Treść książki
Przejdź do opcji czytnikaPrzejdź do nawigacjiPrzejdź do informacjiPrzejdź do stopki
18
1.Introduction
thatitmeanstheabsenceofthatboundary.Totakethistackistocommit
thecontinuumfallacy,whichinvolvesarguingthatiftwoextremesare
connectedbysmallintermediatediferencesandifatnostepcanone
indicateadecisivediference,thentheextremesarethesame.Tousean
analogy,inabilitytospecifyatwhattemperaturecoldturnstohotshould
notleadtotheconclusionthatcoldisreallythesameashot.Butthis
ismoreorlesswhathappenswhenthefuzzinessofthedistinctionis
takenasajustificationofviewingalllanguageformsasconstructions
andgrantingthemequalsemanticpotential.
Onecouldevensuspectthattherevisionistatmospheresurrounding
discussionsofthelexiconofsyntaxmaybeanartifactoftheemphasis
onfuzzinessasacrucialfeatureofthemostfundamentalconceptsin
linguisticssuchasdegreeofgrammaticalness(Chomsky,1961)orproto‑
type(Ross,1972;Rosch,1975a;1975b).Applicationsofprototypeandits
inherentfuzzinessarenotlimitedtoanalysesofconceptualcategories
withinsemantics(Geeraerts,1989),butarealsoinvokedinstudiesin
phonology(Jaeger,1980)orsyntax(Ross,1973;Kalisz,1981).Indeed,the
useofprototypesasalinguistictoolissowidespreadthatWierzbicka
(1996)andPosner(1986)warnagainstitsoveruse.Posneradmitsthat
linguistswereHenamoredoftheprototypeidea”(1986,p.55)becauseof
itspromiseofexplanatorypower.Wierzbickaremarksthatprototypehas
beenHtreatedasanexcuseforintellectuallazinessandsloppiness,”and
goesontoatemptthepredictionthatHif[fuzzyprototype]istreatedas
amagicalkeytoopenalldoorswithoutefort,thechancesarethatitwill
causemoreharmthangood”(1996,p.167).Onecouldexplainawayany
exception,anomalyorcontradictionasacaseoffuzziness;constructions
couldalsobetreatedasfamilyresemblancestructureswithnonecessary
conditionsfornaturalusage,andthiswaynoviolatedconstraintwillbe
aproblem,butthiswouldefectivelymakeitunnecessarytoatemptto
describeconstruction.Theprototypeideaandthefamilyresemblance
structurebothpresupposethatconstructionsinvolveahighdegreeof
inherentimprecision,whichisaplausiblehypothesis,butahypothesis
nonetheless.Ibelievethatprototypeandfamilyresemblanceshouldbe
invokedonlyasalastresort,whennootherdescriptionsarecapableof
capturingthenatureofaconstructionwithsomeprecision.
Here,inthecontextofthelexiconandsyntaxtoo,thecaseoffuzzi‑
nessisprobablybeingoverplayed.Putmoresimply,Ibelievethatafuzzy
distinctionisbeterthannoneatall.Theinsightfromthedistinctionthat
closed-classformshavefunctionsratherthanmeaningsisstillvalid;
thatis,althoughclosed-classformscaneasilybedemonstratedtocarry
meanings,thesecanandshouldbeexpectedtobeconstrained.Atthe
moment,manyconstructioniststudiesofsyntacticpaternssuchasthe
