Kobe Shoin Women’s University Repository
Title
What Happens to Their English ? A Study of the
English Proficiency of Japanese Elementary School
Returnees
Author(s)
Alan E. Jackson/Kikuchi Mari
Citation
Shoin Literary Review，No.28：107-133
Issue Date
1995
Resource Type
Bulletin Paper / 紀要論文
Resource Version
URL
Right
WhatHappenstoTheirEnglish?
AStudyoftheEnglishProficiencyof
JapaneseElementarySchoolReturnees
AlanE.JacksonandKikuchiMari
1.Introduction Thisrepresentsapreliminarystudyinafieldlittleexplored‐what happenstotheEnglishofreturneesfromEnglish-speakingcountries whentheysettlebackintoJapanesesociety.Thetopic,whichis increasinglyimportantasthenumberofreturneescontinuestoincrease (alreadyover11,000in1989},comprisesthreeaspects‐howdosuch returneestrytomaintaintheirEnglish,towhatextentdotheysucceed inmaintainingtheirsecondlanguage,andwhatfactorsareimportantin determiningdegreeofmaintenanceorattrition. Thispaperhasthreemainparts;abriefsurveyofwhatisknown aboutreturneemaintenanceactivitieswithanoverviewofattrition/ maintenancetheory;areportofastudyundertakenbytheauthorsinto themaintenanceactivitiesandEnglishproficiencyofagroupof returnees;andaconclusionsummarizingandcommentingonthe resultsofthestudy. 2,Background 2.1ReturneeEnglishLanguageMaantenance TheliteratureabouthowEnglish-speakingreturneessetabout maintainingtheirEnglish,andthesuccesstheyachieve,isnotvery -107-一extensive,Whatisknowncanbesummarizedunderfourheadings: Attitudes,Schools,MaintenanceClasses,andResults. A.Attitudes Evenlimitedacquaintancewithreturneesandtheirparentsquickly revealsthatreadaptationtoJapanesesocietyisthegreatestconcernof thereturnees.OnoHiroshi,aspecialistinthecorrelationbetween intellectualdevelopmentandlanguageacquisition,lendssupporttothis parentalview.Heseesaparticularproblemwithstudentswhohave beenabroadforalongtimetryingtomaintaintheirEnglishskills.If attentionisnotpaidtoJapaneseskills,thereisachancethatboth languageswillbeinadequate.Inhisview,studentsonreturnshould focusonJapanesetoensurefluencyintheirnativelanguage.Iftheydo this,thebasicsoftheirforeignlanguagewillstillberetainedandcan bepickedupandputtouselatertomakeatrulybilingualchild. (ReportinDailyYomiuriontalktorepresentativesofcompanieswith overseasemployees‐Nozawa1993). Yashiro(1990)agreesthatreadaptationisinfactgivengreatest emphasisbutpointsoutthatthisdoesnotmeanthatyoungreturneesdo notcaretoretaintheirforeignlanguageabilitynorthatattemptsat maintenancearerare.Hisviewisreflectedinstatisticscollectedin twomajorsurveys: ResearchcarriedoutbyJapanOverseasEducationalServices {DOES)foundthat92%ofthe1447returneesrespondingtoaquestion-naireexpressedtheirdesiretoretaintheirforeignlanguage,and85% oftherespondentswereemployingsomemeanstodoso.Ofthese,68% attendedspecialmaintenanceclassesoutsideschool. AmorerigorousstudyconductedbytheManagementandCoordina-tionAgencyfoundthat56%ofarandomsampleofreturneespursued maintenanceactivities.Ofthese,44%tunedintolanguageprogramson
TVandradio,41%attendedmaintenanceclasses,35%continued correspondencewithfriendsabroad,and13%conversedinaforeign languageathomeorwithfriends. Twofactors‐ageandsex‐seemtohaveabearingonenthusiasm formaintenanceactivites Age‐Children,whofrequentlywishtoshedtheiroverseasimage,often donotcontinuelonginspecialoutsideclasses,atleastnotbeyondthe ageof12,becausetheycanstudysomeEnglishinjuniorhighand anywaybecometoobusy(Goodman1990).TheManagementand CoordinationAgencystudyshowedmaximuminvolvementamongst elementaryschooluppergraders.Thereasonwhythereislesspartici-pationamongjuniorhighschoolstudentscanbeexplainedbythefact thatEnglishbecomesaschoolsubjectatthatlevel,andthereisa generallyheavieracademicloadforthestudentstostudyinprepara-tionforentranceexaminations.Onemightaddthattheneedfor readjustmenttoJapanesesociety,togetfirmlyontheeducational ladder,suddenlybecomesmorepressingandthemaintenanceofspoken Englishmaynolongerbeseenasapriority. Sex‐therateofmaintenanceclassparticipationwashigherforgirls thanboys.Thereseemstobeageneralfeelingthatgirlsaremoreable inlanguagestudiesthanboysandalsoageneralattitudethatwhat countswithbaysisnothowwelltheycanspeakaforeignlanguagebut whattheycanactuallydowithit.Boysareconsequentlylessencour-agedandThuslessmotivatedtomaintaintheirforeignlanguagethan girlsduetowhatseemstobeasocialbias.Thedifferenceinattitudes toboysandgirlsisalsoreflectedininternationalschoolattendance (seebelow). B.Schools ThemostnotablebookinEnglishonthesubjectofreturnees,R. -109一
Goodman's"Japan'sInternationalYouth"(1990)pointsout(p.150) thatregularelementaryschoolsprovidenolanguageretentionworkat a11andthatparentsmustfindprovisionoutsideschool.Somereturnees gotointernationalschoolsbuttheyarerelativelyfewinnumber(there arenotmanyschools‐only22inJapan‐andtheyareexpensive)and manysuchstudentsaregirls,whotendtospendlongeroverseasand withwhomparentsarewillingtotakemorerisks .Suchchildren become"Thirdculturekids"butitisdoubtfulwhethertheycanreally actasaninterfacebetweenJapanandtheworld.Theyclearlyhavean easiertimeadjustingininternationalschoolsbuttheireventualre-a〔加stmenttoJapaneselifeisnecessarilyand,intheviewofsome , undulydelayed.Suchchildrencanperhapsbeseenasexceptionaland notexemplarsofthetypicalreturnee. Acompromisebetweenthetotallackofprovisionforlanguage maintenanceinregularelementaryschoolsandthedelayedadjustment toJapanesesocietyproducedbyinternationalschoolattendanceisthe "ukeireko" 一一usually"fuzoku"schools(universityaffiliated)whi ch eitherhaveaspecialsegregated"kikokushijo"(returnee)classorstart thatwayandgraduallyintegratewithregularstudents.Suchschools areusuallyattachedtonationaluniversitiesthoughsomeprivatejunior andseniorhighschoolsserveasimilarfunctionandhaveeitherspecial kikokushijoclassesoratleastaspecialselectionsystemwithsome segregationforEnglish.NearlyalItheseareaboveelementarylevel . Theydifferfromordinaryschoolsinclasssize,specialEnglishstream-ingandsomeintensiveJapaneseinstruction,butotherwisefollowthe samecurriculumasregularschools.Themainaimofthespecial treatmentprovidedtoreturneesinsuchschoolsisclearlytoeasethe readjustmentprocess,andasGoodmanpointsout,reassimilationof suchstudentsintoJapanesesociety,includingachievementofadequate
Japaneselanguageability,seemstobecompletedfairlyquickly.Asa by-productofgroupingreturneestogetherorstreamingforEnglish,it islikelythatEnglishmaintenanceistosomeextentpromotedalso. Goodman,however,saysnothingabouthowwellEnglishismaintained. C.MaintenanceClasses Asmentionedabove,Yashiro(1990)quotesManagementandCoordi-nationAgencyfiguresrevealingthatmoststudentsattendprivate foreignlanguagemaintenanceclasses,thoughmanystudentsdonot takeadvantageoftheopportunitiesavailable.Studentsfeelassociat-ingwithforeignstudentsisthemosteffectivewayofmaintainingtheir Englishbutintheabsenceofopportunitiestomeetnativespeakers, specialmaintenanceclassesofferasecond-choicealternative.These classesarerarelyveryintensive,typicallyanhourortwo,onceortwice aweekinaprivatelanguageschoolorinstitutionsuchastheYMCA. With68%ofstudyrespondantsattending,maintenanceclassesappear tobethemostprevalentmeansavailableforsecondlanguagemainte-nance,thoughitisinterestingtonotethatattendancerarelycontinues forlong,with,intheManagementandCoordinationAgencystudy,only 16%ofstudentscontinuingforthreeyears. D.Results Astohoweffectiveallthisactivityisalmostnothingisknown.Even theextensivestudiesquotedbyYashirosaynothingaboutEnglish proficiencyasnomeasurementsweretaken. 2.2Attrition/MaintenanceTheory Sinceabout1980agreatdealofworkhasbeendoneonlanguage attritionandmaintenance,thoughasyetfewveryconvincingempirical studieshavebeendonetoelucidatetherelationshipbeweenrateof attrition/maintenanceandarangeofpredictorvariables.Whilst -111一
acknowledging,then,thatmuchoftheworkisnomorethanspecula-tive,wecanatleasttentativelylistfactorswhichprobablyinfluence whathappenstotheEnglishofreturneestoJapan.Factorsincludethe following A.PersonalCharacteristics Age‐Youngerchildrenwouldappeartohavemoredifficultyin maintainingaforeignlanguage(Cohen89,WeltensandCohen89) , whichmayhavesomethingtodowithlearninghistory .Itmaybethat olderchildren,forwhomconsciouslearningmayplayasignificantrole alongsidemarenaturalacquisition,mayhavebetterretention(Oxford 82). Sex‐Sincefactorsimportantinacquisitionmayalsobeimportantin retention/attrition,sexmaybeakeyfactorasgirlsgenerallydobetter thanboysinallareasoflanguageskill(Burko-Gleason80}. B.LanguageRelatedFactors ProficiencyLevel--Viewsontherelationshipbetweenmaintenanceand proficiencylevelseemtoconflict.Oxford(82)quotesastudyshowing retentiontobebetteramongthosemoreproficenttobeginwith ,aview supportedbyCohen(74),whileWeltens,ElsandSchils(89)foundthat attrition,whichsetinquicklyandthenleveledoff,wasindependentof trainingleve1,i.e.therewasafixedamountofattrition ,notafixed proportion.Athirdview(WeltensandCohen89)suggeststhepossibil-ityofacriticalthresholdabovewhichlanguageskillisdurable . Use‐Asmightbeexpected,maintenanceofproficiencyappearstobe betteramongthosewhomakeactiveuseofthelanguage(Oxford82 , Gardner80). LIcompetence‐Twowriters{Oxford82andGardner80)havesug-gestedalinkbetweenmaintenanceandcompetenceinthesubject'sfirst language.
Literacy‐Onefinallanguage-relatedpredictorvariable,whichmaybe relatedtoage,isliteracy.Cohen(89),Olshtain(89),andWeltensand Cohen(89)havealltakentheviewthatlanguageretentionisimproved ifthesubjectisliterate. C.AffectiveFactors Attitude/Motivation‐Thereseemstobegeneralagreementwith Oxford(82)thatlearnerswithhighmotivation,highneedachievement, andpositiveattitudestowardsthelanguageunderconsiderationare bothbetteracquirersandbetterretainersofasecondlanguage. Gardner(80}seesfourmajorsocialfactorsatworkhere a)Integrativeness‐strengthofintegrativeorientationattitudes towardsethnicgroup,andinterestinforeignlanguages. b)Attitudestowardsthelearningsituation. c)Motivation‐motivationalintensity,desiretolearn,andattitude towardslearningthelanguage. d)Aptitude. Whereindividualsarehighonintegrativenessand/orattitudestoward thelearningsituation,itisreasonabletopredictthatoncelanguage traininghasterminated,theywillstillbeorientedtowardmaintaining competenceinthelanguage,andthus,evenintheabsenceofopportu-pitiestousethelanguage,therateoflanguagelosswillbelesssevere thanforindividualslackingtheattitudinalsupport.Moreover,since attitudinal/motivationalcharacteristicsarerelatedtoindicesofparti-cipationinlanguagerelatedsituations,theywillrelatetoattemptsto maintainsecondlanguageskillsoncetraininghasterminated.This mightinvolveseekingoutfurthertraining,colleagueswhospeakthe language,orsimplyself・practiceetc.,butwouldresultingreaterlan-guageretention.Aninterestingexampleoftheeffectsofattitudinal/ motivationalfactorshasbeengivenbyOlshtain(86),whostudied -113一
youngEnglishspeakersacquiringHebrewinIsrael.Shefoundthatthe needtorebuildHebrewhadanadverseaffectonthesubjects'English, afindingthatmayberelevanttothesituationofJapanesereturnees. D.AttritionPath Production/Reception‐Bothaphasiastudies(Obler80)andlanguage attritionresearch(Cohen89)pointtoproductiveskillsbeingmore difficulttoretainthatreceptiveskills. Fluency/Quantity--Onestudy,conductedbyMoorcroftandGardner ($7),showednolossinfluencywhenalanguageuserbegantolosehis languagebutagradualdeclineinthequantityofproduction . LinguisticElements‐Wherelanguagelosstakesplace ,thepatternof lossseemstobecharacteristic.Amongpossiblelinguisticfeaturesof attritionarethefollowing Attritionwillresultinasmallernumberandsmallervarietyof lexicalitemsavailable,andtheretainedlexiconwillbeofcommon , highfrequency,unmarkedlexicalitems.Gapswillbeoflesscommon , lessfrequent,markeditems.Theattriterwillspeakataslowerrate withmorefrequenthesitations,pauses,repairs,falsestartsetc . {Anderson80}. Inadditiontolossoflexis,youngersubjectswillexhibitsomerever-salofacquisitionorderofgrammaticalmorphemes.(Olshtain89). 3.ThePresentStudy 3.10bjectives Intheabsenceofpreviouslanguagemaintenancestudiesamongst Japanesereturneesthestudydiscussedherewasdesignedverymuchin theformofapilotstudythattheauthorshopedwouldprovideinsights onwhichamorerigorousfuturestudycouldbebased.Theobjectives werefive-fold
a)Todescribethelanguagemaintenanceactivitiesofagroupof returnees. b}Tofindoutwhatwouldhappentothesubjects'Englishlanguage proficiencyoveraperiodofoneyear(inthecaseofsomesubjects, twoyears). c)Torelatechangestopredictorvariables . d)Todescribechangesinsubjectinterlanguageasproficiencyfell . (Note:Itwasassumedattheoutsetthatmaintenanceactivities wouldprobablynotbeadequateforthesubjectstoretaintheir initialproficiency.) e)Toassessthevalueofarangeoftestinstruments . Giventhedifficultyofobtainingenoughsubjectsandthevarietyin personalcharacteristicsandexperienceexhibitedbythosesubjects,it wasclearlyimpossibletocarryoutastudywithatruescientificdesign, i.e.oneinwhichextraneousvariablesarecontrolledandchangesin proficiencyarerelatedtoapredictorvariable.Instead,itwasdecided toadoptadescriptivecasestudyapproachintheformofapilotstudy withsubjectlanguageproficiencymonitoredbyasimplepre-test , post-testpre-scientificdesign.Thoughclearlylessthanstatistically valid,itwasfeltthatsuchastudywouldneverthelessyieldresults valuableenoughtoatleastindicatehowamorethoroughstudymight beconducted. 3.2Subjects Thestudentsinourstudy,exceptfortwoinregularelementary schools,wereallinaspecialreturneeclassinauniversity-affiliated elementaryschool.ThesestudentsweregivennospecialEnglish Iessons-indeedthecurriculumwasthesameasthatofferedtonon-returneestudentsinotherclassesexceptfortwoextraJapaneselessons -115一
i
andanadditionalmathclass-andallteachingwasinJapanese.The'
schooldefinedreturneesasJapanesenationalswhohadspentatleast
ayearabroadandhadreturnedtoJapanlessthanoneyearbefore
enteringthereturneeclass,whichwasprovidednottomaintainEnglish
l
buttoeasereentryintotheJapanesemainstream.Studentscould
continueintheclassuntilgraduationattheendofthesixgrade.We
includedinthestudyonlythosestudentswhohadspentmorethanfive
yearsabroadandwho,ontheirreturn,werereportedtohaveanative-likeEnglishproficiency.(Astheresultsinsection3show,evenfive
yearsisprobablynotenoughtoattainnative-likeproficiency,apoint
whichraisesdoubtsaboutsomepastattritionstudiesthathaveviewed
twoyearsabroadasareasonablebasisforstudy.)
Ourinterviewswithstudentsandtheirparentsrevealedthefollowing
personalcharacteristicsandinformationregardingeducation
TableI:Subjects'PersonalCharacteristics
No.SexlAgewentlYearslAgeatInterval**IGradeat
Amerlcaabroadreturnstartstudy 1.M1892.16 2.1FIIIgl1011.101fi 3.IFI41711111,11fi 4.M65110.96 5.1MI51511010.915 6.1FI3151810.913 7.1MIOI71710.913 8.1MIOIl111110.916 9.1FIII1011110.316 10.IFI51611110.215 11.IMI51611110.216 12.IFI31811111.116 *Onestudent(No .3)IivedinCanada. **IntervalbetweenreturntoJapanandstartofstudyTableII:SubjectsEducationBeforeandDuringStudy No.IAbroadIOnreturn)TlITZIT31T4 1、BilingualReturneeReturnee*ReturneeRegularRegular schoolIP**IPLPIJH***IJH 2.AmericanReturneeReturneeReturneeRegularRegular PIPIPIPIJxIJx 3.CanadianReturneeReturneeReturneeRegularRegular PIPIPIP13HIJH 4.IAmericanIReturneeIReturneeIReturneeIReturnee PIPIPIPIJH 5.IAmericanIReturneeIReturneeIReturneeIReturnee PIPEPIPIP 6.AmericanRegularRegularRegularRegular PIPIPIPIP 7.AmericanRegularRegularRegularRegular PIPIPIPIP 8.1AmericanIReturneelReturneeIReturneeIReturnee PIPIPIPIJH 9.AmericanReturneeReturneeReturneeRegularRegular PPPPJHJH 10.IAmericanIReturneeIReturneeIReturneeIInternational PIPIPIPIschool 11,AmericanReturneeReturneeReturneeReturnee PIPIPIPIJH 12、AmericanReturneeReturneeReturneeRegular PIPIPIPIJH *Returnee=classcomprisingreturnees**P=primaryschool ***JH=juniorhighschool Toprovideadegreeofperspectiveseveraloftheinstrumentsused wereadministeredtothreenative-English-speakingchildrenofapproxi-matelythesameageastheJapanesereturneesubjects.Thesechildren hadthefollowingcharacteristics NS16thgradeMaleMiddletohighverbalskills#orage NS26thgradeFemaleMiddletolowverbalskillsforage NS36thgradeFemaleHighverbalskillsforage -11?一
3.31nstruments Testswereadministeredatthestartofthestudy(T1),6months Iater(T2),12monthslater(T3),and24monthslater(T4).Thefinal administration,carriedoutnotaspartofthemainstudybutasafollow up,wascarriedoutwithonlyfoursubjects.Forcomparativepurposes , threetests‐1(1),1(2)andI(3)‐wereadministeredtothethreenative speakersmentionedin3.2.I Fivedifferentinstrumentswereusedinthisstudytotapdifferent aspectsofthestudents'(evolving)interlanguage,asfollows I(1)InterviewI i
Atwentyminuteinterviewguidedbyquestionsoneightorninetopics
wascarriedoutinaflexiblewaysoastoelicitanaturalsampleofeach
subject'slanguageproduction.Inrepeatedadministrationsofthe
interview,thetopicswerealteredtosuitthetimeofadministration.
1(2)Readingaloud
Atwo-hundredwardselectionfromafairytalesuitablefor4/5th
igraderswasreadaloud.
1(3)Storytelling
Guidedbypicturessubjectstoldawellknownchildren'sstoryoftheir
choice,eitherRedRidingHoodorUrashimaTaro.Subjectsweretold
atthebeginningthattheeventsdepictedhappenedalongtimeago,andI
weretoldtotellthestoryasinterestinglyaspossibleinthepasttense
asifentertainingayoungerchild.Thistypeoftesthasoftenbeenused
instudieswithchildren(e.g.Cohen89)asitseemstohavegreatface
ivalidityandencouragespositivesubjectparticipation.
I
I(4)Storyrepetitionl
Thetraditionalstoryofthegrasshopperandtheantwasrecordedon
tapeinl2to20syllablefragments(10ngenoughtopreventsimpleI
irepetitionandobligesubjectstorecreatefrommemory)
.Studentsfirst
i i -118listenedtothewholestoryandthenhadtorepeateachfragmentafter thetape.ThistypeoftaskwasusedsuccessfullyinthestudyofFrench acquisitioncarriedoutbyHarley(86). 1(5}Translation AtotalofI7sentencescomprisingexemplarsofawidevarietyof increasinglycomplexgrammaticalstructureswasrecordedontapein Japanesetranslationinacontext.Subjectshadtolistentoeachtarget sentenceembeddedinacontext,andthenthetargetsentenceinisola-tion,beforetranslatingitintospokenEnglish.Againthistypeoftask wasusedbyHarley{86). Theabovefiveinstrumentsrepresentacombinationofdiscrete- point/integrativeanddirect/indirectteststosamplesubjectinterlan-guageinasthoroughawayaspossible.Whileallthetesttypeshave beenusedbeforeinonestudyoranother,wecouldnotbesureatthe outsetthattheactualtestitemswechoseforeachinstrumentwere appropriateornot.Nor,sinceresearchofthistypewithJapanese returneeshasneverbeendonebefore,wouldwebeabletointerpretthe resultsofourtestsinthecontextofpreviousstudies.Thesewere seriousproblems,butitwasfeltthatwehadtoworkinthedarkat leasttosomeextentandtryouttheinstrumentsinourpilotstudyasa meansofeitherestablishingtheirusefulnessorfindingouthowto amendthemforuseinmorerigorousstudiesinthefuture.(Adiscus-sionofourfindingswithregardtothesemeasurementinstrumentsis includedinaseparatepaper-KikuchiandJackson(1995}‐forth-coming.) Inallcases,theinstrumentswereappliedinaconsistentwayand subjectoutputwasrecordedontape,transcribedandexaminedforthe followingpurposes -119一
a)Overview Interview‐Transcriptswereexaminedforoutputcharacteristics thatmightformthebasisofafuturestudy.Datawasalsoelicited regardingthesubjects'maintenanceactivities . b)Overallproficiency Storytelling‐outputwasmeasuredfortotalwords,totaldifferent wards,andrangeandvarietyofvocabulary. Storyrepetition‐outputwasgradedaccordingtoamarkingprotocol basedonthenumberofmorphemescorrectlyrepeated . Readingaloud‐outputwasmeasuredforspeedanderrorfrequency . Translation‐outputwasgradedaccordingtoerrorsmadeonascale oflto6. c)Readingability Readingaloud‐outputwasmeasuredforspeedanderrorfrequency . d)Linguisticeffects Translation‐outputofthetranslationtaskwasexaminedforthe linguisticeffectsofattrition. 3.4Results A.MaintenanceActivities TableIII:Subjects'MaintenanceActivities Tlt°T3N o.lnstructionRead/WriteT4S peakingVideo l.YMCAlyrBrotherJHEng . tutorlhr/wkC/matesclass 2・YMCAlyrSisterJHEng. C/matesclass 3・C/matesNewspaperJHEng . Books3/wkclass Books 3/wk
Comment:Thenumberofstudentsattendingmaintenanceclasses outsideschool(5/12)agreedcloselywiththe41%reportedbythe ManagementandCoordinationAgencystudyquotedabove.Relatively fewstudentsmadedeliberateattemptstomaintaintheirEnglishother thangoingtotheseclasses,possiblybecausemoststudentscouldchat inEnglishwithclassmatesduringrecessandafterclass(9/12subjects) orwithsiblings(5/12subjects).Onlytwostudents,numbers5and10, seemedtobemakinggreateffortsatretentionandoneofthese,number 10,movedontointernationalschoolafterelementaryschool,indicating clearlyherdesiretogrowupbilinguallyandbiculturally.Theimpres-sionconveyedbymoststudentswasthatintegrationintoJapanese societywastheirfirstpriority,withEnglishmaintenanceonlybeing pursuedifitdidnotentailtoomuchextraeffort. -121一
B.ReadingAloudTask TableIV:ReadingAloudTaskScores Scoring:Readingspeed-totalwordsperminute(W) Errors,hesitationsandrepeatsinIOOwords(E) No.TIT2T3T4 1.(W)1231221?1? (E)4220?? 2.120135145131 914161s 3、133138138170 $171912 4.19111Q81117 1487 5.1107111fiI108 911014 6.107111116 81513 7.171185193 10138 8.143160146 81317 9.150153163170 4'61313 10.162163145 766 11.11221481138 81415 12.117111511163 21311 i Mean125.18133。45133 .82 (Excl.1)7.916.275.361 NativeSpeakers NS1160/2 NS2145/5 NS3199/1
Comment:Thereadingaloudtaskrevealedonestudentwhowas clearlydeficientinreadingabilityforhisage,andeventhisability deterioratedovertime.Allotherstudents,exceptonewhowasyounger thantheothers,exhibitedareadingabilitynotfarbelownativespeaker normsandthisabilitydidnotattriteovertime.Interestingly,thetwo subjectswhoreadagreatdealofEnglishandwereavailablefortesting atT4showedamarkedincreaseinreadingscoreswellintotheregion occupiedbyourcontrolsampleofnativespeakers. C.StoryTellingTask TableV:StoryTellingTask-TotalWordsandErrors Scoring:Errorcount(allerrorsincludingtenseshifts)(E) Wordcount(totalminusrepeats}(W) (*=UrashimaTaro) TlIT21T31T4N o.EWw1EEWWIEEWWIEEWW/E 1.*1421754.31519038.01316856.01212160.5 2.Ilfi26816.811326420.311122120.111522715.1 3.12549319.711532721.811430521.811228523.8 4.*1423659.01315752.311143143.0 5.2132915.72139418.81632020.0 6,1232827.31036536.51841122.8 7.11170170.01415739.31421353.3 8.102071nf.1221410?.0102531nf, 9.*219497.02250125.02269134.5625242.0 10.11201201.011272272.012217108.5 11.*1324581.711250250.OIO2321nf. 12.11fi262.0101131nf.11130130.O Meanl7.5245.832.816.4246.138.4×6.0240.240.O NativeSpeakersEWW/E NS104371nf. NS2219698.O NS3239?198.5 -123一
Comment:Performancewasfairlystableoverthefirstyearwithonly
aslighttendencytowardslossoftotalvocabulary.Thislossbecame
marked,however,withthefoursubjectstestedaftertwoyears
.
TableVI:StoryTellingTask-RangeandVarietyofVocabulary
VocabularyRange=Numberofdifferentwords(R)I
VocabularyRange/Totalnumberofwords(R/W)
TIT2T3T4N
o.RR/WRR/WR
R/WRR/W
1.770.355730.384690
.411420.347
2.870.325880.333790
.357610.269
3.1270.2581050.321950.311660
.232
4.850.360680.433580.406
5.1460.4441470.3731300
.406
6,1060.3231140.3121170
.285
7.1500.2941770.4901690.324
8.800.386990.463800,316
9.820.423960.3841080
.401760.302
i.10.680.338990.364830.382
11,1480.6041060.424880.379 12.(notmeasuredduetoinabilityofsubjecttotelleitherofthe teststorieswell.) Mean96.00.39197.50.39688,70 .369 NativeSpeakersRR/W NS210ao.229 NS21050.536 NS3640.161 Comment:Therangeofvocabularyusedbysubjectsshowedadecline overthefirstyearaccompaniedbyarathersteeperdropintheratioof newwordstototalwords.Bothtendenciesacceleratedwiththefour studentstestedaftertwoyears.Interestingly,thesubjectsexhibiteda similarrangeofvocabularytonaivespeakers,exceptaftertwoyears , buttheirrange/totalwordsvalueswerehigherbecausetheyusedfar fewerwordsintellingtheirstories.,D.StoryRepetitionTask TableVII:StoryRepetitionTaskScores ScoringMethod:Onepointawardedforeachlexicalitemorgram-maticalmorpheme Onepointdeductedforomission,substitution,addi-tionorreconstructionwithlossofmeaningelement. Nopointsdeductedforreconstructionorsubstitu-tionwithnochangeofmeaning No.ITIITZIT31T4 MeanI202.251210.501214.17 NativeSpeakers NS1256 NS2241 NS3252 Comment:Ifanything,proficiencyonthistasktendedtoincrease slightlyinthefirstyearbeforefallingoffslightlyinthesecondyear.It maybethattheitemschosenweretooeasytocaptureanyproficiency changes. -125一
I f TableVIII:StoryRepetitionTaskScoresvTimeSinceReturn TimesincereturningtoJapan(T) ScoreonrepetitiontaskatTl(S} No.123456789101112 T2.11.101.10.90.90 .90.90.90.30.20.21.1 S144129205227182170206234237242236215 Comment:Thereseemstobeastrongcorrelationbetweentimesince returningandscoresonthefirstofthefourtestsadministered .This resultmayreflectthefactthattwosubjects{numbersland2)returned wellbeforethestudybeganandspentmuchoftheinterveningtimein averysmallreturneeclassthatofferedlittleopportunityforinterac- tioninEnglish.Thesetwosubjectsseemedtohavesufferedconsider-ableattritionuntiltheyjoinedthemuchmorenumeroussixthgrade classwhichprovidedampleopportunitiesforEnglishinteraction . iE .TranslationTask TableIX:TranslationTaskScores ScoringMethod:Eachsentencegradedonascaleoflto6. No.TIT2T3T4 1.54484747 2.150145150152 s.Is81sgIssI7i 4.172173178 5.166161164 s.1641571so 7.155162159 8.707374 9.173179181176 10.179179174 11.171174173 12.159169162 Mean65.0865 ,7565.83
Comment:Performanceonthistaskprovedremarkablystable.Again itmaybethattheitemsusedweretooeasytocaptureproficiency changes. TableX:TranslationTask‐CharacteristicErrorsinSubjectOutput a)SubjectswhostayedforarelativelyshorttimeinanEnglish-speakingcountrymademoredevelopmentalandintralingualerrors thanthosewhowereyoungerbutspentlongerabroad. E.g.IthinkIleftedintherestaurant. Italreadypassedahundredyearsafteritwasbuilted. Ifyoudidn'tleftanyvegetable,1'llboughtyouanice-cream. MaybehealwaysdoitonSunday. IthinkhedoesitinSundays. Iwasworry... Shedoesn'teattoomuch.(somuch) b)Ayoungersubject(X),thoughhavingstayedabroadlonger,em-ployedsimplerandshallowerstructuresthananoldersubject(Y) withlessoverseasexperience. E.g.(X)Ibrokethenextperson'swindowandtheguywasso mad. (Y)Thepersonwhoseglasswasbrokenwasveryangry. (X)Thedoctorsaid,"Youshouldstopsmoking"tomy father. (Y)Thedoctorsaidtomyfathernottosmoke. (X)Hedoesn'tgivemealetter.Ithinkitstrange. (Y)It'sverystrangethathedoesn'tsendmealetter. c)Withthelapseoftimethefollowingotherchangeswereobserved Simpler,shallowerstructuresappear. Subject3T1:ThisisthehouseIusedtolivewhenIwaslittle. T2:ThisisthehouseIlivedinwhenIwassmall. T3:ThisisthehousewhenIwaslittle. T1:Thatisthepersonwhomadeatelephones. T2:Heinventedtelephone. T3:Hehasinventedthetelephone. Subject10T1:Heistheonewhoinventedthephone. T2:Hethoughtofthephone. T3:He'stheonewhomadethetelephone. -127一
i Tl:Thepersonwhoownedthehousethathadthewin-dowbrokenwasverymad. "T2 :Themanwhogothiswindowbrokenwasverymad .r' T3:Theguywhogothiswindowbrokenwasverymad . SubjectllTI:Ithinkit'sweirdthathedoesn'tgivemeanyletters. T2:Ithinkit'sweirdthathedoesn'tgivemeanyletters .I T3:It'sweirdthathedoesn'tgivemeanyfettersto. Well-formedsentencesdisappear. SubjectllT1:Theownerofthehousethatthewindowwasbroken wasverymad. T2:ThepersonwholivedwhereIbrokethewindowwas verymad. T3:Theownerofthewindowthatbrokewasverymad . LIinterferenceerrorsbegintoappear.
Su切ect11麗:ll::翻;:dye
dye盤:舗:builtbuilt:l
T3:It'sbeenalmostahundredyearsbeforeit'sbeenbuilt . TZ:Heinventedthetelephone. T2:Heisthepersonwhoinventedthephone.I T3:Heisthepersonwhoinvented(Q)telephone . 4.Conclusion Aspointedoutin3.1,itwasdecidedtogoaheadwiththepresent studyeventhoughitwasimpossibletoadoptatruescientificdesigndue toalackofsubjectsandagreatdealofvarietyintheirpersonal characteristicsandbackgrounds.Thesimplepre-test,post-testpre-scientificdesignthatwasadoptedyieldedresultsthatcouldnoteasily becharacterisedbecausetheyfailedtoproducetheclearpatternsthat apurerdesignmighthaveyielded.Nevertheless,anumberofgeneral pointscanbemadewithsomeconfidenceaboutthethreequestions posedatthebeginningofthispaperWhileoursubjectsandtheirparentshadallclearlyshownadesireto attendaspecialreturneeclass(oftentravellinggreatdistancestodo so),theirprimarypurposewastoassistinreintegrationintoJapanese society.Generally,thesubjectsexpressedawishtomaintaintheir English,but,exceptintwoorthreecases,theirdesirewasnotmatched byacorrespondingefforttodoso.Otherconcernssoonbegantotake precedence,andwhilesomestudentsattendedspecialmaintenance classes,theirattendanceusuallylastedforashortperiodonly.Onlya fewstudentsseemedtogooutoftheirwaytomaintaintheirEnglishby readingorpositivelyseekingoutinteractionwithnativespeakers.In ourstudythetwoorthreestudentswhofellintothiscategorywereall girls,whichseemstosupporttheview,notonlythatgirlsaremore positivelyorientedtowardslanguagelearningandretention,butalso thatparentsarewillingtotakegreaterriskswithgirls. 4.2Towhatextentdotheysucceedinmaintainingtheirsecond language? Whiletheresultsofthestorytellingtestshowedadegreeofattrition inoverallproficiency,theothertestsseemedtoindicatethatthe subjects'L2waslargelybeingmaintained(thoughthereisapossibility thattwoofthetests,thestoryrepetitionandtranslationtasks,were tooeasytorevealattrition).Ourconclusion,then,isthattheregimeof returneeclassattendancewithsomedegreeofoutsidemaintenance activityislargelysufficienttomaintainEnglishproficiencyatleast overthetwoyearperiodofourstudy.Indeed,overthestudyperiod manyofthesubjectsgavethe(albeitsuperficial)impressionofactually gaininginconfidenceintheirspeakingability.Thisseemstocontradict thosewriterswhoseereturneesasasomewhatneglectedlinguistic resourcewhosesecondlanguageskillsarebeingsadlywasted.Infact, -129一
theprovisionofareturneeclasswiththeexpressaimofassisting readjustmenttoJapanesesocietyratherthanassistinglanguagemain- tenanceseemsneverthelesstoprovidereturneeswithsufficientoppor- tunitytointeractinEnglishtoavoidsignificantattrition.Sincestu-dentswhograduatefromsuchprogramsthenmoveontojuniorhigh andhighschools,inwhichEnglishisanimportantcomponentofthe curriculum,therearegroundsforthinkingthatlanguageskillwillin factberetainedintothefutureinmanycases. Thisisnottosaythatnochangestakeplace,andourstudyreveals severalsignsofinterlanguagechange,asfollows a)Alossofquantity‐onsuccessiveadministrationsofthestory tellingtaskthetotalnumberofwordstendedtofall. b)Alossoflexicalvariety‐againonthestorytellingtest,theratio ofdifferentwordstototalwordswasseentofall. c}Alossofsyntacticcomplexityandwell-formedness‐severalexam-piesofboththesephenomenawereshownbythetranslationtask. Innocase,however,wasthereasenseoflossoffluency.Thecasual observeroverhearingstudentresponsestosuccessivetestadministra-tionswouldprobablyconcludethatnolosswasoccurringsincethe subjectsseemedtocontinuefluentcommunicationwithoutapparent restraint.Infact,beneaththisdeceptivesurfaceakindofhollowingout andsimplificationdoesseemtooccurthoughwithouthinderingcommu-nicativeeffectiveness.ThisseemstoaccordwithOlshtain'sview (1986:p.192}that"...unlesswelookatspecificlanguagefeatures,we maynotseeanyobviousattritiontakingplace:onthecontrary ,it seemsthatthesubjects'secondperformanceismoremotivatedand relaxedthanthefirst." 4.3物'ノra`toysareZ吻octantindetermining4昭7θ6(ゾmaintenance
orattrition? Giventhecrudenessofourpilotstudyitisdifficulttocometoany veryfirmconclusionsaboutrelationshipsbetweenattrition,mainte-nanceandpredictorvariables.Wecanhowevermakethefollowing points a)Literacyandapositiveinvolvementinreadingseemshighlybenefi-cial. b)Olderchildren,perhapsbecauseofliteracyand/orgenerallinguis-ticmaturityseemnottosuffersimplificationoftheirutterancesover timeasmuchastheiryoungercounterparts. c}ChildrenwhowerebornorwholivedfromayoungageintheL2 environmentseemnottoexhibit,overtime,asmanydevelopmental errorsasothersubjects. d)Highproficiency,highmotivationtowardsretention,femalegen-derandactualretentionmaybemutuallyrelated. 」 Bibliography AndersonR.W.(1980)"DeterminingthelinguisticattributesofIan-guageattrition."inLambertR.D.andB.F.Freed(eds.)(1982)The LossofLanguageSkills.Rowley,Mass.:NewburyHouse . Burko-GleasonJ.(1980)"lnsightsfromchildlanguageacquisitionfor secondlanguageloss."inLambertR.D.andB.F.Freed(eds .) (1982)TheLossofLanguageSkills.Rowley,Mass.:Newbury House. CohenA.D.(1974}"CulverCitySpanishimmersionproject:Howdoes summerrecessaffectSpanishspeakingability?"inLanguage Learning.24:55-68. CohenA.D.{1989)"AttritionintheproductivelexiconoftwoPor-tuguesethirdlanguagespeakers."inStudiesinSecondLanguage -131一
Acquisition.11:135-149. GardnerR.C.(1980)"Socialfactorsinlanguageretention."inLambert R.D.andB.F.Freed(eds.)(1982)TheLossofLanguageSkills . Rowley,Mass.:NewburyHouse. GoodmanR.(1990)ノapan'sInternationalYouth.Oxford:Oxford UniversityPress. HarleyB.(1986)AgeinSecondLanguageAcquisition.Clevedon MultilingualMatters. KikuchiM.andA.E.Jackson(1995)"Measurementinstrumentsfor languageattritionstudies."ShoinReview(」 κ2η々ソuKiyo).37. LambertR.D.andJ.S.Moore(1986)"Problemareasinthestudyof languageattrition."inWeltersB.,K.DeBotandT.VanEls(eds.) LanguageAttritioninProgress.177-186Dordrecht:Foris. MoorcroftR.andR.C.Gardner(1987)"Linguisticfactorsinsecond languageIoss."LanguageLearning.37:327-340. NozawaS.(1993}"Bilingualism:Notjustkids'stuff"inDailyYomiuri (Tokyo}.18thMarch,1993. OblerL.K.(1980)"Neurolinguisticaspectsoflanguagelossasthey pertaintosecondlanguageattrition."inLambertR.D,andB.F. Freed(eds.)(1982)TheLo∬ 〔ゾ 加 η8初gθSkills.Rowley,Mass .: NewburyHouse. OlshtainE.(1986)"TheattritionofEnglishasasecondlanguagewith speakersofHebrew."inWeltensB.,K.DeBotandT.VanEls (eds.)LanguageAttritioninProgress,187-204Dordrecht:Foris. OlshtainE.(1989)"lssecondlanguageattritionthereversalofsecond languageacquisition?"inStudiesinSecondLanguageAcquisition. 11:151-165. OxfordR.L.{1982)"Researchonlanguageloss:Areviewwithimplica-lionsforforeignlanguageteaching."inModernLanguageJournal.
66:160-169. WeltensB,andA.D.Cohen(1989)"Languageattritionresearch:An introduction."inStudiesinSecondLanguageAcquisition.11: 127-133. WeltensB.,T.J.M.VanElsandE.Schils(1989}"Thelongterm retentionofFrenchbyDutchstudents."inStudiesinSecond LanguageAcquisition.11:205-216. YashiroK.(1990)"Foreignlanguagemaintenanceclassesfor returnees."TsukubaUniversityJournal.12:129-145. -133一
