OntheImportanceofImitationandRepetition
inForelgnLanguageLeamlng
TbshihikoYAMAOKA
l恒・ヾ′一J−′′fl、（ハ小。／、1i仙イJt・rh山川Jiり〃 Abstract Inthisarticle，imitationandrepetitionofexemplarSOfthetargetforeignlanguageusage is argued to be essentialinlearnlngthelanguage・Firstly，itis claimed that procedural knowledgeofthelanguageratherthanitsdeclarativeknowledgeisthetargetoflearnlngandthatthelearnlng Ofthe former consists ofestablishingform一meanlng COnneCtions ofthe
language・Itisalsoclaimedthatproceduralknowledgeofthetargetlanguageisbestlearned
throughthelearnlngOfexemplarsofthelanguageusageasinstancesdemonstratlngform一
meanlngCOnneCtionsofthelanguage．Theargumentsherearebasedontworecenttheories
relevanttolanguagelearning：AcTRbyAnderson（1993）and Usage−BasedModelby
Tomasello（2α）3）・Secondly，itisarguedthatininput−POOrleamingenvironmentsdeclarative knowledge ofthelanguage can be a facilitator ofthelearnlng Ofthe target procedural knowledge becauseit can assistlearnerS tOidentify the connections between formand meanlng・Thirdly，itisclaimedthatimitationandrepetitionplayacrucialrolerespectivelyin establishingform−meanlngCOnneCtionsofthelanguage・Finally，itispointedoutthatthe notorioustechniqueofoldpatternpracticecanandmustbereconceptualizedasacognltive typeofpracticepronodngthelearningoftheconnecdons・1．DeclarativeandProceduraIKnowledgeofLanguage
ltis gellerally acceptedthatthere are two difrerent kinds oflanguage knowledge：
declarativeandproceduralknowledge・Theformeristhetypeoffactualknowledgeabout
language anditis usually conscious or explicit．lIn contrast，thelatteris thetype of
perfbrmativeknowledgewhichisthedirectsourqeforApplicationintheuseoflanguageandSOitismosttypicallysubconsciousorimplicit・Itisclaimedthatitisverydifncultandtime COnSumlngaSWelltoemploythefbrmertypeofknowledgeinlanguageuse，eSpeCiallyinoral SpOntaneOuSuSe．
languagewithouthavingitsproceduralknowledge・Thisisprobablybecausetheselearners usually receiveformalinstructionofthelanguageand，aS aCOnSequenCeOfit，they have COnSCious declarative knowledge offbrmalru1es ofthelanguage，butthey do not have Sufficientopportunities fordeveloplng prOCeduralknowledge ofthelanguage・This may explain，atleastpartly，thegeneralsituationmostofthemfindthemselvestobein：Whilethey haveLtheknowledgeofthelanguage，theycannotengageeveninarudimentalconversationin
thelanguage・Theproblemwiththemissimpleandplain：theydonothavethenecessary
knowledgeofthetargetlanguageforitsspontaneoususe，thatis，PrOCeduralknowledgeofthe language． 2・DevelopmentofProceduralKnowIedgeofLanguageAnderson（1993），Whoisthe originator ofthe distinction between declarativeand PrOCeduralknowledge，StateSthatproceduralknowledgewhichiscomposedofproduction rulesisformedthroughtheprocessofproductioncompilation．Accordingtohim，thisisthe PrOCeSSOfestablishingconnectionsbetweenanteCedentsandconsequences．Inthecaseof language，anteCedentsare mearungsintendedtobe conveyedand consequences aretheir realizationaslinguisticfbrmsinproduction，Ontheonehand，andincomprehension，Onthe Other，theyarereversed，thatis，linguisticformsareantecedentsandmeanlngStObeextracted 丘omthemareconsequences．Productioncompilationinlanguageleamlngis，therefbre，the PrOCeSSOfestablishingformーmeamngCOnneCtionsinbothdirections．
Importantly，inhisACR−Rtheory，Anderson（1993：294）claimsthat“productionsrules
are created’’byleamersinitially by finding and experienclngthe connections between
antecedentsandconsequencesinindividualexemplarsasinstancesoftheseconnections，and thenbygeneralizingthemthroughrepeatedexperiencewithmoreexemplars．
Applyingtheexemplar−basedviewofproductioncompilationtolanguagelearnlng，itcan beclaimedthatproductionrulesofalanguageareeXemPlifiedinexemplarsofthelanguage useandthatlanguagelearners canfindand experience therulesin these exemplars individuallyitembyitemand，aSareSultofrepeatedexperiencewithmoreexemplars，they Cangeneralizethemasabstractrulesandleamthemassuch．Ormoreprecisely，itisonly throughexperiencingtheconnectionsbetweenformandmeamnglneXemPlarsofuseofa language thatlanguageleamerS Canlearn productions rules ofthelanguage．Itisalso necessary to recognizetheimportance ofrepeated experiencewith more exemplarsin generalizingtheseproductionrules．
AsimilarideaoflanguageleaminghasbeenproposedinUsage−BasedModel（UBM henceforth）byTomasello（2003）．2BasicallythismodelclaimsthatlanguageCOnStruCtions （notethatthismodelemploysthetermCOnStruCtionsinsteadoflinguisticrulesclaimingthat
peopledonothavesuchrulesbutconstruCtionsaspatternS）areleamedinitiallyasitem−based construCtionsinindividualexemplarsand that they are generalized as abstractlinguistic
construCtionsbyaccumulatingsimilaritemsand丘ndingcommonpatternSinthem．
ItisalsoclaimedinUBMthatfrequencyintwodif臨rentmeanlngSiscrucialinthe processofleaming：tOken丘equencyandtypefiequency・Whiletoken丘equency（howo鮎n the same exemplars withparticular construCtions qppearintheinput）contributes to the entrenchmentofitem−based constructions，typefrequency（howoften difEbrentexemplars WiththesameparticularconstruCtionsappearintheinput）isofservicetotheabstractionof theseconstruCtions．
AnemphasisshouldbeplacedhereonthefhctthatconstruCtionsinUBMare，infact，
actualizedthroughtheprocessoflingulStlCrealizationofmeamngswhichisgovernedby
PrOductionsrulesconnectingformwithmeanlnginAnderson’sterm．Thusthelearn1ngOf COnStruCtionsofthetargetlanguagecanalsobeseenasthe processofestablishingfbrm− meanmgCOnneCtionsofthelanguage・ CommontoACTLRtheoryandUBMaretheideasthatlearnersstarttheleamlngOfthe targetknowledgewithitem−basedleamlngOfexemplarSandthattheygeneralizeandabstract theknowledgethroughaccumulatingsimilarexemplarsand丘ndingcommonfbaturesamong them． 3・La独guageLearnungi血PoreignLa血guageLcar払hgE皿Viro放me鵬S UBMisamodelfor丘rstlanguageleamlngandACTLRisalsopertinenttofirstlanguage leamlng．Itshouldbestressed，however，thattheseviewsoflanguageleamlngareaPPlicable infbreign／secondlanguageleamlngaSWell，Sincethereseemstobenootherwayoflearnlng PrOCeduralknowledge ofthe targetlanguage as productionrules or construCtions than experienclngtheconnectionsbetweenfbrmandmeanlngOfthelanguagebasedonexemplarS initiallyasindividualitems（token丘equency）andthenbyaccumulatingthemfbrabstraction totakeplace（type丘equency）．Procedumlknowledgeoflanguageisdynamicinnatureand， there丘）re，is not something to be describedinwords：they areleamable only through experienclngtheform−meanlngCOnneCtionsbylearnersthemselves．Caremustbetaken，however，tOthesituationaldi飴renceoflearningbetweennattFal
丘Btandsecondlanguagelearnlngandforeignlanguageleaming・NaturalsituationsarelnPut− rich，butforeignlanguageSituationsarelnPut−POOr・Ininput−richsituations，1eamershave abundantopportunitiesofleamingbothintermSOftokenandtypefiequency・Itisnotthecase， however，inforeignlanguageleamlng・Somekindofactivemeasures，therefore，muStbe takeninfbreignlanguagelearnlngtOenrichtheinputandincreasethe丘equencyofexemplarlemigbothastokenandtype．
4．WeakInterfhcePositioninForeignLanguageLearnug
Concemlng the developmental relationship between declarative knowledge and
proceduralknowledge oflanguage，there are three di飴rent theoreticalpositions：Non− interfhce Position，StrongInterfhce position，and WeakInterfhce Position・Non−interfhce Positioninsiststhatthesetwotypesofknowledgearemutuallyindependentdevelopmentally （Krashenl982）．StrongInterfhce Position contendsthatdeclarativeknowledge somehow converts directlyinto proceduralknowledge（Stevickl980，Bialystok1981，McLaughlin
1978）．Incontrast，WeakInterfhcePositionclaimsthatwhiledeclarativeknowledgeitself doesnotturnintoproceduralknowledge，itcanindirectlyassisttheleamngofprocedural knowledgeasafhcilitatoroflearning（Seligerl979）・
AsEllis（1997：114）pointsout，WhilebothNon−interfhcePositionandStrongInterfhce Positionaredoubtfu1，WeakInterfhcePositionistenable．ItisinthisperspectiveofWeak Interfhce Positionthatthe currenttopics ofconsciousness−raising（CR）（Sharwood Smith 1991），nOticing（Schmidtl990），andthustheroleofformalinstruction，Shouldbediscussed andinterpreted．Firstly，CRisdennedasadeliberateattempttoraiseleamerS’consciousness Onfbrmalpropertiesofthetargetlanguagewithaviewtofhcilitatlngthedevelopmentofthe targetlanguageknowledgeorproceduralknowledgeofthelanguage（SharwoodSmith1991： 118）．NotethatCRtofbrmalpropertiesentai1S丘ndingtheirconnectionstothemeaningsthey embody．Secondly，Noticinghypothesisclaimsthatasecondlanguageleamerwi1lbeginto leamthetargetlikefbrmifandonlyifitispresentincomprehendedinputandnoticedinthe normalsenseoftheword，thatis，COnSCiously（SchmidtandFrota1986：311）．Inthiscasealso， COnneCtingtheformwithitsmeanlnglSaSSumedsince，aSisclearlystated，thefbrmisnoticed incomprehendedinput．
Theargument above makesthe point clearthat declarativeknowledge ofthetarget languagelearnedasaresultofhavingformalinstruCtionofanykindishelpfulandnecessary inlnPut−POOrleamlng enVironments．Thisis becausein order fbrleamersinthese environmentstosuccess魚llly丘ndandexperiencetheconnectionsbetweenformandmeanlng ase伍cientlyaspossible，theyneedtoknowtheformsorformalpropertiesofthelanguage beforehand．Actually，ittakesmuchtimefbrleamersingeneralto丘ndformal丘aturesifthey
areleftontheirownaSisthecaseinnaturalleaming．CRisanattemPttOreducethetime
neededfbridentiLyingformalfiatt汀eSaSmuChaspossible．ItshouldberememberedthatCRissaidtobeashortcuttoleaming（SharwoodSmith1991）．
Inconnectionwiththis，itmightbenoteworthythat“apuremeaning−focusedapproach tofbreignlanguageteachingismlSguided，butthereAremanyPOSSiblewaystocombine exposure tolnPutand communicativepracticewithafocus on fbrmand consciouspess raising”（Schmidt1995：47−48）．5．ImitationandRepetitioJlinForeignLanguageLearnlng
Foreignlang血gelearnerS at beginming stagesin particular need tbintensively experienceexemplarsofthetargetlanguageusageinordertoleamprOCeduralknowledgeof thelanguageWhichiscomposedofproductionru1esconnectingfbrmwithmeanlngwiththe helpofdeclarativeknowledgeofthelanguage．Thisissueisconcemedwithhowtoraisethe 丘equencyofexperiencewithexemplarsnotonlyastokenbutalsoastypeanditisherethat imitationandrepetitionbecomecruCial． BoththetermSOfimitationandrepetition，however，haveanunhappyhistoryinforeign languagelearn1ngandteaching．ThesetermShavebeenassociatedwithabehavioristviewof languagelearnlngandthusitisassertedthattheycannotsuccessfu11yexplaintheprocessof languageleammg． Inspiteofthissimplisticandmechanicalinterpretationofimitationandrepetition，these two processes，in fhct，Can be quite complexand active whenthey are seen丘omthe PerSPeCtiveofconnectingfbrmwithmeaninginlanguageleaming・Speidel＆Nelson（1989：
4），forexamPle，Observethat”…fhrfrombeingsimple，imitationistheresultofmany
transfbrmationsofenergyand activation ofmany neuralpattems．Itisthe resultofnew
incomingformationmeshingwithinformationalreadyinmemory．’’Imitadonconceivedofin
thiswaylSnOtaParrOtlikeresponsetoamodelbutanaCtiveattemptonthepartofleamersto experiencetheprocessoflinguistic realiz戚ionofspeCi丘c meamngsoncemore．Inother WOrds，thisisanOPpOrtunityfbr experiencing the encoding processaneW bylearners
themselvcs． Thusadistinctionmustbemadecarefullybetweentwodi飴renttypesofimitation：the typeofroteimitationandthetypeofencodingimitation．Theformerisapamtlikere坤OnSe andthelatteranaCtiveself：confirmationoftheprocessofencoding・Thedenialofthefbrmer asadrivingforceoflearnlngmuStnOteXCludethepotentialacceleratingroleofthelatter． RepetitioncanalsobedividedintwodiffbrenttypeS：rePetitionofthe sameidentical utterance and repetition ofdiffbrenttypes ofutterances havingthe samefom1−meaning COnneCtion．Evenintheformer，itcanbequiteactiveinthatitistheprocessofrepetitive experience ofthe encoding process ofthe utteranCe eaCh time anew．This undoubtedly
COntributestoentrenchingtheencodingprocessasUBMproposestobethefunctionoftoken
丘equency．
Repetitioninthelatter serveS the role of abstracting particular fbrm−meamng
COnneCtions．ForexamPle，byrepeating”Theboyisrunmng，”“Thegirlisswimmlng，”“They aredoingsports，のandsoon，leamerscaneXPeriencethecormectionbetweentheprogressive
fbrmwithitscorrespondingprogressivemeamgindividuallyandthuscangeneralizethe COnneCtion・Thismechanismclearlycorrespondstotheroleoftype丘equencyinUBM・
Tosumup・InsplteOftheunhappyassociationwithabehavioristandmechanicalview Oflanguageleamlng，lmitationandrepetitioncanbeseentobequiteactiveinthattheyare opportunitieもfortheleamertoexperienceconnectingformwithmeanlng・Thusircanbe Claimedthattheyareindispensableinlanguagelearnlng，eSPeCiallysoinfbreignlanguage learnlng．Notethatpriordeclarativeknowledgeofformishelpfulinimitationandrepetition・ 6．01dPatternPractice Thediscussionssofhrmightremindusofpatternpracticeprevalentinthe1950sand 1960S．Thisnotoriouswayofpracticewascondemnedtobetheoreticallyunwarrantedand practicallyquiteunsuccessfu1．Althoughcarefulreadingoftherationaleofpattempractice
reveals a sound reasonlng aSin“PatternPraCtice c坤italizes onthemind’s capaclty tO
perceiveidentityofstruCtureWherethereisdiffbrenceincontentanditsquicknesstoleamby
analogy．”（Brooks1964：146）andinspiteofthefhctthatsomeleamersdidsucceedwiththis
PraCtice，itresultedinageneralfhilure．
Withhindsight，SeVeralproblemscanbepointedoutwithpatternpractice・Firstitwas basedonamechanisticviewofbehavioristpsychologylnWhichlearnerswerenotconsidered
to be active agents oflearnlng andit was claimed thatleamlng WaS detemined by
environmerrtalfhctors only．Second，leamers were e坤eCtedto pay atterrtion only to the meanlngSOfcuesglVentOthemattheexpenseofrecognizingtheconnectionsbetweenform andmeanlngaSisclearinthefollowlngquOtation：”PatternPraCticeisrapidoraldrillon PrOblempatternswithattentiononsome血hgotherthantheproblemitseは”（Lado1964：105）．
Thirdly，PatternPraCticewaslimitedtooralproductiononlyandrdectedthepossibilityof
PracticethroughComprehension．Fourthly，PattemPraCticerequiredleamerstomake oral
responsesasquicklyaspossibleinthehopethattheirresponsescouldbecomehabitualand automatic． Becauseofal1theseproblems，aSeaSilycanbeexpected，PattemPraCticeresultedina r慮hermechanisticandmeaninglessandtherefore丘uitlesswayofpractice．7．CognitivePa触mPractice
Beforegolngintothediscussionofanewinterpretationofpattempracticeasateclmiqu6
for promotingleamers’leamlng Ofform一meamng COnneCtions bothin terms oftheir
entrenchmentandabstraction，itmightbeinterestlngtOrefbrtoaninsightfuldiscussiononthe
relationshipbetween‘‘hbit’’and’’ru1e．’’Inhislongneglectedarticle，Carroll（1971），rqjecting
th￠behavioristinterpretationof“habit”byChomsky（1959），Claimsthatthenotionof“habit”
…I would emphasizethatthe notion of“habit”is much morefundamental， psychologically，thanthenotionofはrules・”Aはru1eHissimplyafbrmal，usually，Verbal， statementoftheconditionsunderwhichsomethingisexpectedtooccurornottooccur， usuallyundercertainsanctions・Assuchitisanabstractionoraconstructinsomesense independentofactualbehavior・WbarefhmiliarwiththefhctthatpeoplecanSpeaka languagewithoutanyCOnSCiousknowledgeorapplicationoftheHrulesMthataresaidto −underlietheirlanguage．A“habit”however，isarealthingthatsomehowresidedina person．Itiswhathehaslearned・AslanguageteaCherswearetrylngtOPrOduceand changecertainkindsofhabits．Ifpsychologyhasanythingtosaytolanguageteachers，it willsayitabout“habits，”notabout”rules，”exceptpossiblyassecond−Orderphenomena． Youmaylooktolinguisticsforinformationabout“rules．” （103−4） AccordingtoCarroll（1971），ahabitisde鎖nedasanyleameddispositiontoperceive， behave，OrPerforminacertainmannerunderspeci丘edcircumstanCeS．Itisundoubtedlythe casethatahabitisfbrmedthroughrepeatedexperiencewiththesamepattem．Itshouldbe noted，however，thatapatteminpattempracticere缶rstoastruCturalpattern．Itshouldalsobe rememberedthatastruCturalpatternisconceivableonlyincolt）unCtionwiththemeaningthe Pattemembodies．Thus，theprocessofpattem丘ndinginvoIvestheprocessofidentitying fbrm一meanlngCOnneCtionsinexemplarsentences．Inadditiontothis，COmmOnmeanlng−form
COnneCdons must beidentinedand generalizedby experienclng eXemplarswithdi蝕rent
SentenCemeamngS．Usingthesameexamplescitedearlierinthisarticle，inexperienclng“The boyisru月山ng，”“Thegirlisswimig，”’’Theyaredoingsports，”andsoon，eaChofwhich hasadifrbrentsentencemeanlng，learnerSareinvoIvedinthetasktofindthecommonfbrm− meanlngCOnneC60ninthem：theconnectionbetweenthepresentprogressiveformandthe PreSentPrOgreSSivemeanlng．Notethattheprocessofpattemfindinganditsgeneralizationis theveryprocessofhabitformationinthatform−meanlngCOnneCtionsareestablished． Theprocessofhabitformationis，therefore，fhr丘ommechanicalandmeamngless：itis COgnitiveand meaningfulin natureinvoIvingidentification ofparticularform−meanlng COnneCtionsandtheirestablishmentasgeneralpatternsofconnections．
It must be stressedthatidentitication ofpartiCularform−mean1ng COnneCtions，their abstractionaspatterns，andthusmakingthemhabitsarealldonebasedonexemplars・Inthis SenSe，eXemPlarsoflanguageuSageareaSinequanonforlanguagelearnlng：nOlanguage
leamngcanoccurwithoutexperienclngeXemPlarsoflanguageusage．WhenaneXemPlarofa
SPeCificform−meanmgCOnneCtionisexperiencedbythelearnerasanitemdemonstratingtheCOnneCtion，itcan1eadtohisorheridentificationoftheconnection．WhenthesameeXemPlar
isexperiencedrepeatedlyeitherreceptivelyorproductivelybytheleamer，itmayresultinthe entrenchment ofthat connection by theleamer．And when di飴renttypes Ofexemplarsembodyingthe sameform−me弧ng COnneCtion are experienced by theleamer，itwi11 COntributetoitsabstractionasapattern，thatis，ahabit．
Tosum叩SOfhr・Languqgeleammgconsistsinexemplarleamlng・Exemplarsprovide theleamerwiththekeytolanguagelearmnglnSeVeralimportantwaysatleast：identification
Of speci丘Cform−meanmg COnneCtions，entrenChment of these connections，and their
generalization．The血stidentificationisdoneasitemleamlngOftheexemplar・Thesecond
entrenchmentisattainedbyexperiencingthesameeXemplarrepeatedly（token丘equency）． ．Thethirdgeneralizationiscompletedthroughexperiencingdiffbrenttypesoftheexemplar （type丘equency）．By going throughthese steps，learners can succeedinthelemig of PrOCeduralknowledgeofthetargetlanguage，thatis，PrOductionrulesconnectingformwith meamngorconstruCtionsofthelanguage．Itgoeswithoutsayingthatimitationandrepetition areessentialinthesestepsofleamlng． 01dpatternPraCticecan，therefore，bereconceptualizedinthisnewperspectiveasatype OfpracticeforpromotingthelearnlnglnVOIvingidentificationofform一meanlngCOnneCtionsOfthetargetlanguage，theirentrenclment，andtheirgeneralization．Pattempracticeinthis
newgulSeCanbeca11edcognitivepatternpractice．Ithasthefbllowlngcharacteristics： 1．Itiscognitiveratherthanmechanical． 2・Leamers’attentionlSPaidbothtofbrmandmeadllngratherthanOnCueSalone．3・LeamerSCanimitateandrepeatexemplarsattheirownPaCe．
4，lmitationandrepetitioncanbebothreceptiveandproductive．
There seems to be no need to elaborate more on the nrst two characteristics above．
Theaimofcognitivepatternpracticeisontheestablishmentofform一meanlngcOnneCtionsof
thetargetlanguage・Automaticityinexecutingtheconnectionsisnottheaimofpractice；itis aresultofpractice・LearnerSCan，therefbre，PraCticeattheirownPaCedevotlngtOthetaskof
establishingfbrm−meむungCOnneCtions．
lmitationcanbe bothoverand covert・Itisnotessentialtoimitateovertly丘）rthe establishment ofform−meanlng COrmeCdons・Repetidon canalso be both receptive and PrOductive・TheideaofinputprocessingbyⅥmPattenandCadiemo（1993）correspondsto receptiverepetitionofdi飴renttypeSOfexemplarshavingthesamemeaming−formconnec銭on asinthefbllowingdefinition：InputprocesslngHinvoIvesthosestrategiesandstrategiesand mechanismsthatpromoteform一meaningconnectionsduringcomprehension’’（Wattenand Cadierno1993：226）． AsⅤ彿Patten，WilliamS，＆Rott（2004：4）claims，theestablislmentoffbrm−memig COnneCtionsisahndamentalaspectofbothnrstandsecondla喝uageaCquisition．Asisalso
Claimed bythem，aform一meaJung COmeCdon must gothrough each ofthe fbllowlng
processesinordertobefullyleamed：makingtheinitialconnection，SubsequentprocesslngOf theconnection，aJldaccesslngtheconnectionforuse．Thesearereallyessentialprocessesof languageleamlng，andimitationandrepetitioncanPlayadecisiverolerespectivelyinthese processeswhichpatternPraCticeinanewcognitiveinterpretationcanacCOrrmOdate・ 8．ConClusions ThefollowlngSCanbestatedasconclusionsofthisarticle．Proceduralknowledgeofthe targetlanguageconsistsofproductionsrulesofthelanguageCOnneCtingformwithmeanlng orknowledgeofconstruCtions embodyingtheirrespective meanlngS・ThelearnlngOfthe targetlanguageproceduralknowledgeisestablishingits丘）rm−meanlngCOnneCtionsanditis
donebasedontheleamingofexemplarsexemplifyingspecificfbrm−meahingconnections・It
isstressedthatimitationandrepetitionarecruCialintheprocessofleamlng．Oldpattem
PraCticecanbereconceptualizedasacogmitivetypeOfpracticeseenfromtheperspectiveof
establishingform一meanlngCOnneCtionsofthetargetlanguage． Notes l．Thestatusofconsciousnessofdeclarativeknowledgeisdi鮪rent−inACT＊modelandACTLR model byAnderson，0riginator of the declarative−PrOCeduraldistinction．While
PrOCeduralknowledgeis defined to be consciousinACT＊（Anderson1983），it canbe
SubconsciousinACTLR（Anderson1993）．Subconsciousdeclarativeknowledgeconsistsof
“declarativememoryforexamplesofhowtheproceduresshouldbeexecuted’’（Anderson＆
Fincham1994：1322）・Theclaimthatdeclarativeknowledgecanbesubconsciousdoes．not denytheexistenceofconsciousdeclarativeknowledge．Importantly，however，Subconscl0uS declarative knowledgeis a stepplng StOne fbr developlng PrOCeduralknowledgejust as COnSciousdeclarativeknowledgeis． 2・Usage−BasedModelclaimsthatlanguagestruCtureemergeS丘omactuallanguagepseor usageandthatitisofnousetopresupposeanykindofimatefhcultyfortheleamngof languageStruCture． Re鮎remces
Anderson，J．R．1983．TheAtT：hitecture QlCognition・MallWah，NJ；Lawrence，Erlbaum，
’Associates．Anderson，J．R．1993．RulesQftheMindHillsdale，NJ：LawrenceErlbaumAssociates・
Anderson，J．R．，＆Fincham，J．M．1994．AcquisitionofProceduralSkills丘omExamPles・ 血〟r乃α1〆物eri椚e肋J極力0わ紗ご上eαr〃f喝地肌0堺。〃dCo伊仙〝20：1322−1340・ Bialystok，E．1981．SomeEvidencefbrtheIntegrityandInteractionsofTwoKnowledgeSources．InAndersen，R．（Ed．）Nbw Dimensionsin Sbcond Language Aequisition
Brooks，N．1964．LanguageandLa77guLqeLearning：TheofyandPractiee・（SecondEdition）・ NewYork：Harcourt，Brace＆World．
Carroll，J．B．1971．CurrentIssuesin Psycholinguisticsand Second Language Teaching． 7篭ざ0⊥2〟αr絶r少5：101−114． Chomsky，N．1959．AReviewofB．F．Skinner’S柁rbalBehaviorlLanguage35：26−58・ Ellis，R．1997．SLAResearchandLanguage7baching・Oxfbrd：OxfordUniversityPress・ KraShen，S．D．1982．Pr乃Cわわが〟〝dPmCrce 〃腿co〃d上。〃g〝明野dc〟誠わ乃・0Xbrd： PergamonPress． Lado，R．1964．Language乃aching：AScientyic4zpTDaCh・NewYork：McGrawHill・ McLaughlin，B・1978・TheMonitorModel‥SomeMethodologlCalConsiderations・Language 上eαr乃乃g28：309−332．
Schmidt，R．W．1990．TheRole ofConsciousnessinSecondLanguage Learnlng．A”lied
上物相r短11：129−158．
Schmidt，R．W．1995．ConsciousnessandForeignLanguageLearnmg：ATutorialontheRole
OfAttention and Awarenessin Leamig．In Schmidt，R・W・（Ed・）・Attention and
A14ノ〝ene．WinFbreWILa7mLearning．fIonolulu：SecondLanguageTeachingand CurriculumCenter，UniversityofHawai’i． Schmidt，R．W．，＆Frota，S．N．1986．DevelopingBasicConversationalAbilityinaSecond Language：ACaseStudyofanAdultLeamerofPortuguese・InDay，R．D・（Ed．）．TbLking toLearn：ConversationinSecondLanguageAcquisition・Rowley，MA：NewburyHouse． Seliger，H．・1979．OntheNatureandFunctionofLanguageRdesinLanguageTeaching． 閻0上砂′αrねゆ13：359−369． SharwoodSmidl，M．1991．SpeakingtoManyMinds：OntheRelevanceofDi飴rentTypesof LanguageInformationfortheL2Leamer．SbcondLanguageResearch7：118二132． Speidel，G・E．，＆Nelson，K．E．1989．AFreshLookatImitadoninLanguageLeammg，In
