A Study on Teaching Inanimate Subjects

全文

(1)A Study on Teaching lnanimate Subjects. A Thesis. Presented to. The Facuity of the Graduate Course at. Hyego University of Teaeher Education. In Partial Fulfiiment. of the Requirements for the Degree Master of Education. by. Shigeki 1（amarno to. （ Student Nunber ： 82237 ）. Hyogo University of Teacher Educatien. Deeember・ t 985.

(2) PREIiiACE. ；t is often said that one of the pinrposes of English. teaching is to eultivate intermational understqnding in the iearner．，Through teaehing English， Engiish teaehers have to try to make learners understand how the 1inglish think and feel． lt is doubtfui， however， the present textbooks are. cornpesed se as to make Japanese learners understand English way of thinking．， An EngZish inanimate subject sentenee is a peeul． iar， sentenee strueture to us not oniy frora’ @the point. of view of ・the sentence st？ucture but alse from the point. ef view of the ”eoneeption”． Cenception is the notion that. substsntialize a theught into an aetual language form． lf we emphasize the necessity of EngXish teaehing for the un・一. derstanding◎f Engli8h way of thin腱ng， it seems essen七ial. that an inanimate subjeet sentenee has to be introduaed as a step to the goai in eur English teaehing． ln aetuaiity，. however， sueh a point is not eonsidered in a textbook making． In this thesis， 1 Xooked for the method to utiiize the. eomparisen between English and Japane＄e at the ievel of eon−. cep ti◎パby using臨glish inanimate subject sen七ences and their Japanese equivaients． 1 made a brief description of what an inanimate subjeet sentenee is and of the ，difference between EltgMsh and its Japanese equivalents at the level ，／’／）. of eoneeption． Then 1 listed implieationS obtained from the. 一i一.

(3) comparisen． Ot eourse， this thesis is not perfeet and has rnany defeets because there are many new hypotheses in it；. the construetion of hierarehy of conception， the eomparison’ between two Zanguages at the level ef conception， the new markedness dif：ferential hypothesis，．and s◎ on． So，工 raust. try to search for a better・eontrastive analysis hypothesis and， by verifying it in practiee， have a better eomparison. between Engiish ahd 」apanese at the levei 6f eoneeption for. the better English teaching． ’Lastly， 1 would ，like to express my hearty ，thanks to. Prof． Masamichi Tanaka， whe gave me a iot of preeious advice and help． Bu’t fer his suppert， 1 could net have cempieted. this七hesis・Andエw叫d like t◎thank Pr。f・Sh。r。ku A。ki・ Prof．工chiro Mange， P：pef． To shihiko Yarnaoka， and tho se wh◎. gave me heip direetiy or indireetiy．，. Deeember， 1985．. 一 ii e. S．’ Yamamo to.

(4) CON①E聾1刃S. PREIE“ACE. CHAPTER r． CONTRASTZVE ANALYSTS 1． 1Alhat the Contrastive Analysis is ．．．．．．．．．．．．．．．．． 1. 2． The Mstorical Sketch of the Contrastive Analysis ・・”一一e一… 一・ 9. 3． The New Merkedness Mfterentiai Hypothesis and i七s ：P◎s8ibili七y 。・・●の。●σ。9・●・の。．●○・e・●●●。のの・i。●。・ 20. CNIAPTER II． CONTRASYVE ANALYS：S BETWEEN ENGIL，ISH INANIMATII SUBJECTS AND THEIR JAPANESE EQUIVALENTS ＄一． Thought， Language， and Conception ．．．．．．．．．．．．．．； 26 2． 11he Hierarehy of Coneeption ．．．．．．．．．．．．．．．．．．．．． 30. 3． l11he Contrastive Analysis and its lmPlieations ．．． 33. CHAP田ER I工工． HOW TO TEACH INAN工MAτE SUBJECTS l． An Examinatior｝ of English Textbeoks ．．．．．．．．．．．．． 47. 2． Proposal for the Effeetive Teaching of lnanimate Su，bjec七s．．．．．．．。．．．．・。．・．。。。．．．．．．c。．。．。．．。．・．．。50. BIBL1OGRAPE［Y e．．．i．．．．．．．．．・．．．．．．．e．．．．．．．．．．．．．i．．．．．．． 55. APPENDrXES Appendix A．．．．．．．．．．．．．e．．．．．．．．．．．e．．．．．．．e．．．．．．． 59 Appendix B．．．．．．．．ee．．．．．．．．．．．．．．e．．．．．．．．．．．．．．．． 61 App endiX C．．．e・．．．．．．．．．．．e．．．．．．．．．．．．．．．．．e．．．．． 101.

(5) CHAPTER I. CONTRASTIVE ANALYSIS. t． W’hat the Contrastive AnaZysis is The eomparative study ef two or more languages is not a. new study but the one which has been done mainly sinee the late. 19 eentury as eomparative linguistic＄． The eontrastive linguis− ties， however， to which paany peop，ie who engage themseXves in. second or foreign language teaehing pay very mueh attention from the late 50｝s to the eariy 60，s， is different from the. comparative linguistics． The purpose Qf both the eontrastive iinguisties and the eomparative linguisties is to elueidate the simiiarities and the differences of two languages by eon− trasting and eomparing them， but we ean know the diffeteenee. of the twe iinguisties by the toUowing quotation．［lboo eases are eensidered in the eomparisQn of two er rnere languages． One is the case that’・language A and language B are the cognate． 1n this case， we ean expect by the eorapari son that the twe languages hold systematieaZly common parts in terms ot voeaburary， the sound sYstems， the syntax structu？es and so on． Or even if eommon parts are not found，． at ieqst there exist systentatieal eerrespendenees． The other is the case that lariguage A and Zanguage B are not the eog−ts nate・． ln thiS ease， we ean not expect to find the s灘弓懸亀鎗蝕1蓼．！Gの㎜駐P碑紬。卸。◎鞭㊤・⑬蜜三顧“e醇・．s㊤◎. the cemparisen i・s done to investigate ＄imiiarities and differences between the two languages． ff1he eern− eg：it．SO．n．Rggg．e．？Y．．gXgkehaand Japanese eomes under the. ‘一・一・. @1. Taishi ｛Vagiri，. Nli chi−ei o 】ユ。 h o. en ne hikak． u （The Cern一田h◎8e o f Ehl藝茎茎：亘ア丁. arison between Japanese Exnressions and. Gendai eige kyoiku ko za 7. ： Mchi−eigo no hikaku，’ 〈Tok；yo ：. Kenkyusha， 1965）， p． 85．. 一 1 一.

(6) In七his quota七ion，七he fomer is七he c◎mpamtive iinguistics and 七he Za七むer is t；he c◎n七：Pa8tivg linguisties・. Taking 七he above statedもhings ：into conside：ration， le七 us の. 。◎nsider．wha七 the cont：rじast：ive analy8is is and what kind o：f lin−. guis七ic and psycho↓◎gica．：L theories a：ne behind i七：in mo：r臼 detail. in this section． First，七he contrastive ana↓ysis uses七he con七ras雛ve lin。 guis七ics， name：Ly， i七u4しil：Lzes 七he results of c◎mp， aring and c◎n一七：rastin9 七w◎ ：Languages （：Lea：rner曾s mo七her ：Language and his 七a：r−. ge七 1anguage） f◎r：fo：Peign language 七eaching． It has been di：e鱒 fused from the la七e 50薯8 七〇七he early 60響s． In the nex七 qu◎七a−. tion we can find．the、：eundamen七al ：peason why lingu：Ls七8 and Zan−. guage teachers鵬ade the prac七ical apPlication o：e the con七：rastiマe analysis 七〇 the seeond o：p 七h6 fo：ee：ign language 七eaching．田his hyp◎thes：is， deeply：ro◎七ed in behavi◎rism and st：ructuralism， clai＝ned 七ha「』 the principa⊥ bar：rie：r. to second language acquisition is the interference of七he fir8t language syste醜wi七h七he secgnd iangu− age system， and that a scien七ific， struetura：L anal− y8is o：e the 七w◎ ⊥anguages in question wou：Ld yield a taxonomy of linguistic cOnt：【・a8t getween 七hem. 画き。諸iき畿i麟d、：￡誌面、言’二面七と9．P｛edict As 七his quota七i◎n clearly showsg the reason．why language teach噸 ers and ＝Linguists paid attention ’bo rbhe con七ra8七iv臼 anaユ．ysis. in 七he field o：f ：eoreign language teaching is．七he・idea that 七he. 1 K．. Teaehin. Dougias Brown， Principles of lsanguaRe lliearning and ，（New Je：【唱ey 3. P’r”entieed−Hall， lne．， Engiewoed Cliffs，. t980）， p． 148．. 鱒 2 一.

(7) most important obstaeie of learning a fotseign language is inter− ference frorn learnerts mether tongue． So in erder to Zessen. the intepference effect as mueh as possibie， they thought・it. was necessayy to eiucidate the similarities and differences o￡．leapne． r’s ！n．oV．hen tongue and hi s target ianguage by eom一，. paring end eentrasting the two languages， to prediet in which. points ef the target language interference takes plaee， and to make the best use of the predietion in the actual situation of torsign language teaehing． There￡ore， seienti￡ic eornparison between twe languages i s required in order te put the eontras−. tive anaiysis into practiee eorreetly． At the very Xeast this version demands ef linguists that they have available a set Q￡ linguistiq univer・一 sals formulated within a comprehensive linguistic theery which deals adequately with symtax， semantics， and phonQlogy．］ivrthermore， it requires that they have a theory o￡ contrastive linguisties into whieh they Gan ’plug eomplete linguistic deseriptions of the two ianguages being eontrasted so as to produce vag．．eogreet set ot eentrasts between the two Xangu− ages．. As this quotation shows， in order to eompare two ianguages sci− entifieally， following iterns have to be satisfied．（1＞ Seien一・. tific and descriptive analysis of the target language to be ユearne（1。（2） Scientific and de8c：pip七ive analy8is of 七he mo七her language．（3） Seientifie eomparison o￡（1） and （2） so as，to. qleariy show the differenee＄ of strueturai’patterns whieh the. 4. RenaZd VI ardhaugh，. mpi ， Vo l．. itThe Contrastive Analysis Hypothesis，t’ 4， No． 2， Jun．，（1970）， p． 125．. 一5一.

(8) two languages have． On this point， Banathy expressed the same idea．. Prerequisite to the’ preparation of a eontrastive lin− gui． stic analy．sis aye adequate linguistie deseriptions of”the’native language o￡ the student and of the tar− get language． A eareful c6mparisen・of the two systerns. wiU furnish the idea for the eontrastive analysts． The． data thus acerued eonstitute the rnost signifieant infbrmation’the 1inguist can o’f￡er to’language teaeh−. ing sinett．． they identify the target elements of a lan−. guage course． Again， what the student has to learn equals the sun of the qifferenees estabiished by the co− 獅狽窒≠唐狽奄魔?@anaXysis． t. And it was the structural lingui stics that brought us the tech一. nique to eompare two languages aeeurately and seientifieaUy．. It was possible beeause the struetural linguists estabiished their theories by acquiring data oniy from observable language. materiaZs． Struetural Zinguists made an assunption that aU. language items in a language system are set up ac一・. eerding to their ewn funetiens in eaeh ianguage system， and made rnany trials to deseribe a ianguage oniy by the’ @given， observable， and limited Zanguage data to prove ，their assunption ’seientifically and positiveiy． For example， Fries （1952） reeorded 50 hours telephone calls which ineluded about 250，000 words as voiced data to deseribe the strueture ef modemb English， and observed and classi￡ied them． He．adopted sueh a way o￡ eoUecting data because of the premise that itaLZ structurai signs are the 蓋霧晦。躍整h’ch ean Pe desc「’bed by f◎「m。「. 1 Bela Banathy， Edkh erewell Wrager， and Carl D． doddle， ”The Use ef Contrastive Data in georeign Language Course Devel−. opment，” in rwt d L T h ， u q d i Aibeit Vaidm an，（New Yo rk ： ttgraw−HUI Book C ovap any， 1966）， p． 37．. 七．識・＆識E蓋畿登㌔凱撒，a衰艦b￥至￡h麗盤藍玉n評幽 Eigogaku taikei A 2，. tokyo ： Tai shukan， e 982）， p．. 一4一. 9．.

(9) W，’e can say that the way of colleeting data e￡ the gtrueturai. linguistiese nameZy， deseription of ianguages by anaiyzing and elasstt￡ying ， eaeh ianguage mateyial oniy from aetuai utteranee，. brought an aecurate Xanguage description or a teehnique to do so． And thi．s is the main aentwi・but，ion of iinguistics te the. eontrastive analysis． Next， ieV s eonsider the p， syehologieal eontribution to the. C◎ntrastive an母y8is・。ne。f七he main c。ntributi。n◎f psy⑳。1− ogy is the introduetion of the notion’of stimulus−response to. the eontrastive analysis． This notion wss firstly eontrived by a Russian psyeheiogist， lvan Paviov． He ．came to the eonciu− 8ion that ユ．earning is t；o．form and reinforce the stiP宝ulUSor欝愈幽. s：po捻se connection from a series of exper加ents in which he trained a deg to salivate to the tone of a tuning fork． The. next q．uotation teUs us the details． Drawing on Paviov，s findings， Jehn B． Watson （193S） ceined the term behaviorism． ln the empirieal tra一・ dition o￡ John Locke， Watson eontented that human. behavier should be・studied objeetiveiy， rejeeting ！粧entalis七ic no七i◎ns of inna七enes8 and i捻8七i．nct●. Taking an ttenvironmentalist” positio’n， following Pavlov， he adopted elassicai eonditioning theory as the ．explanat，ion fer all learning’： b， y the process Qf ◎Qnditi◎ning， we buiユ．．t an array of s七imu：Lus齢re帽. 皇麟含dC霧婁n設回忌紹d、灘sC：雰Pき盤塞h含茎’呈ξ§P謙。」1. As we ean see from this quotation， Paviev，s experiment whieh used a dog firstly introdu￠ed the notion of stimulus−response. ’S” ’tt”rown， op． eit・， p． 67e. 一一 5一.

(10) to learning and W’atson named it behaviorism． A few years later，. B． F． Skinner improved the notion and reaehed the eenclusion that as human beings learn any behaviors by the eS￡eet （ren−. spense）， if・they could not get adequate responses to his own behaviors， the behaviors can net be reinferced， and they ean not. learn any behaviors aecordingly． So when we eonsider ttlearning” from the point of view of Skinney， we ean reach the eonelusiion that i：f a 七eacher c◎uid giYe tea◎hing鵬ateriaユ8 which are. pla皿ed．to have an adequa七e re8P◎nse be：eorehand according to a learner， s abiXity， a learner can learn the materiais effee一一七ively．. Seund and satisfactly materials through whieh to develop the necegsity oe Engiish struetural pstterns must yest upon an adeqgate seienti￡ie analysis of English and a careful co瓢pa：r旨i80n of 七he．：re亭uユ七8 0f. this analysis with those of a similar，analysis of the native language ot the students． But these anal− yses． and their eQmparison will be of little practieal aid to ordinary students unless they are built into 蓋翻講6hぎ鶴s吾。蓋hεee蓋謡＆ご㍉七h「。ugh wh’ch the 工n sho：rt， in ：eo：reignユanguage teaching， in orde：【・七◎ 七each a. target language ef￡eetiveiy a teaeher must prepare，teaehing materials frern whieh a Zearner can get sound responses and by whieh a iearner ean peinforee the responses by dividing ian− guage items to be learned． in erder to make an appropriate. teaehing materials， a teaeher must know in which iearning. t Charles C． Fries，. Teachin ＆ Learnin En ユ．ish as a F◎rei n. 舗7．. The Univ． ot Miehigan Press， 1945＞，. ，（Ann Arber ：. 一6一.

(11) po：Ln七＄ a ：Learner has a dif吟ficuユ七y 七〇 lea：rn． And as a！臓eans 七σ. know trouble spets， a teacher has much reliance on the eone. trastive analysis． Another eont−ributien of psyehoiogy is the introduetion ef the notion t’transfer”． Brown expiains ．trans￡er as fQUows． Transfer is a generaX term deseribing the earryovey of Preeious performanee or knewiedge to subsequent iearning． Pesitive transfer eeeurs when the prior knowledge benefits the Iearning task 一 that is， when a previous item is correetly appiied to present subjeet matter． Negative trans￡er oceurs when the previous perfommance etisyupts the performanee on a seeend task． The iatter ean be reffered to as v interferenGe， in that previeusly learned materiai interfeyed with subsequent materia｝一 a previous 醗。とSW鑑。；識き読ぎa器f慧nε琶．igCO「「ect’y aSS。禰. When we appiy the netion ’， transfer” te the foreign language. iearning， it is evident that any lessons have previeus leayning experiences． Narneiy a learney already has acquired his mother language tsefore he starts te leam a foreign language． So aceording to the・ not’ion ”transfert’， iearn’ ?窒堰吹@mother language. infiUenCeS hiS fOreign language leaming．工n O七her wOrdS， a leamer can learn effec七ively七he◎rganiza七ions of the targeポ．1∴：11．／1．．. language whieh are ndentieai or very simiZar to those of his mother Xanguage beeavtse of positive transfer． A learner， however，. will have a difficulty in・ iearning organizations of the tapget language whieh are not similar to those of hi s mother language because of interference． E［2ierefore， in ￡oreign language teaching，. 1 Brown， op． ’ ?奄煤D， pp． 84−p85．. 一7 ．一.

(12) if a teaeher would eorcpare and eontrast the mother language. and the target ianguage accurately， and would grasp，troubie spots correetly， and wouXd give a learneT advices adequateiy， a lea：rne：p couユ．d iea：rn 七he target language m◎re e：C：eectively．. 1七 is obvious that 七he 捻oti◎n 響暑t：【・ansfer脚 supP◎rts the contras鱒. tive anaiysis theoretieally， and that the behavioristie p＄y一一. ehology has influeneed the eontrastive analysis te a great extent． Behaviorism contributed to the notion that human behavior is the sun of its smalXest parts and eom一・. ponents， and thepefere that language learning couid be described as the aequisition of aU of these dis一一 erePe units． Moreover， human learning theeries high− iighted interferimg elements of learning， eoneluding that where ne interferenee eeuid be predicted， no difficulty would be experimented sinee one eouid transfer Positively ali other items in a language．1. As this quotation indieates， the eontrastive analysis is supperted by tt stimulus−regponsptt and ”transferH theories from. behavieristie psyehoZegy． Xlrtthen we eonsider language learning from the viewpoints. of linguistics and psyehology， we can reach the eenelusion that iearning a foreign language i s fundanaentaUy overeoming. the differences between the mother language and the target language． Learning a ￡oreign−language means tha t tho＄e who have a l anguage custom of their raother tongue ac− quires a language system of a ￡oreign ianguage. a Brewn， ibid．， p． 148．. 鱒8 0.

(13) whieh is quite different from that of the ncther tongue． rn o・ther words， ieavning a foreign・ language is nothing but the．．，mutuaX fviction of two different Zanguage eustom＄． S. As the above quotation indicate＄， what lingui sts and language. teaehers intend to by the contrastive analysis is to pursuit an effeetive method to evereeme the di￡ferenees between the rao ther ianguage and the tayget language．. 2． The Historieal Sketeh of the Contrastive Analysis Aceording to the idea d￡ the eentvastive analysis， the most・ impoytant task of a foyeign Zanguage Xearner is to con御 que．r the diffepenees between the raother language and the tar一. get ianguage．．．． the change that has to take plaee in the lan−b. guage behavioT of a foreign language studen＃ ean be equated with the differences between the strueture ef the studenV s native ianguage and culture and 七hat o：e 七he target ：LangU．age a箆d c導蛮1む鷺re．［［1he task. of the iinguist， the cuXtural anthropolegist， and the socieiogist is to identify these differences． The task of the writey o￡ a foreign ianguage teaehing progyam is te develop material＄ whieh wiU be based on a statement o￡ these dif￡erenees： the task of the. tereign language teaeher is，to be aware of these SkgSgr8nees； the task ot the student is to iearn ln the cempari sen of the two languages， however， how ean we. ：Fi s o h shin 七． shi七a ei ok o’uho （工塾皇一「［漏Yanbe， ased ’on. Enalish Teaehin Gendai eigo kyo. 斑e七hod. t4“el．gl，i一．q“tT t 1．．．．A．．ttJS．ua．一．p−ti9．i’」；. uh6…ko Za”2”” F Eigb’一kyb’ 鰍獅??G. ．1964）， p． 97．. 2. Bana七hy，. Tokyo ： Kenkyusha，. Trager， and IAIaddle， op． cit．， p． 57．. 一9麟.

(14) distinguish what are the diffieult points 一to iearn and，what. are七he easy poin誌to iearn？Ru七h Hok：sta七e8 a8 foU◎ws． Working on the theory that learning a language i s acquiring a set of habits and that a student is prone to earry habits from his own language into the new one， it foUows that those．habits to be learned for the new language which most neariy eo− ineide with thoge he uses in his own iauaguage wiii cause the ’ 奄?≠唐煤@dif￡ieuity； and， eonversely， those. that do not ceincide in ons aspect or another wUl be the hardest to iearn． Henee， if the Veaeher ean ．aeeurately predict those areas of ease and diffi− eulty， he ean rco？e efCieiently budget the time and effork of his elass． 1. In shovt， easy le．arning points are the ones that are the same o：【塾8i孤i：Lar 七〇セhe st：puc七ure ◎f his 1ロ〇七her languageg an．d diffi騨. euXt iearning points are the ones that are different from the. structure ef his mother language． Wardhaugh calls this idea as ’ithe strong version” ef the eentrastive analysis． ARd ac一 coyding to this version， we ean determine the degree of dif一・ fiCUユ．七y． as foユ．ユ．OWS●. Stockwell， Bowen， and Martine （1965） who are also preponents of this version of the CAH ・indieate that. the m◎s七difficUlt i七e斑s加甑e targe七language will. be those for whieh there．is ne counterpart in the native language・工f we a8su搬e thei：e vantage pointg the CAH seerns te predict that the greater the dif− ference七he greater七he響撃in七erference”◎f七he na七ive with the target l anguage．，2. Like this， in the styeng versien， the more the ditfevenee. t Ruth Hok， ’tContrast ： An Ef￡ective Teaching Deviee，it gtLg，lli：sl｝mpt reeaLgltlgng， Vei． t 7， No． 3， Apr．，（1963）， p． 120．. 2 」ohn w oUer，」r．， and Seid Me. Ziahosseiny， ’tThe Contras−. Vo］．． 20， Ne． 2，（G970）， p． 184．. 御 10一. ’. Langua”e Learnin tive Analysis Hypothesis and SpeUing Error，”.

(15) between the st？ucture of the rno ther language and that of the. target ；anguage there is， the more diffieuity the learner has． It has been eiear that the strong version has not always been，. applicable in the aetuai teaehing situation． We got the results whieh were opposite te the predictien ef the gtreng version． Let me show an example． Foliewing the strong version，，，in. learning English speliing， a learner whose．native ianguage is not Rornan−seript language has more difftcuity than a learner whese native language is Reman．・seript ianguage sinee・Engligh is a Roman一一seript language．，」ehn W． Oiier and S．eid M． Ziahosseiny’. showed， however， that the assumption was not eerrect in their. experiment． With the strong versien of the CAH we would prediet more speliing errors for students whose native ian− guage was written in a non・一Roman seript， than for students who had experiene e wi th the Roman system． But， the rqsuZts e￡ this study indieate that we wouZd b／ee’ itrong． t. Mke this， the theory of the strong version was denied by the experiment． urhen， what was the eause ef the discrepancy between. the theery and the result ef the experiment？ We ean find three reasons． First， as we see in section t， in order to eayry out. the contrastive anaiysis eorrectly， accurate deseriptiens of two languages and their a￠cur，ate contrast are required． But in actuaよi七y， it is diffiCUユ七七〇eli鵬inate 8ubjectivi七y co燃一. pletely， beeause the eontrastive analysis is carried out by. 1 oUer， and ziahpsseiny， ibid．， pp． t 84−185．. 一 11鱒.

(16) a man himseif．. If one looks specieiealZy at how phonelogieai preb−p iems have been deaZt with in the strong version， he ean easily find evidence to support the assertions just made． Many a linguist has presented eontrastive statements of the phoneraie system of two languages Witheut asking． whether・it ’is possible to eontrast 七he phone鵬ic lsys七ems o：f 七wo ＝Lang犠ages by p：で◎（箆edUre8. whieh atterapt te relate tin lingZish p to a French P，. beeause iinguists have chosen to syrnbolize some not weli・一de￡ined similarity between the two languages in the same way， in this ease by the ietter p， er. 謙昔s：f朧塞i二七宅瀧藍琶。豊言茎ごdtw’th ce翻p m。ve鱒 As there is a tendency that researeher ￡orms a eonelusion only fpom data whieh are in favoum o￡ his researeh conseieusly or. unconseiously， it is not Mkeiy tQ elirninate subjeetivity in earrying out the strong version． Seeondary， even if． the eon．. trastive anaiysis was carried out eornpletely objectivaly and seientifieaily， it has been reeognized that sueh a eemparison does not neeessariiy iead to eortect predietion of trouble spots． ln other words， the eauses ef treuble spots are not. always attributable only to the differenee between the mother language and the tayget language．．．．， he now firmZy di stinguishes between the eom− pe’ 狽?獅??@handled by．a contrastive analysis and the. erpers that oecu？ in a iearnert s performance： t Even. if we eouid obtain accurate and compiete description ef eompetenees，・comparing them wouXd not lead to predietion ef aetual instanees of errort． Errors， he points out， are a matter of performanee： t A sys−. tematie CA of two language competenee， souree and. goal， provides the basis for potentiai error eorn．・. 1 w’ardhaugh， op． eit．，’ 吹D S 25．. 齢 12 一.

(17) mi s sion．t What he is presunably saying here is ’that when theye are ctifferenees ， between the gramer of．. one『S簸a七ive la箆guage andもha七〇f．1anguage o箆e is iearning， one rnay make mi s takes in speaking． But no t aXX mistakes will be caused in this way：， t not every ．error made by the language learner has its or±gin．. in the eentyast 一between， native language and ianguage being learned’．． Dip thus sides with those who believe. eontrastive linguisties wiii help explain some ef. 七h￠．．．’er？6rs．融d6 by． a language learner， ra七her than㌃し. with these who beliqve’ it wiU be the ’basis for pve− dietion of eprors． 1. From this quotation we can understand that the contrastive analysis does not have the ability tu make a eonelusive pre”. dietion of troubie spots， bub has the abiiity only toimply so鵬七rouble spo七s．. Lastly， the third eause is that if the. eontrastive analysis has the ability to predict troubie spots， the theory to interpret the resuLts oe eontrast itselt might be wrong．. For exarnple， it vaay be possible to e＄tablish that two segmental sounds in， the mother tongue and the. target language are very simiiar， and that anQther sound in the’target language has no near e，quivalent in．七he鵬◎七her七〇ngu6．工t do es．no t f◎ユ10w ’bhat the. first new language sound will be easy te teach and the second sound dif￡icuit． The strange is often more readiiy reeognized and ？ealized than the ap−p. 撃霊四七毛9｛織罐t野宮蝿撃hea◎鵬ln。n’s n。七con一. 工n 七h：Ls way，七he deglree of 遺i：eference does not neces8a．rily. refZeet the degree of diffieulty， but there might be the ease. 1 Rebert 」． Di pietro， t’Language structurel in Contaet， it 幽， Voユ． 49， No． 3，（1975）， pp。 7：59−74◎．. 2 Aiexander Baird， ’tContrastive Studies and the Language Teacher， ti 1tllggg，Si h L T hi ， Vol． 21， No． 2， Jan．，（1967）， p． ift．. 葡13一.

(18) that a learning point ean easiiy be understood beeause ef the stru，c七ural difeerenee be七ween 七he 】醗〇七her ユanguage and the tar輪. get language． For these reasons， the strong version was denied， but iも d‡d n◎t 鵬ean the deniεし：L of the cont：【・astive anaユysis i七駒. seif・And as a鵬ah8七〇co噸ensa七e for the defects of the strong version，一 the notion of ’tthe weak version” was devised． The weak. versien is known by the name o￡ eontrastive anaiysis a poster− iori．（The strong versien is known by the narne of eentrastive. analysis a・prieri．） And its fundamental eoncept is as follows． War（ihaugh noted， however， thst， eontrastive， analysis. has intuitive appeal， and that teaehers and linguists have suceessfuUy used ”the best lingUistie knowledge ava21abie ．．． in order te aceount for ebserved dif−. fieulties in seeond language learning．tt He termed sueh ob＄ervational use of contrastive anaiysis the weak versi’on of the contrastive anaiysis hypethesis．［the weak、 version d。es not加Ply七he a priori predic− tion o￡ eertain ￡ine degrees of dif￡ieUlty．・lt ree−. ogni zes the signifieanee of interterenee aeross lan． guages， the ￡aet that． sueh interferenee does not ex一・ ist ’and can expiain diffieulties， but it aiso rec”. ognizes that’ iinguistie−diffieulties can be more profitably expiained a posteriowi一一after the faet． As’the letirnefo i s learn−ing the Zanguage and errors. appear； the teaeher ean utUize his knowledge of the target and．native languages te understand sourees ef 6yrer． 1. Ats・ this quotation shows， the pur’pose oS the weak ver＄ion is to serTu． tinize the eause＄ of errots a learner makes． by using. the knowledge ef the mother isnguage． and the target lariguage．. In the ease of strong version， the eomparison between’the mother ianguage and’ the target language eemes tirst， then the. 1 Brown， op． eit．， p． 457．. 頓14一.

(19) pr’edic㌻ion of iea：rnipg も：rouble spot8 comes．工n the case o：e．the. weak version， however， the factst error＄， which are considered ／i，，，. as troubie spots eeme． fir＄t， and then the eomparison of the 鵬othe：r ユa捻guage a鎗d 七he ta：Pget lang腿age i8 Carried out 七◎make. elear the causes of errer．工t starもs wi七h 七he evidence provided葦y lingui8tic interferenee and uses sueh evidenee to explain the ＄imilarities and diSferences between systems． There shouid be no mistake，，，，：3 abovtt the emphasis on sysVems．. エn七his version systems』． are importan七， becau8e the，e is ne regyession to any pre−systemie view ef languqge，. nor dees the appreaeh resuit in merely classifYing e’rrers in any way that oecurs to the investigato．r． However， the sVarting point in the eontrast is pro− vided by actual evidenee，￡rom sueh phenomena as fauユty trans．ユ．a七ion，．：Xearni】ng difficu1七ie＄， re8idual. foreign aecents， and so on， and reSerenee is made 一to the two systems only in order to qxpZain aetually observed interfprenbe phenomena． 1 Again in the case ef the weaK veysion） facts （errors＞ corne ．．i’iii ’t，，／，，z“’1，. first， and then， their eauses are．expiained by using・linguistie knowledge of・ the mother language．and Vhe target language or. the comparison of the two languages． Theretore a researeher is not ailowed to piek up the language faets whieh are faver・一 able to his researeh． So in the wea・k version， subjgyetivity ean be reMeved． But you can alyeady guesss the weak versien ，，／／／／，r’i’・，i’i’i−t−. gives up prediction ef troubie spots． The weak version of the CAH makes ne predictions whatever as to whieh errors will be made． ln view of 七his i七 see黎1S Cユ．ear七ha七七he weak versi◎n． of the. CAH is not intended as a principle of seeond langU一. 3 wardhaugh， ep． e i t．， pp． 126−127．. 鱒朽一.

(20) age aequisition， sinqe it is not falsi￡iable． Rather， the weak version appears to be a heuristic which ean 職豊蓋罐含a鑑藍量a註彰畿i二XYdrn七e「「o「s’n a 8ec。nd零. Sinee the weak versien is not accompanied with a prediction， it can not resolve the seeond and the third problem in the. strong versien． Namely， the weak version dees not expiain all of the eauses of troubie spots， and does not gi−ve us the answer. whether the degpee of diffieulty corresponds to the degree of. differenee between two languages． Beeause these two probiems eerne from predietion o￡ treuble spots． Wardhaugh ealied thz’s period ”the period o￡ quiesceneett during whieh the．sb’rong ver−. sion was denied and the weak version eame to the front． Langu−. age teaehers and ianguage researchers， however， were not sat− is￡ied with the weak version， whieh oniy expiain the causes ef erro？s． And the zeai of such peopie produeed a new version. of the eontrastive anaXysis． Whe new version is caUed． etthe. moderate version，f． The fundamentai theory’ef mederate veysion i＄ again supported by the rtotion t’transfer’i between the pre一・. vious iearning and subsequent Xearning as the strong version is．. llEhe categ。riza七i。n◎f abs七rg．C七and c。ncre七e p織七七erns. a’eeerding to their pereeived similarities and dif− fe：【・ence8 i＄七he basi8 ：for 鑑6．骸r氣蓬甑91｝・う・：む算β：「毎塗’㊤：診￠茎う備聡参t痢．’．. ever patteyns aye ntnimaiiy distinet in for’m or meaning in ene op rnope systems， eonfusion may result．t，. 1 Fred R．. Eck鵬an，奪警Markedness aぬd 七he Contras七ive Ana：Lysi8. Hypothesis“t. L，LemsyEegs2．：L， i ， VoZ． 27s No． 2， “g77），. pp． 516−317．. 卿16顧.

(21) Conversely， where patterns are ￡unetionally or per一. 躍二藍。器ま盛鷲nき，誌rご警ys纈ρ「8ys七ems。◎”「ec七. Zn the moderate version the’ @basie assunption is that ．however smali the dif￡erenee is， if there is a differenee bettseen the. mother ianguage and the target language， interference takes Plaeei and if theTe i＄ no ．dif￡erenee．between thera， ’positive. tran＄fer takes plaee． But in the moderate verston the assunp−. tion that the degree of difficulty eoincide＄ with the degree of dif￡erencp was given up， and feUowing idea is adopted． This vex。sio箆 o：f 七｝le GAH sugges七s 七hat 七he lea拒ning. ot sounds， sequenees and meanings will be the most di：fficul’b where the 鵬os七 sub七：Le dis’binctions，1；1こ盈蟄e. required either between the target and native langu− age， or within the target language． 2. Thi s way of thinking of the mederate version eornes from the peint of view that． Zearning i s done by a man himsei￡， and. foliewing psychoiogical considepation supperts the version． OUer and Zi’ahesseinyt s moderate version put the eentrastive analy＄is ’hypothesis into some perspec． tive． They rightly emPha．sized the generaiizing’na− tu．re of．hu蹴an lea：pni㎎● 工七 is cemmon to overgener− a↓ize 七〇七he exterガb．七hat mini搬al differences aエ，e. overieoked； at the same time gress difference−be− eause of their salieney−are often more easily. 器「識v匿￡w躍遊里蓋程二二詮濃雲竜dき二二：嶺呈器．5 Thi＄ way of thinking can be the solution to the third problem. a ouer， and Ziahesseiny， op’ D oit．， pp． 185・一t86．. 2 ouer， and Ziahesseiny， ibid．， p． a86， 3 Brovgn，. op．. eit．， p． 159．. 纏t7一.

(22) inh．erent in the strong version that the less degree of differ−b enee i s， the more the degree ef difficulty is． WhaV s more，. the moderate version ean solve not oniy the thi，rd probiem but the ・seeond problem．［Che second problem is that treuble spots. de net neeessarily derive frorn ．the differenees between the mother language and the target language． Of eourse， thi s opin一一. ion is abseluteXy eorrect． Because ￡oreign language learning ineludes many eiements sueh as a language itself， a iearner， Zearning environnent， and se on， and it ean be probable that. some trouble spots come fpom such non−Unguistieai elements． Furthermore even if we focus our eyes en the field， oS trans−. fer between previous learning and subsequent learning， the opinion is important eneugh， for in the strong version only. the mother ianguage is eonsidered as previous learning， namely foeus is only en an interlingual’transfer， but in actuality. previous learning experienee eonsists of not onlY tihe mother language but also previous learning of the target language． Se we must net ignore an intraUnguai transfer． Herein lies a d“erma， a iegiesl ineonsistency in 七he conむras㌻ユV臼．．繰naiysis hypothesis● As 七he studen．t ．begins to learn・ the new language， his linguistic. habit strueture ehanges， and it is thi s altered habit structure which win detβ猟ine the amo㎜七and． loeus of negative transfer in subsequent learning． After any graall inerement of leayning， the student is ne longer the pure native speaker assuned by the eontrastivG anaiysis ot the native and target Jan− guages． All of what he has learned w“1 have facil一” itatien er interferqnce effeets upon what has not yet been 七augh七。皿ie units of 七he鵬毎歌・s．登難ing sca：L⑧，. 韓 18 一.

(23) pesponses to be iearned， have been ehanged． 1. 1n the moderate versien） hewever， this seeond problem whieh the strong version eoulct not eonquinr was got over and the foユ10wing iδea wa8 adop七edの Sueh a peTspeetive underscores the significance of intraiingual errors， which ave as rnueh a faetor in seeond ianguage leayning as interlingUal’errors． The forns within one language−are often perceived to be minimaXly distinct in comparison to the vast di＄tinetions between the native and target language，朧灘宅。畿霊器呈毛iぎ§曾t塾「s．can lead七◎．SO瓢e。f． As this quotatien shows，／iin’1／g・the moderate version， language. yesearehers and language teachers tried to grasp trouble spots eorreetZy by foeusing their attentien on intralingual differ− enees as weil as interiingual differenee． So we． ean’say that. the moderate version is the idea that has deveieped itself on the strong version if we argue frora the point of view o￡. transfer． The versien， however， has the same problem as the strong versipn， that is， the version eouiq not exelude the subjeetivity， for in the raodevate version prediction o， f， trou”. ble spots comes first， and atter that， eonerete erroys are examined． So r．esearehers ean．u．g，le”’／・？roniy data which are in favour. of their own researehes intentionaUy er unintentienaUy． Therefere， the moderate version eould not methodologieally. 1 J． A． Upsher， ttLanguage Profieieney Testing and the tLg；ggagajtmpAgg， VoX． 12，. G◎n七：【・as七ive Ana：Lysis Di：Le㎜a，響響. No． 2，（1962）， p． 126．. 2 Brown， op． eit．， p． 159．. 一19一.