Pragmatics in the False-Belief Task: Let the Robot Ask the ...

Experiment: Explicit False Belief Task in the Mentor-Child Context ThisarticleispartoftheResearchTopic TheoryofMindinRelationtoOtherCognitiveAbilities Viewall 8 Articles Articles HiromiTsuji OsakaShoinWomen'sUniversity,Japan LauraMacchi DepartmentofPsychology,UniversityofMilan-Bicocca,Italy FedericoManzi DepartmentofPsychology,CatholicUniversityoftheSacredHeart,Milan,Italy Theeditorandreviewers'affiliationsarethelatestprovidedontheirLoopresearchprofilesandmaynotreflecttheirsituationatthetimeofreview. Abstract 1.Introduction 2.TheAmbiguityoftheToMQuestion 3.ChangingtheContexttoDisambiguatetheToMQuestion 4.Experiment:ExplicitFalseBeliefTaskintheMentor-ChildContext 5.Discussion 6.Conclusion DataAvailabilityStatement EthicsStatement AuthorContributions Funding ConflictofInterest Acknowledgments Footnotes References SuggestaResearchTopic> DownloadArticle DownloadPDF ReadCube EPUB XML(NLM) Supplementary Material Exportcitation EndNote ReferenceManager SimpleTEXTfile BibTex totalviews ViewArticleImpact SuggestaResearchTopic> SHAREON OpenSupplementalData ORIGINALRESEARCHarticle Front.Psychol.,23November2020 |https://doi.org/10.3389/fpsyg.2020.593807 PragmaticsintheFalse-BeliefTask:LettheRobotAsktheQuestion! JeanBaratgin1,2*,MarionDubois-Sage1,2,BaptisteJacquet1,2,Jean-LouisStilgenbauer1,2,3andFrankJamet1,2,4 1LaboratoireCognitionHumaineetArtificielle,UniversitéParis8,Paris,France 2Probability,Assessment,ReasoningandInferencesStudies(P-A-R-I-S)Association,Paris,France 3FacultésLibresdePhilosophieetdePsychologie(IPC),Paris,France 4CYCergy-ParisUniversité,ESPEdeVersailles,Paris,France Thepoorperformancesoftypicallydevelopingchildrenyoungerthan4inthefirst-orderfalse-belieftask“Maxiandthechocolate”isanalyzedfromtheperspectiveofconversationalpragmatics.Anambiguousquestionaskedbyanadultexperimenter(perceivedasateacher)canreceivedifferentinterpretationsbasedonasearchforrelevance,bywhichchildrenaccordingtotheirageattributedifferentintentionstothequestioner,withinthelimitsoftheirownmeta-cognitiveknowledge.Theadultexperimentertellsthechildthefollowingstoryofobject-transfer:“Maxiputshischocolateintothegreencupboardbeforegoingouttoplay.Inhisabsence,hismothermovesthechocolatefromthegreencupboardtotheblueone.”ThechildmustthenpredictwhereMaxiwillpickupthechocolatewhenhereturns.Tothechild,thequestionfromanadult(aknowledgeableperson)mayseemsurprisingandcanbeunderstoodasaquestionofhisownknowledgeoftheworld,ratherthanonMaxi'smentalrepresentations.Inourstudy,withoutanymodificationoftheinitialtask,wedisambiguatethecontextofthequestionby(1)replacingtheadultexperimenterwithahumanoidrobotpresentedas“ignorant”and“slow”buttryingtolearnand(2)placingthechildintheroleofa“mentor”(theknowledgeableperson).Sixty-twotypicalchildrenof3years-oldcompletedthefirst-orderfalsebelieftask“Maxiandthechocolate,”eitherwithahumanorwitharobot.Resultsrevealedasignificantlyhighersuccessrateintherobotconditionthaninthehumancondition.Thus,youngchildrenseemtofailbecauseofthepragmaticdifficultyofthefirst-ordertask,whichcausesadifferenceofinterpretationbetweentheyoungchildandtheexperimenter. 1.Introduction Foralmost40years,theexplicitquestioninfalsebelieftasks(FBT)ofWimmerandPerner(1983),inwhichthechildmustexpressthefalsebeliefofacharacteronthestateoftheworld,hasbeenthecommonlyacceptedtasktostudytheTheoryofMind(ToM).Understandingthefalsebeliefsofothersisofconsiderableimportanceforthecognitiveandsocialdevelopmentofchildren.Itisrequiredtograspthatothershavementalstates,subjectiverepresentationsconditionedtospecificknowledgeandexperiences,distinctfromours.Thus,understandingthatbeliefscanbedifferentfromonepersontoanother(Perner,1991).SabbaghandBowman(2018)highlightthatexplicitFBTareasimpletestparadigmperfectlyrepresentativeofthisunderstanding.Inthesetasks,childrenmustrecognizethatsomeoneelsewillbehaveinawaythatdoesnotcorrespondtohowtheyunderstandthestateoftheworld. ExplicitFBTrequireadirectverbalanswertoanexplicitquestionoftheexperimenter.Theexpectedanswerseemstobeveryintuitiveandistraditionallyconsideredtobeareliableindicatoroftheunderstandingoffalsebeliefs.TheexplicitFBTofWimmerandPerner(1983)isthefollowingtask:Theexperimentertellsthechildparticipantastoryofobjecttransferthroughtheuseofclips1:Beforegoingouttoplay,thechildMaxiputshischocolateintothegreencupboard.Whileheisoutside,hismothermovesthechocolateandputsitintothebluecupboard.Maxithencomesbacktogethischocolate(seeFigure1)2. FIGURE1 Figure1.Thestory“MaxiandChocolate”ofWimmerandPerner(1983)inclips(takenfromDuvaletal.,2011,p.45).Leftclip:Maxicomeshomefromshoppingwithhismother,andputsthechocolateintothegreencupboardbeforegoingoutsidetoplay.Middleclip:WhileMaxiisgone,Maxi'smothertakesthechocolatefromthegreencupboardtomakeacakeandputsitbackintothebluecupboard.Rightclip:Maxicomeshomeforasnack.Hestillrememberswhereheputthechocolate. ThechildmustpredictwhichcupboardMaxiwillopentotrytogethischocolate.Togetthechild'sanswer,theexperimenterasksthefollowingtestquestion(ToMquestion):“WherewillMaxilookforthechocolate?”Inthisquestion,childrenareinvitedtoindicatethatMaxiwilllookforthechocolatewherehebelievesitis(i.e.,whereheleftit)insteadofwherethechildrenknowitreallyis.Toanswercorrectly(greencupboard),thechildmustactivateintheirmindthefalsebeliefofthecharacterMaxi,whodoesn'tknowthechocolatehasbeenmoved,whileinhibitingtheirownknowledgeoftheworld(thechocolateisinthebluecupboard).AcontrolquestionisthenaskedbytheexperimenterfollowingtheToMquestiontomakesurethechildunderstoodthestory.IfthechildanswerscorrectlytotheToMquestion,a“realityquestion”isthenaskedregardingthetruelocationofthechocolateattheendofthestory:“Whereisthechocolatereally?”IfthechildinsteadfailstoanswertheToMquestion,thenextquestionisthena“memoryquestion”toseeiftheyrememberwherethechocolatewasatthebeginningofthestory:“DoyourememberwhereMaxiputthechocolateinthebeginning?”Resultsofnumerousstudiesdonewithneuro-typicalchildrenofvariouscultures(Callaghanetal.,2005)indicatethatthemajorityof4years-oldchildrenanswerthebluecupboardtotheToMquestion(wherethechocolateactuallyis).Itisnecessarytowait4–5yearstoseechildrenansweringcorrectlythatMaxiwilllookintothegreencupboard(WimmerandPerner,1983;Baron-Cohenetal.,1985;Wellmanetal.,2001;SabbaghandBowman,2018).Theexplanationbeingthatchildrenbetween3and5learnconceptualknowledgenecessarytomakeexplicitdecisionsabouttherepresentativementalstateofothers3. Yet,theseresultsseemtobeincontradictionwithbehaviorobservedin3years-oldchildrenrequiringfirstorderabilities,suchasthegameofHideandSeek.Inthisgame,thechildmustgosomewheretheywillnotbeseenbyothers.Tosucceedthechildmustunderstandthedifferencebetweentheirknowledgeandwhatotherswillperceive.Childrenyoungerthan4areabletoevaluatewhatcanbeperceivedbyothers,andthustoadoptapointofviewdifferentfromtheirown(Shatzetal.,1983;Reddy,1991,2007,2008;BartschandWellman,1995).ThefirstorderToMthenseemstobeanabilityacquiredbeforetheageof4(Baillargeonetal.,2010;WestraandCarruthers,2017).Halaetal.(1991)showthatchildrenwhofailedanexplicitFBT,inanecologicalsituation,canunderstandandusethefalsebeliefstoexplainthementalstateoftheprotagonistofthestory.Thereasonsofthesystematicfailureof3years-oldchildrenwouldbe,fortheseauthors,duetothespecificityoftheexplicitFBT.Thechildmustgiveconsciousanddeclarativeanswerstothequestionsoftheexperimenter.Explicittaskswithverbalanswerswouldrequireimportantcognitiveresources.Thesetaskswouldgreatlyinvolveexecutivefunctions;suchastheabilityofthechildtoinhibittheirownpointofviewtoconsiderthatofothers.Theseexecutivefunctionswouldstillbeimmatureattheageof4(Leslie,2005;Baillargeonetal.,2010;WestraandCarruthers,2017;Oktay-Güretal.,2018,foradiscussion).InimplicitFBT,inwhichtheanswersofchildrenarededucedfromactionsorgazesandnotfromexplicitpointingorlinguisticreplies,amuchmoreprecocioussuccess(startingfrom15monthsold)isobserved(OnishiandBaillargeon,2005;Southgateetal.,2007;Surianetal.,2007;Baillargeonetal.,2010;Scottetal.,2010;Heyes,2014). Inconsequence,thereisa“developmentalparadoxoftheunderstandingoffalsebeliefs”(DeBruinandNewen,2012;NewenandWolf,inpress):toddlerssucceedinimplicitFBTusingbehavioralresponsesbutkidsbelowtheageof4generallyfailtheexplicitFBTinwhichtheymustexplicitlyanswertheexperimenter'squestion.Somefollowinga“nativist”approachargueinfavorofanearlyabilitytodetectfalsebeliefs(basedonaninnatemodule)allowingtoddlerstosucceedinimplicitFBT(Leslieetal.,2004).Othersfollowingamore“empiricist”approacharguethattheabilitytounderstandfalsebeliefsisduetothedevelopmentofcognitiveabilities.ItisthatdevelopmentwhichisresponsibleforthechangeofperformanceinexplicitFBTattheageof4(seeNewenandWolf,inpress,forarecentreview).NewenandWolf(inpress)pointoutadistinctiondividingbothnativistsandempiricistsintothosewhogiveacognitiveexplanationandthosewhogiveasituationalexplanationtothefailureofchildren.Fortheformer,explicitFBTwouldbedifficultbecausethecorrectanswerwouldrequirecognitiveresourcesnotyetdevelopedforchildrenbetween3and4.Forthelatter,thefailureinexplicitFBTwouldactuallybetheresultoftheprocedureitself;whichwouldbeasourceofthemisunderstandingofthequestionforthesechildren.Ourstudyfocusesonthissituationalexplanation(andinparticularthepragmaticexplanation)ofthefailureoftoddlersinexplicitFBT.WesuggestanewprocedureabletocanceloutsituationalfactorswithoutmodifyingthestructureofexplicitFBTthemselves.Still,webelievethatthesituationalexplanationisprofoundlycognitiveaswellastheRelevanceTheory(SperberandWilson,1986)weusetoexplaintheinfluenceofthesituationisafundamentallycognitivetheory. Helmingetal.(2014,2016),Westra(2017),andWestraandCarruthers(2017)allconsiderthefailureofchildrenyoungerthan4tobecausedbyadefectiveunderstandingoftheexpectationsoftheexperimenterinthequestion.ThecorrectinterpretationoftheToMquestionwouldrequireacognitiveefforttoogreatforchildrenofthatage.Furthermore,sincediscussionsonbeliefsarenotcommon,childrenwouldsystematicallyinterprettheexpectationsoftheexperimentertobeabouttestingthechild'sknowledgeaboutthestateoftheworld(i.e.,indicatingwherethechocolatereallyis)comparedtothebeliefsofafictivecharacter.ThisincorrectinterpretationoftheToMquestionwouldbecausedbytheconversationalcontext:theattributionofthestatusofteachertotheexperimenter,andtheirownstatusofpupil.Wesuggesttransformingthiscontext(1)Byswitchingtherolesandspecificstatusesoftheexperimenterandofthechildparticipantand(2)Byreplacingtheexperimenterwithan“ignorantandslowentity.”Thiscontext,wecall“mentor-child,”disambiguatestheToMquestionaskedbytheignorantentitybymakingitclearthatitexpectstounderstandthefalsebeliefofMaxi.Todothis,wereplacetheexperimenterwithahumanoidrobotNAO.Thisworkwillbeorganizedinthefollowingway:Afterrecallingtheobligationtoconsiderthepragmaticimplicaturesinallactsofcommunication,wewillexposethosedrivingthechildtoproduceanincorrectanswerintheexplicitFBT.Wewillthenexplainanewproceduretodiminishtheambiguityofthequestions.Afterdescribingtheresults,wewilldiscussthemandconcludewiththesuggestionoffutureareasofresearch. 2.TheAmbiguityoftheToMQuestion SperberandWilson(1986,2002)haveshownthatallcommunicationisinevitablyofapragmaticnature.Acommunicatorperformsinaway,suchasproducingaspeechactoragesture,andthereceivingaudiencemustunderstandtheintenthiddenbeneaththesurface.Itisespeciallyimportanttounderstandthatmostoftheexperimentalparadigmsincognition,socialcognitionanddevelopmentalcognitioncorrespondtoanactofcommunicationbetweenanexperimenterandparticipants.Therearemanyexamplesinthepsychologicalliteraturethatanswersgiven,consideredtobeincorrectbytheexperimenter,byadultparticipantsareactuallytheresultoftheparticipants'misunderstandingoftheintentionsoftheexperimenter.Theutterancesusedandthecontextoftheexperimentaltasktriggerimplicaturesintheparticipantsthatcaninduceanswersthataredifferentfromthoseexpectedbytheexperimenter(seeDulanyandHilton,1991;Sperberetal.,1995;BaratginandNoveck,2000;Macchi,2000;PolitzerandMacchi,2000;Baratgin,2002,2009;BagassiandMacchi,2006;BaratginandPolitzer,2006,2007,2010;MacchiandBagassi,2012;Macchietal.,2019,2020,forexamples).Manydevelopmentalstudiesalsogivepiecesofevidencefortheabilityofchildren,giventheirage,torecognizetheintentionsofthecommunicator(seeBraineandShanks,1965a,b;McGarrigleandDonaldson,1974;RoseandBlank,1974;MarkmanandWachtel,1988;Politzer,1993,2004,2016;GelmanandBloom,2000;DiesendruckandMarkson,2001;Bagassietal.,2020,forexamples). Sperber(1994)suggeststhatthechildusesthesimplestprocedureofinterpretationwhichconsistsininferringfromthecommunicativestimulusthemostrelevantintentioninrelationtotheirownpointofview.However,whatisrelevantforthechildmaybedifferentfromwhattheexperimenteractuallyintendstocommunicate.Thus,byanalyzingtheexperimentaltaskofPiagetandSzeminska(1941)ontheclassinclusionquestion,Politzer(1993,2004,2016)hasshownthattheperformancesofchildreninrelationtotheiragecouldbeexplainedbythedifferencesoftheirinterpretationofthequestion.Theexperimentershowedfiveastersandthreetulips.Thechildwasthenaskedwhether“therearemoreastersormoreflowers.”Thetypicalanswerofchildrenunder8is“Therearemoreasters.”Politzerdemonstratesthatthequestioncanbecharacterizedbyanambiguityattherootoftheresponseoftheyoungestchildren.Whilethequestionofclassinclusionisenunciated,accordingtotherelevanceprinciple(SperberandWilson,1986),childrenwilltrytoinfertheexpectationsoftheexperimenterandtoadapttheiranswersothatitfeelsrelevanttothem.Questionsarerelevantwhentheymakethepersontowhomtheyareaskedanswerinarelevantway(i.e.,questionsthatrequiretheleastcognitivecostforthemostcontextualeffect).Theseassumptionsdependontherepresentationalattributionsofthechildfortheexperimenterwhichareafunctionoftheirdevelopment(Hayes,1972).AccordingtoPolitzer,youngchildrendonotmakemistakesofclassinclusion.Theysimplyhaveadifferentrepresentationofthequestion,makingthemgiveanincorrectanswer. “Thisisafundamentalinsight.Oncethisviewisadopted,thedisambiguationofthequestionmustbeenvisagedinrelationtothechild'sdevelopment.Fromthenotionthatthechildrenattempttorenderthequestionoptimallyrelevantitfollowsthatthewaytheydosowillvarywiththeircognitivedevelopment.Inotherwords,theinterpretationchosenbythechildrenisconstrainedbytheirlevelofdevelopment.Therefore,theinterpretationcanbepredictedbasedonwhatislikelytobethechildren'sestimationoftherelevanceofthequestion”(Politzer,2016,p.3). Politzerobservedthatwhenhedisambiguatedthequestionofclassinclusionthesuccessofparticipantswassignificantlyimprovedandcameearlier:between5and6years-old(seealsoJametetal.,2018). Itisthenlegitimatetowonderif,likewiththequestionofclassinclusion,theincorrectreplygivenbyyoungchildreninexplicitFBTcouldalsobetheresultofadifferentinterpretationoftheToMquestionwhichwouldbecausedbyanincorrectinferenceoftheexperimenter'sexpectations.Withtheyears,andwiththeacquisitionofthepragmaticskillsofthechild,theambiguityofthequestionwouldlaterdecrease.ThispragmatichypothesiscouldexplaintheearlysuccessinimplicitFBT,whicharesimplificationsofexplicitFBTinwhichtheToMquestionisnotexplicitlyasked.Tosucceedinthesetests,thechilddoesnotneedtocorrectlyinterpretthequestionortocorrectlyinfertheintentionoftheexperimenter.Theyonlyneedtounderstandthefalsebeliefs.ForSiegalandBeattie(1991);Westra(2017);WestraandCarruthers(2017),sincethebeginningoftheirdevelopment,youngchildrencancreaterepresentationsofothers'beliefsandunderstandthefalsebeliefs.However,3years-oldchildrendonotexpectbeliefstobealikelytopicofconversation(Westra,2017).Itisdifficultforthemtoinducethatfactsrelativetosomeone'sbeliefscanbearelevanttopicintheconversationwiththeexperimenterandthatthisiswhatthequestionisabout.Despitethefactthatyoungchildrenconstantlyattributepropositionalattitudestootheragents,understandingwhenthesepre-linguisticconceptsplayapartintheconversationisnotonlyaquestionofacquisitionoftheadequatevocabularybutwouldalsobeaquestionofthedevelopmentofpragmaticskills(WestraandCarruthers,2017).Thechildmustbeexposedtoconversationsforthesesocialstimulitoplayacrucialpartinthestrengtheningoftheirlinguisticandpragmaticskills(AstingtonandOlson,1995;CarpendaleandLewis,2004;Antoniettietal.,2006;Westra,2017)4. ThislackofpragmaticskillisevenmoresalientinexplicitFBTastheconversationalinteractionhappensbetweenthechildandastranger(theexperimenter).Attheageof3,eveniftheyoungchildhashadnumerousinteractionswiththeirparentsandfamily,interactionswithadultsaregenerallylimited,exceptfortheteacherwhichisformoststillarecentinteraction(3isusuallytheageatwhichchildrenstartschool).Theteacheriscertainlyanimportantreferencefortheyoungchildduringtheexperiment.After2monthsofclass,preschoolchildrenhaveintegratedthedidacticcontractwantedbytheteacher.Theteacherexplicitlyinvitesthepupilstoworkwell,toshoweverythingtheyknow.Eachtimethechildreturnstoclass,aftercompletinganactivity,theteacherwillaskthemiftheyworkedwell.AsWestraandCarruthers(2017)explained,childrenarereadilyabletoconsiderthattheinteractionwiththeexperimenterhasaneducationalintention.Indeed,educationalcluesarealmostalwayspresentinanexplicitFBT.Theexperimenter,forthechild,isinasocialpositionmuchsuperiortotheirsand,justliketheirteacher,hastheencyclopedicknowledge.Theexperimenter-childrelationshipreinforcesthisimpressionofsuperioritysincetheexperimenterisintroducedtothechildasanauthorityfiguretowhomtheymustobey.Thisattributionofteacherisfacilitatedevenmorebythefactthattheexperimentsmostoftenhappenatschool,duringschooltime.Thissuppositionofaneducationalintentioninthetaskimplies,forthechild,thataneducationalbehaviorisexpectedofthem,asitisusuallythecaseinthiscontext.Therefore,theyareinapositionofpupilduringtheexperiment.Howuncommonthesituationis,anadultreplacingtheteacherforaneducationalexercise,canstrengthentheideathatthisexerciseisreallyimportantandthatthisnewteachermaybespecialandknowsmorethantheusualteacher.Thisattributionisalltheeasiersincetheexperimenterisoftenpresentedasaresearcher,aspecialist.Preschoolchildrenindeedseemtobealreadysensitivetotheknowledgeoftheinformantineducationalactivities(JeongandFrye,2018b). Whenteachinganewconceptinanexampleorastory,theteacherlaterchecksthechildunderstoodcorrectlythroughsimpleanddirectquestionslinkedtowhatwasjusttold.Thesequestionsareveryrarelyambiguous.Thecorrectanswerexpectedbytheteacherisusuallymeanttoprovethattheyunderstoodthestorycorrectly.Thus,tothechild,thesamecanbeexpectedofthequestionsaskedbytheexperimenter.ThemaindifficultyinexplicitFBTisthefactthattheyinvolvefourdifferentelementsofknowledge:(1)WhereMaxiinitiallyputthechocolate,(2)ThechangeoflocationdonebyMaxi'smother,(3)ThefactthatthischangeoflocationhappenedinMaxi'sabsence,and(4)ThefactthatMaxiislookingforhischocolate,probablyinthewrongplace.Forthechild,therearemultiplepossibleinterpretationsoftheexperimenter'sexpectationswhentheyasktheToMquestion.Theycanbe:tryingtoassesswhetherthechildunderstoodthechangeoflocationofthechocolate(steps1and2),orassesswhetherthechildunderstoodthefactthatMaxiwasnotthereduringthechangeoflocation,andthatinconsequencehewilllookforthechocolateinthewrongplace:theinitiallocation(steps3and4).Alongtheseinterpretations,theonewhichconcernstheattributionofbeliefstosomeoneelsehasagreatercognitivecostforyoungchildren.Theyaregenerallynotexperiencedenoughininteractingwithadultstograsptherelevanceofthisexpectation.Childrenof3years-oldwillinsteadusethemorefamiliarinterpretation:theywillthinkthattheexperimenterexpectsthereplytobeaboutthechild'sunderstandingofthechangeoflocation(SiegalandBeattie,1991;Hansen,2010;Lewisetal.,2012;Westra,2017). Helmingetal.(2014,2016)offeramoreelaboratepragmaticexplanationofyoungchildren'sanswer.Forthem,explicitFBTforcethechildrentoadopttwopointsofviewsatthesametime.Oneismoredetached:“spectating”inthethirdpersontheactionofthemaincharacterofthestory,inparticularfocusingonthecharacter'sbeliefs;andtheotherismorecommunicative:interactingwiththeexperimenterinthesecondperson.Thisfirstpointofviewbeingdisruptedbythesecond.TheToMquestionthengeneratestwobiases:one“referential”andone“cooperative.”ChildrenhavethepossibilityofmentallyrepresentingthereallocationofthechocolateorwhereMaxiwronglybelievesitis.Usingthewordchocolateinthequestioncanbiaschildrentowardansweringwiththereallocation(referentialbias).Theinteractionwiththeexperimenterwouldbringthechildtofocusontheknowledgetheyshare(i.e.,thereallocationoftheitem).Thiswouldthendisrupttheabilityofthechildtotrackthefalsebeliefofthemaincharacterfromthethirdpersonpointofview.Inessence:whentheexperimenterreferstothetargetitem,theydirecttheattentiontowardthereallocation.Thecooperativebiasistheresultofthetendencyoftoddlerstowanttomakethemselvesusefulbyspontaneouslyhelpingothers(evenadultstrangers)toreachtheirgoals,evenifitrequiresagreatereffortandiftheyarebusywithataskoftheirown(seeWarnekenandTomasello,2007,2009,2013;Liszkowskietal.,2008;Buttelmannetal.,2009,2014;Warneken,2015).Thishelpfulnessseemstobemainlymotivatedbyanintrinsiccarefortheotherandnotforanypersonalreward(Hepachetal.,2012,2016).ThistendencytohelpothersmadeitpossibletocreateimplicitFBT.Thetaskgiventotoddlersconsistedinhelpinganadultreachtheirgoal.Yettoinferthisgoal,thetoddlersneededtoconsiderwhattheadultbelieved.Thistendencywoulddrivechildrentoadoptasecondpersonpointofviewtowardthemaincharacterofthestory,ratherthanaspectatingpointofviewinthethirdperson,thisinturndrivingthemtoincorrectlyinterprettheexpectationsoftheexperimenter.Childrenunderstandthatthemaincharacterneedshelp,becausehehasfalsebeliefs,toavoidpickingthewronglocation.Theyspontaneouslywanttohelphimbytellinghimthecorrectlocationandcanreadilyexpecttobeinvitedtodoso.This,forthechild,wouldstrengthentheinterpretationoftheToMquestion“WherewillMaxilookforhischocolate?”asaninvitationtohelpthemaincharacterfindtheitem.Thismeansinterpretingthequestionasanormativequestion“Whereshouldhelookforhischocolate?”oreven“CanyoutellMaxiwheretofindhischocolate?”AsNewenandWolf(inpress)pointout,thispragmaticexplanationisnotincontradictionwiththecognitiveexplanation(intermsof“mentalfiles”byRecanati,2012)suggestedbyPerneretal.(2015),PernerandLeahy(2016),andHuemeretal.(2018).Thesementalfiles,ormentalrepresentations,includethe“informationmanagementtoolsaboutanobjectintheworld”andthelinksbetweenthedifferentfileswhichmakeitpossibletoshareinformationbetweenthem.In“Maxiandthechocolate,”thechildhastwomentalfilesofthesituation:one“regular”filewiththeinformationthatthechocolateisinthebluecupboard,andone“indirectbyproxy”fileindexedonMaxiwiththeinformationthatthechocolateisinthegreencupboard.AccordingtoPernerandLeahy(2016),whenchildrenbelowtheageof4arefacedwiththeToMquestion,theyarenotyetabletoswitchbetweentheindirectmentalfileandtheregularmentalfileinacontrolledandsystematicway.ItisonlyoncethementalfilesarelinkedthatthechildcanaccesstheinformationaboutMaxi'sbeliefs.Thepragmaticexplanation,throughtheRelevanceTheory,allowsustounderstandwhichmentalfilewillbeactivated.Inatraditionalcontext,thementalfilewhichhastheleastcognitivecostandthegreatestcontextualeffectistheregularmentalfilewhichanswerswhatthechildbelievestobetheexperimenter'sexpectation. Thus,asWestraandCarruthers(2017)pointedout,therearetwointerpretationsatstakeinadditiontothecorrectinterpretationoftheToMquestionforatotalofthreepossibleinterpretations:(1)The“helpfulness-interpretation”wherethequestioncorrespondstoaninvitationtohelpthecharacter,(2)The“knowledge-exhibiting-interpretation”wherethequestioncorrespondstoaninvitationtoshowone'sknowledgeoftheeventsinthestory(steps1and2asdescribedinthepreviousparagraphs),and(3)The“psychologicalknowledge-exhibiting-interpretation”wherethequestioncorrespondstoaninvitationtoreportthecharacter'sfalsebeliefsaboutthelocationoftheobject(steps3and4)5.Thechild'staskistodeterminewhichofthesethreecompetinginterpretationsismostlikelytomeettheexperimenter's.Interpretation(3)istheoneexpectedbytheexperimenter.Eachoftheothertwoleadstotheincorrectanswerofindicatingtheactuallocationofthechocolate.Asindicatedabovetoddlersdonotyethavethepragmaticexperiencerequiredtounderstandthatpeople'sbeliefsareavalidtopicofconversation.Inconsequence,theyaremoreinclinedtointerprettheToMquestionasakindofindirectlanguageacttoverifytheirknowledgeofthereallocationofthechocolate(interpretation1).Thiswillalsohelpthecharacterfindthechocolate(interpretation2).Aschildrengainexperienceindiscourseaboutthebeliefsofothers,theybegintobeabletorecognizethetruepurposeofthequestionandtheirtrueexpectation(interpretation3):reportingexplicitlythefalsebeliefofthecharactercalledMaxi(WestraandCarruthers,2017;Frank,2018).Theythenunderstandthatthequestion“WherewillMaxilookforthechocolate?”implicitlymeans“WhatdoesMaxifalselythinkaboutthelocationofthechocolate?” AnumberofauthorshavetriedtodirectlydisambiguatetheToMquestion.SiegalandBeattie(1991)givethefollowingquestion[reformulatedtofitMaxiandtheChocolate]:“WherewillMaxilookforthechocolatefirst?”whichdirectlyexplainstheexperimenter'sexpectation.Theauthorsobserveasignificantincreaseincorrectanswers(Yazdietal.,2006;Białecka-Pikuletal.,2019).Hansen(2010)alsoobservesmuchbetterresultswhentheexperimenterdirectlyspecifiesintheirquestionthattheyarenotinterestedinthechild'sknowledgeofthestateoftheworld[reformulatedtofitMaxiandtheChocolate]:“YouandIknowwhereMaxi'schocolateis,butwheredoeshethinkitis?” Anothersolutionistoexplicitlyandconceptuallyexplaintheimportantcluesinthestorytomakethecorrectinterpretation(3)oftheToMquestionmoreconceptuallyrelevant(NewenandWolf,inpress),forexamplebymakingthefalsebeliefofthemaincharactermoresalient.MitchellandLacohée(1991)noticedthatchildrenparticipatinginexplicitFBTwhokeptanexplicitaide-memoireoftheirpriorbelief(thecupboardwherethechocolatewas[reformulatedtofit“MaxiandtheChocolate”])wasmuchmoresuccessfulatavoidingalaterdeformationofthisbelief.Lewisetal.(2012)showedthattheexplanationofthefalsebeliefsofanotherpersonisimprovedifweaddanothercharactertothestorywhoisalsoobservingobject'schangeoflocation.Thepresenceoftheotherpersonconceptuallyhighlightsthepossiblepointofviewsinthestory.InthissituationtheToMquestion,beingexplicitlydirectedatthecharacterwhodidnotseethechangeoflocation,increasestherelevanceofinterpretation(3)onthefalsebeliefsofthecharacter.Rubio-FernándezandGeurts(2013,2016)demonstratedthattoddlerscanalsosucceedinexplicitFBTifthetaskismodifiedinsuchawaythat,first,thepointofviewoftheotherpersonisfrequentlyrepeatedtothechildduringtheexperimentand,second,theToMquestionaskedtothechildistransformedinto“Whathappensnext?”Herethedisruptioninducedbytheexperimenterfocusingontheitemisnolongerpossible.Itisalsopossibletomakeinterpretation(2),ofexposingthechild'sknowledgeaboutthereallocationoftheitem,lesscontextuallyrelevant.WellmanandBartsch(1988),MascaroandMorin(2015),andMascaroetal.(2017)indeednoticebetterperformanceswhenthechildrenthemselvesdonotknowwheretheactuallocationoftheitemisoriftheitemisremovedfromthescene. Finally,itispossibletochangetheexperimentalproceduretomakethespontaneoustendencyofchildrentobeuseful,whichusuallydriveschildrentowardthe“helpfulness-interpretation”(1),tobecomeanindicatoroftheeffectivefalsebeliefofthecharacter.MatsuiandMiura(2008)showedthattoddlerssucceededmoreeasilywhenthetaskwaschangedtohavethemchooseacharacterwhomtheyhadtohelpfindtheitem(pro-socialcontext). Tosumup:whetherchildrencandisambiguatetheToMquestiondependsontheirmeta-cognitivedevelopment.Toddlersmakethequestionmorerelevantbyinterpretingitasaquestionabouttheirknowledgeofthestoryor,withthesameresult,aquestionabouttheirknowledgeofthestorytohelpthemaincharacter.Olderchildreninterpretitcorrectlytobearequestforthemtoreporttheirknowledgeofthefalsebeliefofthecharacterinthestory. 3.ChangingtheContexttoDisambiguatetheToMQuestion Inalltheseexperimentstheoriginaltaskismodified.TheToMquestionissometimesmodified,theparticipantissometimesaskedtokeepinmemorytheinitialbelief,acharacterissometimesaddedorsomeinformationissometimesremoved.Ourobjectiveistodecreasethesalienceofincorrectinterpretationswithoutchangingneitherthestorynorthequestionasked:byplayingwiththeglobalcontextoftheexperimentitself.Agoodexampleisthelengthandnumberconservationtask(PiagetandSzeminska,1941).Assessingtheconservationofnumberisdonebypresentingtwolinesoftokens,equalinnumberandarrangedinaone-to-onecorrespondence,infrontofachildwhojudgesthemtobethesame.Whentheexperimenterrearrangesoneoftherowsthenon-conservingchildchangestheirjudgmentinfavorofthelongerrow.McGarrigleandDonaldson(1974)showedthatwhenthetransformationoftherowoftokensistheindirectresultofanactionwithadifferentgoal,suchasatransformationeffectedbya“naughtyteddybear”whowantsto“spoilthegame,”childrenaremoreconservant.Inthis“accidentaltransformation,”therearenostructuralmodificationofthetask. Asexplainedabove,thewaythechildinterpretsthequestionsoftheexperimenterinexplicitFBTdepends,inpart,ontheirunderstandingofthenatureofthecommunicativeexchange(i.e.,itstopicandgoal).Fortoddlers,thecontextofthetask,asshownabove,stronglyexpressesthatofaschoolactivitywiththestatusoftheexperimenter-teacher,abletojudge,andthelocation.Thus,thechildinferseffortlesslytheirroleinthistaskwillbetheonetheyalreadyknowandareusedtoduringclasses:thatofapupilwiththegoaloflearningandshowtheirknowledge.Theseassumptionsmadebythechildfortheexperimentertobetestingtheirknowledgearetheoriginofinterpretations(2)and(3).The“helpfulness-interpretation”(1)canbeconsideredtobethedesiredexpectationinordertohelpthecharacterinthestory(numerousstudiescitedaboveindicatehowspontaneously,andwithoutulteriormotives,thetoddlerdisplaysanaltruisticbehavior).Thus,ifwehadanexperimentalcontextinwhichexposingtheknowledgeofthefalsebeliefsofthecharacter(interpretation3)couldalsosatisfya“helpfulness-interpretation”(interpretation1),theninterpretation(2),abouttheactualstateoftheworlddescribedinthestory,couldbeinhibited. 3.1.AMentor-ChildandanIgnorant,Naive,andSlowPupil TodothiswemustconsiderasituationwhichwouldchangetheassumptionsofthechildaboutthepersonaskingtheToMquestion;asituationinwhichthechildcouldspontaneouslyinferthatananswerindicatingthefalsebeliefofthecharacterwouldhelpthepersonaskingthisquestion6.AfirstmodificationofthecontextwouldthenhavethepersonaskingtheToMquestiondisplayanexplicitneedtoknowthefalsebeliefofthecharacter.Thepersonwouldhavetroubleunderstandingthestory,asforthemtheanswertotheToMquestionisfarfromobvious,eveniftheyaskedit.Thispersonmusthavelessknowledgethanthechildandmustconsiderthechildtobesomeonewhoknowsmore.Thus,wemustconsideracontextinwhichthestatusofthechildandoftheexperimenterareswitchedcomparedtotheoriginalcontext. Wecanimaginea“mentor-child”contextinwhichtheyoungchildmustanswerthequestionsofanignorantentityintroducedbyanauthorityfigure:“Youaretheteacherandthisisyourpupil.Itdoesn'tknowmuch.Itneedsyou7.”Intheconversationalact,theexpectationofthechildregardingthequestionsoftheentityistobeabletohelpitlearnnewthings.Letusimaginethatthisignorantbeingtellsthechild:“IwastoldastorythatIdidn'tunderstandverywell.I'lltellittoyouandthenpleaseexplainittome.”Aftertellingthestory,theignorantentityaskstheToMquestioninanaivetone.Thechildansweringcorrectlyshowstheirknowledgewhilehelpingtheentity.Thequestionhereisdisambiguatedandreliablydrivesthechildtowardinterpretation(3).Thequestionaskedinthiscontextbecomesnatural,forthechildknowstheentitytobeignorantandthatitcanasktrivialquestions.Thisisnotthecaseinthetraditionalcontextwhereitcanseemsurprisingthata“knowing”adultcouldasksuchaquestion. Anotherimportantaspectistohighlightthe“naive,”“unsureofitself,”and“slow”traitsoftheignorantentity.Thisaspecthelpsthechildconsiderthemselvesknowledgeablecomparedtoit.Italsohelpsthechildfeelusefulwhenhelpingit.Moreimportantly,the“slow”aspectoftheentitycanfavortheinterpretationofthecontrolquestionsaskeddependingonthesuccessorfailureintheToMquestion(respectivelytherealityquestionandthememoryquestion).Toourknowledge,thepragmaticanalysisofthesequestionshasneverbeenexplicitlydoneintheliterature.ThiscancertainlybeexplainedbythefactthatinWimmerandPerner(1983),allchildrensucceedingintheToMquestionalsocorrectlyansweredtherealityquestion8.Inthestandardcontext,theinterpretationoftherealityquestionisindeedcompletelyobviousforthechild.Itcorrespondstothequestionthatisthemostexpected;whichhasthestrongestcontextualeffectandrequirestheleastcognitivecosttoanswerit.AfterindicatingthefalsebeliefofthecharacterintheToMquestion,therealityquestionmakesitevenmoreexplicitbyindicatingwheretheitemactuallyis.Thissecondquestiondoesnotseemtobeincongruousinthestandardcontextinwhichthechildassignsthestatusofteachertotheexperimenter.Ateacheroftenasksmultiplequestionstotesttheknowledgeofthechild.Inour“mentor-child”context,therealityquestionaskedbytheignorantentitycanseemtobeabitoddtothechildintheirroleofteacher.Indeed,askingthissecondquestionrequiresunderstandingthatthechocolateisinadifferentplacethanwherethecharacterbelievesitis9.Thus,theignorantentity,ifitdidunderstandcorrectlytheanswergivenbythe“mentor-child”totheToMquestion,shouldhavealsounderstoodthatMaxihasafalsebeliefandwilllookintothegreencupboardwhichisnowempty.Frequently,whenapupilasksasecondquestiontotheteacherjustafterreceivingananswertoanother,itisoftenbecausetheyneedmoreprecisionorbecausetheydidnotunderstandtheanswer.Inthiscase,therearetwopossibleanswersforthe“mentor-child”:(1)ThinkingthattheywerenotclearenoughwiththeirfirstanswerandbeinclinedtorepeatthesameanswerasintheToMquestion,or(2)AccuratelyanswertherealityquestiontogivesomenewinformationtohelptheentityunderstandtheirfirstanswertotheToMquestion.Inordertoincreasethechancesofthissecondoption,theentitymustnotsimplybeperceivedbythe“mentor-child”as“ignorant”butalso“abitslow.” Inasimilarfashion,thememoryquestion,askedafteranincorrectreplytotheToMquestion,canalsobeinterpretedasarequestforconfirmationoftheunderstandingofthefirstanswer.Yet,inthestandardcontext,thememoryquestioncanseemtobedisruptingforchildrenyoungerthan4incorrectlyansweringtheToMquestionasthefinallocationoftheitemisgivenattheendofthestory.Indeed,thisanswerimplieshavingfollowedthechangeoflocationofthechocolateduringthestoryandtoremembertheinitiallocationofthechocolate.Theweakperformancesobserved(37.8%ofsuccessinthememoryquestion)inWimmerandPerner(1983)forchildrenbetween3and4maynotbetheresultofthedifficultyofthetaskbutinsteadbetheresultoftheambiguityofthequestionfortheirage.Olderchildren,becauseoftheirconversationalexperience,maymorereadilyreinterpretthememoryquestiontobecontrollingtheirinitialanswertotheToMquestion10. 3.2.TheRobot-PupilSolution ThereisanimportantliteratureshowingtheadvantagesofusingahumanoidNAOrobotinsocialinteractionswithyoungchildren,especiallyinsituationsoflearningbyteaching(seeJametetal.,2018).Studieshaveshownthatinconversationalinteractionwithanartificialagent,evencompletelyvirtualones,humansautomaticallydetectpragmaticviolationsoftheirspeaker(Jacquetetal.,2018;Jacquetetal.,2019a,b,c;LockshinandWilliams,2020).Itwasshownthatchildrenasyoungastwocanbesusceptibletotheconversationalviolationsofarobot(Ferrieretal.,2000).Recentstudies(Yasumatsuetal.,2017;Martinetal.,2020a,b)alsoshowedthatthenaturalandspontaneouspropensityofyoungchildrentotrybeingusefulextendstohumanoidrobotsseemingtobeindifficulty.Itseemsthat3years-oldchildrenassignmentalstatestoarobot(DiDioetal.,2018,2020a;Marchettietal.,2018).DiDioetal.(2020a)observedin3years-oldchildrenwhohadalreadydevelopedafirst-orderToMskillatendencytorepresenttheemotionalstateofarobotintermsofmentalstates.Fortheseauthors,therecouldbeanattempttoanthropomorphizetherobotontheemotionaldimensionwhich,attheageof3,couldbeparticularlysalient.Thissuggeststhatyoungchildrenareeagertothinkabouttherobotmindinthesamewaytheydoaboutthehumanmind(DiDioetal.,2018).TheNAOrobotwasalsousedtostudytheendowmenteffectinadults(Massonetal.,2015,2016;Massonetal.,2017a,b). Theeffectivenessofour“mentor-child”context11wastestedwithchildrenbetween5and6intheclassinclusiontaskandsuccessfullymadethequestionofclassinclusionmorerelevantforthechildwhenitwasasked(Massonetal.,2017a;Jametetal.,2018).Wehypothesizethatthe“mentor-child”contextshouldsimilarlydecreasetheambiguityoftheToMquestiontomakeitclearerthatitisarequestaboutthementalstatesofthecharacterMaxi.TheperformanceofpreschoolchildrenshouldthenbesignificantlyimprovedwithoutchangingtheoriginalexplicitFBT.Shouldthisbeobserved,wewouldconcludethattheunderstandingoffalsebeliefsdevelopsbeforetheageof4andthattheabilitiesofyoungchildrenareunderestimatedduetopragmaticfactors. Wealsobelievethatthe“mentor-child”contextcankeepthecontrolquestionsunambiguous.Therefore,weexpecttohavearateofcorrectresponsestothecontrolquestionsthatshouldberoughlyequivalenttothatofolderchildreninthestandardcontext. 4.Experiment:ExplicitFalseBeliefTaskintheMentor-ChildContext 4.1.MaterialsandMethods4.1.1.Participants Werecruited62nativeFrenchchildreninpreschoolat“LesPetitsPrinces”inVersailles,France.Thesamplechosenintheclasseswascomposedof34girlsand28boys,from38months-old(3yearsand2months)to49months-old(4yearsand1month)12.Themeanageofchildrenwas44months-old(N=62,M=44months-old,SD=2.82months-old)13.Thechildrenwererandomlyassignedaconditiondependingontheirageandgender.Theseconditionswere“humanexperimenter”(“human”condition)and“robotexperimenter”(“robot”condition).Eachconditioncontained31childrenbetween38months-oldand49months-old(N=31,M=44months-old,SD=3.47months-oldforthe“human”conditionandN=31,M=44months-old,SD=3.09months-oldforthe“robot”condition). 4.1.2.Materials Thestory“Maxiandthechocolate”wasshowntothechildwiththeclipsdisplayedinFigure1.Eachclipwas6.4×5.8cm(2.5×2.3in).TheclipswereshowninablackandopaquefoldercontainingacardboardspacerinA4format(21×29.7cmor8.3×11.7in).Allthreeclipsofthetaskwereattachedtothecardboardspacerinadvance.Therobotusedinthisexperimentwasa58cmtall(23in)NAOrobotcreatedbyAldebaranRobotics(Aldebaranversion4—“Evolution”).Ithasamovinghead,armsandhands,eachwiththreefingers,allowingittopointattheclipsofthestorytopunctuateitsdiscoursewithgestures.NAOisalsoequippedwithamicrophoneandspeakerstocommunicatewithhumans.Therobotwasremotelycontrolledbytheexperimenterusingacomputer,butitsgesturesandspeechwererecordedinadvance.Theycouldseethechildthankstoacameraintheeyesoftherobot.Themovementsandthespeechsectionstriggeredinrealtimeavoidedhavingtoomuchvariabilitybetweenthedifferentparticipants,whilestillmakingitpossibletomaketheanswersoftherobotfitthoseofthechild.Wechosetoremotelycontroltherobotforlogisticalreasons:eventhoughNAOdoeshavetheabilitytorecognizespeechmakingitpossibleforittoautonomouslyreact,thebehaviorsofchildrencansometimesbeunpredictable.Someflexibilitywasneededtoreproducewithfluidityanaturalconversationwithahuman.Moreover,childrencouldsometimesspeaktoolowtobeunderstoodbytherobot,whichwouldhavemadetheinteractionimpossible.Finally,therobotalsoallowedabetterstandardizationoftheenunciationcontextthankstoitsintonationsandutterancesbeingstrictlyidenticalacrossallparticipants.TomakeNAOmorechildishandlessintimidating,itsvoicewasmanipulatedsothatithadahigherpitchandspokemoreslowly.NAOwasprogrammedtoblinkrandomlyduringtheexperimenttostrengthenitshumanness. 4.1.3.Procedure Beforethebeginningoftheexperimentonememberoftheresearchteam,thatwecallthecompanion,waswelcomedintotheclassandgavetheirname.Childrenweresittinginacircleinfrontoftheteacher.Sheexplainedthatthisnewpersonwastheretomakeallthechildrenoftheclassworkonatask,abitliketeacher.Theprocedureinbothconditionswassubdividedintotwosequentialsteps:theprimingstepandtheexplicitFBT. 4.1.3.1.HumanCondition Inthisconditionthecompaniontoldthechildrentheywouldbeparticipatinginanactivityiftheyagreed.Afterthisintroductioneachchildwasguidedtothelocationoftheexperiment,inaquietmulti-purposeroomoftheschool.Duringthewalk,thecompaniontoldthechildthedidacticcontract:“You'reabouttolistentoastory,likeinclass,andmycolleague[nameoftheexperimenter]willaskyousomequestions.Youwillneedtoanswerthem.”Thecompanionthenaskedfortheagreementofthechild.Ifthechildagreed,thechildthenenteredtheroomwithoutthecompanionandstayedwiththeexperimenter. Theexperimenterthenintroducedthemselvestothechildwhowasseatedonachairinfrontthem.Thechild'sabilitytocorrectlynamethetwocolors(blueandgreen)wascheckedbeforethemaintask14.Thefalsebeliefstorywasthenverballytoldandillustratedwithclips,whichallowednon-verbalanswersforchildrenwhichpreferredtopointattheiranswerinsteadofsayingthem. Ifthechild'sanswertotheToMquestionwasthegreencupboard,theexperimenterpointedatitontheclipandsaid“ahitisthere.”Ifthechilddidnotchangetheirinitialanswer,theanswerwasconsideredtobecorrect.Whenthechildinsteadgavetheincorrectanswer,noconfirmationwasrequired,andtheanswerwasimmediatelyconsideredtobefalse. Therealityquestionandthememoryquestionwerethenasked(respectivelyfollowingthesuccessandfailuretotheToMquestion). Finally,theexperimenterthankedthechild,andthecompanionguidedthembacktotheclassroomwhilecongratulatingthem. 4.1.3.2.RobotCondition Inthiscondition,thecompanionexplainedthattheycamewithaNAOrobot.TheytoldtheclassNAOneededthechildren'shelpbecauseitknewnothingwhiletheyallknewalot.Ifonechilddoubtedoftheirknowledge,thecompaniontoldthemthattheywerelearningmanythingsinclassbutalsothattheyalreadyknewalot.Moreimportantly:theyknewmorethantherobot.ThecompanionthenaskedifthechildrenagreedtoteachthingstoNAO15. Likeinthe“Human”condition,thecompanionguidedeachchildindividuallyfromtheclasstothelocationoftheexperimentandtoldthemthedidacticcontract:“YourjobistoteachlotsofnewthingstoNAO.NAOisalittlerobotwhoknowsnothing.NAOneedsyoutolearnnewthings.NAOdoesn'tknowanything.Youwillbehisteacher.Doyouagreetobehisteacher?”TomakethechildunderstandNAO'signorancethecompanionpointedatthechild'sclothes,orvariousitemsinthelocationoftheexperiment.Theyaskedthechildtonamethem,whichwasdonewithoutanydifficulty,andthentheytoldthechild: “Yousee,NAOdoesn'tknowallthat.IfNAOasksyouweird,strangequestions,youmustanswerhim.Rememberthatheknowsnothing.IfNAOtellsyoustrangethings,orifhemakesmistakes,youcorrecthim16.Youarehisteacher.DoyouagreetobeNAO'steacher[nameofthechild]?” Ifthechildagreed,thecompanionletthechildentertheexperimentroomandleftthechild“alone”withtherobot(seeFigure2).Thisisanespeciallyimportantdetailwiththerobot.Indeed,shouldthecompanionremainintheroom,thechildmaybetemptedtoanswertherobotinthesamewaytheywouldwithahumanexperimenterbecauseofthepresenceofanadultintheroom.Pragmaticinterpretationswouldthenbemodified.Theactualexperimenterwashiddenbehindascreen,withoutthechildknowingaboutit,andremotelycontrolledNAOusingalaptop. FIGURE2 Figure2.Robotexperimenterandmaterialsforthe“Robot”condition.TheNAOrobotisseatedonatableinfrontofthechild. NAOintroduceditselftothechild.Itaskedifthechildagreedtobeitsteacherbecausehewasthereto“learnmanythings.”Onceagain,ifthechilddidnotagree,theexperimentstopped.Therobotthenaskedthechildiftheycanhelpitlearncolors.NAOthenpointedthecoloredcardboardsheetsandmademistakes(forexample:NAOsaid“That'syellow?”whiledesignatingthebluecardboard,makingitmorebelievablethefactthatNAOdidnotknowmuchandthusstrengtheningtheroleofthechildasateacher).Tofurtherstrengthenthenaiveaspectoftherobot,NAOinsistedonitsignoranceallalongtheexperiment(e.g.,“Alright,Ihadnotunderstoodthat.Iamreallystupid.”Itisimportanttonotethatgreatcarewasgiventonotoverdoingthe“stupidity”oftherobot.Indeed,ifitsmistakesbecametoopredictable,therewasagreatriskoflosingthechild'sinterestinteachingitanything.Achildcouldquicklyhaveinferredthat“NAOwillmakemistakesnomatterwhatItellhim.”whichcouldhavebiasedthechild'sexperienceinthetaskifithadnotbeencontrolled. NAOinitiatedthestory“MaxiandtheChocolate”bytellingthechild“AmantoldmeastorythatIdidnotunderstand.Doyouwanttohelpme?”Identicallytothe“Human”condition,NAOtoldthestoryandthenaskedtheToMquestionandthecontrolquestion. TheanswertotheToMquestionwasconsideredtobecorrectiftherobotwascorrectedbythechildwhenitmadeamistaketryingtorepeattheanswer.Forexample,ifthechildansweredthatMaxiwilllookforthechocolateinthegreencupboard(initialposition),NAOsaid:“Ahthanks,soifIunderstoodwellthechocolateisinthebluecupboard.”IfthechildcorrectedNAOandsaid:“No,thechocolateisthere.”(indicatingthegreencupboard)or“No,itisinthegreenone.”theanswerwasconsideredtobecorrect.Notethat,justlikeinthehumancondition,thechildcouldalsopointattheclipsdirectlyinsteadofansweringverbally. Oncethetaskwasover,NAOthankedthechildforbeingitsteacher“Thankyou,you'vebeenanawesometeacher.I'velearnedmanythingsthankstoyou!”andtoldthemgoodbye.Thecompanionthencametobringthechildbacktotheclassroom.Sometimestheteacheraskedhowthetaskwent.Thecompanioncongratulatedthechildforthequalityoftheirteaching.TheytoldtherestoftheclassthatNAOstillneededtoworktolearnthings.Thisway,thefactthatNAOneededhelpwasprogressivelyverywell-communicatedtothewholeclasswhiletheydidtheirusualclassactivities. 4.2.Results Inthisexperiment,thedependentvariablewasadichotomousvariablewhichmodalitieswereinterpretedintermsofsuccessorfailure.AccordingtotheprocedureusedbyWellmanandLiu(2004),thechild'sresponsewasasuccesswhentheyproducedcorrectanswerstoboththeToMquestionandtherealityquestion.Thisvariablewillbenotedbelow:TR(ToMandReality).Wealsoanalyzedthedatafromalessconservativeperspective(WimmerandPerner,1983)byinterpretingthesuccessasbeingsimplyacorrectanswertotheToMquestion.ThissecondversionofthedependentvariablewillbesymbolizedbytheletterT.Finally,wealsoanalyzedtheanswerstothememoryquestionforchildrenwhohadfailedtoanswertheToMquestioncorrectly.Thisvariablewillbedesignatedby¬TM(NotToMandMemory)17. Theindependentvariable(notedC)hadtwomodalities:“Human”vs.“Robot.”Wealsotestedtheinfluenceoftwoothervariables:thesexofthechild(notedS,withtwomodalities:Girlsvs.Boys),andtheageofthechildinmonths(notedA,numericalvariablerangingfrom38to49months-old).Inthefirststepoftheanalysis,weadjustedalinearmodelonourdatawithalinklogitfunctionandabinomialdistributionoftheerrors.WeappliedthistreatmenttoallthreeversionsofthedependentvariableTR,Tand¬TM.Forallofthemweincluded,inthelinearpredictorofthemodels,themaineffectsofeachofthethreefactorsC,S,andA,aswellasallthepossibleinteractionswhichincludesthetripleinteraction.Werefertothesesaturatedmodelsbyusingthefollowingexpressions:firstlyTR↫C*S*A,secondlyT↫C*S*Aandthirdly¬TM↫C*S*A.The“↫”symbolreferstotheinfluence,supposedorreal,oftheindependentvariablesonthedependentvariablewhilethe“*”symbolindicatesthatallthepossibleinteractionsbetweentheindependentvariablesaretakenintoaccount.Wethenusedaprocedureofautomaticbackwardsimplificationonallthesaturatedmodelstoleadtothecorrespondingfinalmodels. TheprincipalcharacteristicsofthesefourmodelsareshowninTable1.Resultsshow,regardlessoftheversionofthedependentvariable(TR,T,and¬TM),thatthesimplificationmodelsystematicallyterminatedonafinalmodelcontainingonlytheCfactor.ThismeansthatthesuccessrateforToM,definedeitherastheconservativemodel(TR)orasamorepermissivemodel(T),remainedcompletelyexplainablebythecondition(i.e.,“Human”vs.“Robot”).Thesamewasalsoobservedforthefinalmodelofthememoryquestion(¬TM).Therefore,inourstudyneitherthesex(S)northeage(A)ofthechildrencansignificantlyimprovethepredictionofthesuccessweobserved.Thus,intherestofthepaperthesetwovariableswillbeomitted. TABLE1 Table1.Maincharacteristicsofthedata-adjustedmodels. AsummaryofthedatawecollectedisshowninTable2. TABLE2 Table2.Distributionofthechildren'sanswersdependingontheexperimentalcondition(N=62). Table3showsthecoefficientsassociatedwitheachconditionforthethreeresultingmodelsrequiredtoestimatetheeffectsizeofthecondition(C). TABLE3 Table3.Estimatedβcoefficientsassociatedwithbelongingtotheconditionsforthethreemodels. ThemodelTR↫Chasasignificantcoefficient(β=1.23,p<0.05).ThiscoefficientisalsosignificantforthemodelT↫C(β=1.38,p<0.05).WeshowinTable3theoddsratio(OR)correspondingtotheβcoefficients.WeobtainedOR=3.43forthemodelTR↫C,whichmeansthatthechancesofsuccessforachildinthe“Robot”conditionarealmost3.5timesgreaterthanthatofchildreninthe“Human”condition.RegardingthemodelT↫C,weobtainedOR=3.98indicatingthat,whensimplifyingthesuccesscriterion,achildwasfourtimesmorelikelytosucceedinthe“Robot”conditionthanoneinthe“Human”condition.WealsoobservedatendencyforchildrenwhofailedtheToMquestiontoanswerthememoryquestionwithmoresuccesswhenitwasaskedbytherobot(β=1.62,p=0.06).Whilenotsignificantwecanstillpointoutthatchildrenwerefivetimesmorelikelytocorrectlyanswerwiththerobotthantheywerewiththehuman(OR=5.04). Table2showsthedistributionoftheparticipantsdependingontheconditionandonwhethertheysucceededinthetask(dependingonthecriterionusedtodefinesuccess).AnunilateralproportiontestwithnocontinuitycorrectionrevealsthatthesuccessrateforTRissignificantlydifferentfromchance(χ2=2.77,df=1,p<0.05).WhenonlylookingattheToMquestion(T)thetestdoesnotshowasignificantdifference(χ2=0.4,df=1,p>0.05).However,asexplainedabove,ananswerscoredas“correct”fortheToMquestionneededtobeconfirmed.Thus,wecanprobablythinkthatthe58%ofchildreninthe“Robot”conditiondidnotsimplygivethecorrectansweratrandom.Only2childrenchangedtheirchoicesforToMquestioninthe“Robot”condition(andwerecountedasawronganswerforToM)andnoneinthe“Human”condition. 5.Discussion ThegoalofthisstudywastoproposeanewmethodologyoftheexplicitFBT.Withit,wehopedtoinhibittheerroneousinterpretationsmadeby3years-oldchildrenregardingtheToMquestion.Our“mentor-child”contextseemstohavechangedtheprevailinginterpretationoftheToMquestioninthewaywehoped:arequesttoreportthefalsebeliefofthecharacter.Theyoungchildrenwhoparticipatedinourstudydidbetterinthe“Robot”conditionthanthoseinthe“Human”condition.Thisresulthasthreeimportantconsequences: 1.ItprovidesnewexperimentalargumentsforapragmaticexplanationofthefailureofyoungchildreninexplicitFBT(Cummings,2013;Helmingetal.,2014,2016;Westra,2017;WestraandCarruthers,2017;Frank,2018). 2.Itindicatesthatasignificantproportionofpre-schoolchildrencancorrectlyanswertheoriginalToMquestion. 3.Thisresult,followingthoseofJametetal.(2018)onPiaget'sclassinclusiontask18,supportstherelevanceofourmethodologytodisambiguatetheexperimenter'sexpectationsthroughtheirquestionindevelopmentaltasks. Inourstudy,theperformanceofchildrenintheconjunctionoftheToMandrealityquestionswassignificantlyimproved,withchildreninthe“Robot”conditionbeingabout3.5timesmorelikelytosucceedthanthoseinthe“Human”condition.Moreover,inthe“Human”conditiontheperformanceswerecomparabletothoseobservedintheliterature(WimmerandPerner,1983;Hogrefeetal.,1986;Perneretal.,1987).Thepreviousresultisalsoamplifiedif,asWimmerandPerner(1983),oneadoptsalaxerinterpretationofthesuccess.Indeed,lookingonlyattherecordedresponsestotheToMquestion,ourresultsshowthatchildrenbelongingtothe“Robot”conditionare4timesmorelikelytosucceed.Furthermore,althoughwefocusedontheperformanceofpre-schoolchildrenwithparticipantsbetween3and4years-old,itisinterestingtonotethatthesuccessrateinour“Robot”condition(58%)issimilartowhatWimmerandPerner(1983)considerstobeasuccessfulcompletionofthetaskforchildrenbetween4and6years-old(57%).However,thisproportionremainslowerthantheonerecordedbythesameauthorsforchildrenbetween6and9years-old(89%).Wecanpointoutthatthissuccessrateisnotassensationalasthoseobservedinsomestudies(suchasthe90–100%observedinRubio-FernándezandGeurts,2013,2016).Toourknowledge,wearethefirsttofindsuchaperformancewithoutanymodificationbeingmadetotheinitialparadigmwiththesamescenario,thesamequestionsandthesameprocedureforanalyzingtheanswersgivenbythechild.Besides,theexperimentalprotocolalsoconsidersseveralmethodologicalcriticismsmadeinthestudiescitedabove(Wellmanetal.,2001;WellmanandLiu,2004;KammermeierandPaulus,2018;Priewasseretal.,2020).Indeed,weconsideredacorrectanswertobewhenthechildansweredbothquestions(ToMandreality)correctly.Ourparticipantswerealsorandomlyassignedtothe“Robot”and“Human”conditionsinahomogeneousway. Replicatingourprocedurewithchildrenbetween4and9wouldbeimportantandinterestinginordertoseeifourmethodologyproducesasimilarimprovementforthe4–6years-oldagegrouporifthislevelofperformancecorrespondstoaplateauforchildrenbelowtheageof6.Inthefirstcase,thetraditionalresultsfoundintheliteratureofexplicitFBT,showingaprogressionwithage,wouldnotqualitativelychangebutsimplybeshiftedtowardyoungeragegroups.Itwouldthenbeessentialtoreplicateourprocedurewithchildrenbetween2and3todecideatwhatagetheexplicitFBTcanstarttobecomesuccessful.Inthesecondcase,withalimitedsuccessratebeforetheageof6,the6–7agegroupwouldbethepivotalageforreachingalmosta90%successratewithexplicitFBT.Thiswouldimplythatimportantpragmaticand/orcognitivecapacitieswouldstillbelackingattheageof5,preventingatotalsuccessatthisage.Thiswouldnotnecessarilycontradictourpragmaticapproach.Indeed,numerousstudiesreportthat6isthepivotalagetobeabletocorrectlygeneraterelevanceimplicatures(BoscoandGabbatore,2017;GrigoroglouandPapafragou,2017).Asexplainedabove,explicitFBTarecomplexastheyrequirea“tripleattributionofmentalstates”(Helmingetal.,2014,2016;WestraandCarruthers,2017).Theyimplynotonlythatthechildmusttakeintoaccounttheperspectiveofthecharacterofthestorybutalsothatoftheirinterlocutor,whoisanadultexperimenterinthestandardtest,sincethechildinfersexpectationsfromthemandfinallytheirownperspective.Consequently,thistaskwouldnotbeafirstordertask,butratherasecondordertask,thusexplainingthethresholdofa60%successrate. Itisalsointerestingtolookspecificallyathowthechildrenrespondedtothecontrolquestionsinour“mentor-child”context.AswasshownbyPernerandWimmer(1985)thetwotypesofsuccesscoding(withorwithouttherealityquestion)slightlymodifytheresultsdownwardswithoutchangingtheinterpretation:thechancesofsuccessinthe“Robot”conditionrelativetothatinthe“Human”conditionwentfrom4timeshighertoabout3.5timeshigher.Thesuccessratedecreasedwhentheanswertotherealityquestionwasconsidered.Intermsofproportions,bothconditionshadasimilarsuccessrateintherealityquestion(75%in“Human”and77.7%in“Robot”conditions).Thismayconfirmthattheemphasisonthe“slow”traitoftherobotallowsustodisambiguatealargepartoftherealityquestion.Forthememoryquestion,aspredicted,wefoundahighersuccessrate(85%)inthe“Robot”conditionwhichissimilartothe83.7%observedinWimmerandPerner(1983)withchildrenbetween4and5years-old.Thisresultseemstoconfirmourhypothesisthatthisquestionisnoticeablyambiguousinthestandardcontextforpreschoolchildren. Thefactthatthis“mentor-child”contextworkswith3years-oldchildrenalsoprovidesnewargumentsinfavoroftheuseofahumanoidrobotasatoolinexperimentalresearchonchildrenandadults.Ourstudydidnothaveasitsmainobjectivetomeasuretheimportanceoftherobottoolitselfbutrathertheinfluenceofthecontextitallowedtoproduce.However,itwouldbeimportantinafuturestudytoseeifthereisaspecificroboteffectinourresultsthatcanstandoutonitsown.Wecanruntheexperimentwithpuppetsorotherobjectsrepresentingan“ignorant,”“naive,”and“slow”entity(inanunpublishedexploratorystudyontheclassinclusiontaskJamet,Saïbou-DumontandBaratgin(2018)obtain,fromchildrenofFrenchGuyana,similarperformancestothoseobtainedinJametetal.(2018)withtheuseofapuppetoramandisguisedasarobotinsteadoftheNAOrobot19).Asecondpossibilitywouldbetorunthestudywithaknowledgeableandintelligent“NAOteacher”inadditiontothehumanexperimenterandtotheslowrobot.Manystudieshaveshownthatchildrenasyoungas3years-oldacceptedtheNAOrobotasapossibleteacher(RosandaandIstenicStarcic,2020).OrançandKüntay(2020)observedinchildrenfrom3to6years-oldaclearpreferencetoasktherobotquestionsaboutmachines,andlessaboutbiologyandpsychology.Thus,onecouldexpectthatinthissituationchildrenwouldbeevenlessinclinedtointerprettheToMquestioncorrectlyasbeingaquestionaboutMaxi'sbeliefs.Allthisseemstoindicatethatourresultsarelargelytheconsequenceofthe“mentor-child”context. Ourstudyalsobringstwoimportantnewelementsonchild-robotinteraction.Firstly,ourstudyseemstoconfirmthatpreschoolchildrenattributebeliefstotherobotaswasalsoindicatedinrecentstudies(DiDioetal.,2018,2020a;Marchettietal.,2018).Secondly,inourstudythechildcanbehavelikeamentor,withthemotivationtohelparobotunderstandastory.Thishelpingbehaviorstillhappenedeventhoughphysicalinteractionsarequitelimited.Indeed,therobotdidnothaveagreatautonomyofmovementwhenseatedinfrontofthechildanditdisplayedfewexpressions(theNAOrobotcannotsmileanditsfacialexpressionsareverylimited:onlyitseyescanchangecolorstosignifyanemotion).ThisiscoherentwithresultsfromMartinetal.(2020a,b)whichindicatethatthehelpingbehaviorofchildrendoesnotseemtobeconditionedtothelevelofanimatedautonomynortothefriendlyexpressionsoftherobot'svoice. Whileourmethodologyseemstoworkforaninteractionwithchildrenolderthan3yearsand2months,childrenbetween5and6years-old,andalsowithadults(Massonetal.,2015,2016;Massonetal.,2017a,b),childrenundertheageof3didnotagreetostay“alone”withNAO.Itispossiblethatthechoiceofahumanoidrobotmaytroubleyoungchildren.DiDioetal.(2020b)showsthat3years-oldchildrentendtotrusthumansmorethanrobots,asopposedto7years-oldchildren.Manzietal.(2020)showedthatchildrenof5,7,and9years-olddifferentlyassignmentalstatestotwohumanoidrobots,NAOandRobovie,differingontheirlevelofanthropomorphism.Itispossiblethat,forveryyoungchildrenunder3years-old,theNAOrobotmaynotbethemostadequatetool(seeDamianoetal.,2015,forareviewofthedifferenttypesofrobots).Thiswouldexplainthelownumberofstudieswithchildrenofthisage.Recentreviewsontheinteractionsbetweenneuro-typicalchildrenandarobot(Jametetal.,2018;Neumann,2020;vanStratenetal.,2020)indicatethatonlyonestudywasconductedusingNAOandagroupofchildrenfrom2to8years-old(Yasumatsuetal.,2017).Thefewotherstudiesconductedon2years-oldeitherusedthetinyhumanoidrobotQRIOthatissmallerthana2years-oldchild(Tanakaetal.,2007),theiRobiQrobotthatlooksmorelikeatoy(Hsiaoetal.,2015),orrobotsspecificallydesignedtobeenjoyedbyyoungchildrenlikethestuffeddragonrobotDragonbot(KoryWestlundetal.,2017)andtheRUBI-4(Movellanetal.,2009).Thus,shouldwedecidetodoalongitudinalstudyfrom2to9years-oldusingourcontextualprocedurewewouldneedtostudywhichrobotisthemostrelevanttoplaytheroleofaratherslowandignorantbeingforallages. 6.Conclusion TheessentialpropositionthathasbeendevelopedandtestedinourstudyisthattheanswertotheToMquestioncruciallydependsonthe“conversationallogic”atplayinthecontextualizedinteractionsbetweentheexperimenterandthechild.Thisinteractionshapesthechild'sinterpretationofthequestion.OurcontextualmodificationpragmaticfilterstheToMquestion,removingirrelevantinterpretations.Thestandardparadigmforcesthechildtoperformarelevancesearchtointerpretanambiguousquestionaskedbyanexpert(withastatuslikethatofateacher)withinthelimitsofthechild'sownmeta-cognitiveknowledge.Inour“mentor-child”contextthechildanswersanunequivocalquestionaboutthebeliefsoftheprotagonistofthestoryaskedbyasomewhatslowentitywhoneedstheirhelp.Here,the3years-oldchildcananswercorrectlyeveniftheirmeta-cognitiveknowledgeispoorlydeveloped.Thisprocedurehelpsusbecomemore“competent”speaker-experimenters(Sperber,1994)asitoffersatooltoplaceourselvesattheleveloftheyoungchild'sinterpretationstrategy.Thisallowsthemtorealizewhatisrelevanttoanswerthequestioncorrectly.Forsimilarreasonswebelievethatthisproceduremayalsohelpwiththeunderstandingofthesecond-orderToM(PernerandWimmer,1985).Itcouldreducetheambiguityofthequestionoftheexperimenterwhichexistsinmanyexperimentalparadigms.ResultsofLombardietal.(2018)indeedindicate,usingadialogicalperspective,thataconsiderablepartofthesupposedfailuresobservedwithchildreninthesecondordertaskareinfacttheresultofanadversepragmaticcontext.InadditiontothePiagetiantasksoflengthandnumberconservation(McGarrigleandDonaldson,1974),volumeconservation(Jametetal.,2014),orclassinclusion(Politzer,2016),thereareavarietyofexperimentalparadigmsthatlendthemselveswelltoourdisambiguationmethodology.The“mentor-child”contextcouldalsofacilitatesomestudieswithatypicallydevelopingparticipants,suchasindividualswithanAutismSpectrumDisorderwhoshowbothdeficientperformanceonthefalsebelieftask(Baron-Cohen,1997)andinlanguagepragmatics(Angelerietal.,2016).Finally,ourmethodologyalsooffersnewcluesontherelevanceofhuman-robotinteraction,andinparticularonchild-robotinteraction.Morestudiesshouldmostcertainlyfocusontheinteractionbetweenchildrenandrobots,takinginconsiderationthebeliefstheyassociatetothesetools,andtheireffectonwell-knownpsychologicalresults. DataAvailabilityStatement ThedatasetsanalyzedforthisstudycanbefoundintheOpenScienceFrameworkrepositoryatthefollowingaddresshttps://osf.io/ey4n5/?view_only=d8d2e16f39ea4186b994e2468a7408cd. EthicsStatement ThestudiesinvolvinghumanparticipantswerereviewedandapprovedbyM.CharlesEl-nouty,ProfesseurdesUniversitésenMathématiques,LAGAUMR7539,UniversitéParis13:PresidentoftheCommittee.M.Jean-YvesHenry,Chirurgien-Dentistediplômédel'UniversitéParis7;M.MichelDionnet,Chefdecuisine,Membretitulairedel'AcadémieCulinairedeFrance;M.FabriceGutnick,MCFassociéenSciencesdel'Éducation,UniversitéJulesVernesAmiens,Psychologuedutravail;MmeDominiqueKlarsyMédecindutravail.Writteninformedconsenttoparticipateinthisstudywasprovidedbytheparticipants'legalguardian/nextofkin. AuthorContributions JBandFJ:conceptualelaboration.JB,FJ,andMD-S:designofthestudy.MD-SandFJ:datacollection.J-LS:dataanalysis.JBandMD-S:draftofthemanuscript.JB,BJ,andFJ:criticalrevisionofthemanuscript.Allauthorscontributedtothearticleandapprovedthesubmittedversion. Funding WethanktheP-A-R-I-SAssociationforthetechnicalandfinancialhelpwereceivedaswellastheCHArtlaboratorywhichparticipatedinfinancingthepublicationofthearticleinopenaccess.P-A-R-I-SFundingnumber:2020-0301728-5CHArt-Paris8Fundingnumber:2020-0331087-0. ConflictofInterest Theauthorsdeclarethattheresearchwasconductedintheabsenceofanycommercialorfinancialrelationshipsthatcouldbeconstruedasapotentialconflictofinterest. Acknowledgments WewouldliketoexpressourgratitudetotheNationalEducationInspectorEugénieMontesofVersaillesandtothepedagogicalteamofLesPetitsPrincesdeVersaillesschoolforwelcomingus,fortheirinvolvementandfortheirinterestinthisresearchproject.WewouldliketothankinparticulartheheadmasterMrs.Wiklacz,aswellastheteachersoftheschool:Mrs.Moussette,Combe,Dupuy,andDelehaye.WealsothankNataliaObrochta,OlivierMasson,andYouriMinnefortheirhelpduringthepre-testsofthisexperiment.WealsothankResearchDirector(DR)BéatriceDesgrangesandAnneChevalier,copyrightmanagerofRevuedeNeuropsychologie,fortheirauthorizationtorepublishFigure1,originallyfromDuvaletal.(2011).WewouldfinallyliketothankAndrewHromekforproof-readingthepaperandGuyPolitzerforhiscarefulreviewofthefirstdraftofthisdocument. Footnotes 1.^Thesituationismoreorlesscomplexdependingontheleveloffalsebeliefevaluated.Threelevelsofrepresentation(threegradualordersofdifficulty)obtainedatdifferentagescanbedistinguished(Duvaletal.,2011).Orderzeroisautomaticallyacquired.Itcorrespondstowhatwearecurrentlythinkingabout.Firstordercorrespondstotheinferenceofthementalstateofsomeoneelseandwouldonlybeacquiredattheageof4.Thesecondorderreferstotheinferenceofthementalstateofanotherpersonaboutanotherpersonandshouldbeacquiredbetween6and7.ThispaperismainlyinterestedintheageatwhichchildrenacquirethefirstorderToM.Forthesakeofsimplicity,wewillomittospecify“firstorder”whenwereferto“explicitFBT”intherestofthispaper. 2.^InWimmerandPerner(1983)Maxiwouldputthechocolateinthe“blue”cupboardandhismotherwouldmoveittothe“green”cupboard.We'veinterchangedthecolorsinthisarticletomatchtheclipsusedinourexperimentthatweretakenfromDuvaletal.(2011,p.45). 3.^Seeforarecentargumentinfavorofthishypothesis(DohertyandPerner,2020). 4.^CluesexistseemingtoindicatethatthelatesuccessinexplicitFBTmayindeedbetheresultoflearningfromrepeatedsocialexperiences(WangandSu,2009).Studiesshowthatacorrelationexistsbetweenthenumberofbrothersandsistersofasimilarageandthecomprehensionoffalsebeliefs(Perneretal.,1994;Ruffmanetal.,2012;JenkinsandAstington,2014).Fromtheageof3thechildcanuselanguagefromameta-cognitivepointofviewtolie(LewisandSaarni,1993)andstarttobeabletousecontextualinformation(Salomoetal.,2013).Itisnecessarytowaittheageof4toseechildrenabletoadapttheirdiscoursebytakingintoaccounttheageofthelistener,theirstatusandtheirgender;andcanadapttheirdiscoursetoyoungerpeople.Theycanalsoaskaconversationpartnertoreformulateanutteranceiftheydidnotunderstandit(ClarkandAmaral,2010). 5.^Thesetwoadditionalinterpretationswerealreadyevokedin(Perneretal.,1987,p.126):“Theymayhavemisinterpretedthetestquestion:‘Wherewilltheprotagonistlookforthechocolate?'asmeaning,‘Whereshouldhelook?'or‘Helphimtofindit!”'Theseauthorschangedthequestionto“Wheredoeshethinkthechocolateis?”However,aspointedoutinWestraandCarruthers(2017),theterm“think”requiresmorecognitiveresourcesthantheterm“tolook.”Also,thisversioncomplicatesratherthansimplifiestheissue,whichexplainswhyitdoesnotimprovetheperformanceofyoungchildren. 6.^Severalpiecesofexperimentalevidenceindicatethat3years-oldchildreneasilydistinguishbetweenwhatanotherpersonknowsanddoesnot(Hogrefeetal.,1986;PernerandLeekam,1986). 7.^Youngchildrenseemtobetterunderstandtheirroleineducationalactivitieswhentheyareexplicitlyformulated(JeongandFrye,2018a). 8.^Itwasonlychildren4years-old(noyoungerchildrenhadansweredtheToMquestioncorrectly). 9.^Therealityquestion,inthiscontext,lookslikeaviolationoftheprincipleofinformativeness(Grice'sMaximofQuantity,1975;Ducrot'slawofexhaustiveness,1980/2008)whichrequiresthateachparticipantinaconversationanswertheirpartner'sutterancewithanappropriatequantityofinformation(neithertoolittlenortoomuch).Ifmultipleexperimentalcluesquestionthecompleteacquisitionofthisprincipleattheageof3(ContiandCamras,1984;Noveck,2001;Eskrittetal.,2008),otherstudiesindicatethatsomechildrenofthatageshowskillslikeadaptingtheircommunicativebehaviortothestateofknowledgeoftheirpartners(O'Neill,1996;Dunhametal.,2000;Ferrieretal.,2000).PernerandLeekam(1986)showthatfromtheageof3,childrenprefermentioningfirstthemostinformativeelementandavoidmentioningelementsalreadyknownbytheirlistener. 10.^Itcanbenotedthatthisinterpretationofthememoryquestion(anexpectationoftheexperimentertobecontrollingtheanswertotheToMquestion)requiresskillsofsecondorderToM. 11.^Sincethe“ignorant,”“naive,”and“slow”robotwasonlytheretostrengthenthechild'simpressiontobetheonewiththeknowledge,itspresencewillbeimpliedeachtimewerefertothe“mentor-child”context. 12.^Writteninformedconsenttoparticipateinthisstudywasprovidedbytheparticipants'legalguardian/nextofkin.Alldatawascollectedanonymously.TheexperimentwasreviewedandapprovedbytheEthicsCommitteeoftheCHArtLaboratory.TheEthicsstatementcanbeobtainedhere:https://osf.io/wk4af?view_only=d8d2e16f39ea4186b994e2468a7408cd. 13.^Theinitialsamplecontainedfiveclassesofpreschoolchildrenforatotalof70childrenfrom34to49months-old.Wehadatleastone34months-oldchild,one35months-oldchild,andsoon,ineachofthetwoconditions.Duringtheexperiment,wenoticedthat3childreninthe“robot”condition(theyoungest:34,35,and36months-old)becamereallyscaredwhentheNAOrobotstartedmoving,liftingitsheadandlookingatthechild.Themovementoftherobotisnotasfluidasthatofahuman,andthenoiseofthemotorsisquitenoticeable.Inconsequencethesethreechildrenhadtoberemovedfromthecondition.Inordertokeeptheconditionshomogeneousintermsofage,weremovedonechildunder3years-oldinthe“robot”conditionwhohadsucceededinthetask,andall4youngestchildreninthe“human”condition,whohadallfailedinthetask. 14.^Inbothconditionsallchildrencorrectlynamedthetwocolors. 15.^Ascriptoftheinteractionbetweena“mentor-child”andtheNAOrobot,writtenasaclinicalandcriticalPiagetianinterview(Ducret,2016)canbefoundathttps://osf.io/z5s7k/?view_only=d8d2e16f39ea4186b994e2468a7408cd. 16.^Thecompanioninsistedonthisspecificpoint. 17.^ThecompleteRscriptoftheanalysesisavailableathttps://osf.io/34hzn/?view_only=d8d2e16f39ea4186b994e2468a7408cdandthedataitselfisavailableathttps://osf.io/wzx7g/?view_only=d8d2e16f39ea4186b994e2468a7408cd. 18.^Jametetal.(2018)randomlyassigned40children(between5and6years-old)totwoconditionssimilartothosewecreatedforthepresentstudy(“HumanExperimenter”vs.“IgnorantNAORobot”).Theauthorsobservedaclearimprovementinperformanceinthe“IgnorantNAORobot”condition:onechildoutoffiveansweredcorrectlyinthe“HumanExperimenter”condition,andmorethansixchildrenoutofteninthe“IgnorantNAORobot”condition. 19.^ExperimentswerecarriedoutduringtheMINformationofteachersrequestedbytherectoroftheacademyofFrenchGuyana. References Angeleri,R.,Gabbatore,I.,Bosco,F.M.,Sacco,K.,andColle,L.(2016).Pragmaticabilitiesinchildrenandadolescentswithautismspectrumdisorder:astudywiththeabacobattery.Miner.Psichiatr.57,93–103. GoogleScholar Antonietti,A.,Sempio,O.,andMarchetti,A.(2006).TheoryofMindandLanguageinDevelopmentalContexts.TheSpringerSeriesonHumanExceptionality.NewYork,NY:SpringerNewYork. GoogleScholar Astington,J.W.,andOlson,D.R.(1995).Thecognitiverevolutioninchildren'sunderstandingofmind.Hum.Dev.38,179–189.doi:10.1159/000278313 CrossRefFullText|GoogleScholar Bagassi,M.,andMacchi,L.(2006).Pragmaticapproachtodecisionmakingunderuncertainty:thecaseofthedisjunctioneffect.Think.Reason.12,329–350.doi:10.1080/13546780500375663 CrossRefFullText|GoogleScholar Bagassi,M.,Salerni,N.,Castoldi,V.,Sala,V.,Caravona,L.,Poli,F.,etal.(2020).Improvingchildren'slogicalandmathematicalperformanceviaapragmaticapproach.Front.Educ.5:54.doi:10.3389/feduc.2020.00054 CrossRefFullText|GoogleScholar Baillargeon,R.,Scott,R.M.,andHe,Z.(2010).False-beliefunderstandingininfants.TrendsCogn.Sci.14,110–118.doi:10.1016/j.tics.2009.12.006 CrossRefFullText|GoogleScholar Baratgin,J.(2002).Isthehumanminddefinitelynotbayesian?Areviewofthevariousarguments.Curr.Psychol.Cogn.21,653–682. GoogleScholar Baratgin,J.(2009).Updatingourbeliefsaboutinconsistency:themonty-hallcase.Math.Soc.Sci.57,67–95.doi:10.1016/j.mathsocsci.2008.08.006 CrossRefFullText|GoogleScholar Baratgin,J.,andNoveck,I.A.(2000).Notonlybaseratesareneglectedintheengineer-lawyerproblem:aninvestigationofreasoners'underutilizationofcomplementarity.Mem.Cogn.28,79–91.doi:10.3758/BF03211578 PubMedAbstract|CrossRefFullText|GoogleScholar Baratgin,J.,andPolitzer,G.(2006).Isthemindbayesian?Thecaseforagnosticism.MindSoc.5,1–38.doi:10.1007/s11299-006-0007-1 CrossRefFullText|GoogleScholar Baratgin,J.,andPolitzer,G.(2007).Thepsychologyofdynamicprobabilityjudgment:ordereffect,normativetheories,andexperimentalmethodology.MindSoc.6,53–66.doi:10.1007/s11299-006-0025-z CrossRefFullText|GoogleScholar Baratgin,J.,andPolitzer,G.(2010).Updating:apsychologicallybasicsituationofprobabilityrevision.Think.Reason.16,253–287.doi:10.1080/13546783.2010.519564 CrossRefFullText|GoogleScholar Baron-Cohen,S.(1997).Mindblindness:AnEssayonAutismandTheoryofMind.Cambridge:MITPress. GoogleScholar Baron-Cohen,S.,Leslie,A.M.,andFrith,U.(1985).Doestheautisticchildhavea“theoryofmind?”Cognition21,37–46.doi:10.1016/0010-0277(85)90022-8 CrossRefFullText|GoogleScholar Bartsch,K.,andWellman,H.M.(1995).ChildrenTalkAbouttheMind.Oxford:OxfordUniversityPress. GoogleScholar Białecka-Pikul,M.,Kosno,M.,Białek,A.,andSzpak,M.(2019).Let'sdoittogether!Theroleofinteractioninfalsebeliefunderstanding.J.Exp.ChildPsychol.177,141–151.doi:10.1016/j.jecp.2018.07.018 PubMedAbstract|CrossRefFullText|GoogleScholar Bosco,F.M.,andGabbatore,I.(2017).Theoryofmindinrecognizingandrecoveringcommunicativefailures.Appl.Psycholinguist.38,57–88.doi:10.1017/S0142716416000047 PubMedAbstract|CrossRefFullText|GoogleScholar Braine,M.D.S.,andShanks,B.L.(1965a).Theconservationofashapepropertyandaproposalabouttheoriginoftheconservations.Can.J.Psychol.19,197–207.doi:10.1037/h0082903 PubMedAbstract|CrossRefFullText|GoogleScholar Braine,M.D.S.,andShanks,B.L.(1965b).Thedevelopmentofconservationofsize.J.VerbalLearn.VerbalBehav.4,227–242.doi:10.1016/S0022-5371(65)80025-1 CrossRefFullText|GoogleScholar Buttelmann,D.,Carpenter,M.,andTomasello,M.(2009).Eighteen-month-oldinfantsshowfalsebeliefunderstandinginanactivehelpingparadigm.Cognition112,337–342.doi:10.1016/j.cognition.2009.05.006 PubMedAbstract|CrossRefFullText|GoogleScholar Buttelmann,D.,Over,H.,Carpenter,M.,andTomasello,M.(2014).Eighteen-month-oldsunderstandfalsebeliefsinanunexpected-contentstask.J.Exp.ChildPsychol.119,120–126.doi:10.1016/j.jecp.2013.10.002 PubMedAbstract|CrossRefFullText|GoogleScholar Callaghan,T.,Rochat,P.,Lillard,A.,Claux,M.L.,Odden,H.,Itakura,S.,etal.(2005).Synchronyintheonsetofmental-statereasoning:evidencefromfivecultures.Psychol.Sci.16,378–384.doi:10.1111/j.0956-7976.2005.01544.x PubMedAbstract|CrossRefFullText|GoogleScholar Carpendale,J.I.M.,andLewis,C.(2004).Constructinganunderstandingofmind:thedevelopmentofchildren'ssocialunderstandingwithinsocialinteraction.Behav.BrainSci.27,79–96.doi:10.1017/S0140525X04000032 PubMedAbstract|CrossRefFullText|GoogleScholar Clark,E.V.,andAmaral,P.M.(2010).Childrenbuildonpragmaticinformationinlanguageacquisition.Lang.Linguist.Compass4,445–457.doi:10.1111/j.1749-818X.2010.00214.x CrossRefFullText|GoogleScholar Conti,D.J.,andCamras,L.A.(1984).Children'sunderstandingofconversationalprinciples.J.Exp.ChildPsychol.38,456–463.doi:10.1016/0022-0965(84)90088-2 CrossRefFullText|GoogleScholar Cummings,L.(2013).“Clinicalpragmaticsandtheoryofmind,”inPerspectivesonLinguisticPragmatics.PerspectivesinPragmatics,Vol.2,Philosophy&Psychology,edsA.Capone,F.LoPiparo,andM.Carapezza(Dordrecht:Springer),23–56. GoogleScholar Damiano,L.,Dumouchel,P.,andLehmann,H.(2015).Towardshuman-robotaffectiveco-evolutionovercomingoppositionsinconstructingemotionsandempathy.Int.J.Soc.Robot.7,7–18.doi:10.1007/s12369-014-0258-7 CrossRefFullText|GoogleScholar DeBruin,L.C.,andNewen,A.(2012).Anassociationaccountoffalsebeliefunderstanding.Cognition123,240–259.doi:10.1016/j.cognition.2011.12.016 PubMedAbstract|CrossRefFullText|GoogleScholar DiDio,C.,Isernia,S.,Ceolaro,C.,Marchetti,A.,andMassaro,D.(2018).Growingupthinkingofgod'sbeliefs:theoryofmindandontologicalknowledge.SAGEOpen8.doi:10.1177/2158244018809874 CrossRefFullText|GoogleScholar DiDio,C.,Manzi,F.,Peretti,G.,Cangelosi,A.,Harris,P.L.,Massaro,D.,etal.(2020a).Comeibambinipensanoallamentedelrobot.ilruolodell'attaccamentoedellateoriadellamentenell'attribuzionedistatimentaliadunagenterobotico[Howchildrenthinkabouttherobot'smind.theroleofattachmentandtheoryofmindintheattributionofmentalstatestoaroboticagent].Sist.Intell.1,41–46.doi:10.1422/96279 CrossRefFullText|GoogleScholar DiDio,C.,Manzi,F.,Peretti,G.,Cangelosi,A.,Harris,P.L.,Massaro,D.,etal.(2020b).ShallItrustyou?Fromchild-robotinteractiontotrustingrelationships.Front.Psychol.11:469.doi:10.3389/fpsyg.2020.00469 PubMedAbstract|CrossRefFullText|GoogleScholar Diesendruck,G.,andMarkson,L.(2001).Children'savoidanceoflexicaloverlap:apragmaticaccount.Dev.Psychol.37,630–641.doi:10.1037/0012-1649.37.5.630 PubMedAbstract|CrossRefFullText|GoogleScholar Doherty,M.J.,andPerner,J.(2020).Mentalfiles:developmentalintegrationofdualnamingandtheoryofmind.Dev.Rev.56:100909.doi:10.1016/j.dr.2020.100909 CrossRefFullText|GoogleScholar Ducret,J.(2016).Jeanpiagetetlaméthode“clinicocritique”[jeanpiagetandthe“clinico-critical”method].J.FrançaisPsychiatr.44,79–84.doi:10.3917/jfp.044.0079 CrossRefFullText|GoogleScholar Ducrot,O.(1980/2008).Direetnepasdire:principesdesémantiquelinguistique[ToSayandNottoSay:PrinciplesofLinguisticSemantics].CollectionSavoir.Paris:Hermann. GoogleScholar Dulany,D.E.,andHilton,D.J.(1991).Conversationalimplicature,consciousrepresentation,andtheconjunctionfallacy.Soc.Cogn.9,85–110.doi:10.1521/soco.1991.9.1.85 CrossRefFullText|GoogleScholar Dunham,P.,Dunham,F.,andO'Keefe,C.(2000).Two-year-olds'sensitivitytoaparent'sknowledgestate:mindreadingorcontextualcues?Br.J.Dev.Psychol.18,519–532.doi:10.1348/026151000165832 CrossRefFullText|GoogleScholar Duval,C.,Piolino,P.,Bejanin,A.,Laisney,M.,Eustache,F.,andDesgranges,B.(2011).Lathéoriedel'esprit:aspectsconceptuels,évaluationeteffetsdel'âge[Thetheoryofmind:conceptualaspects,evaluationandeffectsofage].Rev.Neuropsychol.3,41–51.doi:10.3917/rne.031.0041 CrossRefFullText|GoogleScholar Eskritt,M.,Whalen,J.,andLee,K.(2008).Preschoolerscanrecognizeviolationsofthegriceanmaxims.Br.J.Dev.Psychol.26,435–443.doi:10.1348/026151007X253260 PubMedAbstract|CrossRefFullText|GoogleScholar Ferrier,S.,Dunham,P.,andDunham,F.(2000).Theconfusedrobot:two-year-olds'responsestobreakdownsinconversation.Soc.Dev.9,337–347.doi:10.1111/1467-9507.00129 CrossRefFullText|GoogleScholar Frank,C.K.(2018).Revivingpragmatictheoryoftheoryofmind.AIMSNeurosci.5,116–131.doi:10.3934/Neuroscience.2018.2.116 PubMedAbstract|CrossRefFullText|GoogleScholar Gelman,S.A.,andBloom,P.(2000).Youngchildrenaresensitivetohowanobjectwascreatedwhendecidingwhattonameit.Cognition76,91–103.doi:10.1016/S0010-0277(00)00071-8 PubMedAbstract|CrossRefFullText|GoogleScholar Grice,H.P.(1975).“Logicandconversation,”inSpeechActs(SyntaxandSemantics3),edsP.Cole,andJ.Morgan(NewYork,NY:AcademicPress),41–58. GoogleScholar Grigoroglou,M.,andPapafragou,A.(2017).“Acquisitionofpragmatics,”inOxfordResearchEncyclopediaofLinguistics,edsR.ClarkandM.Aronoff(Oxford:OxfordUniversityPress).doi:10.1093/acrefore/9780199384655.013.217 PubMedAbstract|CrossRefFullText|GoogleScholar Hala,S.,Chandler,M.,andFritz,A.S.(1991).Fledglingtheoriesofmind:deceptionasamarkerofthree-year-olds'understandingoffalsebelief.ChildDev.62,83–97.doi:10.2307/1130706 CrossRefFullText|GoogleScholar Hansen,M.(2010).Ifyouknowsomething,saysomething:youngchildren'sproblemwithfalsebeliefs.Front.Psychol.1:23.doi:10.3389/fpsyg.2010.00023 PubMedAbstract|CrossRefFullText|GoogleScholar Hayes,J.R.(1972).“Thechild'sconceptionoftheexperimenter,”inInformationProcessinginChildren,edS.Farnham-Diggory(NewYork,NY:AcademicPress),175–181.doi:10.1016/B978-0-12-249550-2.50018-3 CrossRefFullText|GoogleScholar Helming,K.A.,Strickland,B.,andJacob,P.(2014).Makingsenseofearlyfalse-beliefunderstanding.TrendsCogn.Sci.18,167–170.doi:10.1016/j.tics.2014.01.005 PubMedAbstract|CrossRefFullText|GoogleScholar Helming,K.A.,Strickland,B.,andJacob,P.(2016).Solvingthepuzzleaboutearlybelief-ascription.MindLang.31,438–469.doi:10.1111/mila.12114 PubMedAbstract|CrossRefFullText|GoogleScholar Hepach,R.,Vaish,A.,Grossmann,T.,andTomasello,M.(2016).Youngchildrenwanttoseeothersgetthehelptheyneed.ChildDev.87,1703–1714.doi:10.1111/cdev.12633 PubMedAbstract|CrossRefFullText|GoogleScholar Hepach,R.,Vaish,A.,andTomasello,M.(2012).Youngchildrenareintrinsicallymotivatedtoseeothershelped.Psychol.Sci.23,967–972.doi:10.1177/0956797612440571 PubMedAbstract|CrossRefFullText|GoogleScholar Heyes,C.(2014).Falsebeliefininfancy:afreshlook.Dev.Sci.17,647–659.doi:10.1111/desc.12148 PubMedAbstract|CrossRefFullText|GoogleScholar Hogrefe,G.-J.,Wimmer,H.,andPerner,J.(1986).Ignoranceversusfalsebelief:adevelopmentallaginattributionofepistemicstates.ChildDev.57,567–582.doi:10.2307/1130337 CrossRefFullText|GoogleScholar Hsiao,H.-S.,Chang,C.-S.,Lin,C.-Y.,andHsu,H.-L.(2015).“Irobiq”:theinfluenceofbidirectionalinteractiononkindergarteners'readingmotivation,literacy,andbehavior.Interact.Learn.Environ.23,269–292.doi:10.1080/10494820.2012.745435 CrossRefFullText|GoogleScholar Huemer,M.,Perner,J.,andLeahy,B.(2018).Mentalfilestheoryofmind:whendochildrenconsideragentsacquaintedwithdifferentobjectidentities?Cognition171,122–129.doi:10.1016/j.cognition.2017.10.011 PubMedAbstract|CrossRefFullText|GoogleScholar Jacquet,B.,Baratgin,J.,andJamet,F.(2018).“Thegriceanmaximsofquantityandofrelationintheturingtest,”in11thInternationalConferenceonHumanSystemInteraction(HSI)(Gdansk).doi:10.1109/HSI.2018.8431328 CrossRefFullText|GoogleScholar Jacquet,B.,Baratgin,J.,andJamet,F.(2019a).Cooperationinonlineconversations:theresponsetimesasawindowintothecognitionoflanguageprocessing.Front.Psychol.10:727.doi:10.3389/fpsyg.2019.00727 PubMedAbstract|CrossRefFullText|GoogleScholar Jacquet,B.,Hullin,A.,Baratgin,J.,andJamet,F.(2019b).“Theimpactofthegriceanmaximsofquality,quantityandmannerinchatbots,”in2019InternationalConferenceonInformationandDigitalTechnologies(IDT)(Zilina),180–189.doi:10.1109/DT.2019.8813473 CrossRefFullText|GoogleScholar Jacquet,B.,Masson,O.,Jamet,F.,andBaratgin,J.(2019c).“Onthelackofpragmaticprocessinginartificialconversationalagents,”inHumanSystemsEngineeringandDesign(IHSED),Volume876ofAdvancesinIntelligentSystemsandComputing,edsT.Ahram,W.Karwowski,andR.Taiar(Cham:Springer),394–399.doi:10.1007/978-3-030-02053-8_60 CrossRefFullText|GoogleScholar Jamet,F.,Baratgin,J.,andFilatova,D.(2014).Globalwarmingandtheriseofthesealevel:astudyofintellectualdevelopmentinpreadolescentsandadolescentsfrom11to15yearsold.Stud.Pedag.24,361–380. GoogleScholar Jamet,F.,Masson,O.,Jacquet,B.,Stilgenbauer,J.-L.,andBaratgin,J.(2018).Learningbyteachingwithhumanoidrobot:anewpowerfulexperimentaltooltoimprovechildren'slearningability.J.Robot.2018:4578762.doi:10.1155/2018/4578762 CrossRefFullText|GoogleScholar Jenkins,J.M.,andAstington,J.W.(2014).Cognitivefactorsandfamilystructureassociatedwiththeoryofminddevelopmentinyoungchildren.Dev.Psychol.327,70–789.doi:10.1037/0012-1649.32.1.70 CrossRefFullText|GoogleScholar Jeong,J.,andFrye,D.(2018a).Explicitversusimplicitunderstandingofteaching:doesknowingwhatteachingishelpchildrentolearnfromit?Teach.Teach.Educ.71,355–365.doi:10.1016/j.tate.2018.02.002 CrossRefFullText|GoogleScholar Jeong,J.,andFrye,D.(2018b).Informationaboutinformants'knowledgestatesaffectschildren'spredictionsoflearningandtheiractuallearning.Cogn.Dev.48,203–216.doi:10.1016/j.cogdev.2018.08.008 CrossRefFullText|GoogleScholar Kammermeier,M.,andPaulus,M.(2018).Doaction-basedtasksevidencefalse-beliefunderstandinginyoungchildren?Cogn.Dev.46,31–39.doi:10.1016/j.cogdev.2017.11.004 CrossRefFullText|GoogleScholar KoryWestlund,J.M.,Dickens,L.,Jeong,S.,Harris,P.L.,DeSteno,D.,andBreazeal,C.L.(2017).Childrenusenon-verbalcuestolearnnewwordsfromrobotsaswellaspeople.Int.J.ChildComput.Interact.13,1–9.doi:10.1016/j.ijcci.2017.04.001 CrossRefFullText|GoogleScholar Leslie,A.M.(2005).Developmentalparallelsinunderstandingmindsandbodies.TrendsCogn.Sci.9,459–462.doi:10.1016/j.tics.2005.08.002 PubMedAbstract|CrossRefFullText|GoogleScholar Leslie,A.M.,Friedman,O.,andGerman,T.P.(2004).Coremechanismsin‘theoryofmind'.TrendsCogn.Sci.8,528–533.doi:10.1016/j.tics.2004.10.001 PubMedAbstract|CrossRefFullText|GoogleScholar Lewis,M.,andSaarni,C.(Eds.).(1993).LyingandDeceptioninEverydayLife.NewYork,NY:TheGuilfordPress. GoogleScholar Lewis,S.,Lidz,J.,andHacquard,V.(2012).Thesemanticsandpragmaticsofbeliefreportsinpreschoolers.Semant.Linguist.Theory22,247–267.doi:10.3765/salt.v22i0.3085 CrossRefFullText|GoogleScholar Liszkowski,U.,Carpenter,M.,andTomasello,M.(2008).Twelve-month-oldscommunicatehelpfullyandappropriatelyforknowledgeableandignorantpartners.Cognition108,732–739.doi:10.1016/j.cognition.2008.06.013 PubMedAbstract|CrossRefFullText|GoogleScholar Lockshin,J.,andWilliams,T.(2020).““Weneedtostartthinkingahead”:theimpactofsocialcontextonlinguisticnormadherence,”inCogSci2020(Zilina).doi:10.31234/osf.io/2wtzq CrossRefFullText|GoogleScholar Lombardi,E.,Greco,S.,Massaro,D.,Schär,R.,Manzi,F.,Iannaccone,A.,etal.(2018).Doesagoodargumentmakeagoodanswer?Argumentativereconstructionofchildren'sjustificationsinasecondorderfalsebelieftask.Learn.Cult.Soc.Interact.18,13–27.doi:10.1016/j.lcsi.2018.02.001 CrossRefFullText|GoogleScholar Macchi,L.(2000).Partitiveformulationofinformationinprobabilisticproblems:beyondheuristicsandfrequencyformatexplanations.Organ.Behav.Hum.Decis.Process.82,217–236.doi:10.1006/obhd.2000.2895 PubMedAbstract|CrossRefFullText|GoogleScholar Macchi,L.,andBagassi,M.(2012).Intuitiveandanalyticalprocessesininsightproblemsolving:apsycho-rhetoricalapproachtothestudyofreasoning.MindSoc.11,53–67.doi:10.1007/s11299-012-0103-3 CrossRefFullText|GoogleScholar Macchi,L.,Caravona,L.,Poli,F.,Bagassi,M.,andFranchella,M.A.(2020).SpeakyourmindandIwillmakeitright:thecaseof“selectiontask.”J.Cogn.Psychol.32,93–107.doi:10.1080/20445911.2019.1707207 CrossRefFullText|GoogleScholar Macchi,L.,Poli,F.,Caravona,L.,Vezzoli,M.,Franchella,M.A.,andBagassi,M.(2019).Howtogetridofthebeliefbias:boostinganalyticalthinkingviapragmatics.Eur.J.Psychol.15,595–613.doi:10.5964/ejop.v15i3.1794 CrossRefFullText|GoogleScholar Manzi,F.,Peretti,G.,DiDio,C.,Cangelosi,A.,Itakura,S.,Kanda,T.,etal.(2020).Arobotisnotworthanother:exploringchildren'smentalstateattributiontodifferenthumanoidrobots.Front.Psychol.11:2011.doi:10.3389/fpsyg.2020.02011 CrossRefFullText|GoogleScholar Marchetti,A.,Manzi,F.,andItakura,S.A.M.D.(2018).Theoryofmindandhumanoidrobotsfromalifespanperspective.Z.Psychol.226,98–109.doi:10.1027/2151-2604/a000326 CrossRefFullText|GoogleScholar Markman,E.M.,andWachtel,G.F.(1988).Children'suseofmutualexclusivitytoconstrainthemeaningsofwords.Cogn.Psychol.20,121–157.doi:10.1016/0010-0285(88)90017-5 PubMedAbstract|CrossRefFullText|GoogleScholar Martin,D.U.,MacIntyre,M.I.,Perry,C.,Clift,G.,Pedell,S.,andKaufman,J.(2020a).Youngchildren'sindiscriminatehelpingbehaviortowardahumanoidrobot.Front.Psychol.11:239.doi:10.3389/fpsyg.2020.00239 PubMedAbstract|CrossRefFullText|GoogleScholar Martin,D.U.,Perry,C.,MacIntyre,M.I.,Varcoe,L.,Pedell,S.,andKaufman,J.(2020b).Investigatingthenatureofchildren'saltruismusingasocialhumanoidrobot.Comput.Hum.Behav.104:106149.doi:10.1016/j.chb.2019.09.025 CrossRefFullText|GoogleScholar Mascaro,O.,andMorin,O.(2015).Epistemologyforbeginners:two-tofive-year-oldchildren'srepresentationoffalsity.PLoSONE10:e140658.doi:10.1371/journal.pone.0140658 PubMedAbstract|CrossRefFullText|GoogleScholar Mascaro,O.,Morin,O.,andSperber,D.(2017).Optimisticexpectationsaboutcommunicationexplainchildren'sdifficultiesinhiding,lying,andmistrustingliars.J.ChildLang.44,1041–1064.doi:10.1017/S0305000916000350 PubMedAbstract|CrossRefFullText|GoogleScholar Masson,O.,Baratgin,J.,andJamet,F.(2015).“NAOrobotandthe“endowmenteffect”,”in2015IEEEInternationalWorkshoponAdvancedRoboticsanditsSocialImpacts(ARSO)(Lyon),1–6.doi:10.1109/ARSO.2015.7428203 CrossRefFullText|GoogleScholar Masson,O.,Baratgin,J.,andJamet,F.(2017a).“NAOrobotasexperimenter:socialcuesemitterandneutralizertobringnewresultsinexperimentalpsychology,”inInternationalConferenceonInformationandDigitalTechnologies(IDT-2017)(Zilina),256–264.doi:10.1109/DT.2017.8024306 CrossRefFullText|GoogleScholar Masson,O.,Baratgin,J.,andJamet,F.(2017b).“NAOrobot,transmitterofsocialcues:whatimpacts?”inAdvancesinArtificialIntelligence:FromTheorytoPractice.IEA/AIE2017,Volume10350ofLectureNotesinComputerScience,edsS.K.T.BenferhatandM.Ali(Cham:Springer),559–568.doi:10.1007/978-3-319-60042-0_62 CrossRefFullText|GoogleScholar Masson,O.,Baratgin,J.,Jamet,F.,Ruggieri,F.,andFilatova,D.(2016).“Usearobottoserveexperimentalpsychology:someexamplesofmethodswithchildrenandadults,”inInternationalConferenceonInformationandDigitalTechnologies(IDT-2016)(Rzeszow),190–197.doi:10.1109/DT.2016.7557172 CrossRefFullText|GoogleScholar Matsui,T.,andMiura,Y.(2008).Pro-socialmotivepromotesearlyunderstandingoffalsebelief.Nat.Prec.doi:10.1038/npre.2008.1695.1 CrossRefFullText|GoogleScholar McGarrigle,J.,andDonaldson,M.(1974).Conservationaccidents.Cognition3,341–350.doi:10.1016/0010-0277(74)90003-1 CrossRefFullText|GoogleScholar Mitchell,P.,andLacohée,H.(1991).Children'searlyunderstandingoffalsebelief.Cognition39,107–127.doi:10.1016/0010-0277(91)90040-B PubMedAbstract|CrossRefFullText|GoogleScholar Movellan,J.R.,Eckhardt,M.,Virnes,M.,andRodriguez,A.(2009).“Sociablerobotimprovestoddlervocabularyskills,”in20094thACM/IEEEInternationalConferenceonHuman-RobotInteraction(HRI)(LaJolla,CA),307–308.doi:10.1145/1514095.1514189 CrossRefFullText|GoogleScholar Neumann,M.M.(2020).Socialrobotsandyoungchildren'searlylanguageandliteracylearning.EarlyChild.Educ.J.48,157–170.doi:10.1007/s10643-019-00997-7 CrossRefFullText|GoogleScholar Newen,A.,andWolf,J.(inpress).Thesituationalmentalfileaccountofthefalsebelieftasks:anewsolutionoftheparadoxoffalsebeliefunderstanding.Rev.Philos.Psychol.doi:10.1007/s13164-020-00466-w CrossRefFullText|GoogleScholar Noveck,I.A.(2001).Whenchildrenaremorelogicalthanadults:experimentalinvestigationsofscalarimplicature.Cognition78,165–188.doi:10.1016/S0010-0277(00)00114-1 PubMedAbstract|CrossRefFullText|GoogleScholar Oktay-Gür,N.,Schulz,A.,andRakoczy,H.(2018).Childrenexhibitdifferentperformancepatternsinexplicitandimplicittheoryofmindtasks.Cognition173,60–74.doi:10.1016/j.cognition.2018.01.001 PubMedAbstract|CrossRefFullText|GoogleScholar O'Neill,D.K.(1996).Two-year-oldchildren'ssensitivitytoaparent'sknowledgestatewhenmakingrequests.ChildDev.67,659–677.doi:10.2307/1131839 CrossRefFullText|GoogleScholar Onishi,K.H.,andBaillargeon,R.(2005).Do15-month-oldinfantsunderstandfalsebeliefs?Science308,255–258.doi:10.1126/science.1107621 PubMedAbstract|CrossRefFullText|GoogleScholar Oranç,C.,andKüntay,A.C.(2020).Children'sperceptionofsocialrobotsasasourceofinformationacrossdifferentdomainsofknowledge.Cogn.Dev.54:100875.doi:10.1016/j.cogdev.2020.100875 CrossRefFullText|GoogleScholar Perner,J.(1991).Learning,Development,andConceptualChange.UnderstandingtheRepresentationalMind.Cambridge:TheMITPress. GoogleScholar Perner,J.,Huemer,M.,andLeahy,B.(2015).Mentalfilesandbelief:acognitivetheoryofhowchildrenrepresentbeliefanditsintensionality.Cognition145,77–88.doi:10.1016/j.cognition.2015.08.006 PubMedAbstract|CrossRefFullText|GoogleScholar Perner,J.,andLeahy,B.(2016).Mentalfilesindevelopment:dualnaming,falsebelief,identityandintensionality.Rev.Philos.Psychol.7,491–508.doi:10.1007/s13164-015-0235-6 PubMedAbstract|CrossRefFullText|GoogleScholar Perner,J.,andLeekam,S.R.(1986).Beliefandquantity:three-yearolds'adaptationtolistener'sknowledge.J.ChildLang.13,305–315.doi:10.1017/S0305000900008072 PubMedAbstract|CrossRefFullText|GoogleScholar Perner,J.,Leekam,S.R.,andWimmer,H.(1987).Three-year-olds'difficultywithfalsebelief:thecaseforaconceptualdeficit.Br.J.Dev.Psychol.5,125–137.doi:10.1111/j.2044-835X.1987.tb01048.x PubMedAbstract|CrossRefFullText|GoogleScholar Perner,J.,Ruffman,T.,andLeekam,S.R.(1994).Theoryofmindiscontagious:youcatchitfromyoursibs.ChildDev.65,1228–1238.doi:10.2307/1131316 CrossRefFullText|GoogleScholar Perner,J.,andWimmer,H.(1985).“Johnthinksthatmarythinksthat…”attributionofsecond-orderbeliefsby5-to10-year-oldchildren.J.Exp.ChildPsychol.39,437–471.doi:10.1016/0022-0965(85)90051-7 CrossRefFullText|GoogleScholar Piaget,J.,andSzeminska,A.(1941).Lagenèsedunombrechezl'enfant[Theoriginofnumberinchildren].Neuchâtel:DelachauxetNiestlé. GoogleScholar Politzer,G.(1993).Lapsychologieduraisonnement:Loisdelapragmatiqueetlogiqueformelle[Thepsychologyofreasoning:lawsofpragmaticsandformallogic](Thèsed'état),UniversitéParis8,Paris,France. GoogleScholar Politzer,G.(2004).“Reasoning,judgementandpragmatics,”inExperimentalPragmatics.PalgraveStudiesinPragmatics,LanguageandCognition,edsI.NoveckandD.Sperber(London:PalgraveMacmillan),94–115.doi:10.1057/9780230524125_5 CrossRefFullText|GoogleScholar Politzer,G.(2016).Theclassinclusionquestion:acasestudyinapplyingpragmaticstotheexperimentalstudyofcognition.SpringerPlus5:1133.doi:10.1186/s40064-016-2467-z PubMedAbstract|CrossRefFullText|GoogleScholar Politzer,G.,andMacchi,L.(2000).Reasoningandpragmatics.MindSoc.1,73–93.doi:10.1007/BF02512230 CrossRefFullText|GoogleScholar Priewasser,B.,Fowles,F.,Schweller,K.,andPerner,J.(2020).Mistakenmaxbefriendsduplogirl:Nodifferencebetweenastandardandanacted-outfalsebelieftask.J.Exp.ChildPsychol.191:104756.doi:10.1016/j.jecp.2019.104756 PubMedAbstract|CrossRefFullText|GoogleScholar Recanati,F.(2012).MentalFiles.Oxford:OxfordUniversityPress. GoogleScholar Reddy,V.(1991).“Playingwithothers'expectations:teasingandmuckingaboutinthefirstyear,”inNaturalTheoriesofMind,edA.Whiten(Cambridge,MA:BasilBlackwell),143–158. GoogleScholar Reddy,V.(2007).Gettingbacktotheroughground:deceptionand‘socialliving'.Philos.Trans.R.Soc.BBiol.Sci.362,621–637.doi:10.1098/rstb.2006.1999 PubMedAbstract|CrossRefFullText|GoogleScholar Reddy,V.(2008).HowInfantsKnowMinds.Cambridge,MA:HarvardUniversityPress. GoogleScholar Rosanda,V.,andIstenicStarcic,A.(2020).“Therobotintheclassroom:areviewofarobotrole,”inEmergingTechnologiesforEducation,edsE.Popescu,T.Hao,T.C.Hsu,H.Xie,M.Temperini,andW.Chen(Cham:SpringerInternationalPublishing),347–357.doi:10.1007/978-3-030-38778-5_38 CrossRefFullText|GoogleScholar Rose,S.A.,andBlank,M.(1974).Thepotencyofcontextinchildren'scognition:anillustrationthroughconservation.ChildDev.45,499–502.doi:10.2307/1127977 CrossRefFullText|GoogleScholar Rubio-Fernández,P.,andGeurts,B.(2013).Howtopassthefalse-belieftaskbeforeyourfourthbirthday.Psychol.Sci.24,27–33.doi:10.1177/0956797612447819 PubMedAbstract|CrossRefFullText|GoogleScholar Rubio-Fernández,P.,andGeurts,B.(2016).Don'tmentionthemarble!Theroleofattentionalprocessesinfalse-belieftasks.Rev.Philos.Psychol.7,835–850.doi:10.1007/s13164-015-0290-z CrossRefFullText|GoogleScholar Ruffman,T.,Taumoepeau,M.,andPerkins,C.(2012).Statisticallearningasabasisforsocialunderstandinginchildren.Br.J.Dev.Psychol.30,87–104.doi:10.1111/j.2044-835X.2011.02045.x PubMedAbstract|CrossRefFullText|GoogleScholar Sabbagh,M.A.,andBowman,L.C.(2018).Theoryofmind.StevensHandb.Exp.Psychol.Cogn.Neurosci.4,1–39.doi:10.1002/9781119170174.epcn408 CrossRefFullText|GoogleScholar Salomo,D.,Lieven,E.,andTomasello,M.(2013).Children'sabilitytoanswerdifferenttypesofquestions.J.ChildLang.40,469–491.doi:10.1017/S0305000912000050 PubMedAbstract|CrossRefFullText|GoogleScholar Scott,R.M.,Baillargeon,R.,Song,H.,andLeslie,A.M.(2010).Attributingfalsebeliefsaboutnon-obviouspropertiesat18months.Cogn.Psychol.61,366–395.doi:10.1016/j.cogpsych.2010.09.001 PubMedAbstract|CrossRefFullText|GoogleScholar Shatz,M.,Wellman,H.M.,andSilber,S.(1983).Theacquisitionofmentalverbs:asystematicinvestigationofthefirstreferencetomentalstate.Cognition14,301–321.doi:10.1016/0010-0277(83)90008-2 PubMedAbstract|CrossRefFullText|GoogleScholar Siegal,M.,andBeattie,K.(1991).Wheretolookfirstforchildren'sknowledgeoffalsebeliefs.Cognition38,1–12.doi:10.1016/0010-0277(91)90020-5 PubMedAbstract|CrossRefFullText|GoogleScholar Southgate,V.,Senju,A.,andCsibra,G.(2007).Actionanticipationthroughattributionoffalsebeliefby2-year-olds.Psychol.Sci.18,587–592.doi:10.1111/j.1467-9280.2007.01944.x PubMedAbstract|CrossRefFullText|GoogleScholar Sperber,D.(1994).“Understandingverbalunderstanding,”inWhatisIntelligence?edJ.Khalfa(Cambridge:CambridgeUniversityPress),179–198. GoogleScholar Sperber,D.,Cara,F.,andGirotto,V.(1995).Relevancetheoryexplainstheselectiontask.Cognition57,31–95.doi:10.1016/0010-0277(95)00666-M PubMedAbstract|CrossRefFullText|GoogleScholar Sperber,D.,andWilson,D.(1986).Relevance:CommunicationandCognition,Vol.142.Cambridge,MA:HarvardUniversityPress. GoogleScholar Sperber,D.,andWilson,D.(2002).Pragmatics,modularityandmind-reading.MindLang.17,3–23.doi:10.1111/1468-0017.00186 CrossRefFullText|GoogleScholar Surian,L.,Caldi,S.,andSperber,D.(2007).Attributionofbeliefsby13-month-oldinfants.Psychol.Sci.18,580–586.doi:10.1111/j.1467-9280.2007.01943.x PubMedAbstract|CrossRefFullText|GoogleScholar Tanaka,F.,Cicourel,A.,andMovellan,J.R.(2007).Socializationbetweentoddlersandrobotsatanearlychildhoodeducationcenter.Proc.Natl.Acad.Sci.U.S.A.104,17954–17958.doi:10.1073/pnas.0707769104 PubMedAbstract|CrossRefFullText|GoogleScholar vanStraten,C.L.,Peter,J.,andKühne,R.(2020).Child-robotrelationshipformation:anarrativereviewofempiricalresearch.Int.J.Soc.Robot.12,325–344.doi:10.1007/s12369-019-00569-0 PubMedAbstract|CrossRefFullText|GoogleScholar Wang,Y.,andSu,Y.(2009).Falsebeliefunderstanding:childrencatchitfromclassmatesofdifferentages.Int.J.Behav.Dev.33,331–336.doi:10.1177/0165025409104525 CrossRefFullText|GoogleScholar Warneken,F.(2015).Precociousprosociality:whydoyoungchildrenhelp?ChildDev.Perspect.9,1–6.doi:10.1111/cdep.12101 CrossRefFullText|GoogleScholar Warneken,F.,andTomasello,M.(2007).Helpingandcooperationat14monthsofage.Infancy11,271–294.doi:10.1111/j.1532-7078.2007.tb00227.x CrossRefFullText|GoogleScholar Warneken,F.,andTomasello,M.(2009).Therootsofhumanaltruism.Br.J.Psychol.100,455–471.doi:10.1348/000712608X379061 PubMedAbstract|CrossRefFullText|GoogleScholar Warneken,F.,andTomasello,M.(2013).Parentalpresenceandencouragementdonotinfluencehelpinginyoungchildren.Infancy18,345–368.doi:10.1111/j.1532-7078.2012.00120.x CrossRefFullText|GoogleScholar Wellman,H.M.,andBartsch,K.(1988).Youngchildren'sreasoningaboutbeliefs.Cognition30,239–277.doi:10.1016/0010-0277(88)90021-2 PubMedAbstract|CrossRefFullText|GoogleScholar Wellman,H.M.,Cross,D.,andWatson,J.(2001).Meta-analysisoftheory-of-minddevelopment:thetruthaboutfalsebelief.ChildDev.72,655–684.doi:10.1111/1467-8624.00304 PubMedAbstract|CrossRefFullText|GoogleScholar Wellman,H.M.,andLiu,D.(2004).Scalingoftheory-of-mindtasks.ChildDev.75,523–541.doi:10.1111/j.1467-8624.2004.00691.x PubMedAbstract|CrossRefFullText|GoogleScholar Westra,E.(2017).Pragmaticdevelopmentandthefalsebelieftask.Rev.Philos.Psychol.8,235–257.doi:10.1007/s13164-016-0320-5 CrossRefFullText|GoogleScholar Westra,E.,andCarruthers,P.(2017).Pragmaticdevelopmentexplainsthetheory-of-mindscale.Cognition158,165–176.doi:10.1016/j.cognition.2016.10.021 PubMedAbstract|CrossRefFullText|GoogleScholar Wimmer,H.,andPerner,J.(1983).Beliefsaboutbeliefs:representationandconstrainingfunctionofwrongbeliefsinyoungchildren'sunderstandingofdeception.Cognition13,103–128.doi:10.1016/0010-0277(83)90004-5 PubMedAbstract|CrossRefFullText|GoogleScholar Yasumatsu,Y.,Sono,T.,Hasegawa,K.,andImai,M.(2017).“Icanhelpyou:altruisticbehaviorsfromchildrentowardsarobotatakindergarten,”inProceedingsoftheCompanionofthe2017ACM/IEEEInternationalConferenceonHuman-RobotInteraction,HRI'17(NewYork,NY:AssociationforComputingMachinery),331–332.doi:10.1145/3029798.3038305 CrossRefFullText|GoogleScholar Yazdi,A.A.,German,T.P.,Defeyter,M.A.,andSiegal,M.(2006).Competenceandperformanceinbelief-desirereasoningacrosstwocultures:thetruth,thewholetruthandnothingbutthetruthaboutfalsebelief?Cognition100,343–368.doi:10.1016/j.cognition.2005.05.004 CrossRefFullText|GoogleScholar Keywords:theoryofmind,preschoolchildren,pragmatics,humanoidrobot,mentor-childcontext,ignorantrobot,humanrobotinteraction,first-orderfalsebelieftask Citation:BaratginJ,Dubois-SageM,JacquetB,StilgenbauerJ-LandJametF(2020)PragmaticsintheFalse-BeliefTask:LettheRobotAsktheQuestion!Front.Psychol.11:593807.doi:10.3389/fpsyg.2020.593807 Received:11August2020;Accepted:28October2020;Published:23November2020. Editedby:HiromiTsuji,OsakaShoinWomen'sUniversity,Japan Reviewedby:FedericoManzi,CatholicUniversityoftheSacredHeart,ItalyLauraMacchi,UniversityofMilan-Bicocca,Italy Copyright©2020Baratgin,Dubois-Sage,Jacquet,StilgenbauerandJamet.Thisisanopen-accessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense(CCBY).Theuse,distributionorreproductioninotherforumsispermitted,providedtheoriginalauthor(s)andthecopyrightowner(s)arecreditedandthattheoriginalpublicationinthisjournaliscited,inaccordancewithacceptedacademicpractice.Nouse,distributionorreproductionispermittedwhichdoesnotcomplywiththeseterms. *Correspondence:JeanBaratgin,[email protected] COMMENTARY ORIGINALARTICLE Peoplealsolookedat SuggestaResearchTopic>