与论元实现中的力量传输和定向变化.docx

与论元实现中的力量传输和定向变化.docx

《与论元实现中的力量传输和定向变化.docx》由会员分享，可在线阅读，更多相关《与论元实现中的力量传输和定向变化.docx（28页珍藏版）》请在冰豆网上搜索。

与论元实现中的力量传输和定向变化

牛津认知语言学手册（GeeaertsandCuyckens2010再版）

Repost:

事件词汇化与论元实现中的力量传输和定向变化（第二部分，Croft:

forthcoming）

Thetransmissionofforcetakesplacefortheprofiledtemporalphaseoftheevent,butforconvenienceitisonlyrepresentedbythecausalarrowsatthebeginningandtheendoftheprofiledphase.

TheeventinFigure12containstwosubevents:

theoneinvolvingJaneisanundirectedactivityandtheoneinvolving“WarandPeace”isanaccomplishment.Theoveralleventprofiledbytheverbisanaccomplishment.Ingeneral,iftheverborpredicateprofileincludesadirectedchangesubevent,thentheaspectualtypeofthepredicateasawholeisadirectedchangeoftherelevanttype.Therealsoappearstobeanotherconstraintonpredicateprofiles,atleastforsimplepredicates:

onlyonedirectedchangesubeventoccursinapredicateprofile.ThisistheequivalentprincipleinthemodelpresentedheretoTenny’sconstraintonverbs（i.e.,eventlexicalizations）:

‘therecanbenomorethanonemeasuringoutforanyeventdescribedbyaverb’（Tenny1994:

11）.

ThemodelpresentedherehasseveralimportantadvantagesovertherepresentationinCroft（1991）aswellastherepresentationsproposedbyotherauthors.First,itclearlydistinguishestheaspectualandthecausalstructureofevents.Besidesthesemanticcleannessofthisfeature,italsoallowsustomoreclearlyrecognizethedistinctcontributionsthataspectualstructureandcausalstructuremaketothegrammarofpredicatesandarguments.Second,itallowsustoemploythefine-grainedaspectualanalysispresentedin§3alongwiththecausalstructureanalysispresentedin§2.Third,itrepresentsbothtypesofcausalstructure,namelythetransmissionofforcerelation—participantsactingonotherparticipants—andthestandardunderstandingofcausationintermsofeventscausingotherevents.Thisisbecauseoftheprinciplethateachparticipanthasitsownsubeventinthecausalstructureoftheoverallevent.Asnotedabove,thecausalsubeventscanbeinformallythoughtofaswhateachparticipantdoesorundergoesintheoverallevent.Finally,thismodeldemonstratesthateventscanbedecomposedinthreedistinctways:

temporally,intermsoftemporalphases;qualitatively,intermsofthestatesdefinedontheqdimensionforeachparticipant’ssubevent;andcausally,intermsofthesegmentsofthecausalchain.

Thethree-dimensionalgeometricalrepresentationconstitutesthesemanticstructureofthelinguisticrepresentation.Outsideofthatrepresentationarethemorphosyntacticstructuresthatarelinkedtothesemanticstructure.Iuseaconstructionalmodelofsyntaxhere（Fillmoreetal.1988;Goldberg1995;Croft2001;CroftandCruse2004）.Theleftmostcolumn,withsyntacticrolesinallcapitals,indicatesthesyntaxoftheargumentstructureconstruction,whichisassociatedwiththesemanticsofthecausalchain.Thenextcolumntotheleftofthesemanticstructure,initalictypeface,indicatesthesyntaxoftheargumentphrasesthatinstantiatetheargumentrolesoftheargumentstructureconstruction.Theyareassociatedwiththeindividualcausalsubevents,whichdescribewhateachparticipantdoesorundergoesintheevent.Thecolumntotherightofthesemanticstructure,alsoinitalictypeface,indicatesthesyntaxofthepredicateandsatellitephrasesthatinstantiatethepredicateandsatellite（ifany）rolesintheargumentstructureconstruction.Theyareassociatedwiththeparticipantsubeventsthattheyeachprofile,includingthecausal（ornoncausal）relationsbetweenthesubeventsthattheyalsoprofile（indicatedbytheverticallinksinthesemanticstructure）.

Acrucialbutincompletelyanalyzedelementinthesemanticstructurearethesubeventsthemselves.ThediagramsinFigures11and12givesuggestivelabelsforeachparticipant’ssubevent.Unlikethelinguisticformsinitalics,thesearepartofthesemanticstructure,andaregiveninromantypeface.Infact,theyareonlysuggestivelabels.Aproperdescriptionwouldbebasedondefiningallthewell-definedstatesonthequalitativedimensionforeachparticipant’ssubevent.Thiswouldbeofcourseananalysisofwhathasbeencalledthesemanticrootofapredicate（AR,71-72）.

Inthenextsectionofthischapter,Iwillusethisrepresentationtoprovideananalysisofseveralphenomenathathavebeenassociatedwithaproposedcontrastbetweenmannerandresultpredicates.Iarguethatinfactthecrucialsemanticdistinctionisthatbetweendirectedchangeandundirectedchange.

5.Directedchangeineventlexicalization 

LevinandRappaportHovav（1991）addressthequestionofwhatargumentstructurealternationsareallowedforindividualpredicatesinEnglish.Theyobservethatsomepredicatesallowformanyargumentstructurealternations,usingtheexampleofwipe:

 （15）Transitive:

Kaywipedthecounter.[surfacecontactmeaning]

（16）Resultative:

Kaywipedthecounterclean.

[resultfromsurfacecontact]

（17）Removal:

Kaywipedthefingerprintsfromthecounter.

[removebymeansofsurfacecontact]

（18）Application:

Kaywipedthepolishontothetable.

[applybymeansofsurfacecontact]

Incontrast,otherpredicatessuchasbreakandopenappeartoallowveryfew

argumentstructurealternations（RappaportHovavandLevin1998:

100-3）.

LevinandRappaportHovavdescribeverbsofthewipetypeasmannerverbs,and

verbsofthebreaktypeasresultverbs.Theytreatthisasamajordistinctioninevent

lexicalizationpatterns,andtracethedistinctionbacktoFillmore（1967）.Talmy

（1991/2000）makesessentiallythesamedistinctionbetweenverb-framing（result）and

satellite-framing（manner）realizationsofevents.

RappaportHovavandLevin（1998）proposeanexplanationintermsofthestructural

complexityofevents（seealsoAR,115-17）.Mannerverbsbasicallydescribesimple

events,consistingofonesubeventintheirrepresentation.Furthersubeventscanbe

combinedwithasimplemannersubevent,leadingtotheargumentstructuresfoundin

（16）-（18）.Resultverbs,ontheotherhand,formcomplexsubevents,includingamong

otherthingstheresultstate.Forthemostpart,furthersubeventscannotbecombinedwith

acomplexresultevent.Asaconsequence,resultverbsdonotallowforasmany

argumentstructurealternations.

LevinandRappaportHovav’sexplanationisbasedonahighlyabstractpropertyof

eventstructure,namelyitscomplexityintheireventdecompositionmodel,ratherthanon

specificsemanticpropertiesoftheeventoranyofitssubevents.Itisthereforehighly

sensitivetothewayinwhichaneventisdecomposedinasemanticanalysis.For

example,inthemodelofeventdecompositionpresentedhere,virtuallyeveryeventis

complexinatleastoneofthethreedimensions（time,qualitativestates,andcausal

chain）.Theonlytypeofeventthatissimple（=consistsofonlyonesubeventonall

dimensions）isaone-participantinherentpermanentstate.Thuswemustfindamore

specificsemanticpropertyofeventsthatwilldistinguishmannerandresultverbs.That

propertyappearstobedirectedchange.

Mannerverbsdonotprofileadirectedchangeintheirmost“basic”orlowestvalency

construal.Forexample,thelowestvalencyconstrualforwipeisthetransitive

constructioninexample（15）;thediagramrepresentingitssemanticstructureisinFigure

13:

Figure13.Representationofexample（15）.

Highervalencyconstructionsaddadirectedchangetotheverbalprofile.Sincethere

isnodirectedchangeinthelowest-valencyprofile,onecanconstruedifferenttypesof

directedchangesforthemannerevent.Aresultativeconstructionforwipeasin（16）

construesthelocusofsurfacecontactaspossessingascalarpropertythatisgradually

broughtaboutbythemanner:

Figure14.Representationofexample（16）.

Theremovalconstructionin（17）construestheentitythatisremovedasafigureina

directedchangeoflocation,andtheapplicationconstructionin（18）construesanapplied

entityasafigureinanotherkindofdirectedchangeoflocation:

Figure15.Representationofexample（17）.

Figure16.Representationofexample（18）.

Incontrast,aresultverbprofilesadirectedchangeinitslowestvalencyform.Hence

itcannotoccurinaconstructionthataddsanotherdirectedchange,becauseofthe

constraintallowingonlyonedirectedchangesubeventinaverbalprofile.

ThesuggestivelabelfortheundirectedactivitysubeventinFigures13-16isa

combinationthatincorporatesmanner（wipe）andasecondpredicatethatdescribeswhat

isbeingdoneinthatmanner（contact,remove,apply）.Themannersemanticcomponent

isincorporatedintoasubeventwhichotherwiseappearstodescribearesultofthemanner

activity:

thewipingappearstocausecontact,removalorapplicationintheexamples.Ina

causalmodel,onemightexpecttwoseparatesubevents,whichinthiscasewouldshare

thesameparticipant.

However,thisisnotgenerallythecase.Forexample,whilethemannerofmotion

causesthedirectedmotionin（19）,themannerofsoundemissionin（20）doesnotcause

thedirectedmotion,nordoesthemannerofsoundemissionin（21）causetheproduction

ofthelinguisticutterance:

（19）Sheswamtotheotherside.

（20）Thecarscreechedaroundthecorner.

（21）Hegrowledhisanswer.

Thus,itappearstha