Ridgeview Publishing Company

The Intrinsic Quality of Experience

Author(s): Gilbert Harman
Reviewed work(s):
Source: Philosophical Perspectives, Vol. 4, Action Theory and Philosophy of Mind (1990), pp.
31-52
Published by: Ridgeview Publishing Company
Stable URL: http://www.jstor.org/stable/2214186 .
Accessed: 17/03/2012 15:10

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

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 support@jstor.org.

Ridgeview Publishing Company is collaborating with JSTOR to digitize, preserve and extend access to
Philosophical Perspectives.

http://www.jstor.org
PhilosophicalPerspectives,4
ActionTheoryand Philosophyof Mind,1990

QUALITYOF EXPERIENCE*
THE INTRINSIC

GilbertHarman
PrincetonUniversity

ABSTRACT

Thereare threefamiliar and relatedarguments againstpsycho-

physicalfunctionalismandthecomputer modelofthemind.The
firstis thatwe are directlyaware of intrinsicfeaturesof our
experienceand arguesthatthereis no way to accountforthis
awarenessina functional view.The secondclaimsthata person
blindfrombirthcan knowall aboutthefunctional roleofvisual
experience withoutknowing whatitis liketo see somethingred.
Thethird cannotaccountforthepossibility
claimsthatfunctionalism
ofan inverted spectrum.Allthreearguments can be defusedby
oftheobjectofexperience
properties
distinguishing fromproperties
oftheexperience ofan object.

The Problem

Many philosophers, and artificialintelligence

psychologists,
researchersaccept a broadlyfunctionalistview of the relation
betweenmindandbody,forexample,viewing themindinthebody
as something likea computerin a robot,perhapswithmassively
(as inRumelhart
parallelprocessing and McClelland1986).Butthis
viewofthemindhasnotgoneunchallenged. Somephilosophers and
othersobjectstrenuouslythatfunctionalismmustinevitablyfailto
accountforthe mostimportant partof mentallife,namely,the
subjectivefeelofconsciousexperience.
The computermodelof mindrepresents one versionof func-
32 / Gilbert
Harman

tionalism,although itisnottheonlyversion. Initsmostgeneralform,

functionalism defines mentalstatesandprocessesbytheircausalor
functional relations toeachotherandtoperceptual inputs from the
worldoutsideandbehavioral outputsexpressed inaction.According
tofunctionalism, itisthefunctional relationsthatareimportant, not
the intrinsic qualitiesof the stuffin whichthese relationsare
instanced.Justas the same computerprogramscan be runon
differentcomputers madeoutofdifferent materials,so functionalism
allowsforthesame mentalstatesand eventsin beingswithvery
different physicalconstitutions, since the verysame functional
relationsmight be instantiatedinbeingswithverydifferent physical
makeups. According tofunctionalism,beliefs,
desires, thoughts,and
feelingsare notlimited to beingsthatare materially likeourselves.
Suchpsychological statesandeventsmight alsooccur,forexample,
in siliconbased beings,as longas the rightfunctional relations
obtained.
Functionalism can allowforthepossibility thatsomething about
siliconmakesit impossible fortherelevantrelations to obtainin
siliconbasedbeings,perhapsbecausetherelevant eventscouldnot
occurfastenoughinsilicon.Itisevenconceivable thattherelevant
functional relations mightobtainonlyin thesortofmaterialthat
makesup humanbrains(Thagard1986;Dennett1987,Chapter9).
Functionalism impliesthatinsucha case thematerial is important
onlybecauseitis neededfortherelevantfunctional relationsand
notbecauseofsomeothermoremysterious or magicalconnection
betweenthatsortofmatterand a certainsortofconsciousness.
Variousissuesarisewithin thegeneralfunctionalist approach.For
one thing, thereis a disputeabouthowto identify theinputsto a
functional system.Shouldinputsbe identified witheventsin the
externalenvironment (Harman1988)or shouldtheyinsteadbe
identifiedwitheventsthataremoreinternal suchas thestimulation
of an organism's sensoryorgans(Block1986)?Thereis also the
possibilityof disagreement as to how deterministic the relevant
functional relationshave to be. Do theyhave to be completely
deterministic,orcantheybe merely Ormight
probabilistic? theyeven
be simplynondeterministic, not even associatedwithdefinite
probabilities(Harman1973,pp. 51-53)?
I willnotbe concernedwiththeseissueshere.Instead,I will
concentrate onthedifferent andmorebasicissuethatI havealready
mentioned, namely,whetherthissortoffunctionalism, no matter
 QualityofExperience/ 33
The Intrinsic

howelaborated, can accountforthesubjective feelofexperience,

for"whatitis like"(Nagel1974)toundergothisorthatexperience.
Furthermore, I willnotconsiderthegeneralchallenge, "Howdoes
functionalismaccountforX?"forthisorthatX. NorwillI consider
negativearguments againstparticular analyses.I will
functionalist
insteadconsiderthreerelatedarguments thatpurport todemonstrate
thatfunctionalismcannotaccountforthisaspectofexperience. I will
arguethatall threearguments I willsaylittlethatis
are fallacious.
andwillforthemostpartmerely
original elaboratepointsmademany
yearsago (Quine1960,p. 235,Anscombe1965,Armstrong 1961,
1962,and especially1968,Pitcher1971),pointsthatI do notthink
havebeen properly The threearguments
appreciated. are these:
First,whenyou attendto a painin yourleg or to your
experienceoftherednessofan apple,you are awareofan
intrinsicqualityofyourexperience, wherean intrinsic
qualityis a qualitysomething apartfromits
has in itself,
relationsto otherthings. Thisqualityofexperiencecannot
be capturedin a functional definition,sincesucha definition
is concernedentirely withrelations, relationsbetween
mentalstatesand perceptual input,relations amongmental
states,and relationsbetweenmentalstatesand behavioral
output.Forexample,"Anessentialfeatureof [Armstrong's
functionalist]analysisis thatit tellsus nothing aboutthe
intrinsicnatureofmentalstates...He nevertakesseriously
thenaturalobjectionthatwe mustknowtheintrinsic nature
ofourown mentalstatessincewe experiencethemdirectly"
(Nagel1970).
Second,a personblindfrombirthcouldknowall about
thephysicaland functional factsofcolorperception without
knowingwhatit is liketo see something red.So, whatit is
liketo see something redcannotbe explicatedin purely
functional terms(Nagel1974,Jackson1982,1986).
Third,it is conceivablethattwopeopleshouldhave
similarly functioningvisualsystemsdespitethefactthat
thingsthatlookredto one personlookgreento theother,
thingsthatlookorangeto thefirst personlookblueto the
second,and so forth (Lycan1973,Shoemaker1982).This
sortofspectrum inversion in thewaythingslookis possible
butcannotbe givena purelyfunctional sinceby
description,
34 / Gilbert
Harman

hypothesisthereare no functional
differences
betweenthe
peoplein question.Sincetheway thingslookto a personis
an aspectofthatperson'smentallife,thismeansthatan
importantaspectofa person'smentallifecannotbe
explicatedin purelyfunctional
terms.

Intentionality

Inordertoassessthesearguments, I beginbyremarking on what

is sometimes calledtheintentionality ofexperience. Ourexperience
oftheworldhas content-that is,itrepresents things as beingina
certainway. In particular, perceptualexperiencerepresents a
perceiveras in a particular environment, forexample,as facinga
treewithbrownbarkandgreenleavesfluttering ina slight breeze.
One thingthatphilosophers meanwhentheyreferto thisas the
intentionalcontent ofexperience isthatthecontent oftheexperience
maynotreflect whatis reallythere.Although itlooksto me as if
I amseeinga tree,thatmaybe a cleverillusion produced withtilted
mirrors andpainted backdrops. Oritmaybe a hallucination produced
by a drugin mycoffee.
Thereare manyotherexamplesofintentionality. Poncede Leon
searchedFloridafortheFountainofYouth.Whathe was looking
forwasa fountain whosewaters wouldgiveeternal youth towhoever
woulddrinkthem.In fact,thereis no suchthingas a Fountainof
Youth,butthatdoes notmeanPoncede Leonwasn'tlookingfor
anything. He was lookingforsomething. We cantherefore saythat
hissearchhadan intentional object.Butthething thathewaslooking
for,theintentional objectofhissearch,didnot(anddoesnot)exist.
A paintingofa unicorn is a painting ofsomething; ithasa certain
content. Butthecontent doesnotcorrespond toanything actual;the
thingthatthepainting represents does notexist.The painting has
an intentional contentin therelevantsenseof "intentional."
Imaginingormentally picturing a unicorn isusefully compared with
a painting ofa unicorn. In bothcasesthecontent is notactual;the
objectpictured, theintentional objectofthepicturing, doesnotexist.
It is onlyan intentional object.
Thisis notto supposethatmentally picturinga unicorninvolves
an awarenessof a mentalpictureof a unicorn.I am comparing
mentally picturingsomething witha picture ofsomething, notwith
 The Intrinsic
QualityofExperience/ 35

a perception of a picture.An awarenessof a picturehas as its

intentional objecta picture. Thepicturehasas itsintentional object
a unicorn. Imagining a unicorn isdifferentfrom imagining a picture
ofa unicorn. The intentional objectoftheimagining is a unicorn,
nota pictureofa unicorn.
It is veryimportant to distinguish betweenthe properties of a
represented objectand theproperties of a representation of that
object.Clearly, theseproperties can be verydifferent. Theunicorn
ispictured as havingfourlegsanda singlehorn.Thepainting ofthe
unicorndoes nothavefourlegsand a singlehorn.The painting is
flatand coveredwithpaint.The unicornis notpictured as flator
coveredwithpaint.Similarly, an imaginedunicornis imaginedas
havinglegsand a horn.The imagining oftheunicornhas no legs
or horn.The imagining of theunicornis a mentalactivity. The
unicornis notimaginedas eitheran activity or anything mental.
The notorious sensedatumtheoryofperception arisesthrough
failing to keeptheseelementary pointsstraight. According to that
ancienttheory, perception ofexternalobjectsin theenvironment
is alwaysindirect and mediatedby a moredirectawarenessofa
mentalsensedatum.Defenders ofthesensedatumtheory arguefor
itbyappealing totheso-called argument from illusion.
Thisargument
beginswiththe uncontroversial premisethatthe way thingsare
presented inperception is notalwaysthewaytheyare.Eloisesees
somebrownandgreen.Butthereisnothing brownandgreenbefore
her;it is all an illusionor hallucination. Fromthistheargument
fallaciously thatthebrownandgreenEloiseseesisnotexternal
infers
toherandso mustbe internal ormental. Sinceveridical, nonillusory,
nonhallucinatory perception can be qualitatively indistinguishable
fromillusory or hallucinatory perception, theargument concludes
thatinall casesofperception Eloiseis directly awareofsomething
innerand mentaland onlyindirectly awareofexternalobjectslike
treesand leaves.
An analogousargument aboutpaintings wouldstartfromthe
premisethata painting can be a painting ofa unicorn eventhough
thereare no unicorns. Fromthisit mightbe concludedthatthe
painting is "inthefirst instance" a painting ofsomething else that
is actual,forexample,thepainter's idea ofa unicorn.
In orderto see thatsucharguments are fallacious,considerthe
corresponding argument appliedto searches:"Poncede Leonwas
searching fortheFountain ofYouth.Butthereis no suchthing.So
36 / Gilbert
Harman

he musthavebeensearching forsomething mental."Thisis justa

mistake.Fromthefactthatthereis no FountainofYouth,itdoes
notfollowthatPoncede Leonwassearching forsomething mental.
Inparticular, hewasnotlooking foran ideaoftheFountain ofYouth.
He alreadyhad theidea.Whathe wantedwas a realFountainof
Youth,notjusttheidea ofsucha thing.
Thepainter haspainteda picture ofa unicorn. Thepicture painted
is nota pictureofan idea ofa unicorn. The paintermightbe at a
losstopainta picture ofan idea,especially ifhe is notfamiliar with
conceptual art.Itmaybe thatthepainterhasan idea ofa unicorn
and triesto capturethatidea in hispainting. Butthatis to say his
painting is a painting ofthesamethingthathisidea is an idea of.
Thepainting is nota paintingoftheidea,buta painting ofwhatthe
idea is about.
In thesameway,whatEloisesees beforeheris a tree,whether
or notitis a hallucination. Thatis tosay,thecontent ofhervisual
experienceis thatshe is presented witha tree,notwithan idea of
a tree.Perhaps,Eloise'svisualexperienceinvolvessome sortof
mental pictureoftheenvironment. Itdoesnotfollow thatsheisaware
ofa mentalpicture. Ifthereis a mentalpicture, itmaybe thatwhat
she is awareofis whateveris represented by thatmentalpicture;
butthenthatmentalpicture represents something intheworld,not
something in themind.
Now,we sometimes countsomeoneas perceiving something only
ifthatthingexists.So, ifthereis no treebeforeherand Eloiseis
sufferingfrom a hallucination,
we might describe thiseitherbysaying
thatEloiseseessomething thatis notreallythereorbysayingthat
she does not reallysee anything at all but onlyseems to see
something. Thereis nota use of"searchfor"corresponding to this
seconduse of"see"thatwouldallowus to say that,becausethere
was and is no suchthingas theFountainofYouth,Poncede Leon
was notreallysearching foranything at all.
Butthisambiguity in perceptual verbsdoes notaffect thepoint
I am trying to make.To see thatitdoesnot,letus use "seet"("see-
dagger")forthesenseof"see" inwhichtheobjectseen mightnot
exist,as whenMacbethsaw a daggerbeforehim.'Andletus use
itsee*"("see-star")forthesenseof"see" inwhichonlythingsthat
existcan be seen. Macbethsawt a daggerbuthe did notsee* a
dagger.
Theargument from illusion
startsfrom a caseinwhichEloise"sees"
 QualityofExperience/ 37
The Intrinsic

something brownand greenbeforeher,although thereis nothing

brownand greenbeforeherintheexternalphysicalworld.From
this,theargument infersthatthebrownand greenshe sees must
andmental.
be internal Now,if"see"is "seet"here,thisisthefallacy
alreadynoted,like thatof concluding thatPonce de Leon was
searchingforsomethingmental from thefactthatthereisnoFountain
ofYouthintheexternal world.On theotherhand,if"see"is "see"
here,thenthepremiseoftheargument simplybegsthequestion.
Noreasonatall hasso farbeengivenfortheclaimthatEloisesees*
something brownandgreeninthiscase.Itistruethatherperceptual
experiencerepresents her as visuallypresentedwithsomething
brownandgreen;butthatistosaymerely thatsheseestsomething
brownand green,notthatshesees*anything at all. (Fromnowon
I willsuppressthet and * modificationofperceptual verbsunless
indicationofwhichsenseis meantis crucialto thediscussion.)
Here,some philosophers (e.g. Jackson1977) wouldobjectas
follows:
You agreethatthereis a sensein whichEloisesees
something greenand brownwhenthereis nothing green
and brownbeforeherin theexternalworld.You are able to
denythatthisbrownand greenthingis mentalby takingit
to be a nonexistent object.Butit is
and merelyintentional
surelymorereasonableto supposethatone is in thiscase
awareofsomething mentalthanto supposethatone is
awareofsomething thatdoes notexist.How can therebe
anything thatdoes notexist?The verysuggestionis a
in terms,since"be" simplymeans"exist,"so
contradiction
thatyou are reallysayingthatthereexistssomething that
does notexist(Quine1948).Thereare no suchthingsas
nonexistentobjects!
In reply,letme concedeimmediately thatI do nothave a well
workedouttheory ofintentionalobjects.Parsons(1980)offersone
suchtheory, althoughI do notmeantoexpressan opinionas tothe
successofParson'sapproach.Indeed,I am quitewilling to believe
thatthereare notreallyanynonexistent objectsandthatapparent
talkofsuchobjectsshouldbe analyzedawaysomehow.I do notsee
thatitis myjob toresolvethisissue.Howeverthisissueis resolved,
thetheory thatresults
hadbetter endupagreeing thatPoncede Leon
was lookingforsomething whenhe was lookingfortheFountain
 Harman
38 / Gilbert

ofYouth,eventhough thereis noFountain ofYouth,andthetheory

hadbetter nothavetheconsequence thatPoncede Leonwaslooking
forsomething mental.Ifa logicaltheorycan accountforsearches
forthingsthatdo not,as it happens,exist,itcan presumably also
allowfora senseof"see"inwhichMacbeth can see something that
does notreallyexist.
Another pointisthatEloise'svisualexperience doesnotjustpresent
a tree.It presentsa treeas viewedfroma certainplace.Various
featuresthatthetreeispresented as having arepresented as relations
betweentheviewerandthetree,forexample,features thetreehas
fromhere.Thetreeispresented as "infrontof"and "hiding" certain
othertrees.It is presented as fulleron "theright."It is presented
as thesamesize "fromhere"as a closersmallertree,whichis not
to saythatitreallylooksthesameinsize,onlythatitis presented
as subtending roughly thesameanglefrom hereas thesmallertree.
To be presented as thesameinsizefromhereis nottobe presented
as thesame in size,period.
I do notmeantosuggest thatthewaythetreeisvisually presented
as beingfromhereis something thatis easilyexpressedinwords.
In particular,I do notmeanto suggestthatthetreecan thusbe
presented as subtending a certainvisualangleonlytosomeonewho
understands wordslike"subtend"and "angle"(as is assumedin
Peacocke1983,Chapter1). I meanonlythatthisfeature ofa tree
fromhereis an objectivefeature ofthetreein relationto here,a
featureto whichperceivers are sensitiveand whichtheirvisual
experiencecan somehowrepresent thingsas havingfromhere.
Now,perhaps,Eloise'svisualexperienceevenpresents a treeas
seenbyher,thatis,as an objectofhervisualexperience. Ifso,there
is a senseafterall in whichEloise'svisualexperiencerepresents
something mental:itrepresents objectsin theworldas objectsof
visual experience.But thisdoes not mean that Eloise'svisual
experience inanywayrevealsto hertheintrinsic properties ofthat
experienceby virtueofwhichit has thecontentit has.
I wantto stressthispoint,becauseitis veryimportant. Eloiseis
awareofthetreeas a treethatsheisnowseeing.So,we cansuppose
she is awareofsomefeatures ofhercurrent visualexperience.In
particular,she is awarethathervisualexperiencehasthefeature
ofbeingan experienceofseeinga tree.Thatis to be awareofan
intentional featureof her experience;she is aware that her
experience hasa certaincontent. Ontheotherhand,I wanttoargue
 / 39
QualityofExperience
TheIntrinsic

thatshe is notawareofthoseintrinsic features of herexperience

byvirtueofwhichithasthatcontent. Indeed,I believethatshehas
noaccessatall totheintrinsic features ofhermentalrepresentation
thatmakeita mentalrepresentation ofseeinga tree.
Thingsaredifferent withpaintings.Inthecaseofa painting Eloise
can be awareofthosefeatures ofthepainting thatare responsible
foritsbeinga painting ofa unicorn. Thatis,shecanturnherattention
to thepatternof thepainton thecanvasby virtueofwhichthe
painting representsa unicorn.Butinthecaseofhervisualexperience
ofa tree,I wanttosaythatsheisnotawareof,as itwere,themental
paintbyvirtueofwhichherexperienceis an experience ofseeing
a tree.She is awareonlyoftheintentional or relational featuresof
herexperience, notofitsintrinsic nonintentional features.
Somesensedatumtheorists willobjectthatEloiseis indeedaware
oftherelevant mentalpaintwhenshe is awareofan arrangement
ofcolor,becausethesesensedatumtheorists assertthatthecolor
she is awareofis innerand mentaland nota property ofexternal
objects.But,thissensedatumclaimis counterto ordinary visual
experience.WhenEloise sees a treebeforeher,the colorsshe
experiencesare all experiencedas featuresof the treeand its
surroundings. Noneofthemareexperienced as intrinsicfeaturesof
herexperience. Nordoes she experienceanyfeatures ofanything
as intrinsicfeaturesofherexperience. Andthatis trueofyoutoo.
Thereis nothing specialaboutEloise'svisualexperience. Whenyou
see a tree,youdo notexperience anyfeatures as intrinsicfeatures
ofyourexperience.Lookat a treeand tryto turnyourattention
tointrinsic featuresofyourvisualexperience. I predictyouwillfind
thattheonlyfeatures theretoturnyourattention towillbe features
ofthepresented tree,including relationalfeatures ofthetree"from
here."
Thesensedatumtheorists' viewaboutourimmediate experience
ofcoloris definitely notthenaiveview;itdoes notrepresent the
viewpoint ofordinary perception.Thesensedatumtheory isnotthe
resultof phenomenological study;it is ratherthe resultof an
argument, namely,theargument fromillusion.Butthatargument
is eitherinvalidor question-begging, as we haveseen.
Itisveryimportant todistinguishwhatareexperienced as intrinsic
features oftheintentional objectofexperience from intrinsic
features
of theexperienceitself.It is notalwayseasy to distinguish these
things, buttheycan be distinguished. Considertheexperienceof
40 / Gilbert
Harman

havinga paininyourright leg.Itisverytempting toconfuse features

of whatyou experienceas happeningin yourleg withintrinsic
features ofyourexperience. Butthehappening inyourlegthatyou
are presented withis theintentional objectofyourexperience; it
is nottheexperienceitself. The contentofyourexperience is that
thereisa disturbance ofa certain specific sortinyourright leg.The
intentionalobjectoftheexperience isan eventlocatedinyourright
leg. The experienceitselfis notlocatedin yourrightleg. Ifthe
experienceis anywhere specific,itis somewhere in yourbrain.
Noticethatthecontent ofyourexperience maynotbe truetowhat
isactuallyhappening. A slippeddiscinyourbackmaypressagainst
yoursciaticnervemakingitappearthatthereis a disturbance in
yourright legwhentherereallyisnot.Theintentional objectofyour
painful experience maynotexist.Ofcourse,thatis nottosaythere
is no paininyourleg.You do feelsomething there.Butthereis a
senseinwhichwhatyoufeelinyourlegisan illusion orhallucination.
Itis truethat,ifMelvinhallucinates a pinkelephant, theelephant
thatMelvinsees does notexist.Butthepaininyourleg resulting
froma slippeddiscin yourbackcertainly does exist.2The painis
notan intentional objectinquitethewaytheelephantis.The pain
inyourlegcausedbytheslippeddiscinyourbackis morelikethe
afterimage ofa bright light.Ifyoulookat a blankwall,yousee the
imageon thewall.The imageison thewall,thepainisinyourleg.
Thereisnophysical spotonthewall,thereisnophysical disturbance
inyourleg.Theafterimage exists,thepainexists. Whenwetalkabout
afterimages or referredpains,someof whatwe say is aboutour
experience andsomeofwhatwe sayis abouttheintentional object
ofthatexperience. Whenwe saythepainor afterimage exists,we
meanthattheexperience exists.Whenwe say thattheafterimage
is on thewallor thatthepainis in yourleg,we are talking about
thelocationoftheintentional objectofthatexperience.

Assessmentof the FirstObjection

We arenowina position ofthethreearguments

torejectthefirst
againstfunctionalism
whichI nowrepeat:
Whenyou attendto a painin yourleg or to your
experienceoftherednessofan apple,youare awareofan
intrinsic
qualityofyourexperience,wherean intrinsic
 QualityofExperience
TheIntrinsic / 41

qualityis a qualitysomething apartfromits

has in itself,
relationsto otherthings.Thisqualityofexperiencecannot
be capturedin a functional sincesucha definition
definition,
is concernedentirely withrelations,
relations between
mentalstatesand perceptual input,relations amongmental
states,and relationsbetweenmentalstatesand behavioral
output.
We can now see thatthisargument failsthrough confounding a
qualityoftheintentional objectofan experiencewitha qualityof
theexperienceitself. Whenyouattendto a painin yourleg or to
yourexperienceoftherednessofan apple,youare attending to a
qualityofan occurrenceinyourlegora quality oftheapple.Perhaps
thisqualityis presentedto you as an intrinsic qualityof the
occurrence inyourlegoras an intrinsic qualityofthesurface ofthe
apple.Butitis notat all presented as an intrinsic qualityofyour
experience. And,sinceyouare notawareoftheintrinsic character
ofyourexperience, thefactthatfunctionalism abstracts fromthe
intrinsic
characterofexperience doesnotshowitleavesoutanything
you are awareof.
To be sure,thereare possiblecomplications. SupposeDavid
undergoes brainsurgery whichhewatchesina mirror. Supposethat
features
he sees certainintrinsic ofthefiring ofcertainneuronsin
hisbrainandsupposethatthefiring oftheseneurons istherealization
ofpartoftheexperience he is havingat thatmoment. Inthatcase,
Davidis awareofintrinsic features ofhisexperience. Butthatway
ofbeingawareofintrinsic features ofexperience isnotincompatible
with functionalism. Given a functionalist account of David's
perception oftrees,tables,andthebrainprocessesofotherpeople,
thesameaccountapplieswhentheobjectperceivedhappenstobe
David'sownbrainprocesses.TheawarenessDavidhasofhisown
brainprocesses ispsychologicallysimilartotheawareness anyother
sightedperceiver mighthaveofthosesamebrainprocesses, including
perceiversconstructedina verydifferent wayfrom thewayinwhich
David is constructed.
According tofunctionalism,thepsychologicallyrelevant properties
of an internalprocessare all functional properties. The intrinsic
natureoftheprocessis relevantonlyinasmuch as itis responsible
fortheprocess'shavingthefunctional properties ithas.I havebeen
consideringtheobjectionthatcertain features
intrinsic ofexperience
42 / Gilbert
Harman

must be psychologically relevantpropertiesapart fromtheir

contribution to function,sincetheseare properties we are or can
be awareof.The objectionis notjustthatwe can becomeaware
of intrinsic featuresof certainmentalprocessesin the way just
mentioned, thatis,byperceiving ina mirror theunderlying physical
processesthatrealizethosementalprocesses. Thatwouldnotbe an
objectionto functionalism. The objectionis ratherthatall or most
consciousexperiencehas intrinsic aspectsofwhichwe are or can
be aware in sucha waythattheseaspectsoftheexperienceare
psychologically significantover and above thecontribution they
maketo function.
Ofcourse,tosaythattheseaspectsarepsychologically significant
is nottoclaimthattheyareoroughttobe significant forthescience
of psychology. Rather,theyare supposedto be psychologically
significant in thesenseofmentally significant,
whether or notthis
aspectofexperienceis susceptible ofscientific
understanding. The
objectionis thatanyaccountofourmentallifethatdoes notcount
theseintrinsic as mental
properties orpsychologicalpropertiesleaves
out a crucialaspectofourexperience.
Myreplyto thisobjectionis thatitcannotbe defendedwithout
confusing intrinsic featuresoftheintentional objectofexperience
withintrinsic featuresoftheexperience. Apartfrom thatconfusion,
there is no reasonto thinkthatwe are everawareoftherelevant
intrinsic features ofourexperiences.
Thereare otherwaysin whichone mightbe awareofintrinsic
featuresof our experiencewithoutthatcastingany doubton
functionalism. Forexample,onemight be awareofintrinsicfeatures
ofexperiencewithout being aware of themas intrinsic
features of
experience, justas Ortcuttcanbe awareofa manwho,as ithappens,
is a spywithout beingawareofthemanas a spy.WhenEloisesees
a tree,sheis awareofherperceptual experienceas an experience
witha certainintentional content.Supposethatherexperienceis
realizedby a particular physicaleventand thatcertainintrinsic
features oftheeventareinthiscaseresponsible forcertainintentional
features ofEloise'sexperience. Perhaps thereisthena senseinwhich
Eloiseis awareofthisphysical processandawareofthoseintrinsic
features, although sheis notawareofthemas theintrinsic features
thattheyare.
Evenifthatis so, itis no objection to functionalism.
Theintrinsic
featuresthat Eloise is aware of in that case are no more
 / 43
QualityofExperience
The Intrinsic

psychologically significantthanis theproperty ofbeinga spyto

Ortcutt'sperception ofa manwhohappensto be a spy.The case
givesno reasonto thinkthatthereis a psychologically significant
difference betweenEloise'sexperienceand theexperienceofany
functionalduplicateofEloisethatismadeofdifferent stufffromwhat
Eloiseis madeof.
Similarly,ifEloiseundertakes thesortofeducation recommended
by Paul Churchland (1985)so thatshe automatically thinksofthe
intentionalaspects of her experience in terms of their
neurophysiological causes,thenshe may be aware of intrinsic
features of herexperienceas theveryfeatures thattheyare. But
againthatwouldbe no objectiontofunctionalism, sinceitgivesno
reasontothink thatthereisa psychologicaldifferencebetween Eloise
aftersuchtraining and a robotwhois Eloise'sfunctional duplicate
and who has been givensimilartraining (Shoemaker1985).The
duplicate nowwrongly thinksofcertainaspectsofitsexperience as
certain featuresofcertainneurological processes-wrongly, because
therelevant processes intheduplicate arenotneurological processes
at all.
Observe,by theway,thatI am notoffering anysortofpositive
argument thatEloiseandherduplicate musthaveexperiences that
are psychologically I am onlyobserving
similarinall respects. that
thecasesjustconsidered arecompatible withthefunctionalistclaim
thattheirexperiences are similar.
Theobjections tofunctionalism thatI amconsidering inthispaper
claimthatcertainintrinsic properties ofexperienceso inform the
experiencethatanyexperiencewithdifferent intrinsicproperties
wouldhavea different psychologicalcharacter.WhatI haveargued
so faris thatthisobjectionis notestablished bysimpleinspection
ofourexperience.

Perceptionand Understanding

whichI repeat:
Now,let me turnto thesecondobjection,
A personblindfrombirthcouldknowall aboutthe
factsofcolorperception
physicaland functional without
red.So, whatit is
knowingwhatit is liketo see something
liketo see somethingredcannotbe explicatedin purely
functionalterms.
44 / Gilbert
Harman

In ordertoaddressthisobjection, I haveto saysomething about

thefunctionalisttheoryofthecontent ofmentalrepresentations and,
more particularly, something about the functionalist theoryof
concepts.I have to do thisbecauseto knowwhatit is liketo see
something redistobe capableofrepresenting toyourself something's
beingred.You can represent thatto yourself onlyifyouhavethe
relevantconceptofwhatit is forsomething to be red.The blind
personlacksthefullconceptofrednessthata sightedpersonhas;
so theblindpersoncannotfullyrepresent whatit is fora sighted
persontosee something red.Therefore, theblindpersoncannotbe
said to knowwhatit is liketo see something red.
Onekindoffunctionalist accountofmental representation supposes
thatmentalrepresentations are constructed fromconcepts, where
the contentof a representation is determined by theconceptsit
containsand thewaytheseconceptsare puttogether to formthat
representation(Harman1987).Inthisview,whatitistohavea given
conceptis functionally determined. Someonehas theappropriate
conceptof something's beingred if and onlyif the personhas
availablea conceptthatfunctions in the appropriate way. The
relevant functioningmayinvolveconnections withtheuse ofother
concepts,connections to perceptual input,and/orconnections to
behavioraloutput.In thiscase,connections to perceptual inputare
crucial.Iftheconceptis to function in sucha waythattheperson
has thefullconceptofsomething's beingred,thepersonmustbe
disposed toform representationsinvolving thatconceptas thenatural
andimmediate consequence ofseeingsomething red.Sincetheblind
personlacksanyconceptofthissort,theblindpersonlacksthefull
conceptofsomething's beingred.Therefore, theblindpersondoes
notknowwhatit is liketo see something red.
Itis noteasyto specify therelevant functional relationprecisely.
Someonewhogoesblindlaterinlifewillnormally retaintherelevant
conceptofsomething's beingred.Sucha personhasa conceptthat
he or she wouldbe able to use in forming suchimmediate visual
representations exceptforthecondition thatinterferesinhisor her
case withnormalvisualperception. So,theright functional relation
holdsforsucha person.I am supposing thatthepersonblindfrom
birthhas no suchconcept;thatis, thepersonhas no conceptof
something's beingredthatcouldbe immediately brought intoservice
in visualrepresentations of the environment ifthe personwere
suddenlyto acquiresight.
 The Intrinsic
QualityofExperience/ 45

We arenowina position toassesstheclaimthatthepersonblind

from birth
couldknowallthephysical andfunctional factsaboutcolor
perception without knowing whatitis liketo see something red.I
claim thatthereis one important functional factabout color
perception thattheblindpersoncannotknow,namely,thatthere
is a conceptR suchthatwhena normalperceiver sees something
redingoodlighting conditions,theperceiver hasvisualexperience
witha representational structure containing thisconceptR. The
personblindfrombirthdoes notknowthatfact,becausein order
toknowitthepersonneedstobe able torepresent thatfactto him
or herself, whichrequireshavingthe relevantconcepts.A key
conceptneededtorepresent thatfactis theconceptofsomething's
beingred,becausethefactinquestionis a factaboutwhathappens
whena normal perceiverseessomething red.Sincethepersonblind
frombirthdoesnothavethefullconceptofsomething's beingred,
thepersoncannotfully understand thatfactandso cannotknowthat
fact.
The blindpersonmightknowsomething resembling this,for
example, thatthereisa concept R suchthat,whena normal perceiver
sees something thatreflectslightofsuchandsucha frequency, the
perceiverhas visualexperiencewitha representational structure
containing thisconceptR. Butthatis to knowsomething different.
The personblindfrombirthfailsto knowwhatit is liketo see
something redbecausehe or she doesnotfully understand whatit
is forsomething tobe red,thatis,becausehe orshedoesnothave
thefullconceptofsomething's beingred.So, contrary to whatis
assumedinthesecondobjection, thepersonblindfrombirthdoes
notknowallthefunctional sinceheorshedoesnotknowhow
facts,
theconceptR functions withrespecttotheperception ofthingsthat
are red.
Thisresponsetothesecondobjection appealsto a functionalism
thatrefersto thefunctions ofconcepts, notjustto thefunctions of
overallmentalstates.Thereareotherversions offunctionalism that
tryto makedo withreferences to thefunctions ofoverallmental
states,without to
appeal concepts. Some ofthese versionsidentify
thecontents ofsuchstateswithsetsofpossibleworlds(orcentered
possibleworlds).Theseversions offunctionalism cannotrespondto
theobjection in thewaythatI haveresponded. Itis unclearto me
whetherany satisfactory responseis possibleon behalfof such
theories.Forexample,Lewis(1983)is forcedto say thatalthough
 Harman
46 / Gilbert

thepersonblindfrombirthlackscertainskills,e.g.,theabilityto
redobjectsjustbylooking
recognize at theminthewaythatsighted
aboutvisualperception.
peoplecan,thispersonlacksno information
I amnothappywiththatresponse, sinceitisclearlyfalsetosaythat
thepersonblindfrombirthdoes notlackanyinformation.

InvertedSpectrum

whichI repeat:
tofunctionalism,
I nowturntothethirdobjection
It is conceivablethattwopeopleshouldhavesimilarly
functioning visualsystemsdespitethefactthatthingsthat
look red to one personlook greento theother,thingsthat
look orangeto thefirst personlookblue to thesecond,and
so forth. Thissortofspectrum in theway things
inversion
lookis possiblebutcannotbe givena purelyfunctional
description, thereare no functional
sinceby hypothesis
differences betweenthepeoplein question.Sincetheway
thingslookto a personis an aspectofthatperson'smental
life,thismeansthatthereis an important aspectofa
person'smentallifethatcannotbe explicatedin purely
functional terms.
In orderto discussthisobjection, I need to saysomething more
abouthowperceptual statesfunction. In particular, I have to say
something abouthowperceptual statesfunction inrelation tobelief.
Perceptualexperience represents a particular environment ofthe
perceiver.Normally, a perceiver usesthisrepresentation as hisor
herrepresentationoftheenvironment. Thatis tosay,theperceiver
usesitinorderto negotiate thefurniture. In stillotherwords,this
isusedas theperceiver's
representation beliefabouttheenvironment.
Thissortofuse ofperceptual representations is thenormalcase,
although thereare exceptions whena perceiver inhibitshisor her
naturaltendency andrefrains from usinga perceptual representation
(or certainaspects of that representation) as a guide to the
environment, as a beliefaboutthesurroundings. The contentof
perceptualrepresentationis functionallydefined inpartbytheways
inwhichthisrepresentation normally arisesinperception andinpart
bythewaysin whichtherepresentation is normally usedto guide
actions(Armstrong 1961,1968;Dennett1969;Harman1973).
 The Intrinsic
QualityofExperience
/ 47

Theobjection hasusconsider twopeople,callthemAliceandFred,

withsimilarly functioning visualsystems butwithinverted spectra
withrespecttoeachother.ThingsthatlookredtoAlicelookgreen
to Fred,things thatlookblueto Alicelookorangeto Fred,and so
on. We are to imaginethatthisdifference betweenAliceand Fred
is notreflected in theirbehaviorin anyway.Theybothcall ripe
strawberries "red"and call grass"green"and theydo thisin the
effortlesswaysin whichnormalperceivers do who have learned
Englishin theusualways.
Consider whatthismeansforAliceina normalcase ofperception.
She looksat a ripestrawberry. Perceptual processing resultsin a
perceptual representationof that strawberry,includinga
representation ofitscolor.Sheusesthisrepresentation as herguide
to theenvironment, thatis,as herbeliefaboutthestrawberry, in
particular,herbeliefaboutitscolor.She expressesherbeliefabout
thecolorofthestrawberry byusingthewords,"itis red."Similarly,
forFred.His perception ofthestrawberry resultsin a perceptual
representationofthecolorofthestrawberry thatheusesas hisbelief
aboutthecolorand expresseswiththesamewords,"itis red."
Now,inthenormalcase ofperception, therecanbe nodistinction
betweenhowthingslookand howtheyare believedto be, since
how thingslook is given by the contentof one's perceptual
representationandinthenormal caseone'sperceptual representation
is usedas one'sbeliefabouttheenvironment. Thehypothesis ofthe
inverted spectrum objectionis thatthestrawberry looksdifferent
in colorto Aliceand to Fred.Sinceeverything is supposedto be
functioningintheminthenormal way,itfollows thattheymusthave
differentbeliefsaboutthecolorofthestrawberry. Iftheyhad the
samebeliefswhilehavingperceptual representations thatdiffered
in content,thenat least one of themwouldhave a perceptual
representation thatwas notfunctioning as hisor herbeliefabout
thecolorof thestrawberry, whichis to say thatit wouldnotbe
functioning in whatwe are assuming is thenormalway.
A furtherconsequence oftheinverted spectrum hypothesis isthat,
sincein thenormalcase Aliceand Fredexpresstheirbeliefsabout
thecolorofstrawberries and grassby saying"itis red"and "itis
green,"theymustmeansomething differentbytheircolorwords.
By"red"Fredmeansthewayripestrawberries lookto him.Since
thatis theway grasslooksto Alice,whatFredmeansby "red"is
whatshe meansby "'green."
48 / Gilbert
Harman

It is important to see thatthesereallyare consequencesofthe

inverted spectrum hypothesis.IfAliceandFredmeantthesamething
bytheircolorterms, theneither(a) one ofthemwouldnotbe using
thesewordsto expresshisor herbeliefsaboutcoloror (b) one of
themwouldnotbe usinghisor herperceptual representationsof
coloras hisor herbeliefsaboutcolor.In eithercase,therewould
be a failure ofnormalfunctioning, contrary tothehypothesis ofthe
inverted spectrum objection.
According ifAliceandFredusewordsinthesame
tofunctionalism,
waywithrespect tothesamethings, thentheymeanthesamethings
bythosewords(assuming also thattheyare members ofthesame
linguisticcommunity and theirwordsare takenfromthecommon
language).But this is just commonsense. SupposeAlice and
Humphrey arebothmembers ofthesamelinguistic community, using
wordsinthesameway,etc.Aliceis an ordinary humanbeingand
Humphrey is a humanoid robotmadeofquitea different material
fromAlice.Commonsensewouldattribute thesame meanings to
Humphrey's wordsas to Alice's,giventhattheyuse wordsin the
sameway.Somesortofphilosophical argument is neededtoargue
otherwise. Nosuchargument hasbeenprovided bydefenders ofthe
inverted spectrum objection.
Shoemaker(1982) offersa different versionof the inverted
spectrum objection.He has us considera singleperson,call him
Harry,at two different times,at an initialtimeof normalcolor
perception and at a latertimeafterHarryhas suffered through a
highly noticeablespectrum inversion (perhapsas theresultofthe
sortofbrainoperation described inLycan1973,inwhichnervesare
switchedaroundso thatred thingsnow have the perceptual
consequencesthatgreenthingsusedto have,etc.)and has finally
completely adaptedhisresponses so as torestorenormal functioning.
Shoemaker agreesthatHarrynowhasthesamebeliefsaboutcolor
as beforeand meansthesame thingsby hiscolorwords,and he
agreesthatthereisa senseinwhichstrawberries nowlooktoHarry
the same as theylookedbeforeHarry'sspectrum inversion.
But
Shoemaker takesittobe evident thatthereisanother senseof"looks"
inwhichitmayverywellbe truethatthings do notlookthesame
as theylookedbefore,so thatin thissecondsenseof "looks"red
thingslooktheway greenthingsusedto look.
Inotherwords,Shoemaker thinks itis evidentthattheremaybe
a psychologically relevantdifference between thesortofexperience
 QualityofExperience/ 49
The Intrinsic

Harryhad on lookingat a ripestrawberry at theinitialstageand

theexperiencehe hason lookingat a ripestrawberry at thefinal
stage(afterhe hascompletely adaptedto hisoperation). Thatis,he
thinksit is evidentthattheremaybe a psychologically relevant
difference betweenthese experienceseven thoughthereis no
functional difference and no difference in the contentof the
experiences.
Now,thismayseemevidentto anyonewhohas fallenvictimto
thesensedatumfallacy, whichholdsthatone'sawareness ofthecolor
ofa strawberry ismediatedbyone'sawareness ofan intrinsicfeature
ofa perceptual representation. Butwhyshouldanyoneelseagree?
Twoperceptual experienceswiththesameintentional content must
be psychologicallythesame.Inparticular, therecanbe nothing one
is awareofin havingtheone experiencethatone is notawareof
in havingtheother,sincetheintentional content ofan experience
comprises everything one is awareofin havingthatexperience.
I suggestthatShoemaker's inverted spectrum hypothesis willseem
evidentonlytosomeonewhobeginswiththepriorassumption that
peoplehavean immediate anddirectawarenessofintrinsic features
oftheirexperience, including thoseintrinsicfeatures thatfunction
to represent color.Sucha personcan thengo on to supposethat
theintrinsic featureofexperiencethatrepresents redforAliceis
theintrinsic featureofexperiencethatrepresents greenforFred,
and so forth.Thispriorassumption is exactlytheviewbehindthe
first whichI havearguediscontrary
objection, toordinary experience
and can be defended onlybyconfusing qualitiesoftheintentional
objects of experiencewith qualitiesof the experienceitself.
Shoemaker'sinvertedspectrumhypothesis therefore offersno
independent argument againstfunctionalism.3

Conclusion

To summarize briefly,
I havedescribedandrepliedtothreerelated
claimsthatwearedirectly
Thefirst
tofunctionalism.
objections aware
ofintrinsic ofourexperienceand arguesthatthereis no
features
view.To this,I
way to accountforthisawarenessin a functional
replythatwhenwe clearlydistinguishpropertiesoftheobjectof
experiencefrompropertiesoftheexperience, we see thatwe are
notawareoftherelevantintrinsicfeaturesoftheexperience.The
50 / Gilbert
Harman

secondobjection claimsthata personblindfrom birthcan knowall

aboutthefunctional roleofvisualexperience withoutknowing what
itis liketo see something red.To thisI replythattheblindperson
doesnotknowall aboutthefunctional roleofvisualexperience; in
particular, theblindpersondoes notknowhowsuchexperience
functions in relationto the perception of red objects.The third
objection claimsthatfunctionalism cannotaccountforthepossibility
of an invertedspectrum. To thisI replythatsomeonewiththe
relevant sortofinverted spectrum wouldhavetohavebeliefs about
thecolorsofthingsthatare different fromthebeliefsothershave
and wouldhave to meansomething different by hisor hercolor
terms, despitebeinga functionally normalcolorperceiver whosorts
thingsby colorin exactlythewayothersdo and who uses color
terminology inthesamewaythatothers do.Functionalism's rejection
ofthispossibility is commonsensical and is certainlynotso utterly
implausible orcounter-intuitive
thatthesecasespresent an objection
tofunctionalism. On theotherhand,toimaginethattherecouldbe
relevant casesofinverted spectrum without inversionofbeliefand
meaningis to fallbackontothefirst objection and notto offerany
additionalconsideration aaainstfunctionalism.

Notes
*Thepreparation ofthispaperwas supported in partby researchgrants
to Princeton
University fromtheJamesS. McDonnell Foundation and the
NationalScienceFoundation.
1. W. Shakespeare, Macbeth, ActII,SceneI: Is thisa daggerwhichI see
beforeme,The handletowardmyhand?Comeletme clutchthee.I
havetheenot,andyetI see theestill.Artthounot,fatalvision, sensible
To feeling
as to sight?or artthoubutA daggerofthemind,a false
creating,
Proceeding from theheatoppressed ... ...I see theestill;
brain?
Andonthybladeanddudgeon goutsofblood,Whichwasnotso before.
There'sno suchthing; itis thebloodybusinesswhichinforms Thusto
mineeyes.
2. I am indebted to SydneyShoemaker foremphasizing thisto me.
3. I shouldsay thatShoemakerhimself does notofferhiscase as an
to whathe callsfunctionalism.
objection He claimsthathisversionof
iscompatible
functionalism withhiscase.ButI amconsidering a version
offunctionalismthatisdefinedina waythatmakesitincompatible with
sucha case.
 The IntrinsicQualityof Experience/ 51

References

Anscombe, G E. M. (1965)"Theintentionality ofsensation:

a grammatical
feature,"AnalyticalPhilosophy, secondseries,editedbyR. J.Butler
(Oxford, Blackwell);reprinted inAnscombe, G.E.M.,Metaphysicsand
the Philosophy of Mind:CollectedPhilosophical Papers,VolumeII
(Minneapolis,Minnesota; University ofMinnesota Press:1981)pp.3-20.
Armstrong, DavidM. (1961)Perception and thePhysicalWorld(London:
Routledge and KeganPaul).
Armstrong,DavidM.(1962)Bodily Sensations (London:RoutledgeandKegan
Paul).
Armstrong, David M. (1968) The Materialist Theoryof Mind(London:
Routledge and KeganPaul).
Block,Ned(1986)"Advertisement fora semantics forpsychology,"Midwest
Studiesin Philosophy 10: 615-678.
Churchland,Paul(1985)"Reduction, qualia,andthedirectintrospection of
mentalstates,"Journal ofPhilosophy 82: &28.
Dennett,DanielC. (1969)Content and Consciousness (London:Routledge
and KeganPaul).
Dennett,DanielC. (1987)TheIntentional Stance(Cambridge,Massachusetts:
MITPress).
Harman, Gilbert
(1973)Thought (Princeton,NewJersey: Princeton
University
Press).
Harman,Gilbert (1987)"(Nonsolipsistic)conceptualrolesemantics," New
Directionsin Semantics, editedby ErnestLePore,London,Academic
Press(1987)55-81.
Harman, Gilbert
(1988)"Widefunctionalism," CognitionandRepresentation,
editedby StephenSchiffer and Susan Steele (Boulder,Colorado:
WestviewPress)11-20.
Jackson,Frank(1977)Perception: A Representative Theory(Cambridge,
England:Cambridge University Press).
Jackson,Frank(1982)"Epiphenomenal qualia,"Philosophical
Quarterly32:
127-32.
Jackson,Frank(1986)"WhatMarydidn'tknow,"Journal ofPhilosophy 83:
291-295.
Lewis,David K. (1983) "Postscript to 'Mad pain and martianpain',"
PhilosophicalPapers,Volume1, (NewYork:OxfordUniversity Press)
130-132.
Lycan,WilliamG. (1973)"Inverted spectrum" Ratio15.
Nagel,Thomas(1970)"Armstrong on themind," Review79,
Philosophical
reprintedinReadinginthePhilosophy ofPsychology Volume1,edited
by NedBlock(Canbridge, Massachusetts: HarvardUniversity Press).
Nagel,Thomas(1974)"Whatis it liketo be a bat?"Philosophical Review
83: 435-450.
Parsons,Terence(1980)Nonexistent Objects(NewHaven:Yale University
Press).
Peacocke,Christopher(1983)SenseandContent (Oxford:OxfordUniversity
Press).
52 / GilbertHarman

Pitcher,George(1971)A Theoryof Perception (Princeton,

New Jersey:
PrincetonUniversity Press).
Quine,W. V. (1948)"Onwhatthereis,"ReviewofMetaphysics, reprinted
inFroma LogicalPointof View(Cambridge, Harvard
Massachusetts;
UniversityPress:1953).
Quine,W.V. (1960)WordandObject(Cambridge, Massachusetts:
MITPress).
Rumelhart, DavidE., and McClelland,JamesL. (1986)ParallelDistributed
Processing, 2 volumes(Cambridge, Massachusetts:MITPress).
Shoemaker, Sydney (1982)"Theinvertedspectrum,"JournalofPhilosophy
79: 357-81.
Shoemaker,Sydney(1985) "Churchlandon reduction,qualia, and
FSA1984,Volume2 (Philosophy
introspection," ofScienceAssociation)
pp. 799-809.
Thagard, PaulT. (1986)"Parallelcomputation
andthemind-body problem,"
CognitiveScience10: 301-318.