PhilosophicalPerspectives,20,Metaphysics,2006INDEFENSEOFESSENTIALISML. - PDF document

PhilosophicalPerspectives,20,Metaphysics,2006INDEFENSEOFESSENTIALISML.A.PaulUniversityofArizonaAustralianNationalUniversity/RSSSIntroductionIfanobjecthasapropertyessentially,ithasthatpropertyineverypossibleworldaccordingtowhichitexists.Ifanobjecthasapropertyaccidentally,itdoesnothavethatpropertyineverypossibleworldaccordingtowhichitexists.Claimsaboutanobjectsessentialoraccidentalpropertiesarederemodalclaims,andessentialandaccidentalpropertiesarederemodalproperties.Takeanobjectsmodalprofiletospecifyitsessentialpropertiesandtherangeofitsaccidentalproperties.NotethatworldŽasIamusingitisatermofart:amodalrealistbelievesthattherearemanyconcreteworlds,whilethebelievesinonlyoneconcreteworld,theactualworld.Theersatzistisanactualistwhotakesnonactualpossibleworldsandtheircontentstobeabstracta.istheviewthatobjectshavepropertiesessentially,butoneshoulddistinguishessentialismfromshallowessentialism.Deepessentialiststakethe(nontrivial)essentialpropertiesofanobjecttodetermineitsnature„suchpropertiesgivesensetotheideathatanobjecthasauniqueanddistinctivecharacter,andmakeitthecasethatanobjecthastobeacertainwayinorderforittoatall.AsStephenYablo(1987,297)describesit,theessenceofathingisanassortmentofpropertiesinvirtueofwhichitistheentityinquestion,Žaswellasameasureofwhatisrequiredforittobethatthing.ŽIntuitively,onthedeepessentialistpicture,anordinaryobjecthasessentialproperties,anditmusthaveitsessentialpropertiesinorderforittoexist.Onthisview,objectsessentialpropertiesare,i.e.,arenotdeterminedbycontextsofdescribing(orthinking,etc.)abouttheobject,andtruthsaboutsuchpropertiesareabsolutetruths.Shallowessentialistsopposedeepessentialists:theyrejecttheviewthatobjectscanbesaidtohaveessentialpropertiesindependentlyofcontextsofdescriptionorevaluation,andsosubstitutecontext-dependenttruthsforthedeepessentialistscontext-independentones. L.A.PaulAccordingtothedeepessentialist,itcannotbethecasethatonecontextofdescriptioncanmakeittruethat,e.g.,isessentiallyblondeŽwhileanothercontextofdescriptioncanmakeittruethatisaccidentallyblondeŽ(wherereferstotheverysameobjectinbothsentences).Truthsaboutanobjectsmodalpropertiesarenotdeterminedbythewaywedescribeit.Thedeepessentialistalsoholdsthatmostobjects,andespeciallyordinaryobjects,haveessentialpropertiessuchthattherearemanywaystheycouldnotbe,ormanychangesthroughwhichtheycouldnotpersist.Inthisway,objectsmodalprofilescharacterizetheirnatures.Intheusualcase,aswithordinaryobjectslikepersons,mindsandbooks,objectsarenotpossiblyformanyproperties.(Forexample,Iamnotpossiblyafriedegg,Iamnotpossiblyatree,etc,independentlyofwhatcanbeassertedaboutmerelativetocertainsortsofcontexts.)Inaddition,objectshavearestrictedrangeofpropertiessuchthattheyarepossibly.(Forexample,ifIampossiblyblonde,thisisindependentofwhatcanbeassertedaboutmerelativetocertainsortsofcontexts.)Becausewethinkoftheworldasfullofnumericallydistinctobjectsthathaveuniqueandabsolutemodalpersistenceconditions,deepessentialismdoesthebestjobofcapturingthewaywewanttomakesenseoftheworldandtheordinaryobjectsitcontains.Deepessentialistscanholdthat,independentlyofacontext,abrainstateisessentiallyphysical,apainstateisessentiallypainful,abookisessentiallywrittenbyitsauthor,personsareessentiallysentient,andstallionsareessentiallymale.AsIwillarguebelow,deepessentialismistheonlysortofessentialismthatpreservesthiselementofourordinaryunderstandingofeverydayobjects.Somehavethoughtthatdeepessentialismfalterswhenfacedwithaskepticalobjectionconcerningarbitrarinessandtheneedforareductiveaccountofderemodality,andhaveadoptedshallowessentialisminresponse.Butweneednotresorttosuchadesperateact:Illargueforawaytoanswertheskepticalobjectionthatpreservesthecentralelementsofdeepessentialism.Thecrucialmoveofmyapproachistodevelopareductiveaccountofderemodalitythatiscongenialtotheabsolutismofthedeepessentialist.Oncedeepessentialistshavearesponsetotheskepticalobjection,theycanbegainfullyemployedingivingbroadlyreductivetreatmentsofrelatedtopicssuchasmaterialconstitution,nonreductivephysicalism,supervenience,andcausaloverdeter-mination.Suchtreatmentsshouldbemoresatisfactorythanthosecurrentlyonoffer.1.SubstanceEssentialismHowdoesanobjecthaveitsessentialandaccidentalproperties?Whatistheontologicalbasisforanobjectsmodalprofile?Theneedtoexplaintheunderlyingontologicalstructureofhowanobjecthasitsderemodalpropertiesisresistedbymanydeepessentialists.Instead,thefundamentalstructureofanobject,which InDefenseofEssentialismincludesstructurethatdetermineshowithasitsmodalproperties,istakenasanunanalyzableprimitive.Onepopularversionofthisviewtakesobjectstofallundersortalconcepts.Calltheadvocateoftheviewthat(a)objectsaresubstancesthatfallundersortalsand(b)weneednofurtherontologicalexplanationofhowobjectshavederemodalpropertiesasubstancetheorist.ŽDeepessentialistswhodefendsubstancetheoryholdthatthederemodalandotherpersistencepropertiesofordinaryobjectscorrespondtowhichsortstheybelongto,butthatnomoredevelopedontologicalaccountisavailable.Substancetheory,asIamcharacterizingit,isadescendentoftheAristoteliantheoryofprimarysubstancesasindividualobjectswhichareneitherinasubjectnorsaidofasubject,forexample,theindividualmanorindividualexamplesofsubstanceareman,horse.ŽOnthisview(whichisseparablefromthesubstance-as-substratumviewthatwemaytaketodescendfromtheworkofDescartesandLocke),substancesarebasicthingsthatdonotadmitoffurtherreduction,andtheconceptofsubstanceisnotamenabletofurtherphilo-sophicalanalysis.Atbest,wemaybeabletocharacterizetheconceptindirectlybycitinginstances,forexample,bysayingthatsubstancesaresimplyordinaryobjectsthatendurethroughtimeincharacteristicwaysandaremetaphysicallyindependent(insomesense)ofotherthings.Onthefaceofit,theviewhasintuitiveappeal.Commonsensically,werecognizeobjectsbytheirsortalpersistenceconditions,andwedowanttosaythatthereissomethingconceptuallybasicaboutthewaywedistinguishbetweenordinaryobjects.Butthetroublewithsubstancetheoryisthatalongwiththeseappealingthesescomethestipulationthatlittle,ifanything,isknownabouthoworwhyobjectshavetheirderemodalproperties.Anysubstancetheorist,bydefinition,lacksadeepexplanationoftheontologicalstructurethatdetermineshowobjectshavetheirderemodalproperties(sincehetakesitasprimitive).Substancetheoristswhoaredeepessentialistsaremytargethere,andIllcallsuchtheoristssubstanceessentialists.Bytakingsubstance,sorthood,andthewayobjectshavederemodalpropertiesasunanalyzableandhenceprimitive,anyseeminglycounterintuitiveconsequencesofsubstanceessentialismaredifficulttoexplainormakepalatable.Sinceweknowsolittleaboutthewayobjectshavederemodalandotherpersistencepropertiestheydo,substanceessentialistscannotanswerobjectionsrelatedtohoworwhyanobjecthasitsmodalproperties.Laudingourabilitytorecognizeobjectsinsomespecialwaydoesnottellenoughofthestory.ThisbecomesclearerinthecontextofthefamousQuineanchallengetodeepessentialists,whereQuinedemandsanexplanationofhowanobjecthassomeofitspropertiesessentiallyandotherpropertiesaccidentally.(Quine1960,esp.p.199,and1963,p.155)AsKitFine(2005,4)putsQuinesobjection:Themetaphysicalnotionofnecessity[anobjectsnecessarilybeingacertainway]capableofdiscriminatinginaninterestingwaybetweendifferentobjects.Thenumber9,forexample,isnecessarilyanumberthoughnotnecessarilythenumber L.A.Pauloftheplanets;andSocratesisnecessarilyapersonthoughnotnecessarilyaphilosopher.Fromwhencederivesthesedifferencesinthenecessaryfeaturesofanobject?ŽIntheabsenceofanaccountofwhatitisabouttheobjectthatdeterminesthemodalstatusofthepropertiesithas,askepticaboutessencecanarguethatthereisnothingthatdeterminesthatanobjecthassomepropertiesessentiallyandsomeaccidentally,sothedistinctionbetweenabsoluteessentialandaccidentalpropertiesis,asQuine(1963,p.155)putsit,invidious.ŽIfnothingdeterminesderemodalstatusofsuchproperties,itseemsthatpropertiessimplyarenthad(absolutely)essentiallyoraccidentally„sincesurelymodalpropertiesarentarbitrarilyItmayseemthatadeepessentialisthasareplytothisargument,forshecanarguethatQuineisaskingfortoomuch.Theworldisacertainway,andwhenwereadoffthefeaturesofreality,thesefeaturesincludeobjectshavingsomepropertiesessentiallyandsomeaccidentally.CompareQuinesobjectiontosomeoneobjectingthatwecannotdefendtheviewthatthereareforcesunlesswecanexplainwhythelawsofnatureholdthatforceisproportionaltoaccelerationratherthaninverselyproportional.Forcesjustdoexist,andforcejustisproportionaltoacceleration,andwhenwelearnaboutforcewearelearningaboutwhatisintheworld.ThisreplytoQuineamountstothepointthatitisntthatdeterminesanobjectsderemodalproperties,rather,whatdeterminesthemisontologicallyfundamental,andassuch,isprimitiveandunanalyzable.Everyontologyhasitsprimitives,anddeepessentialismisnoexception.Thedeepessentialistissurelycorrecttopointoutthattakingamattertobeprimitiveisdifferentfromtakingittobearbitrary.Moreover,rejectingdeepessentialismsimplybecauseitinvolvesameasureofprimitivismwouldbeunwarranted,and(settingasideQuinehimselfandtakingtheQuineantobeacontemporaryQuinean)theQuineanisnotjustifiedinrejectingdeepessentialismonthesegroundsalone.Unfortunately,however,theQuineanwillnotbesatisfiedwiththedeepessentialistsdefenseofprimitivism.TheQuineanwillrespondthatyes,everyontologyhasitsprimitives,butthereareprimitivesandthenthereareprimitives.Takingittobeafundamental,unanalyzablefactthatforceisproportionaltoaccelerationisacceptableonlyifthereisnoreductiontobehad.Thesamegoesfortakingthedeterminationofanobjectsderemodalpropertiestobeprimitive:itisonlyjustifiedifthereisnoreductiontobehad.Thedeepessentialistwhoholdsthatderemodalpropertiesareunanalyzablefeaturesoftheworldneedstoshowthatnosuitablereductionisavailable.Inotherwords,primitivismperseisnotwhattheQuineanobjectsto„hisobjectionistounmotivatedprimitivism,inparticular,unmotivatedprimitivismabouthowobjectshavetheirderemodalproperties.AccordingtotheQuinean,forprimitivismaboutanentitytobeacceptable,itmustbenonarbitrary,plausibleandwell-motivated,ratherthanstipulatedinanadhocmanner.Thismeansthatprimitiveentitiesareonlyacceptablewhentheyareincludedinanotherwise InDefenseofEssentialismwell-developedontology,andonlywhennofurtherreductionisavailable.But,asitturnsout,theQuineanwillargue,well-motivated,acceptablereductionsofhowobjectsgettheirderemodalpropertiesareavailable,asarereductionsofderemodalitytopurelygeneralorqualitativefacts.Moreover,theQuineancanadoptanaturalŽexplanationofwhywewanttoascribecertainderepropertiestocertainobjects:followingtheleadoftheshallowessentialist,itissimplyamatterofprivilegingcertaindescriptionsorevaluationsoftheobjects.Atthispointinthedialectic,wecanthinkoftheQuineanaspresentingthedeepessentialistwiththreerequirementsforanadequatetheory.(i)Anyprimitivesofthetheoryofessentialismneedtobewell-motivated,i.e.,notamenabletofurtherreduction,andnotadhoc.(ii)Thetheorymustgiveareductiveaccountofhowobjectshavetheirderemodalproperties.(iii)Thetheorymustreducederemodalfactstopurelyqualitativeorgeneralfacts.ThecontemporaryinterpretationofQuinesobjectiontodeepessentialismisthatdeepessentialismisuntenablebecauseitcannotsatisfy(i)-(iii):Illcallthisinterpretationtheskepticalobjectiontoessentialism.Requirement(i)seemsstraightforwardlyacceptable.Oneviewthatfailsthetestfor(i)isDavidLewiss(1968)originalversionofcounterparttheory(revisedinhis1983PostscriptŽ),accordingtowhichobjectshavetheirderepropertiessolelyinvirtueofbeingsimilarinspecialwaystopossibleindividuals.Onthisview,anobjecthasthesamecounterpartshoweveritisreferredto,andthuscanbeascribedpropertiesessentiallyoraccidentallyindependentlyofcontextsofdescription.Thetroubleisthatsincetherearemanyrespectsinwhichobjectsaresimilartopossibleindividuals,but(intuitively)onlysomeofthosepossibleindividualsshouldcountascounterpartsoftheobject(andthuscounttowardsdeterminingwhichpropertiesareessentialoraccidental),certainsimilaritiesarespecial,i.e.,mustreceivepriority.Butwhatisresponsibleforthispriority?Anunjustifiableprimitivismseemstobetheonlyoption.Requirements(ii)and(iii)arereductiverequirements.TheQuineansdemandforawell-developedontologytosupporttheoriesofderemodalityinvolvesademandforareductiveaccountofderemodality,especiallygiventhefactthattherearewidelyacceptedreductionsavailable.(ii)requiresareductiveexplanationofhowobjectsgettheirderemodalpropertiesand(iii)isbasedontheideathatafull-blownreductionofthederemodaltothequalitativeallowscertaindesirabletheoreticalandmethodologicalvirtuestobemaximized.Requirement(ii)ismetbycontemporarymetaphysiciansbytakingderemodalpropertiesofobjectstobedeterminedbyrelationstheobjectshavetopossibleindividuals(wheresuchindividualsarepartsofpossibleworlds).Suchpossibleindividualscouldbeconcretecounterparts(undermodalrealism)orersatzindividualswithcounterparts(underersatzism).Ifallthecounterpartsof-object(ortheersatzindividualthatrepresentsthe-object)are,thisdeterminesthatthe-objectisessentially.Ifnotallthecounterpartsofthe-object(ortheersatzindividualthatrepresentsthe-object)are,thisdeterminesthatthe-objectisaccidentally L.A.PaulSayingthatobjectshavetheirmodalpropertiesinvirtueofstandinginrelationstopossibleindividualsmakesstandinginsuchrelationswhatdeter-minesmodalproperties.Deremodalpropertiesofobjectsarethereforenotmysteriousconstituentsofthem:theyaredetermined(andthusexplained)bythecounterpartsanobjectishasorisrepresentedashaving.Counterpartrelationsarequalitativesimilarityrelations(orersatzsimilarityrelationssomehowbasedonthequalitativesimilaritiesthatwouldexistbetweenactualobjectsandtheircounterparts,shouldthosecounterpartsexist).AlthoughIhavefleshedthingsoutintermsofqualitativecounterparts,notethattherequirementforanexplanationofhowobjectshavetheirderepropertiesisnotinthefirstinstanceanobjectiontoprimitivismaboutallsortsofunreducedmodalitydere.Inparticular,itmaybeconsistentwith(ii)toholdthatitisaprimitivefactabouttheworldthatactualobjectscanhavetheirmodalpropertiesinvirtueofstandinginsomesortofnonqualitativerelationtopossibilia(i.e.,invirtueofstandinginsomesortofrelationtopossibiliathatsupervenesonspecificdere(ii)isademandforareductiveexplanationofhowobjectsgettheirderemodalproperties,whichmightfallshortofafull-blownreductionofderemodality.AnObjectionmayariseatthispoint.TheObjectionrunsasfollows:takinganobjectsessentialandaccidentalpropertiestobedeterminedbytherelationsithastootherworldlyindividuals(evenifthedeterminingrelationsare,say,nonqualitativecrossworldidentityrelations)getstheexplanationthewrongwayaround.Intuitively,itistheobjectsessentialandaccidentalpropertiesthatdeterminewhetheranobjectexistsaccordingtoaworld,nottheotherwayaround.Thisobjectionhasprimafacieintuitivepull.Thereductionistmaysimplygranttheforceoftheintuition,butarguethatthereductionhasbenefitssufficienttooutweightheintuitivecostofrejectingit.ButcloserinspectionofthemotivationbehindtheObjectionweakensitsintuitivepull.Howisthedeterminationofanobjectsrelationstopossibleindividualssupposedtoworkifitrunsinthedirectiontheobjectorsaysitdoes?Whatisresponsibleforit?ItisasthoughthereissomesortofmysteriousModalForcethatobjectsaresupposedtohavewhichdetermineswhethertheyexistaccordingtopossibleworlds.Again,theprimitivismofthisessentialiststoryfailstoprovideasatisfactoryexplanationofthederemodalfacts.AfurtherconsiderationarisesifoneacceptsareductiveinterpretationofwhatSider(2003)callsLeibnizianbiconditionalsŽ:propositionispossibleiffistrueinsomepossibleworldandpropositionisnecessaryiffistrueineverypossibleworld.TheLeibnizianbiconditionalsinterpretmodalclaimsintermsoftruthsaboutpossibleworlds,whethersuchworldsareconcreteworldsofakindwithourown,asthemodalrealistwouldhaveit,orabstractentitieslikepropositions,sets,orcombinationsofstatesofaffairs,suchastheerstazistandmanyotheractualistswouldhaveit.Fortheoreticalaswellasintuitivereasons,theideathatapropositionispossiblytrueiffitistrueinsomepossibleworld InDefenseofEssentialismandnecessarilytrueiffitistrueineverypossibleworldhasbecomeubiquitousincontemporaryphilosophy.Onthereductiveinterpretationofthisapproach,modaltruthsarereducibletofactsaboutpossibleworlds:sbeingpossiblytruejustisPsbeingtrueinsomepossibleworld.ThereductionofderemodalpropertiesofobjectstorelationsorrelationalpropertieshadtopossibiliacanbeseenasanextensionofthereductiveversionoftheLeibnizianframework.Ifoneacceptsthegeneralideathatpossibilityandnecessityreductivelysuperveneuponfactsaboutpossibilia,whateverclaimtheObjectionhasonourintuitionsabouthowobjectshavederemodalpropertiesshouldberecognizedasdeeplymisguided.ThedefenderoftheObjectioneitherneedstoexplicitlyrejectthereductiveversionoftheLeibnizianframeworkandprovidereasonsfordoingsoorneedstoexplainwhyessencesmustbedetermineddifferentlyfromhowpossibilitiesmoregenerallyconstruedaredetermined.Requirement(iii)involvesthereductionofderemodalitytosomethingmorefundamental.(Ideally,thiswouldinvolveareductionofthederemodaltothenonmodal.Infact,asInotebelow,whatwemayhavetobesatisfiedwithisareductionofthederemodaltopurelyqualitativefactsinvolvingthenonmodalandthededicto.)Lewis(1986)arguesthatareductivetreatmentofdereshouldbepreferredbecauseitmaximizesthevirtuesofsimplicity,elegance,andontologicalparsimony(intermsofthekindsofentitieswemustposit).Sider(2001,2003&unpublished)arguesthatacceptingungrounded(i.e.,primitive)deremodalortensedfactsabouttheworldismethodologicallyinferiortogivingmorereductiveaccounts.AccordingtoSider,thefundamentalpropertiesoftheworldshouldbecategoricalratherthanhypothetical(2001p.41,2003p.185),wherecategoricityinvolvestheactualnatureofthingsandhypotheticalityinvolvesoutwardlookingfactssuchasmerelypossiblefacts.Others,suchasHeller(1990),Sosa(1987)andZimmerman(1995)addtothechorusforthereductionofthederemodalbyarguingthatitseemsintuitivelyrighttosaythatanobjectsderemodalcharacteristicsaredeterminedbyitsnonmodalfeatures.Sosa(1987,173…4)arguesthatjustasweshouldbeabletoexplainobjectsdispositionsbytheirnonmodalproperties(e.g.,amarblesdispositiontorollisexplainedbyitsroundness),weshouldbeabletoexplainobjectsmodalpropertiesbytheirnonmodalproperties(e.g.,amarblesbeingunabletosurvivesquashingshouldbeexplainedbyitshardness,shape,andotherphysicalproperties).Hellertellsusto[c]onsidertwoobjectsthatexistontwodistinctplanets.Supposethatthesetwoobjects[sharealltheirnonmodalproperties]andhaveexactlythesameforcesactingonthemateverytimeduringtheirexistence.Moreover,supposethatifeitherofthemweretohavebeeninadifferentsituationthantheoneitisin(havedifferentforcesactingonitthandoinfactactonit)thentheotherwouldhavebeeninacompletelyanalogoussituation(thetwoobjectswouldstillhavehadthesameforcesactinguponthem).Giventhesesuppositions,itseemstomeincredibletosupposethatoneoftheobjectscouldhavegoneoutofexistencebeforetheother.Ž(1990,31) L.A.PaulThepointoftheseargumentsaboutderemodalpropertiesbeingdeterminedbynonmodalpropertiesistosuggestthatanyacceptableontologyofderepropertieswilltakethemtoreductivelysuperveneonnonmodalproperties,oratleastonpurelyqualitative,nondereproperties.(Inmanycases,dependingonassociatedviewsaboutmodality,thereductionwillbetothenonmodalplusafewassociateddedictofacts,ratherthantothenonmodal.Thisisbecause(asLewis1986argues)thosewhorejectmodalrealismusuallyneedtoacceptcertaindedictomodalfactsasprimitive.)IfindthesortsofargumentsgivenbySider,Heller,SosaandZimmermanplausible,butIammostinfluencedbyLewissviewthatthesavingsintermsoftheoreticaleconomymakesthereductionworthhaving.Thevalueofthissortofeconomybecomesclearerwhenweconsiderthatanumberofcentraltopicsinmetaphysicsinvolvereduction:supervenience,causation,dispositions,laws,physicalism,naturalism,andmaterialisminmind,justtonameafew.Moreover,whenthefanofderemodalityispushedtoexplaintheconsequencesofherviews,shecangiverealanswers,ratherthanbeforcedtofallbackonclaimsaboutprimitivism.So,Iamconvincedthatthedeepessentialistmustrespondtoparts(i)and(ii)oftheskepticalobjection,andthinkthereisafairlystrongcasefor(iii)aswell.Theskepticalobjectionpressesparticularlyhardonthesubstanceessen-tialist,whoseeveryjointisswollenwithprimitivism.Considerrequirement(i):thatatheoryofessencenotrelyonadhocorgratuitousprimitivism.Thesubstanceessentialistmaybeintuitivelyjustifiedinholdingthatthereexistsortsorcategoriesintheworld,andinholdingthatknowingthesortalconceptthattheobjectfallsundercantelluswhatitsmodalpropertiesare.Butthecontemporaryskepticdoesnotdisputethisdatum(atleast,nothere).Rather,hearguesthatthesubstanceessentialistneedsastoryaboutwhatmakesitthecasethatanobjectisthekindorsortofthingthatitis.Anaturalansweristhatanobjectsmodalandotherpersistencepropertiesare(partof)whatmakeitthesortofthingitis.Ifthesubstanceessentialistdisagreesandholdsthatthesortanobjectbelongstodeterminesitsderemodalproperties(ratherthantheotherwayaround),thenheneedstogivean(ontological,notconceptual)explanationofwhatdeterminesanobjectssort.Simplystipulatingthatitisaprimitivefactthatbeingofacertainsortdeterminesthemodalprofileofanobjectisunacceptablyadhocespeciallygiventheavailabilityofthenaturalexplanationthatanobjectshavingthepersistencepropertiesthatitdoesarewhatmakeitaparticularsortofthing.Similarconsiderationsapplytothesubstanceessentialistwhoholdsthatsomeothersortofprimitiveentity,suchasaform,determinesanobjectssort(orperhapsdeterminesbothitssortanditsmodalproperties).Inthecontextofareductiveaccountofderemodality,especiallyonethatcanexplainhowobjectsderemodalpropertiesdetermineitssort,suchprimitivismisnotdefensible.Attheveryleast,itisnotdefensiblewithoutareductiveaccountofhowanobjectsformdeterminesitssortandareductiveaccountofhowanobjectsderepropertiesaredetermined. InDefenseofEssentialismRequirements(ii)and(iii)fitneatlyintothisobjectiontosubstanceessen-tialism,sincewhetherornotthesubstanceessentialistagreesthatanobjectsmodalpersistencepropertiesmakeitthesortofthingitis,hisviewmakesnoroomforareductionofhowobjectsgettheirderemodalpropertiesorforareductionofthederetothepurelyqualitative.Thetroublewithnonreductivetheorieslikesubstanceessentialismisthatinthecontextofwell-developedtheoriesofmodality,wherederemodalpropertiesaredeterminedbyobjectsrelationstopossibilia,simplystatingthattheworldissuchthatobjectshaveprimitiveessentialandaccidentalproperties(orsimplyfallundersorts,orhavecertainforms)justisntgoodenough.Tomakemattersworse,suchtheoriesdontseemtofitintolargerreductiveorpartlyreductiveprojectsinvolvingsupervenience,analysesofcausation,dispositions,lawsandpersistence,andmereology,orwithreductionsofpossibilia,mentalstatesandpropertiestomorefundamentalentities,whichgreatlyincreasesthecostoftheview.Thelackofeconomyofsubstanceessentialismshowsupinsubstanceessentialistsinabilitytodefendmanyoftheconsequencesoftheirviews.Suchconsequenceshavebeenwell-discussedintheliteratureonmaterialconstitution.Theproblemofmaterialconstitutionarisesbecauseofdifferencesbetweenpersistenceconditions,whichincludedifferencesinderemodalprop-erties.Intherelevantcases,therecanexistmultiplematerialobjects„suchasapersonandthehunkoftissuethatconstituteshim„inthesameplaceatthesametime.Theexplanationofmaterialconstitutiongiven,forexample,bymanyversionsofsubstanceessentialismisthattherearemultipleobjectsbecauseaperson(Person)isessentiallyhuman-shapedinvirtueoffallingunderthehuman-sort,whilethehunkoftissue(Hunk)constitutingPersonisaccidentallyhuman-shapedinvirtueoffallingunderthetissue-sort.Butthisdoesnotexplainenough.HunkandPersonsharethesamesub-stance(andthesamematerial)andoccupypreciselythesamespatiotemporalregion.IfHunkandPersonsharetheirmaterialandregionandthussharepropertiesliketheirshape,color,location,mass,weight,texture,etc.,thenthepropertiesthatseemtoberelevanttodetermininganobjectssortarejustthepropertiesthatHunkandPersonshare.Inotherwords,itseems,atleastprimafacie,thatthereisnotenoughdifferencebetweenHunkandPersontosupporttheirdifferencesofsort(ordifferencesofform)andtheirensuingmodaldifferences.HowcanHunkandPersonbeofdifferentsorts,orfallunderdifferentsortalconcepts,orhavedifferentformsiftheyshareeverythingthatdetermineswhichsortstheyare,whichsortalconceptstheyfallunderorwhichformstheyhave?Insubstancetheoryterms,thepuzzleisaboutwhatgroundswehaveforsayingthatanobjectfallsunderthepersonsortandnotthetissuesort,orvice-versa.Substanceessentialists,asprimitivists,mustsayitisamatterofbrutefact.Theywillgivenofurtherexplanationofhowobjectsrelatedbyconstitutionhavetheirderemodalpropertiesandhencenoexplanationofthemodaldifference.Thishighlightstheadhocnatureoftheirprimitivism. L.A.PaulArelatedworryinvolvestheabilityofthesubstanceessentialisttoexplainthecoincidenceimpliedbyconstitution.WhenHunkconstitutesPerson,theycoincide:twonumericallydistinctmaterialobjectssharetheirmatterandoccupytheverysamespatiotemporalregion.Butwithoutanyexplanation,thisseemstocontradictourusualwayofthinkingaboutmaterialobjects,i.e.,asobjectsindividuatedbytheirmatterandregion.Onwhatgroundsdowedistinguishthetwoobjects,andhowcanwemakeconceptualsenseofthissituation?Asidefromemphasizingthatthematterandtheirmaterialsaresharedinsomewaywhiletheirformsorsortsarent,thesubstanceessentialisthaslittletosayinresponse„justisaconsequenceofhisview.Again,hisprimitivismpreventshimfromhavingsufficientmaterialstoconstructanadequateexplanation.Thereisafurtherproblem:substanceessentialistswhotakethepersonandthehunkoftissuetobenumericallydistinctwhilesharingtheirmatterandregioncannotendorseclassicalextensionalmereologyandthetheoryofobjectsitimplies(thatobjectsaresimplyhunksofmatterinspacetime).Themereologyofthecoincidentobjectsofthesubstanceessentialistisdarkandmysterious,sincecoincidentobjectsseemtoshareverysmallspatiotemporalpartsbutnotlargespatiotemporalparts.Howcansuchadistinctionbedefended?Whatisthemereologyofsubstancetheoryaccordingtothesubstanceessentialist?Substanceessentialists,anddeepessentialistsgenerally,mustalsofaceob-jectionstothecoherenceoftheirposition.Chisholm(1979)arguesthatShipofTheseus-styleexamplesinvolvingseeminglyplausiblesmallchangesinessentialpropertiesofobjectsacrossmanyworldsrequiredeepessentialiststoaccepthaecceitiesorotherunpalatableconsequences.NathanSalmon(1986)developsarelatedobjectioninvolvingoriginessentialismhecallsTheFourWorldsParadox.ŽAddtotheseobjectionsthenotoriousvariabilityofmodalintuitions:ImightthinkthatSandyisessentiallyhuman,butrevisethatviewwhenremindedof(letsimagine)scientificadvancesthatcanreplaceherbodywithsophisticatedplasticswhileretainingmentalcontinuity,memoriesandotherpsychologicaltraits.Doesntsuchwafflingmakedeepessentialistintuitionsunreliable?Anydeepessentialistwithoutadevelopedstoryofhowobjectshavederemodalpropertieslackstheresourcestoaddressthesepuzzles.Suchessentialistsareintheunfortunatepositionofbeingabletomarshalstrongandplausiblecommonsenseintuitionstosupporttheviewthatobjectshaveessentialprop-ertiesbutofbeingunabletoaccommodatetheseintuitionsinaphilosophicallyrespectableway.Forthesereasons,whileIembracethecommonsenseplausibilityofdeepessentialism,Irejecttheamountofprimitivismneededtosupporttheviewthatthehavingofderemodalpropertiesisanontologicallyprimitivematter.AlthoughIhavefocusedonsubstanceessentialismasthebest-knownversionofdeepessentialismthattakesthisstance,myrejectionisablanketrejection.Anydeepessentialistviewthatdoesnotgiveasufficientlydevelopedaccountofhowobjectshavetheirderemodalproperties,i.e.,thatdoesnotaddressatleast(i)and(ii)oftheskepticalobjection,willfaceproblemswithgivingsatisfactoryaccountsofrelatedontologicaltopics.(Inordertothoroughlyrespondtotheskeptic,one InDefenseofEssentialismshouldalsoaddress(iii).)Theskepticalobjectionhasmerit,andanyacceptabletheoryofdeepessentialismmustrespondeitherbyaccomplishingthereductionsitcallsfororbyshowingwhyprimitivismisacceptableafterall.2.ShallowEssentialismEntertheshallowessentialist.ThissortofessentialistrespondstoQuinesoriginalobjectionbygivinganaccountofthederemodalclaimswemakeaboutanobjectintermsofitsfallingundercertainsortsofdescriptionsincertaincontexts,fleshedoutwithinthecontextofareductiveapproachtomodality.Theshallowessentialistthusoffersareductiveand(whathetakestobea)naturalexplanationforthedifferencesbetweenobjectsthatthesubstanceessentialistortheold-stylecounterparttheoristtookasprimitive,andfullyendorsestheskepticalobjection.Viz.Fine:[t]heQuineanwillrespondthat[differencesbetweenthenecessaryfeaturesofanobject]derivesfromourprivilegingcertaindescriptionsoverothers.Ž(2005,4)Therearethreeimportantfacetsoftheshallowessentialistsposition.Forsimplicityssake,startwiththeshallowessentialistwhoendorsesgenuinemodalcounterparttheory(i.e.,endorsesmodalrealismwithcounterparttheory).Onsuchaview,thefirstfacetofshallowessentialismisadevelopedaccountofhowobjectshavetheirderemodalpropertiesintermsofpossibilia:accordingtotheshallowessentialist,objectshavetheirderemodalpropertiesinvirtueofhavingcounterparts.Thesecondfacetisthereductionofderetopurelyqualitative,generalfacts:objectshavetheircounterpartsinvirtueofbeingsimilartotheminnonmodal,purelyqualitativerespects.(AsInotebelow,onemightaddafurtherrequirementinvolvingcontexthere.)Thethirdfacetisthetreatmentofdererepresentationasdescription-dependentorinconstant.Objectsaresaidtohavetheirmodalpropertiesdependingonhowtheyaredererepresentedbytheircounterparts,butcanonlybesaidtohavetheircounterpartsinvirtueofbeingnonmodallysimilartotheminrelevantways,wherewhatisrelevantisdeterminedbydescribingorevaluatingtheobjectinacertainway.Sotherecanbemanydifferentdererepresentationrelations,dependingonwhichcounterpartsareincludedinthesetthatdoestherepresenting.Thedescriptionornameusedtodenoteanobjectevokesacontextthatmakescertainoftheobjectsnonmodalpropertiesrelevant,whichthendeterminestherelevantcounterpart-theoreticsimilaritiesandsoselectsadererepresentationrelation.Mattersaresomewhatmorecomplexfortheshallowessentialistwhoen-dorsesersatzmodalcounterparttheory,sincedererepresentationfortheersatzistismorecomplex.Ontheersatzistview,thefirstfacetofshallowessentialismisalsoanaccountofhowobjectshavetheirderemodalpropertiesintermsofpossibilia,butthistime,objectshavetheirderemodalpropertiesinvirtueofdererepresentedbyersatzindividualsthatstandincounterpart-theoreticrelationstocounterparts(otherersatzindividuals).Onthisview,tosayanersatz L.A.Paulindividualhasared-hairedcounterpartistosaythattheersatzindividualhasacounterpartthatrepresentssomethingred-haired,andtosaythatan(actual)objecthasacounterpartistosaythatitisdererepresentedbyanersatz(actual)individualthathasacounterpart.Thesecondfacetagaininvolvesareductionderemodalitytopurelyqualitative,generalfacts:theersatzindividualsthatdererepresentobjectshavetheircounterpartsatleastinpartbystandinginsomesortofersatzsimilarityrelationtotheircounterparts.(Intuitively,ersatzsimilarityisdefinedbyhowtheobjectrepresentedbytheersatzindividualwouldbesimilartotheobjectsrepresentedbytheersatzindividualscounterpartsiftherepresentedobjectshadexisted.)Finally,dererepresentationisagaindescription-dependent:theersatzindividualsthatdererepresentobjectsareonlyascribedcounterpartsrelativetocertaincontextsofdescriptionorevaluation,sotherecanbemanydifferentdererepresentationrelationsdependingonwhichcounterpartsareincludedinthestructure(ofanersatzindividualplussomecounterparts)thatdoestherepresenting.Shallowessentialism,whetherrealistorersatzist,thusmeetstheskepticalobjectiononitsownterms.Itprovidesareductiveaccountofderemodalityandofhowobjectshavetheiraccidentalandessentialpropertiesaccordingtowhichderemodalclaimsaboutobjectsaredeterminedinpartbyhowsuchobjectsaredescribed.Onsuchaview,modaldererepresentationisinconstant:thereisnosettledanswer,fixedonceandforall,aboutwhatistrueconcerningacertainindividualaccordingtoacertain(genuineorersatz)worldŽ(Lewis1986,198).Thecharacterizationofthedererepresentationofshallowessentialismasinconstantadmitsoftwointerpretations.Onthefirstinterpretation,shallowes-sentialismtakesmodalpropertiestobecontextdependent.Onthisinterpretation,counterparthoodisdeterminedinpartbyqualitativesimilarity(orersatzsimi-larity),andinpartbycontextsofdescriptionorevaluation.Dererepresentationofanobjectisinconstantbecausethemodalpropertiesareinconstant.Callthisinterpretationofshallowessentialismtheevaluativeinterpretation.Onthesecondinterpretation,shallowessentialismtakesmodalpropertiestobecontextindependent,buttakesmodalpredicationtobecontextdependent.Onthisinterpretation,counterparthoodisdeterminedbyqualitativesimilarity(orersatzsimilarity),buthowcounterpartrelationsareisgovernedbycontextsofdescriptionorevaluation.Here,correctlyascribingmodalpredicatesisdeterminedpartlybywhatwecansayrelativetoacontext(whilecounter-parthooditselfisindependentofcontext),andsodererepresentationofanobjectisinconstantsimplybecausemodalpredicationisinconstant.Callthisinterpretationtheinterpretation.Onbothviews,thetruthofisessentiallyblondeŽcanvarywithcontext,butitvariesfordifferentreasons.Ontheevaluativeinterpretationofshallowessentialism,thecounterpartsandthusthemodalpropertiesofanobjectaredeterminedinpartbycontextsofdescriptionorevaluation,andsoanobjectsmodalpropertiescanvarywithcontext.Ontheantiessentialistinterpretationofshallowessentialism,themodalpropertiesofanobjectaredeterminedby InDefenseofEssentialismcontextindependentcounterpartrelations,butwhatmodalpredicationswecanmakeofanobject(whatcounterpartswecantruthfullyascribetoit)varieswithcontext.Oneitherinterpretation,thereisahugecosttoshallowessentialism:welosetheintuitivepowerofthedeepessentialistviewthatordinaryobjectshaveasubstantialnumberof(absolutely)essentialproperties.Eachviewcarrieswithitabizarreviewofthemodalpropertiesthatobjects,especiallyordinaryobjects,have.Oneachview(butfordifferentreasons)modalprofilesofobjectsnolongercapturewhatwecommonsensicallytakeanobjecttobe,eitherbecause(ontheevaluativeview)modalpropertiesarepartlydependentuponcontextsofevaluation,orbecause(ontheantiessentialistview)objectshave(almost?)noessentialpropertiesandfartoomanyaccidentalproperties.EvaluativeessentialismdeniesthatSandyhasessentialpropertiesinanyabsolute,context-independentsense,forshe(ortheersatzindividualthatrepre-sentsher)doesnothavemodalcounterpartsinanyabsolute,context-independentsense.Thisisfundamentallyimplausible:theevaluativeessentialisthassoldhissoultoQuinebymakingseeminglydeepfactsaboutanobjectsessencepartlydependentuponmeremattersofevaluation.Forexample,accordingtothegenuinecounterparttheoristsversionofevaluativeshallowessentialism,givenmysimilaritytoafriedeggoratree,IcouldhavebeenafriedeggorIcouldhavebeenatreeifacontextexiststhatcanmakeafriedeggoratreeoneofmycounterparts.Similarconsequencesderivefromtheersatzistversionoftheview.Sincesuchsuitablyphilosophicalcontextssurelyexist,thenaccordingtoevaluativeshallowessentialismIampossiblyafriedegg.Thiscannotberight,forIamnotpossiblyafriedegg.Moregenerally,evaluativeshallowessentialismisfalsebecausethemodalpropertiesofanobjectdetermine„independentlyofacontextofevaluation„whetheranobjectcouldsurviveachangeinitsshapeoritscolororwhetheritcouldbeafriedegg.Whatareisdeterminedbytheirmodalproperties,andthesearedeterminedinacontextindependentmatter.Theproblemwithevaluativeshallowessentialismcomesdowntothefactthatiftheessentialandaccidentalpropertiesofanobjectarenotabsolute,thentherearenoabsolutefactsaboutwhatanobjecthastobelikeinorderforittoexist.Andthisseemswrong.Theevaluativeversionofcounterparttheoryanditsinconstantdererepresentationshouldberejectedforthisreason.Theantiessentialistversionofshallowessentialismisevenlessacceptable.Thisviewtrivializesessentialism,preservinglittleabouttheessentialistpositionotherthanthepermissiontomouthcertainessentialistattributionsincertaincontexts.Forexample,accordingtothegenuinecounterparttheoristsversionofantiessentialistshallowessentialism,sinceanythingissimilartoalmosttoanythingelseinsomeway(speakinginthebroadestmetaphysicalsense),everyobjecthasanimmensevarietyofcounterparts.Onthisview,sincefor(almost)anyproperty,anobjectwillhaveacounterpartthatis,anobjectispossiblyfor(almost)anyproperty L.A.PaulOnthisview,Ihavefriedeggsandtreesascounterparts(sinceIamsimilarincertainrespectstofriedeggsandtotrees),butwecanonlytruthfullythatIampossiblyafriedeggŽorIampossiblyatreeŽifcontextpermits.Onsuchaview,thereisnosubstantialwayforanobjecttohaveauniqueanddistinctivemodalcharacter,sinceanythinghasalmostanythingasacounterpart,andsothereisverylittlecontentintheideathatobjecthastobeacertainwayinorderforittoexist.Ifthereisalmostnowayanobjecttobeinorderforittoexist,thewhat-it-isŽ-nessofanobjecthasbeeneviscerated.Similarconsequencesderivefromtheersatzistversionofantiessentialistshallowessentialism.Again,weshouldrejecttheview.Whetherevaluativeorantiessentialist,theshallowessentialistturnsouttobeanessentialistonlyinthesensethathegrantsthatobjectshavederepropertiesofsomesortandwillgrantthetruthofderemodalclaimslikeSandyisessentiallyfemaleŽastruewithrespecttoacertaincontextorclassofpossibilia.Thetrue-heartedessentialistmightwellthinkmeafalsefriend,aQuineanskepticinessentialistsclothing.Ž(Lewis1968[1983],42)Lewisisright:shallowessentialismisnosortofacceptableessentialismatall.(Notethatthedeepessentialistmaystillacceptaversionofderemodalityasdeterminedbyconstantdererepresentationandcounterparttheory.Suchviewwouldholdthatthereasettledanswer,fixedonceandforall,aboutwhatistrueconcerningacertainindividualaccordingtoacertain(genuineorersatz)world.See4forjustsuchanaccountofdeepessentialismintermsofconstantdererepresentationandconstantersatzmodalcounterparttheory.)Theissuesherearemirroredbyanothersortofcounterparttheorythathasbeenrecentlybeendefended:temporalcounterparttheory(Hawley2001,Sider2001).Examiningtemporalcounterparttheorywillhelptofurtherclarifywhatissoobjectionableaboutshallowessentialism.Temporalcounterparttheoryholdsthatobjectshavetheirthis-worldlytemporalpersistencepropertiesinvirtueofhavingtemporalcounterparts.Iftemporalcounterparttheoryisanalogoustomodalcounterparttheory,thenpresumably,iftemporalcounterparttheoryistrue,objectshavetheirtemporalpersistencepropertiesinvirtueoftheirtemporalcounterparts.Forexample,SandywillbeblondeatŽistrueiffSandyhasablondetemporalcounterpartatUndertemporalcounterparttheory,itisnaturaltoadoptastagetheoryofhowobjectspersist.Accordingtothestageview,anobjectpersistsbybeinganinstantaneoustemporalstageofatemporallyextendedcrosstimeworm.Instan-taneousstagesareordinaryobjects,inthesensethattheyarethereferentsofordinaryterms,membersofordinarydomainsofquantification,subjectsofordinarypredications,andsoon.Ž(Sider2001,60)Temporalcounterpartrelationscanthenunitethestagesofspacetimeworms.Stagetheoryhasmuchtorecommendit,andsomeversionofstageorwormtheorymaybecorrect.Butjustaswehaveshallowanddeepessentialism,wecanhaveshallowanddeeppersistencetheory.Forthesamereasonsthatweshouldrejectshallowessentialism,weshouldrejectshallowpersistence.Theobjectionis InDefenseofEssentialismnottostageortemporalcounterparttheory,buttostagetheorybuiltontheshiftingfoundationofinconstantderetemporalrepresentation.Accordingtoshallowpersistencetheory,thetemporaldererepresentationrelationisinconstantinjustthewaytheshallowessentialisttakesthemodalrepresentationrelationtobeinconstant:foranystage,whetherandhowthatstageisdererepresentedbyatimeisrelativetodifferentcontextsofevaluationordescription.AnytemporalcounterparttheoristwillholdthatSandywillbeblondetomorrowŽistrueiffSandyhasablondetemporalcounterparttomorrow.Butaccordingtotheshallowpersistencetheorist,whichtemporalcounterpartsSandycanbesaidtohavedependsonourcontextofevaluation.Shallowpersistenceadmitsofmultipleinterpretationsinthesamewaythatshallowessentialismdoes:onemaytaketemporalcounterparthoodtobecontextdependent,oronemaytaketemporalpredication(butnottemporalcounterparts)tobecontextdependent.Eitherway,accordingtotheshallowpersistencetheorist,thetruthofIwasatthebarlastnightŽmightvarybecausethereisnosettledanswer,onceandforall,aboutwhatistrueconcerningmeaccordingtoatime,sinceingeneral,thereisnosettledansweraboutwhatistrueaboutindividualsaccordingtotimes.Thismaybebecauseourtemporalpropertiesdependuponcontextsofevaluation,orbecausewhichsuchpropertieswecanascribetoanobjectdependsuponcontextsofevaluation(andwehavemanymoretemporalpropertiesthanwecommonsensicallythought,sincetemporalcounterparthoodisdeterminedbysimilarityalone).Ontheevaluativeversionofshallowpersistence,whetherornotIwasatthebarlastnightdependsonwhatmytemporalcounterpartsare,andthisdependsoncontextsofevaluation.Ontheantiessentialistversionofshallowpersistence,thingshaveasmanytemporalcounterpartsassimilaritywillallow,buttemporalpredicationsarecontext-dependent.Onthelatterview,Ihaveabar-goingcounterpart,andasleepingcounterpart,andafried-eggcounterpart,etc,allofwhichexistatappropriatetimes„whatstopsitfrombeingtruethatIwasatthebarlastnightŽ(orIwasafriedegglastnightŽ)isnotthemetaphysicsofthetemporalfacts,butmere:thereisnocontextofevaluationinwhichtomaketheclaim.Eitherversionofshallowpersistenceisimplausible.Sothetruthaboutwhetherornothaveatemporalcounterpartthatwasatthebarlastnight(orwhetherwillhaveablondecounterparttomorrow,etc.)isnotdependentonacontextofdescriptionorevaluation.Moreover,Idonthavesuchacounterpart,fullstop(thoughIhaveonethatwasatthebartwonightsago).Suchthingsareamatterofconstanttemporaldererepresentation:anotherwaytoputthepointistosaythatthegen-identityrelationisabsolute.Shallowviewsfailtorespectintuitivelycorrectcontext-independentmodalandtemporalpersistencefactsaboutobjects.Whetherthedebateisovertemporalpersistenceoroverderemodality,Iconcludethatitisbettertobedeepthantobeshallow.Ifdeepessentialismcanbejustifiedinthefaceoftheskepticalobjection,itmustbepreferred:weshouldnotbedriventoshallowwatersunlessthereisnohopeofotherwisesurviving. L.A.PaulTheworldissuchthatthereisanabsolutefactofthematterwhetherIwouldsurviveifIchangedmyhaircolor,whethersomethingwithadifferentorigincouldhavebeenme,andthatIcouldnotsurvivebeingturnedintoafriedegg.Byrecognizingordinaryobjectsasdiscrete,persistingindividualswithcharacteristichistoriesseparablefromtherestoftheworld,wearerecognizingthatobjectshaveanumberofdistinctive,absoluteessential(andtemporal)propertiesandthatthetruthsaboutsuchpropertiesarecontextindependent.Thesefactsaboutthepropertiesofordinaryobjectsarefactsaboutthenatureoftheworld,andweneedanaccountthatcanpreservethesefacts.Shallowexplanationsareunacceptable.3.DeepEssentialismReduxShallowessentialismeliminatestheveryheartofwhatmotivatesessentialisminthefirstplace,butithasseemedtomanytobetheonlygameintownforfansofreductive(butnoteliminativist)treatmentsofderemodality.Theskepticalobjectionshowsthattheprimitivismofsubstanceessentialismand(toalesserextent)old-stylecounterparttheoryisuntenable.Manyseemtothinkthat,inthefaceoftheskepticalobjection,adoptingshallowessentialismorsomeequallytoothlessviewistheonlyoptiontheessentialisthas.Mustthosewhowishtopreserveanexplanationofourderemodalattributionsresorttoshallowessentialism?No.Realessentialistscandobetter.Tojustifyareturntodeepessentialism,weneedadeepessentialistaccountthatavoidsunmotivatedoradhocprimitives,givesareductiveaccountofhowobjectshavetheirderemodalproperties,takesdererepresentationtobeconstant,and,ideally,allowsderemodalitytobereducedtopurelyqualitativefacts.Ishalldevelopsuchanaccountbelow.Likemanyothers,Ipreferanersatzismthatcharacterizesthehavingofmodalpropertiesintermsof(modal)dererepresentation,i.e.,intermsofobjectsbeingdererepresentedbyersatzpossibilia.Themostpopularversionsofersatzismtakeersatzpossibiliatobemaximalconsistentsetsofsentences,orpropositions,orstatesofaffairs,orthelike,andtakemodaltruthstobeevaluableintermsofsuchabstractobjects.Forexample,accordingtoRobertAdams,Formetofeelapaininsomepossibleworldisjustforaproposition,totheeffectthatIfeelapain,tobeamemberofacertainsetofpropositions.Ž(1979,p.205)Liketheshallowessentialist,Ithinkthebestreductiveapproachtohowobjectshavetheirderemodalpropertiestakesthehavingofderemodalpropertiestobedeterminedbyrelationstoersatzpossibilia.Therefore,Ishallunderstanddererepresentationintheseterms,butbroadly,intermsofobjectsbeingdererepresentedbysomesortofabstractobjectsorrelationalcomplexesofabstractobjects,whatevertheymaybe.(Inotherwords,Iwillnotchoosebetweenersatzpossibiliaaspropositions,setsofsentences,statesofaffairs,etc.,norwillIspecify InDefenseofEssentialismthemethodofdererepresentation,althoughlinguisticersatzismisprobablytheleadingcandidate.)Now,adeepessentialistsuchasmyselfneedsaresponsetotheskepticalobjectionthatwillsatisfyitsreductiverequirements.If(forexample)anaccountderemodalpropertiesintermsofersatzpossibiliathatare(setsof)proposi-tionsiscorrect,thenIampossiblyinpainthepropositionthatIaminisamemberofacertainsetofpropositions,nottheotherwayaround.Correspondingly,ifanaccountofderemodalpropertiesintermsofersatzpossibiliathatare(possible)statesofaffairsiscorrect,thenIampossiblyininamaximalnonactualstateofaffairsinwhichIamincluded,Iaminpain„nottheotherwayaround.Thispointstoafundamentaldifferencebetweenmyviewsandthoseofmanycontemporarydeepessentialists,includingmanyersatzistones,sinceIrejecttheviewthatderemodalpropertiesarehadprimitivelyandIrejecttheviewthatanobjectsderemodalpropertiesdetermineitsrelationstopossibilia.Mostdeepessentialistsrejectreductionismabouthowanobjectgetsitsdereproperties:eithertheytakederemodalityasflat-outprimitiveandhavenotruckwithpossibilia,ortheytakeanobjectsderemodalpropertiestodeterminehowitisrepresentedbypossibilia.AsIdiscussedin1,theseassumptionscreateproblemsthatleadtotheskepticalobjection.So,liketheersatzistshallowessentialist,mystoryabouthowobjectshavederemodalpropertieswillbeconductedundertheauspicesofanersatzistreduction.Alsofollowingthe(ersatzist)shallowessentialist,letustakeobjectstobedererepresentedbyersatzindividualsthathavecounterparts,wherecounterparthoodisbasedonersatzsimilarity.Toavoidskepticalworriesaboutunmotivatedselectivityprinciplesforcounterparts,liketheersatzistshallowessentialist,Iwillassumethatersatzsimilarityisontologicallycheap:ersatzsimilarityis(somehow)ascheapassimilarity,andsinceeverythingissimilartoeverythingelse,ontologicallyspeaking,similarityextremelycheap.Thismeansthatthecounterpartrelationisextremelyeasytoget.(Twocaveats.One:attheendofthissectionIllcomebacktoanimportantissuethatrelyingonersatzsimilarityraises.Two:onewayIamnotfollowingshallowessentialistsisthatItakedererepresentationtobeconstant.Idiscussthisbelow.)Myversionofersatzismwilltakeobjectstohavetheirderemodalpropertiesinvirtueofhavingcounterpartsinthefollowingway.Anobjectdererepresentedbyanersatzindividual,wherehascounterparts(usually,otherersatzindividuals).Whendererepresentedbyastructuredcomplexofplussomeofitscounterparts,hascertainderemodalproperties.Forexample,theactualHumphreyisdererepresentedbyanersatzactualHumphrey,suchthattheersatzactualHumphreyhasacounterpartthathasthepropertyofbeingawinner.BecausetheactualHumphreyisdererepresentedbyanersatzactualHumphreythathasacounterpartthatisawinner,theactualHumphreyispossiblyawinner.Moresimply:theactualHumphreyisdererepresentedas L.A.Paulhavingawinningcounterpart(becausetheactualHumphreyisdererepresentedbyanersatzactualHumphreythathasawinningcounterpart),soheispossiblyawinner.Evenmoresimply:ifHumphreyisdererepresentedasawinner,heispossiblyawinner.Thisapproachisdifferentfromthatoftheersatzistshallowessentialistwithregardtothenatureandtheconstancyofthedererepresentationinvolved.Giventheshallowessentialistsaccountofdererepresentation,itiseasyfortheersatzactualHumphreytohavemanycounterparts,andthuseasyforHumphreytobedererepresentedinmanydifferentways„contextpermitting,ofcourse.Forexample,theshallowessentialistholdsthateventhoughwecansaythatHumphreyispossiblyawinner,wecanalsosaythatHumphreyisessentiallyaloser.Wecansaythis,claimstheersatzistshallowessentialist,becausethereisacontextinwhichtheactualHumphreycanbedererepresentedbyanersatzactualHumphreythathasnowinningcounterparts.Inadifferentcontext,holdstheshallowessentialist,wemaysaythatHumphreyispossiblyawinnerbecauserelativetothiscontext,theactualHumphreyisdererepresentedbyanersatzactualHumphreythathasawinningcounterpart.Thus,ontheshallowessentialistview,anobjectlikeHumphreycanbedererepresentedinmanydifferent,conflicting,ways.Asadeepessentialist,Ithinkthisisallwrong.Objectsaredererepresentedconstantly,sothetruthaboutwhetheranobjectisessentiallyoraccidentallydoesnotdependoncontext.SupposethatthepersonIrefertoasHumphreyŽisessentiallydescendedfromhisparents,RagnildandHubert.Anersatzistdeepessentialistwillthusholdthatthisperson,Humphrey,isnotdererepresentedashavingcounterpartswithdifferentparents,nomatterwhatthecontext.Likewise,Humphreyisonlyaccidentallyaloser.Inersatzistterms,thismeansthattheersatzactualHumphreydoesnotdererepresenttheactualHumphreyashavingcounterpartswithdifferentparents,anddoesdererepresenttheactualHumphreyashavingwinningcounterparts.Period.OrconsiderSandy:shehasheroriginessentially,soshecouldnothavedevelopedfromadifferentzygote.Sandyisalsoablonde,buthadshedecidedtodyeherhairred,shewouldhavepersistedthroughthechange,andinthissenseshecouldhavebeenaredhead.(Shealsocouldhavebeenaredheadbecauseofamistakemadebyanincompetentstylist,butthisisamoretroublesomesenseofcouldhavebeenaredhead.Ž)IfSandycouldnothavehadadifferentorigin,thensheisnotdererepresentedwithoutthatorigin,butifshecouldhavebeenaredheadinsteadofablondethensheisdererepresentedashavingredhair.Inotherwords,ifSandycouldnothavehadadifferentorigin,thenshehasnocounterpartsthathavedifferentorigins,i.e.,sheisnotdererepresentedashavingcounterpartswithdifferentorigins.Thatis,inmyersatzistterms,theersatzindividualthatdererepresentsherhasnocounterpartswithdifferentorigins.IfSandycouldhavebeenaredheadinsteadofablondethenshehasred-headedcounterparts,i.e.,sheisdererepresentedashavingredhairbybeingdererepresentedbyanersatzindividualthathasred-headedcounterparts.(From InDefenseofEssentialismnowon,IwillassumethatclaimslikeSandyhasaredheadedcounterpartŽaretobeunderstoodinmyersatzistwayunlessotherwisespecified.)Ofcourse,asInotedabove,counterparthoodisextremelyeasytoget,sinceersatzsimilarityisextremelycheap.How,then,cantheersatzist(oreventherealist)deepessentialistmaintainthatHumphreyhasnocounterpartswithdifferentparentsorthatSandyhasnocounterpartswithdifferentorigins?Mustsheaddfurtherrequirementstothecounterpartrelationtomakeitextremelyselectiveornonqualitativeinsomemysteriousway,justasLewisdidinhisoriginalcounterparttheory?Doesnonreductionismthreaten?Doestheskepticalobjectionloom?Thedeepessentialistdoesnotneedtoendorseaselectivecounterpartrelation,buttoseewhywilltakesomehonestmetaphysicallabor.Inordertohaveareductiveaccountofhowobjectshavetheirderemodalpropertiesthatcanaddresstheskepticalobjection,Ineedmorethanjustersatzismandcounterparttheory:Ialsoneedtohaveareductiveaccountofobjects.Weneedtoknowjusthowobjectshavetheirderemodalproperties,justhowobjectsaredererepresented,andjusthowdererepresentationplaysaroleinwhatanobjectisinordertoshowhowwecanhaveacheapcounterpart-theoreticrelationwhilekeepingdererepresentation(andthusderemodalpropertiesandderepredications)constant.DeReModalPropertiesandDeReRepresentationMydeepessentialistexplanationofwhatitisforanobjecttohavederemodalpropertieswillcharacterizeanobjectasaofproperties,soobjectshavepropertiesasqualitativeparts.ThismeansthatIembraceapropertymereologyinadditiontoaspatiotemporalmereology.Compositionwithrespecttoqualitativepartsisrestricted.Thistheorytakesordinaryobjectstobenothingmorethanbundlesofproperties,suchthatbundlingisatypeofmereologicalfusion.MypropertymereologyhasanantecedentinNelsonGoodmans(1951)interpretationofRudolfCarnaps(1928)phenomenalisticconstructionofqualityclassesintheAufbau.Goodmandevelopsaversionofapropertymereologyintermsofspatiotemporallylocatedtrope-likepropertyappearances.Goodmanssystemexpandsuponthelittle-noticedpointthatLeonardandGoodmans(1940)formalcalculusofindividualsincludedpropertyinstancesaspartsalongwithspatialandtemporalparts.AmereologicalversionofbundletheorycanalsobefoundintheworkofD.C.Williams(1953,1986).(AsWilliamsisnotevenminimallyexplicitabouthowatrope-theoreticmereologicalapproachistobeformulated,hecannotbeseenasventuringbeyondmorethanastraightforwardadoptionofanonphenomenalistversionofGoodmanssystem.)Unlikemanyadvocatesofbundletheory,Iamnotdefendingaconceptualanalysisofthetermordinaryobject,Žbutratherareductivetheoryofthe L.A.Paulontologyofobjects,especiallyordinaryobjects,intheactualworldandinworldssuitablylikeours.Iamalsoperfectlyhappytoallowforothersortsofentitiestoexistintheworld,suchasspatiotemporalrelationsorpoints,sothereisnoneedtoconstructeverythingintheworldoutoffusionsofpropertiesalone.Finally,Ipreferarelativelysparseapproachtoproperties:notjustanypredicatedefinesaproperty,andtherearenonegativeproperties,merelynegativepredicates:tosayanobjectincludesisforittofailtoincludeinitsfusion.Thisshouldbeenoughofasketchofmyviewforcurrentpurposes:therearemanyotherfeaturesofmymereologicaltheoryofobjects,buttheyarenotdirectlyrelevanttowhatfollows.InterestedreadersshouldconsultPaul(2002),Paul(2004),andPaul(2006forthcoming)formoredetails.Nowthatwehaveatheoryofobjectsinhand,wecanreturntotheissueofhowobjectshavetheirderemodalproperties.ThewaymyviewreductivelyaccommodatesSandysessentialandaccidentalpropertiesistoreducethemtofusionsofSandysbasicnonmodalproperties,suchasherpropertiesofblondebeingoforiginOhavingmassm,etc.,plustherelationalpropertiesofdererepresentedincertainways.Thiscombinesmymereologicalbundletheorywithanersatzistreductiveaccountofhowobjectshavetheirderepropertiesinawaythatisperfectlyconsistentwiththespiritofanersatzistaccountofderemodality.Illstartwithaconcreteexampleinordertofocusthediscussion.RecallSandy:shehasheroriginessentially,soshecouldnothavedevelopedfromadifferentzygote.Sandyisalsoablonde,butshecouldhavebeenaredhead.Thissupervenesoncertainsortsofdererepresentationalfacts:ifSandycouldnothavehadadifferentorigin,thensheisnotdererepresentedwithoutthatorigin(shehasnocounterpartswithdifferentorigins),butifshecouldhavebeenaredheadinsteadofablondethensheisdererepresentedashavingredhair(shehasredheadedcounterparts).Onmyview,ifSandyisaccidentallyblonde,thesumthatisSandyincludesthepropertyofbeingblondeplusthemonadicrelationalpropertyofbeingdererepresentedasred-haired:inthisway,sheisdererepresentedasbeingred-haired.IfSandyhasheroriginessentiallysheincludesthepropertyoforiginOanddoesnotincludeanyrelationaldererepresentationalpropertyofbeingrepresentedwithsomeotherorigin.Thishighlightsanimportantsuperve-nienceprinciplethatispartofmytreatmentofderemodalproperties:derepropertiessuperveneonfusionsofnonmodalpropertieswithrelationalpropertiesofbeingdererepresentedincertainways.(CallsuchrelationalpropertiesdererepresentationalpropertiesInparticular,foranyobjectaccidentallyFiffandincludesthedererepresentationalpropertyofbeingrepresentedasnot-essentiallyFiffandlacksthedererepresentationalpropertyofbeingrepresentedasnot-.(Recallthatnegativepropertiesarereallyabsences InDefenseofEssentialismofcontraryproperties.BeingessentiallyFŽmeansonelackstheabilitytobeF,soonesfusionlacksacertaindererepresentationalproperty,i.e.,representedasItisimportanttobeclearaboutwhatthesedererepresentationalpropertiesare.Callthesumofthebasicnonmodalpropertiesofanobjectitscore(Roughly,suchbasicnonmodalsarephysicallyfundamentalpropertyinstances.Forsimplicityssake,Illassumethatthecorecanbespecifiedusingsomewhatlessfundamentalphysicalpropertyinstanceslikethoseofhavingcolor,havingorigin,havingmass,etc.,andtakepropertiesŽtorefertopropertyinstances.)Callthepropertiesincludedinthesumthecoreproperties.dererepresentationalpropertiesaremonadicrelationalpropertiesthatareontologicallygeneratedbythecoreoftheobjectstandinginamodaldererepresented-byrelationtopossibiliaandcanbeincludedintheobject:Idefinethisastheobjectsdererepresentedbythesepossibilia.SoSandyisdererepresentedbysomepossibiliawhenhercorestandsinamodaldererepresented-byrelationtothesesheincludeshercoreplusthedererepresentationalpropertiesgeneratedbyhercoresstandingintheserelations.Thepossibilia,asisusualforersatzviews,areersatzindividualsthathavecounterparts:thus,onmyersatzistview,acoreofanactualobjectstandsindererepresented-byrelationstoanersatzactualindividualthatcanhaveotherersatzindividualsascounterparts.Invirtueofstandinginsuchadererepresented-byrelation,thecoreoftheobjectgeneratesarelationaldererepresentationalproperty.deremodalpropertiesreductivelysuperveneoncorepropertiesplusrepresentationalproperties,andforanobjecttohaveaderemodalproperty,boththecorepropertiesandthecore-generateddererepresentationalpropertiesthatthederemodalpropertysupervenesonmustbeincludedinthesumthatistheobject.Thishasaveryimportantconsequence:sincederemodalpropertiessuper-veneoncorepropertiespluscore-generateddererepresentationalproperties,modalpropertiesareincludedinthesumthatistheobject.WecanpicturetheunderlyingontologyofhowSandyisaccidentallyblondeusingFigure1(forsimplicity,IveassumedthattheersatzindividualthatdererepresentsSandyhasonlyonecounterpart).Ifanobjectsderemodalpropertiesareincludedinwhatitis,whenweusenamesandotherreferringexpressions,theypickoutindividualsthataresumsthatincludebasicnonmodalpropertiesandcore-generateddererepresentationalproperties(andperhapsotherrelevantproperties).Intheusualcase,theydonotpickoutsumsthatincludetheersatzindividualsthathelptogeneratethedererepresentationalproperties:thesearenotpartsofobjectssuchasSandy.Myaccountofhowobjectshavetheiressentialandaccidentalpropertiesgivesthedeepessentialisteverythingsheneedstorespondtotheskepticalobjection.Itisworthbeingexplicitabouthowthisresponseshouldgo.Firstofall,inresponsetorequirement(i)oftheskepticalobjection,Irejectthe L.A.Paul SandyErsatz individual with a redheaded counterpart. De re representational property of being represented as red haired representation relation Figure1.TheDeepEssentialistAccountofhowSandyisAccidentallyBlonde.needforanyadhocorprimitiveprivilegingofersatzsimilaritiesinordertorestrictthecounterpartsofersatzindividuals.Ersatzindividualshaveafullrangeofcounterparts,restrictedonlybytheweaklimitations(ifany)imposedbyersatzsimilarity.However,analogouslytoshallowessentialism,differentdererepresentationrelationsexistforthedifferentwayscounterpartsareincludedinthestructurethatdoestherepresenting.(Theanalogyisincompletebecausecontextdoesnotplayarole:thereisafullrangeofdererepresentationrelationscorrespondingtoeverywayofcarvingupthestructureoftheersatzindividualanditscounterparts.)Anersatzindividualthusstandsinmanydifferentdererepresentationrelationstoacore,sinceithasmanycounterparts,andsomanydifferentdererepresentationpropertiesaregeneratedbythatcore.However„andhereisthekey„sinceanobject(suchasSandy)iscarvedatitsqualitativejoints,itneednotincludeallofthesecore-generateddererepresentationalpropertiesjustbecauseitincludesthecore.So:Sandyincludesthedererepresentationalpropertiesofhavingredheadedcounterparts,butneednotincludeotherdererepresentationalproperties(e.g.,shedoesnotincludecore-generateddererepresentationalpropertiesofhavingcounterpartswithdifferentorigins).Sandyisdererepresentedasbeingredheadedifsheincludeshercoreandthecore-generateddererepresentationalpropertiesofhavingredheadedcounterparts,andnototherwise.Sandyisaccidentallyblondeiffsheincludeshercoreandtherelevantcore-generateddererepresentationalpropertiesofhavingcounterpartswithdifferentlycoloredhair,andnototherwise. InDefenseofEssentialismSoitisnotacontext-dependentmatterthattheobjectthatwearecallingSandyŽisaccidentallyblonde.Objectsthatincludemorecore-generateddererepresentationalpropertieswillusuallyhavemoreaccidentalproperties.IwilldiscussthisinmoredetailinMyresponsestorequirements(ii)and(iii)intheskepticalobjectionshouldbeclearfromtheforegoingdiscussion:objectshavetheirderemodalpropertiesbyincludingcore-generatedpropertiesofbeingdererepresented,inthewayIdescribedabove,suchthatthesedererepresentationalpropertiesareontologicallygeneratedbytheobjectscorestandingindererepresented-byrelationstoersatzThus,deremodalityisreducibletothepurelyqualitative,inaperfectlyacceptableersatzist,counterpart-theoreticway.Toseewheremyviewislocatedontheessentialistspectrum,itwillbehelpfultoreviewDavidLewissessentialisttreatmentsofderemodalityandcomparemyviewtohis.Lewissoriginalcounterparttheorywasadeepessentialist,partlyreductiveaccountofhowobjectshadtheirderemodalproperties.OnLewissoriginalaccount,objectshadtheirderemodalpropertiesinvirtueofhavingcounterparts,andanobjectscounterpartsweredeterminedbysimilaritiestothebasicnonmodalpropertiesoftheobject.Thepropertiesoftheobjectscounterpartsthendeterminedwhichpropertieswereessentialandwhichwereaccidentaltotheobject.However,sinceobjectscanbesimilartomanydifferentindividualsinmanydifferentways,iftheintuitivecontentofdeepessentialismistobepreserved,notallsimilaritiescanbetreatedasequal.Unfortunately,asImentionedin1,theprivilegingofsimilaritiescanbegivennofurtherexplanation,andthusseemsarbitrary.QuineanskepticsrejectLewissoriginalcounterparttheoryforthisreason.Inresponsetotheproblemofarbitrariness,Lewisjettisonedhisoriginalcounterparttheoryandadoptedshallowessentialism,accordingtowhichclaimsaboutcounterpartsarerelativetocontextsofevaluationordescription.Thus,Lewissoriginalreductionofhowobjectshaveandcanbesaidtohavetheirmodalproperties,spelledoutintermsofconstantdererepresentation,wasreplacedbyareductionintermsofinconstantdererepresentation.LetmeemphasizeakeymoveimplicitinLewissapproachthatshowswhyonestheoryofobjectsmatterstoonesaccountofessentialismandderemodality.WhatItaketobethecoreofanobject„itsfusionofbasicnonmodalproperties„iswhatLewistakestotheobject.Lewissmistakeinbothhisoriginalandhisrevised(genuine)counterparttheoryistooverlooktheexistenceofmonadicrepresentationalproperties,andthustoexcludederemodalpropertiesfromwhatanobjectis.AccordingtoLewis,ifSandyispossiblyred-haired,thisisbecausewhatLewistakestobeSandy(butwhatItaketobehercore)standsinacounterpartrelationtosomethingthatisred-haired.ForLewis,whenanobjectstandsinacounterpartrelationtootherobjects,neitherthecrossworldcounterpartrelationnorthecounterpartitselfisincludedinwhattheobjectis,norisanymonadicrelationalpropertysuchasbeingdererepresentedasFincluded.SinceforLewis, L.A.Paul SandyRedheaded counterpart De re representation relation Figure2.TheLewisianAccountofhowSandyisAccidentallyBlonde.SandysderemodalpropertiesreductivelysuperveneonmorethanwhatSandyis„theysuperveneonrelationalcomplexesthatincludeSandyaswellashercounterparts„thenwhateverSandyis,sheisnotsomethingthatincludeshermodalproperties.Thisisparticularlyclearwhenwerecallthat,despitehismodalrealism,LewisinsiststhatindividualssuchasSandyareworldbound.IfSandyisworldbound,shecannotincludehercounterpartrelations,sincesuchrelationsarecrossworldrelations.Thus,Lewissviewimplicitlyseparatesoutderemodalpropertiesfromtherestoftheobject.TheLewisianapproachispicturedinFigure2.Inordertobringoutthecomparisonbetweenmyviewandtheshallowessentialistevenmoreexplicitly,consideranersatzistwhofollowsaLewisianshallowessentialistprogram.Thissortofshallowessentialistinterpretscounter-parttheoryintermsofersatzindividuals,butinotherimportantrespectsfollowsLewissleadasshowninFigure3.LewisandhisfollowerscouldacceptthatSandyincludesthis-worldlyrelationalpropertiessuchasmodaldererepresentationalpropertiesgeneratedbySandyscorepropertiesstandingindererepresentationrelationstopossibilia(whichwouldkeepSandyworldbound),buttheydonot.ForLewisandhisfollowers,thepropertyofbeingpossiblyred-hairedcanbepredicatedSandy,butitisnotconstitutiveofwhatSandyThisviewisextremelycounterintuitivefromthedeepessentialistperspective.deremodalproperties,superveninginpartonsimilarityrelationstocounterparts,arelikespatiotemporalrelationsandpointsinspacetime:theyarenotincludedinwhatwetakeanobjecttobe.Bykeepingderemodalpropertiesexternaltotheobjectsthathavethem,Lewisalienatesobjectsfromtheirdere InDefenseofEssentialism SandyErsatz individual with a redheaded counterpart De re representation relation Figure3.TheErsatzistShallowEssentialistAccountofhowSandyisAccidentallyBlonde.modalnatures.KeepingderemodalpropertiesexternaltoobjectsmakesiteasyforLewis(andhisfollowers)todropdeepessentialisminfavorofashallowessentialismthattakestruthsaboutobjectsnaturestobedependentonhowtheyaredescribedorevaluated.Oncewehavemadethefalsestepofalienatingobjectsfromtheirmodalproperties,itiseasytocontinuedownthedangerouspathtoshallowessentialism.Takingderemodalpropertiestobepartlyexternaltotheobjectsthathavethemisstrangeifwewanttoholdontotheintuitiveideathatanobjectsderemodalpropertiescaptureitsnature,orthatderemodalpropertiesbear,inthemetaphysicallysignificantsenseofthephrase,onwhatanobjectis.Ž(Fine1994,p.2)Foritdoesnotseemrighttosaythatsomethingthatistheessenceofanobjectispartlyexternaltotheobject:suchthingsshouldbeincludedinwhattheobjectis,likebeingblondebeingfemale.Forthedeepessentialistwhotakestheskepticalobjectionseriously,thismustbetrueevenwhilethemodalcharacteroftheobjectisdeterminedbyobjectsoutsideofitself.Objectsnaturesareitspotentialities,thatis,anobjectsnatureisdeterminedbywhatitpotentiallycanorcannotdo(orcanorcannotbe),thatis,determinedbywhatitis(context-independently)dererepresentedasbeingordoing.Butsinceitsnatureispartofwhatanobjectis,weneedtoincludeitinthesumthatistheobject.Iagreewiththeshallowessentialistthat(ersatz)similarityiseasytoget,andthatweshouldnotprimitivelyrestrictwhichcounterpartsanobject(oranersatzindividual)has.Butinsteadofmakingdererepresentationinconstantwhenfacedwithskepticalworries,weshouldpaymoreattentiontothewayobjectsareconstructedandthewaytheyhavepropertiessuchasderemodalproperties.Itisnotenoughforthecoreoftheobjecttostandinadererepresented-byrelation L.A.Paultoasuitableindividualtosaythattheobjectisaccidentally:theobjectmustalsoincludethecore-generatedpropertyofbeingdererepresentedasnot-Fitsfusion.However,ifwecanbesurethatan-objectexiststhatincludesthecore-generatedpropertyofbeingdererepresentedasinitsfusion,wecanendorseconstancy:thefactthattheobjectisaccidentallywillnotdependonacontextofdescription.(Beingsurethatsuchanobjectexistsisimportant:Ishalldevotethenextsectiontothistopic.)Shallowessentialismmakesobjectsnaturesexternaltothemand(attheveryleast)takesclaimsaboutnaturestobedependentoncontextsofevaluation,whilemydeepessentialismtakesobjectsnaturesandclaimsaboutthemtobeindependentofcontextyetoutward-pointing.Thedeepessentialistmustsomehowgiveanexplanationofanobjectsnaturethatcapturesthewayitisoutward-pointing(i.e.,hypothetical)whilekeepingitinternaltotheobject:thewaytocapturebothelementsofanobjectsderemodalnatureistograntthatcertainpropertiesincludedinwhattheobjectisarepropertiesthatarepartlydeterminedbyexternalfacts,i.e.,bythewaystheworldmightbe.Thereisonefinalcomplicationwiththispicture.Ihavebeendescribingmydeepessentialist,ersatzisttreatmentofcounterparttheoryassupervenientonersatzsimilarity.ThisisbecauseIamnotamodalrealist,so,strictlyspeaking,Icannottakecounterparthoodtosuperveneonsimilarity.Instead,justasanyersatzerwhodefendsasimilarity-basedaccountmustdo,Imustholdthatcounterparthoodsupervenesonsomesortofqualitativerelationofersatzsimilarity(intuitively,thisisapurelyqualitativerelationofrepresented-similarity).Lewis(1986,238)andSider(unpublished)discussthisfeatureoftheersatzistontology.Presumably,asIassumedinearliersections,sinceersatzsimilarityismodeledonsimilarity,itisasontologicallycheapasrealsimilarity.(Ithastobe,ifersatzersaretoavoidQuineanworriesaboutselectivity.)Buthowisersatzistsimilaritytobefleshedout?Ersatzistshavenotyetdevelopedanadequateresponsetothisworry.(Siderunpublisheddiscussestheproblemandsomeoptionsfortheersatzist.)Inanycase,howeverersatzismaddressestheproblem,mypointstillholds:theersatzistdeepessentialistcanrespondtotheskepticalobjectionaswellastheersatzistshallowessentialist,andtherealistdeepessentialistcanrespondtotheskepticalobjectionaswellastherealistshallowessentialist.5.ModalCompositionHavingmadeitthisfar,youmaybecontentwithmyanswerstoparts(ii)and(iii)oftheskepticalobjection,butfeelthatIhavebeenglossingmyreplytopart(i).Youarecorrect.Ihavenotyetsaidenoughaboutwhymyaccountshouldbethoughttoavoidadhocprimitivism.Theplacetolocatetheworryaboutadhocprimitivismisinresponsetomyassertionthatthereisnoneedtoprivilegecertaindererepresentationalproperties InDefenseofEssentialismaseligibletobeincludedinanobject.Theobjectioncanbeformulatedinthefollowingway:whatdetermineswhetheranobjectincludesthespecificdererepresentationalpropertiesthatitdoes?Ifthereplyisthatitisaprimitivematterofeligibility,itmightseemasthoughwearebackwherewestarted,oratleastnofartheralongthanLewissoriginalcounterparttheory.(Thisisntquiteright„wewouldntbebackwherewestarted„Illcomebacktothispointlater.)However,theanswerdoesnotrequireanyspecialeligibility.Thedeepessentialistcanreplywithoutresortingtoprimitivism,butonlybecauseshehasdevelopedherviewwithinthecontextofacleartheoryofobjects,andinparticular,becauseshehasdevelopedherviewintermsofmereologicalbundletheory.Thereplyisstraightforward:thedeepessentialistcanholdthatforanyclassofacoreandsomeofthedererepresentationalpropertiesitgen-erates,thereisasumofthemembersofthatclass.Inotherwords,ifthesortofcompositionbetweenacoreandthedererepresentationalpropertiesitgeneratesisunrestricted,thenthereisnoadhocprimitiverestrictiononwhichobjectsexist.Theyalldo.Wehaveobjectswithdifferentmodalprofilesbecausewehavedifferentobjectsthatareproperpartsofthesumofallthedererepresentationalpropertiesgenerates.(Eachobjectincludesbutmayincludedifferentdererepresentationalproperties.)Ifthereisnoprimitivestipulationrestrictingwhichdererepresentationalpropertiesarepartsofanobject,deepessentialismdoesnotfallvictimtopart(i)oftheskepticalThisistheshortanswertotheobjection.Butalongeranswerwillhelpustoseehowthedeepessentialistcanacceptthatcompositionbetweenacoreandthedererepresentationalpropertiesitgeneratesisunrestrictedwhilepreservingthenaturalideathatordinaryobjectssuchaspersonsandcatshavesomepropertiesessentiallyandotheraccidentally.Thelongeranswerdevelopstheworryaboutadhocprimitivismintermsofwhetherasumthatistheobjectthatwearetalkingorthinkingabouthastherelevantdererepresentationalpropertiesasparts.This,inturn,canbedividedintotwoquestions:(1)aquestionaboutwhichsumwearereferringtowhenwenameanobjectand(2)aquestionaboutwhichsumsexist.Inanswerto(1),Itakeitthatitispartlyacontextualmatterandpartlyamatterofwhichpropertieswedefinitionallyassociatewithnames(includingdescriptivenames)ofobjects.(Seemy(2004)fordiscussion.)Inanswerto(2),wemustdeterminewhether(andifso,howmuch)mereologicalcompositionwiththedererepresentationalpropertiesitgeneratesisrestricted.Thedeepessentialistcanreplythatthissortofcompositionisunrestricted,andthusthatthereisnoadhocselectionofcertaindererepresentationalpropertiestobesummedwiththecore.Showingwhytheanswerto(2)isplausiblerequiresCallthesortofcompositiondescribedby(2)composition:itisaspeciesofqualitativecomposition.Modalcompositionoccurswhenanobjectscore„asumofmaterialproperties„isfusedwith(modal)dererepresentational L.A.Paulpropertiesithasgenerated.Adiscussionofrestrictionsonmodalcompositionputsussquarelyinthefamiliarterritorycoveredbyextantdiscussionsofwhetherthemereologicalrelationofcompositionisrestrictedorunrestricted.FollowingvanInwagen(1990),theGeneralCompositionQuestionaskswhatmereologicalcomposition,andtheSpecialCompositionQuestionasksaboutthecircumstancesinwhichcompositionoccurs.(SeevanInwagen(1990)andMarkosian(1998)fordiscussion.)WhatIllcalltheModalGeneralCompositionQuestionaskswhatmodalcomposition,andwhatIllcalltheModalSpecialCompositionQuestionasksaboutthecircumstancesinwhichcompositionofacorewithsomedererepresentationalpropertiesitgeneratesoccurs.Mostdefendersofmereologytakecompositiontobeaprimitive,unanalyz-ablerelation,andIamnoexception:justaswithspatiotemporalcomposition,qualitativecompositionisaprimitive,unanalyzablerelation.ThisdodgesboththeGeneralCompositionQuestionandtheModalGeneralCompositionQuestion.ButthereisdebateoverSpecialComposition,i.e.,whethercompositionisunrestrictedorwhetheritisrestricted,i.e.,whetheranyclassoftherelevantsortofentitieshasasumorwhetheronlysomeclassesaresummed.(Iamassumingthatcompositionoccurs,atleastsometimes.)Inthecaseofclassesofthingsthatcouldbespatiotemporalparts,commonsenseintuitionsupportsmoderatecomposition,i.e.,theviewthatnotjustanyclassissummed.However,therearemanycasesofpossiblespatiotemporalcompositionwhereourintuitionsaresilentorconfused.Asthereisnoclearcaseforrestrictingspatiotemporalcompositioninaparticularway,manyavoidtheproblemofdeterminingwhensummingoccursbyholdingthatspatiotemporalcompositionisunrestricted,thatis,anyclassofspatiotemporalobjectsissummed.Othersholdthatspatiotemporalcompositionmustberestricted(thereisnoobjectthatisthesumofmymicrowaveovenwiththetopofMt.Everest)butadmitthattheylackpreciseorcomprehensiveguidelinesfortheconditionsunderwhichspatiotemporalcompositiondoesordoesnotoccur.Theintuitivecaseforwhetherthereisrestrictedcompositionofpropertiesisdifferentfromthespatiotemporalcase.Althoughwemightnotbeabletocomeupwithnecessaryandsufficientconditionsforwhenqualitativecompositionoccurs,itisclearthatitmustberestricted,sinceabsolutelyunrestrictedqualitativecompositionwouldimplythatobjectswithincompatiblepropertiesandobjectssuchaswingedpigsorgoldenmountainswereactual.Averyplausibleviewisthatqualitativecompositionisrestrictedinsuchawaysoastoensurethattherearenosumsofincompatiblepropertiesorhornedhorsesandthelike.Notethatinansweringtheskepticalobjection,thedeepessentialistisonlytryingtodeterminewhethercompositionisunrestricted,i.e.,whetherqualitativecompositionisunrestrictedalongthedimensionofwhichdererepresentationalpropertiesaresummedwiththecoresthatgeneratedthem.(Iamignoringcaseswherethevaguenessofqualitativecompositionpiggybacksonthevaguenessofspatiotemporalcomposition.) InDefenseofEssentialismIfmodalcompositionisentirelyunrestricted,everypossiblesumofacoredererepresentationalpropertiesitgeneratesexists.Friendsofunrestrictedspatiotemporalcompositionwillembracethisviewsincetheythinkthatinprinciplecompositionmustalwaysbeasunlimitedaspossible,sotheywillarguethatthereisnoontologicaljustificationforlimitingmodalcomposition.Thus,qualitativecompositionisonlyminimallyrestricted,andtherearemanymoreobjectsthancommonsensewouldendorse,sincetherearemany,manysums,andtherearealsomany,manyproperpartsofthesesumsthatarealsoobjects.Theviewthatthereexistlotsofobjectsthathavealloftheircorepropertiesaccidentallyisnoexceptiontothegeneralrulethatanysortofminimallyrestrictedmereologicalcompositiongeneratesmanyobjects:ifwehavesuchobjects,wealsohavelotsofproperpartsofthoseobjectsthataresimilartothembutdonothavealltheircorepropertiesaccidentally.Thefriendofunrestrictedmodalcompositiondefendsthisimplicationbyarguingthatmostoftheseobjectsareregularlyignoredandthatourquantifiersdescribingwhatthereisareimplicitlyrestricted.Theobjectsthatdonothavealltheircorepropertiesaccidentallywillbesumsofthecoreoftheobjectwithsomebutnotallofthedifferentpropertiesofbeingdererepresentedindifferentwaysindifferentpossibleworlds.Suchsumsareproperpartsofthesumthatconsistsofthecoreoftheobjectandallofthedifferentpropertiesofbeingdererepresentedindifferentwaysindifferentpossibleworlds.Istheviewofthedeepessentialistwhoembracesunrestrictedmodalcom-positionbizarre?Itimpliesthattherearemanyobjectswiththeactualcoreparts(properties),say,ofbeingblondehavingoriginO,wheretheseobjectsdifferwithrespecttowhethertheyareaccidentallyblondeoressentiallyblonde,whethertheyaccidentallyhaveoressentiallyhave,andsoon,foreachcoreproperty.Tobesure,inallormostnonphilosophicalcontextswewillrefusetoassenttotheexistenceofsuchobjects,butaccordingtothedeepessentialist,theynonethelessexist.(Theshallowessentialistwhotriestomakethispointintoanobjectionisonespeciallyweakground,sincehealreadygrantsthatthereareasmanymodalprofilesastherearecontexts.)Mydeepessentialistviewisonlybizarretotheextentthatmereologyandthefactthatweareoftenambiguousorindeterminateinourreferenceisbizarre,foritisthemereologicalontologythatcreatesnewobjectsforeverypartofanobject,andtheambiguityorindeterminacyofreferringthatallowsustoforgetthattheworldcontainsmanymoreobjectsthanwecommonsensicallywanttogrant.Bybetterunderstandingmereologyandthewayswerefertoobjects,wecanseehowtounderstandanddiscounttheexistenceoftheseunwantedobjects.Thisproblemofunwantedobjectsisanexplicitlyrecognizedandacceptedconsequenceofanyapproachtomereologythattakesproperpartsofobjectstobeobjectsintheirownrightwhileallowingformorethanaminimalamountofcomposition.Ithasbeenwell-discussedwithrespecttotheconsequencesof L.A.Paulpopularinterpretationsofspatiotemporalmereology,whereithasbeendubbedTheProblemoftheMany.Ž(Unger1980,Lewis1999)Consideraroughlycat-shapedcloudoverParisthatoccupiesavaguelyspecifiedspatiotemporalregion.CallthecloudClaude.ŽForanybound-arywecandrawthatdefinesClaudesedges,wecanshifttheboundaryatinybit(say,byawatermoleculeortwo)andhaveanewedgethatdefineswhereClaudeendsandtherestoftheworldbegins.EachtimeweshifttheboundaryofwhatwetaketobeClaude,wepickoutaslightlydifferentsumofwatermolecules(weaddorsubtractamoleculeortwo).AsLewisandothershaveargued,thismeanstherearemanysumsofwatermoleculesthatareequallygoodcandidatestobecalledClaude.Ž(Infact,sincewecanshifttheboundaryinfinitesimallysmallamountstoincludeorexcludesmallerandsmallerspatiotemporalpartsofwatermolecules,therearemoresuchsumsthantherearerealnumbers.Anyonewhoacceptsclassicalmereologyisthuscommittedtotheexistenceoffar,farmoreobjectsthancommonsensegrants.)Eachcandidatesumoccupiesaslightlydifferentspatiotemporalregionthatpartiallyoverlapsalmostalloftheothercandidateregions,andindependentlyofsomewaytohighlightit,nocandidatestandsoutasthesumthatespeciallydeservestobecalledClaude.ŽItisimportanttonotethatalthoughwemayhavemanydifferentobjects,thedifferencesarepartial,i.e.,thereismuchoverlap.Onewidely(butnotuniversally)acceptedtreatmentofthisconsequenceofmereologyistograntthatthereexistmanypartlydifferentbutpartlyidenticalobjectsinaregion,buttorecognizethatforthemostpartsuchobjectsareignoredorreceiveattentiononlywhenwemustdecidewhichobject(orobjects)arepickedout(forexample,bythatcat-shapedcloudoverParisŽorClaudeŽ).Moreover,whenwechoosebetweencandidatesums,sincewearechoosingbetweenobjectsthatoverlapwithrespecttonearlyalloftheirparts,strictlyspeaking,therearemanyobjects,butasLewisputsit,therearealmostone.Ž(Lewis1999)Wecanthinkoftheideainformallylikethis:imagineaperson,Sandy,sittinginfrontofyouinachair.Mereologytellsusthattherearemanyspatiotemporalsumsintheregioninfrontofyou,eachsumminimallydifferentfromtheothers,andeachsumincludingenoughmaterialtoconstituteaperson.Itmakesperfectsensetograntthattherearethesedifferentsums,sincetheydifferinwhattheyinclude,eveniftheyarenotentirelydifferentduetooverlappedparts.Butjustbecausetherearemanyminimallydifferentsumsdoesnotmeanthatwecansaytherearemanydifferentpersonsinfrontofyou.Weneedmorethanaminimaldifferenceinpartstosaythatwehavedifferentpersons,evenifaminimaldifferenceinpartsisallittakestosaythatwehavedifferent.Somereologydoesnotentailthatwecansaytherearemanypersonsinthechair.Instead,mereologyentailsthattherearemanycandidatesforthereferentofthepersoninfrontofyou,ŽorSandy,Žwhilethewayweuselanguageentailsthatweeitherreferindeterminatelytomanyofthosecandidateswhenweusethenameand(or)usecontexttohelpustonarrowdownthefield. InDefenseofEssentialismThesituationisexactlythesamewithunrestrictedmodalcomposition,exceptthatthedifferentcandidatesumsoccupypreciselythesamespatiotemporalregioninsteadofroughlythesamespatiotemporalregion(thedifferentcandidatesumsoccupyroughlydifferentregionsofmodalspaceŽinstead).Presumably,inthechairinfrontofme,therearemanyalmost-identicalsumsofparticlesthatarecandidatesforthenameSandy.ŽLikewise,therearemanyalmost-identicalsumsofcorepropertieswithdererepresentationalpropertiesthatarecandidatesforthenameSandy.ŽButthisdoesnotmeanthatinanyacceptablecontextwewillgrantthattherearemany,ormanySandys,intheregionthatthesesumsoccupy.Thereisonlyoneperson,buttherearemanycandidates.Theskepticwhorejectsmodalcompositionsimplybecauseitimpliestheexistenceofmanyalmost-identicalsumsmustrejectordinarycasesofspatiotemporalcompositionaswell,sincethereasoningisparallel.Fewarewillingtodenytheexistenceofpersonsinordertosolvethespatiotemporalversionoftheproblemofthemany.Forthesamereason,fewshoulddenytheexistenceofpersonswithessentialandaccidentalproperties.Ifmodalcompositionisunrestricted,thereisnoprimitiveprivilegingofcertainpropertiesorcertainsums,andthedeepessentialistcanrebutpart(i)oftheskepticalobjection.Therearemanyobjects,itistrue,butthedeepessentialistcanexplainthisintheusualpragmaticorsemanticway,usingreasoningthatisacceptedbymostQuineansinregardtotheparallelcaseoftheProblemoftheMany.SinceQuineanstendtobeextremelysanguineabouttheproliferationofobjectsgeneratedbymereology,theyhavelittleornogroundsforrejectingtheproliferationofobjectsendorsedbythedeepessentialist.(Theyeventendtoaccepttheviewthatspatiotemporalcompositionisentirelyunlimited,thusallowingformanymoreobjectsthanjustthosewegetfromtheProblemoftheMany,sinceforsuchQuineansthereanobjectthatisthesumofmymicrowaveandthetopofMt.Everest,andmanymorebizarrethings.)Notealsothattheproliferationofobjectsandmodalprofilesgoesalongwaytoexplainingthevaguenessofourmodalintuitions.(Idiscussthisindetailinmy(2004).)Sothedeepessentialistcanthoroughlyanswertheskepticalobjection.Now,mydefenseoftheProblemoftheManynotwithstanding,IadmitthatacceptingtheManycomesatacertaincosttocommonsense,bothforQuineansandfordeepessentialists.Quineansshouldntmind,partlybecausetheytendtograntunrestrictedcompositioninothercontexts,butalsobecausetheyshouldnotobjecttothepossibilityofhavingmanymodalprofiles„sincetheQuineangrantsthatthereareasmanymodalprofilesastherearecontextsofdescriptionorevaluation.TherereallyisntmuchleftherefortheQuineantoobjectto.Butsomedeepessentialistsmightfeelaggrievedattheideathatthereexistsomanyobjectswithsomanydifferentmodalprofiles,evenifweusuallyignoremostofthem.Itwouldbeniceifsuchadeepessentialistcouldrebuttheskepticalobjectionwhileholdingthatmodalcompositionisrestrictedinamoderatewaysuchthatthereareonlyasmanyobjectsandmodalprofilesasshewantsto L.A.Paulsanction(allowing,ofcourse,formanyborderlinecaseswherethereisindecision).Suchafanofmoderatemodalcompositioncouldallowenoughcompositiontoexplaintheseemingvariabilityofourmodalintuitions,orcoulddefendasparseviewaccordingtowhichthereisjustenoughmodalcompositiontoallowforafewobjectstohavesomepropertiesaccidentally.Therightchoicewoulddependontheaccountgivenofthecircumstancesunderwhichmodalcompositionoccurs,andwouldrequireacarefulbalancingofconsiderationsofontologicaleconomyandtheexplanationofmodalintuitions.Itwouldbenicetoendorsesuchaview,butImnotconvincedthatoneshouldbewillingtodoit.Ithinktheprospectsforamoremoderatesortofdeepessentialismareunclear,andIdonotthinkoneshouldrejecttheversionofdeepessentialismIdefendedabovesimplybecauseofthemodalcompositionalversionoftheProblemoftheMany.However,becauseamoderateversionofdeepessentialismisattractive,itisworthexploringitinmoredetail.Somemaythinkitworththeontologicalexpense.Ifmodalcompositionismoderate,thennoteverycombinationofasumofcorepropertieswiththedererepresented-bypropertiesitgeneratesexists.IfSandyisblonde,andthereisnosumofhercorepropertieswithadererepresentationalpropertyofhavingadifferenthaircolor,thensheisessentiallyblonde.Butifthereissuchasum,thensheisaccidentallyblonde.Havingfewersumsreducesthenumberofobjectsintheworld,andreducesthedistributionofessentialandaccidentalpropertiestheyhave.Thisisaniceresult,andonethatmany,especiallythosewhothinkthatspatiotemporalcompositionshouldberestricted,willfindappealing.Theissueiswhethermoderatemodalcompositionviolatestheskepticsprohibitionagainstarbitraryoradhocprimitivism.Beforeweevaluatethisissuefurther,itwillbeusefultocomparethecaseofrestrictedspatiotemporalcomposition,sincetheissuesatstakeinadebateovermoderatemodalcompositionarefamiliarfromdiscussionsofmoderatespa-tiotemporalcomposition.Ifprimitivelyrestrictingspatiotemporalcompositionisadhocbytheskepticslights,thenprimitivelyrestrictingmodalcompositionwillbeaswell.However,whilesomeskepticswillthinkthatmoderatespatiotemporalcompositionisadhoc,othersmaynotagree.Friendsofunrestrictedspatiotemporalcompositionarguethatsincewelackanacceptableaccountabouthowspatiotemporalcompositioncouldberestricted,thereisnowaytosatisfycommonsenseintuitionsaboutjusthoworwheretolimitspatiotemporalcomposition.Thetroubleisthattherearegoodintuitivereasonsforthinkingthatspatiotemporalcompositionismoderate,butthereisnogen-erallyacceptedanswertotheSpecialCompositionQuestionforspatiotemporalFriendsofunrestrictedspatiotemporalcompositionconcludethattheonlymetaphysicallyrespectablemoveistoacceptthattherearenolimitations.(Forexample,Lewis1986,211…213.)Inreply,friendsofmoderatespatiotemporalcompositioncanpointoutthatleavingspatiotemporalcompositionunrestrictedviolatesatleastasmanyintuitionsasrestrictingspatiotemporalcomposition.Somearguethatwecan InDefenseofEssentialismsimplyholdthatwhencompositionoccursisbruteorunanalyzable(Markosian1998),andsomeargueforaparticularansweratthecostofviolatingmanyofourcommonsenseviewsaboutwhatobjectsexist(e.g.,vanInwagen1990holdsthatsomescomposeawhentheactivityofthesconstitutealife).mostsimplytaketheissueofwhenspatiotemporalcompositionoccurstobeanunresolvedquestion:itisrestrictedsomehow,butjusthowisyettobesettled.(Isuspectthatifnoacceptableanswerisforthcoming,Markosiansbrutalapproachmaywintheday.)Sothesituationofthefriendofmoderatespatiotemporalcompositionisthis:shemaytakeittobeprimitiveandunanalyzablewhenspatiotemporalcom-positionoccurs,orshemayacceptsomesortofextremelycontroversialaccountsuchasvanInwagens.Thefriendofunrestrictedspatiotemporalcompositionmayrejectthemoderatesprimitivistmoveasarbitraryandothermoves(suchasvanInwagens)asimplausible,buttheimplausibilityofhisownviewmakeshispositionjustasuncomfortable.HowdoesmoderatespatiotemporalcompositionlookfromtheperspectiveofthecontemporaryQuineandescribedin1?Rathergood,allthingsconsidered.Hereisthecrucialdetail:giventhestateofthedebate,unlesssomeonediscoversagenerallyacceptablereductiveaccountofhowspatiotemporalcompositionisrestricted,takingittobeaprimitivematteraboutwhenmoderatespatiotemporalcompositionoccursshouldnotbeobjectionableinprincipletosuchaQuinean.(OurQuineanmightfindotheranswerstothespecialspatiotemporalcompositionquestionmoreplausible,butthisisdifferentfromoutrightrejectionoftheprimitivismofthemoderateview.)WecanseethisoncewerecallthatprimitivismshouldnotberejectedasadhocbythecontemporaryQuinean,sinceherecognizesthatweareallprimitivistsaboutsomethingorother„forexample,manyotherwiseontologicallyconservativephilosophersacceptprimitivenat-uralness,primitivedistinctionsbetweencertainclasses,orprimitiveeligibilityforreferents.WhattheQuinean(whoembracestheskepticalobjection)rejectsunmotivatedprimitivism,i.e.,primitivismdefendedinspiteofasuccessfulreductiontosomethingmorefundamental.Butthereisnogenerallysuccessfulreductionoftheconditionsunderwhichspatiotemporalcompositionoccurs(alloftheanswers,includingthealwaysŽanswer,areintuitivelyunacceptable).ThereisnonaturalexplanationŽorevenmoderatelyplausiblereductivestoryofwhenspatiotemporalcompositionoccurs.Thismeansthatgiventhecurrentstateofthedebate,theprimitivismofmoderatespatiotemporalcompositionisnotevenbythecontemporaryQuineanslights.Whatworksforthefriendofmoderatespatiotemporalcompositionworksforthefriendofmoderatemodalcomposition.Ifitisnotadhoctosatisfyourcommonsenseintuitionsaboutspatiotemporalcompositionbyprimitivelyrestrictingit(sincethereisnoreductionoralternativenaturalexplanationonthecards),itisnotadhoctosatisfyourcommonsenseintuitionsaboutmodalcompositionbyprimitivelyrestrictingit,sincethereisnoreductionoralternativenaturalexplanationonthecards.(Iamassumingthatfromtheperspective L.A.Paulofthemoderatethatunrestrictedcomposition,eitherspatiotemporalormodal,doesnotprovideanaturalexplanation.Ofcourse,ifunrestrictedcompositionisacceptedthentheskepticalobjectionhasalreadybeenrebutted.)ThissuggeststhatthedeepessentialistwhodefendsmoderatecompositionisinarelativelygoodpositionvisavisthecontemporaryQuinean.Notasgood,perhaps,asthedeepessentialistwhoendorsesunrestrictedmodalcomposition,butsignificantlybetterthanthesubstanceessentialistortheoriginalcounterparttheorist,bothofwhomrelyonunmotivatedprimitivism.SowhileIdonotofficiallyendorsemoderatecomposition,Ithinkithasenoughgoingforitthatitdeservesfurtherconsideration.Inanycase,whetheroneprefersunrestrictedmodalcompositionwithadoseofpragmatismorrestrictedmodalcompositionwithaleaner,meanerontology,theskepticalobjectionhasbeendealtwith.Deepessentialismisbackinblack.1.Ithanktheparticipantsofthe2006ArizonaOntologyConferenceandJohnDiversforhelpfuldiscussion.IamindebtedtoAgustšnRayoforalongandinterestingdiscussionaboutshallowessentialism,andamespeciallyindebtedtoTedSiderforcommentsonseveraldrafts.2.ThereisadistinctionbetweenanobjectstriviallyessentialpropertiesanditsrealŽessentialproperties:anobjectstriviallyessentialpropertiesarepropertiesthathavenothingtodowiththenatureofthatparticularobject,butneverthelessarehadbyitineverypossibleworld.Examplesincludethepropertyofbeingself-orthepropertyofbeingeitherfemaleornotfemale.Sincetrivialessentialpropertieshavelittleornothingtodowithwhatanobjectis(inthemetaphysicallyinterestingsenseofdetermininganobjectsnature),theyarenottrulyessentialproperties.Theymightbebettercalledtriviallynecessarypropertiesandtakentobesomehowprimitivelydifferentfromessentialproperties.(HereIamimpressedbyKitFines(1994)pointaboutthedifferencebetweenessentialpropertiesandothersortsofnecessaryproperties.)3.Afterthispoint,IlldropthenontrivialŽqualificationforeaseofexposition.4.Iargueinmy(2004)thatessentialistsshoulddefendacertainsortofsemanticindeterminacyconsistentwiththisviewinordertoexplainthevaguenessofsomeofourmodalclaims.5.DavidWigginsisthebestknownproponentoftheviewthatordinaryobjectsarethingsthatfallundersortalconcepts.Wiggins(inhis2001andelsewhere)combinesanAristotelianapproachtowardssubstancewiththeKantianideathatourabilitytotrackthepersistenceofobjectsthroughchangerequiressubstanceinordertomakesenseofourabilitytoindividuateobjectsandnavigatetheworld.AlthoughWigginstellsusthatweknowobjectshavecertainmodalpropertieswhenwerecognizethattheyfallundercertainsortalconcepts,thisisnotanexplanationofhowobjectshavetheirderemodalproperties.ItiscertainlynotsufficienttoaddresstheobjectionsIdiscussbelow.6.Aristotle,Categories,sec.Ilines4…5,&28.(Trans.,Ackrill,1987.) InDefenseofEssentialism7.HoffmanandRosenkrantz(1997)developacontemporarynotionofsubstanceasindependent.8.Thereisadevelopingliteratureinvisualpsychologythatsuggeststhatthewayweusesortalconceptsismorecomplexthanthisbreezygeneralizationsuggests.E.g.,seeXu(1997).IhopetodiscussthisinmoredetailinPaul(unpublished).9.Ofcourse,Quinehimselfwasopposedtoallsortsofmodality,andwouldhaveobjectedtoanymetaphysicalprojectthatpurportedtoexplainmodalityintermsofpossibleworlds.ThecontemporaryQuineanwillusuallyacceptsomeprimitivededictomodalityandwillengageinprojectsinvolvingmodalmetaphysics.10.Thus,anactualistcouldinsteadtaketherelationbetweenanobjectandpossibleindividualsthatdeterminestheobjectsmodalpropertiestobebasedoncross-worldidentityrelationsinsteadofcounterpartrelations.Thecrossworldidentityrelationswouldbeprimitiverelationssomehowbasedonthesortsofcrossworldidentityrelationsthatwouldexistbetweenactualobjectsandotherworldlyobjects,hadthoseotherworldlyobjectsexisted.Crossworldidentityfortheactualististhus:itissomeother(ontologicallybasic)relation.11.ImindebtedtoSusannaSiegelforpressingmetoaddressthisobjection.12.Irecapturesomeoftheelementsofthisintuitionin3bydeveloping,aspartofmyreductiveaccountofhowobjectshavederemodalproperties,theideathatderemodalpropertiesmustbeinternaltoanobject.Forexample,Iholdthatwhataperson,issomethingthatisdererepresentedinacertainway,andthusapersonincludesoutwardpointingŽdererepresentationalproperties.13.NotethatembracingtheLeibnizianframeworkisindependentofwhetherornotafullreductionofderemodalityisaccomplished:possibleworldsorrelationstopossibleindividualsmaystillbetakentoinvolveprimitivemodalnotions.Sider(2003,184)emphasizesthis.14.AnotherwaytorespondtotheObjectionistoarguethatthebestreductiveexplanationofhowobjectshavetheirderemodalpropertiesoccursintwosteps:(1)weconvertderemodalclaimstodedictomodalclaims,then(2)reducemodalclaimstoclaimsaboutpossibilia.Onthiswayofseeingthereduction,theObjectionmakeslittlesense,sincethereisnodirectreductionofthederepropertiesofanobjecttorelationstopossibilia.(IamindebtedtoDavidChalmersforsuggestingthisresponse.)15.Whymustderefactsaboutparticularindividualsmustbereducedawaywhilefactsaboutparticularproperties(thoughtofasclassesofobjects,tropes,universalsorwhat-have-you)arekept?Whileanyreductionisattractiveforreasonsofparsimonyandtheargumentsdescribedinthetextseemtosupporttheideathatitisbettertohaveapurelyqualitativecharacterizationofmodality,moreworkneedstobedonetosortoutthemotivationfortheintuitivedistinctionsthatarebeingdefendedhere.(IamindebtedtoJohnHawthorneforbringingthispointtomyattention.)16.ImindebtedtoMichaelReaforsuggestingIconsiderthisresponse.17.Wiggins(1967,1980,2001)isonewell-knownproponentofthisview.18.Forexample,asWiggins(1997,2001)holds.19.Bennett(2004),Sosa(1987)andZimmerman(1995),amongothers,havepressedversionsofthisworry.20.Iamassumingthatwemustaccountforcaseswherethepersonandthehunkoftissueexistforexactlythesameamountoftime. L.A.Paul21.Idevelopthesepointsindetailinmy(2006,forthcoming).22.InthispaperIfocusonaddressingtheskepticalobjection.Iaddressthepuzzlesoftheambiguityofouressentialistintuitions,materialconstitutionandthecross-worldparadoxesinPaul(2004),Paul(2006forthcoming),andPaul(unpublished),respectively.23.AnantiessentialistshallowessentialistcouldinsistonspeakingwiththevulgarŽatalltimes,andsodenythatwecaneversay,eveninthestrangestofcontexts,that,e.g.,anobjectispossiblyafriedegg.SinceIthinkwecanperfectlywellestablishtherelevantcontexts(e.g.,inthemetaphysicsseminarroom),Ifindthismaneuverunconvincing.24.Iwillassumegenuinetemporalcounterparttheoryhere.25.Fromthispointon,Iwillnotdistinguishbetweenevaluativeandantiessentialistinterpretationsofshallowviewsunlessnecessary.26.Theshallowpersistencetheoristcanunifyandsolveanumberofdifferentonto-logicalpuzzlesinvolvingcoincidence,theproblemofthemany,personalidentityfissioncases,andmore.Siderarguesthat[w]henweconsiderthesecases,ouroverwhelmingfeelingissurelythattheanswerstothequestionsdependonhowweconceptualizetheobjectsinvolved,thattheworlddoesnotforcesingleanswersuponus.Ž(2001,207)Iagreethatthereissemanticindeterminacyintheoffing.Butwhatsortofindeterminacyiscontroversial,sincethedeep(temporal)persistencetheoristcangiveherownaccountofthesemanticindeterminacyinvolved.Inparticular,IthinkmanyofthecasesthatSider(2001,ch.5)discusses,suchasthecaseoftheShipofTheseus,thecaseofhowtwodifferentestablishmentseachlaidclaimtobeingtheoriginalBookbindersrestaurant,oracasewhereapileoftrashundergoespartialreplacementofrefuse,canbebetterhandledbypostulatingsemanticindeterminacywithrespecttowhicharereferredtoratherthansemanticindeterminacywithrespecttohowthesameobjectistemporallydererepresented.Thisgivesusthedesiredconstanttemporaldererepresentation:thedifferentobjectsreferredtomightbedifferentworms,orslightlydifferentstages(stagesthatdifferonlywithrespecttosometemporalrepresentationalproperties).IfIamright,thepuzzlementengenderedbySidersparadoxicalcasesdoesnotlicensetheinferencetoanysortofshallowpersistence.Thesamepointholdswithregardtopuzzlesaboutthevariabilityofourmodalintuitions.(Inmy(2004)Iarguethattheinconstancyofourmodalintuitionscanbeexplainedbysemanticindeterminacywithregardtotheobjectswepick27.Forsimplicity,Iveignoredthepossibilitythatobjectsmighthavesomepropertiesthatarecontext-dependent,wheresuchpropertiesarepropertiesthatdefinewhatanobjectis.Ifobjectshavesuchproperties,theyarenottherelevantmodalpropertiesthatareunderdiscussionbetweenthedeepessentialistandtheshallowessentialist.ThanksareduetoKarenBennettandTedSiderfornotingthispoint.28.Tosomeextent,wehaveempiricalaccesstothesepersistenceconditions,asWiggins(2001)andAyers(2005)argue.Thisisnottodenythatsignificantepistemicindeterminacyabouttheprecisedetailsofobjectspersistenceconditionsremains.29.Fromnowon,unlessIsayotherwise,IshallassumethatdererepresentationŽreferstomodaldererepresentation.30.Otherdiscussionsofersatzismcanbefoundin,e.g.,Plantinga(1974),Lewis(1986),Heller(1990),Divers(2002),Sider(2002)andSider(2003). InDefenseofEssentialism31.Thus,onewell-knownersatzistviewIrejectisAlvinPlantingas.Plantingas(1974)accountofersatzpossibiliatakesworldstobemaximalpossiblestatesofaffairs,onlyoneofwhich(theactualworld)obtains.ForPlantinga,tosaythatSocratesexistsinaworldistosaythatifhadbeenactual,Socrateswouldhaveexisted,andSocratesisessentiallyhumaniffSocratesishumanineveryworldinwhichheexists.32.WhileIhavenotexplicitlyendorsednon-ersatzistwaysofbeingactualist,ifsuchwaysaresufficientlyreductiveandcanallowforobjectivederepredication,suitablemodificationsofmyargumentshouldbeavailable.33.Idevelopanearlierversionofthistheoryofobjectsandtheirpartsinmy(2002),callingqualitativepartslogicalparts.ŽThereIindicatethatIprefertoholdthatcompositionofpropertypartsisrestricted,butdeveloptheaccountintermsofunrestrictedpropertycompositioninordertokeeptheviewassimpleandbroadaspossible.Iveborrowedheavilyfrommy(2006forthcom-ing)forthenextfewparagraphsonmymereologicalbundletheory:interestedreadersmaywishtoconsultthatpaperforamoredetailedtreatmentofmyview.34.Itisunclearwhichontologyofobjectsisbestsupportedbycontemporaryphysics,inpartbecausetheontologyofthereigningview,quantumfieldtheory(QFT),issopoorlyunderstood.Myviewiswellsupportedbyclassicalfieldtheory,oranytheorythatallowssomeofthebuildingblockstobeproperties(asQFTseemstoallow).35.Moreover,nothingIvesaid(yet)impliesthatqualitativepartsmustbequalitativeproperties,i.e.,non-haecceitisticproperties.WhileIwillultimatelyrejectsuchproperties,nothingaboutmytreatmentofobjectsassumsofqualitativepartsrequiressucharejection.36.Forsimplicity,Ivedescribedsamplepropertiesinless-than-fundamentalterms.Iamassumingthat,e.g.,ifSandyhasaninstanceofthepropertyofblondethenthisinstancereducestosomemorecomplexconjunctionofsparsefundamentalproperties.37.Onmyview,SandyincludesthemonadicrelationalpropertyofhavingoriginOmuchthesamewaythatsheincludesthemonadicdererepresentationalpropertiesIdescribebelow.38.Iamassumingthatisanonmodalpropertyandisnotadererepresentationalproperty.39.Whatifapropertyisasingle-casedeterminate?Thatis,isadeterminatepropertyofthedeterminableandtherearenootherwaystoinstantiateexceptbyinstantiating.Itisnotclearthatsuchapropertyexists:determinablesareusuallyinstantiatedbymanydifferentdeterminates(e.g.,rednessbluenessgreennessetc.asdeterminates).Butifsuchapropertydoesexist,howcananobjectbedererepresentedas?Mustalwaysbehadessentially?„Notnecessarily.Tobesure,acommonwayforanobjecttobedererepresentedistobedererepresentedas,wherehavingthedeterminateincompatiblewithhaving.Butanotherwayfortobedererepresentedasisfortobedererepresentedaslacking,wherethereisnoreplacementpropertyfor.ImindebtedtoJohnHawthorneforraisingthisquestion.40.Imgoingtoignorethefactthatanobjectssummayincludemorethanjustthecoreanddererepresentationalpropertiesinordertoavoidintroducingunnecessarycomplexity. L.A.Paul41.Inotherwords,Istipulatethatanobjectisdererepresentedasinvirtueofthatobjectsincludingthedererepresentationalpropertiesgeneratedbyitscorestandinginadererepresented-byrelationtoanersatzindividualthathasacounterpartthatis.(Amodalrealistcouldstipulatethatanobjectisrepresentedasinvirtueofthatobjectsincludingdererepresentationalpropertiesgeneratedbyitscorestandinginadererepresented-byrelationtoacounterpartthatis42.Analternative(actualist)viewIfindattractivethatcouldalsobeusedtosolvetheproblemsforessentialistsisadverbialismaboutparthood,i.e.,takeparthoodtobespatiotemporallyandmodallyindexed.Onthisview,objectshavetheirderemodalpropertiesbybeingrepresentedashavingpropertypartsindexedtoworlds,sotosayobjectsaresumsofpartshadessentiallyoraccidentallyistosaythatobjectscanberepresentedashavingpropertypartswithrespecttoworlds.43.Thereisastipulativeelementhere,asinmoreconventionalLewisianandersatzistpictures,aboutwhatcountsasdererepresentationofanobject.Isaythatwhatcountsasdererepresentationofanobjectisthedererepresentationofthecoreoftheobjectbypossibilia.Lewisandhisfollowersineffectholdthatdererepresentationofanobjectisthedererepresentationofthecoreoftheobjectbypossibilia,butforthem,thecorejustistheobject.Lewis(1986,195-7),respondingtoobjectionsfromKripke(1972),defendstherightofthetheoristofderetorelyonstipulativedefinitionsofdererepresentation.Therefore,Itakeitthattheskepticwhoacceptsshallowessentialismcannotobjecttothissortofstipulation.44.IamnotsurewhetherLewisshouldbedescribedasanevaluativeshallowessentialistorasanantiessentialistshallowessentialist.Arguably,Lewis(1986)endorsesevaluativeshallowessentialism,butmuchofhisearlierworkseemstoendorseantiessentialistshallowessentialism.45.PresumablyfollowingArmstrong1978,Lewisholdsthatsuchrelationalpropertiesreducetotherelationalcomplexes(therelataplustherelation).AccordingtoLewis,relationalpropertiesarestructuredproperties:propertiestakentohaveaquasi-syntacticstructurewherebytheyareconstructedfromtheirconstituents.Therelationalproperty[bearing-bent-at-to-t]hasadyadicrelationasoneofitsconstituents,andasuitablerelatumasanother.Ž(Lewis2002,p.3)46.HereIamemphasizingthatforanobjecttobeessentiallyoraccidentallymustincludetherelevantdererepresentationalproperties.TheremightbesomesortofhavingŽsuchthatanobjecthasŽapropertybybeingaproperpartofalargerfusionthatincludesorbystandinginarelationtoa,butitisnotthesortofhavingthatisneededfortheobjecttohaveaderemodalproperty.47.Thesituationseemstobemetobeparalleltocasesofcontentexternalism.48.ItsworthnotingthatIbelievetheersatzistcouldmakeequallygooduseofsomesortofpurelyqualitativeersatzcrossworldidentityrelationinsteadofersatzsimilarity,whichmeansthedeepessentialistmightbeabletodispensewithcounterparttheory.(First,crossworldidentityforanactualistisntidentity,whateveritis.Second,itisntcrossworldidentitybutersatzcrossworldidentitythatweneed,sowhycantthisrelation,whateveritis,bebasedonarelationthatispurelyqualitative?)IhopetoexplorethispointinmoredetailinPaul(unpublished).49.Torecapforthesakeofclarity:onmyview,whetheranobjectisdererepresentedisabsolute,notcontextdependent.Thereisasettledanswer,onceandforall,about InDefenseofEssentialismwhatistrueconcerningacertainobjectaccordingtoaworld.Icanendorseconstancybecauseonmyviewtheoccurrenceofdererepresentationaloneisnotenoughforanobjecttohavederemodalproperties:foranobjecttobeaccidentally,itmustincludeplusthepropertyofbeingdererepresentedas.Wehaveobjectswithdifferentmodalprofilesbecausewehavedifferentobjectsthatareproperpartsofasumofcorepropertiesplusadditionaldererepresentationalproperties.(Eachobjectincludesacorebutmayincludedifferentdererepresentationalproperties.)50.Especiallybecauseevenifmodalcompositionisrestricted,unlessitissorestrictedthatalmostnosumsexistorweendorsecontroversialviewsaboutparthood,someversionoftheProblemoftheManyarises.(Aswithspatiotemporalcomposition:allweneedisonenormal-sizedcloudtohavemanycloud-candidates.)51.AsvanInwagen(1990)shows,holdingthatrelationssuchascontact,physicalbondingorthelikearesufficientforspatiotemporalcompositionfailtorespectcommonsenseviewsaboutwhatsumsexist.52.Fewarguethattakingspatiotemporalcompositiontobeaprimitiverelationorourinabilitytosatisfycommonsenseintuitionswithrespecttolimitsonspatiotemporalcompositionjustifiestheclaimthatcompositiontakesplace.ThosefewwhodoarguethismightfeeltheneedtoadoptvariantsofnihilismlikethosedescribedinRosenandDorr(2003),vanInwagen(1990)andUnger(1979).53.Ontheotherhand,ifthespecialcompositionquestionforspatiotemporalcompositionisadequatelyansweredbydiscoveringareduction,e.g.,thatitoccursunderwhentheactivityofcertainsconstitutealifeorsomeotherrelationthatsstandin,wemaydiscoveracorrespondingacceptablereductiveaccountoftheconditionsunderwhichmodalcompositionoccurs.ReferencesAdams,Robert(1979).TheoriesofActuality,ŽinLoux,ed.,ThePossibleandtheActual:ReadingsintheMetaphysicsofModality(Ithaca:CornellUP),190…209.Aristotle(1987).Categories,inANewAristotleReader(PrincetonUniversityPress),Trans.:J.L.Ackrill.Armstrong,D.M.(1978).ATheoryofUniversals(CambridgeUniversityPress).Ayers,Michael(2005).OrdinaryObjects,OrdinaryLanguageandIdentity,ŽTheMonistBennett,Karen(2004).Spatio-TemporalCoincidenceandtheGroundingProblem,ŽicalStudies118,339…371.Carnap,Rudolf(1928).DerlogischeAufbauderWelt(Berlin:Weltkreis-Verlag).Chisholm,Roderick(1979).IdentityThroughPossibleWorlds:SomeQuestions,ŽinLoux,ed.,ThePossibleandtheActual:ReadingsintheMetaphysicsofModality(Ithaca:CornellUP),1979,80…87.Divers,John(2002).PossibleWorlds(London:Routledge).Fine,Kit(1994).EssenceandModality,ŽPhilosophicalPerpectives8,1…16.Fine,Kit(2005).ModalityandTense(Oxford:OxfordUniversityPress).Forbes,Graeme(1985).TheMetaphysicsofModality(Oxford:ClarendonPress).Goodman,Nelson(1951)[1966].TheStructureofAppearance,TheBobbs-MerrillCompany,Hawley,Katherine(2001).HowThingsPersist(Oxford:ClarendonPress). L.A.PaulHeller,Mark(1990).TheOntologyofPhysicalObjects(CambridgeUniversityPress).HoffmanandRosenkrantz(1997).Substance:itsNatureandExistence(Routledge).Kripke,Saul(1972).NamingandNecessity,(HarvardUniv.Press).Leonard,H.S.andGooman,Nelson(1940).TheCalculusofIndividualsanditsUsesŽ,JournalofSymbolicLogic5,45…55.Lewis,David(1968)[1983].CounterpartTheoryandQuantifiedModalLogic,ŽJournalofPhilosophy65,113…26,reprintedinhisPhilosophicalPapersI,26-46(1983withPostscriptŽ).Lewis,David(1986).OnthePluralityofWorlds(BasilBlackwell).Lewis,David(1999).Many,ButAlmostOne,ŽinPapersinMetaphysicsandEpistemology(CambridgeUniversityPress),164…182.Lewis,David(2002).TensingtheCopula,Ž111,1…13.Markosian,Ned(1998).BrutalComposition,ŽPhilosophicalStudies92,211…249.Paul,L.A.(2002).LogicalParts,Ž36,578…596.Paul,L.A.(2004).TheContextofEssence,ŽAustralasianJournalofPhilosophy82,170…184.Paul,L.A.(2006forthcoming).CoincidenceasOverlap,ŽPaul,L.A.(unpublished),TheOntologyofObjectsPlantinga,Alvin(1974).TheNatureofNecessity(Oxford:ClarendonPress).Quine,W.V.(1960).WordandObject(MITPress).Quine,W.V.(1963).FromaLogicalPointofView(HarvardUniv.Press).Rosen,GideonandDorr,Cian(2003).CompositionasaFiction,ŽTheBlackwellGuidetoMetaphysics,ed.RichardM.Gale(Oxford:Blackwell).Salmon,Nathan(1986).ModalParadox:PartsandCounterparts,PointsandCounterpoints,ŽMidwestStudiesinPhilosophy,75…120.Sider,Theodore(2001).Four-Dimensionalism(Oxford:ClarendonPress).Sider,Theodore(2002).TheErsatzPluriverse,ŽJournalofPhilosophy99(2002):279…315.Sider,Theodore(2003).ReductiveTheoriesofModality,ŽinLouxandZimmerman,eds.,OxfordHandbookofMetaphysics(OxfordUniversityPress,2003):180…20.Sider,Theodore(unpublished).BeyondtheHumphreyObjection.ŽSider,Theodore(unpublished).ReducingModality.ŽSosa,Ernest(1987).SubjectsAmongOtherThingsŽPhilosophicalPerspectives1,155…187.Unger,Peter(1979).WhyThereAreNoPeople,ŽMidwestStudiesinPhilosophy4,177…222.Unger,Peter(1980).TheProblemoftheMany,ŽMidwestStudiesinPhilosophy5,411…67.vanInwagen,Peter(1990).MaterialBeings(Ithaca:CornellUniversityPress).Williams,D.C.(1953).OntheElementsofBeing,ŽTheReviewofMetaphysicsVII,3…18&Williams,D.C.(1986).UniversalsandExistents,ŽAustralasianJournalofPhilosophy64,1…14.Wiggins,David(1967).IdentityandSpatio-TemporalContinuity(Oxford:BasilBlackwell).Wiggins,David(1980).SamenessandSubstance(Cambridge,Mass:HarvardUniversityPress).Wiggins,David(1997).OnBeingintheSamePlaceattheSameTime,ŽinMichaelRea(ed),MaterialConstitution:AReader(RowmanandLittlefield),3…9.Wiggins,David(2001).SamenessandSubstanceRenewed(CambridgeUniversityPress).Xu,Fei(1997).FromLotsWifetoaPillarofSalt:EvidencethatPhysicalObjectisaSortalConcept,Ž365…392.Yablo,Stephen(1987).Identity,EssenceandIndiscernibility,ŽJournalofPhilosophy84,293…Zimmerman,Dean(1995).TheoriesofMassesandProblemsofConstitution,ŽReview104,53…110.