Bilingual education enhances third language acquis

AppliedPsycholinguistics21(2000),23–44PrintedintheUnitedStatesofAmericaBilingualeducationenhancesthirdlanguageacquisition:EvidencefromCataloniaCRISTINASANZGeorgetownUniversityADDRESSFORCORRESPONDENCECristinaSanz,DepartmentofSpanishandPortuguese,GeorgetownUniversity,Washington,DC20057.Email:[e-mail]cquisitionofathirdlanguage(L3)inabilingualcontexthaveshownthatliteracyintwolanguagesfacilitatestheacquisitionofathird(Cenoz&Valencia,1994;Swain,Lapkin,Rowen,&Hart,1990).ThepresentstudyseekstocontributetothislineofresearchbycomparingtheacquisitionofEnglishasanL3byCatalan/Spanishbilingualhighschoolstudentsinanimmer-sionprogramwiththeacquisitionofEnglishbySpanishmonolinguals.Datafrom201participantsweresubmittedtoahierarchicalmultipleregressionanalysis,renderingresultsthatshowthatbilin-gualismindeedhasapositiveeffectontheacquisitionofanL3.Theevidenceisdiscussedfromacognitiveperspective.Sincethe1960s,researchontheimpactofbilingualismoncognitionhasassoci-atedbilingualismwithpositiveeffectsonanumberofinternalvariables,includ-ingintelligence(Peal&Lambert,1962),metalinguisticawareness(Ben-Zeev,1977;Bialystok,1991),cognitiveflexibilityandprocessingmechanisms(McLaughlin&Nayak,1989;Nation&McLaughlin,1986;Nayak,Hansen,Krueger,&McLaughlin,1990),andevenamoredemocraticdisposition(Pan-dey,1991).Bilingualism–orbiliteracy,tobeprecise–resultsinmoreefficientlanguagelearning,asshownbystudiescomparingbilinguals’andmonolinguals’acquisitionofaforeignlanguage,intermsofbothgenerallanguageproficiency(Cenoz&Valencia,1994;Swainetal.,1990)andtheacquisitionofspecificparameters(Klein,1995).1Thisresearchhasalsoshownthattheeffectsofbilingualismoninternal,cognitivevariablesaremediatedbykeyexternalfactorsrelatedtoparticularsociolinguisticsituations.Thisdoublefaceofbilingualism,clearlyexpressedinLambert’s(1981)distinctionbetweenadditiveandsubtractivebilingualism,warrantsfurtherinvestigationofadditionalevidencefromdifferentsociolinguis-ticcontexts.ThepurposeofthisstudyistocontributetothislineofresearchwithevidencefromCatalonia.Twostudiesontherelationshipbetweenbilingualismandtheacquisitionof©2000CambridgeUniversityPress0142-7164/00$9.50AppliedPsycholinguistics21:124Sanz:BilingualismandL3acquisitionathirdlanguage(L3)haveproducedcontradictoryresults.Ontheonehand,Wagner,Spratt,andEzzaki(1989)concludedthatL1(Berber)illiteracydoesnothinderL2(Arabic)orL3(French)literacy,basedontheirownevidencefromMorocco.Therefore,theauthorssuggestedthatUNESCO’srecommenda-tiontoteachchildrentoreadintheirL1wouldbeanunnecessaryburdenoncountrieslikeMorocco,whichlacksastrongeconomicbaseandwhereasub-stantialportionofthepopulationspeaksalanguagewithoutawritingsystem.Ontheotherhand,CenozandValencia(1994)foundthatintheBasqueCountrystudentsinBasqueimmersionschoolsoutperformedstudentsinmajoritylan-guage(Spanish)schoolsintheirEnglishproficiencytest.Cenozetal.’sresultsareinconformitywiththoseofSwainetal.(1990),whoinvestigatedtheimpactofL1literacyonL3(French)proficiencyintheTorontometropolitanarea.ResultsfromtheBasqueandtheCanadianstudiesdisprovethegeneralbeliefthatliteracyintheL1interfereswithorretardsachievingcontrolovertheL2andL3(Swainetal.,1990).Obviously,thepositiveevidenceprovidedbytwooutofthreestudiesdoesnotallowforuniversalgeneralizationsontheeffectsofbilingualeducationonL3acquisition.2Therearemanydifferencesamongthesestudies,whichmayaccountfortheircontradictoryresults.Forone,thereistheissueofmethodology.Wagneretal.’s(1989)wasa6-yearlongitudinalstudyofprimaryschoolliteracyinthreelanguages.BothSwainetal.’s(1990)andCenozandValencia’s(1994)designswerecross-sectional,focusedongenerallinguisticability(oralandwritten),andincludedolderstudentsasparticipants.Aproblemforbilingualismandappliedlinguisticsscholarsistheirinabilitytoformulategeneralconclusivestatementsbecauseofthesmallnumberofstudies;thevarietyofmethods,samples,andvariablesused;andthelackofreplicationstudies(Sanz,1997a).Akeydistinctionmadeinthesethreestudiesisthatofbilingualismversusbiliteracy.Wagneretal.(1989)comparedbilingualsliterateintheL2withmon-olinguals,whileCenozandValencia(1994)comparedbilingualsliterateinboththeL1andL2withmonolinguals.Noneofthestudiesseparatedoralproficiencyfromwrittenproficiencyintwolanguagesorbilingualismfrombiliteracy.Incontrast,Swainetal.’sstudyisolatedbothvariablesbycomparingbiliteratebilingualswithbilingualsliterateonlyintheL2.Basedontheirevidence,Swainetal.concludedthatthekeytothecognitiveadvantagereflectedinmoreeffi-cientL3acquisitionisbiliteracyandnotbilingualism.BilingualsilliterateintheirL1didnotperformsignificantlybetterthanthemonolingualgroup.Inthepresentstudy,asinCenozandValencia(1994),allparticipantsinthebilingualsamplewerebiliterate,whileallmonolingualparticipants,unlikethoseinthatstudy,weretruemonolingualswholivedinamonolingualarea.CenozandValencia’smonolingualparticipantslivedinabilingualareaandattendedaschoolwhereBasquewastaughtfor4hoursaweek.Anyadvantagefoundinfavorofthebilingualgroupcouldbeassumedtobetheresultoftheir“fullydevelopedbilinguali