國(guó)外心理語言學(xué)

PSY369:PsycholinguisticsLanguageAcquisition第1頁AnnouncementsOn-lineBlackboardquizforchapter4isnowup.Youmaytakeit5times,topscorecountsIampushingExam2backaday.WewillhavetheexamonFeb.28.LanguagedevelopmentsectionincludesinformationfromChapter3,pages72-87Homework#2duetoday第2頁Thelanguageexplosionisnotjusttheresultofsimplesemanticdevelopment;thechildisnotjustaddingmorewordstohis/hervocabulary.Childismasteringbasicsyntacticandmorphologicalprocesses.Languageexplosioncontinues第3頁Take100utterancesandcountthenumberofmorphemesperutteranceSyntaxDaddycoming.Hi,car.Daddycarcomed.Twocaroutside.Itgettingdark.Allgoneoutside.Bye-byeoutside.#morphemes:3,2,4,3,4,2,2‘-ing’and‘-ed’separatemorphemes‘a(chǎn)llgone’treatedasasinglewordMLU=morphemes/utterances=20/7=2.86Meanlengthofutterance(MLU)inmorphemes第4頁P(yáng)roto-syntax(??)Holophrases(around1-1.5years)Single-wordutterancesmaybeusedtoexpressmorethanthemeaningusuallyattributedtothatsinglewordbyadultsLanguageexplosioncontinues“dog”

mightrefertothedogisdrinkingwaterTypicallyidiosyncratic,butsomeconventional/common(e.g.,indicatetheexistenceofanobject,requestrecurrenceofobjectorevent)OftencombinedwithintonationorgestureControversialclaim:Mayreflectadevelopingsenseofsyntax,butnotyetknowinghowtouseit(e.g.,seeBloom,1973)第5頁SyntaxRogerBrown(1973)proposed5stages(determinedbyMLUnotage)Stage1:Telegraphicspeech(MLU~1.75;around24months)ChildrenbegintocombinewordsintoutterancesLimitedtoasmallset(11)ofsemanticrelations(75%oftheirutterances)LanguageexplosioncontinuesChildrenintelegraphicspeechstagearesaidtoleaveoutthe‘littlewords’andinflections:e.g.“Mummyshoe”NOTMummy’sshoe“Twocat”NOTtwocatsDebate:learningsemanticrelationsorsyntactic(positionrules)“babysleep”agent&actionorNounVerbNomination“Thatball”Recurrence“Moreball”Nonexistence“Allgoneball”Agent&action“Daddyhit”Action&object“hitball”Agent&object“Daddyball”Action&locative“gostore”Entity&locative“booktable”Entity&attribute“bighouse”第6頁MorethantwowordsStages2through5Stage2(MLU~2.25)

begintomodulatemeaningusingwordorder(syntax)Modulationsfornumber,time,aspectGradualacquisitionofgrammaticalmorphemes(“-ing”,“-s”Laterstagesreflectgenerallymorecomplexuseofsyntax(e.g.,questions,negatives)LanguageexplosioncontinuesSyntaxRogerBrown(1973)proposed5stages第7頁InnatenessaccountsSemanticbootstrapping(e.g.,Pinker1984,1989)AcquiringSyntaxLearnedaccountsAcquiredfromthelinguisticinputfromtheenvironment(e.g.Bates,1979)ChildhasinnateknowledgeofsyntacticcategoriesandlinkingrulesChildlearnsthemeaningsofsomecontentwordsChildconstructssomesemanticrepresentationsofsimplesentencesChildmakesguessesaboutsyntacticstructurebasedonsurfaceformandsemanticmeaningSpeechtochildrenisnotimpoverished(Snow,1977):Child-directedspeechSimilarwordsoccurinsimilarlinguisticcontextsAcousticinformation(e.g.,prosody)mayprovidesyntacticcuesChildrenlearngrammarbymappingsemanticroles(agent,action,patient)ontogrammaticalcategories(subject,verb,object)第8頁MorphologyTypicallythingslikeinflectionsandprepositionsstartaroundMLUof2.5(usuallyin2yrolds)Kidsacquirethe“rules”forapplyingmorphologyWugexperiment(Berko-Gleason,1958)AcquiringMorphologyThispersonknowshowtorick.Shedidthesamethingyesterday.Yesterdayshe________.Typicallychildrensaythatshe“ricked.”第9頁AcquiringMorphologyAge(yrs)MorphemeExample(s)2PresentprogressiveIdriving2ArticlesAdog,thedoctor2PluralBalls2UncontractibleCopulaHeisasleep,am,are3ThirdpersonsingularHewantsanapple3FullprogressiveBe+ing,Iamsinging3RegularpasttenseShewalkedMorphology:orderofacquisitionisrelativelyconsistent(James&Kahn,1982)第10頁AcquiringMorphologyChildrensometimesmakemistakes.Myteacherholdedthebabyrabbits.YesSheholdedthebabyrabbits.Didyousayyourteacherheldthebabyrabbit?Whatdidyousayshedid?No,sheholdedthemloosely.Didyousayheldthemtightly?第11頁AcquiringMorphologyThisisungrammaticalintheadultlanguageShowsthatchildrenarenotsimplyimitatingInthiscase,whattheyproducesomethingthatisnotintheirinput.Childrensometimesmakemistakes.Myteacherholdedthebabyrabbits.第12頁Whydotheymakeerrorslikethese?Inthecaseathand,wehavewhatiscalledoverregularizationTheverbholdhasanirregularpasttenseform,heldBecausethisformisused,theregularpasttense--thatwith-ed--isnotfound(*hold-ed)AcquiringMorphologyChildrensometimesmakemistakes.Myteacherholdedthebabyrabbits.第13頁AcquiringMorphologyExamples:HortonhearedaWhoIfindedRenéeThealligatorgoedkerplunkThecaseofverbpasttense:Regularverbformsrequirenostoredknowledgeofthepasttenseform(wugtest)Pasttenseisaccomplishedbyapplyingapasttenserule(e.g.,add-ed)totheverbstemWithirregularverbssomethingmustbememorized第14頁StagesintheacquisitionofirregularinflectionsAcquiringMorphologyWithregularverbs,thedefaultform-edisusedWithirregulars,listsassociatingtheverbwithaparticularformofthepasttensehavetobememorized:Pasttenseis-twhenattachedtoleave,keep,etc.Is->wasDig->dugHas->hadThecaseofverbpasttense:第15頁AcquiringMorphologyStagesintheacquisitionofirregularinflectionstimeOnthefaceofit,learningthesemorphologicalquirksfollowsapeculiarpattern:Early:correctirregularformsareusedMiddle:incorrectregularformsareusedLate:correctformsareusedagain第16頁Memory&RulesWhydowefindthistypeofpattern?MemoryandrulesTheuseofoverregularizedformsstartsataroundthesamethatthatthechildisbeginningtoapplythedefault-edrulesuccessfullyEarly:Allforms--whetherregularorirregular--arememorizedMiddle:Theregularruleislearned,andinsomecasesoverappliedLate:Irregularsareusedbasedonmemory,regularsusetherule(theideaisthatifthewordcanprovideitsownpasttensefrommemory,thenthepasttenseruleisblocked)第17頁Memory&RulesWhydowefindthistypeofpattern?MemoryandrulesOtheraccountsMaratsos(2023)–frequencyexplanationItispossibletopredictwhichverbswillbesubjecttooverregularizationThemoreoftenanirregularformoccursintheinput,thelesslikelythechildistouseitasanoverregularizationThisisevidencethatsomepartofoverregularizationoccursbecauseofmemoryfailuresSomethingaboutirregularsisunpredictable,hencetheyhavetobememorized第18頁Whatkindof“teaching”dokidsget?Iflanguageislearned(andnotinnate),howdokidsdoit?Whatkindoffeedbackdotheyget?Positiveevidence:KidsheargrammaticalsentencesNegativeevidence:informationthatagivensentenceisungrammaticalPovertyofthestimulus

Claim:Positiveevidenceisnotsufficientforlearningalanguage.第19頁Whatkindof“teaching”dokidsget?Arethekidsevenawareofmistakes?Thechildrenareapparentlyawareofthefactthattheirformsarestrange:Parent:Where’sMommy?Child:MommygoedtothestoreParent:Mommygoedtothestore?Child:NO!Daddy,Isayitthatway,notyou第20頁Whatkindof“teaching”dokidsget?HowmuchPositiveEvidenceisthere(inChild-Directedspeech)?Estimated5000–7000utterancesadayBetween?and?arequestionsOver20%arenot“full”adultsentences(typicallyNounorPrepositionalphrases)Onlyabout15%havetypicalEnglishSVOformRoughly45%ofallmaternalutterancesbeganwithoneof17words(e.g.,“what”,“that”,“it”,“you”)Cameron-Faulkner,etal(2023)Sowhatkidsdohearmaybesomewhatlimited.第21頁Whatkindof“teaching”dokidsget?Negativeevidencecouldcomeinvariousconceivableforms.Uponhearing“Billacookieate”,anadultmightGrammarteacherparentfeedback?“ThesentenceBillacookieateisnotasentenceinEnglish,Timmy.NosentencewithSOVwordorderis.”MorerealisticNotunderstandLookpainedRephrasetheungrammaticalsentencegrammatically“Billatacookie.”第22頁Kidsresistinstruction…McNeill(1966)Child:

Nobodydon’tlikeme.Adult:

No,say‘nobodylikesme.’Child:

Nobodydon’tlikeme.

[repeatseighttimes]

Adult:

No,nowlistencarefully;say‘nobodylikesme.’Child:

Oh!Nobodydon’tlikesme.第23頁Kidsresistinstruction…Cazden(1972)(observationattributedtoJeanBerkoGleason)Child:

Myteacherholdedthebabyrabbitsandwepattedthem.Adult:

Didyousayyourteacherheldthebabyrabbits?Child:

Yes.Adult:

Whatdidyousayshedid?Child:

Sheholdedthebabyrabbitsandwepattedthem.Adult:

Didyousaysheheldthemtightly?Child:

No,sheholdedthemloosely.Sotheredoesn’tseemtobealotofexplicitnegativeevidence,andwhatthereisthekidsoftenresist

第24頁Negativeevidenceviafeedback?Dokidsget“implicit”negativeevidence?Doadultsunderstandgrammaticalsentencesandnotunderstandungrammaticalones?Doadultsrespondpositivelytogrammaticalsentencesandnegativelytoungrammaticalones?第25頁Brown&Hanlon(1970):Casestudyof“Adam”-lookedatthingsthatweresaidtohimbyadults,andwhathesaidtothemAdultsunderstood42%ofthegrammaticalsentences.Adultsunderstood47%oftheungrammaticalones.Adultsexpressedapprovalafter45%ofthe

grammaticalsentences.Adultsexpressedapprovalafter45%oftheungrammaticalsentences.Suggeststhatthereisn’talotofgoodnegativeevidence.Negativeevidenceviafeedback?第26頁Inaway,it’smootanyway…Oneofthestrikingthingsaboutchildlanguageishowfewerrorstheyactuallymake.Fornegativefeedbacktowork,thekidshavetomaketheerrors(sothatitcangetthenegativeresponse).Buttheydon’tmakeenoughrelevantkindsoferrorstodeterminethecomplexgrammar.Pinker,Marcusandothers,concludethatmuchofthisstuffmustbeinnate.Butthisisn’ttheonlyview.Thereisanongoingdebateaboutwhethertherearerules,orwhetherthesepatternsofbehaviorcanbelearnedbasedonthelanguageevidencethatisavailabletothekids第27頁Critical(sensitive)periods第28頁Critical(sensitive)periodsCertainbehaviorisdevelopedmorequicklywithinacriticalperiodthanoutsideofit.Thisperiodisbiologicallydetermined.Examples:Imprintinginducks(Lorenz,;Hess,1973)DucklingswillfollowthefirstmovingthingtheyseeOnlyhappensiftheyseesomethingmovingwithinthefirstfewhours(after32hoursitwon’thappen)ofhatchingBinocularcellsinhumansCellsinvisualsystemthatrespondonlytoinputfrombotheyes.Ifthesecellsdon’tgetinputfrombotheyeswithinfirstyearoflife,theydon’tdevelop第29頁Critical(sensitive)periodsSomeenvironmentalinputisnecessaryfornormaldevelopment,butbiologydetermineswhentheorganismisresponsivetothatinput.That“when”isthecriticalperiodCertainbehaviorisdevelopedmorequicklywithinacriticalperiodthanoutsideofit.Thisperiodisbiologicallydetermined.第30頁CriticalperiodforlanguageItassumesthatlanguageacquisitionmustoccurbeforetheendofthecriticalperiodEstimatesrangefrom5yearsuptoonsetofpubertyLenneberg(1967)proposedthatthereisacriticalperiodforhumanlanguage第31頁EvidenceforcriticalperiodforlanguageFeralChildrenChildrenraisedinthewildorwithreducedexposuretohumanlanguageWhatistheeffectofthislackofexposureonlanguageacquisition?TwoclassiccasesVictor,theWildBoyofAveyronGenie第32頁Victor,TheWildBoyofAveyronFoundin1800neartheoutskirtsofAveyron,FranceEstimatedtobeabout7-years-oldConsideredbysometobethefirstdocumentedcaseofautismNeitherspokeorrespondedtospeechTakentoandstudiedbyDr.Jean-Marc-GaspardItard,andeducatorofdeaf-muteandretardedchildrenNeverlearnedtospeakandhisreceptivelanguageabilitywaslimitedtoafewsimplecommands.DescribedbyItardas“analmostnormalboywhocouldnotspeak”

第33頁GenieFoundinArcadia,Californiain1970,wasnotexposedtohumanlanguageuntilage13.5.RaisedinisolationasituationofextremeabuseGeniecouldbarelywalkandcouldnottalkwhenfoundDr.SusanCurtissmadegreateffortstoteachherlanguage,andshedidlearnhowtotalk,buthergrammarneverfullydeveloped.Onlycapableofproducingtelegraphicutterances(e.g.MikepaintorApplesaucebuystore)Usedfewclosed-classmorphemesandfunctionwordsSpeechsoundedlikethatofa2-year-old第34頁GenieByageof17(after4yearsofextensivetraining)Vocabularyofa5yearoldPoorsyntax(telegraphicspeechmostly)ExamplesMamawashhairinsinkAtschoolscratchfaceIwantCurtissplaypianoLikegorideyellowschoolbusFathertakepiecewood.Hit.Cry.第35頁WhatDoTheseCasesTellUs?Suggestiveofthepositionthatthereisacriticalperiodforfirstlanguagelearning(inparticularforsyntaxandphonologicaldevelopment)Ifchildisnotexposedtolanguageduringearlychildhood(priortotheageof6or7),thentheabilitytolearnsyntaxwillbeimpairedwhileotherabilitiesarelessstronglyaffectedNotuncontroversial:VictorandGenieandchildrenlikethemweredeprivedinmanywaysotherthannotbeingexposedtolanguageGeniestoppedtalkingafterage30andwasinstitutionalizedshortlyafterward(Rymer,1993)第36頁WhatDoTheseCasesTellUs?Suggestiveofthepositionthatthereisacriticalperiodforfirstlanguagelearning(inparticularforsyntaxandphonologicaldevelopment)Why?Nativistexplanation(seepg79oftext)Maturationalexplanation:“l(fā)essismore”第37頁SecondlanguagelearningLearninganewlanguageWhatifwealreadyknowonelanguage,butwanttolearnanother?Adultslearninganotherlanguagetypicallyhaveapersistentforeignaccent–perhapsacriticalperiodforphonology(Flege&Hillenbrand,1984)Adultstypicallydobetterinitiallyatlearninganewlanguagecomparedtokids,butkidstypicallydobetteroverthelongterm(Krashen,Long,&Scarcella,1982)第38頁SecondlanguagelearningJohnsonandNewport(1989)NativeChinese/KoreanspeakersmovingtoUSTask:ListentosentencesandjudgewhethergrammaticallycorrectTestscoreAgeofarrival217R=-.87TestscoreAgeofarrival1740R=-.16第39頁SecondlanguagelearningJohnsonandNewport(1989)NativeChinese/KoreanspeakersmovingtoUSTask:ListentosentencesandjudgewhethergrammaticallycorrectConcludedthataroundtheageof16somethinghappensDifferentfactorsoperateonlanguageacquisitionbeforeandaftertheageof16BirdsongandMolis(2023)ReplicatedtheJohnsonandNewportstudyinSpanish/Englishspeakers.Didnotfindadiscontinuityaroundtheageof16第40頁EffectsoftheCriticalPeriodLearningalanguage:Under7years:perfectcommandofthelanguagepossibleAges8-c.15:PerfectcommandlesspossibleprogressivelyAge15-:ImperfectcommandpossibleButtheseclaimsarefarfromuniversallyaccepted第41頁Bilinguals&PolyglotsManypeoplespeakmorethanonelanguageTucker(1999)-multilingualsoutnumbermonolingualsWhatistheimpactofknowing/usingmorethanonelanguage?Factorsaffectingsecondlanguageacquisition?Whatdoesthelexiconlooklike?InterestingeffectsinbilingualsInterferenceCodeswitchingCognitiveadvantages第42頁SecondlanguageacquisitionContextsofchildhoodbilingualismSimultaneousBothlanguagesareacquiredatthesametimeVocabularygrowthofbilingualsissimilartothatofmonolingualsSomeaspectsofacquisitionmaybeslowed,butbyageof4typicallycaughtupDoesn’tseemtomatterwhetherlanguagesare“related”ornot(e.g.,English-FrenchversusEnglishJapanese)Canachieve“fluency”inbothlanguagesSequentialacquisitionThesecondlanguageislearnedafterafirstlanguageWhenthesecondlanguage(L2)isacquiredisimportantEarlyversuslatelearning(e.g.,seetheJohnsonandNewportstudy)第43頁SecondlanguageacquisitionFrequencyofusageofbothlanguagesHowoftenandinwhatcontextsdoyouusethetwolanguages“Useitorloseit”-languageattritionModeofacquisitionNativebilingualism-growingupinatwolanguageenvironmentImmersion-schoolingprovidedinanon-nativelanguageSubmersion-onelearnersurroundedbynon-nativespeakersLanguagedominanceeffectsRelativefluencyofL1andL2mayimpactprocessing第44頁Howdowerepresentlinguisticinformationinabilinguallexicon?ProbablydependsonmanyofthefactorsjustdiscussedLet’slookatsomemodelsandresearchfocusingonthesituationwhereL1isdominantrelativetoL2BilingualRepresentations第45頁ModelsofthebilinguallexiconsL1=FirstLanguageL2=SecondLanguagePotteretal(1984):SeparateStoresModels–separatelexiconsforeachlanguageL1L2CONCEPTSWordAssociationModelL1L2CONCEPTSConceptMediationModel第46頁ModelsofthebilinguallexiconsL1=FirstLanguageL2=SecondLanguagePaivio,Clark,&Lambert(1988):CommonStoresModels–wordsfrombothlanguagesinsamestoreL1&L2CONCEPTS第47頁RevisedHierarchicalModelL1L2conceptslexicallinksconceptuallinksconceptuallinksKroll&Stewart(1994)ProposedthatthefluencyofL2needstobeconsideredintheprocessingmodelTheresultsaremixed,supportingmorecomplexmodelsMaybedifferentindifferentbilingualsdependingonthingslikeageofacquisition,relativeproficiency,etc.第48頁InterestingeffectsinbilingualsInterferenceCodeswitchingCognitiveadvantages第49頁InterferenceDoesknowingtwolanguagesleadtointerference?Whenfound,interferenceisatmultiplelevels

Phonological-leastamountofinterferenceLexical-mixingwordsfromdifferentlanguagesInitially,appeartouseaonewordperthingstrategyButastheyrealizetherethatthey’respeakingtwolanguage,thenthey’llusewordsfrombothlanguagessimultaneouslySyntacticUntilyeartwo,mayuseonlyonesyntacticsystemwhichiscommontobothlanguagesThenabriefperiodwithtwosetsoflexicalitems,butstillacommonsyntaxFinally,twolexiconsandtwosetsofsyntaxInterestingeffectsinbilinguals第50頁InterestingeffectsinbilingualsDeterminewhoorwhatistheoneperformingtheaction.Thewaitresspushesthecowboys.Thetelephonespushesthecowboys.Kissesthetabletheapple.Thebasketstheteacherkicks.AsanativespeakerofEnglishwecanusemanycues:WordorderAnimacyVerbagreementNotalllanguagesusethesamecuestothesameextente.g.,Germandoesn’trelyasmuchonwordorder,butreliesmoreonagreementprocesses第51頁InterestingeffectsinbilingualsDeterminewhoorwhatistheoneperformingtheaction.Thewaitresspushesthecowboys.Thetelephonespushesthecowboys.Kissesthetabletheapple.Thebasketstheteacherkicks.Kilborn(1989,1994)Foundthatbilinguals(Englishassecondlanguage)typicallycarryoverthedominantprocessingstrategiesfromtheirnativelanguages.Thisinteractswiththeirleveloffluencyinthesecondlanguage第52頁CodeswitchingWhenbilingualssubstituteawordorphrasefromonelanguagewithaphraseorwordfromanotherlanguage

“IwantamotorcycleVERDE”Switchingissystematic,notrandom第53頁Whenbilingualssubstituteawordorphrasefromonelanguagewithaphraseorwordfromanotherlanguage“Iwantamotorcy