认知语言学

IntroductiontoCognitiveLinguistics•Thereisnoobjective,disembodiedtruth,andconsequentlytheworldisnotobjectivelyreflectedinlanguage.Languageismuchmorethanjustamirror,itdescribesourindividualandcollectiveexperiencesoftheworld.Conceptualandlinguisticuniversalsarisefromthefactthatwehavesimilarbodiesandbrains,thatweinhabitsimilarenvironmentsandthatwecommunicatewitheachother.•CognitiveLinguisticsisanewapproachtothestudyoflanguagethatemergedinthe1970’sasareactionagainstthedominantgenerativeparadigm(RuizdeMendoza1997).–Someofthemainassumptionsunderlyingthegenerativeapproachestosyntaxandsemanticsarenotinaccordancewiththeexperimentaldatainlinguistics,psychologyandotherfields.–E.g.,Mentalimages,generalcognitiveprocesses,basic-levelcategories,prototypephenomena,theuseofneuralfoundationsforlinguistictheoryandsoon,arenotconsideredpartofthesegrammars.•Aim:toexaminetherelationoflanguagestructuretothingsoutsidelanguage.•Threemainapproaches:–Experientialview–theProminenceview–theAttentionalviewoflanguage(UngererandSchmid,1997)1TheExperientialview•Thisviewpursuesamorepracticalandempiricaldescriptionofmeaning•Itistheuserofthelanguagewhotellsuswhatisgoingonintheirmindswhentheyproduceandunderstandwordsandsentences.•Thefirstresearchwithinthisapproach-thestudyofcognitivecategoriesledtotheprototypemodelofcategorization(EleanorRoschetal.,1977,1978)•Theknowledgeandexperiencehumanbeingshaveofthethingsandeventsthattheyknowwellistransferredtothoseotherobjectsandevents,whichtheyarenotsofamiliarwith,andeventoabstractconcepts.•Categorizationisbasictothought,perception,action,andspeech.•Everytimeweseesomethingasakindofathing,wearecategorizing.Notonlythings,wealsocategorizeactions(motoractivities).•Almosteverythingwedoinvolvescategories.Thinkofanyutterance:categoriesofspeechsounds,ofwords,ofphrasesandclauses,aswellasconceptualcategories.•“Thingsareinthesamecategoryifandonlyiftheyhavecertainpropertiesincommon.Thesepropertiesdefinethecategory.”(Classicalcategorizationtheory)•HumancategorizationisoneofthemajorissuesinLinguistics.Theabilitytocategorize,i.e.,tojudgethataparticularthingisorisnotaninstanceofaparticularcategory,isanessentialpartofcognition.Categorizationisoftenautomaticandunconscious,exceptinproblematiccases.•This“firstlevel”categorization(Brown)hasthefollowingproperties:•􀁻Itisthelevelofdistinctiveactions(wethinkofplayingwiththedog,notthequadruped).•􀁻Itisthelevelwhichislearnedearliestandatwhichthingsarefirstnamed.•􀁻Itisthelevelatwhichnamesareshortestandusedmostfrequently.•􀁻Itisanaturallevelofcategorization,asopposedtoalevelcreatedby“achievementsoftheimagination”.EleanorRosch’spioneeringworkmadecategorizationtheoryamajorfieldofstudy.•Thetheoryhascometobecalled“thetheoryofprototypesandbasic-levelcategories”or“prototypetheory”.•Roschfocusedontwoimplicationsoftheclassicalviewoncategorization:“Ifthecategoriesaredefinedonlybypropertiesthatallmembersshare,thennomembersshouldbebetterexamplesofthecategorythananyothermembers.Ifcategoriesaredefinedonlybypropertiesinherentinthemembers,thencategoriesshouldbeindependentofthepeculiaritiesofanybeingsdoingthecategorizing;thatistheyshouldnotinvolvesuchmattersashumanneurophysiology,humanbodymovement,andspecifichumancapacitiestoperceive,toformmentalimages,tolearnandremember,toorganizethethingslearned,andtocommunicateefficiently.•HerearlystudieswereoncolorcategorizationinspeakersofDani(aNewGuinealanguage).InDani,thereareonlytwobasiccolorterms:mili(fordarkcool,incl.black,green,andblue)andmola(forlight-warm,incl.white,red,yellow).Herhypothesiswasiflanguagealonedeterminedcolorcategorization,thentheDanishouldhaveequaldifficultylearningnewwordsforcolors,nomatterwhetherthecolorrangeshadprimarycoloratthecenteroranonprimarycolor.HerstudiesprovidedempiricalevidencethatprimarycolorcategorieswerepsychologicallyrealforspeakersofDani,andfocalcolorshada“specialcognitivestatus”,whichwouldcorrespondtowhatshewillcallcognitivereferencepointsorprototypesinherlaterresearch.•Amongtheexperimentalparadigmssheusedare:•􀁻Directrating:howgoodanexampleofacategory(e.g.•bird)variousmembersare(e.g.arobin,achicken,etc.)•􀁻Reactiontime:Subjectsaskedtopressabuttonto•answertrue/falsequestionsoftheform“An[example]is•a[categoryname]”(e.g.,“Achickenisabird.”)•􀁻Productionofexamples:Subjectswereaskedtolist•ordrawexamplesofcategorymembers.Itwasmore•likelythatmorerepresentativeexamplesweredrawn.•􀁻Asymmetryinsimilarityratings:Lessrepresentative•examplesareoftenconsideredtobemoresimilarto•morerepresentativeexamplesthantheconverse.Inone•study(involvingAmericans),subjectsconsideredMexico•tobemoresimilartotheUSthantheUSistoMexico.•Asymmetryingeneralization:Informationaboutarepresentativememberismorelikelytobegeneralizedtonon-representativemembersthanthereverse.Inonestudy(Rips1975)thesubjectsbelievedthatadiseasewasmorelikelytospreadfromrobinstoducksthanduckstorobins.•Roschandherassociatesfoundthatthepsychologicallymostbasiclevelisinthemiddleofthetaxonomichierarchies.•Theyfoundthatbasiclevelis:Thehighestlevelatwhichcategorymembershavesimilarlype