资源描述
Transformational-GenerativeGrammar,TheBirthofTGGrammarAndtheInnatenessHypothesis,1.Founder:AmericanlinguistNoamChomsky,FullnameAvramNoamChomskyBorn7December1928(1928-12-07)Philadelphia,PennsylvaniaEra20th/21st-centuryphilosophyRegionWesternPhilosophySchoolLinguistics,Analytic,MaininterestsLinguisticsPsychologyPhilosophyoflanguagePhilosophyofmindPoliticsEthicsNotableideasGenerativegrammar,universalgrammar,transformationalgrammar,governmentandbinding,X-bartheory,Chomskyhierarchy,context-freegrammar,principlesandparameters,theminimalistprogram,languageacquisitiondevice,povertyofthestimulus,Chomsky-Schtzenbergertheorem,ChomskyNormalForm,propagandamodel,2.Themarkofthebeginning,ThepublicationofSyntacticStructuresin1957InSyntacticStructures,Chomskytriestoconstructaformalizedtheoryoflinguisticstructureandplacesemphasisonrigorousformulationsandpreciselyconstructedmodels.,3.Fivestagesofdevelopment,1)TheClassicalTheory,whichaimstomakelinguisticsascience.SyntacticStructures,19752)TheStandardTheory,whichdealswithhowsemanticsshouldbestudiedinalinguistictheoryAspectsoftheTheoryofSyntax,19653)TheExtendedStandardTheory,whichfocusesdiscussiononlanguageuniversalsanduniversalgrammarStudiesonSemanticinGenerativeGrammar,19724)TheRevisedExtendedStandardTheory(orGB),whichfocusesonGovernmentandBindingLecturesonGovernmentandBinding,19815)TheMinimalistProgram,whichisafurtherrevisionoftheprevioustheory.TheMinimalProgram,1995MinimalistInquiries:TheFramework,1998,4.TheInnatenessHypothesis(WhyshouldtherebeaTGGrammar?),WhenasastudentofHebrewwiththestructuralistmethodology,hefoundthattheclassificationofstructuralelementsoflanguageaccordingtodistributionandsubstitutionhaditslimitationhebelievesthatlanguageissomewhatinnate,andthatchildrenarebornwithwhathecallsaLanguageAcquisitionDevice,whichisauniquekindofknowledgethatfitsthemforlanguagelearning.Accordingtohim,thestudyoflanguage,orthestructureoflanguage,canthrowsomelightonthenatureofthehumanmind.,Hypothesisisbasedonsomeimportantfacts:1).Childrenlearntheirnativelanguageveryfastandwithlittleeffort.2).Childrenlearntheirmothertongueinverydifferentenvironments,buttheyfollowmoreorlessthesamestagesinacquisition:thebabblingstagenonsensewordstageholophrasticstagetwo-wordutterancedevelopinggrammarnear-adultgrammarfullcompetence,3)Thechildrenlearnsthetotalgrammarofthelanguageduringalimitedperiodoftime,fromlimitedexposuretospeech(whichisoftendegeneratedata)Allthesesuggestedthatalthoughbabiesarenotbornknowingalanguage,theyarebornwithaprepositiontodevelopalanguage.Ifchildrenarepredisposedforlearninganylanguage,humanlanguagesintheworldmusthavethesameunderlyingprinciplesincommon.ThisiswhatChomskycalledlanguageuniversalsorlinguisticuniversals.,TheClassicalTheoryandtheStandardTheory,1.HowtobeginthestudyofLanguageUniversal?,1)Chomskybelievesthateveryspeakeroflanguagehasmasteredandinternalizedagenerativegrammarthatexpresseshisknowledgeofhislanguage.Andthegenerativegrammar,means“asystemofrulesthatinsomeexplicitandwell-definedwayassignsstructuraldescriptionstosentences”.2)Inordertoreachhisultimategoal,Chomskyputsforwardthreedifferentlevelstoevaluategrammarson:theobservationaladequacylevelthedescriptiveadequacyleveltheexplanatoryadequacy3)ContrarytoBloomfieldsdata-orienteddiscoveryprocedure,ChomskyinsistsontheHypothesis-deductionmethodandhisresearchiscalledevaluationprocess,4)Whatisagenerativegrammar?Byagenerativegrammar，Chomskysimplymeans“asystemofrulesthatinsomeexplicitandwell-definedwayassignsstructuraldescriptionstosentences”.Hebelievesthateveryspeakerofalanguagehasmasteredandinternalizedagenerativegrammarthatexpresseshisknowledgeofhislanguage.“Thusagenerativegrammarattemptstospecifywhatthespeakeractuallyknows，notwhathemayreportabouthisknowledge.”(Chomsky,1965)Togenerateistopredictwhatcouldbethepossiblesentencesoflanguage.Transformationisakindofprocessthattransformsonesentenceintoanother.Itusesafinitenumberofrulestoproduceinfinitenumberofsentences.,2.TheclassicalTheory,1)Itaimstomakelinguisticsascience:2)Itsfeatures:emphasisongenerativeabilityoflanguageintroductionoftransformationalrulesgrammaticaldescriptionregardlessofmeaning3)Chomskyputforwardthreekindsofgrammar:finitestategrammarphrasestructuregrammartransformationalgrammar,Threekindsofgrammar,Thefinitestategrammarisappropriatetodescribeprocessoflearning,butitisinappropriatetoexplainthecomplexitiesofthehumancognitivesystem.Itcontainsdeepstructureandsurfacestructure.Deepstructure:theabstractrepresentationofthesyntacticpropertiesofaconstruction.Surfacestructure:thefinalstageinthesyntacticderivationofaconstruction,Thephrasestructurerulesareasfollows:a.SNP+VPb.VPVerb+NPc.NPNP(single)NP(plural)d.NP(s)D+Ne.NP(p)D+N+sf.Dtheg.Nman,ball,door,dog,book,h.Verbaux+Vi.Vhit,take,bite,eat,walk,j.AuxTense(+M)(+have+en)(+be+ing)k.TensePresentPastl.Mwill,can,may,shall,must,NPArtNPPPNPArtNPPPNPArtNPPPNPArtNPP,Thehouseinthewoodsbythemountainsneartheriver,embedding,Transformationalrules(T-rules)aretheoperationsthatadd，deleteorchangeelementsinonestructuretoproduceanotherstructure.Itreferstoalltherulesthatapplyafteralllexiconitemshavebeeninsertedintothestructures.TwoobligatoryT-rules:Flip-floprule(FFrule，affix-transformational/shifting/hoppingrule):theaffixandverbareshiftedinorder.2.Lexicalformationrule(LFrule):Itrequiresthelexiconthatdeterminethelastappearanceofthesentence.,TheGenerativeProcessofASentence(NP(Det(the)N(man)VP(V(hit)NP(Det(the)N(ball)SNPVPDetNVerbNPDetNthemanhittheball,4)SomedefectsinthetheoryThetransformationalrulesaretwopowerful.Therulesmaygenerateill-formedsentencesaswellaswell-formedones.Forexample,withtherulesSNP+VP,andVPV+NP,theremightbegeneratedthefollowingtwo:SSNPVPNPVPVNPVNPJohndrinkwinewinedrinkJohnThetransformationalrulesforthepassivevoicecannotbeusedatwill,forsomeoftheEnglishverbsdonothavepassivestructures.,SNPAuxVPDetNPPThemotheroftheboyandthegirlwillarrivesoon.SNPAuxVPNPNPThemotheroftheboyandthegirlwillarrivesoon.,phrasestructureistooflexible,3.ThestandardTheory,1)Chomskysaysthatagenerativegrammarshouldconsistofthreecomponents:syntactic,phonologicalandsemantic.Therelationshipsofthethreecomponentscanbeillustratedbythefollowingdiagram:,BaseComponent,RewritingRules,Lexicon,DeepStructure,T-Rules,SurfaceStructure,SemanticComponent,PhonologicalRules,SemanticRepresentation,PhonologicalRepresentation,Transformationalrules(T-rules)aretheoperationsthatadd，deleteorchangeelementsinonestructuretoproduceanotherstructure.Itreferstoalltherulesthatapplyafteralllexiconitemshavebeeninsertedintothestructures.TwoobligatoryT-rules:1.Flip-floprule(FFrule，affix-transformational/shifting/hoppingrule):theaffixandverbareshiftedinorder.2.Lexicalformationrule(LFrule):Itrequiresthelexiconthatdeterminethelastappearanceofthesentence.,2)ItsadvantagesovertheClassicalTheoryTransformationcanonlychangetheformsofsentencesandarenotallowedtoalterthemeaning.ToruleoutthegenerationofsentenceslikeWinedrinksJohn,thereisnowaselectionalrestrictiontoensurethattheanimatenounappearsbeforetheverbandinanimatenounappearaftertheverb.Restrictionsareputontransformationsinordernottogenerateill-formedsentences.Accordingtore-writingrules,sentencescanbeembedded,thus,thetheorynowcoversnotonlysimplesentencesbutalsocomplexones.Theruleareproperlyorderedandthereisasetorderinwhichtherulesapply.,3)ProblemsunsolvedinthetheoryThetransformationalrulesarestilltoopowerful,fortheycanmoveordeletelinguisticsegments,changethecategories,keeptheoriginalmeaningintact,andaccordingtospecificcircumstances.Derivednounsandtheircorrespondingverbsareirregularnotonlyintermsofsyntacticfeatures,butalsoinphonologicalandsemanticrelations,whichisignoredintheStandardTheory.,TheStandardTheoryholdsthatsemanticinterpretationsaredeterminedbythedeepstructure,andtransformationalprocesswillnotchangethesentencemeaning.Laterthiswasfoundtobeimpossible.TheStandardTheorycannotexplaingappedstructuresas:Johnatesomespaghetti,andMarysomemacaroni.Investigationsofmoretypeofrulesshowedthatmanytransformationalrulesmusthavecomplexconstraintsinorderthattheydonotproduceungrammaticalsentences.,.TheExtendedStandardTheoryandLaterTheories,1.TheExtendedStandardTheory,1)IntheExtendedStandardTheory,ChomskyrevisedhisStandardTheorytwice.2)Inhisfirstrevision,Chomskymovedpartofsemanticinterpretationtothesurfacestructure.3)Thesecondrevisioninvolvesthewholetheoreticalframework,whichcanbeillustratedbythefollowingdiagram.*ThemostremarkablechangeisthatChomskynowcompletelyputssemanticinterpretationinthesurfacestructure.,Phrasestructure;LexiconTheTransformationalComponentPhonologicalRulesSemanticInterpretativeRules,TheBaseComponent,DeepStructure,SurfaceStructure,PhonologicalRepresentation,LogicalFormRepresentation,2.GovernmentandBinding,1)ThistheoryconsistsofX-barTheory,-Theory,BoundingTheory,GovernmentTheory,CaseTheory,ControlTheory,andBindingTheory.2)InthetheoryofGovernmentandBinding,thereisanewconcept:theemptycategory.Chomskybelievesthatthroughtheemptycategory,wecanfurthergettoknowaboutthemechanismoflanguage.Butthistheoryisnotyetmature.,C-command(Constituentcommand),ABCDEF,c-commandsifdoesnotdominateandeverythatdominatesalsodominates.,(Bc-commandsC,E,F,butnotD;Cc-commandsB,D,butnotE,F.),VPVPPPNPDetNspeakaboutthelanguage(Vc-commandsPP,PPalsoc-commandsV.)“c-command”referstotherelationbetweenanelementandanotherofthesamenodeinatreediagram,andanyothersunderthelatterelementsaswell.,BindingTheory,A.Ananaphorisboundinitsgoverningcategory.B.Apronominalisfreeinitsgoverningcategory.C.Anr-expressionisfreeBindingreferstotherelationbetweendifferentreferringwordandthesubjectofasentencecontainingit.,Anaphorisusedinanarrowsensetoincludereflexiveslikemyselfandreciprocalslikeeachother.Pronominalrefertopronounsotherthatreflexivesandreciprocals.R-expressions(referential-expressions)coveralltheotherr-expressionsexceptanaphorsandpronominals,e.g.John,Bill,theman.,(a)John1likeshimself1.(b)They1hiteachother1.(c)John1doesntlikeBills2criticismofhimself2.(d)John1likeshim2.(e)Bill1saysJohn2likeshim2.(f)Bill1saysJohn2likestheman3.,3.TheMinimalistProgram,1)TheMinimalistProgramismotivatedbytworelatedquestions:Whatarethegeneralconditionsthatthehumanlanguagefacultyshouldbeexpectedtosatisfy?Towhatextentisthelanguagefacultydeterminedbytheseconditions,withoutaspecialstructurethatliesbeyondthem?,2)Remarkablechangesinthenewtheory:SomeofthediscreteanalyticalmodelsintheGBarediscardedandthetwolevelsofanalysis.thedeepstructureandthesurfacestructure,areleftout.Theimportantconceptof“government”intheprevioustheoryisrejectedandthefactsinterpretedbythetheoryofgovernmentarereplacedbyseveralrevisedconcepts,thusthetheoryofgovernmenthasturnedfromasubsystemofuniversalgrammarintotheinterpretativeconstraintoftheoutputcondition.,3)Inthe1990s,ChomskyfurtherdevelopedtheminimalisttheoryinhisMinimalistInquiries:TheFramework(1998).Inhiswork,ChomskyholdsthattheinitialstatesofhumanlanguagesarethesamewhereasthestatesofacquiringdifferentlanguagesarenotWhilethefacultyoflanguageconsistsofacognitivesystemthatstoresinformationsuchassound,meaning,andstructures,theperformancesystemretrievesandusestheinformation.Heraisesaprofoundquestion:Howwellisthelanguagefacultydesigned?ChomskyimaginesacaseinwhichacertainprimatewhichhasbeenrecognizeditsmindbygivingitthefacultyoflanguageThus,heputforwardthestrongestminimalistthesis:Linguisticmechanismistheidealsolutiontotheproblemoflegibilityconditions.,ReconsiderationonTGGrammar,1.Itsconsistentaimsandpurposes,ThedevelopmentofTGGrammarcanberegardedasaprocessofconstantlyminimalizingtheoriesandcontrollingthegenerativepowers.Althoughithasinvolvedputtingforward,revising,andcancelingofmanyspecificrules,hypotheses,mechanisms,andtheoreticalmodels,itsaimsandpurposeshavebeenconsistent,i.e.toexplorethenature,originandtheusesofhumanknowledgeonlanguage.,2.ItsMainFeatures,1）Chomskydefineslanguageasasetofrulesorprinciples.2)Hebelievesthattheaimoflinguisticistoproduceagenerativegrammarwhichcapturesthetacitknowledgeofthenativespeakerofhislanguage.3)Chomskyandhisfollowersareinterestedinanydatathatrevealthenativespeakerstacitknowledge.Theyseldomusewhatnativespeakersactuallysay;theyrelyontheirownintuition.4)Themethodologyishypothesis-deductive.5)Itfollowsrationalisminphilosophyandmentalisminpsychology.,3.Theinfluence,Inthelate1950s,theappearanceofTGGrammarinAmericaviolentlypunchedtheprevailingstructuralistdescriptivelinguistics.Chomskystheoriesarepopular,particularlyintheUnitedState,buttheyhaveneverbeenfreefromcontroversy.Criticismhascomefromanumberofdifferentdirections.Somecritics(seelanguagelearning)havequestionedwhetheritisnecessarytopositUniversalGrammarinordertoexplainchildlanguageacquisition,arguingthatdomain-generallearningmechanismsaresufficient.Todaytherearemanydifferentbranchesofgenerativegrammar;onecanviewgrammaticalframeworkssuchashead-drivenphrasestructuregrammar,lexicalfunctionalgrammarandcombinatorycategorialgrammarasbroadlyChomskyandgenerativeinorientation,butwithsignificantdifferencesinexecution.,Thankyou!,chomsky.info:TheNoamChomskyWebsite,
展开阅读全文