David Hume A Treatise of Human Nature ed by L A SelbyBigge second edn - PDF document

1 David Hume, A Treatise of Human Nature,
David Hume, A Treatise of Human Nature, ed. by L. A. Selby-Bigge, second edn revised byP. H. Nidditch (Oxford, Clarendon Press, 1978). Hereafter cited parenthetically as ÔTreatiseÕ.The distinction between ideas of memory and ideas of imagination is not taken up in theEnquiries. 2D. G. C. MacNabb, David Hume: His Theory of n McDonoughHumeÕs commentators have traditionally been highly critical of his accountof memory in the Treatise.1D. G. C. MacNabb simply states as a matter offact that Ôthe subject of memory is one on which Hume is notoriously weakÕ.And after suggesting that ÔHumeÕs whole account of memory is infectedwith the confusion between mental images and ÒideasÓ considered as unitsof thought,Õ the kindest thing he can !nd to say on HumeÕs behalf is that infailing to give an intelligible account of memory H

2 ume Ôwas no worse...than other philosoph
ume Ôwas no worse...than other philosophers of his own time and earlier timesÕ.2John Passmore agrees and imaginationhas been upheld by epistemologists recently concerned with these concepts, ISSN 0960-8788 print/ISSN 1469-3526 nite descriptions closely) resembles mÓ, where mis a variable ranging over positive ideas(mental images)Õ.5Since Hume never says that his memory ideas arerelative ideas, which Hume would be committed to by the content of FlageÕs de!nitedescriptions. That thesis Ð letÕs call it the Ôcausal thesisÕ Ð is that it is a neces-sary condition for a representationÕs being a memorial representation thatit be causally connected with whatever it represents.The causal thesis is, of course, almost universally accepted by contem-porary philosophers, and it is therefore tempting to defend HumeÕs accoun

3 tof memory by reading it back into his w
tof memory by reading it back into his work. Nonetheless I think that if wewish to sympathetically understand HumeÕs work on memory Ð and notmerely assimilate main sections. The !rst section iso-lates three puzzles in HumeÕs account of memory. The second sectionargues that while HumeÕs dif!culties Õs work by considering the roles Hume assigns to Hume on Memory and CausationÕ, Hume Studies Tenth Anniversary Issue,Supplement (1985) p. 168. For FlageÕs reading paradigmatically on rememberingsensory experiences. Within this restricted scope of cases, Hume further focuses on what heconsiders to be the central question an account of memory must answer:how are ideas of memory distinguished from ideas of imagination? Thatcentral question is, however, ambiguous. It might ask either (a) What con-stitutes the difference

4 between an idea of memory and an idea of
between an idea of memory and an idea of imagin-ation? or (b) What are the marks which we, in practice, use to distinguish(what we take to be) ideas me thinkthat I am remem-bering turning off the stove rather than imagining doing so?Õ is an instanceof (b).Hume, as we will see, The constitutive kinds of ideas...the imaginationis not restrainÕd to the same am to remember am simply to imagineaChristmas party happening last year, I am less constrained in my repre-senting. Õs party even if I cannot recall the brand of wine served,or whether I sat down before or after the madrigal singers performed.Hume, as we might expect, also recognizes that we may have such imper-fect memories. He writes: Ô with the following example:A musician somewhere in Siberia !rst made explicit in C. B. Martin and Max Deutscher, ÔRemembe

5 ringÕ The Philosophical absolutely non-c
ringÕ The Philosophical absolutely non-circular justi!cationfor its reliability it be a peculiar property of the memory to preserve the original orderand position of its ideas, while the imagination transposes 85)If memory defend his reason by reason; and by the same rule he must assent tothe principle concerning the existence of body, thoÕ he cannot pretend by anyarguments of philosophy to maintain its veracity. Nature has not left this to hischoice, and has doubtless esteemÕd it an affair suggested that Hume is concerned to not onlyanswer (a) what constitutes the rst sight, the ideas of memory are much more lively andstrong than those of the imagination, and that explain by calling it asuperior force, or vivacity, or solidity, or !rmness, or steadiness...I confess,that Õtis impossible to explain perfectly t

6 his feeling or manner of conception.We m
his feeling or manner of conception.We may make use of words, that express something near it. But its true andproper name is belief, which is a term that every one suf!ciently understands incommon life.(Treatise, 629) The sensible properties which Hume identi!es in giving his phenomenalcriterion strongly suggest that, in practice, ideas of memory are distin-guished from ideas of imagination on the basis of belief. Given that Humeapproaches the problem of classifying ideas egocentrically, that is aplausible suggestion. We are likely to think that our own ideas are mem-ories only if we believe that what they represent actually occurred in thepast. For this reason, we would have a hard time understanding someonewho claims that she remembers going to last yearÕs Christmas party, butdoes not believe that she went to it

7 . A criterion based on belief could not,
. A criterion based on belief could not, of course, be completely suf!cient,if for no other reason than the fact that we do not always approach HumeÕsquestions about memory egocentrically. We may judge that BobÕs belief thatMartians attended last yearÕs Christmas party is due to his Ð perhapseggnog-induced Ð imagination and not his memory. As a defeasible guideentering into our evaluations of our own representational suggestion. And there are passages whichsuggest that Hume offers his phenomenal criterion as a reliable (internal)heuristic and nothing more:We are frequently in doubt concerning the ideas of memory, as they becomevery weak and feeble; and are at a loss to determine whether any image pro-ceeds from the fancy or the memory, when it is not drawn in lively colours asdistinguish that latter facult degree

8 , as to be taken for an idea of the imag
, as to be taken for an idea of the imagination; so on the otherhand an idea of the imagination may acquire such imagination. The following passages, among others, suggest this alternativestrategy:Since therefore the memory is known neither by the order of its complex ideas,nor the nature of its simple ones; it follows, that thedifference betwixt it and theimagination lies in its superior force and vivacity.(Treatise, 85, italics added)Thus it appears, that the beliefor assent, which alwaysattends the memory andsenses, is nothing but the vivacity of those perceptions they present; and thatthis alonedistinguishes them from the imagination.(Treatise, 86, italics added)It frequently happens, that when two men have been engagÕd in any scene ofaction, the one shall remember it much better than the other, and shall hav

9 e allthe dif!culty in the world to make
e allthe dif!culty in the world to make his companion recollect it. He runs overseveral circumstances in vain...till at last he hits on some lucky circumstance,that revives the whole, and gives his friend a perfect might suppose that the manifest tangle itself"ows from philosophical carelessness, poor exposition, or both. revolutions a point withrespect to HumeÕs account of memory. Indeed, as we into supposing that it must have been obvious to Hume.10Clearly it was notobvious to Locke. Nor does the inference rule from Ôobvious to usÕ to Ôobviousto earlier generationsÕ have an impressive track record. With regards to thecausal thesis, however, not, I suggest, at all obvious. Considerthe seemingly similar phenomenon of referring to past historical !gures andevents. Many contemporary philosophers now endorse th

10 e view that causalchains may play an imp
e view that causalchains may play an important role in referring I suggest that it isfar from obvious Ð even if inescapably true Ð that a causal link is a necessarycondition for remembering a the causal thesis to Hume not only introduces a particularlytroublesome new together (a) and (b). For such a causal connection would be as clearly appli-cable in answering the constitutive question as it discussion in ÔLecture IIÕ, Naming criterionwould not be introspectable, the prospects for skepticism with culties the absenceofthe causal thesis presented for Hume in marking the distinction betweenideas of memory and ideas of imagination. Recall that Hume is concerned with sorting three different kinds of ideas:(a) accurate memories, (b) inaccurate memories, and (c) mere imaginings think she is inaccurately (1) or (2)

11 clearly in mind.13We saw earlier with t
clearly in mind.13We saw earlier with the ÔYesterdayÕ example that without the analyticconnection, Hume did not have the resources to fully sort memories (accu-rate or inaccurate) from mere imagininations. Without a causal criterion the musicianÕs ÔY inaccurate) mere imagininations. Furthermore, one might think that Hume does have the conceptualresources necessary to fully grasp the (1) distinction. Indeed, David Pearsspeci rather thanremembering? You do not need any criterion for that. You just meanyourmental image in one way rather than the other. This is the point Hume missescompletely.14We might think, however, that it is not a point that Hume misses completely.For Hume does at some stage suggest that we might be able to determinewhether we a memorial representation or an imaginativerepresentation by immediat

12 e introspection. We might nonetheless th
e introspection. We might nonetheless thinkthat example from C. B. Martin and MaxDeutscher: HUMEÕS ACCOUNT OF MEMORY8113As should become clear later, such an assumption appears entirely justi!ed in this case.Which is not to say that it might not be thought to be unjusti!ed Ð either because it is toostrong or too weak Ð in other cases. So, for example, one might think that the assumption istoo strong if one is impressed scene which the painter saw just the amazed observers have all the evidence needed toestablish that in being a memory without, as it were,improving upon its accuracy. (Treatisepassages suggest that Hume is concerned with distinguishing accu-rate memories from either inaccurate memories or mere imaginations, aswhen he claims thatthe imagination is not restrainÕd to the same order and form with th

13 e originalimpressions; while the memory
e originalimpressions; while the memory is in a manner tyÕd down in that respect, withoutanypower of variatioit impossible to recal the past impressions, in order to compare them with ourpresent ideas, and see whether their arrangement be exactlysimilar.16(Treatise, 9, 85, emphasis added in both)Once we recognize that Hume did not have either (1) or (2) clearly in mind,a solution to our !rst puzzle is readily forthcoming. Hume fails implausibly Ð be supposed toresult from careless oversight, but rather from his !nding the phenomenalcriterion compelling when thinking broadly in terms of (1), and !nding theconstitutive criterion compelling when thinking broadly in terms of (2). The!rst puzzle arises because Hume inaccurate memories (as well as mere imaginations) on the other, the needfor a non-introspectable poss

14 ibility of skepticismwith regards to mem
ibility of skepticismwith regards to memory.Solutions to both the second and third puzzles now !t snugly into place.We might reasonably suppose that in shifting from (roughly) the (1) dis-tinction to the (2) distinction, Hume would have no longer found the needfor a non-introspectable criterion necessary. That is, like Pears, he mighthave thought that we can simply tell from the inside (as it were) whetherwe are remembering (correctly or incorrectly) or imagining. Indeed withoutthe causal thesis that thought might well seem inescapable. That explainsHumeÕs temptation to shelve the before the work of Dedekind and Cantor, mathematiciansdid not have a perspicuous grasp of the difference between mere imaginations on the other. Since it appeals only to introspectablefeatures of our own mental cult passage reading of

15 Hume of a rational Being: And as far a
Hume of a rational Being: And as far as this consciousness any past Action or Thought, so far reaches the Identity of that Personit isthe same selfnow it was then; and Õtis by the same selfwith this present one thatnow re"ects on it, that that Action was done.18Memory is important in LockeÕs account because it is what makes diachronicidentity can recall theexperiences of Bob at T1.While agreeing with Locke that memory persons, Hume with an incon and movemen perceptions successively make their is properly no simplicityin it at one time, continued existence is the resultof a smooth transition from one idea (perception/impression) the next.That transition, Hume claims, is determined by the relations of resem-blance, causation, and contiguity in the case of objects; and (because hebelieves that not all ideas

16 constitutes his mindor thinking principl
constitutes his mindor thinking principle, and suppose that he always preserves the memory of aconsiderable part of past perceptions; Ôtis evident that nothing could more con-tribute resembling perceptions in the chain of thought, conveyHUMEÕS ACCOUNT OF MEMORY8519Hume, for example, writes:The identity which we ascribe to the mind of man, is only a !ctitious one, and of a likekind with that which we ascribe to vegetables and animal bodies. It cannot, therefore,have a different origin, but must proceed one object? In this particular, then, the memory...con-tributes to its production, HumeÕs account of how memory contributes to therelation of personal identity, however, would appear to reverse those com-mitments. First, Hume would make resemblance links essential to memory,and thus squander the resources of the c

17 ausal thesis to account for non-isomorph
ausal thesis to account for non-isomorphic memories. Second, Hume not only emphasizes the resemblanceof memorial representations, but does so to the exclusion of causal links.Memory and causation are both important in HumeÕs account of personalidentity, but not, it would seem, because memory is itself essentially causal. Hume nonetheless does y, as the sourceof personal identity. Had we no memory, we never should have any notion ofcausation, nor consequently of that chain of causes and effects, which consti-tute our self discovering perceptions, sonal identity precisely because it producesthe causal links which tie ourperceptions together. Again, however, Hume suggests just the opposite.Although causal relations are central to personal identity, Hume not with Hume, but with our own tendency to misunder- Philo