacuum is the result of misunderstanding. the character ofan idea that we really do have in his view. We do have perceptions that we might call “spatially nondense” —that is, spatially complex perceptions, some of whase component parts are not contiguous but have no other perception between them. Hume is concemed only to deny that the unfilled “fictitious dis- ous perceptions is itself another perception. While there are perceptions—both impressions and ideas—of spatially nondense situa- tions, there are no perceptions of the vacuum itself, as something that occupies the unfilled locations. In other words, we represent spatially nondense situations entirely with colored and tactile representations that are themselves spatially non- dense; we do not and cannot utilize any uncolored and nontactile representations of the empty space itself Itis potentially misleading, therefore, to describe Hume as denying that we ean‘The Copy Principle 55 conceive any such thing as a vacuum. For as the passage cited shows, he is inclined to agree (and thereby implies that he can conceive) that there are real situations that would permit the interjection of additional bodies without either “impulsion or penetration” of the bodies already present just as our spatially nondense percep- tions suggest there might be. And this is precisely what the defenders of the vac- tuum have maintained. Hume's position is best expressed not as the claim that there is no such thing as a vacuum but rather as the claim that there is no such (conceivable) thing as a vacuum—that is, empty absolute space— playing a sub- stantive role in our ontology. ‘Similar considerations apply to the alleged idea of “time without a changeable existence,” or changeless time. According to Hume, we cannot represent anything to ourselves as the enduring content of emply time, because we represent time by representing a succession of changing objects. Hence, no impression can be an impression of changeless time, and no idea can serve as the abstract idea of change- less time, Our mistaken belief that we have an idea of time without a changeable existence, however, isin Hume's view the result of misunderstanding the character cof an idea that we really do have. For we do sometimes have perceptions of objects that remain unaltered while other objects change. By thinking of an unchanging ‘object fist in conjunction with one state of a changing object, and then in con- junction with a later state of the same object, we seem to perceive some change, and hence some alteration and temporal sequence, in the unchanging perception itself. In fact, however, the only perception of temporal sequence is our perception of the changing object. This produces the appearance of a “fictitious duration” applying to the unchanging object itself (THN 65), analogous to the “fctit tance” between noncontiguous spatial pars that have no object separating them, Just as there are perceptions of spatially nondense situations but no perceptions of, the vacuum as something existing between the objects, so there are perceptions of, situations in which one thing changes while another does not, but no perception of the duration ofthe unchanging object as something that the unchanging object goes through. Just as Hume grants the possibility of new matter being created between ‘wo objects without impulsion or penetration, so he grants the possibilty that an object that in fact did not change might instead have done so (THN 65). Hume's position about time is best expressed not as the claim that there cannot be time without change but rather as the claim that there is no such (conceivable) thing as cchangeless time playing a substantive role in our ontology. ‘Conceptual Empiricism and the Copy Principle Hume's Copy Principle and his other views about the cognitive role of ideas, when taken together, imply conceptual empiricism. He does not treat the Copy Principle 8 a priori or as necessary, however, but instead as an empirical generalization — just as his own methodological and nomologieal empiricism require him to do. Hence, although the Copy Principle serves to justify Hume's conceptual empiri «ism in something lke the way in which the verification principle justified the con- ceptual empiricism of the twentieth-century logical positivist, its bass and status56 Cognition and Commitment in Hume's Philosophy are entirely different. Accordingly, Hume cannot and does not treat either the Copy Principle or the methodological directive that it justifies (i.e., to seek original impressions) as sufficient to demonstrate the nonexistence of a particular idea in the absence of careful investigation of the specific characteristics of the alleged idea Such investigations show, Hume believes, that there is no idea of a vacuum (ie., ‘empty absolute space), of changeless time (ie., empty absolute time), of necessary causal connection in nature, ofa substantial subject of inherence, ofa unified and identical self, and of real moral qualities and relations existing inthe objects alone In each case, however, this isnot the end of the story but the beginning of a careful investigation of the cognitive phenomena that are mistakenly described as the hav- ing of these ideas. Searches for impressions of space, time, and existence, in con- trast, show for Hume that we have (many) impressions of each. The intial plausi- bility of the claim that we do not have such impressions results from the fact that these impressions are also of other things as well, so that there is no separate or dis- tinct impression of them. Hume's theory of abstract ideas, however, shows how we ‘ean form ideas of space, time, and existence without their impressions being sepa- rate and distinet from impressions of other things. Hume's philosophy would lose much of its distinctive character without the Copy Principle and the conceptual empiricism to which it leads. Yet conceptual ‘empiricism, as a doctrine about concepls and their content, is subject to consider- able vagueness and ambiguity. Hume's doctrine of the essential role of images in representation and meaning—the doctrine that provides the context in which the ‘Copy Principle, as a claim about images, implies conceptual empiricism—serves to impose greater determinacy on the doctrine of conceptual empiricism. But it does 0 chiefly by oversimplifying the nature of concepts and human cognition through, an identification of concepts with images (or, as in the case of abstract ideas, of ‘concepts with images used as associative exemplars). Hume’s methodological direc- tive to clarify difficult or problematic concepts by looking for particular kinds of impressions seems, in ths light, to be particularly misguided. For of all the many ways to deepen one’s insights into the nature of important concepts, the mere intensification of images through concentration on particular experiences is surely neither one of the most important nor one of the most effective. ‘Nevertheless, Hume's discussions of such crucial yet problematic philosophical concepts as “space,” “time,” “cause,” “substance,” “person,” and “virtue” played an important role in opening up both the eighteenth century and succeeding eras to new ways of thinking about these concepts, and in each of these Humean discus- sions the Copy Principle undeniably plays a central role. We will examine Hume's treatment of the latter four concepts in later chapters, and so I will not comment on them farther here. But Hume's discussion of space and time—though some- what less influential than the others—is illustrative enough for our purposes. His use of the Copy Principle calls into question both (i) the soundness of the Carte sins’ alleged demonstrations that the world must be a plenum (since attention to the source of our idea of space shows that the denial of a plenum is perfectly repre sentable and hence consistently conceivable); and (ii) the intelligibility and explanatory value of Newtonian metaphysicians’ conceptions of emply absolute‘The Copy Principle 7 space and empty absolute time as independent metaphysical entities or quasi-enti- ties. He thereby seeks to subject both Cartesian and Newtonian treatments of space and time to critical examination in the light of reflection on our cognitive eapaci- ties. Successful or not, this ia serious and substantive undertaking, ‘As this example suggests, and the case of other philosophical concepts will con- firm, the primary importance of Hume's Copy Prineiple lies neither in its putative capacity to help establish conceptual empiricism as a general doctrine nor in its potential use as a weapon to attack particular words, propositions, or philosophical problems as “meaningless” Rather, its primary importance lies in the motivation that it provided to Hume for more detailed investigations into the cognitive processes underlying the use of central yet problematic concepts, and, more specifi cally, its tendency to focus those investigations in the areas of human experience and practice on which these concepts are based. The insights we gain from such particular investigations as these, Hume quite plausibly supposes, can have impor- tant consequences for the nature of our commitments to modes of both belief and action,