認(rèn)知心理學(xué)語(yǔ)言課件

1認(rèn)知心理學(xué)

CognitivePsychology

語(yǔ)言

Language張亞旭yxzhang@1認(rèn)知心理學(xué)

CognitivePsychology

語(yǔ)2為什么研究語(yǔ)言認(rèn)知...2為什么研究語(yǔ)言認(rèn)知...3RichInformationinWordsMintGrapheme&phonemem,in,tWordlengtheffecteat<treatRegularityeffectmint[mint]<pint[paint]grapheme-phonemecorrespondence,GPC字素——表征音素的字母或字母組合3RichInformationinWordsMint4RichInformationinWordsTreatBodyorsub-syllabicunitseatConsistencyeffecttreat,sweat,greatclean<treat4RichInformationinWordsTrea5RichInformationinWordsTreatmentMorphemetreat,mentLexicaldecisionspellcung flurbpair thrimnade真假 假真 假假sized<raked(frequencyeffect)5RichInformationinWordsTrea6Morphology形態(tài)學(xué)Morphemescanbeaddedtoderiveonewordfromanother派生sad-sadnesssad-sadlyortochangetheinflectionofaword(e.g.,numberortense)屈折變化walk-hewalks…walk–Yesterdayhewalked…6Morphology形態(tài)學(xué)Morphemescanb7RichInformationinWords怡Semantic&phoneticradicalsRegularityeffect抬

<怡

(哪些額外變量需要控制???)Consistencyeffect煌

<枝蝗、惶、凰；技、歧7RichInformationinWords怡8Phonemes音素/音位Thesound

unitsoflanguage200phonemeshavebeenidentified;46areusedinEnglishPhonemesaredefinedrelativetoaparticularlanguageThedistinctionbetweensoundsmustmakeadifferencebetweenwords8Phonemes音素/音位Thesoundunits9Phonology音韻學(xué)Phonemesareanimportantunitinspeechrecognitionandproduction,sincetheyarethesoundsthatmakeupwordsStudyofphonemes=phonologyInmany(butnotall)theoriesofspeechrecognition,thespeechinputistranslatedintophonemesbeforebeingmatchedagainstthewordsinthelexicon9Phonology音韻學(xué)Phonemesarean10PhoneticsandPhoneticFeaturesPhonetics:ThestudyofthesoundsthemselvesEachphonemecanbedescribedintermsofasetofphoneticfeatures,whichdescribethewaythatthesoundisarticulatedbythevocalorgansSoundsaremadebychangingtheshapeofthevocaltractinsomewayConsonantsaremadebyrestrictingtheflowofairthroughthevocaltractatsomepointVowelsaremadebychangingtheshapeofthevocaltractbutitisnotclosedDifferentialneuralmechanismsforprocessingconsonantsandvowels10PhoneticsandPhoneticFeatu11Caramazzaetal.(2000).Separableprocessingofconsonantsandvowels.Nature,403,428-43011Caramazzaetal.(2000).Sep12心理詞典(MentalLexicon)一種知識(shí)庫(kù)儲(chǔ)存詞匯的正字法(orthography)、語(yǔ)音(phonology)、語(yǔ)義(semantics)和句法(syntax)等信息所謂的正字法是使拼寫合乎標(biāo)準(zhǔn)的方法閱讀過(guò)程中，正字法、語(yǔ)音和語(yǔ)義等各種表征激活之間的相互關(guān)系，以及語(yǔ)義和語(yǔ)音激活的相對(duì)時(shí)間進(jìn)程(timecourse)12心理詞典(MentalLexicon)一種知識(shí)庫(kù)Levelt,J.M.W.,Roelofs,A.,&Meyer,A.S.(1999).Atheoryoflexicalaccessinspeechproduction.BehaviouralandBrainSciences,22,1-75.Levelt,J.M.W.,Roelofs,A.,14心理詞典的三個(gè)層次Lexemelevel:表征一個(gè)詞的詞形和音素Lemmalevel:表征一個(gè)詞的語(yǔ)法（如陰陽(yáng)性）和語(yǔ)義特征Conceptuallevel:表征一個(gè)詞所意指的概念層次間相互聯(lián)系，詞條之間通過(guò)語(yǔ)義相互聯(lián)系語(yǔ)義啟動(dòng)（semanticpriming）醫(yī)生

--->護(hù)士

vs.面包

--->護(hù)士詞匯判斷（lexicaldecision）14心理詞典的三個(gè)層次15漢字閱讀中的亞詞匯加工青(semantic)/猜(complex)/波(control)--->紫Zhou&Marslen-Wilson,1999,JEP:LMC15漢字閱讀中的亞詞匯加工青(semantic)/猜(com16動(dòng)詞——論元結(jié)構(gòu)

VerbArgumentStructureindicatesthetypeofsyntacticconstituentswithwhichaverbco-occurs,thegrammaticalrolesandthematicroles/題元角色(e.g.,Agent/施事,Recipient/受事,andTheme/主題)ofthoseconstituentsSusangavetheboyabanana.Peterrepairedacaryesterday.Johndepartedlastweek.16動(dòng)詞——論元結(jié)構(gòu)

VerbArgumentStruc17RichInformationinSentencesSyntacticinformationSyntacticcategory(noun,verb,etc.)MorphosyntacticinformationFunctionalwordsWordorder...SemanticinformationWordmeaning17RichInformationinSentence18RichInformationbeyondSentences她是去年生的孩子AB:她是去年生的孩子老張下午沒(méi)去開會(huì)，他去做手術(shù)了AB:老張下午沒(méi)去開會(huì)，他去做手術(shù)了Pragmaticinformation18RichInformationbeyondSent19PragmaticsisasubfieldoflinguisticswhichstudiesthewaysinwhichcontextcontributestomeaningPragmaticsencompassesspeechacttheory,conversationalimplicature,talkininteractionandotherapproachestolanguagebehaviorinphilosophy,sociology,andlinguistics.Itstudieshowthetransmissionofmeaningdependsnotonlyonthelinguisticknowledge(e.g.grammar,lexicon,etc.)ofthespeakerandlistener,butalsoonthecontextoftheutterance,knowledgeaboutthestatusofthoseinvolved,theinferredintentofthespeaker,andsoon.Inthisrespect,pragmaticsexplainshowlanguageusersareabletoovercomeapparentambiguity,sincemeaningreliesonthemanner,place,timeetc.ofanutterance.Theabilitytounderstandanotherspeaker'sintendedmeaningiscalledpragmaticcompetence.Pragmaticawarenessisregardedasoneofthemostchallengingaspectsoflanguagelearning,and,thoughitcanbetaught,oftencomesonlythroughexperience19Pragmaticsisasubfieldof20RichInformationbeyondSentencesA:我不想和你過(guò)了B:他是誰(shuí)20RichInformationbeyondSent

字詞/句子/篇章

聽覺 視覺語(yǔ)言理解

聽 閱讀語(yǔ)言產(chǎn)生

說(shuō) 寫 字詞/句子/篇章 聽覺 視覺22孤立狀態(tài)下的視覺詞匯識(shí)別

直接通達(dá)假設(shè)語(yǔ)義語(yǔ)音詞形燈/rose22孤立狀態(tài)下的視覺詞匯識(shí)別

直接通達(dá)假設(shè)語(yǔ)義語(yǔ)音詞形燈/r23孤立狀態(tài)下的視覺詞匯識(shí)別

間接通達(dá)假設(shè)語(yǔ)義語(yǔ)音詞形燈/rose23孤立狀態(tài)下的視覺詞匯識(shí)別

間接通達(dá)假設(shè)語(yǔ)義語(yǔ)音詞形燈/r24Direct-accesshypothesisdeepdyslexiadap,ish,larDual-routehypothesis24Direct-accesshypothesis25SemanticActivationinReadingChineseZhou,X.,&Marslen-Wilson,W.(1999).Phonology,orthography,andlexicalsemanticactivationinreadingChinese.JournalofMemoryandLanguage,41,579-606Zhou,X.,&Marslen-Wilson,W.(2000).TherelativetimecourseofsemanticandphonologicalactivationinreadingChinese.JournalofExperimentalPsychology:Learning,Memory,andCognition,26,1245–126525SemanticActivationinReadi26LexicalAmbiguityResolutionandWordRecognitioninSentenceContext在去醫(yī)院看望女朋友之前，小宋先去店里買了杜鵑鄭老伯埋怨屋里一點(diǎn)聲音也沒(méi)有，老伴準(zhǔn)備第二天去市場(chǎng)買杜鵑中心問(wèn)題語(yǔ)境作用的時(shí)間點(diǎn)When/how/whether26LexicalAmbiguityResolution27MultipleaccessmodelSelectiveaccessmodelReorderedaccessmodelIntegrationmodel公式公示攻勢(shì)工事公事宮室Zhang,Y.,Wu,N.,&Yip,M.(2006).LexicalambiguityresolutioninChinesesentencereading.InP.Li,LH.Tan,E.Bates,&O.J.L.Tzeng(Eds.),HandbookofEastAsianPsycholinguistics(Vol.1:Chinese,pp.268-278).Cambridge,UK:CambridgeUniversityPress27Multipleaccessmodel認(rèn)知心理學(xué)語(yǔ)言課件29ModularityofCognitionModulartheoryFodor,1983;Forster,1979TheautonomyofthelexicalprocessorInteractivetheoryMarslen-Wilson&Tyler,1980;Marslen-Wilson&Welsh,1978;McClelland&Elman,1986TheoutputofhighersubsystemscaninfluencecomputationoflowersubsystemsFodor,J.A.(1983).Themodularityofmind:anessayonfacultypsychology.Cambridge,Mass.:MITPress29ModularityofCognitionModul30Syntacticparsingandsentencecomprehension30Syntacticparsingandsenten31Thedefendant

examined...31Thedefendantexamined...32Garden-PathSentences①Thedefendant

examined

bythelawyerturnedouttobeunreliable.②Thedefendantthatwasexamined

bythelawyerturnedouttobeunreliable.③Theevidence

examined

bythelawyerturnedouttobeunreliable.④Theevidencethatwasexamined

bythelawyerturnedouttobeunreliable.32Garden-PathSentences①Thede33MinimalAttachmentPrinciple

最小附加原則Thedefendant

examined...Syntax-firstmodelGarden-pathmodelS SNPVP NPVP NPRRS:sentenceNP:nounphraseVP:verbphraseRR:reducedrelative33MinimalAttachmentPrinciple34MinimalAttachmentPrinciple

最小附加原則Theevidence

examined...Syntax-firstmodelGarden-pathmodelS SNPVP NPVP NPRRS:sentenceNP:nounphrase名詞短語(yǔ)VP:verbphrase動(dòng)詞短語(yǔ)RR:reducedrelative關(guān)系34MinimalAttachmentPrinciple35PPAttachmentAmbiguity

介詞短語(yǔ)附加歧義VPThemanpainted

thedoors

withnewbrushesbeforethefestival.

NPThemanpainted

thedoors

withlargecracksbeforethefestival.VP VPverbNPPP verbNP NPPP35PPAttachmentAmbiguity

介詞短語(yǔ)36Puttheappleonthetowelinthebox.Tanenhaus,M.K.etal.(1995).Integrationofvisualandlinguisticinformationinspokenlanguagecomprehension.Science,268,1632-163436Puttheappleonthetoweli認(rèn)知心理學(xué)語(yǔ)言課件38LateClosurePrincipleAfterMaryandJohnhugged

thewomanleft.38LateClosurePrincipleAfter39ModularityofSyntacticProcessingSerial,syntax-firstmodelsFrazier&Fodor,1978;Frazier&Rayner,1982;Friederici,2002Constraint-basedmodelsMacDonaldetal.,1994,PsychologicalReviewConcurrentmodelBoland,199739ModularityofSyntacticProc40Three-Phase

NeurocognitiveModel

AngelaD.Friederici,2002Phase1(100-300msaftertheonsetofthecriticalword)TheinitialsyntacticstructureisbuiltonthebasisofthesyntacticcategoryinformationofawordPhase2(300-500ms)Lexical-semanticandmorphosyntacticprocessesoccurwiththegoalofthematicroleassignment(題元角色指派)Phase3(500-1000ms)ThedifferenttypesofinformationareintegratedFriederici,A.D.(2002).Towardsaneuralbasisofauditorysentenceprocessing.TrendinCognitiveSciences,6,78–84.“Withsome450citationstodate(ISIWebofScience),the2002articlealonehasbeencitedonceeveryweekonaveragesinceitwenttopress.”(Steinhauer&Drury,2012,B&L)40Three-PhaseNeurocognitiveM41ConcurrentModel

JulieE.Boland,1997Althoughsyntacticconstraints,whenavailable,preventadoptionofaplausible,butsyntacticallyill-formed,interpretationsemanticanalysisispermittedtogetaheadofsyntacticanalysis,assigningthematicrolesanddevelopingaprovisionalinterpretationbeforeasyntacticcommitmenthasbeenmadeTheevidenceexamined…Boland,J.E.(1997).Therelationshipbetweensyntacticandsemanticprocessesinsentencecomprehension.LanguageandCognitiveProcesses,12,423–484.41ConcurrentModel

JulieE.Bo42Ferreira&Clifton,1986,JML;

Trueswelletal.,1994,JML;

Cliftonetal.,2003,JML①Thedefendant

examined

bythelawyerturnedouttobeunreliable.②Thedefendantthatwasexamined

bythelawyerturnedouttobeunreliable.③Theevidence

examined

bythelawyerturnedouttobeunreliable.④Theevidencethatwasexamined

bythelawyerturnedouttobeunreliable.42Ferreira&Clifton,1986,JM43咬死 獵人 的 狗咬死 獵人 的 雞咬死 狐貍 的 狗賣掉 獵人 的 狗VP N1 de N243咬死 獵人 的 狗44怠慢客人的孩子偏正囑咐患者的家屬動(dòng)賓撞倒肖明的車子偏正或動(dòng)賓44怠慢客人的孩子偏正45LimitedSyntacticParallelisminChineseAmbiguityResolution①粉刷|公寓|的|房間|之后，|小王|還|打掃|了|客廳|。

②粉刷|臟亂|的|房間|之后，|小王|還|打掃|了|客廳|。③訓(xùn)練|士兵|的|將軍|之后，|總司令|發(fā)表|了|簡(jiǎn)短|演說(shuō)|。

④訓(xùn)練|年輕|的|將軍|之后，|總司令|發(fā)表|了|簡(jiǎn)短|演說(shuō)|。Hsieh,Y.,Boland,J.E.,Zhang,Y.,&Yan,M.(2009).LimitedsyntacticparallelisminChineseambiguityresolution.LanguageandCognitiveProcesses,24(7/8):1227-1264.45LimitedSyntacticParallelis46Exp.2First-fixationdurations粉刷公寓的房間之后，小王還打掃了客廳訓(xùn)練士兵的將軍之后，總司令發(fā)表了簡(jiǎn)短的演說(shuō)46Exp.2First-fixationdurat47Exp.2Gazedurations47Exp.2Gazedurations48Exp.2

Regression-pathdurations48Exp.2Regression-pathdura49Exp.2Probabilityoffirst-passregressions49Exp.2Probabilityoffirst50Exp.2Totaltimes50Exp.2Totaltimes51你們這里有沒(méi)有30的卡賣100的？Themousewaseatenbythecheese.Sentenceinterpretationmaynotalwaysbedeterminedbythesyntax

(e.g.Bever1970;TownsendandBever2001;Ferreira2003;Kolketal.2003;KimandOsterhout2005;vanHertenetal.2005,2006;FerreiraandPatson2007)51你們這里有沒(méi)有30的卡賣100的？52Acentralquestioninthestudyofsentencecomprehensionwhethersyntacticprocessesaretemporallyand/orfunctionallypriortosemanticprocessessuchthatsemanticinterpretationofasentencemustbeestablishedontopofthesyntacticstructureofthatsentence52Acentralquestioninthest53Syntaxfirst,bothtemporallyandfunctionally?Yesserial,syntax-firstmodels(e.g.,Frazier&Fodor,1978;Friederici,2002)Noconstraint-basedmodelsconcurrentmodelunificationmodel(Hagoort,2003,2005)non-syntactocentric,dynamicmodel(Kuperberg,2007)53Syntaxfirst,bothtemporall54“Duringnormalon-lineprocessing,semanticintegrationofwordsdoesnottakeplaceforwordswhicharenotsyntacticallylicensed(havewrongwordcategory)”(Friederici&Weissenbor,2007,BrainResearch,p.54)Functionalprimacyofsyntactic(wordcategory)oversemanticprocessesFriederici,A.D.,&Weissenborn,Y.(2007).Mappingsentenceformontomeaning:Thesyntax-semanticinterface.BrainResearch,1146,50–58.54“Duringnormalon-lineproce55ERPApproach55ERPApproach56ERPEffectsofSemanticProcesses56ERPEffectsofSemanticProc57575858Hagoort,P.Hald,L.,Bastiaansen,M.,&Petersson,K-M.(2004).Integrationofwordmeaningandworldknowledgeinlanguagecomprehension.Science,304,438-441Hagoort,P.Hald,L.,BastiaanAsagreedupon,Janewastowakehersisterandherbrotheratfiveo’clockinthemorning.Butthesisterhadalreadywashedherself,andthebrotherhadevengotdressed.Janetoldthebrotherthathewasexceptionallyquick/slow.Asagreedupon,Janewastowa61Kutas,M.,&Federmeier,K.D.(2011).Thirtyyearsandcounting:FindingmeaningintheN400componentoftheevent-relatedbrainpotential(ERP).

AnnualReviewofPsychology,62,621-647Kutas,M.,&Hillyard,S.A.(1980).Readingsenselesssentence:brainpotentialsreflectsemanticincongruity.Science,207,203-20561Kutas,M.,&Federmeier,K.62ERPEffectsofSyntacticProcessesSyntacticcategoryviolations(E)LAN+P60062ERPEffectsofSyntacticPro63Hahne&Friederici,2002

DasEiswurdeimgegessen.應(yīng)該去掉im

(Theicecreamwasin-theeaten.)語(yǔ)法錯(cuò)誤63Hahne&Friederici,2002

Das64Hahne&Friederici,2002

DasEiswurdeimgegessen.

(Theicecreamwasin-theeaten.)64Hahne&Friederici,2002

Das65DistinctBrainResponsestoSyntacticandSemanticProcesses65DistinctBrainResponsesto66(L)ANtoSyntacticCategoryViolationsHagoort,Wassenaar,&Brown,2003CorrectDehouthakkerontweekdeijdeleschroef

opdinsdag.(ThelumberjackdodgedthevainpropelleronTuesday.)

SyntacticCategoryViolationDehouthakkerontweekdeijdeleschroeft

opdinsdag.(ThelumberjackdodgedthevainpropelledonTuesday.)66(L)ANtoSyntacticCategory676768ELANorLAN?Modality:AuditoryorvisualPresentationrateVisualcontrastSyntacticcategoryuniquenesspoint(prefixorsuffix)68ELANorLAN?Modality:Audito69MorphosyntacticViolationsagreementviolationsofperson,number,gender,ortenseLAN+P600（有的實(shí)驗(yàn)發(fā)現(xiàn)）e.g.,Osterhout&Mobley,1995;Silva-Pereyra&Carreiras,2007P600（其他實(shí)驗(yàn)發(fā)現(xiàn)）e.g.,Hagoort,2003;Nevinsetal.,200769MorphosyntacticViolationsag70Newmanetal.,2007,NeuroImageViolationsofregularpasttenseinflectionevokeaLAN左側(cè)前部負(fù)活動(dòng)*YesterdayIfrownatBilly.規(guī)則變化Butviolationsofirregularpasttenseinflectiondonot沒(méi)有LAN*YesterdayIspeaktohimfrankly.不規(guī)則變化LANreflectsrule-governedcompositionalprocessesatthesyntacticlevel,ratherthanthesyntactictenseviolationitselfLAN表示規(guī)則變化動(dòng)詞的……70Newmanetal.,2007,NeuroIm71LANcanalsobeevokedbyover-regularizationsofirregularverbs(Morris&Holcomb,2005)bringed

71LANcanalsobeevokedbyov72LANhasalsobeenexplainedasreflectingdomain-general,rule-governedprocessingofsymbols,includingnon-linguisticsequencing(Hoen&Dominey,2000,NeuroReport)72LANhasalsobeenexplained73Kim&Osterhout,2005*Themealwasdevouring…飯正在吞咽紅色詞引發(fā)P600,noN400N400反映語(yǔ)義，P600反映語(yǔ)法Kim,A.,&Osterhout,L.(2005).Theindependenceofcombinatorysemanticprocessing:Evidencefromevent-relatedpotentials.JournalofMemoryandLanguage,52,205-225.73Kim&Osterhout,2005*Theme認(rèn)知心理學(xué)語(yǔ)言課件75WhenSemanticsMeetsFailedSyntax跨語(yǔ)言差異75WhenSemanticsMeetsFailed76FunctionalPrimacyinGerman:abriefreviewDasBrotwurdegegessen. Correct(Thebreadwaseaten)DerVulkanwurdegegessen. Semantic(Thevolcanowaseaten)語(yǔ)義異常N400DasEiswurdeimgegessen. Syntactic(Theicecreamwasin-the

eaten)句法錯(cuò)誤ELANP600DasTürschlo?wurdeimgegessen. Double(Thedoorlockwasin-the

eaten)雙重錯(cuò)誤無(wú)N400Hahne&Friederici,2002,CognitiveBrainResearch76FunctionalPrimacyinGerman77N400responsestosemanticviolations

Thevolcanowaseaten.Hahne&Friederici,2002,CognitiveBrainResearch77N400responsestosemanticv78(E)LAN-P600tosyntacticviolations

Theice-creamwasin-theeaten.Hahne&Friederici,2002,CognitiveBrainResearch78(E)LAN-P600tosyntacticvio79(E)LAN-P600,butnoN400fordoubleviolations

Thedoorlockwasin-theeaten.Hahne&Friederici,2002,CognitiveBrainResearch79(E)LAN-P600,butnoN400for80ThisfindingofnoN400suggeststhatthefailedlocalphrasestructureconstructionblockslexical-semanticintegrationsupportingthefunctionalprimacyofsyntacticcategoryoversemanticprocessesHahne&Friederici,2002,CognitiveBrainResearch80ThisfindingofnoN400sugg81NoN400effectswerefoundfordoubleviolations,regardlessofModalityAuditory(Friedericietal.,2004;Hahne&Friederici,2002,Experiment1;Hahne&Jescheniak,2001;Iseletal.,2007)visual(Friedericietal.,1999)Sometaskdemandsoverallcorrectnessjudgment(e.g.,Hahne&Friederici,2002,Experiment1;Iseletal.,2007)probeverification(e.g.,Friedericietal.,1999)unlesssemanticintegrationwasforcedbyinstructions(inGerman:Hahne&Friederici,2002,Experiment2;

inEnglish:

Thierryetal.,2008)SyntacticcategoryuniquenesspointPrefix(e.g.,Hahne&Friederici,2002,Experiment1)suffix(Friedericietal.,2004)81NoN400effectswerefoundf82DasBuchwurdetrotzverpflanztvon…(Thebookwasdespitereplantedby...)Friedericietal.,2004,NeuroReportSuffix82DasBuchwurdetrotzverpfla83Lefauteuilquiestdansladort(Thechairwhichisintheissleeping)Iseletal.,2007French83Lefauteuilquiestdansla84Friederici&Weissenborn,2007,BrainResearch,p.54“Duringnormalon-lineprocessing,semanticintegrationofwordsdoesnottakeplaceforwordswhicharenotsyntacticallylicensed(havewrongwordcategory)”84Friederici&Weissenborn,2085Hetvrouwtjeveegdedevloermeteenoudebedeldegemaaktvantwijgen(Thewomansweptthefloorwithanoldbeggedmadeoftwigs)vandenBrink&Hagoort,2004Dutch85Hetvrouwtjeveegdedevloer86Thierryetal.(2008,NeuroImage)observedanN400effectforthecombinedviolationShakespeareanfunctionalshiftHowever,theirparticipantswereaskedtojudgesentencesformeaningfulness86Thierryetal.(2008,NeuroI認(rèn)知心理學(xué)語(yǔ)言課件88Aretherecross-linguisticdifferencesinsyntacticprimacy?Possiblyyes,andcouldhingeuponmorphology形態(tài)學(xué)InGermanandsomeotherIndo-Europeanlanguages,thesyntacticcategoryofawordcanbemarkedwithgrammaticalmorphology(e.g.,inflectionalorderivationalaffixes)orthographicconventions(suchascapitalizingallGermannouns)88Aretherecross-linguisticd89Incontrast,Chineseisanon-Indo-Europeanlanguagewithverylimitedinflections(e.g.,pluralinflectionofhumannouns;perfectiveandprogressivemarkers)Thereisnogrammaticalmorphologythateithermarkssyntacticcategoryormarkssyntacticfeaturessuchasperson,number(exceptpluralinflectionofhumannouns),gender,case,andtenseGrammaticalrelationsinChinesearenotgenerallycuedmorphologically;rathertheyarecuedbywordorder,functionwords(e.g.,ba[PREP]),andlexical/semanticcontent89Incontrast,Chineseisano90TheselinguisticpropertiesofChinesehaveleadtotheconjecturethatsentencecomprehensioninChinesereliesmoreonsemanticanalysescomparedtoIndo-Europeanlanguages(e.g.,Xu,1997;seeLi,1998;Yeetal.,2006)90Theselinguisticproperties設(shè)計(jì)師制作新衣，把布料裁了。 Correct

伐木工開采森林，把松樹裁了。語(yǔ)義SemanticViolation

設(shè)計(jì)師制作新衣，把裁了。句法 SyntacticViolation

伐木工開采森林，把裁了。雙重CombinedViolationYe,Luo,Friederici,&Zhou,2006,BrainResearch設(shè)計(jì)師制作新衣，把布料裁了。 Correct

伐木工開清潔工/把/大廈的/窗戶/全部/擦/了/一遍。Correct

清潔工/把/大廈的/窗戶/全部/贏/了/一遍。SemanticViolation

清潔工/把/大廈的/窗戶/全部/糖/了/一遍。CombinedViolation語(yǔ)法錯(cuò)誤，可能不會(huì)引發(fā)語(yǔ)義矛盾的N400Yu&Zhang,2008,NeuroReport清潔工/把/大廈的/窗戶/全部/擦/了/一遍。Correct女孩/把/新到的/長(zhǎng)裙/立刻/購(gòu)買/了/下來(lái)。 Correct

女孩/把/新到的/長(zhǎng)裙/立刻/欺詐/了/下來(lái)。 Semantic

女孩/把/新到的/長(zhǎng)裙/立刻/鈔票/了/下來(lái)。 Weak

女孩/把/新到的/長(zhǎng)裙/立刻/鐵鍬/了/下來(lái)。 StrongZhang,Y.,Yu,J.,&Boland,J.E.(2010).Semanticsdoesn'tneedaprocessinglicensefromsyntaxinreadingChinese.JournalofExperimentalPsychology:Learning,Memory,andCognition,36,765-781.女孩/把/新到的/長(zhǎng)裙/立刻/購(gòu)買/了/下來(lái)。 Correc94ViolationReferencePointsDasTürschlo?wurdeimgegessen.(Thedoorlockwasin-theeaten)94ViolationReferencePointsDa95TaskVerificationsentencejudgmentinWeak-StrongexperimentOverallacceptabilityjudgmentorothertasksinGermanandFrenchstudies95TaskVerificationsentenceju女孩/買/了/裙子/和/手套。 Correct

女孩/買/了/很/裙子/和/手套。 Syntactic句法

女孩/吃/了/裙子/和/手套。 Semantic語(yǔ)義

女孩/吃/了/很/裙子/和/手套。 CombinedZhang,Y.,Yu,J.,&Boland,J.E.(2010).Semanticsdoesn'tneedaprocessinglicensefromsyntaxinreadingChinese.JournalofExperimentalPsychology:Learning,Memory,andCognition,36,765-781.女孩/買/了/裙子/和/手套。 Correct

女孩/買/97很男人很淑女很德國(guó)《無(wú)極》之后，有人說(shuō)，陳凱歌不再霸王別姬。97很男人98Discussion&ConclusionSemanticintegrationproceedsevenwhensyntacticcategoryprocessingfailsduringChinesesentencereading,resultinginN400effectsforcombinedviolations,evenunderconditionsinwhichsemanticintegrationisnotforcedThereisnolanguage-universalfunctionalprimacyofsyntacticcategoryoversemanticprocessesInordertoaccountforallthecross-linguisticdata,aprocessingmodelmusthaveameansforsemanticintegrationinthefaceofseveregrammaticalviolations98Discussion&ConclusionSeman9999Friederici,A.D.(2011).Thebrainbasisoflanguageprocessing:Fromstructuretofunction.PhysiologicalReviews,91,1357–1392.Friederici,A.D.(2011).The101“Areviewoftheliteratureonthetimingofsyntacticinformationandsemanticinformationacrossthedifferentlanguagesrevealsthattheabsolutetimingofthesyntax-initialandotherprocessesmayvary,butthattheorderoftheseprocessesintimeisfixedacrossthedifferentlanguageswithsyntacticwordcategoryinformationbeingprocessedfirst”(Friederici,2011,p.1379)Friederici,A.D.(2011).Thebrainbasisoflanguageprocessing:Fromstructuretofunction.PhysiologicalReviews,91,1357–1392.101“Areviewoftheliterature102Bornkessel,I.,&Schlesewsky,M.(2006).Theextendedargumentdependencymodel:Aneurocognitiveapproachtosentencecomprehensionacrosslanguages.PsychologicalReview,113,787–821.Bornkessel-schlesewsky,I.,&Schlesewsky,M.(2008).Analternativeperspectiveon‘semanticP600’effectsinlanguagecomprehension.BrainResearchReviews,59,55–73.Bornkessel-schlesewsky,I.,&Schlesewsky,M.(2009).Theroleofprominenceinformationinthereal-timecomprehensionoftransitiveconstructions:Across-linguisticapproach.LanguageandLinguisticsCompass,3,19–58.102Bornkessel,I.,&Schlesews103103認(rèn)知心理學(xué)語(yǔ)言課件105TheUnificationModelHagoort,P.(2003).Howthebrainsolvesthebindingproblemforlanguage:Aneurocomputationalmodelofsyntacticprocessing.NeuroImage,20,S18–S29.Hagoort,P.(2005).OnBroca,brain,andbinding:Anewframework.TrendsinCognitiveScience,9,416–423.105TheUnificationModelHagoor106Theprocessingofanimacyintheinterpretationofclassifier-nouncombinations106TheprocessingofanimacyiThingsintheworld,andnounsdenotingthem,candifferinanimacyThingsintheworld,andnouns108RapiduseofanimacyinformationinthematicinterpretationSyntacticambiguityresolution(Trueswelletal.,1994;butseeCliftonetal.,2003;Ferreira&Clifton,1986)Thedefendant

examinedbythelawyerturnedouttobeunreliable.Theevidence

examinedbythelawyerturnedouttobeunreliable.108Rapiduseofanimacyinform109RapiduseofanimacyinformationinthematicinterpretationProcessingsyntacticallycomplexsentences(Maketal.,2002,2006;Traxleretal.,2002,2005)Themoviethatthedirectorwatchedreceivedaprizeatthefilmfestival.賓語(yǔ)關(guān)系從句109Rapiduseofanimacyinform110RapiduseofanimacyinformationinthematicinterpretationProcessingsyntacticallysimple,butsemanticallyanomaloussentences...雇傭保鏢...雇傭母雞Li,Shu,Liu,&Li,2006110Rapiduseofanimacyinform111RapiduseofanimacyinformationinthematicinterpretationAnimacyfunctionsasatypeofprominenceinformationthatinfluencestheestablishmentofinterpretiverelationsbetweenagentandpatientargumentseveninsimple,syntacticallyunambiguous,andbothsyntacticallyandsemanticallywell-formedsentences(e.g.,Philippetal.,2008;Roehmetal.,2004;seeBornkessel-SchlesewskyandSchlesewsky,2009forreview)111Rapiduseofanimacyinform112ResearchquestionIsanimacy生命度,abasicsemanticfeature,usedimmediatelyinrealtimeprocessingofwordcombinationswithoutthematicprocessingbeinginvolved?112ResearchquestionIsanimacyDesign&ExamplesCorrect汽車/趙慶豐/看見/一輛/黑色的。Car/QingfengZhao/hadseen/oneCL-liang(classifyinggroundvehicle)/black.

Congruencescore:

6.19(0.57)QingfengZhaohadseenablackcar.Within-categoryviolation臺(tái)燈/趙慶豐/看見/一輛/便宜的。Desklamp/QingfengZhao/hadseen/oneCL-liang(classifyinggroundvehicle)/cheap.

Congruencescore:1.82(0.52)QingfengZhaohadseenacheapdesklamp.Between-categoryviolation海豹/趙慶豐/看見/一輛/笨拙的。生命性違法Seal/QingfengZhao/hadseen/oneCL-liang(classifyinggroundvehicle)/clumsy.

Congruencescore:1.38(0.37)QingfengZhaohadseenaclumsyseal.Design&ExamplesCorrect114Participants&Procedure30participants400-mspresentationplus100-msISITask:Sentenceacceptabilityjudgment114Participants&Procedure30AnN400wouldbeobservedforanimacymismatch,ifanimacyisc