



















Preview text:
The Problem of the External World
UniversityPressScholarshipOnline OxfordScholarshipOnline
TheSignificanceofPhilosophicalScepticism BarryStroud
Printpublicationdate:1984
PrintISBN-13:9780198247616
PublishedtoOxfordScholarshipOnline:November2003
DOI:10.1093/0198247613.001.0001
TheProblemoftheExternalWorld
BarryStroud(ContributorWebpage)
DOI:10.1093/0198247613.003.0001 AbstractandKeywords
Chapter1setsouttheProblemoftheExternalWorldtowhichscepticismisanatural
response:theproblem,firstposedbyDescartesinhisFirstMeditation,ofhowtoshow
thatwehaveanyknowledgeabouttheworldaroundus.
Reflectingonthenatureofhissensoryexperiences,Descartesfindshimselfunableto
ruleoutthepossibilitythatheisdreamingand,onthataccount,driventothedevastating
conclusionthatheknowsnothingatallabouttheworldaroundhim.Stroudarguesthatif
Descartesisrighttoinsistthatinordertoknowsomethingabouttheworldaroundhim
hemustknowthatheisnotdreaming,thenheisalsorightthathehasnosuch
knowledge,becausetheconditionforknowledgethatDescartesacceptscanneverbe
fulfilled:fulfillingitwouldrequireknowledgewhichitselfwouldbepossibleonlyifthe conditionwerefulfilled. Page 1 of 27
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2014.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: University of Sydney; date: 03 February 2015
The Problem of the External World
Themorepromisingstrategyinthefaceofthescepticalargument,therefore,isto
examinemorecarefullytherequirementthatwemustknowthatwearenotdreamingif
wearetoknowanythingabouttheworldaroundus.
Butifthatrequirementisafactofourordinaryconceptionofknowledge,asitseemsto
be,wemustacceptit,becausethereisnonotionofknowledgeotherthantheordinary
onethatisembodiedintheproceduresandpracticesofeverydayandscientificlife;and
unlesswefindawayofrejectingtheproblemaltogether,wewillhavetoacceptwithitthe
conclusionthatnooneknowsanythingaboutworldaroundus.
Keywords:conditionsofknowledge,Descartes,dreaming,FirstMeditation,ordinaryconceptionof
knowledge,Possibilityofknowledge,ProblemoftheExternalWorld,rejectionoftheproblem,sc eptical
argument,sc epticalexperienc e
SinceatleastthetimeofDescartesintheseventeenthcenturytherehasbeena
philosophicalproblemaboutourknowledgeoftheworldaroundus.1Putmostsimply,
theproblemistoshowhowwecanhaveanyknowledgeoftheworldatall.Theconclusion
thatwecannot,thatnooneknowsanythingabouttheworldaroundus,iswhatIcall
‘scepticismabouttheexternalworld’,sowecouldalsosaythattheproblemistoshow
howorwhyscepticismabouttheexternalworldisnotcorrect.Myaimisnottosolvethe
problembuttounderstandit.Ibelievetheproblemhasnosolution;orratherthatthe
onlyanswertothequestionasitismeanttobeunderstoodisthatwecanknownothing
abouttheworldaroundus.Buthowisthequestionmeanttobeunderstood?Itcanbe
expressedinafewEnglishwordsfamiliartoallofus,butIhopetoshowthatan
understandingofthespecialphilosophicalcharacterofthequestion,andofthe
inevitabilityofanunsatisfactoryanswertoit,cannotbeguaranteedbyour
understandingofthosewordsalone.Toseehowtheproblemismeanttobeunderstood
wemustthereforeexaminewhatisperhapsbestdescribedasitssource—howthe
problemarisesandhowitacquiresthatspecialcharacterthatmakesanunsatisfactory
negativeanswerinevitable.Wemusttrytounderstandthephilosophicalproblemofour
knowledgeoftheexternalworld.
TheproblemaroseforDescartesinthecourseofreflectingoneverythingheknows.He
reachedapointinhislifeatwhichhetriedtositbackandreflectoneverythinghehad
everbeentaughtortold,everythinghehadlearnedordiscoveredorbelievedsincehe
wasoldenoughtoknowor(p.2) believeanything.2Wemightsaythathewasreflecting
onhisknowledge,butputtingitthatwaycouldsuggestthatwhathewasdirectinghis
attentiontowasindeedknowledge,andwhetheritwasknowledgeornotisprecisely
whathewantedtodetermine.‘AmongallthethingsIbelieveortaketobetrue,what
amountstoknowledgeandwhatdoesnot?’;thatisthequestionDescartesaskshimself.It
isobviouslyaverygeneralquestion,sinceitasksabouteverythinghebelievesortakes
tobetrue,butinotherrespectsitsoundsjustlikethesortofquestionweareperfectly
familiarwithineverydaylifeandoftenknowhowtoanswer.
Forexample,Ihavecometoacceptovertheyearsagreatmanythingsaboutthe
commoncold.Ihavealwaysbeentoldthatonecancatchcoldbygettingwetfeet,orfrom Page 2 of 27
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2014.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: University of Sydney; date: 03 February 2015
The Problem of the External World
sittinginadraught,orfromnotdryingone'shairbeforegoingoutdoorsincoldweather.
Ihavealsolearnedthatthecommoncoldistheeffectofavirustransmittedbyanalready
infectedperson.AndIalsobelievethatoneismorevulnerabletocoldswhenover‐tired,
understress,orotherwiseinlessthanthebestofhealth.Someofthesebeliefsseemto
meonreflectiontobeinconsistentwithsomeothers;Iseethatitisveryunlikelythatall
ofthemcouldbetrue.Perhapstheycouldbe,butIacknowledgethatthereismuchIdo
notunderstand.IfIsitbackandtrytothinkaboutallmy‘knowledge’ofthecommon
cold,then,Imighteasilycometowonderhowmuchofitreallyamountstoknowledge
andhowmuchdoesnot.WhatdoIreallyknowaboutthecommoncold?IfIwere
sufficientlyinterestedinpursuingthematteritwouldbenaturaltolookintothesourceof
mybeliefs.Hasthereeverbeenanygoodreasonforthinkingthatcoldsareeven
correlatedwithwethairincoldweather,forexample,orwithsittinginadraught?Arethe
peoplefromwhomIlearnedsuchthingslikelytohavebelievedthemforgoodreasons?
Arethosebeliefsjustoldwives'tales,oraretheyreallytrue,andperhapsevenknownto
betruebysomepeople?ThesearequestionsImightaskmyself,andIhaveatleasta
generalideaofhowtogoaboutansweringthem.
(p.3) Apartfrommyimpressionoftheimplausibilityofallmybeliefsaboutthecommon
coldbeingtruetogether,Ihavenotmentionedanyotherreasonforbeinginterestedin
investigatingthestateofmyknowledgeonthatsubject.Butforthemomentthatdoesnot
seemtoaffecttheintelligibilityorthefeasibilityofthereflectiveproject.Thereisnothing
mysteriousaboutit.Itisthesortoftaskwecanbeledtoundertakeforanumberof
reasons,andoftenverygoodreasons,insofaraswehaveverygoodreasonsfor
preferringknowledgeandfirmbelieftoguessworkorwishfulthinkingorsimplytaking thingsforgranted.
Reflectiononorinvestigationofourputativeknowledgeneednotalwaysextendtoawide
areaofinterest.Itmightbeimportanttoaskwhethersomequitespecificandparticular
thingIbelieveorhavebeentakingforgrantedisreallysomethingIknow.Asamember
ofajuryImightfindthatIhavebeenrulingoutonesuspectinmymindbecausehewas
athousandmilesaway,inCleveland,atthetimeofthecrime.ButImightthenbeginto
askmyselfwhetherthatisreallysomethingthatIknow.Iwouldreflectonthesourceof
mybelief,butreflectioninthiscaseneednotinvolveageneralscrutinyofeverythingI
takemyselftoknowaboutthecase.Re‐examiningtheman'salibiandthecredentialsofits
supportingwitnessesmightbeenoughtosatisfyme.IndeedImightfindthatitsreliability
onthosecountsispreciselywhatIhadbeengoingonallalong.
Inpointingoutthatweareperfectlyfamiliarwiththeideaofinvestigatingorreviewing
ourknowledgeonsomeparticularmatterorinsomegeneralareaIdonotmeanto
suggestthatitisalwayseasytosettlethequestion.Dependingonthenatureofthecase,
itmightbeverydifficult,perhapsevenimpossibleatthetime,toreachafirmconclusion.
Forexample,itwouldprobablybeverydifficultifnotimpossibleformetotraceand
assesstheoriginsofmanyofthosethingsIbelieveaboutthecommoncold.Butitis
equallytruethatsometimesitisnotimpossibleorevenespeciallydifficulttoanswerthe
question.Wedosometimesdiscoverthatwedonotreallyknowwhatwepreviously Page 3 of 27
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2014.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: University of Sydney; date: 03 February 2015
The Problem of the External World
thoughtweknew.ImightfindthatwhatIhadpreviously(p.4) believedisnoteventrue
—thatsittingindraughtsisnotevencorrelatedwithcatchingacold,forexample.OrI
mightfindthatthereisnotorperhapsneverwasanygoodreasontobelievewhatI
believed—thattheman'salibiwasconcoctedandthenfalselytestifiedtobyhisfriends.I
couldreasonablyconcludeineachcasethatI,andeveryoneelseforthatmatter,never
didknowwhatIhadpreviouslythoughtIknew.Weareallfamiliarwiththeordinary
activityofreviewingourknowledge,andwiththeexperienceofreachingapositive
verdictinsomecasesandanegativeverdictinothers.
Descartes'sowninterestinwhatheknowsandhowheknowsitispartofhissearchfor
whathecallsageneralmethodfor‘rightlyconductingreasonandseekingtruthinthe
sciences’.3Hewantsamethodofinquirythathecanbeassuredinadvancewillleadonly
tothetruthifproperlyfollowed.Ithinkwedonotneedtoendorsethewisdomofthat
searchorthefeasibilityofthatprogrammeinordertotrytogoalongwithDescartesin
hisgeneralassessmentofthepositionheisinwithrespecttothethingshebelieves.He
comestofindhisputativeknowledgewantingincertaingeneralrespects,anditisinthe
courseofthatoriginalnegativeassessmentthattheproblemIaminterestedinarises.I
calltheassessment‘negative’becausebytheendofhisFirstMeditationDescartesfinds
thathehasnogoodreasontobelieveanythingabouttheworldaroundhimand
thereforethathecanknownothingoftheexternalworld.
Howisthatassessmentconducted,andhowcloselydoesitparallelthefamiliarkindof
reviewofourknowledgethatweallknowhowtoconductineverydaylife?Thequestion
inoneformoranotherwillbewithusfortherestofthisbook.Itisthequestionofwhat
exactlytheproblemofourknowledgeoftheexternalworldamountsto,andhowitarises
withitsspecialphilosophicalcharacter.Thesourceoftheproblemistobefound
somewherewithinorbehindthekindofthinkingDescartesengagesin.
OnewayDescartes'squestionabouthisknowledgediffersfromtheeverydayexamplesI
consideredisinbeing(p.5) concernedwitheverythinghebelievesortakestobetrue.
Howdoesonegoaboutassessingallofone'sknowledgeallatonce?Iwasabletolista
fewofthethingsIbelieveaboutthecommoncoldandthentoaskabouteachofthem
whetherIreallyknowit,andifsohow.ButalthoughIcancertainlylistanumberofthe
thingsIbelieve,andIwouldassenttomanymoreofthemassoonastheywereputto
me,thereobviouslyisnohopeofassessingeverythingIbelieveinthispiecemealway.
Foronething,itprobablymakesnosense,strictlyspeaking,totalkofthenumberof
thingsonebelieves.IfIamaskedwhetheritisoneofmybeliefsthatIwenttoseeafilm
lastnightIcantrulyanswer‘Yes’.IfIwereaskedwhetheritisoneofmybeliefsthatI
wenttothemovieslastnightIwouldgivethesameanswer.HaveItherebyidentified
two,oronlyone,ofmybeliefs?Howisthatquestionevertobesettled?IfwesaythatI
identifiedonlyoneofmybeliefs,itwouldseemthatImustalsobesaidtoholdthe
furtherbeliefthatgoingtoseeafilmandgoingtothemoviesareoneandthesamething.
Sowewouldhavemorethanonebeliefafterall.Theprospectsofarrivingevenata
principleforcountingbeliefs,letaloneatanactualnumberofthem,seemdim.
Evenifitdidmakesensetocountthethingswebelieveitisprettyclearthatthenumber Page 4 of 27
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2014.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: University of Sydney; date: 03 February 2015
The Problem of the External World
wouldbeindefinitelylargeandsoanassessmentofourbeliefsonebyonecouldnever
becompletedanyway.Thisiseasilyseenbyconsideringonlysomeofthesimplestthings
oneknows,forexampleinarithmetic.OnethingIknowisthatoneplusoneequalstwo.
AnotherthingIknowisthatoneplustwoisthree,andanother,thatoneplusthreeis
four.ObviouslytherecouldbenoendtothetaskofassessingmyknowledgeifIhadto
investigateseparatelythesourceofeachoneofmybeliefsinthatseries.AndevenifI
succeededIwouldonlyhaveassessedthethingsIknowabouttheadditionofthe
numberonetoagivennumber;Iwouldstillhavetodothesamefortheadditionoftwo,
andthentheadditionofthree,andsoon.Andeventhatwouldexhaustonlymybeliefs
aboutaddition;allmyothermathematicalbeliefs,nottomentionalltherestofmy
knowledge,wouldremainsofarunexamined.Obviouslythejobcannotbedone
piecemeal,onebyone.(p.6) Somemethodmustbefoundforassessinglargeclassesof beliefsallatonce.
Onewaytodothiswouldbetolookforcommonsourcesorchannelsorbasesofour
beliefs,andthentoexaminethereliabilityofthosesourcesorbases,justasIexamined
thesourceorbasisofmybeliefthatthesuspectwasinCleveland.Descartesdescribes
suchasearchasasearchfor‘principles’ofhumanknowledge,‘principles’whosegeneral
credentialshecantheninvestigate(HR,145).Ifsome‘principles’arefoundtobe
involvedinallorevenmostofourknowledge,anassessmentofthereliabilityofthose
‘principles’couldbeanassessmentofallormostofourknowledge.IfIfoundgood
reasontodoubtthereliabilityofthesuspect'salibi,forexample,andthatwasallIhadto
gooninmybeliefthathewasinCleveland,thenwhatIearliertooktobemyknowledge
thathewasinClevelandwouldhavebeenfoundwantingorcalledintoquestion.Its
sourceorbasiswouldhavebeenundermined.Similarly,ifoneofthe‘principles’orbases
onwhichallmyknowledgeoftheworlddependswerefoundtobeunreliable,my
knowledgeoftheworldwouldtothatextenthavebeenfoundwantingorcalledinto questionaswell.
Arethereanyimportant‘principles’ofhumanknowledgeinDescartes'ssense?Ittakes
verylittlereflectiononthehumanorganismtoconvinceusoftheimportanceofthe
senses—sight,hearing,touch,taste,andsmell.Descartesputsthepointmoststrongly
whenhesaysthat‘allthatuptothepresenttimeIhaveacceptedasmosttrueand
certainIhavelearnedeitherfromthesensesorthroughthesenses’(HR,145).Exactly
whathewouldincludeunder‘thesenses’hereisperhapssomewhatindeterminate,but
evenifitisleftvaguemanyphilosopherswoulddenywhatDescartesappearstobe
saying.Theywouldholdthat,forexample,themathematicalknowledgeImentioned
earlierisnotandcouldnotbeacquiredfromthesensesorthroughthesenses,sonot
everythingIknowisknowninthatway.WhetherDescartesisreallydenyingtheviewsof
thosewhobelieveinthenon‐sensorycharacterofmathematicalknowledge,andwhether,
ifhewere,hewouldberight,areissueswecansetasideforthemoment.Itisclearthat
thesensesareatleast(p.7) veryimportantforhumanknowledge.Evenrestricting
ourselvestothetraditionalfivesenseswecanbegintoappreciatetheirimportanceby
reflectingonhowlittlesomeonewouldevercometoknowwithoutthem.Apersonblind
anddeaffrombirthwhoalsolackedtastebudsandasenseofsmellwouldknowvery Page 5 of 27
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2014.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: University of Sydney; date: 03 February 2015
The Problem of the External World
littleaboutanything,nomatterhowlonghelived.Toimaginehimalsoanaesthetizedor
withoutasenseoftouchisperhapstostretchaltogethertoofarone'sconceptionofa
humanorganism,oratleastahumanorganismfromwhomwecanhopetolearn
somethingabouthumanknowledge.Theimportanceofthesensesasasourceorchannel
ofknowledgeseemsundeniable.Itseemspossible,then,toacknowledgetheir
importanceandtoassessthereliabilityofthatsource,quiteindependentlyofthedifficult
questionofwhetherallourknowledgecomestousinthatway.Wewouldthenbe
assessingthecredentialsofwhatisoftencalledour‘sensory’or‘experiential’or
‘empirical’knowledge,andthat,asweshallsee,isquiteenoughtobegoingonwith.
Havingfoundanextremelyimportant‘principle’orsourceofourknowledge,howcanwe
investigateorassessalltheknowledgewegetfromthatsource?Asbefore,wearefaced
withtheproblemoftheinexhaustibilityofthethingswebelieveonthatbasis,sono
piecemeal,one‐by‐oneprocedurewilldo.Butperhapswecanmakeasweepingnegative
assessment.Itmightseemthatassoonaswehavefoundthatthesensesareoneofthe
sourcesofourbeliefsweareimmediatelyinapositiontocondemnallputativeknowledge
derivedfromthem.Somephilosophersappeartohavereasonedinthisway,andmany
haveevensupposedthatDescartesisamongthem.TheideaisthatifIamassessingthe
reliabilityofmybeliefsandaskingwhetherIreallyknowwhatItakemyselftoknow,andI
comeacrossalargeclassofbeliefswhichhavecometomethroughthesenses,Ican
immediatelydismissallthosebeliefsasunreliableorasnotamountingtoknowledge
becauseoftheobviousfactthatIcansometimesbewronginmybeliefsbasedonthe
senses.Thingsarenotalwaysastheyappear,soifonthebasisofthewaytheyappearto
meIbelievethattheyreallyareacertainway,Imightstillbewrong.Wehaveallfoundat
onetimeor(p.8) anotherthatwehavebeenmisledbyappearances;weknowthatthe
sensesarenotalwaysreliable.Shouldwenotconclude,then,thatasageneralsourceof
knowledgethesensesarenottobetrusted?AsDescartesputsit,isitnotwisernever
‘totrustentirelytoanythingbywhichwehaveoncebeendeceived’(HR,145)?Don't
wehavehereaquitegeneralwayofcondemningasnotfullyreliableallofourbeliefs
acquiredbymeansofthesenses?
Ithinktheanswertothatquestionis‘No,wedonot’,andIthinkDescarteswouldagree
withthatanswer.Itistruethathedoestalkofthesenses‘deceiving’usonparticular
occasions,andhedoesaskwhetherthatisnotenoughtocondemnthesensesingeneral
asasourceofknowledge,butheimmediatelyremindsusoftheobviousfactthatthe
circumstancesinwhichthesenses‘deceive’usmightbespecialincertainascertainable
ways,andsotheiroccasionalfailureswouldnotsupportablanketcondemnationoftheir reliability.
Sometimes,togiveanancientexample,atowerlooksroundfromadistancewhenitis
actuallysquare.Ifwereliedonlyontheappearancesofthemomentwemightsaythatthe
distanttowerisround,andwewouldbewrong.Wealsoknowthattherearemanysmall
organismsinvisibletothenakedeye.Ifthetablebeforemeiscoveredwithsuch
organismsatthemomentbutIlookatitandsaythereisnothingonthetableatall,once
againIwillbewrong.Butallthatfollowsfromthesefamiliarfacts,asDescartespointsout, Page 6 of 27
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2014.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: University of Sydney; date: 03 February 2015
The Problem of the External World
isthattherearethingsaboutwhichwecanbewrong,ortherearesituationsinwhichwe
cangetfalsebeliefs,ifwerelyentirelyonoursensesatthatmoment.Sosometimeswe
shouldbecarefulaboutwhatwebelieveonthebasisofthesenses,orsometimes
perhapsweshouldwithholdourassentfromanystatementabouthowthingsare—when
thingsaretoofarawaytobeseenproperly,forexample,ortoosmalltobeseenatall.
Butthatobviouslyisnotenoughtosupportthepolicyofnevertrustingone'ssenses,or
neverbelievinganythingbasedonthem.NordoesitshowthatIcanneverknow
anythingbymeansofthesenses.Ifmycarstartspromptlyeverymorningfortwoyears
intemperateweatheratsealevelbutthenfailstostart(p.9) onemorninginfreez ing
weatheratthetopofahighmountain,thatdoesnotsupportthepolicyofnevertrusting
mycartostartagainonceIreturntothetemperateloweraltitudefromwhichIso
foolishlytookit.NordoesitshowthatIcanneverknowwhethermycarwilleverstart
again.Itshowsonlythattherearecertaincircumstancesinwhichmyotherwisefully
reliablecarmightnotstart.Sothefactthatwearesometimeswrongor‘deceived’inour
judgementsbasedonthesensesisnotenoughinitselftoshowthatthesensesarenever
tobetrustedandarethereforeneverreliableasasourceofknowledge.
Descartes'snegativeassessmentofallofhissensoryknowledgedoesnotdependonany
suchreasoning.Hestartshisinvestigation,rather,inwhatwouldseemtobethemost
favourableconditionsforthereliableoperationofthesensesasasourceofknowledge.
WhileengagingintheveryphilosophicalreflectionsheiswritingaboutinhisFirst
MeditationDescartesissittinginawarmroom,bythefire,inadressinggown,witha
pieceofpaperinhishand.Hefindsthatalthoughhemightbeabletodoubtthatadistant
towerthatlooksroundreallyisround,itseemsimpossibletodoubtthathereallyis
sittingtherebythefireinhisdressinggownwithapieceofpaperinhishand.Thefireand
thepieceofpaperarenottoosmallortoofarawaytobeseenproperly,theyareright
therebeforehiseyes;itseemstobethebestkindofpositionsomeonecouldbeinfor
gettingreliablebeliefsorknowledgebymeansofthesensesaboutwhatisgoingon
aroundhim.ThatisjusthowDescartesregardsit.Itsbeingabest‐possiblecaseofthat
kindispreciselywhathethinksenableshimtoinvestigateorassessatonefellswoopall
oursensoryknowledgeoftheworldaroundus.Theverdicthearrivesatabouthis
putativeknowledgethatheissittingbythefirewithapieceofpaperinhishandinthat
particularsituationservesasthebasisforacompletelygeneralassessmentofthesenses
asasourceofknowledgeabouttheworldaroundus.
Howcanthatbeso?Howcanhesoeasilyreachageneralverdictaboutallhissensory
knowledgeonthebasisofasingleexample?Obviouslynotsimplybygeneraliz ingfrom
oneparticularexampletoallcasesofsensoryknowledge,asonemightwildlyleaptoa
conclusionaboutallred‐haired(p.10) menonthebasisofoneortwoindividuals.
Rather,hetakestheparticularexampleofhisconvictionthatheissittingbythefirewitha
pieceofpaperinhishandasrepresentativeofthebestpositionanyofuscaneverbein
forknowingthingsabouttheworldaroundusonthebasisofthesenses.Whatistrueof
arepresentativecase,ifitistrulyrepresentativeanddoesnotdependonspecial
peculiaritiesofitsown,canlegitimatelysupportageneralconclusion.Ademonstration
thataparticularisoscelestrianglehasacertainproperty,forexample,canbetakenasa Page 7 of 27
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2014.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: University of Sydney; date: 03 February 2015
The Problem of the External World
demonstrationthatallisoscelestriangleshavethatproperty,aslongastheoriginal
instancewastypicalorrepresentativeofthewholeclass.WhetherDescartes's
investigationofthegeneralreliabilityofthesensesreallydoesfollowthatfamiliarpattern
isadifficultquestion.Whether,orinpreciselywhatsense,theexampleheconsiderscan
betreatedasrepresentativeofourrelationtotheworldaroundusis,Ibelieve,thekey
tounderstandingtheproblemofourknowledgeoftheexternalworld.Butifitturnsout
thatthereisnothingillegitimateaboutthewayhisnegativeconclusionisreached,the
problemwillbeproperlyposed.
ForthemomentIthinkatleastthismuchcanbesaidaboutDescartes'sreasoning.He
choosesthesituationinwhichhefindshimselfasrepresentativeofthebestpositionwe
canbeinforknowingthingsabouttheworldinthesensethat,ifitisimpossibleforhimin
thatpositiontoknowthatheissittingbythefirewithapieceofpaperinhishandthenitis
alsoimpossibleforhiminothersituationstoknowanythingabouttheworldaroundhim
onthebasisofhissenses.Anegativeverdictinthechosencasewouldsupportanegative
verdicteverywhereelse.TheexampleDescartesconsidersisinthatsensemeanttobe
thebestkindofcasetherecouldbeofsensoryknowledgeabouttheworldaroundus.I
thinkwemustadmitthatitisverydifficulttoseehowDescartesoranyoneelsecouldbe
anybetteroffwithrespecttoknowingsomethingabouttheworldaroundhimonthe
basisofthesensesthanheisinthecaseheconsiders.Butifnoonecouldbeinany
betterpositionforknowing,itseemsnaturaltoconcludethatanynegativeverdict
arrivedataboutthisexample,anydiscoverythatDescartes'sbeliefsinthiscasearenot
reliableordonotamounttoknowledge,could(p.11) safelybegeneraliz edintoa
negativeconclusionaboutallofoursensory‘knowledge’oftheworld.Ifcandidateswith
thebestpossiblecredentialsarefoundwanting,allthosewithlessimpressivecredentials
mustfallshortaswell.
Itwillseematfirstsightthatinconcedingthatthewholequestionturnsonwhether
Descartesknowsinthisparticularcaseweareconcedingverylittle;itseemsobvious
thatDescartesonthatoccasiondoesknowwhathethinksheknowsabouttheworld
aroundhim.ButinfactDescartesfindsthathecannotknowinthiscasethatheissitting
bythefirewithapieceofpaperinhishand.Ifthecaseistrulyrepresentativeofour
sensoryknowledgeingeneral,thatwillshowthatnoonecanknowanythingaboutthe
worldaroundus.Buthowcouldheeverarriveatthatnegativeverdictintheparticular
caseheconsiders?Howcouldanyonepossiblydoubtinsuchacasethatthefireandthe
pieceofpaperarethere?ThepaperisinDescartes'shand,thefireisrighttherebefore
hisopeneyes,andhefeelsitswarmth.Wouldn'tanyonehavetobemadtodenythathe
canknowsomethingaboutwhatisgoingonaroundhiminthosecircumstances?
Descartesfirstanswers‘Yes’.Hesaysthatifheweretodoubtordenyonthatoccasion
thatheissittingbythefirewithapieceofpaperinhishandhewouldbenolessmadthan
thosepauperswhosaytheyarekingsorthosemadmenwhothinktheyarepumpkinsor
aremadeofglass.Buthisreflectionscontinue:
AtthesametimeImustrememberthatIamaman,andthatconsequentlyIamin
thehabitofsleeping,andinmydreamsrepresentingtomyselfthesamethingsor Page 8 of 27
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2014.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: University of Sydney; date: 03 February 2015
The Problem of the External World
sometimesevenlessprobablethings,thandothosewhoareinsaneintheirwaking
moments.HowoftenhasithappenedtomethatinthenightIdreamtthatIfound
myselfinthisparticularplace,thatIwasdressedandseatednearthefire,whilstin
realityIwaslyingundressedinbed!Atthismomentitdoesindeedseemtome
thatitiswitheyesawakethatIamlookingatthispaper;thatthisheadwhichImove
isnotasleep,thatitisdeliberatelyandofsetpurposethatIextendmyhandand
perceiveit;whathappensinsleepdoesnotappearsoclearnorsodistinctasdoes
allthis.ButinthinkingoverthisIremindmyselfthatonmanyoccasionsIhavein
sleepbeendeceivedbysimilarillusions,andindwellingcarefullyonthisreflectionI
seesomanifestlythattherearenocertainindicationsbywhichwemayclearly
distinguishwakefulnessfromsleepthatIamlostinastonishment.Andmy
astonishmentissuchthatitisalmostcapableofpersuadingmethatInowdream. (HR,145–6.)
(p.12) Withthisthought,ifheisright,Descarteshaslostthewholeworld.Heknows
whatheisexperiencing,heknowshowthingsappeartohim,buthedoesnotknow
whetherheisinfactsittingbythefirewithapieceofpaperinhishand.Itis,forhim,
exactlyasifheweresittingbythefirewithapieceofpaperinhishand,buthedoesnot
knowwhethertherereallyisafireorapieceofpaperthereornot;hedoesnotknow
whatisreallyhappeningintheworldaroundhim.Herealiz esthatifeverythinghecan
everlearnaboutwhatishappeningintheworldaroundhimcomestohimthroughthe
senses,buthecannottellbymeansofthesenseswhetherornotheisdreaming,thenall
thesensoryexperiencesheishavingarecompatiblewithhismerelydreamingofaworld
aroundhimwhileinfactthatworldisverydifferentfromthewayhetakesittobe.Thatis
whyhethinkshemustfindsomewaytotellthatheisnotdreaming.Farfromitsbeing
madtodenythatheknowsinthiscase,hethinkshisrecognitionofthepossibilitythathe
mightbedreaminggiveshim‘verypowerfulandmaturelyconsidered’(HR,148)
reasonsforwithholdinghisjudgementabouthowthingsareintheworldaroundhim.He
thinksitiseminentlyreasonabletoinsistthatifheistoknowthatheissittingbythefire
hemustknowthatheisnotdreamingthatheissittingbythefire.Thatisseenasa
necessaryconditionofknowingsomethingabouttheworldaroundhim.Andhefindsthat
thatconditioncannotbefulfilled.Oncarefulreflectionhediscoversthat‘thereareno
certainindicationsbywhichwemayclearlydistinguishwakefulnessfromsleep’.He
concludesthatheknowsnothingabouttheworldaroundhimbecausehecannottellthat
heisnotdreaming;hecannotfulfiloneoftheconditionsnecessaryforknowingsomething abouttheworld.
TheCartesianproblemofourknowledgeoftheexternalworldthereforebecomes:how
canweknowanythingabouttheworldaroundusonthebasisofthesensesifthesenses
giveusonlywhatDescartessaystheygiveus?Whatwegainthroughthesensesison
Descartes'sviewonlyinformationthatiscompatiblewithourdreamingthingsaboutthe
worldaroundusandnotknowinganythingaboutthatworld.Howthencanweknow
anythingabouttheworldbymeansofthe(p.13) senses?TheCartesianargument
presentsachallengetoourknowledge,andtheproblemofourknowledgeoftheexternal
worldistoshowhowthatchallengecanbemet. Page 9 of 27
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2014.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: University of Sydney; date: 03 February 2015
The Problem of the External World
WhenIspeakhereoftheCartesianargumentorofDescartes'sscepticalconclusionorof
hisnegativeverdictabouthisknowledgeIreferofcourseonlytothepositionhefinds
himselfinbytheendofhisFirstMeditation.Havingatthatpointdiscoveredandstated
theproblemoftheexternalworld,DescartesgoesonintherestofhisMeditationstotry
tosolveit,andbytheendoftheSixthMeditationhethinkshehasexplainedhowhe
knowsalmostallthosefamiliarthingshebeganbyputtinginquestion.SowhenIascribe
toDescartestheviewthatwecanknownothingabouttheworldaroundusIdonot
meantosuggestthatthatishisfinalandconsideredview;itisnothingmorethana
conclusionhefeelsalmostinevitablydriventoattheearlystagesofhisreflections.But
thosearetheonlystagesofhisthinkingIaminterestedinhere.Thatiswherethe
philosophicalproblemofourknowledgeoftheexternalworldgetsposed,andbeforewe
canconsiderpossiblesolutionswemustbesureweunderstandexactlywhatthe problemis.
Ihavedescribeditasthatofshowingorexplaininghowknowledgeoftheworldaround
usispossiblebymeansofthesenses.Itisimportanttokeepinmindthatthatdemandfor
anexplanationarisesinthefaceofachallengeorapparentobstacletoourknowledgeof
theworld.ThepossibilitythatheisdreamingisseenasanobstacletoDescartes's
knowingthatheissittingbythefire,anditmustbeexplainedhowthatobstaclecaneither
beavoidedorovercome.Itmustbeshownorexplainedhowitispossibleforustoknow
thingsabouttheworld,giventhatthesense‐experienceswegetarecompatiblewithour
merelydreaming.Explaininghowsomethingisneverthelesspossible,despitewhatlooks
likeanobstacletoit,requiresmorethanshowingmerelythatthereisnoimpossibility
involvedinthething—thatitisconsistentwiththeprinciplesoflogicandthelawsofnature
andsointhatsensecouldexist.Themerepossibilityofthestateofaffairsisnotenoughto
settlethequestionofhowourknowledgeoftheworldispossible;wemustunderstand
howtheapparentobstacleistobegotround.
(p.14) Descartes'sreasoningcanbeexaminedandcriticizedatmanydifferentpoints,
andhasbeencloselyscrutinizedbymanyphilosophersforcenturies.Ithasalsobeen
acceptedbymany,perhapsbymorethanwouldadmitorevenrealiz ethattheyacceptit.
Thereseemstomenodoubtabouttheforceandthefascination—Iwouldsaythealmost
overwhelmingpersuasiveness—ofhisreflections.Thataloneissomethingthatneeds
accountingfor.Icannotpossiblydojusticetoallreasonablereactionstothemhere.In
therestofthisfirstchapterIwanttoconcentrateondeepeningandstrengtheningthe
problemandtryingtolocatemorepreciselythesourceofitspower.
Thereareatleastthreedistinctquestionsthatcouldbepressed.Isthepossibilitythat
Descartesmightbedreamingreallyathreattohisknowledgeoftheworldaroundhim?
Isherightinthinkingthathemustknowthatheisnotdreamingifheistoknow
somethingabouttheworldaroundhim?Andisherightinhis‘discovery’thathecan
neverknowthatheisnotdreaming?IfDescarteswerewrongonanyofthesepointsit
mightbepossibletoavoidtheproblemandperhapseventoexplainwithoutdifficultyhow
weknowthingsabouttheworldaroundus.
Onthefirstquestion,itcertainlyseemsrighttosaythatifDescartesweredreamingthat Page 10 of 27
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2014.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: University of Sydney; date: 03 February 2015
The Problem of the External World
heissittingbythefirewithapieceofpaperinhishandhewouldnotthenknowthatheis
sittingbythefirewithapieceofpaperinhishand.Whenyoudreamthatsomethingis
goingonintheworldaroundyouyoudonottherebyknowthatitis.Mostoften,of
course,whatwedreamisnoteventrue;nooneisactuallychasinguswhenwearelying
asleepinbeddreaming,norareweactuallyclimbingstairs.Butalthoughusuallywhatwe
dreamisnotreallyso,thatisnottherealreasonforourlackofknowledge.Evenif
Descarteswereinfactsittingbythefireandactuallyhadapieceofpaperinhishandat
theverytimehewasdreamingthatheissittingbythefirewithapieceofpaperinhis
hand,hewouldnottherebyknowhewassittingtherewiththatpaper.Hewouldbelikea
certainDukeofDevonshirewho,accordingtoG.E.Moore,oncedreamthewasspeaking
intheHouseofLordsandwokeuptofindthathewasspeakingintheHouseof(p.15)
Lords.4Whathewasdreamingwasinfactso.Butevenifwhatyouaredreamingisinfact
soyoudonottherebyknowthatitis.Evenifweallowthatwhenyouaredreamingthat
somethingissoyoucanbesaid,atleastforthetimebeing,tothinkortobelievethatitis
so,thereisstillnorealconnectionbetweenyourthinkingorbelievingwhatyoudoandits
beingso.Atbestyouhaveathoughtorabeliefwhichjusthappenstobetrue,butthatis
nomorethancoincidenceandnotknowledge.SoDescartes'sfirststepreliesonwhat
seemstobeanundeniablefactaboutdreams:ifyouaredreamingthatsomethingisso
youdonottherebyknowthatitisso.
Thisbaldclaimneedstobequalifiedandmorecarefullyexplained,butIdonotthinkthat
willdiminishtheforceofthepointforDescartes'spurposes.Sometimeswhatisgoingon
intheworldaroundushasaneffectonwhatwedream;forexample,abangingshutter
mightactuallycausemetodream,amongotherthings,thatashutterisbanging.Ifmy
environmentaffectsmeinthatway,andifindreamsIcanbesaidtothinkorbelievethat
somethingisso,wouldInotinthatcaseknowthatashutterisbanging?Itseemstome
thatIwouldnot,butIconfessitisdifficulttosayexactlywhyIthinkso.Thatisprobably
becauseitisdifficulttosayexactlywhatisrequiredforknowledge.Weusetheterm
‘know’confidently,wequiteeasilydistinguishcasesofknowledgefromcasesofits
absence,butwearenotalwaysinapositiontostatewhatwearegoingoninapplyingor
withholdingtheterminthewayswedo.Ithinkthatinthecaseofthebangingshutterit
wouldnotbeknowledgebecauseIwouldbedreaming,Iwouldnotevenbeawake.At
leastitcanbesaid,Ithink,thatevenifDescartes'ssittingbythefirewithapieceofpaper
inhishand(likethebangingshutter)iswhatinfactcauseshimtodreamthatheissitting
bythefirewithapieceofpaperinhishand,thatisstillnohelptohimincomingtoknow
whatisgoingonintheworldaroundhim.Herealiz esthathecouldbedreamingthathe
issittingbythefireevenifheisinfactsittingthere,andthatisthepossibilityhefindshe hastoruleout.
Ihavesaidthatifyouaredreamingthatsomethingisso(p.16) youdonotthereby
knowthatitisso,anditmightseemasifthatisnotalwaystrue.Supposeamananda
childarebothsleeping.Isayofthechildthatitissoyoungitdoesnotknowwhatseven
timesnineis,whereasthegrownmandoesknowthat.Ifthemanhappensatthatvery
momenttobedreamingthatseventimesnineissixty‐three(perhapsheisdreamingthat
heiscomputinghisincometax),thenheisamanwhoisdreamingthatsomethingisso Page 11 of 27
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2014.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: University of Sydney; date: 03 February 2015
The Problem of the External World
andalsoknowsthatitisso.Thesamekindofthingispossibleforknowledgeaboutthe
worldaroundhim.Hemightbeaphysicistwhoknowsagreatdealaboutthewaythings
arewhichthechilddoesnotknow.Ifthemanalsodreamsthatthingsarethatwayhecan
onceagainbesaidtobedreamingthatsomethingissoandalsotoknowthatitisso.
Thereisthereforenoincompatibilitybetweendreamingandknowing.Thatistrue,butI
donotthinkitaffectsDescartes'sargument.Heisledtoconsiderhowheknowsheisnot
dreamingatthemomentbyreflectingonhowheknowsatthatmomentthatheissitting
bythefirewithapieceofpaperinhishand.Ifheknowsthatatall,hethinks,heknowsit
onthebasisofthesenses.Butherealiz esthathishavingthesensoryexperiencesheis
nowhavingiscompatiblewithhismerelydreamingthatheissittingbythefirewitha
pieceofpaperinhishand.Sohedoesnotknowonthebasisofthesensoryexperiences
heishavingatthemomentthatheissittingbythefire.Nor,ofcourse,didthemaninmy
examplesknowthethingshewassaidtoknowonthebasisofthesensoryexperienceshe
washavingatthatmoment.Heknewcertainthingstobeso,andhewasdreamingthose
thingstobeso,butindreamingthemhedidnottherebyknowthemtobeso.
Butaslongasweallowthatthesleepingmandoesknowcertainthingsabouttheworld
aroundhim,evenifhedoesnotknowthemonthebasisoftheverydreamsheishaving
atthemoment,isn'tthatenoughtoshowthatDescartesmustneverthelessbewrongin
hisconclusionthatnoonecanknowanythingabouttheworldaroundhim?No.Itshows
atmostthatwewerehastyorwereignoringDescartes'sconclusioninconcedingthat
someonecouldknowsomethingabouttheworldaroundhim.IfDescartes'sreasoningis
correctthedreamingphysicist,evenwhenheisawake,does(p.17) notreallyknowany
ofthethingswewereuncriticallycreditinghimwithknowingaboutthewaythingsare—
oratleasthedoesnotknowthemonthebasisofthesenses.Inordertoknowthemon
thebasisofthesensestherewouldhavetohavebeenatleastsometimeatwhichhe
knewsomethingaboutwhatwasgoingonaroundhimatthattime.ButifDescartesis
righthecouldnothaveknownanysuchthingunlesshehadestablishedthathewasnot
dreamingatthattime;andaccordingtoDescarteshecouldneverestablishthat.Sothe
factaboutdreamsthatDescartesrelieson—thatonewhodreamsthatsomethingisso
doesnottherebyknowthatitisso—isenoughtoyieldhisconclusioniftheotherstepsof
hisreasoningarecorrect.
WhenhefirstintroducesthepossibilitythathemightbedreamingDescartesseemstobe
relyingonsomeknowledgeabouthowthingsareorwereintheworldaroundhim.He
says‘IremindmyselfthatonmanyoccasionsIhaveinsleepbeendeceivedbysimilar
illusions’,soheseemstoberelyingonsomeknowledgetotheeffectthathehasactually
dreamtinthepastandthatheremembershavingbeen‘deceived’bythosedreams.That
ismorethanheactuallyneedsforhisreflectionsaboutknowledgetohavetheforcehe
thinkstheyhave.Hedoesnotneedtosupporthisjudgementthathehasactuallydreamt
inthepast.Theonlythoughtheneedsisthatitisnowpossibleforhimtobedreaming
thatheissittingbythefire,andthatifthatpossibilitywererealiz edhewouldnotknow
thatheissittingbythefire.OfcourseitwasnodoubttruethatDescarteshaddreamtin
thepastandthathisknowledgethathehaddonesowaspartlywhathewasgoingonin
acknowledgingthepossibilityofhisdreamingonthisparticularoccasion.Butneitherthe Page 12 of 27
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2014.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: University of Sydney; date: 03 February 2015
The Problem of the External World
factofpastdreamsnorknowledgeoftheiractualoccurrencewouldseemtobestrictly
requiredinordertograntwhatDescartesrelieson—thepossibilityofdreaming,andthe
absenceofknowledgeifthatpossibilitywererealiz ed.Thethoughtthathemightbe
dreamingthatheissittingbythefirewithapieceofpaperinhishand,andthefactthatif
hewerehewouldn'tknowhewassittingthere,iswhatgivesDescartespause.That
wouldworryhiminthewayitdoesevenifhehadneveractuallyhadanydreamsexactly
likeit(p.18) inthepast—ifhehadneverdreamtaboutfiresandpiecesofpaperatall.In
fact,Ithinkheneedneverhaveactuallydreamtofanythingbefore,andcertainlyneedn't
knowthatheeverhas,inordertobeworriedinthewayheisbythethoughtthathe mightbedreamingnow.
ThefactthatthepossibilityofdreamingisallDescartesneedstoappealtobringsout
anothertruthaboutdreamsthathisargumentdependson—thatanythingthatcanbe
goingonorthatonecanexperienceinone'swakinglifecanalsobedreamtabout.This
againisonlyastatementofpossibility—nosensiblepersonwouldsuggestthatwedoat
sometimedreamofeverythingthatactuallyhappenstous,orthateverythingwedream
aboutdoesinfacthappensometime.Butitisveryplausibletosaythatthereisnothing
wecouldnotdreamabout,nothingthatcouldbethecasethatwecouldnotdreamtobe
thecase.Isayitisveryplausible;ofcourseIcannotproveittobetrue.Butevenifitis
nottruewithcompletegenerality,wemustsurelygrantthatitispossibletodreamthat
oneissittingbyafirewithapieceofpaperinone'shand,andpossibletodreamof
countlessotherequallyobviousandequallymundanestatesofaffairsaswell,andthose
possibilitiesarewhatDescartesseesasthreateningtohisknowledgeoftheworldaround him.
Thereseemslittlehope,then,ofobjectingthatitissimplynotpossibleforDescartesto
dreamthatheissittingbythefirewithapieceofpaperinhishand.Norisitanymore
promisingtosaythatevenifheweredreamingitwouldnotfollowthathedidnotknow
thathewassittingthere.IthinkboththosestepsorassumptionsofDescartes's
reasoningareperfectlycorrect,andfurtherdefenceofthematthisstageisunnecessary.
Ifhisargumentandtheproblemtowhichitgivesrisearetobeavoided,itmightseem
thatthebesthopeisthereforetoaccepthischallengeandshowthatitcanbemet.That
wouldbeineffecttoarguethatDescartes'salleged‘discovery’isnodiscoveryatall:we
cansometimesknowthatwearenotdreaming.
Thiscaneasilyseemtobethemoststraightforwardandmostpromisingstrategy.It
allowsthatDescartesisrightinthinkingthatknowingthatoneisnotdreamingisa
condition(p.19) ofknowingsomethingabouttheworldaroundus,butwronginthinking
thatthatconditioncanneverbemet.Andthatcertainlyseemsplausible.Surelyitisnot
impossibleformetoknowthatIamnotdreaming?Isn'tthatsomethingIoftenknow,and
isn'titsomethingIcansometimesfindoutifthequestionarises?Ifitis,thenthefactthat
ImustknowthatIamnotdreamingifIamtoknowanythingabouttheworldaroundme
willbenothreattomyknowledgeoftheworld.
Howeverobviousandundeniableitmightbethatweoftendoknowthatwearenot
dreaming,IthinkthisstraightforwardresponsetoDescartes'schallengeisatotalfailure. Page 13 of 27
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2014.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: University of Sydney; date: 03 February 2015
The Problem of the External World
IncallingitstraightforwardImeanthatitacceptsDescartes'sconditionsforknowledgeof
theworldandtriestoshowthattheycanbefulfilled.ThatiswhatIthinkcannotbedone.
Toputthesamepointinanotherway:IthinkDescarteswouldbeperfectlycorrectin
saying‘therearenocertainindicationsbywhichwemayclearlydistinguishwakefulness
fromsleep’,andsowecouldnevertellwearenotdreaming,ifhewerealsorightthat
knowingthatoneisnotdreamingisaconditionofknowingsomethingabouttheworld
aroundus.ThatiswhyIthinkonecannotacceptthatconditionandthengoontoestablish
thatoneisnotdreaming.IdonotmeantobesayingsimplythatDescartesisright—that
wecanneverknowthatwearenotdreaming.ButIdowanttoarguethateitherwecan
neverknowthatwearenotdreamingorelsewhatDescartessaysisaconditionof
knowingthingsabouttheworldisnotreallyaconditioningeneralofknowingthingsabout
theworld.Thestraightforwardstrategydeniesbothalternatives.Iwilltrytoexplainwhy
Ithinkwemustacceptonealternativeortheother.
WhenDescartesaskshimselfhowheknowsthatheissittingbythefirewithapieceof
paperinhishandwhydoesheimmediatelygoontoaskhimselfhowheknowsheisnot
dreamingthatheissittingbythefirewithapieceofpaperinhishand?Ihavesuggested
thatitisbecauseherecognizesthatifheweredreaminghewouldnotknowonthebasis
ofhissensesatthemomentthatheissittingthere,andsohethinkshemustknowthat
thatpossibilitydoesnotobtainifheistoknowthatheisinfactsittingthere.(p.20) But
thisparticularexamplewaschosen,notforanypeculiaritiesitmightbethoughtto
possess,butbecauseitcouldbetakenastypicalofthebestpositionwecaneverbein
forcomingtoknowthingsabouttheworldaroundusonthebasisofthesenses.Whatis
trueofthiscasethatisrelevanttoDescartes'sinvestigationofknowledgeissupposedto
betrueofallcasesofknowledgeoftheworldbymeansofthesenses;thatiswhythe
verdictarrivedatherecanbetakentobetrueofoursensoryknowledgegenerally.But
whatDescartesthinksistrueofthisparticularcaseofsensoryknowledgeoftheworldis
thathemustknowheisnotdreamingifheistoknowthatheissittingbythefirewitha
pieceofpaperinhishand.Thatisrequired,notbecauseofanypeculiaritiesofthis
particularcase,butpresumablybecause,accordingtoDescartes,itisanecessary
conditionofanycase—evenabestpossiblecase—ofknowledgeoftheworldbymeansof
thesenses.ThatiswhyIascribedtoDescartesthequitegeneralthesisthatknowingthat
oneisnotdreamingisaconditionofknowingsomethingabouttheworldarounduson
thebasisofthesenses.Sincehethinksthepossibilityofhisdreamingmustberuledout
inthecaseheconsiders,andthecaseheconsidersisregardedastypicalandwithout
specialcharacteristicsofitsown,hethinksthatthepossibilitythatheisdreamingmustbe
ruledoutineverycaseofknowingsomethingabouttheworldbymeansofthesenses.
Ifthatreallyisaconditionofknowingsomethingabouttheworld,Ithinkitcanbeshown
thatDescartesisrightinholdingthatitcanneverbefulfilled.Thatiswhatthe
straightforwardresponsedenies,andthatiswhyIthinkthatresponsemustbewrong.
WecannotacceptthetermsofDescartes'schallengeandthenhopetomeetit.
SupposeDescartestriestodeterminethatheisnotdreaminginordertofulfilwhathe
seesasanecessaryconditionofknowingthatheissittingbythefirewithapieceofpaper Page 14 of 27
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2014.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: University of Sydney; date: 03 February 2015
The Problem of the External World
inhishand.Howishetoproceed?Herealiz esthathisseeinghishandandseeingand
feelingapieceofpaperbeforehimandfeelingthewarmthofthefire—infacthisgettingall
thesensoryexperiencesorallthesensoryinformationheisthengetting—issomething
thatcouldbehappeningeven(p.21) ifheweredreaming.Toestablishthatheisnot
dreaminghewouldthereforeneedsomethingmorethanjustthoseexperiencesorthat
informationalone.Hewouldalsoneedtoknowwhetherthoseexperiencesandthat
informationarereliable,notmerelydreamt.Ifhecouldfindsomeoperationortest,orif
hecouldfindsomecircumstanceorstateofaffairs,thatindicatedtohimthathewasnot
dreaming,perhapshecouldthenfulfilthecondition—hecouldknowthathewasnot
dreaming.Buthowcouldatestoracircumstanceorastateofaffairsindicatetohimthat
heisnotdreamingifaconditionofknowinganythingabouttheworldisthatheknowheis
notdreaming?Itcouldnot.Hecouldneverfulfilthecondition.
Letussupposethatthereisinfactsometestwhichapersoncanperformsuccessfully
onlyifheisnotdreaming,orsomecircumstanceorstateofaffairswhichobtainsonlyif
thatpersonisnotdreaming.Ofcourseforthattestorstateofaffairstobeofanyuseto
himDescarteswouldhavetoknowofit.Hewouldhavetoknowthatthereissuchatest
orthatthereisastateofaffairsthatshowsthatheisnotdreaming;withoutsuch
informationhewouldbenobetterofffortellingthatheisnotdreamingthanhewouldbe
iftherewerenosuchtestorstateofaffairsatall.Tohaveacquiredthatinformationhe
wouldatsometimehavetohaveknownmorethanjustsomethingaboutthecourseof
hissensoryexperience,sincetheconnectionbetweentheperformanceofacertaintest,
orbetweenacertainstateofaffairs,andsomeone'snotdreamingisnotitselfjustafact
aboutthecourseofthatperson'ssensoryexperience;itisafactabouttheworldbeyond
hissensoryexperiences.Nowstrictlyspeakingifitisaconditionofknowinganything
abouttheworldbeyondone'ssensoryexperiencesthatoneknowthatoneisnot
dreaming,thereisanobviousobstacletoDescartes'severhavinggottheinformationhe
needsaboutthattestorstateofaffairs.Hewouldhavetohaveknownatsometimethat
hewasnotdreaminginordertogettheinformationheneedstotellatanytimethatheis
notdreaming—andthatcannotbedone.
ButsupposeweforgetaboutthisdifficultyandconcedethatDescartesdoesindeed
know(somehow)thatthereis(p.22) atestorcircumstanceorstateofaffairsthat
unfailinglyindicatesthatheisnotdreaming.Still,thereisanobstacletohiseverusingthat
testorstateofaffairstotellthatheisnotdreamingandtherebyfulfillingtheconditionfor
knowledgeoftheworld.Thetestwouldhavetobesomethinghecouldknowhehad
performedsuccessfully,thestateofaffairswouldhavetobesomethinghecouldknow
obtains.Ifhecompletelyunwittinglyhappenedtoperformthetest,orifthestateofaffairs
happenedtoobtainbuthedidn'tknowthatitdid,hewouldbeinnobetterpositionfor
tellingwhetherhewasdreamingthanhewouldbeifhehaddonenothingordidnoteven
knowthattherewassuchatest.Buthowishetoknowthatthetesthasbeenperformed
successfullyorthatthestateofaffairsinquestiondoesinfactobtain?Anythingonecan
experienceinone'swakinglifecanalsobedreamtabout;itispossibletodreamthatone
hasperformedacertaintestordreamthatonehasestablishedthatacertainstateof
affairsobtains.And,aswehaveseen,todreamthatsomethingabouttheworldaround Page 15 of 27
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2014.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: University of Sydney; date: 03 February 2015
The Problem of the External World
youissoisnottherebytoknowthatitisso.Inordertoknowthathistesthasbeen
performedorthatthestateofaffairsinquestionobtainsDescarteswouldthereforehave
toestablishthatheisnotmerelydreamingthatheperformedthetestsuccessfullyor
thatheestablishedthatthestateofaffairsobtains.Howcouldthatinturnbeknown?
Obviouslytheparticulartestorstateofaffairsalreadyinquestioncannotserveasa
guaranteeofitsownauthenticity,sinceitmighthavebeenmerelydreamt,sosome
furthertestorstateofaffairswouldbeneededtoindicatethattheoriginaltestwas
actuallyperformedandnotmerelydreamt,orthatthestateofaffairsinquestionwas
actuallyascertainedtoobtainandnotjustdreamttoobtain.Butthisfurthertestorstate
ofaffairsissubjecttothesamegeneralconditioninturn.Everypieceofknowledgethat
goesbeyondone'ssensoryexperiencesrequiresthatoneknowoneisnotdreaming.
ThissecondtestorstateofaffairswillthereforebeofuseonlyifDescartesknowsthathe
isnotmerelydreamingthatheisperformingorascertainingit,sincemerelytodream
thathehadestablishedtheauthenticityofthefirsttestisnottohaveestablishedit.And
soon.Atnopointcanhefindatestfor(p.23) notdreamingwhichhecanknowhasbeen
successfullyperformedorastateofaffairscorrelatedwithnotdreamingwhichhecan
knowobtains.HecanthereforeneverfulfilwhatDescartessaysisanecessarycondition
ofknowingsomethingabouttheworldaroundhim.Hecanneverknowthatheisnot dreaming.
Imustemphasiz ethatthisconclusionisreachedonlyontheassumptionthatitisa
conditionofknowinganythingabouttheworldaroundusonthebasisofthesensesthat
weknowwearenotdreamingthatthethingisso.Ithinkitishisacceptanceofthat
conditionthatleadsDescartesto‘seesomanifestlythattherearenocertainindications
bywhichwemayclearlydistinguishwakefulnessfromsleep’.AndIthinkDescartesis
absolutelyrighttodrawthatconclusion,givenwhathethinksisaconditionofknowledge
oftheworld.ButallIhavearguedonDescartes'sbehalf(heneverspellsouthis
reasoning)isthatwecannotbothacceptthatconditionofknowledgeandhopetofulfilit,
asthestraightforwardresponsehopestodo.Andofcourseifoneofthenecessary
conditionsofknowledgeoftheworldcanneverbefulfilled,knowledgeoftheworld
arounduswillbeimpossible.
IthinkwehavenowlocatedDescartes'sreasonforhisnegativeverdictaboutsensory
knowledgeingeneral.Ifweagreethathemustknowthatheisnotdreamingifheisto
knowinhisparticularcasethatheissittingbythefirewithapieceofpaperinhishand,
wemustalsoagreethatwecanknownothingabouttheworldaroundus.
Oncewerecogniz ethattheconditionDescartestakesasnecessarycanneverbefulfilled
ifheisrightinthinkingitisindeednecessary,wearenaturallyledtothequestion
whetherDescartesisright.Isitreallyaconditionofknowingsomethingabouttheworld
thatoneknowoneisnotdreaming?ThatisthesecondofthethreequestionsI
distinguished.Itistheonethathasreceivedtheleastattention.InaskingitnowIdonot
meantobegoingbackonsomethingIsaidearlierwasundeniablytrue,viz.,thatifoneis
dreamingthatsomethingabouttheworldissoonedoesnottherebyknowthatitisso.
Thatstillseemstomeundeniable,butitisnotthesameasDescartes'sassumptionthat Page 16 of 27
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2014.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: University of Sydney; date: 03 February 2015
The Problem of the External World
one(p.24) mustknowthatoneisnotdreamingifoneistoknowsomethingaboutthe
world.Theundeniabletruthsaysonlythatyoulackknowledgeifyouaredreaming;
Descartessaysthatyoulackknowledgeifyoudon'tknowthatyouarenotdreaming.
Onlywiththestrongerassumptioncanhisscepticalconclusionbereached.
Isthatassumptiontrue?InsofaraswefindDescartes'sreasoningconvincing,oreven
plausible,Ithinkitisbecausewetooonreflectionfindthatitistrue.Isaidthatnotmuch
attentionhadbeenpaidtothatparticularpartofDescartes'sreasoning,andIthinkthat
tooisbecause,ashepresentsit,thestepseemsperfectlyconvincingandsoonlyother
partsoftheargumentappearvulnerable.Whyisthatso?IsitbecauseDescartes's
assumptionisindeedtrue?Isthereanythingwecandothatwouldhelpusdetermine
whetheritistrueornot?ThequestionisimportantbecauseIhavearguedsofarthatifit
istruewecanneverknowanythingabouttheworldaroundusonthebasisofthe
senses,andphilosophicalscepticismabouttheexternalworldiscorrect.Wewouldhave
tofindthatconclusionasconvincingorasplausibleaswefindtheassumptionfromwhich itisderived.
GivenouroriginalfavourableresponsetoDescartes'sreasoning,then,itcanscarcelybe
deniedthatwhatIhavecalledhisassumptionorconditionseemsperfectlynaturalto
insiston.Perhapsitseemslikenothingmorethananinstanceofafamiliarcommonplace
aboutknowledge.Weareallawarethat,eveninthemostordinarycircumstanceswhen
nothingveryimportantturnsontheoutcome,wecannotknowaparticularthingunless
wehaveruledoutcertainpossibilitiesthatwerecogniz eareincompatiblewithour knowingthatthing.
SupposethatonlookingoutthewindowIannouncecasuallythatthereisagoldfinchin
thegarden.IfIamaskedhowIknowitisagoldfinchandIreplythatitisyellow,weall
recogniz ethatinthenormalcasethatisnotenoughforknowledge.‘Forallyou'vesaidso
far,’itmightbereplied,‘thethingcouldbeacanary,sohowdoyouknowit'sa
goldfinch?’.AcertainpossibilitycompatiblewitheverythingIhavesaidsofarhasbeen
raised,andifwhat(p.25) IhavesaidsofarisallIhavegottogoonandIdon'tknow
thatthethinginthegardenisnotacanary,thenIdonotknowthatthereisagoldfinchin
thegarden.ImustbeabletoruleoutthepossibilitythatitisacanaryifIamtoknowthat
itisagoldfinch.Anyonewhospeaksaboutknowledgeandunderstandswhatotherssay
aboutitwillrecogniz ethisfactorconditioninparticularcases.
Inthisexamplewhatissaidtobepossibleissomethingincompatiblewiththetruthof
whatIclaimtoknow—ifthatbirdwereacanaryitwouldnotbeagoldfinchinthegarden,
butacanary.WhatIbelieveinbelievingitisagoldfinchwouldbefalse.Butthatisnotthe
onlywayapossibilitycanworkagainstmyknowledge.IfIcometosuspectthatallthe
witnesseshaveconspiredandmadeupastoryabouttheman'sbeinginClevelandthat
night,forexample,andtheirtestimonyisallIhavegottogooninbelievingthathewasin
Cleveland,ImightfindthatInolongerknowwhetherhewasthereornotuntilIhave
somereasontoruleoutmysuspicion.IftheirtestimonywereallinventedIwouldnot
knowthatthemanwasinCleveland.ButstrictlyspeakinghisbeinginClevelandisnot
incompatiblewiththeirmakingupastorysayinghewas.Theymighthaveinventeda Page 17 of 27
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2014.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: University of Sydney; date: 03 February 2015
The Problem of the External World
storytoprotecthim,whereasinfact,unknowntothem,hewasthereallthetime.Sucha
complicatedplotisnotnecessarytobringoutthepoint;Moore'sDukeofDevonshireis
enough.FromthefactthathewasdreamingthathewasspeakingintheHouseofLords
itdidnotfollowthathewasnotspeakingintheHouseofLords.Infacthewas.The
possibilityofdreaming—whichwasactualinthatcase—didnotimplythefalsityofwhat
wasbelieved.Apossibledeficiencyinthebasisofmybeliefcaninterferewithmy
knowledgewithoutitselfrenderingfalsetheverythingIbelieve.Ahallucinogenicdrug
mightcausemetoseemybedcoveredwithahugepileofleaves,forexample.5Having
takenthatdrug,IwillknowtheactualstateofmybedonlyifIknowthatwhatIseeisnot
justtheeffectofthedrug;ImustbeabletoruleoutthepossibilitythatIamhallucinating
thebedandtheleaves.Buthoweverimprobable(p.26) itmightbethatmybedis
actuallycoveredwithleaves,itsnotbeingcoveredwithleavesdoesnotfollowfromthe
factthatIamhallucinatingthatitis.WhatIamhallucinatingcouldneverthelessbe
(unknowntome)true.Butagoldfinchsimplycouldnotbeacanary.Soalthoughthereare
twodifferentwaysinwhichacertainpossibilitycanthreatenmyknowledge,itremains
truethattherearealwayscertainpossibilitieswhichmustbeknownnottoobtainifIam
toknowwhatIclaimtoknow.
Ithinkthesearejustfamiliarfactsabouthumanknowledge,somethingweallrecogniz e
andabidebyinourthoughtandtalkaboutknowingthings.Weknowwhatwouldbea
validchallengetoaclaimtoknowsomething,andwecanrecogniz etherelevanceand
forceofobjectionsmadetoourclaimstoknow.Thequestionbeforeusistowhatextent
Descartes'sinvestigationofhisknowledgethatheissittingbythefirewithapieceof
paperinhishandfollowstheserecogniz edeverydayproceduresforassessingclaimsto
know.Ifitdoesfollowthemfaithfully,andyetleadstotheconclusionthathecannotknow
whereheisorwhatishappeningaroundhim,weseemforcedtoaccepthisnegative
conclusionaboutknowledgeingeneraljustasweareforcedtoaccepttheconclusionthat
IdonotknowitisagoldfinchordonotknowthewitnesswasinClevelandbecauseI
cannotruleoutthepossibilitieswhichmustberuledoutifIamtoknowsuchthings.Is
Descartes'sintroductionofthepossibilitythathemightbedreamingjustlikethe
introductionofthepossibilitythatitmightbeacanaryinthegardenorthatthealibimight
becontrivedorthatitmightbeahallucinationofmybedcoveredwithleaves?
ThosepossibilitieswereallsuchthatiftheyobtainedIdidnotknowwhatIclaimedto
know,andtheyhadtobeknownnottoobtaininorderfortheoriginalknowledge‐claimto
betrue.DoesDescartes'sdream‐possibilityfulfilbothofthoseconditions?Ihavealready
saidthatitseemsundeniablethatitfulfilsthefirst.IfheweredreamingDescarteswould
notknowwhatheclaimstoknow.Someonewhoisdreamingdoesnottherebyknow
anythingabouttheworldaroundhimeveniftheworldaroundhimhappenstobejust
thewayhedreamsorbelievesittobe.Sohisdreamingisincompatible(p.27) withhis
knowing.Butdoesitfulfilthesecondcondition?Isitapossibilitywhichmustbeknownnot
toobtainifDescartesistoknowthatheissittingbythefirewithapieceofpaperinhis
hand?Ithinkitisdifficultsimplytodenythatitis.TheevidentforceofDescartes's
reasoningwhenwefirstencounteritisenoughtoshowthatitcertainlystrikesusasa
relevantpossibility,assomethingthatheshouldknownottoobtainifheistoknowwhere Page 18 of 27
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2014.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: University of Sydney; date: 03 February 2015
The Problem of the External World
heisandwhatishappeningaroundhim.
WhenthatpossibilitystrikesusasobviouslyrelevantinDescartes'sinvestigationwe
mightcometothinkthatitisbecauseofasimpleandobviousfactaboutknowledge.In
thecaseofthegoldfinchweimmediatelyrecogniz ethatImustknowthatitisnotacanary
ifIamtoknowitisagoldfinch.Anditisverynaturaltothinkthatthatissimplybecauseits
beingacanaryisincompatiblewithitsbeingagoldfinch.Ifitwereacanaryitwouldnotbe
agoldfinch,andIwouldthereforebewronginsayingthatitis;soifIamtoknowitisa
goldfinchImustruleoutthepossibilitythatitisacanary.Theideaisthatthetwo
conditionsIdistinguishedinthepreviousparagrapharenotreallyseparateafterall.As
soonasweseethatacertainpossibilityisincompatiblewithourknowingsuch‐and‐such,it
issuggested,weimmediatelyrecogniz ethatitisapossibilitythatmustbeknownnotto
obtainifwearetoknowthesuch‐and‐suchinquestion.Weseethatthedream‐possibility
satisfiesthatfirstconditioninDescartes'scase(ifheweredreaming,hewouldn'tknow),
andthatiswhy,accordingtothissuggestion,weimmediatelyseethatitisrelevantand
mustberuledout.Somethingweallrecogniz eaboutknowledgeiswhatissaidtomake thatobvioustous.
Butisthe‘simpleandobviousfactaboutknowledge’appealedtointhisexplanationreally
somethingthatistrueofhumanknowledgeeveninthemostordinarycircumstances?
Whatexactlyisthe‘fact’inquestionsupposedtobe?Ihavedescribeditsofar,asapplied
tothecaseofthegoldfinch,asthefactthatifIknowsomethingp(it'sagoldfinch)Imust
knowthefalsityofallthosethingsincompatiblewithp(e.g.,it'sacanary).Iftherewere
oneofthosethingsthatIdidnotknowtobefalse,anditwereinfacttrue,Iwouldnot
knowthatp,sinceinthatcasesomething(p.28) incompatiblewithpwouldbetrueand
sopwouldnotbetrue.ButtosaythatImustknowthatallthosethingsincompatiblewith
parefalseisthesameassayingthatImustknowthetruthofallthosethingsthatmust
betrueifpistrue.Anditisextremelyimplausibletosaythatthatisa‘simpleandobvious
fact’weallrecogniz eabouthumanknowledge.
Thedifficultyisthattherearenodeterminatelimitstothenumberofthingsthatfollow
fromthethingsIalreadyknow.ButitcannotbesaidthatInowknowallthose
indeterminatelymanythings,althoughtheyallmustbetrueifthethingsthatIalready
knowaretrue.EvengrantingthatInowknowagreatdealaboutalotofdifferentthings,
myknowledgeobviouslydoesnotextendtoeverythingthatfollowsfromwhatInow
know.Ifitdid,mathematics,totakeonlyoneexample,wouldbeagreatdealeasierthanit
is—orelseimpossiblydifficult.Inknowingthetruthofthesimpleaxiomsofnumber
theory,forexample,Iwouldtherebyknowthetruthofeverythingthatfollowsfrom
them;everytheoremofnumbertheorywouldalreadybeknown.Or,takingthe
pessimisticside,sinceobviouslynoonedoesknowallthetheoremsofnumbertheory,it
wouldfollowthatnooneevenknowsthatthosesimpleaxiomsaretrue.
Itisabsurdtosaythatweenjoyorrequiresuchvirtualomniscience,soitismore
plausibletoholdthatthe‘simpleandobviousfact’weallrecogniz eaboutknowledgeis
theweakerrequirementthatwemustknowthefalsityofallthosethingsthatweknowto
beincompatiblewiththethingsweknow.Iknowthatabird'sbeingacanaryis Page 19 of 27
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2014.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: University of Sydney; date: 03 February 2015
The Problem of the External World
incompatiblewithitsbeingagoldfinch;thatisnotsomefar‐flung,unknownconsequence
ofitsbeingagoldfinch,butsomethingthatanyonewouldknowwhoknewanythingabout
goldfinchesatall.AndtheideaisthatthatiswhyImustknowthatitisnotacanaryifIam
toknowthatitisagoldfinch.Perhaps,inordertoknowsomething,p,Idonotneedto
knowthefalsityofallthosethingsthatareincompatiblewithp,butitcanseemthatat
leastImustknowthefalsityofallthosethingsthatIknowtobeincompatiblewithp.Since
Iclaimtoknowthatthebirdisagoldfinch,andIknowthatitsbeingagoldfinchimplies
thatitisnotacanary,(p.29) Imustforthatreasonknowthatitisnotacanaryifmy
originalclaimistrue.InclaimingtoknowitisagoldfinchIwas,sotospeak,committing
myselftoknowingthatitisnotacanary,andImusthonourmycommitments.
Thisrequirementasitstands,evenifitdoesexplainwhyImustknowthatthebirdisnot
acanary,doesnotaccountfortherelevanceoftheothersortsofpossibilitiesIhave
mentioned.ThereasoninthegoldfinchcasewassaidtobethatIknowthatitsbeinga
canaryisincompatiblewithitsbeingagoldfinch.ButthatwillnotexplainwhyImustrule
outthepossibilitythatthewitnesseshaveinventedastoryabouttheman'sbeingin
Cleveland,orthepossibilitythatIamhallucinatingmybedcoveredwithapileofleaves.
NorwillitexplainwhyDescartesmustruleoutthepossibilitythatheisdreaming.WhatI
claimedtoknowinthefirstcaseisthatthemanwasinClevelandthatnight.But,aswe
sawearlier,itisnotaconsequenceofhisbeinginClevelandthatnoonewillinventastory
totheeffectthathewasinCleveland;theymightmistakenlybelievehewasnotthereand
thentellwhattheythinkisalie.Norisitaconsequenceofmybed'sbeingcoveredwith
leavesthatIamnothallucinatingthatitis.Butwerecognizethatinordertoknowinthose
casesIneverthelesshadtoruleoutthosepossibilities.Similarly,astheDukeof
Devonshireremindsus,itisnotaconsequenceofDescartes'ssittingbythefirewitha
pieceofpaperinhishandthatheisnotdreamingthatheis.Soifitisobvioustousthat
Descartesmustknowthatheisnotdreamingifheistoknowthatheissittingbythefire,
itcannotbesimplybecausethepossibilityinquestionisknowntobeincompatiblewith
whatheclaimstoknow.Itisnot.
Ifthereissome‘simpleandobviousfactaboutknowledge’thatwerecogniz eandrelyon
inrespondingtoDescartes'sreasoningitmustthereforebemorecomplicatedthanwhat
hasbeensuggestedsofar.Reflectingevenontheuncontroversialeverydayexamples
alonecaneasilyleadustosupposethatitissomethinglikethis:ifsomebodyknows
something,p,hemustknowthefalsityofallthosethingsincompatiblewithhisknowing
thatp(orperhapsallthosethingsheknowstobeincompatiblewithhisknowingthat
(p.30) p).Iwillnotspeculatefurtheronthequalificationsoremendationsneededto
maketheprinciplelessimplausible.Thequestionnowiswhetheritisouradherenceto
anysuchprincipleorrequirementthatisresponsibleforourrecognitionthatthe
possibilitythatthebirdisacanaryorthepossibilitythatthewitnessesmadeupastory
mustbeknownnottoobtainifIamtoknowthethingsIsaidIknewinthosecases.What
exactlyaretheproceduresorstandardswefollowinthemostordinary,humdrumcases
ofputativeknowledge?ReflectiononthesourceofDescartes'sscepticalreasoninghas
ledtodifficultiesindescribingandthereforeinunderstandingeventhemostfamiliar
procedureswefollowineverydaylife.Thatisoneoftherewardsofastudyof Page 20 of 27
PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2014.
All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a
monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: University of Sydney; date: 03 February 2015