01816Number of repetitionsPer cent recalled - PDF document

01816Number of repetitionsPer cent recalled Suppress a 01816Number of repetitionsPer cent recalled Suppress 01816Number of repetitionsPer cent recalled Suppress e 01816Number of repetitionsPer cent recalled Suppress f 01816Number of repetitionsPer cent recalled © 2 001 Macmillan Ma gazines Ltd letterstonatureNATUREVOL41015MARCH2001www.nature.comresponseandtheentranceofanunwantedmemoryintoawareness.Beforethisphase,wetoldsubjectswhichcueswouldrequiresuppressionsothattheywouldrecognizethesewordsonsight.Subjectsperformedsuppressionandrespondtrialsondifferentpairsthatwereintermixed.Oneachtrial,novisualmarkingindicatedwhichcuewordsweresuppressionitems,forcingsubjectstoidentifyeachcuetoknowwhethertorecallorsuppresstheassociatedmemory.Theobjectiveofthethink/no-thinktaskwastodeterminewhetherattemptingtopreventawarenessofanunwantedmemorywouldhinderitslaterretrieval.Toevaluatewhetherthisoccurred,thenextphaserequiredsubjectstorecalltheresponseforeachofthecuewords.Weemphasizedthatthepreviousgoalofavoidingtheassociateditemswasnolongerrelevantandthataresponseshouldberecalledforeverycue.Iftryingtopreventamemoryfromenteringawarenessrecruitsinhibitorycontrolpro-cessesthatimpairthatmemory'sretrievability,recallforsuppres-sionitemsshouldbeworsethanforbaselinepairsonthistest.Inallexperimentsreportedhere,baselineitemswerestudiedpairsthatdidnotappearaseitherrespondorsuppressionitemsduringthethink/no-thinkphase.Inexperiment1,wevariedthenumberofsuppressionorrespondtrialsgivenforeachpair.Subjectsreceived0,1,8or16trialsforeachsuppressionandeachresponditem.Ifexcludingunwantedmem-oriesfromawarenessrecruitsinhibitorycontrol,recallshouldbeworseafter16suppressionattemptsthanafter0attempts(baseline).AsshowninFig.1a,®nalrecallofsuppresseditemswasworsethanofbaselineitems,andimpairmentincreasedlinearlywithsuppres-sionpractice.Incontrast,recallimprovedacrossrepetitionsforresponditems,demonstratingtheestablishedbene®tsofretrievingmemoriesontheirlaterrecall.Thesedivergingpatternsshowthatcontrollingawarenessnotonlyterminatedthepowerfulfacilitativeeffectsofretrieval,butalsoimpairedtherecallofsuppresseditemstobelowtheirbaselinelevel(0suppressionattempts).Theincreas-inginhibitionwithrepetitionfurtherindicatesthatunwantedmemoriesmightbeespeciallyvulnerableinsettingsrequiringprotractedavoidance,unlikethemodesttime(1minover16suppressions)affordedbyourtask.Impairedmemoryforsuppressionitemsindicatesthattheremaybeanexecutivecontrolprocessthatsuppresses(reducesactivationof)theunwantedmemoryitself(forexample,roachinFig.2).However,mechanismsotherthaninhibitionmaybeatwork.Forinstance,onestrategyforavoidinganunwantedmemorywouldbetogeneratediversionarythoughtstoenvironmentalstimulithatremindusofit.Newassociationsbetweenthesestimuliandthediversionarythoughtsmayinterfereduringlaterattemptstorecallthememory.Alternatively,terminatingretrievalmaydegradetheassociationbetweenthecueandtheresponse.Neitherofthesealternativesrequiresustoassumethattheunwantedmemoryitselfisinhibitedandthustheydonotrequirethepostulationofanexecutiveinhibitionprocess.Toisolatethecontributionofinhibi-tion,weusedtheindependentprobemethod.Ifinhibitionimpairstheunwantedmemoryitself,recallshouldbeworseregardlessofwhetherthatitemistestedwiththecueusedtotrainsuppressionorwithanovelcue(Fig.2).However,associativeinterferenceandunlearningpredictthatforgettingshouldbelimitedtotheoriginallytrainedcue.Todistinguishbetweenthesemodels,weretestedsubjectsfromexperiment1withcuesnotpreviouslyencounteredintheexperiment(`independentprobes').Foreachitem,wecuedsubjectswithasemanticcategoryandtheinitialletteroftheresponseword(forexample,forordeal±roach,subjectsreceived`insectr___')andaskedthemtorecallthestudiedwordthat®tthecues.Onthisnewindependentprobe(IP)test,recallofsuppresseditemswasagainworsethanbaseline(Fig.1b),andtheimpairmentwashigherwhenrecallhadbeenavoidedmore.Theamountofforgettingdidnotdifferreliablyfromwhentheoriginallytrainedcuewasusedinthetest(forexample,ordealinthesameprobe(SP)testcondition,Fig.1a).This®ndingrulesoutassociativeinter-ferenceandunlearningandshowsthatimpairmentislocalizedtotheunwantedmemoryitself.ThisstronglysupportstheexistenceofaninhibitorycontrolmechanismItwasnecessarytoshowthatsubjectsdidnotrecallandthenwithholdsuppressionitemsonthe®naltest.Theymighthavedonesooutofconfusion,givenouremphasisonwithholdingresponsesduringthethink/no-thinkphase.Toaddressthis,weexpandedourtestinstructionstoemphasizethatsubjectsshouldrecallanitemtoeverycueregardlessofearlierinstructions,eveniftheywereguessing.Tofurtherencourageresponding,weofferedamonetaryrewardforcorrectanswers.Theseincentiveshadlittleimpactontheinhibitioneffect(Fig.1c,d).Impairmentwassigni®cantoverall,increasedwithrepetition,anddidnotvarywiththetypeoftestcue,againsupportingtheinhibitorycontrolview.Althoughtheinstruc-tionstoguessenhancedrecallofsuppressionandbaselineitemswhenthestudiedcuewasgiven(0.01),thelineartrendforinhibitiondidnotdifferreliablyacrossexperiments(Subjectsmighthaveguessedthehypothesisoftheexperimentandtriedtoconformtotheexpectationofimpairedmemorybywith-holdingitems.Toaddressthis,wealteredthe®naltestinstructions Ordeal AlternativeTrained cueIndependent cuePre-experimental (2)(3) Figure2Threemechanismsthatcanexplainimpairedrecallinthesame-probecondition,illustratedwithastimuluspair.Associativeinterferencepositsthatsuppressiontrainingleadssubjectstogeneratediversionarythoughts(1)tothetrainedcuethatinterfereduringlaterattemptstorecallthetarget.Unlearningassumesthatsuppressiontrainingweakensthecue±targetconnection(2).Thesuppressionhypothesisstatesthatsuppressiontraininginhibitsthetarget(3).Notethattestingthetargetwithanindependentcuecircumventsinterference(1)andunlearning(2).Anyimpairmentfoundwiththistestmaybeattributedtoeffectslocalizedtothetarget. letterstonatureNATUREVOL41015MARCH2001www.nature.com intermixed.Oneachtrial,a®xationcrossappearedfor200ms,followedbyastimulusmemberinthecentreofthescreen.Forrespondtrials,thestimuluswaspresentedforupto4s,oruntilthesubjectresponded,andsubjectshadtoreporttheresponseasquicklyaspossible.Forsuppressiontrials,thestimulusremainedonthescreenfor4s.Ifasubjectresponded,alouderrorbeepsounded.Trialswereseparatedbya400-msintertrialinterval.Suppressionandrespondtrialswereconductedondifferentwordpairs,with®vepairsparticipatingineachofthe0(baseline),1,8and16-repetitionconditionsforboththerespondandsuppressionconditions.Respondtrialson®llerpairswerealsoincludedsothat67%ofthetrialsinthethink/no-thinkphaserequiredaresponse,encouragingastrongmentalsettorespondthathadtobeoverridden,asingo/no-gotasks.Afterthethink/no-thinkphase,wetestedthesubjects'memoryforallofthewordpairsintwoways.Inbothtests,acueforeachwordpairwaspresentedinthecentreofthescreenforupto4s,oruntilsubjectsspoketheresponse.Pairsforthedifferentrespondandsuppressconditionswereintermixedpseudo-randomly,withtheconstraintthattheaveragetestpositionforeachconditionwasequated.Inthesame-probetest,subjects'recallforeachpairwascuedwiththestimulusmemberthatwaspairedwiththeresponsethroughouttheexperiment.Intheindependentprobetest,subjectswerecuedwiththecategorynameforeachexemplaralongwithits®rstletter.Ineachcase,subjectswereaskedtorecallthestudieditemthat®tthecuesandnottowithholdanyitems.Halfofthesubjectsineachexperimentgotthesame-probetest®rst,andhalfweregiventheindependentprobetest®rst.Subjectsinexperiment2weregiven25centsforeachcorrectanswer,uptoamaximumof$4.Subjectsinexperiment3weregivenaquestionnaireaftertheexperimentinwhichtheywereaskedabouttheirimpressionsofour(false)hypothesisthatmemorywouldimprovewithattemptstosuppressanitem.Received10October;accepted21December2000.1.Freud,S.inTheStandardEditionoftheCompletePsychologicalWorksofSigmundFreud1(ed.J.Strachey)117±128(Hogarth,London,1966).2.Chao,L.L.&Knight,R.T.HumanprefrontallesionsincreasedistractibilitytoirrelevantsensoryCog.Neurosci.Neuropsychol.1605±1610(1995).3.Dagenbach,D.&Carr,T.H.(eds)InhibitoryProcessesinAttention,Memory,andLanguage(Academic,SanDiego,1994).4.Smith,E.E.&Jonides,J.Storageandexecutiveprocessesinthefrontallobes.Science1657±16615.Hasher,L.&Zacks,R.T.Workingmemory,comprehensionandaging:Areviewandanewview.Psychol.Learn.Motiv.193±225(1988).6.Anderson,M.C.&Spellman,B.A.Onthestatusofinhibitorymechanismsincognition:Memoryretrievalasamodelcase.Psychol.Rev.68±100(1995).7.Bjork,R.A.inVarietiesofMemoryandConsciousness:EssaysinHonourofEndelTulving(edsRoediger,H.L.&Craik,F.I.M.)309±330(LawrenceErlbaumAssociates,Hillsdale,1989).8.Luria,A.R.HigherCorticalFunctioninMan(BasicBooks,NewYork,1966).9.Logan,G.D.&Cowan,W.B.Ontheabilitytoinhibitthoughtandaction:Atheoryofanactofcontrol.Psychol.Rev.295±327(1984).10.Posner,M.I.&Peterson,S.E.Theattentionsystemofthehumanbrain.Annu.Rev.Neurosci.25±42(1990).11.Knight,R.T.,Staines,W.R.,Swick,D.&Chao,L.L.Prefrontalcortexregulatesinhibitionandexcitationindistributedneuralnetworks.ActaPsychol.159±178(1999).12.Cohen,J.D.&Servan-Schreiber,D.Context,cortex,anddopamine:Aconnectionistapproachtobehaviorandbiologyinschizophrenia.Psychol.Rev.45±77(1992).13.Carter,C.S.,Botvinick,M.M.&Cohen,J.D.Thecontributionoftheanteriorcingulatecortextoexecutiveprocessesincognition.Rev.Neurosci.49±57(1999).14.Mayr,U.&Keele,S.W.Changinginternalconstraintsonaction:Theroleofbackwardinhibition.J.Exp.Psychol.Gen.4±26(2000).15.Casey,B.J.etal.AdevelopmentalfunctionalMRIstudyofprefrontalactivationduringperformanceofago-no-gotask.J.Cogn.Neurosci.835±847(1997).16.Garavan,H.,Ross,T.J.&Stein,E.A.Righthemisphericdominanceofinhibitorycontrol:Anevent-relatedfunctionalMRIstudy.Proc.NatlAcad.Sci.USA8301±8306(1999).17.deZubicaray,G.I.etal.Motorresponsesuppressionandtheprepotenttendencytorespond:AparametricfMRIstudy.Neuropsychologia1280±1291(2000).18.Sakagami,M.&Niki,H.Spatialselectivityofgo/nogoneuronsinthemonkeyprefrontalcortex.Exp.BrainRes.165±169(1994).19.Wegner,D.M.Ironicprocessesofmentalcontrol.Psychol.Rev.34±52(1994).20.Freyd,J.J.BetrayalTrauma:TheLogicofForgettingChildhoodAbuse(HarvardUniv.Press,Cambridge,MA,1996).21.Geiselman,R.E.,Bjork,R.A.&Fishman,E.L.Disruptedretrievalindirectedforgetting:Alinkwithposthypnoticamnesia.J.Exp.Psychol.Gen.58±72(1983).22.Conway,M.A,Harries,K.,Noyes,J.,Racsmany,M.&Frankish,C.R.Thedisruptionanddissolutionofdirectedforgetting:Inhibitorycontrolofmemory.J.Mem.Lang.409±430(2000).23.Norman,D.A.&Shallice,T.inConsciousnessandSelf-Regulation:AdvancesinResearchandTheory(edsDavison,R.J.,Schwardz,G.E.&Shapiro,D.)1±18(Plenum,NewYork,1986).24.MacDonald,A.W.,Cohen,J.D.,Andrew-Stenger,V.&Carter,C.S.Dissociatingtheroleofthedorsolateralprefrontalandanteriorcingulatecortexincognitivecontrol.Science1835±183825.Shimamura,A.P.Theroleoftheprefrontalcortexindynamic®ltering.Psychobiology207±21826.Rowe,J.B.etal.Theprefrontalcortex:Responseselectionormaintenancewithinworkingmemory.1656±1660(2000).27.D'Esposito,M.etal.FunctionalMRIstudiesofspatialandnonspatialworkingmemory.Cogn.Brain1±13(1998).28.Schacter,D.L.&Wagner,A.D.MedialtemporallobeactivationsinfMRIandPETstudiesofepisodicencodingandretrieval.7±24(1999).29.Schacter,D.L.Memoryandawareness.Science59±60(1998).30.Anderson,M.C.ActiveForgetting:Evidenceforfunctionalinhibitionasasourceofmemoryfailure.J.AgressionMaltreatmentTrauma(inthepress).AcknowledgementsTheresearchreportedherewassupportedbyagrantfromtheUSNationalInstituteofMentalHealth.CorrespondenceandrequestsformaterialsshouldbeaddressedtoM.C.A.(e-mail:mcanders@darkwing.uoregon.edu).Maskingunveilspre-amodalcompletionrepresentationinvisualsearchRobertRauschenberger&StevenYantisDepartmentofPsychology,TheJohnsHopkinsUniversity,Baltimore,Maryland21218,USA...............................................................................................................................Whenoneobjectispartlyoccludedbyanother,itsoccludedpartsareperceptually`®lledin',thatis,theoccludedobjectappearstocontinuebehinditsoccluder.Thisprocessisknownasamodalcompletion.Thecompletionofapartiallyoccludedobjecttakesabout200ms(ref.2),andpre-completioninformation(thatis,informationfrombeforeamodalcompletionhasoccurred)existsinthevisualsystemforthatduration.Ithasbeensuggested,however,thatobserverscannotmakeuseofthisinformation,evenwhenitisbene®cialtodoso:visualsearchforatargetthatappearstobepartlyoccluded,forexample,isslowerthanforatargetthatdoesnotundergoocclusion,despitebothtargetsbeingphysically.Hereweshowthatvisualsearchdoeshaveaccesstopre-completionrepresentations,butonlyforalimitedtimethatdependsonthesizeoftheoccludedregion.Earlyinvestigationsofvisualsearchfocusedondiscoveringtheelementaryfeaturesavailableinearlyvision7±12,whereasmorerecentworkhasdemonstratedthattheinputtovisualsearchismuchmorecomplexthanpreviouslyassumed4±6,13,14.Althoughitseemsthattheentrylevelforvision(thatis,entireobjectsorindividualfeatures)canbequitehighinmanycases,questionsremainaboutthenatureoftheinformationavailableatearlierstagesofprocessing.Inthecaseofamodalcompletion,forexample,itseemsthatvisualsearchreliesonapost-completionrepresentationevenwhenthisimpairssearch.This®ndingcanbeinterpretedinatleastthreeways.First,itcouldbethatthereisnopre-amodalcompletionstageinprocessing.Thisisunlikelybecauseotherstudieshaveshownthatpre-completioninformationisavailableforcertainperceptualjudgements.Second,itispossiblethatpre-completioninformationexistsonlyimplicitly,asan`ingredient'inthecomputationofacompletedrepresentation,andisnotexplicitlyavailableforallperceptualjudgements.Examplesofthisincludemonocularinformationduringbinocularrivalry,andveryhighspatialfrequencyinformation,bothofwhicharepresentin Table1Percentageofmeantarget-absenterrorratesExperimentNumberofitemsindisplay2468.............................................................................................................................................................................Mediumnotch(100-msSOA)11.417.720.312.3Mediumnotch(250-msSOA)3.54.65.53.0Smallnotch18.719.822.515.6Largenotch2.04.03.01.9.............................................................................................................................................................................