).

\.
,\ .
.. ;
;
-1j
.J?' ..,
. I
Naturalism; Or, Living Within One's Means
w.v. QVINB
tor Henri Lauener's sixtieth birthday
Abstract
Naturalismholds that there is no JUsher accessto truth thaD empiricaDy testable hypo-
theses. Still it does not repudiate untestablo hypotheses. '!bey fill out iDtendces of1beo1yand
lead tQ furt:helo that uetestable.
Ahypothcisis J8 testedbydccfuc:ing. fromit and abaCkground ofacceptedtheory, some ob-
servationcategorical that dOes not folfOwfromthebackgrounda1oDo. 'lbiscategorical, agener-
aIiZedconditiODal compounded of twoobservation 1SCDteDces, admits intumof apriDdtiveex-
perimental test. . ' .
1'0 sentoDcesthemselves,likoapecriesandbirdcalls, ueinhoJopbrasticUlO-
ciadoD With ranau of noural Jntab. DoqOtadOD of determfnate objects figures neither fa thIa
a.uocfation Doria deducing the eategerical fromthe scientifichypotheses. HeDce tho fndeter.
miJIacy of Ieferaace; is puzeJ.yauxiIiaIy to tho st:l1ICtlml Ofthooxy. 'Iiutb. however, is
seenstiIlutranscendollt at leastiDthis ofasuporsodcdscientifictheorynot that it
ceased to be true, but that it is found to have been false. . ....
Names ofphilosophical positionS are a necessary eviL They are necessary
because we need tQ refer to a stated position or doctrine from time to time,
andit wouldbetireSOme tokeeprestatingit. Theyareevil intbat theycometo
be conceivedas deSignatingschools ofthoupt, objects ofloyaltyfromwithin
and of obloquyfrom without, and hence obstacles, within and with-
out, to the pursuit of truth.
Inidentifyingthe philosophical positionthat I call naturalism, then, I shall
just be describing my oWli p(>sition, without prejudice to possibly divergent
uses ofthe term.ID Theories andThlngs I wrotethat is "therecog"; .
nition that it is science itself, andnot insome prior philosophy, that re-
ality is to be identified and described"; again that it is "abandonment of the
goal ofafirst philosophyprior tonatural science" (pp. 21, 67).
teriz8tions convey thO. right moQd, but they would fare Poorly in adebate.
Howmuch qualifies as "science itself" and not "some prior philosophy"?
Harvard University, USA
DiaIeciica . VoL 49, N 2-4 (1995)
252
W.V.Qaine
Naturalism; Or, Living Within One's Means
253 .
It1 science itself I certainly want to indude the farthest flights of physics
aDd cosmolo81l as well as experimental psychology, history, and the
.sciences. Also insofar at least as it is applied, for it is indispens-
. able tonatural sclence. What theaamIexcludingas "someprior philosOphy,"
and why? Descartes' dualism between miDd and body is called
but it could as weDbe reckoned as science, however false. He even had.a cau-
sal theory of the Interacdoft of mind and body through the pineal gland. If I
saw indirect explanatory benefit in positing sensibilia, possibilia,spirits, a
Creator, I wouldjoyfullyaccord theJn scientificstatus too, ona par with such
avowedlyscientificposits as quarks and blackholes. What then haveI banned
tinder the name of prior philosophy? .
Demarcationisnot mypurpose. Mypoint inthecharacterizations ofnatu-
ralismthat I quotedisjust that the mostwe canreaso:i1ablyseekinsupport of
.an inventory and description of reality is testability of its observable couse-
. quences in the time-honored hypothetico-deductive way - wher.eof m?re
anon. Naturalism need not cast asperSions on irrespoDSlole metaph)'Slcs,.
however deserved, much less onsoft sciences or 011 the speculative reaches of
the hard ones, except molar as a firmer basisisclaimedfor themthantheex"
perimental method itself.
naturalisticrenunciation shows itself most clearlyandsignificantly
is in naturalistic epistemology. VarioUs epistemologists, from Descartes to
Carnap, had'sought a foundation for in mental enti,ties, the
flux of rawsense data. It was as ifwe JDlght first fashion a and
infalliblelore ofsense data, innocent ofreferenCe tophysical things, and then
build out theory of the external worldsomehowon that finished foundation.
Thenaturalisticepistemologist dismisses thisdreamofpriorsense-datumlan-
guage, arguing that the positing of physical things is itself our indispensable
tool for organi?:ing and remembering what is otheJ;Wise, in James' words, a
"blooming, buzzing confusion."
Toaccount for knowledge ofan external thingorevent, accordingly, thenatu-
ralistic epistemologist looks rather to the external thing or event itself and the
causal chain of stimulation from it to one's brain. In a paradigm case, light rays
are reflected from the object to one's retina, activating a.patch of nerve
each of which initiates a neural impulse to one or another center of the bram.
Through intricateprocesses within the brain, finally, and abetted byimitation o.f
other peopleorbyinstruction, acbildcomesin to orassenttos?meru-
dimentarysentence at the endofsuch acausal cbain. I call It an observationsen-
tence. Examples are "It's cold"t "It's raining", "(I'hat's) milk", a) .
Customarilythe experlniental psychologist chooses one or another
or event, from somewhere along such a causal chain, to represent the chain,
. '.
. ,
.: r
.andthis he caDs thestimulus. Usuallyit is anevent ofhisown devising. In one
it will bea flashor a buzzinthe subject's vicinity, andinanother it
will bean Ice cube or a shock at the subject's surface. For our more general
purposes, not linkedto any particular an.economical in
defiDing the stimuIusis to inteIcept the causal chains jUst at the subject's sur-
face. Nothing Is lost, for it Is omy from that pOint Inwardthat the chains con-
tribute to the subject's knowledge of the external world.
.' Indeed evenwhat.reachesthesubject's surfaceisrelevant onlyifit triggers
, ..
neural receptOrs.' So we might for our purposes simply identify the
$timulus, over a given brief moment, with the temporally.orderedset ofsen-
Sory receptors triggered inthat moment . '. . "'.;.
Still further economy might be sought by mtercepting the causal chains
rather at a level- somewherewithinthebrain; for eventhesurfacere-
Ceptors that are triggered on any given occasion are largely without relevant
effect on thesubject's behavior. However, our knowledge of these deeper le-
vels Is still too sketchy. Moreover, as Jncreaslng1y penetrates these .
depths, we become aware of complexity and heterogeneity radicaUy a! va-
riancewiththeneatsimplicityat thesurface. Eachreceptor, afterall, admitsof
just two clean-cut states: triggered or no.t
.'. .Moreover, the behaviorally inelevant triggerings in aglobal can
be defined out anyway, in due course, by appeal to perceptual siJDilarity of
stimuli. The receptors whose firingis sa/ientin a given stimulus are the ones
that it shares with all perceptuallysimilar stimuli. Perceptual itself
'can be measured, for a given individual, .by reinforcement andextincdon of
responses. th b---' "":-ul
Soitseemsbest for present purposestoconstrue esu .n....... usona
giVet1 occasionsimplyas his global intake onthat o,ccaswn. I shaD
refer toit onlyas neural intake, not stimulus, for other notions ofstimulus are
wanted in other studies, particularly where subjects are to get the
. same stimulus. Neural intake is private, for subjects do not share receptors.
Perceptual similarity, then, is a relation between a subject's neural intakes.
Though testable, it is aprivate affair; the intakes are his, and
moreor less similarfor him. Perceptual similarityisthebasisofallleaming, all
habit formation, all expectationbyinductionfrompast experience; forwe are
innately disposed to expect simDar events to have sequels that are similar to
each other.
The association of observation with neural intakes is many-
many. Any one of a iange of perceptually fairly similar intakes may prompt
tho subject's assent to anyone of a range of semanticallykindred sentences.
But incontrast to the privacy ofneural intakes, and the privacy of theJr

sUniIarity, observationseiltelices andtheir sentaDti<:s area publicmat

ter, since, the child has to learn these from her elders. Her learning then de-
pends indeedbothonthepubliccurrencyoftheobservationsentences and on
a preestabDshed harmony of people's priYate scalesof perceptual similarity.
Theharmonyisformal, inthis sense: ifaWitness finds the first ofthreescenes
less similar to the second than to the third; another witness is apt to do like-
wise. This approximate harmonyis preestabDshedina sharedgene pooLDif-
ferent people's feelings still might not match, whatever that might mean.
This much is a naturalistic analogue or counterpart of the traditional epi-
.stemologist's phenomenalisticfoundation in sense data. However, it pretends
'toplausibilityin psychology, in genetics, and even in prehistory. Observation
sentences haVetheir antecedentsinbirdcalls andinthesignal cries oftheapes.
BuDding on this naturalistic founda"tion, then, in parallel to the old epi-
stemologist's proposed construction ofscience on a foun,dation ofsense data,
the natUtaJist would venture a psychologically and historically plausible
sketch of theindiVidual's acquisitionof science and perhaps the evolution of
science down the ages, with an eye primarily to 'the logic of evidence. I will
spareyoumost ofthat, for I havegoneintoit in WordandObjectandbetter in
The Roots of Reference, Pursuit of Truth, and elsewhere. There are just a
couple of aspects that I want to remind you of. .
One is reification, or the positing of objects. Observation sentences com-
monly containwords that refer to objects when used inmature discourse, but
the infant first acquires sucha sentence onlyas a seaIIl1ess whole, conditioned
- like the signal cry of the ape - to an appropriate range of global neural in-
takes. But there is a harbinger of reification already in our innate propensity,
and that of other animals, to confer salience onthose components of a neural
intake that t:ta:nsmit cprporeal patches of the visual field. It is what Donald
Campbell caDs our innate reification of bodies, but I construe reification
rather in degrees., Special ways of compounding observation mark
farther stepsmthe reification and thejobis complete onlywhenthe
speaker bas mastered past and future tense and knows about 1;Jie UDSeQD but .
continuous tI:ans1ation of an identical body through space between
tions. Itis onlythenthat shemakes senseofabody'sbeingthesamebodyfrom
one observation to aIlother despite intervening changes in appearance.
At that point the reification of bodies is full fledged. Reification of less
conspicuous objects, notablyabstract ones suchas numbers and classes, takes
,farther explaiiJing, and admits ofit. A crucial step there, as I seeit, is mastery
of relative and pronouns. ' .. '':'....
This butgeoning language'of science is a direct extension of the faltering
of observation. Segments of carry over and
'become. - of them terms for objects. Convorso!y, SODteDcosleamed
later by synthesis fromasophisticatedvocabularycancome
toqualifyas observationsentencesas welL For, what I takeas definitiveofob-
servation sentenceS is just tbis pair of conditions: first, the speaker must be
disposed to assent to the sentence or dissent from it outright on making the
appropriate pbservation, irrespective ofhis interruptedline of thought ifany,
and second, the verdict must Command the agreement of anywitnesses from
the appropriate language This second requirement,intersubJec-
tivity, is needed inorder that the child be able,to learn observation,sentences
from his elders; and those sentences, some of them, arc his indispeIisable en,;
terillgwedgeinacqUiring cogQitive language. Intersubjectivityofobservation
sentences is likewiieessential at the other end, toassureobjectivityofscience.
The sharing of vocabulary by observation sentences aild sentences of
science was necessary not only for the emergence language; it is
necessaryalso as a Channel for the empirical testing of scientific hypotheses.
The primordial hypotheses are what I call observation c:ategoricals, com-
poundedofpairSofobservationsentences: thus "Whenit snows, it's cold". To
checksucha hypothesis experimentally, we contriveto put ourselvesina situ-
ation where the first component, "Ifs snowing", is observably fulfilled, and
tb;en we check'for fulfillment of the second. component. If it i$ fulfilled, the
categorical remains standinguntil further notice. Ifit is not fulfilled, the cate-
gorical is refuted once for all. " '., :. ,... :
I see this as the key to the empirical testing also of more sophisticated hy-
potheses. We conjointhe hypothesis in,questionto a set ofalreadypreviously
accepted statements, sufficient together to imply some observation categori-
cal that was not implied bythe previous set alone. Thenwe checktheobserva-
tion categorical.
The appeal to logical implicationhere presents no problem. Thobasiclaws
of logic ,are interoaJized in learning the use of the logical particles. For in-
stance, the childlearD$byobservationandparental correctionthat it ismisuse
ofthe conjunction"and" to affirman "aild" colDpoundaIldthen denyone of
thecomponents. The,childhas thusitemalizedonesimplelogicalimplieation,
namely that an "and" compound implies its components, on pain of simply
gettingawordwrong. CorreSpondinglyfor otherbasicimplications, uptoand
including the laws of quantifiers' and identity. Insofar I with Lauener in
recogniZing'analyticity.
.Scientists of course. do not trace all these,Jinks of implication from hypo-
to categorical. It would mean filling in an the logically
reqUisitesuppoitingstatements, most ofwhichare sofamiliar tohimor80 tri-
vial as togowithout saying. Inpractice, moreover, manytacit premisses often
2SS
Natul'll1l1ml Or, Living Within Ono" Moan.
',i
,
,c.,
".
'"
'J
'if
i'
W.V. Qulno 254
256 Naturalism: Or, Living Within One's MeaDs
257
express sta1istical trends or probabilities, which hewill takeinstride un-
less Unexpected results prompt himto reconsider. , ' '
" StiUthedeductionandcheckingofobservationcategoricaJs is theessence,
,surely, of the experimental method, the hypothetico-deduetive method, 'the
method, in Popper's 'Words, of conjedU!e and refutation. It brings out that
prediction of observable events is the ultimate test of scientific theory.
I sPeak of teSt, not The purpose ofscience is to be sought rather
in mteDectual outIOIity and toclmoJogy. In our prehistoric begfnnjnglJ, hoW"
eve-r, the purposo 01tho first gJimmerings ofscientific theory
prediction, insofar as purpose can be despiritualized into natural selection
and survivaJ value. 'Ibis takes us back to our innate sense or standard of per-
ceptual similarity, and the innate expectation that similars will have mutually
similar sequels. Inshort, primitive induction.
Prediction is'verbaIized expectation. Conditional expectation, when COl'-
rect, has survival value. Natural selectionhasaccordinglyfavoredinnatestan-
datds ofperceptUal similarity that have harmonized with trends in our envi-
ronment Natural science, finally, is conditional expectation hypertrophied.
I saidthat predictionis not themainpurposeofscience, but onlythe test. It
,is anegativetest'atthat, atest byrefutation. As afurther disavowal let meadd,
,contrary to positivism, that a sentence does not even need to be testable in
order toqualifyas a respectable sentence ofscience. Asentenceis testable, in
myliberal or holistic sense, if adding it to previouslyaccepted sentences clin"
chesanobservationcategorical that was not impliedbythose'previoussenten-
ces alone; but much good science is untestable even in this liberal sense. We
believe many things because theyfit in smoothlybyanalogy, or theysymme-
, triZeandSimplifythe overall design. Sureiymuchhistoryand social scienceis
of this sort, and some hard science. Moreover, such acceptations are not idle
fancy; their proliferation generates, every here and there, a hypothesis that
canindeedbetested. Surelythisis themajorsoun:eoftestablehypotheses and
the growth of science.
The naturalization of epistemology, as I have been sketching it, is both a
limitationand ah'beration. The oldquest for afoundation for natural science,
firmer than science itself, is abandoned: that much is the limitation. Thelibe-
ration is fiee access to the resources ofnaturaI science, without fear of circu-
larity. The natu,raUstic epistemologist settles for what he can learn about ,the
strategy, logic, and mechanics by which our elaborate theory of the physical
world isin,fact projected, or might be, or shouldbe, fromjust that amorPhous '
neural intake.
Is this'sort of tbin&still philosophy? NaturalismbriDp a salutaryblUI'I'iQg
of' such boundaries. NaturaHstic philosophy is continuous with Il8tural
, I
, .
It undertakes to cIarify, organize, andsimpBfythebroadest'andmost
basic:'Concepts, and to analyze scientific method and evidence within the
framework of science itself. The boundary between naturalistic philosophy
and the rest of science is just a vague matter of degree. ,'.' ,"' :'
Naturalism is naturally associated with physicalism, or materialism. I do
not equatethem, as witness myearlier remark on Cartesian dualism. I doem-
bracephysicali$mas ascientificpoBitioD, but I couldbe dissuaded ofit OD
turelJcJendfic grounds without beJng dissuaded of naturilJam. Quantumme-
chanics today, indeed, in its neoclassical or Copenhagen interpretation, has a
distinctly mentaIistic ring. ' I
My has evidently been boiling down to the claim that in our
pW'Suit of truth about the world we cannot do better than our traditional
scientific procedure, thehypothetico-dedUctive method., A rebuttal suggests
itselfhere: surelymathematicians. Theobvious defense against that rebuttal is
to say that IIlatb,ematiCal truths are not about ,the world. But this is not a
defense of my choosing. Inmyviewapplied mathematics is about the world.
Thus consider again a case where we are testing a scientific hypothesis by
conjoining it to some aIreadyaccepted statements anddeducing an observa-
tion'categorical. Likelyas not, some ofthosealreadyacceptedstatements are
purely mathematicaL This is howpure mathematics gets applied. Whatever
empirical content those already accepted statements cin claim, then, from
beingneeded implyingtheobservationcategorical, is,imbibedinparticular
by the mathematlcaI ones.
Thus itis that I amiodinedtoblurtheboundarybetweenmathematicsand
natural science, noless than the boundary between philosophy and natural
science. If it is that proved mathematical truths are not subject to
subsequent refutation, my answer is that we safeguard them by choOsing to
revoke non-mathematical statements instead, in cases where a set of
ments has been found conjointly to imply a false observation categoric8L
ReaSons can be adduced for doing so; but enough.
That open the vast prolifer8tions of mathematics that there is no
thought or proSPect ofapplying. I see these domains as integral our ove.raU
theoryofrealityonlyonsufferance: theyareexpressedinthesamesyntaxand
lexicon as applicable mathematics, and to exclude themas meaningless by ad
hocgerrymanderingofour syntaxwouldbethanldess at best. Soit is left to us
to trytoassess thesesentences alsoas true orfalse, awecareto. Manyareset-
tledbythe samelaws that settIe applicablemathematics. For therest, I would
settIe themas far as practicable byconsiderations ofeconomy, on a par'with
thedecisioDSwem.innaturallOioncowhontryingtoframoC!mpirJcalhypo-
theses worthy of testing.
258
.W.V.Quine
Naturalism; Or. Living Within One's Means
259
TraditiolJ81epistemologywas inpart nonnativeinintent. Naturalisticepi-
stemology, in contrast, is viewed by Henri Lauener and others as purely de-
sCriptive. I disagree. Just as traditional epistemology on its speculative side
gets naturalized into science, or next of kin, so on its nonnative side it gets
naturaliZed into technology, the teclmology of scientizing.
What might be offeredfirst of anas a normofiJatUtalizedepistemologyis
pte'dietioll ofobservationas atest ofahypothesis. I thinkoftbisas morethana
nonn: as the name ofthegame. Sciencecannot anbetested, andthesofter the
science the spatser the tests; but whenit istested, the test is prediction of ob-
servation. Moteo'\7er, naturalismhas no special claims ontheprinciple, which
is rather the crux of empiricism.
. What are moredistinctivelynaturaliSticand technological are nOrmJbased
onscientiilcfindings. Thus sciencehas prettywell established- subject to fu-
disestablishment, as always - that our information about distant events
andother people reaches us only throughimpact of mys and particles on our
sensoryreceptors. Anormative corollaryisthat weshouldbewaryofastrolo-
gers,palmists, and other soothsayers. Think twice about E.S.P;
. For a richer array of norms, vague in various degrees, we may look to the
heuristics of hypothesis: how to think up a hypothesis worth testing. This is
WbereCObSidetations of conservatismand simplicity come in, and, at a more
. teclmica11evel, probability theory and statistics. In pmctice tbosetechnica1
mattersspillover also, as I remarked, to complicate the hypothetico-deduc-
tive method itself.
I saidat thebeginningofthis paper that accordingto naturalismit iswithin
science itself and not soDie prior philosophy that reality is to bo idontified.
Farther along in a more narrowly scientific spirit, I speculated on how we
roundout our recognitionofobjects asobjects, bit bybit, withour acquisition
of1angU8ge and science. These matters can nowfor somemore broadly
osophical reflections. ..
. Let:us recall, to begin with, that the association of observation sentences
with neural intake is holophrastic. What objects the component words may
designatein other contexts isirrelevant totheassociation. Thisis obviouslyso
if tho observation sentence is to be acquired as a first step in language leam..
ing; but the associationisequallydirect holophrasticinits operationeven
ifthe sentencewas acquiredthroughsynthesis ofits words, andgained its im-
mediacy only through subsequent familiarization. .
the specifics of designation and denotation are not only indif
fetent tothe associationofobservationsentencei toneural intake; thoyarein..
... dlttetellt alSo to the ImpHcatlOl1 of obseMlioll categorlca1s by
ory. Itislogical implication; andlogic, unlikeset theoryandtherest
IDatics, responds to no traits of objects beyond sameness and difference. So
wcmust conclude that objects ofanysort :figure only as neutral nodes in the
structure of scientific theory, so far as empiiica1 evidence is concerned. We
can arbitrarilychangethevalues ofour variables, the designata ofour names,
and the denotata of our predicates without disturbing the evidence, so long
anyWay as the new objects are explicitly correlated one. to one with the old.
Such is the indeterminacy of refere,oce, as I have come to can it
Ai first it is perhaps alarming. We are left with no basis, it would seem, for
judging whether we are talking about famjJjar things or some arbitrary
proxies. The shock subsides, however, when we reflect on a homely example
ot two. Thus tbiDkofabodyinthescientific framework ofspaceand time. In-
sofar as you specify the precise sinuous filament of four-dimensional
timethattbebodytakes upinthe courseofits career, you havefixed theobject
. uniquely.We could go farther and identify the object, a chipmunk perhaps,
withits portionofspace-tiJne, thus sayingthatit is$yat its earlyendand big-
ger at its late end. The move is artificial, but actually it confers abit of econ-
omy, ifwe are going to have the space-time anyway. Subjective connotations
ofbrownness, swift and erratic movement, and the rest simply carry
over. Surely all matters of evidenceremajn undisturbed. We are evenpre-
pared to say that it was whut abody was all along, an appropriately filled-in
.portion ofspace-ime as over against empty ones.
Next we might identifyspace-timeregions intumwith the sets of quadru
pIes ofnumbers that determinetheminsome arbitrarilyadoptedframe ofea-
.ordinates. We cantransfer sensOry connotations nowto this abstractmatbo-
matical object, andstill thereisnoViolence toscientificeVidence. Tospeakin
tuitively, nothing really happened. .
Thuswecancometoterms somewhat withtheindeterminacyofreference,
as applied to bodies and other seDSlble substances, byjust letting the sensory
connQtations of the obserVation sentences cariy over from the old objects to
their proxies.
Inthecaseofabstract objectssuchas numbers, devoidofsensoryconnota-
tions, the indeterminacy ofreference is already familiar. It is seen in Frege's
so-called Caesar problem: tho DUmborfive may be Julius Caesar. We happily
use nU%Dbers without caring whether they be taken.according to .the Frege-
Russell constructionsor Ackermann's or von Neumann's. The point was dra-
matizedlongago F.P. Ramseywith his expedient of Ramseysentences, as
they have como to be called. Instoad.Qf Invoking tho.abstract objects spociB-
wbon QOi1ajg. ofUlelr propertiea are neodedinIIllfgument, the Ramsey
Just says that there areobjects with the propenies, and theninvokes
by variables without further identification. This expedient oDly
.
"Uv w.v. 'J\WIe
t'llUuraiWD; uri .L.I\'mg Within One'. MOIDJ
261
"
works for abstract objects, however, usedasauxiliarieshereandtherewithout
regard to whether theyremain the.same objects from ODe context to another.
indetermiDacy of reference can be seen again in its fuJi generality, as
DaVidsononceremarked, byancmnninationofThrski's cJ8ssical truthdefiDi-
tion.1f asentence comes out true under that definition, it continues to do so
when objects are reassigiledto its predicates in any one-to-one way. .
These reflections on ontology are a salutary that the ultimate
dataofsciencearelimitedtoour neural intake, andthat theverynotionofob-
ject, or abstract, is ofour ownmaking, along with the rest ofnatural
science and mathematics. It is our overwhelmingly ingenious apparatus for
systematizing, predictin& and partially controlling our intake, and we may
take pride. ,
This conventionalist yiewofontologyappeals, I expeCt, toHenri Lauener.
He in his pragIDatismeven settles for a plurality ofscientific specialties, each .
with its working ontology, and no dreaID of an overarcbing, unifying fact of
the matter. ,
Naturalism itself is noncommittal on.this question of unity of science.
NatutaJism just sees it as a question.within science. itself, 'albeit a question
more remotefrom observational checkpoints than the most speculative ques-
tions of the hard and soft sciences ordinarily so called. .
Naturalismcan still.respeet the drive" on the part ofsome of us, for a uni-
-tied, an-piuposeontology. The drive is typical ofthe scientific temper, and of
apiecewiththedrivefor that shapesscientifichypotheSesgenerally.
Physicalismis its!amniar manifestation, and physicalismisbOUnd tohavehad
. side effects in the framing of morc .pecial hypOtheses in various
branches ofscience; for physicalismputs aptemiUlil onhypotheses favorable
to closer integration with physics itself. We have here a conspicuous case of
what I touchedonearlier: scientifichypotheses which, thoughnot themselves
testable, help to elicit others that
In anyevent, we are nowseeing ontologyas more utterly a human option
.than we used' to. We are drawn to Lauener's pragmatism. Must we then con-
clude that true is beyond our ken? No, that would be to forsake natu-
ralism. Rather, the notion of realityis itself part of the apparatus; and sticks,
stones, atoms, quarks, numbers, and classes all are utterlyreal denizens of an
. ultimate real world, except insofar as our present science may prove false on
further testing. . , .
What then is natUralism's line on truth and talsity themselves? The tnith
ptedicate raises no problemin its normal dailyuse as an instrument of what J
have called semantic ascent. 'IBrski's disquotational account accommodates
it, solong as what are called true are sentences in our own language; an4 we
, .
,
., :
: ,
,'.
th
C11
the predicate to sentences of other languages that we accept as
of trutbs of our own. However, paradoxes arise when the truth
predicate is applied to sentences. that contain that very predicate or related
ones; so arecaI1ed upontorecognizerather ahierarchyoftruthpredicates,
each of which behaves properly onlyin application to sentences that do not
containthat predicateitselforhigher ones. It is aheirarchyofbetterandbetter
truthpredicates but nobest. Inpractice, eXcept incontextesuchas these phiJ-
osphica1 ones, occasions seldomarisefor venturingabove thefirst rung of the
ladder. 'liuth olfthe truth, would indeed be transceJident;
bringing it down into scientific theory of the world engenders paradOx. So
naturalism has no plac:e for that.
.Still,' our concept of truth strains at its naturalistic moorings in another
way. Wenaturalists saythat Scienceis thehighest path to truth, stiDwe do
not say that everythirig on which scientists agree is true. 'Nor do we saythat .
sometbip.g that was true became false when scientists changed their minds.
What we say is that they and we thought it was true, but it wasn't. We have
scientistspursuingtruth, not decreeingit,1i:uththus st8ndsforth as anideal of
purereason, inKanfs apt phrase, andtranscendentindeed. Onthis scoreI am
with Lauener. .
C.S. Peirce tried to naturalize truth by identifiying it with'the limit that
scientificprogress This depends on optiJnistic assumptions, but if
w,e reconstrue it as metaphor it does epitomize the scientists' persistent
give and take of conjecture and refutation. 1i:uth as goal remains the estab-
lished usage of the term, and I acquiesce in it as just a vivid metaphor for our
continued adjustment of our world picture to our neural intikc. Metaphor is
perhaps a handycate$oryinwhich toaccommodatetranscendental concepts,
from a naturalist point ofview. '
. .
Yol. 49, N 2-4 (199,)