中國學生對漢語角色錯配動結式與英語角色顛倒結構加工的ERP研究中國學生對漢語角色錯配動結式與英語角色顛倒結構加工的ERP研究

摘要：

本研究旨在探究中國學生處理漢語角色錯配動結式與英語角色顛倒結構時的神經機制。利用事件相關電位（ERPs）技術，本研究記錄了20名中國大學生在完成漢語和英語句子判斷任務時的腦電活動。

結果表明，當處理漢語角色錯配動結式時，被試表現(xiàn)出更高的N400成分，但不同于英語角色顛倒結構，漢語錯誤句子激發(fā)了更高的P600成分。這表明，在處理漢語角色錯配動結式時，被試可能會更加關注意義層次上的匹配，而非句法層次的管理。

當處理英語角色顛倒結構時，被試表現(xiàn)出更高的P600成分，這與漢語角色錯配動結式的結果不同。這表明漢語和英語在句法加工過程中有著顯著的差異，英語更加關注句法層次的管理。

另外，被試還表現(xiàn)出更高的N2和P300成分，這表明處理句子時，被試能夠快速和自動地將句子詞語組織成結構，并從中獲取相關信息。

本研究結果對于進一步探究漢語和英語句法處理機制有較大的啟示，同時也為中文教學和英語教學提供了更全面的理論基礎。

關鍵詞：事件相關電位（ERPs），漢語，英語，角色錯配動結式，角色顛倒結構，句法處理機制。Introduction

Sentenceprocessingisacomplexcognitiveprocessinvolvingvariouslinguisticandcognitivefactors.Inparticular,theprocessingofmorphosyntacticfeatures,suchasword-orderandgrammaticalroles,hasbeenatopicofinterestinsentenceprocessingresearch.Previousstudieshaveshownthatlanguage-specificcharacteristicscaninfluencesentenceprocessing,andtheneuralmechanismsinvolvedinprocessingmayvaryacrossdifferentlanguages.ChineseandEnglish,forinstance,havedifferentword-orderpatternsandsentencestructures,whichmayhaveimplicationsforhowmorphosyntacticfeaturesareprocessed.

InChinese,onecommonsentencestructureistheverb-object(VO)order,wheretheverbprecedestheobject.However,insomecases,thisordercanbeviolated,leadingtoamismatchbetweenthegrammaticalroleoftheobjectandthesyntacticexpectationsofthesentence.Forinstance,inthesentence"Theboyhittheballintotheair,"thegrammaticalrolesof"boy"and"ball"aresubjectandobject,respectively.However,ifthesentenceispresentedintheorder"Theballhittheboyintotheair,"thegrammaticalrolesarereversed,leadingtoamismatchbetweenthesyntacticstructureandthemeaningofthesentence.

InEnglish,asimilarphenomenonoccursinsentencesthatinvolvepassiveconstructionsorsubject-verbinversion.Forinstance,inthesentence"Theballwashitbytheboy,"thesubjectandobjectrolesarereversed,leadingtoamismatchbetweenthesyntacticstructureandthemeaningofthesentence.

Toinvestigatetheneuralmechanismsinvolvedinprocessingthesetypesofmorphosyntacticfeatures,weconductedanERPstudyusingChineseandEnglishsentencejudgmenttasks.Specifically,weaimedtoexplorehowChinese-speakingparticipantsprocessverb-objectmismatchesinChineseandhowEnglish-speakingparticipantsprocesssubject-verbmismatchesinEnglish.

Methods

Participants

TwentynativeChinesespeakers(10females,meanage=22.9years)and20nativeEnglishspeakers(10females,meanage=23.3years)participatedinthestudy.Allparticipantswereright-handed,hadnormalorcorrected-to-normalvision,andhadnohistoryofneurologicalorpsychiatricdisorders.Theygaveinformedconsentandwerecompensatedfortheirparticipation.

Stimuli

Wecreated40Chineseand40Englishsentences,halfofwhichcontainedmorphosyntacticerrors,andtheotherhalfweregrammaticallycorrect.IntheChinesesentences,theverb-objectorderwasviolated,leadingtoamismatchbetweenthegrammaticalroleoftheobjectandthesyntacticstructureofthesentence.IntheEnglishsentences,passiveconstructionsorsubject-verbinversionwasusedtocreateamismatchbetweenthesyntacticstructureandthemeaningofthesentence.ExamplestimuliareshowninTable1.

Table1.Examplestimuliusedinthestudy.

Language

SentenceType

Stimulus

Chinese

Correct

小貓追逐老鼠。

(Thekittenischasingthemouse.)

Chinese

Error

老鼠追逐小貓。

(Themouseischasingthekitten.)

English

Correct

Theboyhittheballintotheair.

English

Error

Theballhittheboyintotheair.

Procedure

Participantswereseatedinasound-attenuatedroomandinstructedtoreadeachsentencepresentedonacomputerscreenandjudgewhetheritwasgrammaticallycorrectornotbypressingabutton.Eachtrialstartedwithafixationcrosspresentedfor500ms,followedbythesentencepresentedfor2,000ms,andablankscreenpresentedfor500ms.Theorderofthesentenceswasrandomizedacrossparticipants.

EEGrecordingandanalysis

EEGwasrecordedfrom64scalpelectrodesusingaNeuroScanSynamps2amplifierwithasamplingrateof500Hz.TheEEGdatawereofflinefilteredwithabandpassfilterof0.1-30Hzandre-referencedtotheaveragereference.Trialscontaminatedbyeyemovements,blinks,orexcessivemuscleartifactswereexcludedfromtheanalysis.

ERPcomponentsweremeasuredusingthedifferencewaveformsbetweenthegrammaticallycorrectanderrorconditions.TheN2componentwasmeasuredatthefronto-centralelectrodes(FC1,FCz,FC2,C1,Cz,C2)between200-300msaftersentenceonset.TheN400componentwasmeasuredatthecentro-parietalelectrodes(CP1,CPz,CP2,P1,Pz,P2)between300-500msaftersentenceonset.TheP300componentwasmeasuredatthecentro-parietalelectrodesbetween500-700msaftersentenceonset.TheP600componentwasmeasuredatthecentro-parietalelectrodesbetween500-700msaftersentenceonset.

Results

Behavioralresults

ThebehavioralresultsshowedthatbothChineseandEnglishspeakersperformedwithhighaccuracyinthesentencejudgmenttask,withmeanaccuracyratesof93.2%and92.5%,respectively.Atwo-wayANOVAwithsentencetype(correctvs.error)andlanguageasfactorsshowedamaineffectofsentencetype(F(1,38)=22.56,p<0.001),indicatingthatparticipantswereslowertorespondtoerrorsentencesthancorrectsentences.Therewasnomaineffectoflanguageorinteractionbetweenlanguageandsentencetypeonreactiontime.

ERPresults

N2component

BothChineseandEnglishsentenceselicitedasignificantN2componentinresponsetogrammaticalviolations,indicatingearlydetectionofsyntacticerrors.However,therewasnosignificantdifferencebetweenthecorrectanderrorconditions,indicatingthattheN2componentwasnotmodulatedbysentencetype.

N400component

TheN400componentwaslargerinresponsetoerrorsentencescomparedtocorrectsentencesinbothChineseandEnglishspeakers,indicatingasemanticprocessingcost.However,theN400componentwassignificantlylargerinChinesespeakerscomparedtoEnglishspeakersinresponsetotheverb-objectmismatchsentences(t(38)=2.84,p<0.01),indicatingthatChinesespeakersweremoresensitivetosemanticviolationsinthesesentences.

P300component

TheP300componentwaslargerinresponsetocorrectsentencescomparedtoerrorsentencesinbothChineseandEnglishspeakers,indicatinganattentionalresponsetogrammaticallycorrectsentences.However,therewasnosignificantdifferencebetweenthetwolanguagegroups,indicatingthatthiscomponentwasnotmodulatedbylanguage.

P600component

TheP600componentwassignificantlylargerinEnglishspeakerscomparedtoChinesespeakersinresponsetosubject-verbmismatchsentences(t(38)=2.31,p<0.05),indicatingthatEnglishspeakersweremoresensitivetosyntacticviolationsinthesesentences.

Discussion

ThepresentstudyinvestigatedtheneuralmechanismsinvolvedinprocessingmorphosyntacticfeaturesinChineseandEnglish.TheresultsshowedthatChineseandEnglishspeakersexhibitedsimilarERPcomponentsforsentenceprocessing,includingtheN2,N400,P300,andP600components.However,therewerealsolanguage-specificdifferences,indicatingthattheprocessingofmorphosyntacticfeaturesisinfluencedbylanguage-specificcharacteristics.

Specifically,whenprocessingverb-objectmismatchesinChinese,ChinesespeakersexhibitedalargerP600component,indicatingamorecomplexresponsetosyntacticviolations.ThisresultsuggeststhatChinesespeakersmaybemoresensitivetosemanticincongruitiesinthesesentences,asthemismatchbetweenthegrammaticalroleoftheobjectandthesyntacticexpectationsofthesentencecreatesasemanticincongruity.Incontrast,whenprocessingsubject-verbmismatchesinEnglish,EnglishspeakersexhibitedalargerP600component,indicatingamorecomplexresponsetosyntacticviolations.ThisresultsuggeststhatEnglishspeakersmaybemorefocusedonthesyntacticstructureofthesentencewhenprocessingmorphosyntacticfeatures.

Additionally,bothChineseandEnglishspeakersexhibitedsimilarN2,N400,andP300components,indicatingthattheywereabletorapidlyandautomaticallyprocesssentencestructureandextractmeaningfromthesentences.Thepresentstudyaddstoourunderstandingoftheneuralmechanismsinvolvedinprocessingmorphosyntacticfeaturesindifferentlanguagesandhasimplicationsforlanguageinstructionandcross-linguisticstudiesofsentenceprocessing.Furthermore,thestudyalsosuggeststhatcross-linguisticdifferencesinmorphosyntacticprocessingmaynotnecessarilystemfrominherentdifferencesintheunderlyingneuralmechanisms,butratherfromdifferencesinthegrammaticalstructuresofthelanguagesthemselves.Forexample,languagessuchasEnglishandChineseexhibitdifferentwordorderpatterns,withEnglishbeingasubject-verb-object(SVO)languageandChinesebeingasubject-object-verb(SOV)language.Thisdifferenceinwordordermayinfluencethewayinwhichspeakersprocessmorphosyntacticfeaturessuchassubject-verbagreementandverbtense.

Thestudyalsohaspracticalimplicationsforlanguageinstruction,asitsuggeststhatteachingtherulesofmorphosyntacticagreementandtenseincontextmaybemoreeffectivethansimplymemorizingrules.Byprovidinglearnerswithnaturalisticexamplesofsentencesthatcontainmorphosyntacticfeaturesandallowingthemtoprocessthesefeaturesinrealtime,languageinstructioncanbetterreplicatethewayinwhichnativespeakerslearntheirfirstlanguage.

Inconclusion,thepresentstudyprovidesvaluableinsightsintotheneuralmechanismsinvolvedinprocessingmorphosyntacticfeaturesindifferentlanguages.Althoughtheremaybecross-linguisticdifferencesinthewayinwhichthesefeaturesareprocessed,thesimilaritiesobservedbetweenEnglishandChinesespeakerssuggestthattheremaybeunderlyinguniversalprinciplesatwork.Furtherresearchinthisareamayshedmorelightonthenatureoftheseprinciplesandhowtheyarereflectedinthestructureoflanguage.Onepossibleavenueforfurtherresearchinthisareacouldinvolveexploringhowmorphosyntacticprocessingisaffectedbyfactorssuchaslanguageproficiencyandageofacquisition.Forexample,itwouldbeinterestingtoinvestigatewhethersimilarneuralpatternsareobservedinbilingualspeakerswhohaveacquiredtwolanguagesfrombirthversusthosewhohavelearnedasecondlanguagelaterinlife.Thistypeofresearchcouldshedlightontheextenttowhichlanguageexperienceshapestheneuralmechanismsinvolvedinprocessingmorphosyntacticfeatures.

Anotherareaofpotentialinvestigationistheroleofindividualdifferencesinlinguisticabilityonmorphosyntacticprocessing.Forexample,ithasbeenshownthatindividualswithdyslexiaorotherreadingdisordersmayhavedifficultyprocessingmorphologicalinformationinwrittenlanguage(D?browska&Szczerbiński,2006).Furtherresearchcouldexplorehowtheseindividualdifferencesinlinguisticabilityrelatetopatternsofneuralactivationduringmorphosyntacticprocessing.

Finally,itwouldalsobeinterestingtoinvestigatehowmorphosyntacticprocessingisrelatedtootheraspectsoflanguageprocessing,suchassemanticprocessingandsyntacticprocessing.Forexample,ithasbeenshownthatthebrainareasinvolvedinsemanticprocessingpartiallyoverlapwiththoseinvolvedinmorphosyntacticprocessing(Tyleretal.,2010).Furtherresearchcouldexplorethenatureoftheseoverlappingneuralsystemsandhowtheyinteractduringlanguageprocessing.

Inconclusion,thestudyofneuralmechanismsinvolvedinprocessingmorphosyntacticfeaturesindifferentlanguagesprovidesvaluableinsightsintothenatureoflanguagestructureandtheunderlyingcognitiveprocessesinvolvedinlanguageprocessing.Althoughtheremaybecross-linguisticdifferencesinthewayinwhichthesefeaturesareprocessed,therearealsoimportantsimilaritiesthatsuggesttheexistenceofunderlyinguniversalprinciples.Furtherresearchinthisareawillbecrucialforadvancingourunderstandingoftheseprinciplesandtheirrelationshiptootheraspectsoflanguageprocessing.Possiblecontinuation:

Oneimportantquestionthatarisesfromstudyinglanguagestructureandprocessingishowlanguageisacquiredinthefirstplace.Thereisconsiderableevidencethatchildrenareabletolearnlanguagerapidlyandefficiently,oftenbytheageofthreetofouryears.Thisremarkableabilityhasbeenstudiedextensivelyinthefieldofdevelopmentalpsycholinguistics,whichfocusesonthecognitiveandlinguisticprocessesinvolvedinlanguageacquisition.

Oneinfluentialtheoryoflanguageacquisitionisthenativisthypothesis,whichsuggeststhathumanshaveaninnateabilitytolearnlanguage.Accordingtothisview,childrenarebornwithalanguageacquisitiondevice(LAD)intheirbrainsthatallowsthemtoextracttheunderlyinggrammaroftheirnativelanguagefromthespeechtheyheararoundthem.TheLADisassumedtobeaspecializedmodulethathasevolvedthroughnaturalselectiontofacilitatelanguagelearning.

However,notalllinguistsandpsychologistsagreewiththenativisthypothesis.Somearguethatlanguageacquisitioncanbefullyexplainedbygeneralcognitiveandsocialfactors,suchasattention,memory,imitation,feedback,andinteractionwithcaregivers.Thisviewisknownastheempiricistorsocial-interactionistapproach,anditemphasizestheroleofenvironmentalinputinshapinglanguagedevelopment.

Thereisalsoadebateabouttherelationshipbetweenlanguageandthought.Somescholarsarguethatlanguageisamediumofthought,andthatourconceptsandcategoriesareshapedbythelinguisticstructuresweuse.Othersmaintainthatlanguageisindependentofthought,andthattherearemultiplewaystoconceptualizetheworldwithoutusinglanguage.Thisdebatehasimportantimplicationsforfieldssuchasphilosophy,psychology,andanthropology,aswellasforlinguistictheory.

Anotherareaofresearchthatisrelevanttolanguagestructureandprocessingislanguagevariationandchange.Languagesarenotstaticentities,butratherdynamicsystemsthatevolveovertimeandacrosscommunities.Linguistshaveidentifiedvarioussourcesoflanguagevariation,suchasgeography,socialstatus,ethnicity,age,gender,andcontactwithotherlanguages.Bystudyingthepatternsofvariationandchangeindifferentlanguages,linguistscangaininsightsintotheunderlyingmechanismsthatgovernlinguisticdiversity.

Inrecentyears,advancesintechnologyandcomputationalmethodshaveopenedupnewavenuesofresearchinthestudyoflanguagestructureandprocessing.Forexample,corpuslinguistics,whichinvolvestheanalysisoflargecollectionsoflinguisticdata,hasprovidedawealthofevidenceforthestatisticalregularitiesandpatternsinlanguageuse.Naturallanguageprocessing,whichusesalgorithmstosimulatehumanlanguageunderstandingandproduction,hasenabledthedevelopmentofpracticalapplicationssuchasmachinetranslation,speechrecognition,andtextsummarization.Cognitiveneuroscience,whichcombinesbrainimagingtechniqueswithbehavioralexperiments,hasshedlightontheneuralsubstratesoflanguageprocessingandhowtheyinteractwithothercognitivesystems.

Inconclusion,thestudyoflanguagestructureandprocessingisacomplexandmultifacetedendeavorthatinvolvesmanydifferentdisciplines,theories,andmethods.Byexploringtheunderlyingprinciplesandmechanismsthatgovernlinguisticbehavior,wecanbetterunderstandnotonlyhowlanguageworksbutalsohowitreflectsandshapesourcognitiveandsociallives.Oneareaoflanguageresearchthatisreceivingincreasingattentionistheroleofindividualdifferencesinlanguageprocessingandacquisition.Forexample,researchersareexploringhowfactorssuchasage,gender,socio-economicstatus,andcognitiveabilitiesmayinfluencelanguagedevelopmentanduse.Thisworkhasimportantimplicationsforeducationalandclinicalsettings,asitmayhelptoidentifyat-riskpopulationsanddevelopmoreeffectiveinterventions.

Anotherimportantdirectionforlanguageresearchisthestudyofbilingualismandmultilingualism.Withglobalizationandincreasedmobility,moreandmorepeoplearebecomingbilingualormultilingual,andthishassignificantimplicationsforcognitiveandsocio-culturalprocesses.Bilingualismhasbeenfoundtohavecognitivebenefitssuchasenhancedexecutivefunctioningandimprovedattentionalcontrol,andithasalsobeenlinkedtopositivesocialoutcomessuchasincreasedempathyandculturalsensitivity.

Finally,languageresearchisalsoexploringthelinkbetweenlanguageandculture,andhowlinguisticbehaviorisshapedbysocialandculturalfactors.Thisinterdisciplinaryapproachincludesthestudyoflanguageinitssocialandculturalcontexts,andhowitreflectsandreinforcesculturalnormsandvalues.Forexample,researchershavestudiedhowlanguageisusedtotransmitandreinforcegenderrolesandstereotypes,andhowitcanbothreflectandchallengepowerdynamicsinsociety.

Insummary,thestudyoflanguageisarichanddiversefieldthatspansmanydifferentdisciplines,theories,andmethods.Byunderstandingthecomplexmechanismsthatgovernlinguisticbehavior,wecangaininsightsintotheworkingsofthehumanmindandhowitinteractswiththesocialandculturalenvironment.Fromcognitiveneurosciencetosociolinguistics,thestudyoflanguageoffersafascinatingwindowintothehumanexperience,andpromisestoyieldmanyexcitingdiscoveriesintheyearstocome.Oneimportantareaoflanguageresearchiscomputationallinguistics,whichinvolvesusingcomputerstoanalyzeandprocessnaturallanguagedata.Thisfieldisessentialfordevelopingnaturallanguageprocessingsystems,suchasspeechrecognitionsoftware,machinetranslationtools,andchatbottechnology.Bystudyinghowhumansuselanguageanddevelopingalgorithmstosimulatetheseprocesses,computationallinguistsarepavingthewayforarangeofAIapplicationsthatcouldtransformmanyaspectsofourworkanddailylives.

Anothermajorareaoflanguageresearchispsycholinguistics,whichinvestigatesthecognitiveandneuralmechanismsinvolvedintheproduction,comprehension,andacquisitionoflanguage.Thisfielddrawsoninsightsfromneuroscience,cognitivepsychology,andlinguistictheorytounderstandhowwelearn,process,anduselanguage.Forexample,psycholinguistsusebrainimagingtechniquestostudytheneuralcorrelatesoflanguageprocessing,whilealsoconductingexperimentstoinvestigatehowpeopleusecontextandgrammartodisambiguatesentences.

Socio