removingvovokan

Millikan'sTeleosemanticsandCommunicativeAgencyPierreJacobMillikan'steleosemanticapproachconstitutesapowerfulframeworkforexplainingthecontinuedreproductionandproliferationofintentionalconventionallinguisticsigns,andtherebythestabilityofhumanverbalintentionalcommunication.
Whilethisapproachneedstobecomplementedbyparticularproximatepsychologicalmechanisms,Millikanrejectsthementalisticpsychologicalmechanisms,whicharepartoftheGriceantraditioninpragmatics.
ThegoalofthispaperistoassessthebalancebetweenMillikan'steleosemanticframeworkandtheparticularproximatepsychologicalmechanismsthatshefavors.
KeywordsAcceptance(compliance)/understanding|Communicative/informativeintention|Conventions|Coordination|Direct/derivedproperfunction|Etiologicaltheoryoffunctions|Imitativelearning|Mindreading|Naturalsigns|PerceptiontheoryofverbalunderstandingAuthorPierreJacobjacob@ehess.
frInstituteJeanNicodParis,FranceCommentatorMariusF.
Jungmjung02@students.
uni-mainz.
deJohannesGutenberg-UniversittMainz,GermanyEditorsThomasMetzingermetzinger@uni-mainz.
deJohannesGutenberg-UniversittMainz,GermanyJenniferM.
Windtjennifer.
windt@monash.
eduMonashUniversityMelbourne,Australia1IntroductionInthispaper,IwishtorevisitatopicthatIad-dressedmanyyearsago(cf.
Jacob1997)fromanovelperspective.
Muchphilosophyofmindofthelatterpartofthetwentiethcenturyhasbeendevotedtonaturalizingintentionalityorthecontentsofmentalrepresentations.
OneofthelandmarksofnaturalisticphilosophyofmindofthepastthirtyyearsisunquestionablyRuthMillikan'steleosemanticframework.
Tele-osemantictheoriesareteleologicaltheoriesthatseektoexplaincontentbyappealingtothefunctionsofrepresentations.
Likemostteleose-manticapproaches,Millikan(1984,2004)em-bracesanetiologicalconceptionoffunction,ac-cordingtowhichfunctionsareselectedeffects(Millikan1984,1989b;Neander1991,1995,2004;Wright1973):thefunctionofatraitistheeffectcausedbythetraitthatexplainsthecontinuedreproduction(survivalorprolifera-tion)ofpasttokensofthistrait.
Millikan'steleosemanticapproachispar-ticularlyimpressivefortworelatedreasons.
First,itappliesinasinglestroketothecon-tentsofintentionalmentalrepresentations,whosefunctionistomediatebetweenpairsofcognitivemechanismslocatedwithinsinglebrains,andalsotothemeaningsofintentionalconventionallinguisticsigns,whosefunctionisJacob,P.
(2015).
Millikan'sTeleosemanticsandCommunicativeAgency.
InT.
Metzinger&J.
M.
Windt(Eds).
OpenMIND:20(T).
FrankfurtamMain:MINDGroup.
doi:10.
15502/97839585701601|22www.
open-mind.
nettomediatebetweenpairsofcognitivemechan-ismslocatedinthebrainsofdistinctindividu-als.
Second,heroverallteleological(orteleo-functional)approach,basedontheetiologicaltheoryoffunctions,ismeanttoofferanaccountoftheproliferationorcontinuedreproductionofbothbiologicalentitiesandnon-biologicalcul-turalthings,suchaslinguisticandnon-lin-guisticconventions.
FollowingMayr(1961),evolutionarybiolo-gistsandphilosophershavelongarguedthatthedistinctionbetweenso-calledultimateandproximateexplanationsofbiologicaltraits(e.
g.
,behaviors)iscentraltoevolutionarytheorizing.
Roughlyspeaking,ultimateexplanationsad-dresswhy-questions:forexample,whydobirdssingWhydoessingingconferaselectionalad-vantage(orgreaterfitness)tobirdsProximateexplanationsaddresshow-questions:forex-ample,whataretheparticularexternalcircum-stanceswhichtriggersinginginbirdsWhataretheinternalbrainmechanismsthatallowbirdstosingThedistinctionbetweenultimateandproximatebiologicalexplanationsraisessomedeepscientificandphilosophicalquestions.
Onesuchquestioniswhetherultimateexplanationsshouldbeconstruedasnon-causalanswerstowhy-questions.
Somephilosophershavearguedthatultimateexplanationsareselectionalex-planationsbasedonnaturalselection.
Naturalselectioncanaccountfortheprevalenceofsometraitinapopulationofindividuals,butitcan-nottrackthecausalprocesswherebythetraitisgeneratedineachindividualinthefirstplace(Sober1984,pp.
147–152;Dretske1988,pp.
92–93;Dretske1990,pp.
827–830).
Otherphilo-sophershaverepliedthatselectionalexplana-tionsarecausalexplanations,onthegroundsthatnotokenofatraitwhosetypehasbeense-lectedforfulfillingits(etiological)functioncouldproliferateunlessitwaslinkedbyacausalchaintotheearlierproductionoftheselectedeffectbyancestortokensofthesametypeoftrait(Millikan1990,p.
808).
1Inthispaper,Iwillnotaddresssuchper-plexingissues.
Iwillsimplyacceptthevalidity1Forfurtherdiscussioncf.
Jacob1997,pp.
256–269.
ofthedistinctionandassumethat(whetherul-timateexplanationsarecausalexplanationsornot)ultimateandproximateexplanationsarecomplementary,notcompeting,explanations.
Giventhatwhy-questionsarefundamentallydifferentfromhow-questions,itislikelythatul-timateexplanationsofferfew(ifany)con-straintsonproximateexplanations,andviceversa.
Iwillfurtherassumethatthedistinctioncarriesoverfrombiologicaltoculturalevolutionandappliestotheevolutionofhumancommu-nication(cf.
Scott-Philippsetal.
2011).
Inpar-ticular,Millikan'sbasicteleosemanticaccountoftheproliferationofintentionalconventionallinguisticsignscanusefullybeconstruedasakindofultimateexplanationofhuman(verbalandnon-verbal)communication.
Itsmaintaskistoaddressquestionssuchas:whatistheevol-utionaryorculturalfunctionofhumancommu-nicationWhydohumansengageincommunic-ationatallAswithotherkindsofultimateex-planations,itneedstobesupplementedbyspe-cificproximateexplanationswhoseroleistodisclosetheparticularhumancognitivecapacit-iesandmentalprocesseswherebyhumanspro-duceandunderstandintentionalconventionalsigns.
Thegoalofthispaperistoassessthebal-ancebetweenMillikan'sbroadteleosemanticap-proachtothecooperativefunctionofhumancom-municationandthechoiceofparticularproximatepsychologicalmechanismsthatsheendorses.
Inparticular,Iwillfocusonheranti-mentalisticview,namelythatverbalunderstandingofan-other'sutteranceisakindofdirectperceptionofwhatevertheutteranceisabout,andhercorrelat-iverejectionofthebasicGriceanpragmaticas-sumptionthatverbalunderstandingisanexerciseinmindreading.
Oneofthedistinctivefeaturesofthehumanmindreadingcapacityisthatiten-ablesindividualstomakesenseoftwokindsofagency:instrumentalandcommunicativeagency.
Inordertomakesenseofanagent'sinstrumentalaction,onemustrepresentthecontentsofbothhermotivationsandepistemicstates.
Inordertomakesenseofanagent'scommunicativeaction,asGricehasbasicallyargued,theaddresseemustinferwhattheagentistryingtoconvey,i.
e.
,hercommunicativeintention,whoseveryfulfilmentJacob,P.
(2015).
Millikan'sTeleosemanticsandCommunicativeAgency.
InT.
Metzinger&J.
M.
Windt(Eds).
OpenMIND:20(T).
FrankfurtamMain:MINDGroup.
doi:10.
15502/97839585701602|22www.
open-mind.
netrequiresthatitisrecognizedbytheaddressee.
Whatisdistinctiveofhumanintentionalcommu-nicationisthatitenablesthecommunicativeagenttocauseheraddresseetoacquirenewpsy-chologicalstates,andtherebytomanipulatehismind.
Thus,IshallexaminethecontrastbetweentheparticularproximatemechanismsfavoredbyMillikanandtheGriceanprag-matictradition.
Inthefirstsection,IshallspelloutthebasicGriceanmentalisticframe-work.
Inthesecondsection,IwillspelloutMillikan'steleosemanticmachinery.
Inthethirdsection,IwillexamineMillikan'sviewthatverbalunderstandingisanextendedformofperception.
Inthefourthsection,Iwillex-aminetheextenttowhichMillikan'saccountofconventionscansupportherrejectionoftheGriceanassumptionthatverbalunderstandingisanexerciseinmindreading.
Finally,inthelastsection,Iwillshowthatrecentdevelop-mentalfindingsintheinvestigationofearlyhumansocialcognitionarerelevanttothecontroversybetweenMillikanandtheGriceantraditionoverthechoiceofproximatemechan-ismsunderlyinghumancommunication.
2TheGriceanmentalisticpictureofcommunicativeagencyTheGriceanmentalistictraditionrestsonthreebasicrelatedassumptions.
2Thefirstistheassumptionthatthecompleteprocesswherebyanaddresseecontributestothefullsuccessofaspeaker'scommunicativeactshouldbedecomposedintotwoseparablepsycho-logicalsub-processes:aprocessofunderstanding(orcomprehension)ofthespeaker'sutteranceandaprocessofacceptance,whichinturncanbeconstruedastheaddressee'sacquiringeitheranewbelieforanewdesireforaction(dependingonthedirectionoffitofthespeaker'sutterance).
I'llcallthistheseparabilitythesis.
2Althoughtherelevance-basedapproachadvocatedbySperber&Wilson(1986)andWilson&Sperber(2004)departsinsomeinter-estingrespectsfromGrice's(1969,1989)ownapproach,Iwillnon-ethelesscalltheirapproach"Gricean"because,inthecontextofthepresentpaper,thecontinuitiesbetweenthetwoframeworksarefarmoreimportantthanthediscontinuities.
TheGriceanmentalistictraditionalsorestsontheassumptionthatverbalunderstandingisanexerciseinmindreading,wherebytheaddresseerecognizesthecomplexpsycholo-gicalstatethatunderliesthespeaker'scom-municativeact.
I'llcallthisthemindreadingthesis.
(Clearly,themindreadingthesisispresupposedorentailedbytheseparabilitythesis.
)Third,theGriceanmentalistictraditionfur-therrestsonafundamentalhypothesizedasymmetrybetweenwhatisrequiredforun-derstandinginstrumentalnon-communicativeagencyandcommunicativeagency.
Anagentintendsherinstrumentalactiontosatisfyherdesireinlightofherbelief,andthedesirableoutcomeofherinstrumentalactioncanberecognizedeveniftheagentfailstofulfilhergoalorintention.
Buttheintendedeffectofaspeaker'scommunicativeaction,whichistheaddressee'sunderstandingofwhatshemeans,cannotbeachievedunlessthespeaker'sin-tentiontoachievethiseffectisrecognized(cf.
Sperber2000,p.
130).
Unlikepurelyinstru-mentalagency,communicativeagencyisos-tensiveinthefollowingsense.
Aspeaker'scommunicativeactisostensivebecauseitsdesirableoutcomecannotbeidentifiedunlesstheaddresseerecognizeswhatthespeakerin-tendstomakemanifesttohim,i.
e.
,whatSperber&Wilson(1986)callthespeaker'sinformativeintention.
Thus,theGriceantra-ditionrestsonthethesisoftheostensivenatureofcommunicativeagency(Sperber&Wilson1986).
2.
1ThemindreadingthesisOnthepictureofpragmaticswhichispartoftheGriceantraditionofthepastfortyyearsbroadlyconceived,ahumanagentcouldnotachieveaverbalornon-verbalactofintentionalcommunic-ationunlessshehadacomplexpsychologicalstate,whichGrice(1957)calledthe"speaker'smeaning"andwhichheconstruedasasetofthreeinterrelatedintentions.
3Firstofall,bypro-ducinganutterance(oranyotherpieceofostens-3Forbrevity,I'lluse"speaker"insteadof"communicativeagent".
Butofcoursenotallcommunicativeactionsareverbal.
Jacob,P.
(2015).
Millikan'sTeleosemanticsandCommunicativeAgency.
InT.
Metzinger&J.
M.
Windt(Eds).
OpenMIND:20(T).
FrankfurtamMain:MINDGroup.
doi:10.
15502/97839585701603|22www.
open-mind.
netivecommunicativebehavior),thespeakermusthavethebasicintention(i)toactonherad-dressee'smind,i.
e.
,tocausehimtoacquireanewbelieforanewdesire(orintention)toperformsomeaction.
Second,thespeakermustintend(ii)heraddresseetorecognizethecontentofherbasicintention.
Third,shemustfurtherintend(iii)heraddressee'srecognitionofherbasicintention(inaccordancewith(ii))toplayamajorroleinhisfulfillingherbasicintention(i).
Inthefollowing,Iwilladopt(Sperber&Wilson's(1986)simplifiedtwo-tieredaccount,accordingtowhichacommunicativeagentwhoproducesanutterancehastwo(notthree)inter-relatedintentions:aninformativeandacom-municativeintention,thefirstofwhichisnestedwithintheother.
Shehastheinformativeinten-tiontomakesomestateofaffairsmanifesttoheraddresseeandalsothefurthercommunicat-iveintentiontomakeherinformativeintentionmanifesttoheraddressee.
Sointhisframework,thespeaker'scommunicativeintentionisful-filledbytheaddresseeassoonasthelatterre-cognizes(orunderstands)whichstateofaffairsitisthespeaker'sinformativeintentiontomakemanifest.
Butmoreisrequiredforthespeaker'sinformativeintentiontobefulfilled:thead-dresseemustfurtheracceptthespeaker'sepi-stemicorpracticalauthority.
Dependingonthedirectionoffitofthespeaker'sutterance,theaddresseemusteitherbelievethefactwhichitisthespeaker'sinformativeintentiontomakemanifesttohim,orhemustacquirethedesiretoactsoastoturnintoafactthepossiblestateofaffairswhichitisthespeaker'sinformativeintentiontomakemanifesttohim.
Inanutshell,muchof(Sperber&Wilson's(1986)relevance-basedframeworkrestsontheirinsightfulrecognitionthat,onthebroadGriceanpictureofthespeaker'smeaning,thetaskoftheaddresseecanbeusefullydividedintotwobasicpsychologicalprocesses:oneistheprocesswherebytheaddresseeunderstands(orrecognizes)thespeaker'sinformativeinten-tionandtheotheristheprocesswherebyhefulfilsthespeaker'sinformativeintention.
Thefirstprocessinvolvestheaddressee'srecognitionofthespeaker'sinformativeintention,wherebytheaddresseefulfilsthespeaker'scommunicat-iveintentionthatherecognizethespeaker'sin-formativeintention.
Byrecognizingthespeaker'sinformativeintention,theaddresseecomesautomaticallytobothfulfilthespeaker'scommunicativeintentionandtounderstand(orcomprehend)thespeaker'sutterance.
Butfortheaddresseetorecognizethespeaker'sinform-ativeintentionisnotipsofactotofulfilit.
Sothesecondprocessneededforthesuccessofthespeaker'scommunicativeactinvolvesthead-dressee'sfulfilmentofthespeaker'sinformativeintention,wherebytheaddresseeeitheracceptsanewbelief(inaccordancewiththecontentofthespeaker'sassertion)orformsanewdesiretoact(inaccordancewiththecontentofthespeaker'srequest;cf.
Jacob2011).
2.
2TheseparabilitythesisWhiletherelevance-basedaccountofcommu-nicationclearlypresupposesthemindreadingthesis,Sperber(2001)hasofferedfurthersup-portinfavoroftheseparabilitythesis.
Follow-ingKrebs&Dawkins(1984),Sperber(2001)hasarguedthatforcooperativecommunicationtostabilizeinhumanevolution,itmustbead-vantageoustobothsendersandreceivers.
Sincetheinterestsofspeakersandhearersarenotidentical,thecooperationrequiredforthesta-bilizationofcommunicationisvulnerabletode-ception.
Whenherutteranceisdescriptive,thespeakercanspeakeithertruthfullyoruntruth-fully.
Theaddresseecaneithertrustthespeakerornot.
Thespeakerisbetteroffifheraddresseetrustsherandworseoffifhedistrustsher,whetherornotthespeakeristruthful.
Iftheaddresseetruststhespeaker,thenheisbetteroffifthespeakeristruthfulandworseoffifthespeakerisnottruthful,whiletheaddresseere-mainsunaffectedifhedistruststhespeaker.
Clearly,noteveryspeakeris(orshouldbe)grantedequalepistemicorpracticalauthorityonanytopicbyeveryaddressee.
AsSperberetal.
(2010)havefurtherargued,giventherisksofdeception,itislikelythathumancooperativecommunicationwouldnothavestabilizedinhu-manevolutionunlesshumanshadevolvedmechanismsofepistemicvigilance,wherebytheyfilterthereliabilityofdescriptiveutter-Jacob,P.
(2015).
Millikan'sTeleosemanticsandCommunicativeAgency.
InT.
Metzinger&J.
M.
Windt(Eds).
OpenMIND:20(T).
FrankfurtamMain:MINDGroup.
doi:10.
15502/97839585701604|22www.
open-mind.
netances.
Focusingonaspeaker'sassertionsattheexpenseofherrequests,aspeaker'sepistemicauthoritydependstoalargeextentonthead-dressee'sevaluationofherreliability(ortrust-worthiness),whichinturndependsjointlyontheaddressee'sevaluationofthespeaker'scom-petenceonthetopicathandandonthead-dressee'srepresentationofhowbenevolentarethespeaker'sintentionstowardshim.
AccordingtoSperberetal.
(2010),anaddressee'sepi-stemicvigilancecanapplytoeitherorboththesourceoftheinformationbeingcommunicatedanditscontent.
3Millikan'steleosemanticmachinery3.
1TeleosemanticsandinformationalsemanticsOneofthefirstattemptsatanaturalisticac-countofcontent(orintentionality)inthephilo-sophyofmindwasDretske's(1981)informa-tionalsemantics,accordingtowhichasignorsignalscarriesinformationaboutpropertyFiffthereisanomic(orlawful)covariationbetweeninstancesofFandtokeningsofs.
AsMillikan(1984,2004)emphasizedshortlyafter,informa-tionalsemanticsfacesthepuzzleofaccountingforthepossibilityofmisrepresentation.
IftheconditionalprobabilitythatFisinstantiatedgivensis1,thenhowcouldsevermisrepresentinstancesofFThispuzzleisneatlysolvedbyteleologicalapproaches:ifarepresentationhasafunction,thenitcanfailtofulfilitsfunctionandtherebymisrepresentwhatitisdesignedtorepresent(Millikan2004,Ch.
5).
AccordingtoDretske's(1988,1995)ownlaterattemptatpre-servinginformationalsemanticsaspartoftele-osemantics,asignorsignalscouldnotrepres-entsomepropertyFunlessshadthefunctionofcarryinginformationabout(orindicating)in-stancesofF.
Millikan's(1984,2004)teleosemanticap-proachsharplydepartsfromDretske's(1988,1995)information-basedframeworkinatleasttwofundamentalrespects.
Firstofall,inMil-likan'searliest(1984)teleosemanticframework,therewasnoroomforinformation-theoreticno-tionsatall.
Inherlater(1989a,2004)work,shearguedthatcarryinginformationcouldnotbeateleologicalfunctionofasignonthefollowinggrounds.
Whetherasigncarriesinformationaboutsomepropertydependsonhowthesignwascausedorproduced.
Butaccordingtotheetiologicaltheoryoffunctions,thefunctionofasignisoneofitsowneffects,i.
e.
,theselectedef-fectthatexplainsthecontinuedreproductionoftokensofsignsofthiskind.
Howasignwascausedcannotbeoneofitseffects,letaloneitsselectedeffect.
Ifandwhenasignhappenstocarryinformationaboutsomething,carryingin-formationcannotbeitsselectedeffect,i.
e.
,itsetiologicalfunction.
4Second,Dretske's(1981)informationalse-manticscouldonlybesuitablynaturalisticintherequiredsenseifinformationisconstruedastheconverseofnomologicalcovariation(orne-cessity),i.
e.
,asanentirelynon-intentionaland/ornon-epistemiccommodity.
ButasMil-likan(2004,pp.
32–34)argues,ifsignalscouldnotcarryinformationaboutFunlesstheprob-abilitythatFisinstantiatedwhensistokenedwere1(inaccordancewithsomenaturallaw),thennoanimalcouldeverlearnaboutFfromperceivingtokensofs.
Inher2004book,Millikanelaboratesanotionofnaturalsignthatismore"user-friendly"preciselybecause"itisatrootanepi-stemicnotion"(Millikan2004,p.
37).
OnMil-likan's(2004)account,naturalsigns(e.
g.
,tracksmadebyquail)arelocallyrecurrentsignswithinhighlyrestrictedspatialandtemporaldomains:relativetoonelocaldomain,suchtracksarenaturalsignsofquail.
Relativetoaneighboringdomain,theverysametracksaremadebypheasantsandarethereforenaturalsignsofpheasants,notquail.
Locallyrecurrentsignsaffordknowledgeoftheworldforanimalswhocanlearnhowtotrackthecircumscribeddomainsrelativetowhichtheycarryreliablein-formation.
Furthermore,locallyrecurrentnat-uralsignscanformtransitivechains(orbepro-ductivelyembedded)withincircumscribeddo-4Forsignificantdiscussionanddefenseoftheviewthatitistheetiolo-gicalfunctionofmentalrepresentationstocarryinformation,inre-sponsetoMilikan'scriticisms,cf.
Neander(1995,2007),Godfrey-Smith(2006)andShea(2007).
Cf.
therecentexchangebetweenNeander(2011)andMillikan(2011)ForacriticismofMillikan'sview,cf.
Pietroski(1992)andseeMillikan's(2000)reply.
Jacob,P.
(2015).
Millikan'sTeleosemanticsandCommunicativeAgency.
InT.
Metzinger&J.
M.
Windt(Eds).
OpenMIND:20(T).
FrankfurtamMain:MINDGroup.
doi:10.
15502/97839585701605|22www.
open-mind.
netmains.
Forexample,retinalpatternscanbeanaturalsignoftracksintheground,whichinturnmaybeanaturalsignofquailwithinacir-cumscribedlocaldomain.
Perceptioniswhatenablesnon-humananimalsandhumansaliketotrackthemeaningsoflocallyrecurrentnat-uralsignsintheircircumscribeddomainofvalidityandtherebytoacquireknowledgeoftheworld(Millikan2004,Ch.
4).
TheapplicationofMillikan's(1984,2004)teleosemanticframeworktothemeaningsofin-tentionalconventionalsignsresultsfromthecombinationofthreerelatedingredients:(i)theetiologicalviewoffunctions;(ii)acceptanceofthesender-receiverstructureasanecessarycon-ditiononthecontentsofintentionalrepresenta-tions;and(iii)anaturalisticaccountofthere-productionofconventions.
3.
2TheetiologicalconceptionoffunctionsAsIsaidabove,ontheetiologicalview,thefunctionofsometraitisitsselectedeffectthatexplainsthecontinuedreproductionofpasttokensofthistrait.
ThisiswhatMillikan(1984)callsadevice'sdirectproperfunction.
Butadevicemayalsohavewhatshecallsade-rivedproperfunction.
Forexample,itisthedir-ectproperfunctionofthemechanismofcolorchangeintheskinofchameleonstomakethemundetectablefromthelocalbackgroundbypredators.
Itisaderivedproperfunctionofthismechanisminaparticularchameleon,Sam,ataparticularplaceandtime,tomakethecolorofitsskinmatchthecolorofitsparticularlocalbackgroundatthattimesoastomakeitun-detectablebypredatorsthereandthen.
WhileMillikan'steleofunctionalframeworkbasedontheetiologicalapproachtofunctionsprimarilyfitsbiologicaltraits,itappliesequallytonon-biologicalitemssuchasnon-bodilytools—includingpublic-languageforms.
Forexample,screwdrivershavethedirectproperfunctionofturning(drivingorremoving)screws.
Thisistheeffectofscrewdriversthatexplainstheircontinuedreproduction.
Clearly,ascrewdrivermayalsobeintentionallyusedforthepurposeofdrivingascrewwithaparticularmetallicstructure,length,anddiameterintoaparticularwoodenmaterialataparticulartimeandplace.
Ifso,thendrivingthisparticularscrewintothisparticularwoodenmaterialataparticulartimeandplacewillbethederivedproperfunctionthatthisparticularscrewdriverinheritsfromtheagent'sintention.
3.
3Thesender-receiverframeworkAccordingtothesender-receiverframework,asignorsignalRcanbeanintentionalrepresent-ation(asopposedtoanaturalsign)onlyifitisarelatuminathree-placerelationinvolvingtwosystems(ormechanisms),oneofwhichisthesender(whoproducesR),theotherofwhichisthereceiver(whousesR).
Byapplicationoftheetiologicalviewoffunctions,thesender(orpro-ducer)andthereceiver(orconsumer)haveco-evolvedsothatwhatMillikan(1984,2004)callstheNormalconditionsfortheperformanceofthefunctionofonedependsontheperformanceoftheother'sfunctionandviceversa.
Inanut-shell,theproducerandtheconsumerareco-operativedevices,whoseinterestsoverlapandwhoseactivitiesarebeneficialtoboth.
Thus,thecooperativeternarysender–receiverstruc-turenaturallyappliestothecontentsofinten-tionalmentalrepresentationsthatmediatebetweencognitivemechanismslocatedwithinasingleorganism.
5Invirtueofthefactthatintentionalmen-talrepresentationscanhavetwobasicdirectionsoffit,theevolvedcooperationbetweenthepro-ducerandtheconsumercantaketwobasicforms.
Ifandwhentherepresentationisde-scriptiveorhasamind-to-worlddirectionoffit,theproducer'sfunctionistomakeasignR,whosecontentmatchessomestateofaffairsS,forthepurposeofenabling(orhelping)thecon-sumertoperformitsowntaskwhenandonlywhenSobtains.
Ifandwhentherepresentationisdirective(orprescriptive)orhasaworld-to-minddirectionoffit,theproducer'sfunctionistoproducearepresentationwhosecontentwillguidetheconsumer'saction,anditisthecon-5Cf.
Godfrey-Smith(2013)andArtiga(forthcoming)forfurtherelab-oratediscussionoftherequirementofcooperationasaconditiononapplicationofthesender-receiverstructure.
Jacob,P.
(2015).
Millikan'sTeleosemanticsandCommunicativeAgency.
InT.
Metzinger&J.
M.
Windt(Eds).
OpenMIND:20(T).
FrankfurtamMain:MINDGroup.
doi:10.
15502/97839585701606|22www.
open-mind.
netsumer'sfunctiontomaketheworldmatchthecontentofthesignbyitsownactivities.
Fur-thermore,Millikan(1995,2004)hasarguedthatthemostprimitivekindsofintentionalmentalrepresentations(sharedbyhumansandnon-hu-mananimals)arewhatshecallspushmi-pullyurepresentations,whichareatoncedescriptiveandprescriptive,withbothamind-to-worldandaworld-to-minddirectionoffit.
3.
4ConventionalpatternsThethirdcomponentofMillikan'steleose-manticapproachtothemeaningsofintentionalconventionalsignsinvolvesher(1998)natural-isticaccountofconventions.
Onheraccount,so-callednaturalconventionalityrestsontwoele-mentarycharacteristics:first,naturalconven-tionsarepatternsthatarereproduced(orthatproliferate).
Second,theyarereproduced(or"handeddown")"owingtoprecedentdeterm-inedbyhistoricalaccident,ratherthanowingtopropertiesthatmakethemmoreintrinsicallyserviceablethanotherformswouldhavebeen"(Millikan2005,p.
188).
Thefactthatconven-tionsrestonhistoricalprecedenttoalargeex-tentaccountsfortheirarbitrariness.
6Onthebasisofhernaturalisticaccountofthecontin-uedreproductionofnaturalconventions,Mil-likanfurtheroffersapurportedlynaturalisticaccountofthecontinuedreproductionofcon-ventionalpublic-languagesigns,whosefunctionistocoordinatethetransferofinformationbetweenspeakersandhearers.
Shetherebyex-tendsthecooperativeternarysender–receiverstructuretothemeaningsofintentionalconven-tionalsigns(orpublic-languageforms)thatme-diatecognitivemechanismslocatedwithinpairsofdistinctorganisms.
Conventionalpublic-languageformsaretoolsormemesinDawkins's(1976)sense:theyhavebeenselectedandhaveaccordinglybeenreproducedbecausetheyservecoordinatingfunctionsbetweenasender(thespeaker)andareceiver(theaddressee),whoseinterestsoverlap.
Butlikeanyothertool,inadditiontoitsdirectmemetic(or"stabilizing")function(whichex-6Includingthearbitrarinessoftherelationbetweenparticularword-typesandwhattheymean(senseand/orreference).
plainsitscontinuedreproduction),aparticulartokenofsomepublic-languageformmayalsohaveaderivedfunctionorpurpose,derivedfromthepurposeofthespeakerwhoproduceditataparticularplaceandtime.
Thus,atokenofapublic-languageformhastwokindsofpur-poses:amemeticpurposeandthespeaker'spurpose,whichmayornotcoincide(cf.
Millikan1984,2004,2005).
4Isverbalunderstandinganextendedformofperception4.
1PerceivingtheworldthroughlanguageOnebasicproblemraisedbyMillikan'saccountoftheproliferationofintentionalconventionalsignsisthatoneandthesamelinguisticformdetachedfromitscontextofusemaybelongtodifferentmemeticfamilies(orchainsofrepro-ductiveevents).
Inthereproductiveprocess,whatgetscopiedfromonepairofsender-receiv-erstothenextisnotmerelyalinguisticform(e.
g.
,"clear"),buttheuseofalinguisticformembeddedinaparticularcontext.
ThisiswhyonMillikan's(2005,Ch.
10,section3)view,theboundarybetweensemanticsandpragmaticsisblurryandtheprocesswherebyahearertracksthememeticlineageofaconventionalsignisapragmaticprocess.
Ontheteleosemanticap-proach,thehearer'staskistoretrievetheap-propriatecontextnecessaryforrecognizingthecorrectmemeticfamily(orlineage)towhichaparticularconventionalsignbelongs.
Inanut-shell,thehearers'taskistotrackthedomainsofintentionalconventionalsigns.
Thus,itwouldappearthatthehearer'staskisquitesimilartowhatisinvolvedintrackingtherestricteddomainoverwhichtheinformationcarriedbyalocallyrecurrentnat-uralsign(e.
g.
,tracksmadeeitherbyquailorbypheasants)isvalid.
Sincetrackingthelocaldo-mainsoverwhichtheinformationcarriedbylocallyrecurrentnaturalsignsisaperceptualtask,itisnotsurprisingthatMillikanhasper-sistentlyurgedthat"inthemostusualcasesunderstandingspeechisaformofdirectpercep-tionofwhateverspeechisabout.
InterpretingJacob,P.
(2015).
Millikan'sTeleosemanticsandCommunicativeAgency.
InT.
Metzinger&J.
M.
Windt(Eds).
OpenMIND:20(T).
FrankfurtamMain:MINDGroup.
doi:10.
15502/97839585701607|22www.
open-mind.
netspeechdoesnotrequiremakinganyinferencesorhavinganybeliefsaboutwords,letaloneaboutspeakerintentions"(Millikan1984,p.
62).
7Millikan(2004,p.
122)nicelyillustratesherviewthatverbalunderstandingisanexten-dedformofperception:raindoesnotsoundthesamewhenheardfallingontheroof,onearth,onsnow,andonthewater,eventhoughitmaybedir-ectlyperceivedasrainthroughanyofthesemedia.
Exactlysimilarly,rainhasadifferentsoundwhenthemediumoftrans-missionistheEnglishlanguage("It'srain-ing!
").
AnditsoundsdifferentagainwhenthemediumoftransmissionisFrenchorGerman.
Inanutshell,"duringNormalconversation,itisnotlanguagethatismostdirectlyperceivedbythehearerbutrathertheworldthatismostdirectlyperceivedthroughlanguage"(Millikan2005,p.
207).
Furthermore,bothordinaryandextendedperceptionrestontranslation,notinference:"thefirststepsinperceptioninvolvereactingtonaturalsignsoffeaturesoftheouterworldbytranslatingthemintoinnerintentionalrepres-entationsoftheseouterfeatures,forexample,ofedges,lines,anglesoflightsourcesinrelationtotheeye"(Millikan2004,p.
118).
Innormalverbalcommunication,translationplaysatwo-foldroleinmediatingtransferfromthespeaker'sbelieftotheaddressee'sbelief.
First,thespeakerofadescriptiveutterancetranslatesherbeliefintoasententialconventionalsign.
Secondly,theaddresseetranslatesthecontentofthespeaker'sutteranceintohisownnewbelief(Millikan1984,2004,2005).
4.
2OrdinaryandextendedperceptionClearly,Millikan'sthesisthatverbalunder-standingisanextendedformofperceptionisnotconsistentwiththeGriceanthesisthatverbalunderstandingisanexerciseinmindreading.
Butonthefaceofit,thethesis7Cf.
Millikan(2000,Ch.
6),Millikan(2004,Ch.
9),Millikan(2005,Ch.
10).
thatverbalunderstandingisanextendedformofperception(ofwhateverspeechisabout)itselfispuzzlingforatleastthreerelatedreasons.
8Firstofall,asMillikan(2004,Ch.
9)herselfre-cognizes,thereisamajordifferencebetweenthecontentofaperceptualrepresentationofsomestateaffairsandtheverbalunderstandingofthecontentofanother'stestimonyabouttheverysamestateofaffairs.
Atanappropriatedistanceandingoodlightingconditions,onecouldnotperceiveacuprestingonatablewithoutalsoperceivingitsshape,size,color,texture,con-tent,orientation,andspatiallocationwithre-specttothetable,toanyotherobjectrestingonthetable,andespeciallytooneself.
AsMil-likan(2004,p.
122)recognizes,unlikethecon-tentoftestimony,thecontentofordinaryper-ceptioncanbeputattheserviceofactionpre-ciselybecauseitprovidesinformationabouttheagent'sspatialrelationtoanobjectthatispo-tentiallyrelevantforaction.
Butifanaddresseelocatedinaroomnexttothespeaker'sroomunderstandsthecontentofthelatter'sutter-anceofthesentence"Thereisacuponthetable",hemayendorsethebeliefthatthereisacuponthetablewithouthavinganydefiniteex-pectationabouttheshape,size,color,texture,content,orientation,andspatiallocationofthecupwithrespecttohimself,thetable,orany-thingelse.
Second,thethesisthatverbalunderstand-ingisanextendedformofperceptionoughttoberestrictedtothehearer'sverbalunderstand-ingofthemeaningsofdescriptiveutterancesofindicativesentenceswithamind-to-worlddirec-tion-of-fit,whichdescribefacts(oractualstatesofaffairs).
Itcannotwithoutfurthermodifica-tionsbedirectlyappliedtothehearer'sverbalunderstandingofthemeaningsofprescriptiveutterancesofimperativesentenceswherebyaspeakerrequestsanaddresseetoactsoastoturnapossible(non-actual)stateofaffairsintoafact(oranactualstateofaffairs).
Prescript-iveutterances,whichhaveaworld-to-minddir-ectionoffit,failtodescribeanyfactthatcouldbedirectlyperceivedatall.
SothequestionariseswhetherMillikanwouldbewillingtoen-8Cf.
Recanati(2002)foradefenseofMillikan'sthesis.
Jacob,P.
(2015).
Millikan'sTeleosemanticsandCommunicativeAgency.
InT.
Metzinger&J.
M.
Windt(Eds).
OpenMIND:20(T).
FrankfurtamMain:MINDGroup.
doi:10.
15502/97839585701608|22www.
open-mind.
netdorsetherevisedtwo-tieredthesisthat(i)averbalunderstandingofaspeaker'sdescriptiveutteranceistheperceptionofwhatevertheut-teranceisaboutand(ii)averbalunderstandingofaspeaker'sprescriptiveutteranceistointendtoperformwhateveractionismostlikelytocomplywiththespeaker'srequest.
Finally,testimonyenablesaspeakertoconveybeliefswhosecontentsfaroutstriptheperceptualcapacitiesofeitherthespeakerorheraddressee.
Forexample,anaddresseemayunderstandthatthespeakerintendstoverballyconveytohimherbeliefthatthereisnogreatestinteger,thatdemocracyistheworstformofgovernmentexceptallthoseotherformsthathavebeentriedfromtimetotime,orthatreligionistheopiumofthepeople.
Butitdoesnotmakemuchsensetoassumethateitherthespeakerorheraddresseecouldperceivewhatthespeaker'sutteranceisabout.
4.
3TrackingthedomainsofintentionalconventionalsignsFurthermore,thethesisthatverbalunderstand-ingisanextendedformofperceptionclearlyrestsontheassumptionthattheprocesswherebythehearerofaspeaker'sutterancetracksthememeticfamilyoftheintentionalconventionalsignusedbythespeakerisbasic-allythesameastheprocesswherebyhumanandnon-humananimalstrackthemeaningsoflocallyrecurrentnaturalsignsintheircircum-scribeddomainofvalidity.
AsImentionedabove,Millikan(2004)arguesthatperceptionisthebasicprocesswherebyanimalstrackthemeaningsoflocallyrecurrentnaturalsignsintheircircumscribeddomainofvalidity.
Cru-cially,onecantrackthemeaningsoflocallyre-currentnaturalsignswithintheircircumscribeddomainofvaliditywithoutrepresentinganagent'spsychologicalstate.
Sothequestionariseswhetherahearerofaspeaker'sutterancecouldalwaystrackthememeticfamilyofthein-tentionalconventionalsignsusedbyaspeakerwithoutrepresentinganyofthespeaker'spsy-chologicalstates.
Inparticular,asRecanati(2007)hasar-gued,thequestionarisesfordescriptiveutter-ancescontainingatleastfourkindsofconven-tionalexpressionsconsideredbyMillikan(2004,Chs.
10–12):so-calledunarticulatedconstitu-entsinPerry's(1986)sense,incompletedefinitedescriptions,quantifiers,andpossessives.
Con-siderfirstanutteranceof(1):(1)Itisraining.
Itisunlikelythatbyanutteranceof(1)aspeakermeanstoassertthatitisrainingsome-whereorotheratthetimeofutterance.
In-stead,sheislikelytomeanthatitisrainingatthetimeofutteranceandattheplaceofutter-ance(whichremainsunarticulatedinthesen-tence).
Ifbyanutteranceof(1),thespeakercouldonlymeanthatitisrainingattheplaceofutterance,thenMillikan'sclaimthatahearerneednotrepresentanyofthespeaker'spsycho-logicalstatesforthepurposeoftrackingthelocaldomainsofintentionalconventionalsignsmightbevindicated.
However,byanutteranceof(1)onthephone,aspeakerlocatedinParismaymeanthatitisraininginChicago,notinParis.
Similarly,aFrenchspeakerlocatedinParismayusetheincompletedescription"thePresident"torefer,nottotheFrenchPresident,butinsteadtothePresidentoftheUS.
Forthepurposeofunderstandinganutter-anceofasentencecontainingauniversalquan-tifier,asshownbyexample(2),thehearermustbeabletoproperlyrestrictthedomainofthequantifier:(2)Everyoneisasleep.
Byanutteranceof(2),thespeakerpre-sumablymeanstoassert,notthateveryoneintheuniverseisasleep,butthateveryoneinsomerestricteddomain(e.
g.
,arelevanthouse-hold)isasleep.
9Therelevantrestricteddomainisthedomainthespeakerhasinmind.
Finally,byusingthepossessiveconstruction"John'sbook",thespeakermayhaveinmindmanydif-ferentrelationsbetweenJohnandthebook:shemaymeanthebookwrittenbyJohn,thebook9Aniceexamplesuggestedbyarefereeis"Thereisnobeerleft",wheretheaudi-encedoesnottakethespeakertomeanthatthereisnobeerleftintheuniverse,butinsteadinsomeproperlyrestricteddomain(e.
g.
,somerelevantfridge).
Jacob,P.
(2015).
Millikan'sTeleosemanticsandCommunicativeAgency.
InT.
Metzinger&J.
M.
Windt(Eds).
OpenMIND:20(T).
FrankfurtamMain:MINDGroup.
doi:10.
15502/97839585701609|22www.
open-mind.
netreadbyJohn,thebookboughtbyJohn,thebooksoldbyJohn,thebookJohnlikes,thebookJohndislikes,thebookJohnjustreferredtointheconversation,thebookJohnlost,thebookJohngavetothespeaker,thebookthespeakergavetoJohn,thebookthehearergavetoJohn,andsoon.
Unlessthehearerhypothes-izeswhatrelationthespeakerhasinmind,hewillfailtounderstandwhatthespeakermeansbyherutteranceof"John'sbook".
Innoneofthesefourcasesdoesitseemasifthehearercouldrecognizethememeticfamilyofinten-tionalconventionalsigns,i.
e.
,tracktheirrelev-antdomains—unlesshecouldrepresentthecon-tentsofsomeofthespeaker'sbeliefsorassump-tions.
5Conventionsandbelief-desirepsychology5.
1TeleosemanticsandtheseparabilitythesisMillikan'sthesisthatverbalunderstandingisanextendedformofperceptionismeantasanal-ternativetotheGriceanthesisthatverbalun-derstandingisanexerciseinmindreading.
ThefurtherquestionarisestowhatextentMillikan'steleosemanticaccountoftheproliferationofpubliclanguageconventionsisconsistentwiththeGriceanseparabilitythesis,i.
e.
,thedistinc-tionbetweenverbalunderstandingandeitheracceptance(belief)orcompliance.
Iwillfirstar-guethatthereisarestrictedsenseinwhichMil-likan'steleosemanticsseemstobeconsistentwiththeseparabilitythesis.
ButIwillfurtherarguethatinabroadersenseMillikan'srejec-tionofthemindreadingthesisunderminestheseparabilitythesis.
OnMillikan'steleosemanticaccount,foraspeaker'sdescriptiveutteranceofanindicativesentencetomeettherequirementofcooperation(andmutualinterest)betweenthesender(orproducer)andthereceiver(orconsumer),itsdirectproperfunctionmustbetocausethead-dresseetoforma(true)belief.
Foraspeaker'sprescriptiveutteranceofanimperativesentencetomeettherequirementofcooperation,itsdir-ectproperfunctionmustbetocausethead-dresseetoactincompliancewiththecontentofthespeaker'srequest.
Intheterminologyoftherelevance-basedframework,aspeakerwhouttersadescriptiveutterancemakesmanifesttoheraddresseehercommunicativeintentiontomakemanifestherinformativeintentiontomakesomefactmani-festtohim.
Theaddresseemayfulfilthespeaker'scommunicativeintentionbyrecogniz-ingherinformativeintentionandyetfailtoful-filherinformativeintentionbyresistingendors-ingtherelevantbelief.
Aspeakerwhouttersaprescriptiveutterancemakesmanifesttoherad-dresseehercommunicativeintentiontomakemanifestherinformativeintentiontomakemanifesttohimthedesirabilityofturningsomepossiblestateofaffairsintoafactbyhisownaction.
Theaddresseemayfulfilthespeaker'scommunicativeintentionbyrecognizingherin-formativeintentionandyetfailtofulfilthespeaker'sinformativeintentionbyresistingen-dorsingtheintentiontoactinaccordancewiththespeaker'srequest.
Origgi&Sperber(2000,pp.
160–161),whosubscribetotheGriceanthesisoftheseparabilitybetweenverbalunderstandingandacceptanceorcompliance,havearguedthatthedirectproperfunctionofeitheradescriptiveut-teranceoraprescriptiveutterancecouldnotbetoreliablyelicittheaddressee'sresponse"atthelevelofbeliefordesireformation"(i.
e.
,"thecognitiveoutputsofcomprehension"),butin-stead"atanintermediatelevelintheprocessofcomprehension".
Millikanmightreplythatac-cordingtoherteleosemanticframework,anut-terancemayhaveadirectproperfunctionandyetremainunfulfilled.
Ifso,thenthefactthatanaddresseemayfulfilthespeaker'scommunic-ativeintention(byrecognizingherinformativeintention)andyetfailtofulfilthespeaker'sin-formativeintentionseemsentirelycompatiblewiththeteleosemanticframework.
However,totheextentthatMillikanex-plicitlyrejectsthemindreadingthesis,whichispresupposedbytheseparabilitythesis,itisunlikelythatshewouldfindtheseparabilitythesisitselfacceptable.
Ontherelevance-basedapproach,itisasufficientconditionforsecuringwhatAustin(1975)calledthe"up-Jacob,P.
(2015).
Millikan'sTeleosemanticsandCommunicativeAgency.
InT.
Metzinger&J.
M.
Windt(Eds).
OpenMIND:20(T).
FrankfurtamMain:MINDGroup.
doi:10.
15502/978395857016010|22www.
open-mind.
nettake"(orsuccess)ofacommunicativeact(orspeechact)thatthespeakercausesthead-dresseetofulfilthespeaker'scommunicativeintentionbyrecognizingherinformativeinten-tion.
Itisnotnecessarythattheaddresseefurtherfulfilthespeaker'sinformativeinten-tion.
Successfulcommunicationdoesnotre-quiretheaddresseetoaccepteitheranewbe-lieforanewdesire,inaccordancewiththespeaker'sinformativeintention.
ButonMil-likan'steleosemanticframework,failureoftheaddresseetocomplywiththespeaker'sgoalofcausingtheaddresseetoaccepteitheranewbelieforanewdesirelookslikeafailureoftheaddresseetocooperatewiththespeaker'sconventionalaction,andthereforelikeabreakdownofthespeaker'scommunicativeac-tion.
Infact,Millikan(2000,2004,2005)hasofferedtwobroadgroundsforrejectingthemindreadingthesis,bothofwhichmakeitun-likelythatshewouldsupporttheseparabilitythesis;thesecondofwhichisbasedondevel-opmentalevidence.
Istartwiththenon-devel-opmentalargument.
5.
2CooperationandsocialconformityFirst,Millikan(2004)rejectsthemindreadingthesisaspartofhercriticismofthereasoningthatleadstotheseparabilitythesis:sherejectsthejointassumptionsthathumanpredictionsofothers'behaviorarebasedonmindreadingandthatcooperationinhumanverbalcommunica-tionisvulnerabletotherisksofdeception.
Ontheonehand,shearguesthat"mostaspectsofsociallivinginvolvecooperationinwaysthatbenefittoeveryone[…]forthemostpart,socialcooperationbenefitsbothorallparties.
Thereisnothingmysteriousaboutitsevolutioninthisrespect"(Millikan2004,pp.
21–22).
Inanut-shell,Millikanarguesthattheurgetoexplainhowthebenefitsofhumancommunicationarenotoffsetbytherisksofdeceptionismisplacedonthegroundsthattheinterestsofspeakersandhearersaresufficientlysimilar,ifnotidentical.
1010AsGodfrey-Smith(2013,p.
45)observes,samenessofinterestsinhumancooperationcanbesafelyassumedinsmallcontemporarycommunities,butnotonalargescale,andnorinanevolutionarycontext.
Ontheotherhand,shearguesthatweusebelief-desirepsychology,notforprediction,but"forexplanationafterthefact"(Millikan2004,p.
22).
Thisisconsonantwithher(1984,pp.
67–69)earlierclaimthatwhilehumanadultshavetheabilitytoreflectonaspeaker'scommu-nicativeintentioniftheautomaticflowofcon-versationisinterruptedforonereasonoran-other,normalverbalunderstandingdoesnotre-quirerepresentingthespeaker'scommunicativeintention.
Instead,normalverbalunderstandingshouldbeconstruedasaconventionaltransferofinformationwherebythespeakertranslatesherbeliefintoanutterance,whosemeaningisinturntranslatedbackbytheaddresseeintoanewlyacquiredbelief.
Thus,Millikanrejectstwoofthemajoras-sumptionsonwhichtheseparabilitythesisrests.
Sheunderestimatesthegapbetweenthein-terestsofspeakersandhearersinhumancom-municationandsheminimizestheroleofbelief–desirepsychologyinthepredictionofothers'behavior.
Interestingly,herrejectionofbothas-sumptionsrestsinturnonherowncompetingaccountofcommunicativeacts.
Assheputsit,"asurpriseofthisanalysisoftheconventionalnatureoftheinformation-transferringfunctionoftheindicativeisthatbelievingwhatyouhearsaidintheindicativeturnsouttobeaconven-tionalact,somethingonedoesinaccordancewithconvention"(Millikan2005,p.
46).
11Firstofall,Millikan(2004,p.
23)arguesthathu-mansexpectotherstobehaveinconformitywithsocialconventions,notonthebasisofoth-ers'beliefsanddesires.
Second,shefurtherspeculatesthattheconventionalbehaviorsthatarecausedbyadispositiontosocialconformitymayderivefromnaturalselectionthememeticfunctionofservingacoordinatingfunction(ibid.
).
Clearly,beingdisposedtosocialconform-ityandexpectingotherstobesimilarlydis-posedmayhelpsolvecoordinationproblems(asshownbydrivingononesideoftheroad).
How-ever,beingdisposedtowardssocialconformityisnotsufficienttocomplywithsocialconven-tions.
Compliancerequireslearning,i.
e.
,theac-11Notethatthisquoteseemstopresupposethenegationofthesepar-abilitythesis.
Jacob,P.
(2015).
Millikan'sTeleosemanticsandCommunicativeAgency.
InT.
Metzinger&J.
M.
Windt(Eds).
OpenMIND:20(T).
FrankfurtamMain:MINDGroup.
doi:10.
15502/978395857016011|22www.
open-mind.
netquisitionofrelevanttruebeliefsaboutthecon-tentsofsocialconventions.
Thus,thebasicchal-lengeforMillikan'sclaimthathumansexpectotherstobehave,notsomuchinaccordancewiththecontentsoftheirbeliefsanddesires,butinconformitywithsocialconventions,istoofferanaccountofhowhumanscometolearnandtherebyknowwhatittakestoactincon-formitywithsocialconventions.
5.
3CounterpartreproductionThisissuehasbeenhighlightedbytheexchangebetweenTomasello's(2006)commentsonMil-likan's(2005)bookandMillikan's(2006)re-sponse,whichfocusesonMillikan's(1998)thesisthatmanyconventions,whosefunctionitistosolvecoordinationproblems,arerepro-ducedbywhatshecallscounterpartreproduc-tion(ornutsandboltreproduction).
Typicalcoordinationproblemsinvolveatleasttwopart-ners,whoshareacommonpurposethatcanbeachievedonlyifeachpartnerplaysitsassignedrole,wherebothpartnerscanberequiredtoperformeitherthesameactortwodistinctcomplementaryacts.
Incounterpartreproduc-tion,whentherespectiverolesofeachpartnerrequirethemtoperformtwodifferentcomple-mentaryacts,onetypicallyadjustsherbehaviortotheother'sandvice-versa.
Counterpartre-productionisexemplifiedby,e.
g.
,handshakere-production,thereproductionoftherespectiveposturesassumedbymenandwomenintradi-tionaldancing,thereproductionofsocialdis-tancesappropriateforconversation,orthere-productionoftheuseofchopsticksforeating.
Similarly,Millikan(2005,2006)arguesthatcounterpartreproductionalsounderliesthecon-tinuedreproductionofconventionalpublic-lan-guagesigns.
Millikan(2005)furthermentionsopen,partiallyorcompletelyblind,conventionalleader–followerco-ordinationsinvolvedinjointactionsbasedonsharedgoals,wherebyoneagent(theleader)introducesacomponentofapatternwhosecompletionrequiresherpartner(thefollower)toperformacomplementarycom-ponent(ibid.
,pp.
12–14).
Oneexampleofopenconventionalleader–followercoordinationisthepatternwherebyoneagentselectsherseatatanarbitrarytableinarestaurantandherpartnerfollowssuitandselectshisaccordingly.
Oneex-ampleofapartiallyblindconventionalleader–followercoordinationisthecouch-movingpat-ternwherebytheleaderaffordsthefolloweran-ticipatorycuesofhernextmovebyostensiblyexaggeratingherownmovements,wherethefol-lower'sfamiliaritywiththepatternenableshimtorecognizetheleader'sostensivecuesandtherebytoreproducethecomplementarypor-tionofthejointaction.
AnotherofMillikan'sexamplesofapartiallyblindconventionalleader–followercoordinationistheUSmailbox-flagconvention,wherebytheleaderputsupaflagaftershehasplacedmailinthemailboxandthepostmanpicksupthemailafterper-ceivingtheflag.
MuchcomparativeworkbyTomaselloandcolleagues(reportedbyTomaselloetal.
2005andsummarizedbyTomasello2006,2008)showsthatwhilemostcommunicativegesturesinchimpanzeesarelearntbyontogeneticritual-ization,mostcommunicativebehaviorsinhu-maninfantsareacquiredbyimitativelearning.
AsTomasello(2006)argues,Millikan'sownre-quirementthatthereproductionofaconven-tionalpatterndependson"theweightofpre-cedent",notonitsperceivedintrinsicsuperiorabilitytoproduceadesiredresult,seemsbetterfulfilledbyaprocessofimitativelearningthanbyaprocessoftrialanderrorwherebyoneindi-vidualadjustsherbehaviortoanother's.
Thereseemstobenothingarbitrary(asthereshouldifitwereconventional)aboutanindividual'sadjustingherbehaviortoanother's.
WhileTo-masello(2006)doesnotdenythatcounterpartreproductionplaysasignificantroleinculturaltransmission,hedisputestheclaimthattheoutputofcounterpartreproductionqualifiesasconventional.
PartofthegapbetweenMillikanandTo-maselloliesinwhattheytaketobetheproperunitfortheanalysisofthemechanismunderly-ingthecontinuedreproductionofconventionalpatternsinvolvedinsolvingproblemsofco-ordination.
WhileTomasellofocusesonthelearningcapacitiesofsingleindividualminds,MillikanfocusesonwhatcanbeachievedbytheJacob,P.
(2015).
Millikan'sTeleosemanticsandCommunicativeAgency.
InT.
Metzinger&J.
M.
Windt(Eds).
OpenMIND:20(T).
FrankfurtamMain:MINDGroup.
doi:10.
15502/978395857016012|22www.
open-mind.
netreciprocaladjustmentsofpairsofcooperativepartners.
Forexample,whenMillikan(2005,2006)arguesthatcounterpartreproductionun-derliesthecontinuedproliferationofthecustomofusingchopsticksforeatinginsomecultures,sheconstruestheconventionofusingchopsticksasasolutiontotheproblemofcoordinationbetweenpairsofpartners,someofwhombuychopsticksandusethemforeatingandotherswhomanufacturechopsticks.
Thelatterwouldnotmanufacturechopsticksunlesstheformerboughtthemandusedthemforeating.
Con-versely,theformerwouldnotbuythemandusethemforeatingunlessthelattermanufacturedthem.
Butofcourse,asMillikanisaware,thisleavesopenthequestionofhowyoungchildrenlearntousechopsticksforeating.
AsMillikan(2006,pp.
45–46)rightlyobserves,younghu-manchildrenunderstandtheirnativelanguagelongbeforetheycanspeakit.
Norcantheylearntounderstandbyimitatingmaturespeak-ers:assheputsit,"theydon'twatchhowotherpeopleunderstandandthencopy".
Shefurtherarguesthatyoungchildrenwouldneverunder-standtheirnativetongueunless"theirteachers"spoketothem,but"theirteachers"wouldneverspeaktoyoungchildrenunless"theyhadhadsomereasonablysuccessfulexperience"withpreviouslisteners.
Thismakesthecontinuedre-productionofconventionalpublic-languagesignsfitthepatternofcounterpartreproduc-tion.
Butstillthequestionarises:howdoyoungchildrenlearntoproducewordsoftheirnativetongueVocalimitativelearningmaywellplayanimportantrole(cf.
Hauseretal.
2002).
Inanutshell,accordingtoMillikanthefunctionofconventionsistosolvecoordinationproblems.
Sheoffersanelegantaccountoftheprolifera-tionofconventionsbasedoncounterpartrepro-duction.
Heraccountmustmakeroomfortheroleofimitativelearninginthewayyounghu-manchildrenlearneithertousechopsticksforeatingortoproduce(andnotjustunderstand)wordsoftheirnativetongues.
AsIshallargueinsection6.
2,evidenceshowsthatimitativelearninginyoungchildrenrestsontheirabilitytoconstruethemodel'sdemonstrationasanos-tensivecommunicativeaction.
Ifso,thenMil-likan'sviewthatcounterpartreproductionun-derliestheproliferationofconventionsmustmakeroomfortheroleofchildren'sabilitytorecognizethemodel'scommunicativeintention.
6Teleosemanticsandthepuzzlesofearlyhumansocialcognition6.
1Millikan'sdevelopmentalpuzzleTofurtherunderminethemindreadingthesis,Millikan(1984,2000,2004,2005)hasalsoap-pealedtofindingsfromthedevelopmentalpsy-chologicalinvestigationofearlyhumansocialcognition,showingthat"childrenyoungerthanaboutfour,althoughfairlyproficientintheuseoflanguage,don'tyethaveconceptsofsuchthingsasbeliefs,desires,andintentions"(Mil-likan2005,p.
204).
Ifsuchchildrendonothavesuchconcepts,then,unlikeadults,theycannotreflectivelyengageintasksofmindreading,i.
e.
,intrackingthecontentsofothers'intentions,beliefs,anddesires.
Totheextentthattheycanengageinverbalunderstanding,thisfurthershowsthatverbalunderstandingcannotrestonmindreading(orbelief–desirepsychology).
AsMillikanemphasizes,muchdevelop-mentalevidenceshowsthatbeforetheyareatleastfouryearsoldthemajorityofhumanchil-drensystematicallyfailelicited-responsefalse-belieftasks.
(Intheterminologyofdevelop-mentalpsychologistsBaillargeonetal.
2010,eli-cited-responsetasksaretasksinwhichaparti-cipantisrequestedtogenerateanexplicitan-swerinresponsetoanexplicitquestion.
)Forexample,intheSally-Annetest,afterSallyplaceshertoyinthebasket,sheleaves.
WhileSallyisaway,AnnemovesSally'stoyfromthebaskettothebox.
WhenSallyreturns,parti-cipants,whoknowthetoy'sactuallocation,areexplicitlyaskedtopredictwhereSally(whofalselybelieveshertoytobeinthebasket)willlookforhertoy.
Theevidenceshowsthatthemajorityofthree-year-olds,"althoughquitepro-ficientintheuseoflanguage"(inMillikan'sterms,Millikan2005,p.
204),typicallypointtothebox(i.
e.
,thetoy'sactuallocation),nottothebasketwheretheagentfalselybelieveshertoytobe(cf.
Wimmer&Perner1983,Baron-Jacob,P.
(2015).
Millikan'sTeleosemanticsandCommunicativeAgency.
InT.
Metzinger&J.
M.
Windt(Eds).
OpenMIND:20(T).
FrankfurtamMain:MINDGroup.
doi:10.
15502/978395857016013|22www.
open-mind.
netCohenetal.
1985andWellmanetal.
2001forameta-analysis).
Millikanassumesthatthefailureofmostthree-year-oldsinsuchelicited-responsefalse-belieftasksdemonstratesthattheylackwhatshecallsa"representationaltheoryofmind".
Inanutshell,sheassumesthatsuccessatelicited-responsefalse-belieftasksisanecessarycondi-tionforcreditinganindividualwitharepres-entationaltheoryofmind(i.
e.
,theabilitytotrackthecontentsofothers'falsebeliefs).
Ac-ceptanceofthisassumptiongivesrisetoMil-likan'sdevelopmentalpuzzle,whichis"toun-derstandhowveryyoungchildrencanbeawareoftheintentionsandofthefocusofattentionofthosefromwhomtheylearnlanguagewithoutyethavingthissortofsophisticatedtheoryofmind"(Millikan2005,p.
205).
BeforeexplainingwhyMillikan'sassumptioniscontentious,IshallbrieflyexamineMillikan'ssolutiontoherownpuzzle.
Millikan'ssolutioninvolvesthreerelatedingredients,themostimportantofwhichisherthesisthatnormalverbalunderstandingisanextendedformofperception(whichdoesnotre-quirethinkingaboutaspeaker'sintentionatall).
Second,shearguesthatyoungchildrencanunderstandthegoal-directednessofaspeaker'scommunicativeactionwithouttrackingthecon-tentofhercommunicativeintention.
Third,shearguesthatyoungchildrencanunderstandthereferentialfocusofaspeaker'sattentionwithouthavingasophisticatedtheoryofmind.
AsIun-derstandit,muchoftheargumentforthepos-sibilityofunderstandingthereferentialfocusofaspeaker'sattentionwithouthavingasophist-icatedtheoryofmindrestsonthethesisthatverbalunderstandingisanextendedformofperception.
AsIhavealreadyexpresseddoubtsaboutthethesisthatverbalunderstandingisanextendedformofperception,Ishallnowbrieflyexaminethesecondthesis:thatyoungchildrencouldunderstandthegoal-directednessofaspeaker'scommunicativewithouttrackingthecontentofhercommunicativeintention.
Millikan(2005,pp.
206–207)offerstwomainreasonsforgrantingyoungchildrentheabilitytorecognizethegoal-directednessofaspeaker'scommunicativeactionwithoutgrant-ingthemafullrepresentationaltheoryofmind.
First,shearguesthattheevidenceshowsthatmammals(dogsandcatsandnon-humanprim-ates,presumably,aswell)lackarepresenta-tionaltheoryofmindbuthavetheabilitytore-cognizethegoal-directednessofeachother'sbe-havior.
Sobyparity,veryyoungchildrenshouldalsobegrantedtheabilitytorecognizethegoal-directednessofothers'actions,includingspeak-ers'communicativeactions.
Second,shearguesthatcommunicativeactionsarecooperativeac-tions.
Whenyoungchildrenareengagedinsomecooperativeaction(includingacommunicativeaction)withacaretaker,theycaneasilykeeptrackofthesharedgoalofthecooperativeac-tion,whiletrackingthefocusofthespeaker'svisualattention,withouthavingafullrepres-entationaltheoryofmind.
Ontheonehand,thereisevidencethatnon-humanprimatesrecognizethegoalsofcon-specificsengagedintheexecutionofinstru-mentalactions(Call&Tomasello2008).
Ontheotherhand,thereisalsoevidencethatnon-hu-manprimates—andbirdsaswell—candiscrim-inateknowledgeableagents(whoknowabout,e.
g.
,foodfromvisualperception)fromignorantagents(whodon'tknowaboutfoodbecausetheirlineofvisionisobstructed)incompetitivesituations(Bugnyar2011;Call&Tomasello2008;Dallyetal.
2006;Hareetal.
2001;To-maselloetal.
2003).
ButthequestionraisedbyMillikan'spuzzleistounderstandwhatenablesveryyounghumanchildrentomakesensejointlyofaspeaker'sgoalandthefocusofhervisualattention,whenthespeakerisperformingacommunicativeaction,notaninstrumentalac-tion,inacooperative,notacompetitive,con-text.
Thefactthatnon-humanprimatescanrepresentthegoalofanagent'sinstrumentalac-tionanddiscriminateaknowledgeablefromanignorantagentinacompetitivecontextfallsshortofprovidingtherequiredexplanation.
Furthermore,twoofMillikan'sassump-tionsarecontentiousinlightofrecentfindingsfromdevelopmentalpsychology.
Oneisheras-sumptionthatyoungchildrencouldrecognizethegoal-directednessofspeakers'communicat-iveactionswithoutarepresentationaltheoryofmind.
TheotherisherassumptionthatsuccessJacob,P.
(2015).
Millikan'sTeleosemanticsandCommunicativeAgency.
InT.
Metzinger&J.
M.
Windt(Eds).
OpenMIND:20(T).
FrankfurtamMain:MINDGroup.
doi:10.
15502/978395857016014|22www.
open-mind.
netatelicited-responsefalse-belieftasksshouldbetakenasacriterionforhavingtheabilitytotrackthecontentsofothers'falsebeliefs(andthereforehavingarepresentationaltheoryofmind).
Ishallstartwiththeformer,whichamountstodenyingtheasymmetrybetweenin-strumentalandcommunicativeagency—whichIearlierdubbedthethesisoftheostensivenatureofhumancommunicativeagency.
6.
2ThepuzzleofimitativelearningThefirstrelevantdevelopmentalfinding,repor-tedbyGergelyetal.
(2002),showsthatapprox-imatelyone-year-oldhumanchildren(fourteen-month-olds)selectivelyimitateanagent'soddaction.
First,infantswereprovidedwithostens-ivecueswherebyanagentmademanifestherintentiontoconveysomevaluableinformationbylookingintotheinfants'eyesandaddressingtheminmotherese.
Shethentoldtheinfantsthatshefeltcoldandcoveredhershoulderswithablanket.
Shefinallyperformedanoddhead-actionwherebysheturnedalightboxinfrontofherbyapplyingherhead,intwoslightlydifferentconditions.
Inthehands-occu-piedcondition,sheusedherhandsinordertoholdtheblanketaroundhershoulderwhilesheexecutedthehead-action.
Inthehands-freecon-dition,sheostensiblyplacedherfreehandsonthetablewhilesheexecutedthehead-action.
Gergelyetal.
(2002)foundthatwhile69%ofthechildrenreplicatedthehead-actioninthehands-freecondition,only21%didinthehands-occupiedcondition.
Inthehands-occu-piedcondition,themajorityofchildrenusedtheirownhandstoturnthelightboxon.
Csibra&Gergely(2005,2006)furtherreportthattheasymmetrybetweeninfants'replicationofthemodel'soddhead-actioninthehands-freeandhands-occupiedconditionsvanishesifthemodelfailstoprovideinfantswithostensivecues.
Gergely&Csibra(2003)havereportedevidencethattwelve-month-oldsexpectagentsengagedintheexecutionofinstrumentalactionstoselectthemostefficientactionasameansto-wardsachievingtheirgoal(orgoal-state),inthecontextofrelevantsituationalconstraints.
SothefindingsonimitationreportedbyGergelyetal.
(2002)raisethefollowingpuzzle.
Manymoreinfantsreplicatedtheagent'shead-actionwhentheteleologicalrelationbetweentheagent'smeansandtheagent'sgoalwasopaque(inthehands-freecondition)thanwhenitwastrans-parent(inthehands-occupiedcondition).
Whydidinfantsreproducetheagent'shead-actionmorewhenitwasalessefficientmeansofachievingtheagent'sgoalofswitchingthelightboxonTheGriceanthesisabouttheostensivenatureofcommunicativeagencyandtheasym-metrybetweeninstrumentalandcommunicativeagencyisrelevanttoansweringthispuzzle.
Ar-guably,receptionofostensivesignalspreparedtheinfantstointerprettheagent'sactionasacommunicative,notaninstrumental,action.
Itmademanifesttotheinfantsthattheagentin-tendedtomakesomethingnovelandrelevantmanifesttothembyhersubsequentnon-verbalcommunicativeaction.
Inthehands-occupiedcondition,theinfantslearnthowcontactwasnecessaryinordertoturnonthelightbulb,whichwaspartofanunfamiliardevice.
Sincethemodel'shandswereoccupied,theinfantswhoseownhandswerefreeassumedthatthattheywerefreetoselectthemostefficientmeansattheirdisposaltoachievethesamegoalasthemodel.
Inthehands-freecondition,themodelcouldhaveusedherhands,butshedidnot.
Sotheinfantslearntfromthemodel'snon-verbaldemonstrationthattheycouldturnthelightonbyapplyingtheirownheads.
Ontheonehand,theevidenceshowsthatinfantsconstrueimitativelearningasaresponsetoanagent'scommunicativeactionandthattheyselectivelyimitateamodel'sactionasafunctionofwhattheytaketoberelevantlyhighlightedbythemodel'scommunicativeact(cf.
Southgateetal.
2009).
Ontheotherhand,furtherevidenceshowsthatnewbornsprefertolookatfaceswithdirectgazeoverfaceswithavertedgaze.
Rightafterbirth,theydisplaysensitivitytoeye-contact,infant-directedspeechormotherese,andinfant-contingentdistalre-sponsivity.
Ifprecededbyostensivesignals,anagent'sgazeshifthasbeenshowntogenerateinpreverbalhumaninfantsareferentialexpecta-Jacob,P.
(2015).
Millikan'sTeleosemanticsandCommunicativeAgency.
InT.
Metzinger&J.
M.
Windt(Eds).
OpenMIND:20(T).
FrankfurtamMain:MINDGroup.
doi:10.
15502/978395857016015|22www.
open-mind.
nettion,i.
e.
,theexpectationthattheagentwillrefertosomeobject(Csibra&Volein2008,cf.
Csibra&Gergely2009,andGergely&Jacob2013,forreview).
OnefurtherintriguingpieceofevidencefortheearlysensitivityofhumantoddlerstotheostensivenatureofhumancommunicativeagencyisofferedbyexperimentsthatshednewlightontheclassicalA-not-Bperseverationer-rorphenomenonfirstreportedbyPiaget(1954).
Infantsbetweeneightandtwelvemonthsareen-gagedinanepisodichide-and-seekgameinwhichanadultrepeatedlyhidesatoyunderone(A)oftwoopaquecontainers(AandB)infullviewoftheinfant.
Aftereachhidingevent,theinfantisallowedtoretrievetheobject.
DuringtesttrialswherethedemonstratorplacestheobjectrepeatedlyundercontainerB,infantscontinuetoperseverativelysearchforitundercontainerAwhereithadbeenpreviouslyhid-den.
ExperimentalfindingsreportedbyTopaletal.
(2008)showthatminimizingthepresenceofostensivecuesresultsinsignificantdecreasesoftheperseverativebiasinten-month-olds.
Thisfindingisconsistentwiththeassumptionthatinfantsdonotinterpretthehide-and-seekgameasagame,butinsteadasateachingsessionabouttheproperlocationofatoy.
Allthisevidencestronglysuggeststhathumaninfantsarepreparedfromthestarttorecognizenonverbalostensivereferentialsignalsandaction–demonstrationsaddressedtothemasencodinganagent'scommunicativeintentiontomakemanifestherinformativeintentiontomakesomerelevantstateaffairsmanifesttotheaddressee.
Butofcoursethisraisesapuzzle:howcouldpreverbalinfantsrecognizeanagent'scommunicativeintentiontomakemanifestherinformativeintentionAnovelapproachtothispuzzlehasbeeninsightfullysuggestedbyCsibra(2010).
AccordingtoCsibra,veryyounginfantsmightwellbeinapositionsimilartothatofaforeignaddresseeofaverbalcommunicativeact,whoisunabletoretrieveaspeaker'sinformat-iveintentionforlackofunderstandingofthemeaningofthespeaker'sutterance.
Nonethe-less,theforeignaddresseemaywellrecognizebeingthetargetofthespeaker'scommunicativeintentiononthebasisofthespeaker'sostensivebehavior.
Furthermore,ostensivesignalstowhichpreverbalhumaninfantshavebeenshowntobeuniquelysensitivecanplausiblybesaidtocodethepresenceofanagent'scommunicativeintention.
Ifthisiscorrect,thenlittle(ifany)furtherworkwouldbeleftforpreverbalinfantstoinferthepresenceofaspeaker'scommunicat-iveintentionafterreceivingostensivesignals.
6.
3Thepuzzleaboutearlyfalse-beliefunderstandingAsMillikanhasemphasized,muchdevelop-mentalpsychologyhasshownthatthemajorityofthree-year-oldsfailelicited-responsefalse-be-lieftasks.
Forexample,whenaskedtopredictwhereanagentwithafalsebeliefwilllookforhertoy,mostthree-year-oldswhoknowthetoy'slocationpointtothetoy'sactuallocation,andnottotheemptylocationwherethemis-takenagentbelieveshertoytobe.
However,inthepasttenyearsorso,developmentalpsycho-logistshavefurtherdesignedvariousspontan-eous-responsefalse-belieftasks,inwhichparti-cipantsarenotaskedanyquestionandthereforenotrequestedtoproduceanyanswer.
Typicalspontaneous-responsetasksinvolvetheuseoftheviolation-of-expectationandanticipatory-lookingparadigms,whichinvolvetwosteps.
Inhabituationorfamiliarizationtrials,parti-cipantsarefirstexperimentallyinducedtoformexpectationsbybeingrepeatedlyexposedtooneandthesameevent.
Second,intesttrialsofvi-olation-of-expectationexperiments,participantsarepresentedwitheitheranexpectedoranun-expectedevent.
Bymeasuringthetimeduringwhichparticipantsrespectivelylookattheex-pectedvs.
theunexpectedevent,psychologistsgetevidenceaboutthenatureandcontentoftheinfants'expectationsformedduringtheha-bituationorfamiliarizationtrials.
Psychologistscanalsousetheanticipatory-lookingparadigmandexperimentallydeterminewhereparti-cipantsfirstlookinanticipationoftheagent'saction,therebyrevealingtheirexpectationaboutthecontentoftheagent'sbelief.
Thus,inaseminalstudybasedonthevi-olation-of-expectationparadigmbyOnishi&Baillargeon(2005),fifteen-month-oldssawanJacob,P.
(2015).
Millikan'sTeleosemanticsandCommunicativeAgency.
InT.
Metzinger&J.
M.
Windt(Eds).
OpenMIND:20(T).
FrankfurtamMain:MINDGroup.
doi:10.
15502/978395857016016|22www.
open-mind.
netagentreachforhertoyeitherinagreenboxorinayellowboxwhenshehadeitheratrueorafalsebeliefabouthertoy'slocation.
OnishiandBaillargeonreportthatfifteen-month-oldslookedreliablylongerwhentheagent'sactionwasincongruentratherthancongruentwiththecontentofeitherhertrueorfalsebelief.
Inastudybasedontheanticipatorylookingparadigm,twenty-five-month-oldswereshowntolookcorrectlytowardstheemptylocationwhereamistakenagentbelievedhertoytobe,inanticipationofheraction(Southgateetal.
2007).
Manyfurthersubsequentstudiesshowthattoddlersandevenpreverbalhumaninfantsareabletotrackthecontentsofothers'falsebeliefsandexpectotherstoactinaccordancewiththecontentsoftheirtrueandfalsebeliefs.
InaclassicalexperimentbyWoodward(1998),six-month-oldswerefamiliarizedwithanagent'saction,whorepeatedlychoseoneoftwotoys.
Inthetesttrials,thespatiallocationsofthetoyswereswitchedandtheinfantseithersawtheagentselectthesametoyasbeforeatanewlocationoranewtoyattheoldlocation.
six-month-oldslookedreliablylongerattheformerthanatthelattercondition.
Luo&Bail-largeon(2005)furthershowedthatinfantsdonotlookreliablylongeratachangeoftargetif,inthefamiliarizationtrials,theagentre-peatedlyreachedforthesameobject,buttherewasnocompetingobject(forfurtherdiscussioncf.
Jacob2012).
Thisresulthasbeenwidelyin-terpretedasshowingthatsix-month-oldsareabletoascribeapreferencetoanagent.
Luo(2011)furtherfoundthatten-month-oldswhoknowthatanagentisinfactconfrontedwithonlyoneobject(nottwo)ascribeapreferencetotheagentifshefalselybelievesthatsheisconfrontedwithapairofobjects,butnotiftheagentknows(astheinfantsdo)thatsheiscon-frontedwithonlyoneobject.
Thus,thepsychologicalinvestigationofearlyhumansocialcognitioniscurrentlycon-frontedwithapuzzledifferentfromthatcon-frontedbyMillikan:ontheonehand,robustfindingsshowthatthemajorityofthree-year-oldsfailelicited-responsefalse-belieftaskssuchastheSally-Annetest.
Ontheotherhand,morerecentfindingsbasedonspontaneous-responsetasksshowthatpreverbalinfantsexpectotherstoactinaccordancewiththecontentsoftheirtrueandfalsebeliefs.
Thepuzzleis:howdowemakesenseofthediscrepancybetweenbothsetsofexperimentalfindingsSofar,psychologistshaveofferedtwobroadstrategiesforthis,oneofwhichassumes(asMillikandoes)thatsuccessatelicited-re-sponsefalse-belieftasksisanecessaryconditionoftheabilitytoascribefalsebeliefstoothers,whichistakentobetheoutputof"aculturalprocesstiedtolanguageacquisition"(Perner&Ruffman2005,p.
214).
Theirburdenistoex-plainawaythefindingsaboutpreverbalinfantswithoutcreditingthemwiththeabilitytotrackthecontentsofothers'falsebeliefs.
Thus,themajorityof"culturalconstructivist"psycholo-gistsofferlow-levelassociationistaccountsofthefindingsaboutpreverbalinfantsbasedonspontaneous-responsetasks.
Otherpsychologists(includingBaillargeonetal.
2010;Bloom&German2000Leslie2005;Leslieetal.
2004;Leslieetal.
2005;Scottetal.
2010)arguethatthefindingsaboutpreverbalinfantsshowthattheycantrackthecontentsofothers'falsebe-liefs.
Theirburdenistoexplainwhyelicited-re-sponsefalse-belieftasksaresochallengingforthree-year-olds.
Theprevalentnon-constructiv-istexplanationistheprocessing-loadaccountofferedbyBaillargeonandcolleagues.
Thecoreoftheassociationiststrategyistoaccountforfindingsaboutpreverbalhumaninfantsbasedonspontaneous-responsetasksonthebasisofathree-wayassociationbetweentheagent,theobject,anditslocation.
Itpostulatesthatinfantswilllooklongerinthetesttrialsateventsthatdepartmorestronglyfromthethree-wayassociationgeneratedbythefamiliar-izationtrials.
Forexample,inthetesttrialsofOnishi&Baillargeon(2005),infantsshouldlooklongerwhentheagentreachesforhertoyintheyellowboxifinthefamiliarizationtrialstheagentplacedhertoyinthegreenboxonthreerepeatedoccasions.
ThemainobstaclefortheassociationistpathisarecentstudybySenjuetal.
(2011)basedontheanticipatory-lookingparadigm.
Inthefamiliarizationstage,eighteen-month-oldsexperiencetheeffectofwearingeitheranJacob,P.
(2015).
Millikan'sTeleosemanticsandCommunicativeAgency.
InT.
Metzinger&J.
M.
Windt(Eds).
OpenMIND:20(T).
FrankfurtamMain:MINDGroup.
doi:10.
15502/978395857016017|22www.
open-mind.
netopaqueblindfoldthroughwhichtheycannotseeoratrickblindfoldthroughwhichtheycansee.
Inthefirsttrialsofthetestphase,thechildrenarefamiliarizedtoseeinganagentretrievehertoyatthelocationwhereapuppethasplaceditinfrontofher.
Theagent'sactionisalwayspre-cededbyapairofvisualandauditorycues.
Inthelasttesttrial,theagentfirstseesthepup-petplacethetoyinoneofthetwoboxes;shethenostensiblycovershereyeswithablindfold,andfinallythepuppetremovesthetoy.
Afterthepuppetdisappears,theagentremovesherblindfoldandthecuesareproduced.
Usinganeye-tracker,Senjuetal.
(2011)foundthatonlyinfantswhohadexperiencedanopaqueblind-fold,notinfantswhohadexperiencedatricksee-throughblindfold,reliablymadetheirfirstsaccadetowardstheemptylocationinanticipa-tionoftheagent'saction.
Senjuetal.
's(2011)findingsareinconsist-entwiththeassociationiststrategy:sinceallin-fantssawexactlythesameevents,theyshouldhaveformedexactlythesamethreefoldassoci-ationbetweentheagent,thetoy,andtheloca-tion,andonthisbasistheyshouldhavegazedatthesamelocationinanticipationoftheagent'saction.
Buttheydidnot.
Onlyinfantswhoseviewhadbeenpreviouslyobstructedbyanopaqueblindfold,notthosewhoseviewhadnotbeenobstructedbyatrickblindfold,expec-tedtheblindfoldedagenttomistakenlybelievethattheobjectwasstillintheopaquecontainerafterthepuppetremovedit.
Theevidenceagainsttheassociationiststrategyisalsoevidenceagainsttheassumption(acceptedbyMillikan)thatsuccessatelicited-responsefalse-belieftasksisanecessarycondi-tionforhavingarepresentationaltheoryofmindandbeingabletotrackthecontentsofothers'falsebeliefs.
Butthisassumptionisun-likelytobecorrectif,asseveralcriticsoftheculturalconstructiviststrategyhaveargued,theabilitytoascribefalsebeliefstoothersisnotasufficientconditionforsuccessatelicited-re-sponsefalse-belieftasks.
Asadvocatesoftheprocessing-loadaccount(Baillargeonetal.
2010)haveargued,anagentcouldhavetheabilitytoascribefalsebeliefstoothersandstillfailelicited-responsefalse-belieftasksforatleastthreereasons:shecouldfailtounderstandthemeaningofthelinguistically-encodedsen-tenceusedbytheexperimentertoasktheques-tion.
Shecouldfailtoselectthecontentoftheagent'sfalsebeliefintheprocesswherebysheanswerstheexperimenter'squestion.
Shecouldfailtohavetheexecutive-controlresourcesne-cessarytoinhibittheprepotenttendencytoan-swerthequestiononthebasisofthecontentofherowntruebelief.
Iwillnowarguethatsolv-ingthepuzzleaboutearlybelief-understandingmaywelldependonacceptanceoftheGriceanthesisoftheostensivenatureofcommunicativeagencyandtheasymmetrybetweeninstru-mentalandcommunicativeagency.
Inowwanttoofferaspeculativesolutiontothepuzzleaboutearlyfalse-beliefunder-standingbasedontworelatedGriceanassump-tions.
Thefirstistheasymmetrybetweenthenon-ostensivenatureofinstrumentalagencyandtheostensivenatureofhumancommunicat-iveagency.
Thesecondrelatedassumptionisthatthehumanabilitytotrackthecontentofthefalsebeliefofanagentofaninstrumentalactionmustbeaby-productoftheabilitytodealwithdeception(e.
g.
,lying)inthecontextofhumancommunicativeagency.
InthetypicalSally-Anneelicited-responsefalse-belieftask,participantsarerequestedtomakesenseoftwoactionsperformedbytwodif-ferentagentsatthesametime:theymusttrackthecontentsofthemotivationsandepistemicstatesofamistakenagentengagedintheexecu-tionofaninstrumentalaction(Sally)andtheymustalsomakesenseofthecommunicativeac-tionperformedbytheexperimenterwhoasksthem"WherewillSallylookforhertoy"Thefindingsbasedonspontaneous-responsetasksstronglysuggestthatmuchbeforetheybecomeproficientinlanguageuse,younghumanchil-drenareabletospontaneouslytrackthecon-tentsofthefalsebeliefsofagentsofinstru-mentalactions.
Sothequestionis:whatisitabouttheexperimenter'squestionthatbiasesthemtowardspointingtothetoy'sactualloca-tionInHelmingetal.
(2014),wehavearguedthattwobiasesareatwork,oneofwhichisareferentialbiasandtheotherofwhichisaco-Jacob,P.
(2015).
Millikan'sTeleosemanticsandCommunicativeAgency.
InT.
Metzinger&J.
M.
Windt(Eds).
OpenMIND:20(T).
FrankfurtamMain:MINDGroup.
doi:10.
15502/978395857016018|22www.
open-mind.
netoperativebias.
Thereferentialbiasitselfturnsontwocomponents.
Ontheonehand,theex-perimentercouldnotaskthequestion"Wherewilltheagentlookforhertoy"unlessshere-ferredtothetoy.
Ontheotherhand,theexperi-mentersharestheparticipants"correctepi-stemicperspectiveonthetoy'slocation.
Inan-sweringtheexperimenter'squestion,parti-cipantshavetheoptionofmentallyrepresentingeitherthetoy'sactuallocationortheemptylocation(wherethemistakenagentbelieveshertoytobe).
Theexperimenter'squestionmaybiasyoungchildren'sanswertowardstheactuallocationbyvirtueofthefactthattheexperi-menterbothreferredtothetoy(whoseactuallocationtheyknow)andsharedtheparti-cipants'correctepistemicperspectiveonthetoy'sactuallocation(attheexpenseofthemis-takenagent'sincorrectperspectiveontheemptylocation).
Whatwefurthercalltheco-operativebiasisthepropensityofyoungchil-drentohelpanagentwithafalsebeliefabouthertoy'slocationachievethegoalofherinstru-mentalactionbypointingtotheactuallocation(cf.
Warneken&Tomasello2006,2007;Knud-sen&Liszkowski2012),inaccordancewiththeirowntruebeliefaboutthetoy'sactualloc-ation.
Ifso,thenyoungchildrenmightinterpretthepredictionquestion"WherewillSallylookforhertoy"asanormativequestion:"WhereshouldSallylookforhertoy"Ofcourse,thecorrectanswertothenormativequestionisthetoy'sactuallocation,nottheemptylocationwherethemistakenagentbelieveshertoytobe.
7ConclusionThegoalofthispaperwastoassessthegapbetweenMillikan'sparticularviewsaboutsomeoftheproximatepsychologicalmechanismsun-derlyinghumancommunicationandthreecoreassumptionsoftheGriceanapproach:themindreadingthesis,theseparabilitythesis,andtheostensivenatureofcommunicativeagency.
IhavecriticizedfiveofMillikan'sbasicclaimsaboutpsychologicalmechanisms:(i)verbalun-derstandingisbestconstruedasanextendedformofperception;(ii)hearerscantrackthedomainsofintentionalconventionalsignswithoutrepresentinganyofthespeaker'spsy-chologicalstates;(iii)theoverlapbetweentheinterestsofspeakersandhearersunderminestheseparabilitythesis;(iv)humanscanpredictoth-ers'behavioroutofsocialconformity;(v)devel-opmentalpsychologysupportstheviewthatneitherverbalunderstandingnorlanguageac-quisitionrequiresarepresentationaltheoryofmind.
Millikan'smajorteleosemanticcontribu-tionhasbeentoopenanentirelynovelap-proachtothecontinuedreproductionofinten-tionalconventionalpublic-languagesigns.
Aswasshownbythediscussionofwhetherherviewoftheproperfunctionofdescriptiveandprescriptiveutterancesisconsistentwiththeseparabilitythesis,thereisroomfordisagree-mentaboutparticularpsychologicalmechan-ismswithinateleosemanticapproach.
IdonotthinkthatMillikan'steleosemanticframeworkforaddressingthecontinuedreproductionofin-tentionalconventionalsignsmandatesthepar-ticularchoiceofproximatepsychologicalmech-anismsthatsherecommends.
Oneofthemajorchallengesforthescientificinvestigationofcul-turalevolutionistomakesurethattheproxim-atepsychologicalmechanismsthatunderliethecontinuedreproductionofhumanculturalcon-ventionsaresupportedbyfindingsfromexperi-mentalpsychologicalresearch,inparticularde-velopmentalpsychology.
AcknowledgementsIamgratefultotheeditorsforinvitingmetowritethisessay,andtoNedBlock,CarstenHansen,GeorgesRey,DanSperberandtwoan-onymousreviewersfortheircomments.
Jacob,P.
(2015).
Millikan'sTeleosemanticsandCommunicativeAgency.
InT.
Metzinger&J.
M.
Windt(Eds).
OpenMIND:20(T).
FrankfurtamMain:MINDGroup.
doi:10.
15502/978395857016019|22www.
open-mind.
netReferencesArtiga,M.
(forthcoming).
Signalingwithoutcooperation.
Austin,J.
(1975).
Howtodothingswithwords.
Oxford,UK:OxfordUniversityPress.
Baillargeon,R.
,Scott,R.
M.
&He,Z.
(2010).
False-beliefunderstandingininfants.
TrendsinCognitiveSciences,14(3),110-118.
10.
1016/j.
tics.
2009.
12.
006Baron-Cohen,S.
,Leslie,A.
&Frith,U.
(1985).
Doestheautisticchildhavea"theoryofmind".
Cognition,21(1),37-46.
10.
1016/0010-0277(85)90022-8Bloom,P.
&German,T.
(2000).
Tworeasonstoabandonthefalsebelieftaskasatestoftheoryofmind.
Cogni-tion,77(1),25-31.
10.
1016/S0010-0277(00)00096-2Bugnyar,T.
(2011).
Knower-guesserdifferentiationinravens:others'viewpointsmatter.
ProceedingsoftheRoyalSocietyofLondonB:BiologicalSciences,278(1705),634-640.
10.
1098/rspb.
2010.
1514Call,J.
&Tomasello,M.
(2008).
Doesthechimpanzeehaveatheoryofmind30yearslater.
TrendsinCog-nitiveSciences,12(5),187-192.
10.
1016/j.
tics.
2008.
02.
010Csibra,G.
(2010).
Recognizingcommunicativeintentionsininfancy.
MindandLanguage,25(2),141-168.
10.
1111/j.
1468-0017.
2009.
01384.
xCsibra,G.
&Gergely,G.
(2009).
Naturalpedagogy.
TrendsinCognitiveSciences,13(4),148-153.
10.
1016/j.
tics.
2009.
01.
00Csibra,G.
&Volein,A.
(2008).
Infantscaninferthepres-enceofhiddenobjectsfromreferentialgazeinforma-tion.
BritishJournalofDevelopmentalPsychology,26(1),1-11.
10.
1348/026151007X185987Dally,J.
M.
,Emery,N.
J.
&Clayton,N.
S.
(2006).
Food-cachingwesternscrub-jayskeeptrackofwhowaswatchingwhen.
Science,312(5780),1662-1665.
10.
1126/science.
1126539Dawkins,R.
(1976).
Theselfishgene.
Oxford,UK:Ox-fordUniversityPress.
Dretske,F.
(1981).
Knowledgeandtheflowofinforma-tion.
Cambridge,MA:MITPress.
(1988).
Explainingbehavior.
Cambridge,MA:MITPress.
(1990).
Replytoreviewersofexplainingbehavior:Reasonsinaworldofcauses.
PhilosophyandPhe-nomenologicalResearch,50(4),819-839.
(1995).
Naturalizingthemind.
Cambridge,MA:MITPress.
Gergely,G.
&Csibra,G.
(2003).
Teleologicalreasoningininfancy:Thenaivetheoryofrationalaction.
TrendsinCognitiveSciences,7(7),287-292.
10.
1016/S1364-6613(03)00128-1(2005).
Thesocialconstructionoftheculturalmind:imitativelearningasamechanismofhumanpedagogy.
InteractionStudies,6(3),463-481.
10.
1075/bct.
4.
19ger(2006).
Sylvia'srecipe:theroleofimitationandpedagogyinthetransmissionofculturalknowledge.
InS.
Levenson&N.
Enfield(Eds.
)Rootsofhumanso-ciality:Culture,cognition,andhumaninteraction(pp.
229-255).
Oxford,UK:BergPublishers.
Gergely,G.
,Bekkering,H.
&Kiraly,I.
(2002).
Rationalimitationinpreverbalinfants.
Nature,415(6873),755-755.
10.
1038/415755aGergely,G.
&Jacob,P.
(2013).
Reasoningaboutinstru-mentalandcommunicativeagencyinhumaninfancy.
InJ.
B.
Benson,F.
Xu&T.
Kushnir(Eds.
)Rationalconstructivismincognitivedevelopment(pp.
59-94).
Waltham,MA:AcademicPress.
Godfrey-Smith,P.
(2006).
Mentalrepresentation,natural-ism,andteleosemantics.
InG.
MacDonald&D.
Papineau(Eds.
)Teleosemantics(pp.
42-68).
Oxford,UK:OxfordUniversityPress.
(2013).
Signals,icons,andbeliefs.
InD.
Ryder,J.
Kinsbury&K.
Williford(Eds.
)Millikanandhercrit-ics(pp.
41-58).
Oxford,UK:Blackwell.
Grice,H.
P.
(1957).
Meaning.
ThePhilosophicalReview,66(3),377-388.
10.
2307/2182440.
(1969).
Utterer'smeaningandintentions.
ThePhilosophicalReview,78(2),147-177.
10.
2307/2184179(1989).
Studiesinthewayofwords.
Cambridge,MA:HarvardUniversityPress.
Hare,B.
,Call,J.
&Tomasello,M.
(2001).
Dochimpan-zeesknowwhatconspecificsknowAnimalBehaviour,61(1),139-151.
10.
1006/anbe.
2000.
1518Hauser,M.
,Chomsky,N.
&Fitch,W.
T.
(2002).
Thefac-ultyoflanguage:whatisit,whohasit,andhowdiditevolveScience,298(5598),1569-1579.
10.
1126/science.
298.
5598.
1569Helming,K.
A,Strickland,B.
&Jacob,P.
(2014).
Makingsenseofearlyfalse-beliefunderstanding.
TrendsinCognitiveSciences,18(4),167-170.
10.
1016/j.
tics.
2014.
01.
005Jacob,P.
(1997).
Whatmindscando.
Cambridge,UK:CambridgeUniversityPress.
(2011).
Meaning,intentionalityandcommunica-tion.
InC.
Maierborn,K.
Heusinger&P.
Portner(Eds.
)Semantics:aninternationalhandbookofnaturallanguagemeaning(pp.
11-24).
Berlin,GER:WalterdeJacob,P.
(2015).
Millikan'sTeleosemanticsandCommunicativeAgency.
InT.
Metzinger&J.
M.
Windt(Eds).
OpenMIND:20(T).
FrankfurtamMain:MINDGroup.
doi:10.
15502/978395857016020|22www.
open-mind.
netGruyter.
(2012).
Sharingandascribinggoals.
MindandLanguage,27(2),200-227.
10.
1111/j.
1468-0017.
2012.
01441.
xKnudsen,B.
&Liszkowski,U.
(2012).
18-month-oldspre-dictspecificactionmistakesthroughattributionoffalsebelief,notignorance,andinterveneaccordingly.
Infancy,17,672-691.
Krebs,J.
R.
&Dawkins,R.
(1984).
Animalsignals:Mind-readingandmanipulation.
InJ.
R.
Krebs&N.
B.
Dav-ies(Eds.
)Behavioralecology(pp.
380-402).
Sunderland,MA:SinauerAssociates.
Leslie,A.
(2005).
Developmentalparallelsinunderstand-ingmindsandbodies.
TrendsinCognitiveSciences,9,459-462.
Leslie,A.
M.
,Friedman,O.
&German,T.
P.
(2004).
Coremechanismsin"theoryofmind".
TrendsinCog-nitiveSciences,8(12),528-533.
Leslie,A.
M.
,German,T.
P.
&Polizzi,P.
(2005).
Belief-desirereasoningasaprocessofselection.
CognitivePsychology,50,45-85.
Luo,Y.
(2011).
Do10-month-oldinfantsunderstandoth-ers'falsebeliefsCognition,121(3),289-298.
10.
1016/j.
cognition.
2011.
07.
011Luo,Y.
&Baillargeon,R.
(2005).
Canaself-propelledboxhaveagoalPsychologicalScience,16(8),601-608.
10.
1111/j.
1467-9280.
2005.
01582.
xMayr,E.
(1961).
Causeandeffectinbiology.
Science,134(3489),1501-1506.
10.
1126/science.
134.
3489.
1501Millikan,R.
G.
(1984).
Language,thoughtandotherbio-logicalcategories.
Cambridge,MA:MITPress.
(1989a).
Biosemantics.
TheJournalofPhilosophy,86(6),281-297.
(1989b).
Indefenseofproperfunctions.
Philo-sophyofScience,56(2),288-302.
(1990).
Seismographreadingfor"ExplainingBe-havior".
PhilosophyandPhenomenologicalResearch,50(4),807-812.
(1995).
Pushmi-PullyuRepresentations.
Philo-sophicalPerspectives,9,185-200.
(1998).
Conventionsmadesimple.
TheJournalofPhilosophy,95(4),161-180.
(2000).
Onclearandconfusedideas:Anessayaboutsubstanceconcepts.
Cambridge,UK:CambridgeUniversityPress.
(2004).
Varietiesofmeaning.
Cambridge,MA:MITPress.
(2005).
Language:Abiologicalmodel.
Oxford,UK:OxfordUniversityPress.
(2006).
ReplytoTomasello.
InM.
Mazzone(Ed.
)Symposiumonlanguage:AbiologicalmodelbyRuthMillikan,.
SWIFPhilosophyofMindReview.
(2011).
ReplytoNeander.
InD.
Ryder&K.
Willi-ford(Eds.
)MillikanandherCritics(pp.
37-41).
Ox-ford,UK:Blackwell.
Neander,K.
(1991).
Functionsasselectedeffects.
Philo-sophyofScience,58(2),168-184.
(1995).
Misrepresenting&malfunctioning.
Philo-sophicalStudies,79(2),109-141.
10.
1007/BF00989706(2004).
Teleologicaltheoriesofmentalcontent.
StanfordEncyclopediaofPhilosophy.
http://plato.
stanford.
edu/entries/content-teleological/(2007).
Biologicalapproachestomentalrepresent-ation.
InM.
Matthen&C.
Stephens(Eds.
)Handbookofthephilosophyofscience:Philosophyofbiology(pp.
548-565).
Elsevier.
(2011).
Towardsaninformationalsemantics.
InD.
Ryder,J.
Kinsbury&K.
Williford(Eds.
)Mil-likanandhercritics(pp.
21-36).
Oxford,UK:Blackwell.
Onishi,K.
&Baillargeon,R.
(2005).
Do15-month-oldin-fantsunderstandfalsebeliefsScience,308(5719),255-258.
10.
1126/science.
1107621Origgi,G.
&Sperber,D.
(2000).
Evolution,communica-tionandtheproperfunctionoflanguage.
InP.
Car-ruthers&A.
Chamberlain(Eds.
)Evolutionandthehumanmind:Language,modularityandsocialcogni-tion(pp.
140-169).
Cambridge,UK:CambridgeUniver-sityPress.
Perner,J.
&Ruffman,T.
(2005).
Infants'insightintothemind:HowdeepScience,308(5719),214-216.
10.
1126/science.
1111656Perry,J.
(1986).
Thoughtwithoutrepresentation.
Pro-ceedingsoftheAristotelianSociety,60(137),137-152.
Piaget,J.
(1954).
Theconstructionofrealityinthechild.
NewYork,NY:BasicBooks.
Pietroski,P.
(1992).
Intentionalityandteleologicalerror.
PacificPhilosophicalQuarterly,73(3),267-282.
Recanati,F.
(2002).
Doescommunicationrestoninfer-enceMindandLanguage,17(2),105-126.
10.
1111/1468-0017.
00191(2007).
Millikan'stheoryofsigns.
PhilosophyandPhenomenologicalResearch,75(3),674-681.
10.
1111/j.
1933-1592.
2007.
00103.
xScott,R.
M.
,Baillargeon,R.
,Song,H.
&Leslie,A.
(2010).
Attributingfalsebeliefsaboutnonobviouspropertiesat18months.
CognitivePsychology,61,366-395.
Jacob,P.
(2015).
Millikan'sTeleosemanticsandCommunicativeAgency.
InT.
Metzinger&J.
M.
Windt(Eds).
OpenMIND:20(T).
FrankfurtamMain:MINDGroup.
doi:10.
15502/978395857016021|22www.
open-mind.
netScott-Philipps,T.
C.
,Dickins,T.
E.
&West,S.
A.
(2011).
Evolutionarytheoryandtheultimate-proxim-atedistinctioninthehumanbehavioralsciences.
Per-spectivesonPsychologicalScience,6(1),38-47.
10.
1177/1745691610393528Senju,A.
,Southgate,V.
,Snape,C.
,Leonard,M.
&Csibra,G.
(2011).
Do18-mont-oldsreallyattributementalstatestoothersAcriticaltest.
PsychologicalScience,22(7),878-880.
10.
1177/0956797611411584Shea,N.
(2007).
Consumersneedinformation:supple-mentingteleosemanticswithaninputcondition.
Philo-sophyandPhenomenologicalResearch,75(3),404-435.
10.
1111/j.
1933-1592.
2007.
00082.
xSober,E.
(1984).
TheNatureofselection:Evolutionarytheoryinphilosophicalfocus.
Chicago,IL:ChicagoUni-versityPress.
Southgate,V.
,Senju,A.
&Csibra,G.
(2007).
Actionan-ticipationthroughattributionoffalsebeliefbytwo-year-olds.
PsychologicalScience,18(7),587-592.
10.
1111/j.
1467-9280.
2007.
01944.
xSouthgate,V.
,Chevallier,C.
&Csibra,G.
(2009).
Sensit-ivitytocommunicativerelevancetellsyoungchildrenwhattoimitate.
DevelopmentalScience,12(6),1013-1019.
10.
1111/j.
1467-7687.
2009.
00861.
x.
Sperber,D.
(2000).
Metarepresentationsinanevolution-aryperspective.
InD.
Sperber(Ed.
)Metarepresenta-tions:Amultidisciplinaryperspective(pp.
117-137).
Oxford,UK:OxfordUniversityPress.
(2001).
Anevolutionaryperspectiveontestimonyandargumentation.
PhilosophicalTopics,29(1/2),401-413.
10.
5840/philtopics2001291/215Sperber,D.
,Clément,F.
,Heintz,C.
,Mascaro,O.
,Mer-cier,H.
,Origgi,G.
&Wilson,D.
(2010).
Epistemicvi-gilance.
MindandLanguage,25(4),359-393.
10.
1111/j.
1468-0017.
2010.
01394.
xSperber,D.
&Wilson,D.
(1986).
Relevance,communica-tionandcognition.
Cambridge,MA:HarvardUniver-sityPress.
Tomasello,M.
(2006).
Conventionsareshared.
InM.
Mazzone(Ed.
)Symposiumonlanguage:AbiologicalmodelbyRuthMillikan.
SWIFPhilosophyofMindRe-view.
http://lgxserver.
uniba.
it/lei/mind/swifpmr.
htm(2008).
Originsofhumancommunication.
Cam-bridge,MA:MITPress.
Tomasello,M.
,Call,J.
&Hare,B.
(2003).
Chimpan-zeesunderstandpsychologicalstates–thequestioniswhichonesandtowhatextent.
TrendsinCognit-iveSciences,7(4),153-156.
10.
1016/S1364-6613(03)00035-4Tomasello,M.
,Carpenter,M.
,Call,J.
,Behne,T.
&Moll,H.
(2005).
Understandingandsharingintentions:Theoriginsofculturalcognition.
BehavioralandBrainSci-ences,28,675-691.
Topal,J.
,Gergely,G.
,Miklosi,A.
,Erdohegyi,A.
&Csibra,G.
(2008).
Infantperseverativeerrorsarein-ducedbypragmaticmisinterpretation.
Science,321(1269),1831-1834.
10.
1126/science.
1176960Warneken,F.
&Tomasello,M.
(2006).
Altruistichelpinginhumaninfantsandyoungchimpanzees.
Science,311(5765),1301-1303.
10.
1126/science.
112144(2007).
Helpingandcooperationat14monthsofage.
Infancy,11(3),271-294.
10.
1111/j.
1532-7078.
2007.
tb00227.
xWellman,H.
M.
,Cross,D.
&Watson,J.
(2001).
Meta-analysisoftheoryofminddevelopment:Thetruthaboutfalsebelief.
ChildDevelopment,72(3),655-684.
10.
1111/1467-8624.
00304Wilson,D.
&Sperber,D.
(2004).
Relevancetheory.
InL.
Horn&G.
Ward(Eds.
)Thehandbookofpragmatics.
Oxford,UK:Blackwell.
Wimmer,H.
&Perner,J.
(1983).
Beliefsaboutbeliefs:representationandconstrainingfunctionofwrongbe-liefsinyoungchildren'sunderstandingofdeception.
Cognition,13(1),103-128.
10.
1016/0010-0277(83)90004-5Woodward,A.
(1998).
Infantsselectivelyencodethegoalobjectofanactor'sreach.
Cognition,69(1),1-34.
10.
1016/s0010-0277(98)00058-4Wright,L.
(1973).
Functions.
ThePhilosophicalReview,82(2),139-168.
Jacob,P.
(2015).
Millikan'sTeleosemanticsandCommunicativeAgency.
InT.
Metzinger&J.
M.
Windt(Eds).
OpenMIND:20(T).
FrankfurtamMain:MINDGroup.
doi:10.
15502/978395857016022|22