Hume's Ontological Commitments Wade L. Robison The Philosophical Quarterly, Vol. 26, No. 102, Hume Bicentenary Issue. (Jan., 1976), pp. 39-47. Stable URL: http://links.jstor.org/sici?sici=0031-8094%28197601%2926%3A102%3C39%3AHOC%3E2.0.CO%3B2-D The Philosophical Quarterly is currently published by The Philosophical Quarterly.
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/philquar.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.
JSTOR is an independent not-for-profit organization dedicated to and preserving a digital archive of scholarly journals. For more information regarding JSTOR, please contact
[email protected].
http://www.jstor.org Tue Apr 24 02:40:14 2007
HURIF'S ONTOLOGICAL COMMITMENTS 'l'l~olnasReid took Hume to be a t the tail end of a chain of diminishing ontological commitments. Locke claimed that all knowledge has its sources in idcas and yet committed himself to the existence of material and imrnatcrial objects as well as God. Berkeley pointed out the inconsistency of being cmpiricist in such a way and yet holding that material objects exist. Hume is read as going the further step of deleting immaterial objects and Qod as well. To quote Reid, as the Bishop undid the n hole material world, this author upon thc same grounds, undoes the world of spirits, and leaves nothing in nature bat ideas and impressions, without any subject on which they may be impressed.l Reid's reading of Hume was a, common one in his day and has remained common. Beattie charged Hume's philosophy with being "totally subversive of science, morality, and religion both natural and revealed" because of, e.g., his claim that "the soul of man has nothing permanent in its nature, nor indeed any kind of existence distinct from its present perceptions, which are continually changing, and will soon be a t an endn.2 I n summing up Hume's philosophy in his 1885 lectures on Scottish philosophy, Andrew Seth said that "the real background of the idea of perceptions, whether material as with Locke, or theological as with Berkeley, is simply wiped out by Hume from his t h e ~ r y " . At ~ the 1973 Oberlin conference on Hume, to mention just one contemporary instance, both Terence Penelhum and his commentator, Vere Chappell, held that Hume denies the existence of such a meta-physical absurdity as the immaterial soul.* The view these diverse philosophers propound is that (a) Hume holds that perceptions exist, and (b) Hume holds that only perceptions exist. (a) is true, and (b) is false. It is false both because Hume commits himself to the existence of another kind of entity and because he does not commit himself to the non-existence of such entities as material and immaterial substances. The importance of establishing the falsity of (b) is that, among lThomas Reid, A n Inquiry into the H u m a n M i n d , ed. Timothy Duggan (Chicago, 1970),p. 14. %JamesBeattie, A n Essay on the Nature and Immutability of Truth, in Opposition to Sophistry and Scepticism (Edinburgh, 177I ) , pp. 538-9. 8Andrew Seth, Scottish Philosophy (Edinburgh, 1885),p. 48; see also p. 50. I t is unclear from Seth's introduction whether the lectures were held in 1883 or in 1884 or 1885. 4Their claims were occasioned b y my asking them whether their views on Hume's theory of personal identity depended upon construing him as denying the existence of an immaterial self, and the complacency of their response is the occasion of this paper. The thesis I arguc for seems obvious to me, but it is clearly not obvious to others.
40
WADE I,. ROBISOX
other things, Hume's scepticism ill other^+ ise bc talien as ontological rather than epistemological, his attack on metaphysicians will be misunderstood and weakened, and he ~villbe committed to answers to a host of problems about meaning and existence that he did not raise, let alone answer. I shall examine the textual evidence in section I, attack the only argument for (b) I can discover in 11, and lay out in I11 some of the considerations which lead me to deny (b).
I. THE TEXTVAL EVIDENCE As is usual in a case involving a serious commitmel~ton Hume's part, the text is ambiguous. Take the case of personal identity. Hume is explicit in the section "Of personal identity" in the Treatise: and were all my perceptions remov'd by death, and cou'd 1 neither think, nor feel, nor see, nor love, nor hate after the dissolution of my body, I shou'd be entirely annihilated (T., p. 252). . . . 1 may venture to affirm of the rest of mankind [excepting somc metaphysicians], that they are nothing but a bundle or collection of different perceptions . . . (ibid., my italics). They are tho successive perceptions only, that constitute the mind (T., p. 253).5 The same view is repeated in the Abstract: . . . it must be our several particular perceptions, that compose the mind. I say, colnpose the mind, not belong to it. The mind is not a substance, in which the perceptions inhere (A., p. 25). All these passages seem to imply that Hume thinks the mind is nothing but However, both the arguments he uses and other passages imply that this is not what he thinks. He gives two arguments in "Of personal identity" regarding an immaterial self. They purpoEt to prove that there is and call be no idea of the self as something that continues the same throughout the course of our lives. The first depends upon the contingent fact that there is no impression which so continues from which such an idea could be derived. The second claims it contradictory to suppose there could be an impression -of the self since the self is what has impressions and is not itself an impression. What is contradictory, it must be noted, is not that there be such a self, but that there be an impression of such a self.' Neither of these arguments 6References to Hume's works use " T " for the Treatise of H u m a n Nature (SelbyBigge edition); " E n for the Enquiry (Selby-Bigge edition); " A " for A n Abstract of A Treatise of H u m a n Nature, ed. Keynes and Sraffa, reprinted by Archon Books, 1965; and "L" for A Letter from a Gentleman to his friend in Edinburgh, ed. E . C. Rlossner and John V. Price (Edinburgh, 1967). 6The Grst is consistent with a body and a set of perceptions, but that interpretation is ruled out by the others. 'These arguments are examined in greater detail in Part I of my paper "Hume on Personal Identity", forthcoming in T h e Journal of the History of Philosophy. It is tho second argument which is the more interesting, and its crucial premise is some version of the claim that in order to have a n impression of X, one must have an X-like impression. It is that premise which makes it impossible to have an impression of the self since ex hypothesi the self is not a n impression.
support the coriclusion that there are nothing but perceptions. They support the far xveal;er conclusion that there is nothing- known about the mind but its perceptions. This same conclusion is supported by the only relevant mgument Hume gives in the section "Of the immateriality of the soul": We have no perfect idea of any thing but of a perception. A substance is entirely different from a perception. We have, therefore, no idea of a substance. . . . What possibility then of answering that question, whether perceptions inhere i n a material or immaterial substance, when we do not so much as understand the meaning of the question? (T., p. 234). The claim is not that there is no immaterial substance, only that we have no idea of any such substance. There is thus a disparity between some claims Hume seems to make and the conclusion which his arguments support and which he explicitly draws. This disparity is blatantly obvious in the Abstract. For after giving the passage quoted above, which states without ambiguity that the mind is not a substance, he concludes by saying, We know nothing but particular qualities and perceptions. As our idea of any body, a peach, for instance, is only that of a particular taste, colour, figure, size, consistency, etc.: So our idea of any mind is only that of particular perceptions, without the notion of any thing we call substance, either simple or compound ( A . ,p. 25, my italics). Such a disparity of claims is repeated in rcgard to material objects, and i t obviously needs settling. One mode of settlement is by appeal to the historiographical principle that where there is ambiguity in a text, one should commit an author to the least one can.8 This principle has its clearest application in such a case as this, for where there is ambiguity, one ought to determine what the arguments used prove, taking all the rest as obiter dicta. The consequence ~ o u l d be that Hume is not committed to the non-existence of immaterial and material objects, his claims to that effect being discounted on grounds of youthful exuberance, inflated rhetoric, his inability to see the point of his own arguments, or what have you. The difficulty with this move, however, is that it cuts short pertinent historical questions: did Hume really think his arguments proved more than they do? Was he cleverly attempting to convince his readers of a point he thought true but was unable to find arguments for? In short, the appeal to such a principle cuts short the crucial question "TVhat did Hume really believe?". Besides, an answer to that seems available. One way of settling ambiguous texts is to get a decisive one, and that is what Hume means to give us in A Letter from a Gentleman to his friend i n Edinburgh. He says that The Author [himself in the Treatise] has not anywhere that I remember denied the Immateriality of the Soul in the common Sense of the $This needs to be filled in with such additions as "consistent with what he elsewhere says", and so on. I shall not attempt such additions here. The principle is clear enough as it st,ands for the small use I make of it.
Word. He only says, That the Question did not admit of any distinct Meaning; because we had no distinct Idea of Substance. This Opinion may be found everywhere in Mr. Lock, as xvcll as in Bishop Be~kley (L.,pp. 29-30). Normally such a remark ~ o u l dbe decisive. I t is later than the original. being penned in 1745. It is clearly directed t o stating what Hume said in the Treatise, not what he ought to have said or even what he meant to say, and i t is consistent with the text and not an aberrant interpretation on Hume's part.$ Unfortunately, this is the only statement Hume gives us about what he meant and his motives in making it are suspect. Ile made it in a response to a charge of "Heresy, Deism, Scepticism, Atheism, etc.", while a candidate for the chair of Ethics and Pneumatical Philosophy a t Edinburgh University.lo One of the duties of the holder of the chair was to reconcile Moral Philosophy with Divinity, and, specifically, to lecture every Monday "upon the Truth of the Christian religion" (L., p. x). Hutcheson opposed Hume's candidacy given such duties (L., p. x), and one must wonder about Hume's disingenuousness in applying. A plausible explanation of Hume's denial in his Letter is that tempted by such a chair, even the most honest soul (forgive me, Hume!) might seize upon ambiguities in his test to make the best case for his orthodoxy. I n sum, although the preponderance of evidence is that Hume did not commit himself to the non-existence of an immaterial self, the text is not decisive: no quotation of what Hume says about tho self will settle the matter
11. BUTCIIVAROV'S ARGUXENT
I call discover only one argument for (b) based on a more enduring featurc of Hume's philosophy than his rhetoric. I t purports to prove that only perceptions can exist, but it needs to be extricated from the ontological commitments of its originator, Panayot Butchvarov. He claims that although Hume's perceptions are extensionally equivalent to Locke's and Berkeley's ideas, there is a fundamental difference between Hume and his empiricist predecessors. They construed perceptions epistemologically, as objects of knowing, whereas the Humean perceptions cannot possibly be defined as an object of linoxving or awareness, unless we completely disregard his explicit rejection of a self that can make such knowing or awareness possible.ll Butchvarov concludes that Hume's "approach is ontological, not epistemological" and proceeds to give "an ontological argument" for the claim that "the class of existents and the class of perceptions must be identical".12 QHume in fact claims that those who read him otherwise "pervert" his meaning
(L., p. xviii). That seems, to put it mildly, a little strong. 'OThe Letters of David Hunze, ed. J. Y. T. Greig, Vol. I (Oxford, 1969), p. 57: letter 24, t o William Mure of Caldwell. 1l"The Self and Perceptions: A Study in Humoan Philosophy", The Philosophical Quarterly, 9 (1959), p. 104. This article has gained wide circulation by being reprinted in the Bobbs-Merrill reprint series in philosophy. 1 8 0 ~ cit., . pp. 101-5.
IIGME'S
ONTOLOGICAL COATiITITMENTS
13
One may havc misgivings about the general framework Butchvarov cr-ects around this argument. The reason he cites for rejecting the claim that perceptions are epistemically defined is that Hume "rejects the self". But his argument for the latter depends upon the claim that there are nothing but perceptions: each claim is uscd to hoist the other.13 Fortunately the "ontological argument" he in fact gives does not depend upon the general framenorli: it is something less than purely ontological. I n Treatise I, ii, G Hume considers how we come to have the idea of existence. If it is derived from some particular impression, we should havc to imagine it conjoined with cvcry perception "since we never renleinber any idea or impression without attributing existence to it" (T., p. 66). But there is no impression 11 hich is "inseparably conjoin'd" with every perception. The consequence is that "the idea of existence, . . . is the very same with the idea of nhat n e conceive to be existent" (ibid.). Butchvarov rightly concludes that Hume thinks not only that whatever we conceive me conceive as existent, but also that mhatever we conceive as existent me conceive. But he also concludes that since "it is meaningless to ever talk about the existence of non-perceptions" . . . "the class of existents and the class of perceptions must he identical".14 The conclusion depends upon a hidden premise: what n-e cannot meaningfully say exists cannot exist. This is a principle based upon a theory of meaning, and if Hume accepts Butchvarov's conclusion, he must accept that principle. Without it the conclusion does not follow: it is one thing to say that whatever we conceive we conceive to be existent and quite another to say that mhatever is or can be existent we conceive. Is Hume committed to Butchvarov's hidden premise? Since he holds that words have meaning only insofar as they are associated with ideas, he would pose the question differently: does it follow from our not being able to have an idea of X that X cannot exist? The anstx-er is typically Humean: it depends. If the reason is that the purported idea contains inconsistent elements, the answer is yes. But there is another reason for our not being able to have an idea. For Hume the only way me humans can come to have ideas is through impressions. An immaterial self is not an impression, but has them. It is thus distinct from any impression. Yet Hume thinks we cannot have impressions of what are not impressions: to have an impression of X it is necessary to have an X-like impression. A self-like impression of an immaterial self, however, is a contradiction: it ~vouldrequire an impression which is distinct from any impression. We thus cannot have an idea of an immaterial self. What is contradictory is not that there might be an imlSThe argument, Butehvarov says, "is not epistemological, but is directly based on the ontology of the first two Parts of the Treatise. The argument is simply that since the self is not and cannot be a perception, but is supposed to be that which has the the concept of existence is, perceptions, therefore there cannot be such a self. For, or is postulated as, equivalent to the concept of perception" (op. cit., p. 111). 140p.cit., pp. 113 fn. 17 and 105 respectively.
...
material self, but that \re might have an irrrpre.vsio~/of it. Does Hume tllink it follows from that contradiction that therc cannot be an immaterial self? To suppose he does would radically alter our conception of his philosophy "The contrary of every matter of fact is still possil~le",Hume says, "becausth it can never imply a coiltradiction . . ." (E., p. 2 5 ) . To claim of Hurnc. that he claims an immaterial self cannot exist is either to deny that it is a matter of fact whether it does or to deny the distinction betw-een ~natters of fact and relations of ideas. Either option is implausible. For Hume all matters of existence are matters of fact. To claim that X exists is either to assert that X is a present perception or memory or to assert that X is causally related to a present perception or memory. But Hume claims both assertions are matters of fact (E., p. 26; T., p. 74), and it is difficult to imagine the resultant system being Humean were that claim denied.l5 The imagination is stretched even more if one denies the distinction bet- eon relations of ideas and matters of fact, and I hardly need to spell out here the consequences oi sucll a denial for our understanding of Hume's system. Butchvarov's argument thus fails; to attribute the necessarx premise to Hume is to alter in too radical a way the structure of his pllilosophical system. To put thr: point in a way directly contrary to Butchvarov's ontological prcdilcctions, thc empiricist principle is an epistemological. not ail ontological principle. I t docs not follon from our failing to 11a-c-ctail idca oi S that S cannot cxist.lG
Ill. H ~ J I E ' S1-IEV-s The cvitlcnce and arguments for denying (b) arc di\.cr.sc. 1,c.t in(. givc four different sorts. I do not incan tlle list to 1x1 complc~tc,or the i~lstancck of equal importance. (1) Hume seems comnlitted to the existence of entities other than percc.1) tions. I n fact, it is one of thc major features of his programme that o t h c ~ Binds of' entities exist, for be thinks he will be the Xcwton of' the mental n orld 1)ecausc of his articulation of principles of association ~ ~ h i c1)irtd ll together some of our perceptions. These principles are not perceptions. Their analysis is a difficult matter, but t o put it briefly, using Hun~c'son11 ~vord,they arc propensities. He thus calls the dctermi~lationof the mind t o conceive or believe the existence of an effect upon perceiving a cause "that propensity, which custom produces, to pass from an object to thc idea of its usual attendant" (T., p. 165). Unfortunately, this evidence of Hurne's denial of (b) is not as decisive as it ought to be. For though Hume's philosopllical system and his Kern 16For example, Hume's attack in Section I X of tho Dialogues Concerni~tgA'atural Religion on the a priori argument for the existence of God would have to he radically u~hoso altered, for it depends crucially upon the premise that "there is no Being, . non-existence implies a contradiction": ed. Nelson Pike (Indianapolis, 1!)70), p. 77: ed. Kemp Smith (Oxford, 1935), p. 233. lBItalso does not f ~ l l o w ~ t hX a tdoes not exist, though I shall not argue that point horo.
..
toniau prc.tc.~isionsn ould suffer a radical loss if these principles of association were rcmovcd, he himself secnis incapable of realizing his commitmcnt to thcni. In the very sentence in which he calls the determination of the mind a propensity he identifies it as an "internal in~pression", or inipression of reflection (T.,p. 165). He uses the associatioris as propensities in elaborating his account of, e.g., how we come to suppose the existence of external objects, but 1ir. c.alls them impressions. The question of whether he is committed to the existence of entities other than perceptions is thus a y~~estion of critericz for rommitn~ent:is a man committed because of what he says or what he tloes? I prefer the latter, c~speciallyil)ccause of the alteratiolis in Hume's system if the principles were not propcrlsities, but I shall only raise, and not :irgne for, my prefiwnccs here. (2) There is 110 ariicl~lation o f sny ontological principle in Hume's writings. This omission is particularly striking given that he was rvriting against the I)ackgronncl of thc forccful articulation of one of the most famous ontological principles, eAse est percipi. I t is not plausible to suppose Hume unaware of Berkeley's writings or the reaction to them or to suppose him Ixnaware of the central importance of that principle to Berkcley's philosophy. i t is thus curious not to find some Humean version of it if Hume asserted both (a) and (b). (3) Hume says that "we can ncver have reason to believe that any object exists, of which we cannot form an idea" (T., p. 172). His argument is that to have a rcason for believing that X exists is to have present those conditio~lswhich allow a causal inference. No causal inference from one object to another can he made unless one has had expcricnce of the constant conjunction of similar o1)jects.l7 Since we experience only perceptions, we can have no reason to believe that anything othcr than perceptions exist (T., p. 172; see, c.g., T., p. 212). However, it seems we can also have no reason to believe that things other than perceptions do not exist. In brief, if the conditions for making a claim of existence do not exist, neither do the conditions for making a claim of non-existence. Take the case of an immaterial self. If one did exist, that we do not have a perception of it is just what me should expc>ct. But with no percclption of such a self, there can be no conjunctions necessary for rnalii~iga causal inference to its existence. Can we infer that it does not exist? On the contrary, everything we do perceive is consistent with its cxistcncc and so can gil-e us no reason to suppose its non-existence. Par from being able to infer the non-existence of a self from lack of a perception and thus from lack of evidence, such a lack is just what we should expect if i t does exist. 17Hurne denies tho necessity of a constant conjunction in the Treatise, saying that "we may attain the knowledge of a particular cause merely by one experiment, provided it be mado with judgement, and after a careful removal of all foreign and superfluous circumstances" ( T . , p. 104; see also pp. 131, 173). This denial is irrelevant to our concorn hero sinco a conjunction is still required, but it is worth noting that Humo takes it hack in thc Et~yuiry( E . ,p. 74).
40
WADE L. ROBISON
(4) The main difficulty with the reading of Reid, Butcllvarov, Penelhurl1 ct al., is that i t fails to t a l e seriously enough Hume's attack on metaphysicians. He argues against two kinds, the kind who posit the existenct. of such things as "sympathies, antipathies, and horrors of a vacuum" and the kind who assert an objective connection between a cause and its effect and the existence of material and immaterial substances (T., p. 224). The former he refuses to take seriously, sajing the inclination to posit such things "only takes place in children, poets, and the antient philosophers" and is due to "changeable, weak and irregular" principles of the imagination "u~hich,ho-\l-evercommon, are neither universal, nor unavoidable in human nature" (T., pp 224-6). The second kind he takes seriously, but according to Reid et al. he charges them with saying false things. They assert the existence of an immaterial substance, for instance, when only perceptions exist. This reading is unjust to the power and breadth of Hume's attack and to his understanding of why metaphysicians say what they say. His attack can be briefly sketched by considering what he thinks true of a metaphysician urbo asserts ".!I self exists": (i) He is trying to say something that has meaning and either failing or succeeding. He is succeeding if he intends to say no more than, e.g., that there are a collection of perceptions, for then there is an idea for the word 'self' to be associated with. But he is failing, Hume claims, if he intends to say that there is an object distinct from perceptions which continues the same through the whole course of one's life. This pattern of response is typically Humean, and i t obviously raises a variety of questions about existence and meaning-partly because Hume himself blithely talks about material and immaterial substances, necessary connections, and God when no acceptable translation seems possible. His discussions of these questions are few and spare, the most thorough being the one Butchvarov appeals to. I t does not support the claim that Hume thinks what is meaningless is false. Indeed, the only direct evidence there seems to be on the matter is Hume's considered charge against metaphysicians in his Letter to a friend, and he claimed to be accusing them not of saying false things, a common enough error, but of thinking they are saying true things which are in fact meaningless. (ii) We cannot infer the existence of an immaterial self from the existence of perceptions, for we can infer only perceptions from perceptions (T., p. 212; see also pp. 193,216-65). If our metaphysician is trying to assert the existence of an immaterial self, he is thus guilty of trying to infer what cannot be inferred as well as of making a claim which has meaning only by meaning something other than what he intends. (iii) Even if he could not infer the existence of such a self, he might be placated if he could believe in its existence. But he cannot. By Hume's theory, belief is "a different MANNER of conceiving an object, something that is distinguishable to the feeling, and depends not upon our will, as all our ideas
do" ( A ., p. 18). But since belief is a manner of conceiving, "belief implies a conception" (ibid.). We thus cannot believe (or disbelieve) what wc do not or cannot conceive. The consequcrlcc is that our metaphysician cannot even believe in what he is trying to assert the existerlce of.18 (iv) What is more, as we sam-, "we can never have reason to believe that any object exists, of which we cannot form an idca" (T., p. 172). S o t only can our metaphysician not believe in the existence of an immaterial self, nithout a reason, he also can have no reason to believe in its existence. To put it mildly, these are serious liabilities for even a metaphysician to operate under. But there is no claim that false things are said. These liabilities are rather to be explained, their interconnections tied, by the understanding that metaphysicians are attempting to explain features of our phenomenal world-coherent and constant impressions, and so on-by underpinning them with extra-phenomenal entities. Hume's charge is that these attempts do and must fail to explain-for the reasons cited. But this still fails to do justice to Hume's theory of philosophical theories. He thinks the naturc of the human mind such that we all must be metaphysicians: we all must suppose a necessary connection between a cause and its effect, a material substance which causes our impressions, and so on. Hume's scepticism and attack on metaphysics is thus not that we say false things, but the more shocking-and, paradoxically, less disrupting-charge that we must suppose things which are meaningless.19 I shall not expand this viehere. Let it suffice t o say that when Hume said we should commit some books of metaphysics to the flames, he was condemning pretensions, not the saying of false things. Ohio Stafe Universify
18This is not to claim that there are no natural beliefs, only that those beliefs are not a kind of belief that Hume's theory of belief allows for. lgSee my "Hume's Scepticism", Dialogue, 12 (1973), 87-99.