Introduction
Asyoumayrecall,whenAlicesteppedthroughthemirror,sheencountered averypeculiarworldinwhichmanyofthepeopleshemetwerechesspieces orcharactersfromnurseryrhymes.Everythingwasinsideout,upsidedown. Soitiswithconsciousness.
Reflectionupontheappearanceofconsciousnessinlivingbeingssuggests thattherearejusttwoalternativeviews.Eitherconsciousnessappeared suddenlysothatitsappearanceislikethatofalightswitchbeingturned onoritarosethroughintermediatestages,notyetdefinitelyinvolving consciousnessbutalsonotdefinitelynotinvolvingit.Ontheformerview, consciousnessisanon/offmatter,butonceitarose,itbecamericherand richerthroughtimeratherasabeamoflightmaybecomebrighterand broaderinitssweep.Onthelatterview,consciousnessisnotanon/off matter.Thereareshadesofgray.Thereisnoonemomentatwhich consciousnessappeared.Itarosegraduallyjustaslifedid,becomingricher throughtimeasanimalbrainsbecamemorecomplex.
Thelatterviewseemsmoreplausibleat firstglance;forifconsciousness suddenlyappearedoutoftheblue,asitwere,thenwhatwasresponsiblefor itssuddenemergence?Presumablytheoccurrenceofsomesuitableneural state.Butneurologicalstatesthemselvesadmitofborderlinecases,sothe relevantneuralstatecannotitselfhavearisensuddenly.Insteaditmusthave appearedgraduallythroughvariousintermediateneurologicalstages.So,if consciousnessoriginallyappearedsuddenlywithoutanyborderlinecases,it cannotbeidentifiedwithanysuchneurologicalstate;norforsimilarreasons canitbeidentifiedwithanycomplexfunctionalorinformationalstate supportedbytheneurologicalarchitecture.Itseems,then,thatifconsciousnessappearedsuddenly,itmustbesomethingspecialandnew,totally differentfromthephysicalpropertiesoftheunderlyingneuralandfunctionalarchitecture.Butifthisisthecase,whatcouldconsciousnessbe?It appearsthatwearedriventothinkofconsciousnessassomethingnonphysicalinnaturethatsuddenlyemergedincertainanimalbrainswithoutany furtherexplanation.Thisisveryhardtoaccept.
Ontheotherhand,ifconsciousnessarosegraduallythenweshouldbe abletodescribeborderlinecasesofconsciousnessjustaswecanforlife. Unfortunately,asIshallargue,thatwecannotdo.Putativeborderlinecases ofconsciousnessareallcasesinwhichthereisindeterminacyin what is experienced,andnotinexperienceorconsciousnessitself.So,akindof paradoxarises.Consciousnesscannotbesharporprecise,butequallyit cannotbevague.
Theparadoxislaidoutfullyinthe firstchapter.Onepossibleresponseto thisparadoxistosaythatitisbasedonamistakenassumptionaboutthe originsofconsciousstates.Consciousstatesdidnotarisewithneurological complexity.Instead,theyarefundamentalfeaturesofmicrophysicalreality (panpsychism)oratleasttheyaregroundedinsuchfundamentalfeatures. Chapter2laysoutthestandardversionofthisview:RussellianMonism. Iarguethattheview,ineitherofitstwostandardelaborations,faces overpoweringobjections.
Inthenexttwochapters,Idiscusstherelationshipbetweenconscious statesandconsciousnessitself.Iarguethatthebasictenetoftherepresentationalistviewofconsciousstatescanbepreservedwithinaframeworkthat takesconsciousnessitself,orratheracentralelementofconsciousnessIcall “consciousness*”,tobesharpbutconsciousstatesvague.Consciousness*, Iclaim,isindeedafundamentalfeatureofmicro-reality,andthusitdidnot evolve,butconsciousstatesarenot.Consciousstatesevolvedgradually,as didlife,througharangeofborderlinecases.TheviewwithwhichIendup presentsnovelsolutionstothreeimportantproblems(theproblemof undirectedconsciousness,theproblemofcombination,andtheproblem oftiny,psychologicalsubjects).Italsotakesupthequestionofhowconsciousnesscanbecausallyefficaciouswithrespecttoanimalbehavior.
ThatIampreparedtoembraceapositionthathassomethingincommon withthepanpsychistworldviewwillcomeasasurprisetomany,givenmy pastwritingsonconsciousness,butasJohnPerryquipped: “Ifyouthink aboutconsciousnesslongenough,youeitherbecomeapanpsychistoryou gointoadministration,” andIhaven’tgoneintoadministration.Icannotsay thatthetransitionhasbeenaneasyone.But,torepeat,Iamstilla representationalistaboutconsciousness.Iamalsostillaphysicalist.Andit isconsciousness*,notconsciousness,Imaintain,thatistobefoundinthe micro-realm.So,thechangeisnotquiteasradicalordramaticasitmay firstseem.
Chapter5turnstothequestionofwhereinthebrainmacroconsciousnessislocatedandwhichanimalbrainssoevolvedastosupport
consciousstates.Itissuggestedherethateventhoughconsciousstates appearedgradually,ontheaccountIamoffering,itmaywellbetruethat inhumanbrainsandthoseofmanyotherspecies,thereisatriggerfor consciousstatesthat typically (thoughnotalways)turnssuchstatesonoroff andsofunctionsinthesamegeneralwayasalightswitch.
Theworldisastrangeplace,ifyoulookintoitdeeplyenough,notasfar removedfromtheworldAliceencounteredthroughthelookingglassaslay peoplesuppose.Weknowthatalreadyfromtheoriesinphysicswhichtellus thatmicrophysicalentitiesarebothwavesandparticles,thattherecanbe actionatahugedistance(onesogreatthattherecannotacausalconnection, asinquantumentanglement),andthattimeisdependentonaframeof reference.Perhapsitisonly fittingthatconsciousnessshouldturnouttobe strangetoo.
IamgratefultoaudiencesattalksintheUSA,theUK,andChinafor commentsanddiscussion.IwouldliketothankspecificallyDerekBall,Zack Blaesi,PaulBoghossian,JaneChen,DavidChalmers,AlexGrzankowski, KeithHossack,CheyenneHowell,JonLitland,DavidPapineau,Simon Prosser,ConnorQuinn,MarkSainsbury,HenryShevlin,andJonathan Simon.IamalsoindebtedtotworefereesforOxfordUniversityPress, whogavememuchfoodforthoughtintheirdetailedreports.
AParadoxofConsciousness
Somephilosophersandscientistshavelikenedtheappearanceofconsciousnessinlivingbeingstothatofalightswitchbeingturnedon. Consciousness,onthisview,suddenlyappearedandthenitbecamericher andricherthroughtimeratherasabeamoflightmaybecomebrighter andbroaderinitssweep(see,forexample,Searle1992).Onsuchapicture, eitherconsciousnessispresentoritisn’t.Therearenoshadesofgray. Butonceconsciousnesshasemerged,therearedifferentdegreesof consciousness.
Othershavesaidthatthelightswitchmodelforconsciousnessisfundamentallymisconceived.Consciousnessis not anon/offmatter.Ratherconsciousnessarosegraduallyjustaslifedid(see,forexample,Lycan1996; Dennett2004).Advocatesofthisviewdenythatconsciousnessisanall-ornothingphenomenon,eveninourowncase.
Onewaytoputthedisagreementhereisintermsofvagueness.Typically vaguenessisunderstoodintermsofborderlinecases.Herearesomerepresentativequotations:
Howeverborderlinecasesshouldbecharacterized,itisadatumthatvague conceptsgiverisetothem.(Wright2003,p.93)
Itisbettertodefineapredicateasvagueifandonlyifitiscapableof yieldingborderlinecases,wherethenotionofborderlinecasesisintroducedbyexamples.(Williamson1994,p.171)
Whatdoesitmeantosaythat ‘bald’ isvague?Presumablyitmeansthatthe predicateadmitsborderlinecases.(Field1994,p.410)
Noticethatintheabovequotations,predicatesandconceptsareclassifiedas vague.Thisispartandparcelofthecommonviewthatvaguenessis ultimatelyalinguisticorconceptualphenomenon.Butwecanalsosensibly askwhether,forexample,thepropertyofbeingbaldorthepropertyof beingredadmitofborderlinecasesandaretherebyvague.Andtheanswer tothesequestionsseemsclearly ‘ yes ’.Theboundarybetweenredandorange
isfuzzy.Objectshavingacolorinthatregionareneitherdefinitelyrednor definitelynotred,andsoareborderlinered.Likewise,somepeoplewith smallpatchesofhairontheirheaddonothavesufficientlyfewhairsto countasdefinitelybaldbutthenneitheraretheydefinitelynotbald.They areinthegrayarea.
Wecannowputthedisagreementaboutconsciousnessintermsofthe followingquestion.Isconsciousnesslikebeingbaldandbeingredinbeing vague,thatis,inallowingborderlinecasesorisconsciousnesslikebeingan evennumberinbeing sharp,thatis,innotbeingcapableofhavingborderlinecases?Searleholdsthatconsciousnessissharp,asdoMcGinn(1982), andSimon(2017).¹DennettandLycanholdthatconsciousnessisvagueas doesPapineau(2002)andasdidIpreviously(Tye1996).
Thereisonefurtherpointofclarificationneeded.Consciousness,for presentpurposes,isexperience.Experiencesarementalstatessuchthat thereisinherentlysomethingitislikesubjectivelytoundergothem. Examplesarefeelingpain,feelinganitch,visualizinganelephant,experiencinganger,andfeelingfearful.Ineachofthesecases,itisincoherentto supposethatthestateexistswithouttherebeingsomephenomenology, somesubjectiveorfeltcharacter.Thustosaythatastateisconsciousin thepresentcontextjustistosaythatitisanexperience;andtoconsider whetherconsciousnessisvagueistoconsiderwhethertherecanbeborderlinecasesofexperience.
Inunderstandingtheterm ‘consciousness ’ inthisway,Idonotmeanto suggestthatthetermhasnothadotherusesbothinscienceandphilosophy.Sometimes,forexample,itisheldthatamentalstateisconsciousjust incaseitisoneofwhichitssubjectisintrospectivelyaware.Thisis sometimescalled ‘higher-orderconsciousness’.Myclaimissimplythat amongthevariousmentalstatesweundergo,manyofwhichareintrospectivelyaccessible(butarguablynotall),areexperiencesandfeelings,and thesestates,unlikebeliefs,forexample,areinherentlysuchthatthey feel a certainway.Differentexperiencesdifferinhowtheyfeel,intheirsubjective character,andthatiswhatmakesthemdifferentexperiences.Inbeingthis way,experiencesareconsciousmentalstatesbytheirverynature.This pointissometimesputbysayingthatexperiencesare phenomenally conscious.²
¹Antony(2006)hasacomplexviewwhichallowsthatconsciousnessmayturnouttobe vagueeventhoughourcurrentconceptofitissharp.Moreonthisbelow,pp.19–20.
²Invariousplacesinthisbook,Iadoptthisusagemyself.
TheparadoxIwishtodiscussisaparadoxaboutphenomenalconsciousness.Itcannowbestatedasfollows:
1)Consciousnessiseithersharporvague.
2)Ifconsciousnessissharp,thenitisn’ta(broadly)physicalphenomenon.
3)Consciousnessisa(broadly)physicalphenomenon.
4) Consciousnessisvague (from1,2,3).
5)Ifconsciousnessisvague,thentherearepossibleborderlinecasesof consciousness.
6)Therearenopossibleborderlinecasesofconsciousness.
7) Consciousnessisnotvague (from4,5,6).
8) Consciousnessisbothvagueandnotvague (from4,7).
Acontradiction!
1.1TheParadoxExplained:PartA
Premise(1)oftheparadoxisaninstanceofthelogicallawoftheexcluded middle.Eitherconsciousnessissharporitisnotsharp,thatis,itisvague. Premise(2)issometimestakentobeanobvious,nonlogicaltruth.Colin McGinn,forexample,says:
Whatevertheexplanation[oftheall-or-nothingcharacterofconsciousness]is whetherindeedtheall-or-nothingcharacterofconsciousnesscan beexplained thisseemstobeafeaturethatanyaccountofconsciousness mustrespect.Andtherearetheoriesofthemind,suchasmaterialismand behaviorism,thatwill findthisfeatureproblematic,sincetheconceptsin termsofwhichtheychoosetoexplainmentalphenomenadonotexhibit thisall-ornothingcharacter.(1982,p.14)
Thisseemstometoofast;forwhyshouldweacceptthat all theconcepts availabletothephysicalistforgivinganaccountofmentalphenomenaare vague?(2)issurelybetterbasedonaconsiderationofthevariousalternativesopentothephysicalistaboutthenatureofconsciousness.
Consider firstthetypeidentitytheoryandthehypothesisputforwardby CrickandKochthatconsciousnessisoneandthesameasneuronaloscillationof40MHz.ItisevidentthatCrickandKochdidnotintendthis hypothesistoruleouteveryneuronaloscillationthatisnot exactly 40MHz.
Whataboutaneuronaloscillationof40.1MHz?Or40.01MHz?Or 40.000001MHz?Theirproposalisthatconsciousnessisoneandthesame asneuronaloscillationof approximately 40MHzorneuronaloscillation sufficiently closeto40MHz.Buttheseformulationsofthehypothesisbring outitsinherentvagueness,andnotjustfromtheuseoftheterms ‘ approximately’ and ‘sufficiently’;fortheterm ‘ neuron ’ isvaguetoo.
Neuronsarecomplexphysicalentitieswithdiversecomponents.Each neuronhasacellbody,dendrites,andanaxon.Electricalimpulsescomein alongthedendritesandgooutalongtheaxon.Imagineremovingatomsone byonefromagivenneuron.Eventually,asonedoesso,therewillbeno neuronleft.Butalongtheway,therewillsurelybearangeofborderline cases entitiesthatareneitherdefinitelyneuronsnordefinitelynotneurons. So,thepropertyofbeinganeuronaloscillationisvague.Itadmitsof borderlinecases.Ingeneral,neurophysiologicalpropertiesarehighlycomplex.Theideathattherelevantneuralpropertiesforconsciousnessaresharp isextremelyimplausible.
Whataboutrepresentationalorfunctionalroleorbehavioralproperties? Theproposalthattherecannotbeborderlinecaseshereisagainvery implausible.Borderlinecasesofrepresentationalpropertiesareeasyto specify.Takethepropertyofrepresenting meat.Historically,theword ‘meat’ meant beingedible:throughtime,itcametomean being flesh.Itis veryhardtoacceptthatthetransitionwassudden.Moreplausibly,therewas agradualdriftandthusaperiodoftimeatwhichthewordneitherdefinitely meantonenordefinitelymeanttheother.Thiswasnot because beingedible andbeing flesharevague(thoughtheyare).Ratheritisbecausetheproperty ofrepresentingitselfadmitsofborderlinecases.Foranonconventional example,consideraneuroninthevisualcortex,the firingofwhichrepresentsthepresenceofanedge.Naturalrepresentationisusuallytakentobea matteroftrackingundernormalconditions.Sincetheconceptofnormality isvagueifitiscashedoutstatistically(what typically happens)andalso vagueifitisunderstoodteleologically(whatissupposed tohappenifthe relevantsystemisoperatingproperly),naturalrepresentationisvaguetoo.
Turningtofunctionalroleproperties,therearephysicalinputsandoutputsandthesewilladmitofborderlinecaseswhethertheyareatthelevelof activityonsensoryandmotorneuronsoratthelevelofenvironmental inputsandbehavioraloutputs.Furthermore,functionalrolepropertiesare propertiesthatinvolve normal conditions,sothepointsjustmadewith respecttonaturalrepresentationcarryover.Similarpointsapplytoany directappealtobehavioralproperties.
Sofar,then,itseemsthatifconsciousnessistakentobeabroadlyphysical phenomenon,itisvague,aspremise(2)asserts.Butperhapsthereare furtherplausiblephysicalcandidatesforidentificationwithconsciousness thataren’tvague,candidatesnottobefoundinneurophysiologyorfunctionalrolesorbehaviororrepresentation.Mightsuchcandidatesbefound withinmicrophysics?Onceagain,thisisimplausible.Weareconscious; rocksandplantsarenot.Butifwearephysicalbeings,thenweandrocks andplantsarebuiltofthesamebasicstuff.So,whyareweconsciousand rocksandplantsnot?
Theobviousansweristhatwehavebrains.Consciousnessrequiresa brain.So,therelevantphysicalproperties,ifanytherebe,foridentification withconsciousnessshouldbeofasortfoundinneurophysiologyorin functioningorrepresentationatalevelofcomplexitythatrequiresabrain. AndsuchcomplexpropertiesIhavearguedarenotsharp.
Perhapsitwillnowbesuggestedthatweshouldlooktothe chemistry of brainprocessesfortherelevantphysicalproperties.Takethefeelingof anxiety,forexample.Thatisassociatedwithadecreaseinserotoninand dopaminelevelsinthebrain.Again,however,inevitablyvaguenesswillarise. Howlargeadecreaseinserotoninanddopamine?Toputanumericalvalue onthedecreaseistoinvitetheresponsethataverytinyamountmoreorless wouldsurelynotmakeadifference.Tosaythatthedecreasemustbe sufficientlylargeistointroducevaguenessrightaway.Itisalsoworthnoting thattheappealtodecreasedserotoninanddopamine inthebrain brings vaguenesswithitsuseofthequalifier “inthebrain”;forthebrainlackssharp boundariesandsothereareborderlinecasesofchanges in thebrain.Itdoes nothelptodropthequalifier “inthebrain,” Imightadd,sinceserotoninis producedwidelythroughoutthebody,butitisonlytheserotoninthatis producedinthebrainthatmattersdirectlytofeelings.Thereisnoreasonto thinkthatthesepointsarenotapplicabletoconsciousnessgenerally.
Supposeitisnowproposedthatintegratedinformationholdstheanswer. Whatitisforaphysicalsystemtobeconsciousisforittohavealargeamount ofintegratedinformation(Phi)init(Tononietal.2016).Thisview,whichcan betakentobeofferingahigh-levelphysicalaccountofconsciousness,has someextremelycounter-intuitiveconsequences.Forexample,asnotedby ScottAaronson(2014),itpredictsthatifasimple2-Dgridhastentimesthe amountofintegratedinformationasmybrain,thegridistentimesmore conscious!Whatexactlyismeantbyonesystembeingmoreconsciousthan anotherhasalsonotbeenmadefullyclearbyadvocatesofthetheory,butfor presentpurposes,itsufficestonotethatwhatitisforanamountofintegrated
informationtobelargeispatentlyvagueandthustheviewisofnohelpto thosewhowanttoholdthatconsciousnessissharpandbroadlyphysical.
Aresponsetothisdifficultyistosaythat some degreeofconsciousness goesalongwith any amountgreaterthanzeroofintegratedinformation.So, consciousnessissharp,afterall.Thisrequiresustoagreethatthermostats areconsciousasarespeedometers,sincetheycontainsomeintegrated information,andthatseemsalinetobeavoided,ifatallpossible!But evenifyoudisagreehere,asnotedabove,thereremainsthequestionasto whatitisforonesystemtohaveagreateramountofconsciousnessthan another.Andsinceadvocatesofintegratedinformationtheoryacceptthat certain2-Dgridsaremoreconsciousthanhumanbrains,itcannothaveto dowiththenumberofexperiencesortheintensityoftheexperiences;for surelynoonewantstoholdthattherelevantgridshavemoreexperiencesor moreintenseexperiencesthanourbrains(Pautz2019).Whatismeantby sayingthattheyaremoreconsciousthen?
Analternativestrategyistoaccept,forthereasonsgiven,thatconsciousnessisnottobereducedtopropertiesfoundinthephysicalsciencesorto functionalpropertieshavinginputsandoutputsbuiltupfromsuchphysical propertiesorinformationalpropertiesbuttoinsistthatconsciousnessis broadlyphysicalnonetheless,sinceitismetaphysically grounded inlowerlevelphysicalproperties.Thisneedsalittleexplanation.FollowingFine (2012),tosaythatapropertyPismetaphysicallygroundedinaproperty Qistosay(a)thatitismetaphysicallynecessarythatwheneverQis instantiated,Pistoo,and(b)thatthenatureofPexplainswhyPis metaphysicallynecessitatedbyQ.Asanillustration,considerthedisjunctive propertyofbeingredorsquare.Thatismetaphysicallygroundedinthe propertyofbeingred.Itissosinceitismetaphysicallynecessarythat wheneverthepropertyofbeingredisinstantiatedthepropertyofbeing redorsquareisinstantiatedandthenatureofthelatterpropertyexplains whythisrelationshipofnecessitationobtains.
Thepointtobemadenowisthatifconsciousnessisheldtobegrounded inlower-levelphysicalpropertiesthentherearetwopossibilities.The firstis thattherelevantlower-levelpropertiesarevagueinwhichcaseyetagain consciousnesswillbevague;forhowcouldasharppropertyhaveanature thatexplainsitsbeingmetaphysicallynecessitatedbyavagueproperty? Sharppropertiesaresharp by theirnature.Theyare necessarily sharp.So, thenatureofasharppropertyPhasnoroomforvaguenesswithinitand thusthatnaturecannot explain thepresenceofPineverypossibleworldin whichthepreferredvaguepropertyispresent.
Perhapsitwillberepliedthatthisistoofast.Howeverpuzzlingitmay initiallyseem,wecanactuallygive examples ofsharppropertiesthatare metaphysicallygroundedinvagueones.Considerthepropertybeingcolored.Thatismetaphysicallygroundedinthepropertyofbeingred.But beingredisvaguewhilstbeingcoloredissharp.
Itisnotclearthatthepropertyofbeingcoloredreallyissharp.Butletus putthattooneside.Thiscaseofmetaphysicalgroundingisunproblematic becausethepropertyofbeingcoloredisjustthesecond-orderpropertyof having a coloranditisapriorithatredisacolor.So,itisaprioriand necessarythatanythingthatisrediscolored.Butnothinglikethisobtainsin theconsciousnesscase;forsupposethatconsciousnessismetaphysically groundedinvaguephysicalpropertyP.Then,onthemodelofbeingcolored andbeingred,therewillhavetobesomesecond-orderphysicalpropertyof havingQsuchthatconsciousnessisoneandthesameashavingQ,where havingQissharp,andPhasthepropertyQandfurtheritisapriorithat PhasQ.Butthereisnosuitable sharp candidateforthe(broadly)physical propertyofhavingQ.Furthermore,itisacrucialfeatureofthephysicalist metaphysicalgroundingproposalforconsciousnessthatconsciousnessnot be reducible tosomebroadlyphysicalproperty.Andthisisbeingdenied withtheclaimthatconsciousnessisoneandthesameasthesecond-order physicalpropertyofhavingQ.
Theconclusiontowhichwearedrivenisthatconsciousnesscannotbea sharppropertythatismetaphysicallygroundedonvaguelower-levelphysicalorfunctionalproperties.Thisbringsustothesecondpossibilitywith respecttothemetaphysicalgroundingofconsciousness,namelythatthe relevantlower-levelpropertiesarecomplexconfigurationsof sharp microphysicalproperties³,inwhichcasetherewillbeareallyhugenumberof groundinglawslinkingconsciousnesstothemicrophysicalrealm.Buteven ifitissupposedthatcomplexconfigurationsofsharpmicrophysicalpropertiesmetaphysicallynecessitateconsciousness,noaccounthasbeenoffered (orevenseemspossible)ofhowthenatureofconsciousness explains whyit issonecessitated.Consideronesuchcase.Calltherelevantconfiguration ‘C’.Nowconsideraminimallydifferentconfiguration,C*thatdoesn’t metaphysicallynecessitateconsciousness.Microphysically,CandC*are almostthesame.YetCnecessitatesconsciousnessandC*doesn’t.If ³Ontheissueofsharpnessinthemicrophysicalrealm,seenote4below.
consciousnessitselfdoesnothaveaphysical/functionalnature,itisatotal mysteryastowhyCdoesthejobbutC*doesn’t.
Ofcourse,itcouldnowbeheldthatthegroundingisbrute.Butleavingaside theissueofwhethertherelationisappropriatelycalled “grounding”,the suggestionthatconsciousnessisbrutelygroundedinthephysicalisvery unsatisfying.Indeed,itputsthephysicalistaboutconsciousnessintheuncomfortablepositionofreplacingthedualist’sbrutepsycho-physicallawsofnature (aboutwhichmorelater)withahostofspecial,inexplicable,metaphysically necessarypseudo-laws.Thatsurelyistobeavoided,ifpossible.
Thereisone finalwayinwhichpremise(2)mightbeputunderpressure. Perhapsconsciousnessisaphysicalpropertythatisnottobefoundinthe physicalsciencesnorismetaphysicallygroundedinanysuchproperty.Ifthis isthecase,thepossibilityopensupthatconsciousnessisasharp,physical propertyafterall.Onthefaceofit,thesuggestionthatconsciousnessmight bephysicalandyetnotliewithinthephysicalsciencesnorbegrounded (only)uponpropertiesreferredtothereinisincoherent.Butthereisaview RussellianMonism underwhichitmakesgoodsense.SinceRussellian Monismisacomplexviewwhichdeservesextendeddiscussion,Iputit asidefornow.IshallreturntoitinChapter2.Forthepresent,Isimply notethatthereisgoodreasontodenythatRussellianMonism,ineitherofits twostandardforms,ultimatelycanbeusedtooverturnpremise(2).
Iturnnexttopremise(3).Considertheoriginoftheuniverseandthe emergenceofmoreandmorecomplexproperties.Takewater,forexample. Wateremergedoncehydrogenandoxygenatomscombinedinacertain way.Further,thereisanexplanationastowhytheycombinedintheway theydotoformwater.Oxygenneedstwomoreelectronstobecomestable.If oxygenatomsweretogototwohydrogenatomsandtakeawaytwo electronsfromthem(onefromeach),theresultwouldbeanO₂ oxideion andtwoH+ ions.Theseionsarenotfullystable.So,insteadoxygenatoms shareelectronswithhydrogenatomsbyformingtwoO-Hbonds.Theresult isH₂O,whichisfullystable.
Whataboutpropertieslikebeingamountainorbeingariver?Again, thesepropertiesseemreducibletofundamentalphysicalandtopicneutral propertiesarrangedintherightways.So,thereisnospecialorinexplicable emergenceinthesecases. ⁴ Thecaseofconsciousnessisradicallydifferentif
⁴ Itmightbesupposedthatthereisapuzzleevenhere,ifitisacceptedthatthesemacropropertiesarethemselvesvague.Butthepuzzlearisesonlyontheassumptionthatrealityatthe levelofmicrophysicsissharp.Thisassumptionishighlycontentious,however.Thesimplest
consciousnessisasharp,nonphysicalpropertythatemergedoutofcertain brainstructures.Herethereisnoreducibilityandrelatedlynoexplanationas towhyitemergedasitdid.So,uniformityinnatureislost.Phenomena graduallygetmoreandmorecomplexandthensuddenlyoutoftheblue somethingradicallydifferentjustoccurs.Why?Thereisnoexplanation.Itis justabrutefactthatoncecertain vague physicalstructuresareinplace, something sharp andnonphysicalemerges.Butthatisverydifficulttoaccept orevencomprehend.TheworryhereisrelatedtotheoneJ.J.C.Smartwas expressinginthefollowingpassagewhichgiveshisreactiontothedualist viewthattherearefundamentalphenomenal-physicallaws:
Icannotbelievethatultimatelawsofnaturecouldrelatesimpleconstituentstoconfigurationsconsistingofperhapsbillionsofneurons(and goodnessknowshowmanybillionbillionsofultimateparticles)allput togetherforalltheworldasthoughtheirmainpurposeinlifewastobea negativefeedbackmechanismofacomplicatedsort.Suchultimatelaws wouldbelikenothingsofarknowninscience.Theyhaveaqueer “smell” to them.Iamjustunabletobelieveinthenomologicaldanglersthemselves, orinthelawswherebytheywoulddangle.(1959,p.143)
Theterm “nomologicaldangler” thatSmartusesinthispassageisdueto Feigl(inanessaynotitselfpublisheduntillater,asashortbook,in1967). MyconcernaboutconsciousnessissimilartothatofSmart’s.Theideathat consciousnessjustsuddenlyemergeswithoutanyexplanationfromcertain neuralconfigurations,themselveswhollywithoutconsciousness,ispuzzling indeed.Onewantstoask:whydid this nonphysicalphenomenonjust suddenlyappearoutoftheblue,given these physicalstates?Whywouldn’t other priorphysicalstateshavedonejustaswell?Butifthelawsconnecting consciousnesstocertainphysicalstatesarefundamental,thentheseseeminglysensiblequestionsareillegitimate.Andthatmakestheviewthat consciousnessisanonphysicalphenomenon “franklyunbelievable” .
Thereisalsoafurtherdimensiontotheworry.Itisnotjustthatconsciousnessisnonphysicalandtiedbyabrutelawtothephysicalrealmthatis difficulttoaccept.Itisalsothatconsciousnessissharpwhereastherelevant
interpretationofquantummechanicshasitthatmicro-realityisvagueorfuzzy,thatproperties ofmicro-entitieslackingtheoreticalvaluesinquantummechanicsarevagueproperties. Exampleswouldincludeenergy,spin,polarization,andspatio-temporallocation.Formore here,seeChibeni2006.
underlyingphysicalstatesarevague.Whyshouldvaguestatesgeneratea sharpone?⁵ Ifanythingcriesoutforanexplanation,thisdoes.Butnoneis forthcoming orevenpossible ifconsciousnessisnonphysicalandso linkedonlybyabrutelawtothephysicaldomain.
So,premise(3)seemsveryplausible.Andoncepremise(3)isaccepted alongwith(2),the firstintermediateconclusionoftheparadoxisestablished:consciousnessisnotsharp.
1.2TheParadoxExplained:PartB
However,ifconsciousnessisnotsharp,itisvagueand,inbeingvague,it permitsborderlinecases.Arethereany?Hereisapossiblecase.Suppose Ihaveonlyjustwokenup,andIamstillgroggy,Iamnotyetfullyconscious. Isn’tthisaborderlinecaseofconsciousness?
ItiscertainlyafactthatIam more consciousoftheworldaroundme whenIamfullyawakethanwhenI firstgroggilyopenmyeyes. What Iexperienceisinitiallyindefiniteandimpoverished.AsIbecomefully awake, what Iexperiencegetsricherandricher.Butthisdoesn’tshowthat experienceorconsciousnessitselfhasborderlinecases.Hereishow Papineauputsthepoint:
Ifthelinebetweenconsciousandnon-consciousstatesisnotsharp, shouldn’tweexpectto findborderlinecasesinourownexperience?Yet whenwelookintoourselvesweseemto findaclearline.Pains,tickles, visualexperiencesandsoonareconscious,whiletheprocesseswhichallow ustoattachnamestofaces,ortoresolverandomdotstereogramsarenot. True,thereare ‘half-conscious’ experiences,suchasthe firstmomentsof waking....But,onreflection,eventhesespecialexperiencesseemto qualifyunequivocallyasconscious,inthesensethattheyarelikesomething,ratherthannothing.(1993,p.125)
Trytothinkofotherclearcut,objectivelyborderlinecasesofconsciousness, thatis,casessuchthatitisobjectivelyindeterminatewhetherconsciousness ispresent.Obviously,withsomesimplercreatures,wemaynot know whethertheyareconscious.Butthatisnotgermanetotheissue.Youcan
⁵ Formorehere,seepp.9–10earlier.
certainlythinkofacaseofconsciousnesswhichisindeterminateasto whetheritisacaseofpain,say.Thinkofsensationsatthedentistasyour teetharebeingdrilled.Someofthesesensationsseemimpossibletoclassify astotheirspecies.Thereisafeelingofpressureperhaps.Isitpain?Not clearlyso,butnotclearlynot.Hereitisindeterminateastowhatyouare feeling,butnotindeterminateastowhetheryouarefeeling.
Alternatively,imaginethatyouareinahospitalbedfeelingpainandthat youcanadjustadialthatcontrolsthedeliveryofmorphinetoyourbody.As youdoso,yourpainbecomeslessintense,graduallytransformingitselfinto afeelingofpleasure.Inthemiddleofthisprocess,theremaywellbe experiencesthatarenoteasytoclassify.Again,thereisindeterminacyat suchtimesastowhatyouarefeeling,butthereisnoindeterminacyasto whetherfeelingcontinuestobepresent.
Considerthecaseofauditorysensations.Supposeyouareparticipatingin anexperiment,listeningtorandomhigh-pitchedsoundsthroughheadphones.Youareaskedtopressabuttonforeachsoundyouhear.Insome cases,youareunsurewhetheryouarehearinganysoundatall.Isn’tthisa borderlinecaseofconsciousness?
Wecanagreethatthereisepistemicindeterminacyhere:youdonotknow whetheryouarehearinganysound.Still,thisisn’tenoughfortheretobea borderlinecaseofconsciousness.Afterall,youarelisteningattentivelyfora sound;areyouhearingasoundornot?Well,evenifyouaren’thearinga sound,youarestillhearingsomething,namelysilence.Thatis,youare hearingtheabsenceofasound;itisnotthatyouarefailingtohearatall! Thereissomethingitislikeforyousubjectivelytohearsilence.So,either way,youarehearingandthusexperiencingsomething.So,thisdoesn’tshow thattherecanbeborderlinecasesofexperience.
Supposesomeoneheldthatbeingtallisprecise,admittingofnoborderlinecases.Wecanquicklyshowthispersonthatsheiswrongbypresenting herwithexamplesofpeoplewhoaren’tdefinitelytallbutwhoalsoaren’t definitelynottall.Wecandothesamewithexperiencingredorfeelingpain orhearingaloudnoiseorfeelinghappy.Butcanwedoitwithbeingan experience(orbeingconscious)?
Idon’tthinkwecan.Wecancertainlyagreethatastheintensityofan experiencediminishes,itbecomeslessandlessdefiniteandrichinits character,buteitheranexperienceisstillthereoritisn’t.Picturingwhatit islikefromthesubject’spointofview,wepicturetheexperiencegradually changinginitsphenomenologyuntilitisso ‘washedout’ andminimalthatit hashardlyanydistinguishingfeaturessubjectively.Butthesubjectisstill
havinganexperience.Thegradualtransitionisinthenumberandintensity ofthesubjectivefeaturesoftheexperience,notinthestate’sbeingan experience(beingphenomenallyconscious).
Themodelthattemptsushereisthatofalightbulbwithadimmer switch.Asonefallsasleep,thelightbecomesdimmeranddimmerandthen suddenlyalllightisgone.Bycontrast,asonecomesoutofanaesthesia,a veryfaintlightsuddenlydawnsandthenitbecomesbright.Butofcourseif thisisthecase,thenconsciousnessissharp.
Lifeisnotlikethis.Thereasonisstraightforward:theconcept living isa functional/behavioral,clusterconcept.Livingthingsuseenergy,theygrow, theyreproduce,theyrespondtotheirenvironment,theyadapt,andthey self-regulate.Whatitisforanentitytobelivingisforittohaveenoughof thesefunctionalandbehavioralfeatures.Borderlinecasesarisewithrespect towhetheragivenentitygenuinelydoeshaveenoughoftherelevant features(perhapsitdoesbymystandardsbutnotbyyours)andalsowith respecttothepossessionofindividualfeatures,as,forexample,withbacteria andcertainkindsoforganicmolecules.Theresultisthatthetransitionin thecaseoflifefromclearlyinanimatetoclearlyanimatebeingsisgradual andcontinuous.
Inthecaseofconsciousness,thereisnofunctionalorbehavioraldefinition.Thisiswhyitisperfectlycoherenttoimagineacreaturethatisnot conscious(azombie)eventhoughitisfunctionallyandbehaviorallyjustlike acreaturethatis.⁶ Asnotedearlier,theconceptofaconsciousstateisjustthe conceptofanexperience,thatis,astatesuchthatthereissomethingitislike toundergoit.So,thereisnothingintheconceptofconsciousnessthat supportstheviewthatthereareintermediatestages,asistrueforlife.
Perhapsitwillnowberepliedthatthefactthatwecannotgiveanyclear examplesofborderlinecasesofconsciousnessandthefurtherfactthatthe concept consciousness isnotliketheconcept life doesnotdemonstratethat theconcept consciousness issharp.Imagine,forexample,thatwelivedina worldwithonlyredobjectsandonlyblueobjects.Insuchaworld,wemight notbeabletoconceiveofaborderlinecaseofredandsowemighttakeitfor grantedthatredisasharpproperty(andtheconcept red asharpconcept) andlikewiseforblue.Butwewouldbewrong.Itismetaphysicallypossible fortheretobeborderlinecaseofredevenifwearenotabletoconceiveofa borderlineredthing.
⁶ Thisover-simplifiesminimally.Formorehere,seeChapter2,pp.7–8.
Thisseemstomecorrectasfarasitgoes.Buttheissueiswhatitis reasonabletobelieveonthebasisofevidenceratherthanwhatisprovable. Conceivabilityisdefeasibleevidenceformetaphysicalpossibilityandinconceivabilityisdefeasibleevidenceformetaphysicalimpossibility.Inthelatter case,thequestioniswhetherwhatweknowbyapriorirationalreflection upontherelevantconcept,giventhatwepossesstheconceptandagood understandingofit,⁷ isgoodgroundsforrulingoutthetruthofthepropositionthatitismetaphysicallypossiblethatthereisanobjectsatisfyingthe concept.Inthecaseoftheconcept tall,wecaneasilyconceiveofaborderline casewhetherornotwehaveencounteredanyborderlinetallmen.Thatis evidencefortheviewthattheconceptisvague.Inthecaseoftheconcept red andgreenallover wecannotconceiveofanobjectthatsatisfiesit.Sincewe possesstheconstituentconceptshereandwehaveagoodgraspofthem,that isevidencethatitismetaphysicallyimpossiblethatthereissuchanobject. Theconcept consciousness issuchthatwecannotconceiveofaborderline caseandthatisprimafacieevidencethatitissharp.Andwhatgoesforthe concept consciousness goesforthepropertyofbeingconscioustoo. Onereactiontotheseobservationsistosaythatitisindeedtruethatour current conceptofconsciousnessissharpbutthatdoesn’tmeanthatitwill remainsharpinthefuture.ThisisapossibilitysuggestedbyMichaelAntony inhis2006.⁸ Hesays:
Onecan,however,distinguishbetweenourcurrentconcept consciousstate andsomefutureversionoftheconcept,claimingthatwhileourcurrent conceptisindeedsharp,afuturedevelopmentofitwillbevague.This occurredwiththeconcept life.Atearlierstagesinitshistorytheconcept wassharp:borderlinelivingcreatureswereinconceivable.However,the conceptdevelopedwiththeadventofmodernbiology,anditnowappears tobevague(viruses,forexample,areoftenplausiblysuggestedasborderline cases).Similarly,itmightbethoughtthatourcurrent,relativelyprimitive concept consciousstate mustundergodevelopmentalchangebeforeitcan correctlyrepresentitssubjectmatter,onesuchchangebeingfromsharpness
⁷ Ratscannotconceiveofthetruthofvarioustruetheoretical,scientificpropositionsbutthis isnoreasontosupposethatthesepropositionsarenotreally(possibly)true.Thisisbecauserats lacktherelevantconcepts.Similarly,ifIpossessagivenconceptbutImisunderstandit(asinthe caseofBurge’sindividualwhopossessestheconcept fortnight butthinksthatafortnightisten days),thefactthatIcannotconceiveofP’sbeingthecaseviamyapriorireflectiononthe conceptisnotgroundsforholdingthatitismetaphysicallyimpossiblethatPistrue.
⁸ IshouldaddthatAntonydoesnotendupendorsingthissuggestion.
tovagueness.Inthisway,onecanadmitthatourcurrentconceptissharp inaccordancewithourintuitions,butmaintainthattheconceptisalsoin asensevaguesinceafuture,morecorrectversionofitwillbevague.
(2006,p.517)
Itisnotfullyclearthattheconcept life reallyhaschangedfromsharpto vague,asAntonysupposesinthispassage.Vitalistsheldthatlifeisavital, immaterialforceanimatinglivingthings.Butthiswasaviewaboutlifeitself, notabouttheconcept life.Thereisnoinconsistencyinholdingboththat whatitisforanyorganismtobealiveisforittohaveenoughofacertain clusteroffunctionalandbehavioralfeaturesandthatinactualfactwhatis responsibleforanyorganism’shavingsuchaclusteroffeaturesisitshaving avital,immaterialforcewithinit.Theforcemightbeheldtobewhat realizes thatcomplexoffeatures.
Still,itmustbeadmittedthatvitalistsseemtohavetakentheirviewto offeranaccountofthenatureofliferatherthanjustarealizationofit.From themodernperspective,theirconceptoflifeisliketheconcept phlogiston in beinganemptynaturalkindconcept.Theclusteroffeatureswenowtaketo definelifeismerelycontingentlyassociatedwithlife,onthevitalistview. Thisisprettyobviouslynotaplausiblemodelforunderstandinghowour concept consciousness mightdevelopintoavagueconceptfromitscurrent sharpform.Foronething,theideathatitcouldturnoutthatconsciousness, thatis,experienceorfeeling,neverexisted,seemsprimafacieabsurd(unlike theideathatthereneverwasanyvitalforce).Toconsciouslythinksuchan ideaistohaveaconsciousexperienceofthinkingandtherebyalreadyto refuteit!Foranotherthing,anyfuturevaguefunctionalistorbehavioral definitionwouldnotbeadefinitionofthephenomenonwecurrentlycall ‘consciousness’;forourcurrentconcepthasnofunctionalistorbehavioral definition.Sincetriviallyitisapriori necessary thatourcurrentconceptof consciousnessiswithoutanysuchdefinition,ourcurrentconceptcannot developinto afuturevaguefunctionalistorbehavioralconcept.Atbest,it couldbereplacedbysuchaconcept(aswasthecasewiththevitalist’ s conceptoflifeonthesecondunderstandingofitaboveandourcurrent conceptoflife).Butweresuchareplacementtooccur,itwouldbeaconcept forsomethingotherthanconsciousnessandthereforethepossibilityofsuch areplacementoffersnosupporttotheviewthatconsciousnessitselfmight turnouttobevague.
Thereisafurtherpointworthmakinghereabouttheabovesuggestion, namelythatitframestheissueinthewrongway.Thequestionofcentral
interestisthequestionastowhetherconsciousnessitselfadmitsofpossible borderlinecases.Thisisnotdirectlyaquestionaboutconceptsatall.So, whetherourcurrentconceptofconsciousnessmightchangethroughtimeis reallynottothepoint.Whatwewanttoknowiswhetherconsciousnessis thesortofthingthatcanhaveborderlinecases.Andtheconclusiontowhich wearedrivenbythereflectionsofthissectionisthatitisnot.Consciousness isnotvague.So,consciousness,itseems,isbothvagueandnotvague.What todo?Houston,wehaveaproblem!
