Rohrer, 1 The cognitive science of metaphor from p

Rohrer,1ThecognitivescienceofmetaphorfromphilosophytoneuropsychologyTimRohrer,May1995AbstractInthispaperIreviewsomeofthetheoreticalissuessurroundingmetaphor,andtracethemthroughthecontextofthecognitiveneurosciencedebate.Metaphor,likeallfigurativelanguage,hasbeenusuallyexplainedasasecondarylinguisticprocesswhichtakesplaceasafunctiontakingplaceonliterallanguage.Howeverthisexplanationdoesnotfitwellwithsomeoftherecentworkonrighthemisphereprocessingoflanguageorrecentcognitivestudies,bothofwhichsuggestthatthefigurativeandliterallanguageareprocessedsimultaneouslyandsharemuchstructure.InseekingwaystooperationalizetheLakoffandJohnsonviewofmetaphorasaconstitutivecognitivephenomenon,IbegintospelloutwhatkindsoftheoreticalpredictionstheLakoff-Johnsonmodelwouldmakeontheneurophysiologicallevelscognitiveinvestigation.Iconcludebyofferingsomerudimentarythoughtsonpossibleproposalsforfurtherinvestigationusingthesemethods.Rohrer,2IntroductionandTheoreticalOverviewWhatevertheskillemployedinthought—thatoflogic,mathematics,language,spatialormusicalsymbols—wemustnotforgetthatitisdrivenbytheJamesianprocesses,undergoesflightsandperchings,issusceptibletogreatvariationsinattention,andingeneral,isfueledbymetaphoricalandmetonymicprocesses.(Edelman,p.174)Thoughithaslongbeenrecognizedthatmostofoureverydayusesoflanguageinvolvemetaphor,withafewnotableexceptions(suchasWinner&Garner,1977)theenterpriseofcognitivesciencehaslargelyignoredtheinvestigationoffigurativelanguageinfavorofinvestigatingliterallanguageuntilrecently.Ofthemanyfactorswhichcontributedtothepaucityofresearchonfigurativelanguagecomprehension,theinstantiationhypothesisisperhapsthemostonerous.Theinstantiationhypothesisarguedthatreason,intelligenceandmindsweresubstrateneutral,thatis,independentofanyspecificembodiment,solongasthebodywasaalgorithmicdevice.Thecomputationaldeviceofchoicewasaserialprocessordrivendigitalcomputer.SincethemathematicianAlanTuringprovedthatalldigitalcomputerswereinprinciplereducibletorecursiveelaborationofafinitestatealgorithm(a'Turingmachine'),minds(andmental'processes')wereinprinciplereducibletofinitestatealgorithms.Finitestatealgorithmshaveapeculiarliteralqualityinthattheirvariablesareeithertrueorfalsewithnoadmixtureoftruthorfalsitypermitted.Uponthisviewthefundamentalproblemoflanguagecomprehensionwasdetermininghowthebrain'srepresentationsandtheworldliterallymatchedup:theworldoutsidethebrainwasthoughttoberepresentedsymbolicallyinsidethebrainbyaseriesofRohrer,3finitestates.Ifthemindwasthekindofsoftwareprogramrunningonthethatkindofhardware,thelackofattentiongiventofigurativelanguagecomprehensionresultsfromanobvioussource:itwouldbeamereafterthoughttosolvingtheproblemofliterallanguagecomprehension,sinceTuringhadproventhatallcomputationalprocessesmustultimatelydecomposeintofinitestates—thatis,intoliterality.Themantraofthisdogmawasclear:solvehowlanguagerepresentsandtheproblemsposedbyfigurativelanguagewillinevitablysolvethemselves.Unfortunatelyforproponentsoftheinstantiationhypothesis,manyoftherecentfindingsincognitivesciencearemotivatedbyitsantithesis:theembodimenthypothesis.Theembodimenthypothesisarguesthatmindsarefundamentallynotdisembodiedalgorithmicprocesseslikeacomputerprogram,butareinsteadconstitutedandconstrainedbythekindsoforganizationreflectedinthebiological,anatomical,biochemical,andneurophysiologicalcharacteristicsofthebodyandthebrain.Whilebothhypothesessharethematerialistassumptionthathigher-levelprocesses,suchasabstractthought,languagecomprehensionandthelike,arebuiltupoutoflower-levelprocesses,theembodimenthypothesisexplicitlydeniesthesubstrateneutralityclaimoftheinstantiationhypothesisproponents.Ratherthanprocesseswhichmanipulatesymbolsinafinite-statefashion,thestudyofmindasabiologicalandneurophysiologicalenterpriserequiresthinkingaboutthemindinaseriesofinterwovenlevelsofinvestigation,manyofwhichappeartobehaveanalogicallyandfrequentlyexhibitakindofadaptiveplasticitynotfoundindigitalcomputers.Rohrer,4Agoodcognitivetheoryrequiresakindofverticalconvergenceacrossanumberofdifferentlevelsofinvestigationrangingfromphilosophyandcognitivepsychologythroughneuropsychologytoneuroanatomyandneurochemistry.Figurativelanguagecomprehensionhasrobustconnectionswiththeembodimenthypothesis,especiallyintheareaofmetaphor.AsGeorgeLakoff(1987)andMarkJohnson(1987)haveargued,ourordinaryuseoflanguageislargelystructuredbymetaphoricandmetonymicprincipleswhichexhibitadirectionality.Humanbeingssystematicallycharacterizeabstractideas—thoughts,religiousbeliefs,politicalandethicalsituations—intermsofbodilymovementsandbodilyfunctions,forexample.Theprimaryclaimoftheirpositionisthatthesemetaphorsandthedirectionalityarenotarbitrary,butinsteadareanaturaloutgrowthofthemannerinwhichourmindsareconstituted.InthispaperIseektotietogetherthreadsfromthephilosophical,cognitive,andneurophysiologicallevels.InoperationalizingtheLakoff-JohnsonhypothesisIhavemadeuseofPosner'sschematizationofcognitivescience(tableI)asaframeworkforseekingaverticallyintegratedcognitivetheory.ConceptualMetaphorTheoryTasksL