Treść książki

Przejdź do opcji czytnikaPrzejdź do nawigacjiPrzejdź do informacjiPrzejdź do stopki
18
1.Tearchitectureofthebilingualmentallexicon
Onefrequentcriticismoftheunitaryposition(cf.Pavlenko1999)isthatit
failstodiferentiatebetweenlinguisticmeaningsandnon-linguisticconcepts.In
response,deGroot(2000)arguesthatthetaskofseparatingconceptualandsemantic
representations,aswellasdefiningthediferencesbetweenthemmaybetedious
andinfeasible.Tisisbecausebotharederivedfromexperiencewithwordsand
thesurroundingenvironment,andassuchbothreflectinternalthoughtprocesses
andinteractionwiththeworldatlarge(Lakof1987).Moreover,attemptstodefine
wordmeaningsassetsofrelevantsemanticfeaturesare,inherview,doomedto
failuresinceitispossibletoprovideconvincingandall-inclusivedefinitionsfor
veryfewwords(deGroot2000).Recentevaluationsofcomponentialanalysisalso
stressthispoint(Kövecses2006).
Toresolvethisdilemma,Francis(2000,2005)proposesacompromisesolution,
statingthatsemanticrepresentationsareasubsetofallconcepts.Shefurther
explainsthateventhoughtherearemoreconceptsthanwords,everyconceptcan
beexpressedinhumanlanguageeitheraswordsorassentences(Francis2005).
Semanticrepresentationsorwordmeaningsarethoseconceptsthatthesewords
referto.Analternativeexplanationisthatwordmeaningsareaparticulartypeof
conceptor“fragmentsofconceptualstructure”(Jackendof1994:131),orindeed
“mappingsofverballabelstotheirconcepts”(Francis2000:14).Accordingly,those
wordmeaningsthatarelinkedtospecificconceptsarereferredtoaslexicalor
semanticconcepts,whileotherconceptsarenon-linguistic(Roelofs2000).
Aclarificationisalsoinorderastotheresearchtechniquesusedtoexplorethis
issue.Tesecomprisedesignsmakinguseofreactiontimestotargetsthatrequire
eitherlexicalorconceptualprocessing.Teformertypeofaccessisinvestigated
throughdata-driventaskssuchasword-decisionorfragmentcompletion.Tests
liketheseconcentrateonthesurfacefeaturesofvocabulary,i.e.phonologyand
morphology,andactivatethosememorystructuresthatareresponsiblefortheir
processing.Conceptualprocessingisexaminedthroughtasksthattapthesemanticl
conceptualinformationassociatedwiththetarget(deGroot2002).Typical
conceptually-driventasksencompassfreerecallofwordspresentedearlier,semantic
categorizationaccordingtoasetofcriteria,e.g.concretelabstract,picturenaming,
semanticpriming,aswellasforwardandbackwardtranslation.Eventhoughthese
tasksaresaidtobeconceptually-driven,theydoneverthelessrequiresubjectsto
makealinguisticresponseinevitablyinvolvingbothsemanticandconceptual
representations.Tisunderstandablyaddsfueltocriticismsthatthemethodology
doesnotallowadistinctionbetweenwordmeaningsandrelatedconcepts.Itisthis
lackofsystemicdiferentiationthat,accordingtoJarvisandPavlenko(2008)and
Pavlenko(2009),makesreactiontimedesignssuitableonlyformonolinguallexis-
orientedresearchwhereonecanexpectdirectconcept-meaningmappings.Yet
anotherlimitationofthemethodologyisthatitcanonlytargetlexizalisedconcepts.
Overall,theviewthatthereisaunitybetweensemanticandconceptual
representationshasbeenvoicedmainlybyclassicalcognitivelinguistsand