The Charm of Naturalism Author(s): Barry Stroud Source: Proceedings and Addresses of the American Philosophical Association, Vol. 70, No. 2 (Nov., 1996), pp. 43-55 Published by: American Philosophical Association Stable URL: http://www.jstor.org/stable/3131038 . Accessed: 31/03/2011 16:43

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp. 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/action/showPublisher?publisherCode=amphilosophical. .

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 a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

American Philosophical Association is collaborating with JSTOR to digitize, preserve and extend access to Proceedings and Addresses of the American Philosophical Association.

http://www.jstor.org THE CHARMOF NATURALISM BarryStroud, University of California,Berkeley

PresidentialAddress delivered before the SeventiethAnnual Pacific Division Meetingof TheAmerican Philosophical Association in Seattle,Washington, April 5, 1996

Iwant to makesome verygeneral observations on whatmany see andapplaud as a broadly"naturalistic" turn in recentphilosophy. There seems littledoubt at firstglance that there is such a thing,at leastjudging from what many now call whatthey are doing. Somethingknown as "epistemologynaturalized" has been withus forsome time. Orat leasta recommendationto that effect was madesome time ago.1 Morerecently we have been encouragedin such enterprisesas "naturalizedsemantics," "naturalizingbelief," and even more generally, "intentionalitynaturalized." And now there is the even moregeneral project (why notgo allthe way?) of "naturalizingthe mind"(the title of a delightfulrecent book)2. I haveeven seen somethingcalled "naturalizing responsibility." And there are no doubtmany other efforts at "naturalization." Is theremore to allthis than just a trendylabel? What, if anything, is behindit? Is it somethingdistinctive, and new? Andif so, is ita goodthing? These questionsare not easy to answer. The idea of "nature,"or "natural" objectsor relations,or modes of investigationthat are "naturalistic,"has been appliedmore widely, at moredifferent times and places, and for moredifferent purposes,than probably any othernotion in the wholehistory of humanthought. The earliest turntowards naturalismthat I have heard of was in the fifthcentury B.C.. Andthey seem to have been happeningevery so oftenever since. Whenwe lookat this most recententhusiasm for what its proponentscall "naturalism,"I think we findthat, whatever they are excitedand optimistic about, itis notnaturalism as such. Withtwo exceptions that I willmention in a moment, I thinkthere is nothingin naturalismalone thatis sufficientlysubstantive to be philosophicallycontroversial. What is usuallyat issue is not whetherto be "naturalistic"or not, but ratherwhat is and whatis not to be includedin one's conceptionof "nature."That is the realquestion, and that is whatleads to deep disagreements.And as faras I can see, thosedisagreements are notthemselves to be settledby what can be recognizedas straightforwardly"naturalistic" means. So one thingthat seems notto have been "naturalized"is naturalism itself. Ifit were, the resulting naturalistic view of the world might be impressively comprehensiveand illuminating, and superior to viewsof otherkinds, but if it had those virtues it would have them on its own merits, not simply because it is an instance of something called "naturalism." "Naturalism"seems to me in this and other respects ratherlike "WorldPeace." Almosteveryone swears allegiance to it,and is willingto marchunder its banner. Butdisputes can stillbreak out about what it is appropriateor acceptableto do in the name of that slogan. And like world peace, once you start specifying concretelyexactly what it involvesand howto achieveit, it becomesincreasingly

- PROCEEDINGSAND ADDRESSES OF THEAPA, 70:2 - 43 Presidential Address of the Pacific Division difficultto reachand to sustaina consistentand exclusive "naturalism." There is pressureon the one handto includemore and morewithin your conception of "nature,"so it loses its definitenessand restrictiveness.Or, if the conceptionis keptfixed and restrictive, there is pressureon the otherhand to distortor even to deny the veryphenomena that a naturalisticstudy-and especiallya naturalistic study of humanbeings-is supposed to explain. The source of these two conflictingmovements of thoughtis whatI wantto illustrate. Butthe firstthing to do withnaturalism, as withany philosophicaldoctrine or "ism,"is to ask whatit is against.What does the so-called"naturalistic turn" turn away from,or deny? Herewe have to distinguishtwo aspects of naturalism.There is naturalismas a view of what is so, or the way thingsare, or whatthere is in the world. Andthere is naturalismas a way of studying or investigatingwhat is so in the world. A naturalisticstudy of humanbeings wouldstudy and understoodthem in relationto the rest of nature. Obviously,what you thinkthe naturalworld is likewill have an effecton howyou investigatethe things in it, and what you think is the best wayto understandthem. Thetwo aspects of naturalismare connected. Underthe firstaspect, as a doctrineabout what is so, or what there is, naturalismsays thatthere is nothing,or thatnothing is so, exceptwhat holds in nature,in the naturalworld. That is notvery informative so far,but even without specifyingit moreprecisely it alreadyseems to excludesome thingsthat many people have apparentlybelieved in. Naturalismon any readingis opposedto supernaturalism.Here we havewhat looks like a substantiveissue, or at any rate somethingcontroversial. Not everyone regards exclusive naturalism as beyond questionor as an unqualifiedgood thing. Thisis the firstof thetwo exceptions I mentioned.By "supernaturalism" I mean the invocationof an agent or forcewhich somehow stands outside the familiar naturalworld and so whose doingscannot be understoodas partof it. Most metaphysical systems of the past included some such agent. A naturalistic conception of the worldwould be opposed to all of them. Supernaturalismas a doctrineabout what is so can have consequences for the study of humanbeings-in particular,how they believe and come to knowthings. In epistemology there have been many supernaturalists.Descartes thoughtthat humanknowledge cannot be accountedfor without a benevolent,omniscient, and omnipotentGod who guarantees the truthof what human beings clearly and distinctlyperceive to be true. For Berkeley,God's agency is the only active force there is in the worldof things we perceive and know about. Withouthim there would be nothingfor us to know. Even Lockerelied on a benevolent agent as the ultimatesource of those cognitivefaculties which are all that humanbeings need to get along in the worldthey findthemselves in. These are not fullynaturalistic accounts of human knowledge. They appeal to something beyond the natural world. Ingoing againstthis supernaturalconsensus, Humeis almostalone among the greats. His credentials as a fully naturalized-or at least as a non- supernaturalized-metaphysicianand epistemologistare impeccable. The same

44 - PROCEEDINGSAND ADDRESSES OF THEAPA, 70:2 Presidential Address of the Pacific Division is probablytrue of JohnStuart Mill, if he countsas one of the greats. Butthere have not been many. Inthe sense inwhich naturalism is opposedto supernaturalism,there has been no recent naturalisticturn in philosophy. Mostphilosophers for at least one hundredyears have been naturalistsin the non-supernaturalistsense. Theyhave taken it for grantedthat any satisfactoryaccount of how humanbelief and knowledgein generalare possiblewill involve only processes and events of the intelligiblenatural world, withoutthe interventionor reassurance of any supernaturalagent. Manypeople regard that as on the wholea good thing. But it is nothingnew. In fact, the long-standingnaturalistic consensus is beingchallenged more directlynow, when the virtuesof naturalismare beingso loudlyproclaimed, than itwas duringthe longperiod when they went more or less withoutsaying. Alvin Plantinga,for example, argues that no satisfactorygeneral explanation of human knowledgecan be given on a naturalisticbasis. He thinksthat justification or warrant,which is essentialto knowledge,can be understoodonly in termsof the properfunction of humancognitive capacities. And that in turn, he thinks,requires a divinedesigner of thosecapacities. Successful epistemology therefore "requires supernaturalism,"3in particular, what he calls"theism."4 He is apparentlynot alone inthat belief. IfPlantinga and his friendsconvince others, there will be a general turnaway from naturalism. That shows thatit is naturalismthat is nowold hat. It is notsomething towards which there has been a recent,glorious turning. Even supernaturalistslike Plantingaand Descartes,Locke, Berkeley, and otherswould still count as "naturalizedepistemologists" in at least one current sense that has been given to that phrase. Epistemologyhas been said to be "naturalizedor naturalistic"as long as ittries to explainonly how human beings do in factarrive at theirbeliefs rather than how they ought to arriveat them.5If that is enoughto makean epistemologynaturalistic, then virtually every philosopher in historyhas been a naturalizedepistemologist. They have all been concernedto describeand understandthe humancondition as it is, to see andto explainhow we actuallyget all the knowledgewe've obviouslygot. If God plays a role in humanbeings' coming to knowthings, that will be partof the answerto the purely "descriptive"question of howhuman beings in fact arrive at theirknowledge. Even supernaturalismas a view of what is so is not incompatiblewith naturalized epistemologyin thiscuriously weak, so-called purely "descriptive" sense. Thisshows that the first aspect of naturalismdominates over the second. Ifyou do not start out withany restrictionsat all on whatthe worldyou are studying contains,studying things only as partof the naturalworld does not amountto anythingvery definite. Some determinateconception of whatthe naturalworld is like is needed to give substanceto the claimthat one's epistemology,or one's studyof any otheraspect of the world,is naturalistic. Thesecond exception to the ideathat there is no realdispute about naturalism is perhapsbest illustrated(at leastin epistemology) by Quine,who after all, as far as I know,is the personwho coinedthe phrase"epistemology naturalized." He

- PROCEEDINGSAND ADDRESSES OF THEAPA, 70:2 - 45 PresidentialAddress of the PacificDivision was respondingto Carnap,with whom he hada realdispute. Carnapsought a reductionof all talk of externalbodies to talk only about possible sense experiences.It was intendedas whatcame to be calleda "rationalreconstruction" of ourscience or knowledge.It would have shown how our conception of theworld couldbe supportedsolely by materialsto be foundin immediate sense experience. But no satisfactorytranslation or reductionwas found. The idea of "rational reconstruction"does notin itselfrequire that statements about the externalworld must be translatableinto statementsabout immediatesense experiences. It requiresonly that it be shownhow our beliefs about the world could be justifiedby informationthat we couldget throughexperience or observation. That general task of what mightbe called"hypothetical (or reconstructed)justification" was pretty muchthe task of analyticepistemology through the middlefifty years or so of this century.One form it took, and perhapsstill takes, is confirmationtheory. Quine's so-called "naturalistic"turn was to say "whyall this creative reconstruction,all this makebelieve?."6 Why ask how statementsof the kinds humanbeings believe could be confirmedby sense experiencesthey could conceivablyhave? As Quineput it:"Better to discoverhow science is in fact developedand learnedthan to fabricatea fictitiousstructure to a similareffect."7 The questionis howscience is "developedand learned."It is notjust a question of the logicalrelations among the propositionshuman beings believe. Somethingis at stakehere betweenQuine and Carnap, but it is notthe merits of naturalism.It is reallya disputeabout what philosophy is or oughtto be doing. Quine obviously has no quarrel with the idea of reducing one domain of discourse to another,if you can do it. Carnapand the positivistsobviously have no quarrel withthe ideaof naturalscientific studies of humanbelief and knowledge,or even of institutionslike science. Naturalismas a way of investigatingthe worldis thoughtby all to be nothingbut a good thing. Butfor logical positivism no such studies couldbe partof philosophy.Philosophy could be onlya priori.Its only subject-mattercould therefore be the"concepts," or the logicalrelations among the "principles,"employed in the sciences. Itsonly task could be "analysis."It could notpronounce professionally on theactual acquisition and development of science, butonly on whatit called its "logic." Quine'srejection of the very notionof the a priorileft himwith no such constraints.Study human knowledge in the same wayyou would study anything else in nature,he says, anddon't worry much about what label you attachto what you are doing. Thatmeant that a taskcontinuous with what epistemologists had attemptedin the pastcould now proceed scientifically. Epistemology would in that sense be partof naturalscience, and itwould study the acquisition,transmission, and growth of natural science. The idea is, in Quine's words, "that knowledge, mind, and meaning are part of the same world that they have to do with, and that they are to be studied in the same empirical spirit that animates natural science. There is no place for a prior philosophy."8 That same empirical spirit is present in the study of the history of science, which could also be described as a form of naturalism in the investigation of human knowledge. The history of science has of course been with us almost as long as

46 - PROCEEDINGSAND ADDRESSES OF THEAPA, 70:2- Presidential Address of the Pacific Division science has, butits flourishing in the 1960swas in partalso a reactionagainst the abstractionsof logicalpositivism. The positivists focussed on whatis known,or on the formof whatis known,rather than on the knowingof it,or on the processes of findingit out. Theydid not study science as a humanenterprise that develops in differentways at differenttimes as a resultof differentsorts of forces. Thatcould notbe partof philosophyfor them. Thegrowth of the historyof science inthe last thirtyyears has changedand enormously enriched the picture.And it certainly has been a verygood thing. Quinehimself at one timeseemed not so sure. He thoughtthe historically- orientedwork of peoplelike Kuhn, Polanyi, and Hanson had "loosed a wave ... of epistemologicalnihilism" (as he put it) and tended to "discreditthe idea of observation,"to "belittlethe roleof evidenceand to accentuatecultural relativism."9 These are curiouscomplaints for a fullynaturalized epistemologist to make. Scientificepistemology must be preparedto acceptwhatever the empiricalstudy of humanbeings actually reveals. Ifit turns out that human knowledge is acquired without there being a firm, fixed line between so-called "observational"and "non- observational" terms, or if what a philosophical "theory of evidence" calls "evidence" is never actually appealed to in the acceptance and rejection of scientifichypotheses, then so be it. Thatwill have to be acceptedas the way knowledgeis infact acquired. If cultural relativism turns out to be the best way to accountfor what happens in humanlife, the committednaturalist has to accept culturalrelativism. (What he shoulddo firstof all,of course,is tryto figureout what the term'cultural relativism' actually means. Butthat is anotherstory.) The pointis thatconclusions of naturalizedepistemology can be drawnonly from the study of what actually goes on with human beings. If it turns out that women's knowledgediffers in certainways frommen's, for instance,or poor southernblacks' knowledge from that of affluenturban whites, that is something thata naturalizedepistemologist should welcome, or at any rateshould not resist. Studiesin the sociology,economics, and politics of knowledgecould also be called "naturalisticepistemology" too. The livelyinterest in such mattersthese days is certainlyon the wholea good thing. Notbecause naturalismis a good thing,but becausecoming to see moreand moredifferences among things in the world-if theyare actuallythere-is almostalways a goodthing. I want to draw attention to a conflict or tension that I think is present in a commitmentto naturalism.It arises most clearly when we movebeyond questions aboutthis or thatculture or this or thatinstitution within a cultureto that more general level at which philosopherstypically ask about apparentlyuniversal featuresof humanlife. NowI meannaturalism inevery area of philosophy,not just epistemology. Naturalismas a view of whatis so, or whatthe worldis like,must be given some determinateand restrictedcontent. That means that anything that human beingsthink about, believe in, care about,or valuethat lies outsidethat restricted conceptioncannot really be seen as partof the naturalworld in which they live.

- PROCEEDINGSAND ADDRESSES OF THEAPA, 70:2 - 47 PresidentialAddress of the PacificDivision But since it cannot be denied that people do have the very thoughts, beliefs, values, and concernsin question,the contentsof those attitudeswill have to be understoodand accounted for in terms of somethingless thantheir possible truth. Whathuman beings think, feel, and care about must be fullyexpressible somehow withinthe restrictedresources available in the naturalist'sworld. And that can lead to distortion.If, to accommodatepsychological phenomena and their contents in alltheir complexity, the restrictionsare lifted, naturalism to thatextent loses its bite. Thisis the basicdilemma I wantto bringout. I can illustrateit by startingwith an extremenaturalist view. I wouldsay thatit is a ridiculouslyextreme position, were it notfor the factthat many philosophers I respect appearto holdit. It says thatthe naturalworld is exhaustedby all the physicalfacts. Thatis alland only what the naturalworld amounts to on thisview; thereis nothingelse in nature.First of all,this view is probablynot itself reached by purelynaturalistic means. It not only states all the physicalfacts, which presumablycan be determinedby broadlynaturalistic means. Butit goes on to say thatthose are allthe factsthere are-that theyare the wholetruth about the world. Andthat claim is morethan the conjunctionof all the physicalfacts. It excludes everythingelse from being true, as they alone do not. Is the exhaustivenessthat is essentialto physicalismsomething that is naturalisticallyor physicalisticallyarrived at? Thatis one question. Second,a naturalworld conceived of onlyas thetotality of allthe physicalfacts obviouslydoes not containany psychologicalfacts. Thereare no truthsto the effectthat someone believes, knows, feels, wants, prefers, or values anything. Of course,anyone who holds that the physicalis allthere is mighthold that everything we thinkalong those linesis reallyjust physical facts in disguise. In any case, that wouldleave no psychologicalfacts for a naturalistictheory of the worldto explain. The studyof humanbeings on such a restrictedphysicalist conception would be justa studyof physicalgoings-on, including some thathappen to go on in human organisms. The case is extremebecause it does not includevery muchfor a studyof humanbeings to explain. Withoutat least biologicalfacts in your naturalistic conceptionof theworld you will not have much to investigatethat is distinctivelyor interestinglyhuman. But if the physicalistconception is expandedto include biologicalfacts as well,what exactly are such factsthought to add? Do biological facts includethe "intentional"facts of humanbeings believing, knowing, feeling, wanting,preferring, and valuing certain things? Some wouldsay not,since these arejust "folk" ways of speaking.Organisms inhabiting the naturalworld are notbe thoughtof as havingany such attitudes, or as actingfrom them, on thatview. That wouldmean that naturalism could never be facedwith the problemof explaining howand why human beings come to believeand feel andwant the thingsthey do. Therewould be no suchfacts. Naturalismas to whatis so wouldbe so restrictive as to leave naturalismas a methodof investigationwith much less to do. Thereis an embarrassingabsurdity in this position which is revealedas soon as the naturalistreflects and acknowledges that he believeshis naturalistictheory of theworld. If persons with attitudes like belief and knowledgeare notreally part

48 - PROCEEDINGSAND ADDRESSES OF THEAPA, 70:2 - Presidential Address of the Pacific Division of nature,he cannotconsistently say thatabout himself. I meanhe cannotsay it andconsistently regard what he is sayingas true. Infact he cannotsay anything and regardit as true,or thinkof himselfas sayingit, if he holdssuch a restricted naturalisticconception. It looksas if any sensible naturalismwill have to acknowledgethat human beingsdo infact have a complexset of attitudes,feelings, evaluations, institutions, and so on. Ifit is goingto explainwhat is so, itwill have to explainhow and why humanbeings think and feel and act in all the ways they do. Itwill offer those explanationsby appealing not only to theways humanbeings are, but also to facts of the naturalworld surrounding and affecting those human beings. To explain why peoplebelieve that there are suchthings as rectangulartables, for example, or red apples, itwill trace the connectionsbetween human beings who perceivethings anda worldthat contains rectangular tables and redapples. Itis because things are as they are in the naturalworld, and because humansare as they are, and interactas theydo withtheir surroundings, that they get the beliefsthey do andare on the wholeright about the naturalworld. Even this simplegeneral picture leaves roomfor humanattitudes directed towardsobjects or states of affairswhich restrictive forms of naturalismcan findno roomfor within their conception of theworld. For example, many philosophers now holdthat things as theyare inthe world of natureare notreally colored. There are rectangulartables in the naturalworld, perhaps, and there are apples in the natural world,but no redapples (and no yellowor green ones, either).This view appears to be heldlargely on the groundsthat colors are notpart of "thecausal order of the world"or do not figureessentially in any purelyscientific account of whatis so. Scientificnaturalism accordingly excludes them. Buteven on thisview those false beliefsand illusoryperceptions of the colors of things must themselvesbe acknowledgedas partof nature. A naturalistic investigatormust somehow make sense of themas the psychologicalphenomena theyare. Since he holdsthat there is no suchfact as an object'sbeing colored, he cannot specify the contentsof those perceptionsand beliefs in terms of any conditionsthat he believes actuallyhold in the world. If he could,that would amountto believingthat there are coloredthings in the worldafter all. Scientific naturalismdenies that. Butstill, the beliefsand perceptionswith those particular contentsmust be accountedfor. An easy way aroundthis difficulty has suggesteditself to manyphilosophers, at least in thiscase. Theytake the apparentlymore sensible scientific naturalist viewthat there really is no systematicerror in our beliefs about the colorsof things. The beliefs are not in generalfalse, since there is somethingin the restricted naturalist'sworld to give contentto them afterall. Beliefsabout the colors of objects,it is said,are reallybeliefs about certain dispositions which those objects have to produceperceptions of certainkinds in certainkinds of perceiversin certainkinds of circumstances.Objects in nature really do havethose dispositions. So the beliefsare preservedas largelytrue. The colorof an objectdepends on what kinds of perceptionsit is disposed to produce.

- PROCEEDINGSAND ADDRESSES OF THEAPA, 70:2 - 49 PresidentialAddress of the PacificDivision A dispositionalisttheory of this kind can succeed only if it can specify the contentsof the perceptionsof colorwhich it says physicalobjects have dispositions to produce.They cannot be identifiedas perceptionsof an object'shaving some qualitythat objects actually have in thatrestricted naturalist's world. They cannot be identifiedsimply as perceptionsof an object'shaving a dispositionto produce just these perceptionsunder certain circumstances. The questionis: which perceptions? There must be some way of identifyingthe perceptions independentlyof the object'sdisposition to producethem. So it looksas if they mustbe identifiedonly in terms of some so-called"intrinsic" quality that they have. Not a qualitythat the perceptionis a perceptionof, butsimply a qualityof the perceptionitself. I doubtthat we can makethe rightkind of sense of perceptionsof colorin this way. So I doubtthat any dispositionaltheory can give a correctaccount of the contentsof ourbeliefs about the colorsof things. Theway we do it in reallife, I believe,is to identifythe contentsof perceptionsof colorby meansof the colorsof the objectsthey are typicallyperceptions of. It is only because we can make intelligiblenon-dispositional ascriptions of colors to objects that we can acknowledgeand identifyperceptions as perceptionsof thisor thatcolor. Butif thatis so, it requiresour accepting the factthat objects in the worldare colored, andthat is whatthe restrictivenaturalist who denies the realityor the objectivityof colorcannot do. Noneof thisis somethingI can hopeto establishhere. Thepoint is onlyto draw attentionto whatI see as a generalproblem of restrictivenaturalism. Exclude coloredobjects in generalfrom the world,and you are in dangerof losingthe capacityto recognizeperceptions of andbeliefs about the colorsof things.Include coloredobjects, and the contentsof those perceptionsand beliefsno longergo beyondwhat is so in the naturalworld. I have found in my experiencethat this tension is not widelyfelt or acknowledged. Mostphilosophers regard it as so obviousand uncontroversialthat colorsare not real,or are in some way only"subjective," that they simply do not recognizewhat I thinkis the distortionor incoherencethey are committed to. That is somethingI continueto ponder,and try to get to the bottomof. Buta problem of this same formis at least sometimesrecognized elsewhere. Twolarge areas of philosophyare problemareas preciselybecause some formof restrictive naturalismlooks like the onlypossibility in those cases. I have in mindthe areas of mathematicsand morality,or evaluationgenerally. Humanbeings have evaluativebeliefs and attitudes; they regard some things as betterthan others, they think that a certainthing is the thingto do on a certain occasion,and so on. To understandand acknowledgethe presenceof these humanattitudes in the world, the naturalistmust understand their contents-what those humanbeings actually think or believe. Naturalismis widelyunderstood to implythat no evaluativestates of affairsor propertiesare partof the worldof nature. On thatassumption, either evaluative thoughts and beliefstake as their "objects"something that is not to be foundin the naturalworld at all, or their

50 - PROCEEDINGSAND ADDRESSES OF THEAPA, 70:2 - Presidential Address of the Pacific Division contentsare equivalent to somethingthat is truein that world, so theyare notreally evaluative. One way to embracethe firstoption would be to say withG. E. Moorethat evaluativestatements are assertionsabout a "non-natural"world, or thatthey ascribe "non-natural"properties to objectsin the naturalworld. We mightthen wonderwhat that "non-natural" domain is like,and howit is relatedto whatgoes on beforeour eyes. Andwhatever it is, we mightwonder why we shouldever take any interestin it. Valuesmight then be other-worldly,and have nothingto do with us. Ifall thatis just too mysterious,we couldkeep to thisfirst option by saying insteadthat evaluative attitudes do nothave contentsthat are trueor false at all. In evaluatingsomething we are prescribing,recommending, approving, or encouragingsomething, but not ascribing any propertiesto it or sayinganything trueof it beyondthe "natural"properties we thinkit has got. Thislast idea, I believe,distorts our actual thought and practice.It cannot give the rightkind of sense to the evaluativethoughts we have or the inferenceswe regardas valid when combiningevaluative and non-evaluativepropositions. Again,that is notsomething I am goingto tryto provehere. I simplydraw attention to the sourceof the pressuretowards some such emotivistor expressivisttheory. It comes froma restrictednaturalistic conception of what the worldcontains. Natureitself, it is said, is value-free.So evaluationscannot be strictlyspeaking eithertrue or false. Thatis one alternative.It is notan inevitableconsequence of a restrictednaturalistic view of the world.The same restrictiveview of natureis whatleads a non-expressivistlike Moore to the ideaof valuesas "non-natural"or in some sense "other-worldly." Dissatisfactionwith both non-naturalism and expressivism leads the restrictive naturalistto the second option,and so to some formof reductionism.Human attitudesthat appear to be evaluativeare to be seen as attitudeswith contents whichcan anddo holdin the restrictednatural world after all. Theycan be trueor false, butthe conditionsof theirtruth are purelynatural and so non-evaluative. Ifsuch a reductionis expressedin terms of the dispositionsnatural objects or states of affairshave to producecertain reactions in humanbeings, it faces the same kindof problemas the dispositionalistview of colors. Those reactions themselves mustsomehow be identified,and if they are left as reactionswith evaluativecontents no naturalisticprogress will have been made. Reductionism threatensto take away the evaluativeaspect of the attitudes,feelings, and reactionsthat objectsare said to produce,just as I thinkit cannotmake the appropriateidentifications in the case of perceptionsof color. Itcannot get the contentsof ourbeliefs or attitudesright. To insistthat evaluative attitudes simply mustbe so reducible,and to restrictoneself to reducedor non-evaluativeterms alone, would be in effect to eliminatethe evaluativevocabulary altogether. Everythingwe say orthink that is intelligibleand either true or false wouldhave to be said or thought without it. Here again it is the restrictivenaturalism that producesthe pressure.

- PROCEEDINGSAND ADDRESSES OF THEAPA, 70:2 - 51 Presidential Address of the Pacific Division The same patternis presentin the philosophyof mathematics,where the quandaryis perhapsmost obvious, and has certainlybeen widely acknowledged. Thereis no questionthat we have mathematicaland logicalknowledge. Could therebe an explanationexclusively in restrictednaturalistic terms of howwe come to have that knowledge? Itwould have to makesense of whatwe believein mathematicsand logic,and could it do so by givingan accountof the conditions under which such things are true or false? If so, would that mean that mathematicaland logicalfacts are to be understoodas partof nature? Many wouldinsist that even if in some sense or otherit is truethat seven plusfive is twelve,it is nota naturalfact, not a factof the naturalworld. But we do all believe it,even knowit to be true.A restrictivenaturalist who holds that what mathematical statementsassert is not partof the naturalworld he believesin wouldhave to explainour knowledge of logicand mathematicswithout himself appealing to any mathematicalor logicalfacts at all. Thishas beentried, or at least proposed.But when we lookat whathas been the most widely-canvassedstrategy for carryingit out I thinkeveryone has to confessto a certaindissatisfaction. The mainidea has been to locatethe source of mathematicaland logical truthsomehow "in us," and not in the world independentof us. Allsuch truthshave been saidto be "analytic"or "truesolely in virtueof the meaningsof theirconstituent terms," something which "we" are in some sense solelyresponsible for. Sincewords mean only what we determineor "decide"they are to mean,logical and mathematical truths are said to be true,if at all,only "by convention." These are allattempts to makesense of mathematical and logical knowledgeon the assumptionthat all of it is "empty"or, in the positivists'phrase, "devoid of factualcontent," and says nothingabout the way the worldis. Anyonewho holds such a viewwould have to accountfor human beings' believingcertain things which he himselfdoes not acknowledgeto be states of affairsthat hold in the worldas he conceivesof it. Thereis goodreason to thinkthat no such theoristwould be equippedeven to identify,let alone explain, the mathematicalknowledge he admitswe allhave. For one thing,no naturalisticreduction looks even remotelyplausible in this case. Facts aboutwhat humanbeings do, how they thinkor speak, even how they decideto thinkand speak, or what conventions or rules they have adopted-all this seems in principleinsufficient to express the contentsof the thingswe believe whenwe believethat seven plusfive equals twelve or thateverything that is both redand round is red. Allof humanbeings' doing or deciding or intendingwhatever theydo is contingent,something that could have been otherwise.But it couldnot have been otherwise than that seven plus five is twelve or that everything that is both red and round is red. No contingent truths, however important, could be adequate to express such necessities. What is more, any naturalism that takes a specifically scientific form, and says that the naturalworld is the world described exclusively in the terms of the natural sciences, would seem forced to accept truths of logic and mathematics anyway. They are needed in the formulation of physical, chemical, and biological theories. And in any case, it is completely unrealistic to expect a naturalistic theorist of any

52 - PROCEEDINGSAND ADDRESSES OF THEAPA, 70:2- Presidential Address of the Pacific Division persuasionto get alongwithout any mathematicaland logicalbeliefs of his own. Theacceptance of some suchtruths might even be essentialto coherentthought; we could not thinkwithout them. Ifthat is so, is thata naturalfact, a fact of the naturalworld? Ifthat meant that it hadto be contingent,it is hardto see how it couldbe. Butif for whatever reason we grantthe indispensabilityof logical truths forthe possibilityof thinkingat all,then we haveto face the consequencesof our reallyaccepting it. Thatis, we mustacknowledge that we do infact think in those ways, thatwe do believethat everything that is bothred and roundis red,that seven plusfive equals twelve, and so on. Wethereby acknowledge that those and othersuch demonstrableor undeniablepropositions are true. Ifthe naturalistdoes or mustaccept logical and mathematicaltruths in orderto have a determinateconception the worldat all, whatbecomes of the idea that those propositionsdo notstate anythingthat holds in the naturalworld? What is the conceptionof naturethat is saidto excludethem? Itcan no longerbe identified as simplythe worldthat a scientificnaturalist believes in, since if he nowaccepts logical and mathematicalpropositions, they are not excludedfrom what he believes. Ifthis still counts as naturalism,itwill be a moreopen-minded or more expansivenaturalism. It does not insiston, or limititself to, a boundaryfixed in advance. Itwill have expanded to includewhatever has been foundto be needed inorder to makesense of everythingthat is so inthe naturalworld. What cannot be avoided is to be accepted. To say that not everythingthat is accepted is acceptedas partof natureraises the questionof howthe naturalistdistinguishes whathe thinksof as the naturalworld from all the restof whathe takesto be the case. And more importantly,what, if anything,now turns on makingthat distinction? The same questionarises in the case of evaluation.If the goodnessor other evaluativeaspect of somethingis not a "natural"quality of it, whatexactly is a naturalquality? After years of effortG. E. Mooreadmitted that the best he could comeup with was thata naturalproperty is a property"with which it is the business of the naturalsciences or of psychologyto deal."10But if thatis whata natural propertyis, then the famous "naturalisticfallacy"-the mistakeof giving a "naturalistic"definition of 'good'-wouldbe simplythe attemptto replaceethics by one of the naturalsciences. "Non-naturalism"inethics would then be nothingmore than the view that ethics is not one of the naturalsciences. Therewould be nothingother-worldly or mysterious about that kind of non-naturalism.Who would notwant to be an ethicalnon-naturalist on thatdefinition? To agree thatethics is not one of the naturalsciences, or thatgoodness or badness is not a scientificmatter, is not to concede thatnothing is betterthan anythingelse, orthat no evaluationsare trueor false. Noteverything that is so is the subject-matterof some naturalscience. Ifit is truethat evaluations cannot be reduced in general to non-evaluativepropositions, then our understandingof evaluations cannot be seen as builtup out of non-evaluativeingredients alone. Anyone who could identifythe presence of evaluative attitudes in the human beings he observes must understandwhat evaluative attitudes are, even if he does

- PROCEEDINGSAND ADDRESSES OF THEAPA, 70:2 - 53 Presidential Address of the Pacific Division not agree with those he discerns in others. That suggests that he must have some evaluative attitudes of his own, on pain of his not being able to recognize them in others. If he acknowledges those attitudes of his, his total view of what is so will contain evaluative states of affairs. He will hold that certain things are better than others, that a certain thing is the thing to do on a certain occasion, and so on. His conceptionof whatis so willhave been forcedto expand,just as I thinkit must expand in orderto recognizebeliefs in logicaland mathematicaltruths, and perceptionsand beliefsconcerning the colorsof things. Itexpands in each case intoa moreopen-minded or less restrictednaturalism. WhatI am callingmore open-minded or expansivenaturalism says we must accepteverything we findourselves committed to inaccounting for everything that we agree is so and wantto explain. We wantto explainthe thoughts,beliefs, knowledge,and evaluative attitudes that we thinkpeople have got. Ifmathematical and logicaltruths have to be acceptedin orderto makesense of those attitudes, then they mustbe accepted,however in some sense "non-natural"they might seem. If some evaluativepropositions must be endorsedin ordereven to recognizethe evaluativeattitudes of others,then evaluative states of affairsmust be includedtoo, howeverdifficult itmight be to decidewhich particular evaluations are correct. Ifwe have to holdthat objects are coloredin orderto specifyand acknowledgeall the perceptionsand beliefsthat we knowpeople have, then the colorsof thingsmust be allowedinto the picture,and notin reductionistform. Thosewho remain committed to a determinateand restricted conception of the naturalworld will have to locatethe contentsof allthose attitudessomehow within that restrictedworld. Ifthat leads to a distortedconception of the attitudesthat peopleon earthhave actually got, as Ithink it does, the determinateand restricted naturalismis what is responsiblefor the distortion.A more open-mindedor expansive naturalismwill admit states of affairs and psychological phenomena that are found problematic from a more restricted naturalistic point of view. With no restrictive commitment in advance, a more open-minded naturalism will feel no pressure to exclude from the picture anything that is needed. By nowit should begin to lookas ifthis expandable or moreopen-minded form of naturalismdoes notamount to anythingvery substantive or controversial.It is "open"because it is notcommitted in advanceto any determinateand therefore potentiallyrestrictive conception of whatis so. Ratherthan calling it open-minded naturalism we could just as well drop the term 'naturalism' and call it open- mindedness. It says that we must accept as true everything we find we have to accept in order to make sense of everything that we think is part of the world. If that is still called "naturalism,"the term by now is little more than a slogan on a banner raised to attract the admiration of those who agree that no supernatural agents are at work in the world.

Notes

1. See W. V. Quine, "Epistemology Naturalized,"in his Ontological Relativity and Other Essays, Columbia University Press, New York, 1968.

54 - PROCEEDINGSAND ADDRESSES OF THEAPA, 70:2- Presidential Address of the Pacific Division 2. FredDretske, Naturalizing the Mind,MIT Press, Cambridge,Mass., 1996. 3. A. Plantinga,Warrant and Proper Function, Oxford University Press, NewYork, 1993, P. 46. 4. Ibid. p. 237. 5. For this description of naturalized epistemology see the Introduction to H. Kornblith(ed.), NaturalizingEpistemology, MIT Press, Cambridge,Mass., 1985. 6. W. V. Quine, op. cit., p. 75. 7.Quine, op. cit., p. 78. 8. Quine, op. cit., p. 26. 9. Quine, op. cit., p. 87. 10. See C. Lewy,"G. E. Mooreon the NaturalisticFallacy" in P. F. Strawson(ed.), Studiesin the Philosophyof Thoughtand Action, Oxford University Press, Oxford, 1968, p. 137.

- PROCEEDINGSAND ADDRESSES OF THEAPA, 70:2 - 55