The Language Planning Situation in Taiwan

Feng-fu Tsao Institute of Linguistics, National Tsing-Hua University, Hsinchu, Taiwan, China

Thismonograph presentsa detailedstudy of thelanguage planning situationin Taiwan.After a generalaccount of thesocio-historical context in which theplanning activitieshave taken place, a briefreview of whathappened in termsof languageplanning in MainlandChina under theNationalist government between1911 and 1945is presented.The following sectionprovides acriticalexamination of thelanguage planning activitiesin both languagepolicy and languagecultivation that havehappened in Taiwansince the island was returned to Chinesejurisdiction in 1945.A turning point in theshort history of languageplanning inTaiwanwas reached in 1987,when martial lawthathad beenin existencefor forty years was lifted. Many changeshave taken place sincethen and many moreare in themaking. The finalsection is thereforea careful examinationof someimportant recent developments in languageplanning. In that sectionan optimisticoutlook forthe future is provided and anexplanation for that optimism is given.

Introduction

What is language planning? FollowingFishman (1974: 79), language planning in the present monograph will be broadlydefined as‘ the organisedpursuit ofsolutionsto language problems’. Asimplied by the definition, the scopeof activitiescovered by language planning israther wide, and within language planning scholarship,an impor- tantdistinction is usually maintainedbetween whatis called policy orlanguage determinationissues, and cultivation or language development issues (Neustupný,1970; Jernudd, 1973;Figueroa, 1988). Paulston (1984: 55) makesthis distinctionmost clear when she states,‘ Ifind ituseful todistinguish between languagecultivation and languagepolicy ,where language cultivationdeals with mattersof language andlanguage policy dealswith matters of society and nation’ (Emphasis in original). In reviewing the language planning effortsin Taiwan,I toofind ituseful to maintainsuch a distinction,as will be madeclear in the following discussion. 1 Furthermore,for ease and convenience ofpresentation,language-in-education issueswill be examined separatelyfrom language planning issues,even though itis very clearthat language educationpolicy andimplementation is a very important part of language planning.

The design of the monograph Before Itakeup the issuesin language planning andlanguage education,it will be useful togive ageneral accountof the socio-historicalcontext under whichthe planning activitieshave taken place (see followingsection). There followsa brief examinationof whathappened in termsof language planning in MainlandChina under the Nationalistgovernment 1911– 1945. The next section providesa criticalexamination of the language planning activitiesthat have

0143-4632/99/04 0328-48 $10.00/0 © 1999 Feng-fu Tsao JOURNAL OFMULTILINGUAL AND MULTICULTURALDEVELOPMENT Vol. 20,Nos. 4&5, 1999 328 The Language Planning Situation in Taiwan 329 happened in Taiwansince the islandwas returned toChinese jurisdictionin 1945.The final sectionis a careful review ofsomeimportant recent developments in language planning inTaiwan.In thatsection I takea lookat possible future developments,explaining atthe sametime why Iamrather optimistic in my outlook.

Socio-historical Context Taiwan,which isseparated from the south-easterncoast of the Mainland Chinaby 150kilometres of the TaiwanStrait, is an island with an area of 35,981 squarekilometres and a populationof about21 million.This population consists mainlyof four ethnic groups:the Taiwaneseor Minnanren (Southern Min people), the Mainlanders,the Hakkaand the Austro-Polynesianaborigines. Accordingto Huang’ s (1991:21) estimate, the percentage ofpopulation of each group is as follows: · Minnanren 73.7%, · Mainlanders 13%, · Hakka 12% and · Austro-Polynesians 1.7%. 2 Taiwan’s complexand bitter historical past has left the countrywith this diverse ethno-linguistic heritage.The Austro-Polynesiansare the aboriginesof the island,who, according to the mostup-to-date researchin linguistics(Li, 1979, 1992,1995), anthropology and archaeology (Chang, 1995)arrived on the island fromthe south-easterncoast of the Asiancontinent about 6000 to8000 years ago. 3 Thoseearly settlers, who are now generally believed tobe the oldestknown ancestorsof the Austronesianpeople, were in timedivided intotwo groups accordingto the placeswhere they resided.Those who live in the coastalplains are called Pingpu Zu (the plain tribes) andthose who live inthe mountainareas are called Gaoshan Zu (the mountaintribes). Unfortunately, very littleis known aboutthe movementsof these people either withinor outside of the island. Chinese historicalrecords on the contactbetween the Mainlandand the island arefew andscattered. The earliestrecord of Chinese contactdates as farback to AD230,when, during the period ofthe Three Kingdoms,Emperor Sun Quan triedwithout success to send troopsto conquer the island.A thousandyears later,Emperor Kubla Khan (1260–1295) of the Yuan(Mongol) dynastymade two similarfutile attempts.Unsuccessful asthey were, itwas these earlyefforts that paved the wayfor the gradualincrease in contactsbetween the Mainlandand Taiwan in later years (Chen, 1996). Butbefore the massivepresence ofChinese onthe islandtook place, the Dutch invadedthe southof the islandin 1624and established colonial rule there (1624–1661). A yearlater, the Spanish, notto beoutdoneby the Dutch,invaded the northof the islandand ruled the areauntil they were driven outin 1648by the colonialgovernment in the south.Even though the Dutch treatedthe indigenous aborigineslike slaves,their language policieswere notparticularly oppressive anddiscriminatory. The missionariesthat came with the colonialgovernment even createda writingsystem for Siraya, an aboriginal language serving asa lingua francain the south.The writingsystem, invented atfirstfor missionary 330 Journal of Multilingual and Multicultural Development purposes,was laterused tokeep recordsand to writecontracts. It was in use for more than a hundred years before Chinese characters eventually replaced it. During the Dutch colonialrule, the islandwas still largely inhabited by the Austro-Polynesianaborigines. According to Tsuchida’ s (1983)and Li’ s (1990, 1992)research, the Gaoshan Zu (the mountaintribes) and Pingpu Zu (the plain tribes) caneachbe further divided intonine tribes.The formerconsists of Atayal, Saisiyat,Bunun, Tsou,Rukai, Paiwan, inhabiting fromnorth to souththe central mountainareas, and Amis, Puyuma andYami in the east.In additionto Siraya justmentioned in connectionwith the Dutch colonialrule, the lattergroup comprises,from north to south, Ketagalan, Kavalan, Taokas, Pazch, Papura, Babuza,Hoanya, and Thao. The exactgeographical distribution of these groups of speakers is shown on the map (adapted from Li, 1992) in Figure 1.

Figure 1 The distribution ofAboriginal languages in Taiwan in the 19th century (adopted from Li, 1992) The Language Planning Situation in Taiwan 331

Havingsuccessfully driven outthe Dutch in 1662,Zheng Cheng-kong (better knownin the westas Koxinga),his Ming royalistsand his family ruled the island for21 years(1662– 1683). As Zheng himself wasfrom Southern Fujian, andhis followerswere mostlyfrom the sameregion, they spoke the Southern Min dialect.Zheng’ s rule wasreplaced by the Manchusof the Qing dynastyand abouttwo hundred yearsof Qing rule ensued (1683–1895). In the lateMing period andthe earlyyears of Qing rule, the coastalprovinces of Fujian and Guandongwere suffering fromeconomic hardship and political turmoil; many inhabitantsin the areawere forcedto leave their hometowns in searchof abetter life in south-eastAsia or Taiwan.Those who went to Taiwan were mostlyfrom the Quanzhouand Zhangzhou districts of Fujian province,speaking the Zhangzhou or Quanzhou variety of the Southern Min. The Hakkamostly from Kuangdong province,speaking either the Hai-lu or Si-hsien varietyof Hakkaaccording to their place oforigin, soon joined thisnew waveof immigration.When these different groupsof immigrants came to Taiwan,they tended tooccupy areason the islandsimilar to their homeregions in the Mainland;Quanzhou people, being shopand factory owners or workers, settledmostly along the coastalareas and ports. Zhangzhou people settled mostlyin the inlandplains and were devotedto agriculture.The Hakka,who were skilled in farmingin hilly areas,settled in tablelands and foothill regions (Shi, 1987: 1–6). The comingof all these different groupsof people led toanumber offierce struggles.The Hansettlers with their largernumbers, better farmingand irriga- tionskills and often withthe implicitsupport of the Qing governmentsoon outmanoeuvredthe Plain tribesof aborigines,who, as aresult,lost their landand were fastassimilated. The Mountaintribes, being separatedby high mountains and deep valleys, were kept out of harm’s way, at least for the time being. In 1895,a yearafter Taiwan was ceded toJapanby the Qing governmentas a resultof losing the firstSino-Japanese war,the Haninhabitants on the island alreadyoutnumbered the aboriginalpeople by aclearmajority, and among the Han,Southern Min speakersaccount for 82%, Hakka speakers for 16% and the restfor only 2%(Lamley, 1981:291– 293). By 1905,the yearwhen the firstcensus wastaken, ethnic Chinese hademerged asthe majoritygroup by asizable margin(2,970,000 Chinese vs.113,000 aborigines) andthey haveremained so ever since. The Japanese rulersmade it clear at the very beginning oftheir occupationthat they intended tointegrate Taiwan fully intothe Japanese Empire. Tothisend, policies ofcomplete Japanisationwere designed. The Japanese rulers,however, were pragmaticenough torealisethat such an end couldnot be achieved in a shorttime. They therefore implemented these policiesin three stagesof educa- tionalplanning which were increasingly assimilatoryin nature.During the first stage(1895– 1919), which is generally referred toas the stageof pacification, privateChinese schoolscalled Shu-fang (book house), where ethnic Chinese sent their children tostudyClassical Chinese withSouthern Min orHakkapronunci- ation,were tolerated,while atthe sametime the Japanese governmenturged people in Taiwanto send their children tothe public elementary school,where Chinese was taught as a required subject. 332 Journal of Multilingual and Multicultural Development

During the secondstage – the stageof assimilation (1919– 1937) – allprivate Chinese schoolswere banned andChinese asa subject wasmade elective. During the final stage– the stageof complete Japanisation(1937– 1945) –notonly wasChinese banned in allpublic domains,but even the few Chinese pages ofthe basicallyJapanese newspaper Xinminbao,the largestin Taiwanat thattime, and allother publications in Chinese were banned. The ethnic Chinese living in Taiwanwere thuscompletely cutoff fromtheir Chinese culturalprint tradition. Earlier,in 1938,a yearwhich saw the beginning ofthe secondSino-Japanese war, the Japanese governmentin Taiwan,in orderto further obliterateChinese influence, launched afierce ‘only-Japanese-speaking-families’campaign whose purpose wasto drive the indigenous languagesout of the family domain,usually believed tobe the beststronghold for language maintenance(Chen, 1996;Tsao, 1997a).With all these repressive anddiscriminatory measures in force,ethnic Chinese andnative Austro-Polynesians were in factsecond-class citizens and their languages could not but be seriously damaged. When Taiwanwas returned toChinaat the conclusionof the Second World War,the people in Taiwanwere overjoyed,firmly believing thattheir socialand politicalstatus would be greatlyimproved. However, as the ironyof life would haveit, their high expectationshave never been realised,as weshallsee in some detail in the following discussion. In 1949,four yearsafter Taiwan was returned toChina, the Nationalist governmentlost Mainland China to the Chinese Communistsand was forced to retreatto Taiwan. Those immigrants and their children, nowknown as ‘ the Mainlanders’, constitutethe secondlargest ethnic group inTaiwan,even though atthe timeof their arrivalthey didnot form a homogeneousgroup asthey origi- natedfrom different partsof Mainland China, speaking avarietyof Handialects orminoritylanguages. Most of the Mainlanderswere atleastreceptive bilinguals when they arrivedin Taiwan.They hadacquired Mandarin either through educationalchannels or during armyservice. The better educatedamong them couldcertainly write Mandarin and probably alsoclassical Chinese. Upon their arrivalMandarin was a naturallingua francafor inter-dialectal communication forthe group. Mostof their children alsoacquired it either asafirstor asecond language. Today,the majorityof the Mainlandersare located in large cities,especially in Taipei, the capital city. Onthe economicfront, things turned outmuch better.In the earlyfifties, aboutten yearsafter the Nationalistgovernment took over Taiwan, a very successfulland reform programme was launched. Thisreform programme not only directlypaved the wayfor agricultural development but alsoindirectly paved the wayfor later industrial development (Huang, 1998).This dynamic transformationstarted in the 1960s.The industrialisationof Taiwanesesociety tookplace between 1961and 1980.Inthe totalworkforce, the proportionof agri- culturalworkers decreased dramatically from 56% in 1953to 19% in 1983,while the proportionof the industrialworkers increased significantly from 18% in 1953 to41% in 1983(Wen, 1985).In the meantime,the per capitagross national productof the countryrose from US$203 in 1950to US$2344 in 1980and to US$12,439 in 1995. 4 Withthe radicaleconomic growth, education was no longer aluxury forthe people ofTaiwan.In 1950,there were 139.64students for every 1000people; in The Language Planning Situation in Taiwan 333

1982,there were 255.18students. In 1950,only 31.99%of primaryschool children continued their middle schooleducation; in 1982,the percentage thatcontinued educationincreased to 98.6%. As literacy became widespreadand opportunities toreceive higher educationincreased (Tsao, 1998), the indigenous residents(the Southern Min, the Hakkaand the Austro-Polynesianpeople) were morelikely to obtainbetter occupationsand higher incomes,which in turnupgraded their socialstatus. Today, with greater socio-economic and political resources, these less powerful groups are better equipped to compete with the Mainlanders. The economicprosperity, however, has not entirely been ablessing. In the courseof development, the aboriginalpeople, whoused tobe protectedby high mountainsand deep valleys,have become moreand more exposed toHan people. Furthermore,well-paved roadsnow lead right into their territoryand televisionnetworks bombard them with Han language andculture. The inevi- tableresult has been arapiddecrease in speakersof the aboriginallanguages. In fact,of alleleven tribesamong whom we canfind speakers,seven havefewer than 10,000 speakers and are in serious danger of extinction. In summary, Taiwan, as it stands today can be characterised as: · amulti-ethnic andmultilingual societywith four majorethnic groups;the Mainlanders,the Southern Min people, the Hakkaand the Austro-Polynesians; · an immigrant society, the latest group being the Mainlanders; · a Chinese society, and above all, · a modern industrialised society.

A Brief Account of the Language Planning Efforts in China from 1911 to 1945 Asmentioned earlier,at the conclusionof the Second WorldWar in 1945, Taiwanwas returned toChina.By thattime a number oflanguage policieshad alreadybeen formedand implemented in MainlandChina and so when the Nationalistgovernment took over Taiwan, the lawsand regulations pertaining tolanguage andlanguage educationwere simply takenover from China with very slightadaptations. Therefore, in orderto fully understandthe language planning situationin Taiwanafter 1945, itisnecessaryto begin onthe Mainland.

A brief history of the national language movement When the Republic ofChinawas established in 1911,it faced two pressing problems:unification and modernisation. Ethnolinguistically, the countrywas composedof morethan 50 ethnic groups,each speaking one ormorelanguages, representing the Sino-Tibetan, Austronesian,Altaic, and Indo-European linguistic stocks.Even though the Hangroup wasby farthe largest,accounting formore than 90% of the population,it actually consists of seven majordialects, dozensof mutuallyunintelligible formsof speech andhundreds ofsub-dialects. Yuan(1960: 22) liststhe majordialects and population percentages as:Mandarin 70%, Wu 8.4%, Xiang 5%, Cantonese 5%, Min 4.2%, Hakka 4% and Gan 2.4%. Itis obvious that in acountrywith so much ethnolinguisticdiversity and complexitylinguistic unity hasbeen aseriousproblem. Actuallytowards the end ofthe Qing dynastyas the Westernpowers encroached ever moreupon China’s 334 Journal of Multilingual and Multicultural Development territory,leaders of the countryrealised that for China to become astrong country, it must have a unified national language and mass literacy. The lackof asingle language foruse amongall Chinese hadlong been takento be aseriousimpediment tonational unification and political, economic and socialdevelopment. Atthe nationallevel, itis reported that in the earlydays of the Qing dynastythe emperor hada hardtime communicating with officials fromthe southernprovinces, especially Fujian andGuandong. The problem actuallybecame soseriousthat the governmenthad to set up aspecialschool to teach them Mandarin. Anotherserious problem wasmassilliteracy. Statistics on illiteracyin Chinain the earlyyears of the Republic havenever been morethan rough approxima- tions,but even aslateas the middle andlate 1950s estimates from a number of sourcessuggest that from one half totwothirds of the adultpopulation were functionalilliterates. The figure couldonly havebeen higher twentyor thirty years earlier. Butto tacklethese twoseriousand urgent problems,two questionsinvolving language planning needed tobe answered:(1) which dialectshould be chosenas the nationallanguage? and(2) howshould it be writtenso that the masscould learn it in the shortest possible amount of time? Realisingthe urgency of the problems,the new republican government workedon them immediately. On 10July, 1912a meeting onnationaleducation washeldatthe Ministryof Education (henceforth MOE)in Peking. Animportant resolutionpassed at the meeting wasto organise the Committeefor the Unifica- tion of Pronunciation (henceforth CUP) whose functions were: · toexamine andauthorise the pronunciationof all the wordsin the national language (NL); · toanalysethe phonemes ofthe NLanddecide onthe number ofphonemes; · toadopt phonetic alphabets:one alphabeticsymbol for each phoneme (Fang, 1965: 131). Establishedofficially on15 February, 1913as a subordinatecommittee of the MOE,the CUPhad45 members, representing different provincesand special districtsof the country.At its first meeting, the importantquestion of which dialectshould be chosenas the nationallanguage wasdiscussed.There were two serious contenders, Mandarin and Cantonese. Consideringeverything, Mandarinshould have been chosenas the national language as it has the following advantages over all other dialects: (1) Mandarin,as previously mentioned,is by farthe largestdialect group, its speakersaccounting for 70% of the totalpopulation, and furthermore, its four majorsub-dialects, namely, Northern, Northwestern, Southwestern, and Lower Yangzi, are said to be mutually intelligible (Chao, 1943: 61). (2) Geographically,Mandarin speakers occupy avery broadterritorial belt running allthe wayfrom the northernmostreaches of Manchuria to the borders of Yunnan and Sichuan in the south-west. (3) Peking hasbeen the nationalculture centre forabout a thousandyears and much ofthe vernacularliterature written in thislong period wasin Northern Mandarin. The Language Planning Situation in Taiwan 335

(4) The final advantageis that Peking hasbeen the capitalof Chinafrom Liao timesthrough the Jin, Yuan,Ming andQing dynastiesup totodaywith only brief interruptions.Since Peking wasthe seatof government,officials from allparts of Chinahad always had to learn a formof Mandarin(called guanhua,‘officialspeech’ ) in orderto be able toconductgovernment busi- ness. This guanhua, or lanqing guanhua ‘blue-green Mandarin’, asit was sometimescalled because of tracesof all kinds of different dialectback- groundsin the speech ofitsspeakers, was regardedby manyChinese asthe lingua franca of all China. However,despite allof these advantages,in the meeting where allprovinces andspecial districts were equally represented acompromisesolution was more likely tobe adopted.The nationallanguage finally chosenwas actually an artifi- cialform of Mandarin containing the maximumdistinctions found in the major dialects:i.e. the entering tone,the jian-tuan distinctions(dental andvelar initials occurringbefore ahigh frontvowel respectively, FT), twomid-vowelphonemes /o/and/ e/ (where mostnorthern dialects have only one). The choiceseemed to be ahappy one asit pleased mostof the partiesconcerned, and that partially accountsfor the factthat in the processof selection (even though there were heateddebates interspersed withskirmishes involving flying teacupsand saucers),when the NLwasthusdecided, there were noseriousriots of the order ofthose in Indiaand Sri Lanka,when Hindi andSinhalese were selected astheir nationallanguage. Anothercontributing factor to the relativesmoothness of selectionis that, as mentioned earlier,there wasa commonconcern among the elites atthattime about the necessityof establishinga nationallanguage tofacili- tate inter-lingual and inter-dialectal communication. Thiscompromise solution, while itseemed tobe able tosatisfy most of the representatives,actually contained an insurmountable difficulty forpropaga- tion.Since itwas anartificiallanguage, there wasnota single teacherwho could claimto speak itnatively. Teachers who had to teachit were soondivided into twocamps according to the strategythey adopted.Those who spoke aNorthern dialectclose to the Peking dialectused the Peking dialectas their basis,with an approximationof the entering toneused in reading pronunciationwhen they readClassical Chinese. Thosewho spoke aSouthern dialectused the entering tonethey hadin their nativedialects, which in actualpronunciation varies from dialectto dialect, and they hadto approximate the otherfour tonesfound in Peking phonology.The twocamps fought aboutwhat standard pronunciation wasfor more than a decadeand this controversy has come to be knownas ‘Jing Gouzhi zhen’ (controversy over Peking pronunciationand national language pronunciation) (Chao, 1976b; Tsao, 1987). Thisserious drawback of trying touse anartificialnational language, coupled withthe factthat a majorityof the Chinese people alreadyspoke sometype of Mandarin,finally tipped the scalein favourof Peking Mandarinas the NLof China.5 In 1932,without publicly announcing anyradical changes, the PronouncingDictionary of the National Language ,which wasauthorised by the MOEin 1919on the basisof the recommendationof CUP, wasquietly revised in the formof the NationalPronunciation ofCommonVocabulary andwas authorised by the MOEasthe dictionaryof standardpronunciation. It included 9920words 336 Journal of Multilingual and Multicultural Development and2299 synonymsand was based exclusively onthe educatedspeech ofPeking (Chao, 1976b). The selection of the writing system Letus nowreturn tothe questionof what writing system should be selected forthe nationallanguage. Asearlyas the lateQing dynastythere were heated debatesas towhether the Chinese characters,as they were atthattime, were a suitablewriting system for Chinese. The mostserious drawback of the systemis thatit is too complex to be learned by agreatnumber ofpeople withina short time.Part of the reasonfor this difficulty isthat a greatmany Chinese characters, chiefly due tohistoricalsound changes, are at thisstage of development devoid of anyassociation with their pronunciation.This lack of associationbetween soundand meaning certainlymakes the learning taskmuch moredifficult as learnersoften haveto resortto rotememory. To put itdifferently, the writing systemis perceived asa greatimpediment tothe promotionof massliteracy, which wasregarded asimperativefor the modernisationof China. On the other hand,it has been the medium withwhich Chinese culture andChinese literature havebeen recordedand any drastic change in the writingsystem entails the dangerof disconnectingthe present fromthe culturalpast. Various proposals were madewhich include keeping the characterwriting system intact but supplementing itwith an auxiliarytranscribing system to indicate the pronunciation,replacing itwith simplified characters,and finally, using aromanised spelling system. Atthe meeting onnational education held atthe MOEin1912(mentioned previously), animportantdecision was taken that Chinese characterswere tobe kept intactbut anauxiliarysystem of phonetic alphabetswas to be adoptedin education. It was the CUP’s duty to devise such a system. In the yearimmediately after the CUPwasconvened, itwas decided thatthe traditionaltranscribing alphabet rather than a Latinalphabet should be adopted asthe officialphonetic (transcribing)device supplementing the characters.This wasessentially a spelling systemthat fell between the Latinalphabet and the Japanese syllablaryin function,but waslike simplified Chinese charactersin shape. After the closeof the firstmeeting, the workof the committeewas suspended because ofpoliticalturmoil. However, in the privatesector, the workcontinued. In the few yearsthat followed, some ‘ transcribingalphabet classes’ were established in Peking andin 1916a periodicalnamed the ‘TranscribedMandarin Paper’was published in whicharticles were printed in characterswith the transcribing alphabetsbeside them(henceforth tobe called‘ transcribedcharacters’ for short). On23 November, 1916 the transcribingalphabets were authorisedby the MOE.The systemconsisted of symbolsfor 24 consonants,3 glides, 12vowelsand 4tones.From a purely linguisticpoint of view,these symbols,with the exception ofthe tonesymbols, are not completely phonemic symbols.Some ofthemrepre- sentsounds (allophones) ratherthan phonemes. The tonesymbols, however, represent phonemic tones,and morphophonemic sandhi rules arestated sepa- rately. The Language Planning Situation in Taiwan 337

Immediatelyafter the Committeefor the Preparationof aUnified National Language (CPUNL) wasfounded in April 1919,its members setto work to improvethe transcribingalphabets. The revised systemformed the basisof the NationalPhonetic Symbols (henceforth NPS), which were officially promulgatedby the MOEin 1930.These symbolswere toplay avery instrumentalrole in the propagation of the national language in Taiwan. In 1928,the MOEonthe recommendationof the CPUNL authoriseda romanisationsystem for transcription, chiefly developed by Chao,Y.R. andLin, Yu-tang,and promulgated the regulationsfor the romanisationof the national language. Fromthen onuntil itsrevision in Taiwanin 1984,this romanised phonetic transcribingsystem, whose chief feature wasthe representationof the tonesin spelling ratherthan by diacriticmarks, was known as the secondform of NPS (NPS2, for short). The change from ‘Chinese Literature’ to ‘Chinese Language’ as a subject in the elementary school Anothermajor contribution of CPUNL wasthe change from‘ Chinese Litera- ture’to ‘ Chinese Language’as asubject in the elementary school.This change mayappear to people unfamiliarwith the traditionalway of teachingChinese as atrivialchange ofname,but in realityit involveda change ofgreatconsequence. TraditionalChinese literaturehad always been writtenin aspecialliterary style closerto Classical Chinese thanto the everyday spoken language. The transcribing alphabets,devised asanaid to the rapidspread of literacy and common education,were actuallybased on the spoken language. So in orderfor the transcribing alphabetsto have the greatesteffect andfor the textbooks,which before the change were uniformly writtenin the literarystyle, to be readilyreadable for elementary school students, the content of the language course had to be changed andthe textbooksre-written. Therefore, atthe firstmeeting ofthe CPUNL,it was recommended thatthe textbookson Chinese forelementary schoolsbe revised sothatall the lessonswere writtenin the colloquialspoken style.In 1920,the required subject ‘Chinese Literature’for the firsttwo grades in the elementary schoolwas changed to‘Chinese Language’by the MOE,andin timethis change wasextended toallsix grades in the elementary school.In this way,the teachingof the literarystyle in elementary educationwent into history. 6

Language planning activities in connection with language development While itseems that much wasgoing onin the areaof language policy and policy implementationduring thisperiod, despite manyinterruptions due to politicalturmoil and the Sino-Japanese War,nothing much seemed tobe happening in the areaof language development. One noticeableexception was the workof the Institutefor Compilation and Translation in the compilationof listsof vocabularyequivalents in scientificand technical fields. The Institutewas founded asabranchoffice ofMOE in 1932,and between itsinception andthe Nationalistgovernment’ s retreatto Taiwanin 1949,it compiled andpublished 25 volumesof word lists covering a varietyof modernscientific disciplines suchas mathematics,physics, chemistry, various specialties in medicine, several branchesof engineering aswell asfive areasin socialscience (including 338 Journal of Multilingual and Multicultural Development economics,psychology and education). However, in spite ofsucha respectable showingin the standardisationin scientificterminology, language development was undeniably a peripheral concern at best during this period.

Summary: Special features of the language planning activities in the period Acareful examinationof the language planning activitiesin thisperiod reveals the following important features. Firstly,the processof the selectionof the nationallanguage wasin general quite smooth.To be sure,there were heatedexchanges of wordsand blows in the meetings, but oncethe nationallanguage wasdecided upon, there were no seriousriots. Two important reasons account for this relative smoothness. First, there wasa general consensusamong the opinion leadersin the lateQing dynastyand the early yearsof the Republic that,in orderfor China to become a modernised,strong country, a unified nationallanguage wasan essential. Second, acompromiseselection was made in the sense thateven though itwas basedon the NorthernMandarin, the nationallanguage alsoincluded some otherfeatures such as the entering tonefound in othermajor dialects and this artificial version of Mandarin was able to satisfy all the parties concerned. Secondly, even though the selectionprocess was, comparatively speaking, rathersmooth, the nationallanguage selected wasnot completely satisfactory. Being anartificiallanguage basedon the NorthernMandarin, the normin some casesexisted only onpaper, there being nonative speakers to exemplify the exact pronunciation.This lack of ‘live norm’in the nationallanguage, soto speak, actu- allypresented aseriouschallenge toitslater propagation. The situationwas not correcteduntil 1932,almosttwo decades after the normwas first conceived. This aspectoffers agoodlesson for those planners whowish to adopt a compro- mise-made language as a national language. Anotherfeature ofthe planning processis its almost exclusive concernwith the pronunciationof the nationallanguage. The firstofficial committee set up for planning purposes wascalled the Committeefor the Unificationof Pronuncia- tionand it set the tonefor the laterprocesses. This exclusive concernwith pronunciationwas probably due toa misconceptionamong the general public thatthe differences between different dialectslay mostly in the areaof phonology.This misconception, in turn,was probably induced by the factthatin the Qing dynastyliterate people in Chinawere able tocommunicate through writingin Chinese characterseven though they readthe charactersdifferently in their owndialect. Whatever the reason,this concentration on phonology in the early stagesof the planning wasprobably justified, but the persistentemphasis ononly the phonologicalaspect of the nationallanguage couldjustifiably be seen as being too restrictive in scope. The fourthfeature isthat,as far as language planning activitiesare concerned, much attentionwas paid to language policy mattersand very littleto language development issues.This bias was probably due in partto the factthat there was nostandingcommittee whose function wasto guide allactivities pertaining to language planning. Every committeewas ad hoc in natureand once the assigned mission was deemed accomplished, it was dissolved. The Language Planning Situation in Taiwan 339

Finally, nosystematic evaluative measures are to be found in anypart of the planning andpropagation process in the period.The lackof an evaluation componentis a feature commonto many language-planning projects throughoutthe worldas Rubin (1971)haspointed out.Since thisfeature alsohas persistedin language planning endeavoursundertaken subsequently in Taiwan, Iwill examine itsconsequence moreclosely in alatersection of the monograph.

Language Planning Activities in Taiwan since 1945 Formore than five decadesthat the Nationalistgovernment has ruled the islandof Taiwan,the mostimportant language policy hasbeen the propagation ofMandarin, the nationallanguage. Iwill therefore begin ourdiscussion in this sectionwith an accountof the so-calledNational Language Movement(NLM) to be followedby abrief evaluationof the movement.However, as madeclear in the previous section,since Taiwanis a multi-ethnic andmultilingual society,the propagationof the nationallanguage hasinevitably affected otherlanguages spoken onthe sameisland. At the sametime,as much ofthe country’s resources havebeen pumped intothe propagationof the nationallanguage inthe educationalsystem, it has also had serious effects onthe otherlanguages, chiefly English,which arebeing taughtin the schoolsystem. I will therefore gointo the policy mattersrelated to these languagesin the secondpart of thissection. Part 3 ofthis section focuses on language development issues,which havebeen gaining in importanceas Taiwanhas become modernised.The final partof thissection sumsup the previousdiscussion by pointing outthe specialfeatures of the language planning activities in this period.

National language movement in Taiwan National language movement: A historical account Atthe closeof World War II in 1945,the Japanese governmentsurrendered unconditionallyto the Republic ofChina(ROC) andTaiwan was returned tothe rule ofthe Chinese government.In the sameyear the TaiwanProvisional Provin- cialGovernment was setup withChen YiasitsAdministrative Head. Although the Nationalistgovernment was not unprepared forthe recoveryof Taiwan(as preparatorywork had started in 1944),when the end tothe warbegan toseem inevitable (Kubler, 1985),the appointmentof Chen asthe administrativehead wasa hasty,ill-considered decision.Chen, aNationalistgeneral, whoonce ruled Fujian Province,turned outto be ratherill-prepared forthe worklying before him torule anisland inhabited by millionsof Southern Min, Hakka,and Austro-Polynesianspeakers, most of whomhad received someJapanese educationand some of whomspoke fluent Japanese asahigh language. Hislack of preparationis clearly revealed in aninterview witha Da-GongNewspaper jour- nalistbefore he tookup hisnew post.He boastedin the interview thatwith his experience in the propagationof the nationallanguage in Fujian Province,he shouldbe able tomakegreat headway in four years.He alsostrongly advocated that strict measures should be taken in promoting the national language. Very littleis known about what Chen wasable toachieve in Fujian Province, but judging fromits present-day much poorershowing in the propagationof Mandarin(called Putonghua (PTH) in MainlandChina) when comparedwith 340 Journal of Multilingual and Multicultural Development thatof Taiwan(Zhou, 1992; Tsao,1997a), Chen’ s statementcannot be takenvery seriously.It was also foolhardy of him toadvocate strict promotional measures because the sociolinguisticsituation of the islandat that time can be roughly characterisedas adiglossiawithout societal bilingualism (Fishman,1967; Tsao, in press).In otherwords, in the Taiwanesesociety, there existeda ruling classof Mainlanders,most of whomcould speak someform of Mandarinand a lower classof people comprisingSouthern Min, Hakkaand Austro-Polynesian speakers,and there wasnoway for these groupsto communicate with each other except throughtranslation. The situationwas extremely delicateand needed to be handled withcare. In thisconnection, one isreminded ofthe well-thought-out andvery cautiouslanguage policy ofappeasement used by the Japanese when they firstarrived on the island(see above).In sharpcontrast to the Japanese, Chen advocatedstrict measures. Indeed in 1946,less than a yearafter he tookup hispost, he banned the use ofJapanese completely in orderto eradicate the Japa- nese influence in Taiwan.While the groundsfor doing somight have been justifi- able, the timingwas unfortunateand the consequences were hardlywhat he had expected. Aspreviously indicated,Japanese wasthe high language thatmany elites used in the public domain.This being the case,banning the Japanese language waslike shuttingtheir mouths,or taking away their voices.Little wonderthat thousands of intellectualsstrongly protested at thisarbitrary act of the government(Hsu, 1991). Ill-considered actslike thiscoupled withthe reports thata number ofpeople fromthe indigenous groupswere either deposed or demotedbecause oftheir poorproficiency in Mandarinsoon turned the indige- nousgroups of people againstthe governmentand the Mandarin-speaking Mainlanders.This anti-government sentiment, enhanced by manyreports of governmentinefficiency andcorruption, reached its peak when the tragic February 28incident broke outin whichthousands of Taiwanese and Main- landerswere killed andthe relationshipbetween the indigenous groupsand the Mainlanderswas greatly traumatised. 7 Soonafter the tragicincident, Chen was deposed andwas eventually executed onthe groundsof conspiring with the Communist Chinese against the government. Onthe nationallanguage promotionside, things were farmore fortunate. The NationalCommittee and Fujian Chapterof the Committeefor the Promotionand Propagationof the NationalLanguage tookactive parts in the deliberationon the reconstructionof Taiwanas earlyas 1944(Kubler, 1985).In November 1945,soon afterthe Japanese surrendered on9 September ofthat year, several dozen members ofthe MainlandCommittee of CPPNL, led by Wei Jiangong, aphilolo- gist,and He Rong,a grammarian,arrived in Taiwanto setup the machineryfor the promotionof the nationallanguage there. Becausethere were stillnot suffi- cient teachersand promoters to staff the variousMandarin centres, some thirty primaryschool teachers with high proficiency in Mandarinfrom Fujian Province were brought in inthe spring of1946. Later,several dozens advanced university studentsin Mandarintraining classes at variousMainland universities were also recruited for the same reason (Fang, 1965). The TaiwanProvincial CPPNL wasestablished in April 1946subordinate to the EducationalDepartment of the ProvisionalProvincial Government. It included amongits charter members, Wei Jian-gong, He Rong,Fang Shiduo, Li Jiannan,Wang Yuchuan, Lin Shaoxian,Zhu Zhaoxiang,and Wu Shouli, several The Language Planning Situation in Taiwan 341 ofwhom were toplay importantroles in the promotionof the nationallanguage in Taiwanfor years to come. However, in additionto the mainoffice ofthe Committeelocated in Taipei, otherbranch offices calledMandarin Promotion Centerswere opened in Taichung,Taitung, Hsinchu, Kaohsiung, Changhua, Chiayi,and Pingtung. They were staffedby the recruitsfrom Mainland China andthey operatedin closecooperation with the localschool systems and city governments(Fang, 1965:133). The Committeeset to workat once; the firstfew urgent tasks that called for immediate attention were: · toset up the standards(chiefly in pronunciation)for the nationallanguage, · to devise the working outline of the NLM in Taiwan, and · tocompile the StandardPronunciation Dictionary of the nationallanguage. In the earlydays after Taiwan’ s restorationto China,enthusiasm for learning the nationallanguage wasextremely high, but thishigh enthusiasm,instead of being fully utilised,was soondampened by the badadministration of the Provi- sionalProvincial Government headed by Chen Yionthe one handand by the lackof qualified teachers,on the other.As can be imagined,teachers were from very different backgrounds,ranging fromthose who were nativespeakers of the nationallanguage tothose who had had very littleexposure tothe national language andcould only speak itwith a very heavy accent.Standard textbooks were alsounavailable. The members ofthe TaiwanCPPNL thusbegan aseriesof effortsto explain tothe public throughthe massmedia (mainly radioand newspaper) the meaning andthe purpose ofNLM,and the definition ofthe national language (Fang, 1965). Anotherachievement ofthe Committeewas to designatean outline forNLM in Taiwan. The following six principles were decided on: (1) torecover the Taiwanesedialect so as to enable the public tolearn the nationallanguage by comparisonbetween the dialectand the national language; (2) to emphasise the standard pronunciation; (3) toeradicatethe influence ofJapanese asreflected in the dailyspeech ofthe people; (4) topromotethe contrastivestudy of morphologyso astoenrich the national language; (5) toadapt the NPSsoas to promote communication among people of different races and origins; and (6) toencourage the intentionof learning the nationallanguage soas tofacili- tate the teaching of it (Fang, 1965: 131). Ofthe sixprinciples, 2,3and6alllook practical and practicable. In fact,I have alreadymentioned someof the workdone inaccordancewith the sixthprinciple. Principle 5wasa very wisedecision, and NPS were toplay avitalpart in the propagationof the nationallanguage. Ihavemore to say aboutthis in connection withthe MandarinDaily Newsin alatersection. Unfortunately, as Taiwanese societyis becoming internationalised,this system is now facing astiff challenge fromcompeting romanisedsystems, especially the one propagatedby the People’s Republic ofChina.It is still too early to say what the outcomeof this competition will be. 342 Journal of Multilingual and Multicultural Development

Principles 1and4 were either controversialor impractical. When Principle 1 wasannounced, it immediately spawned a heateddebate. There were people whoargued thatdialects should be done awaywith once and for all, but there were otherpeople whoargued thatthe nationallanguage canbe bestlearned throughpeople’ s mothertongue, namely the dialect.It wasnot clear whether there wasa consensusamong the committeemembers andif sowhat that consensuswas. Judging fromthe factthat there wasonly one seriesof textbooks calledthe ‘Bridge Series’(which utilised the comparisonmethod to teach the Southern Min speakersthe nationallanguage), and,from the factthat no members ofthe Committeewere knownto takepart in the debate,it seems fair to saythatthe Committeewas notreally tooenthusiastic about this principle. Asfor Principle 4,itlooks more like autopianblueprint, asthere were noexperts on Taiwanat that time who were able toconducta contrastivestudy of Mandarin andSouthern Min, notto mention Hakka and the aboriginalAustro-Polynesian languages. Yetanother major effort ofthe Committeewas the compilationof the Dictio- naryof the StandardPronunciation of the nationallanguage. The firstTaiwan editionwas published in 1952.In the ensuing years,it became sopopular that almostevery teacherhad a copyof it.It certainly played avery importantrole in the standardisation of the national language in Taiwan. The TaiwanProvincial CPPNL wasestablished in April 1946andabolished in 1959.A lower-level committeein the ProvincialDepartment of Educationwas founded toreplace it.Three reasonswere given forthe abolitionof the Committee: (1) the policy objective ofNLM (i.e. standardisationand propagation) was deemed to have been achieved; (2) the cultivationof the NL,along-term enterprise, couldbe continued throughthe jointefforts of the school,the media,and the whole nation;and (3) alower-level committeewas thought to besufficient toguide the development of the national language. Asall three ofthe reasonsgiven were highly questionable,the true reasonsfor itsabolition remain a mystery.Kubler (1985)has argued thatbudget consider- ationsmust have played animportantpart as itwas not too long afterthe battleof Quemoy withthe Chinese Communists,and the Nationalistgovernment could havebeen contemplatingan expensive, large-scalemilitary offensive. (Iamof thisopinion.) In anycase, the feeling thatmuch in termsof language planning remained tobe done andthat a lower-level committeewas insufficient for guiding the operationsoon became widespread.In 1980the governmentwas pressured intore-establishing under the MOEabody equivalent tothe CPPNL, which hadexistedon the Chinese Mainlandbefore the Nationalistgovernment’ s retreat. Scholars’evaluation of the TaiwanProvincial CPPNL seemsto be in general favourable(Tse, 1986; Kubler, 1985).Three tasksin particularare held up as exemplars:first, there wasa step-by-step promotionprogramme. Training of Mandarinpromotion personnel wasthe firststep. Training of primary and secondaryschool teachers was the second.Finally, trainingof studentsstill in schoolas well asthose already working in societywas the third.This The Language Planning Situation in Taiwan 343 step-by-step promotionmethod was deemed tobe effective. Second, the effective PronunciationDemonstrating Broadcasting Program was regarded by manyas anexcellent example asto how the massmedia could be used asan aid in language planning efforts.Finally, the TaiwanProvincial CPPNL shouldbe creditedfor its effort in the formulationof policy andstrategy to teachonly the spoken language throughthe NPS forthe first12 weeks in the firstsemester of the firstgrade in elementary schools.The policy wasmade on the basisof the experimental resultsconducted by the Committeeon the improvementof teachingmethods and teaching materials in the nationallanguage (Tse,1986: 69). In the 1960sand 1970s,with the disbanding ofthe TaiwanProvincial CPPNL, large-scale,organised efforts to promote Mandarin were largely stopped. Language planning activitiesin thisperiod, however, took a new turn.Attention wasnow paid to areas that hitherto had been pretty much neglected –suchas the teachingof reading andcomposition at the elementary schoollevel. There were even plansto construct a Mandarinproficiency testincluding atapedinterview for the oral section for all sixth-grade children in Taiwan (Zhang, 1974: 224). There wasalso a movement,spearheaded by President Yen Jia-gan,to reform the language ofofficial government documents. At that time most of the docu- mentswere composedusing semi-ClassicalChinese. The aimof the movement was to make the language more in line with the vernacular Baihua style. In the late1970s, the workof promotingMandarin in Taiwanby the lower-level committeeunder the ProvincialDepartment of Education and other educationalorganisations included activitiessuch as school spelling bees, pronunciationcompetitions, Mandarin speech contests,as well asMandarin adults’education classes and literacy programmes for employees ofvarious governmentand private institutions. Research activities during the period included scientificcharacter counts of different genres ofcurrent publications for use in reading researchor textbook compiling, Chinese speed reading,Mandarin shorthand,Chinese typewriter development, andChinese charactercomputer technology,some of whichwill be reviewed moreextensively in the sectionon language development (Kubler, 1985). Special mentionmust be madein thisconnection of a movementthat was very activein the late1960s and 70s – the Chinese CulturalRestoration Movement. When the TaiwanProvincial CPPNL wasdisbanded in 1959,some of the language planning activitieswere picked up by the committeein chargeof the Movement.As many of the committeemembers showedgreat concern over whatthey perceived tobe amuch slowerrate of progress in the promotionof the NLsince the disbanding ofthe TaiwanProvincial CPPNL, they passeda six-pointresolution which they presented tothe MOE.The MOEacceptedand announcedthem on 26 November, 1970. These resolutionswere (Chen, 1996; Kubler, 1985): (1) Immediately revive the Committeefor the propagationand promotion of Mandarinin the Ministryof Education to make unified plansand positively oversee the promotion work of the Mandarin committees at every level. (2) Increasefunding forpersonnel in the Committeefor the Promotionof Mandarin in the provincial capital and the chief sites of each county. 344 Journal of Multilingual and Multicultural Development

(3) Toachieve the goalsof the Mandarinmovement, we shouldstart simulta- neously from the following four aspects: (a) Strengthen Mandarineducation in the schoolsand cultivate Manda- rin-teaching personnel. (b) Strengthen Mandarineducation in societyand start supplementary educationprogrammes in the villages,in mines,factories, among adultsin the aboriginaltribes, and for all those who lack formal schooling. (c) Improve radioand television programmes. The amountof foreign language (i.e. English –FFT) anddialect (i.e. Southern Min –FFT) pro- grammingshould be decreasedand Mandarin programmes increased. (d) Strengthen Mandarineducation among overseas Chinese, making use oftextbooks,records, and films,etc. to promotethe Mandarinlan- guage abroad. (4) Askthe people’s representativesto use Mandarinwhen speaking atconfer- ences so as to increase its influence. (5) Require organisations,schools, offices, and all public areasto use Mandarin. Civilservants and, above all, teachers in the public schoolsshould set an example for others. (6) Toincrease interest in speaking Mandarin,various kinds ofcontestsand activitiesshould be employed thatincrease awareness among the people of the importance of speaking Mandarin. During the 1970s,some of these measureswere put intoeffect. Butas previ- ouslymentioned, the revivalof the Committeefor the Promotionand Propaga- tion of Mandarin in the MOE had to wait until 1980. After the Committeewas established, language planning activitiesagain became activeon the nationalscene. The followingare someof the mostimportant things that the Committee has done since its establishment. In 1984,the MOEannounceda revised systemof romanisation of the Chinese characters.The oldersystem, originally developed by Y.R.Chao,Lin Yu-tang andtheir colleaguesmore than half acentury ago,had been felt tobe exception- allycomplicated in thatthe four toneswere represented by lettersinstead of by diacriticmarks, and the rules ofspelling triedto reflect notonly the phonemic system,but alsonarrow phonetic information.The revised systememploys diacriticsfor the four tones:– forhigh level, /forrising, ‘ v’for dipping and\ for high falling. These marks,which arequite iconicin their representationof the actualtone values, are also ones that are used in NPS, andin the Pinyin system used in MainlandChina. In addition,the spelling rules aremade to reflect only the phonemic system,thus greatly simplifying the system.This new revised systemis mainly designed forthe use ofteachingChinese toforeigners andover- seasChinese (i.e. thosewho cannot read Chinese characters)and for local people touse in transcribingtheir namesin romanisedforms (in letters).These laststate- mentsof purpose arefelt tobe necessarybecause the governmentwants to reit- erate its stance of not abolishing Chinese characters. Many critics,this writer included, haveserious doubts about whether the revised system,created after the Pinyin systemin use in MainlandChina since The Language Planning Situation in Taiwan 345 the 1950sand which has gained world-widerecognition, would be extensively employed in the teachingof Mandarinabroad. As for its local application in the areaof the transcriptionof personalnames and place names,it has scarcely been used since mostpeople, including scholarsand linguists, are notfamiliar with it. In aword,many scholars feel thatits creation was moreto satisfy the need ofthe policy than to meet any actual demand at that time. In early1999 the issueof whether NPSorsome romanised spelling system shouldbe used in the teachingof Mandarinwas raised again and has been heat- edly debated.In the discussion,the appropriatenessof the new revised system hasagain been questioned.This issue will be takenup againin thatlarger context in the last section of the monograph. In addition,in line withthe Nationalists’claim to authenticityand its policy of notusing simplified characters,the Committeehas invested a greatdeal of its resourcesto standardise the Chinese orthography(the characters).Before the establishmentof the Committee,an ad hoc committeespent sixyears (from 1973 to1979) in the compilationof alistof standard orthography and another three yearsin itstrial use ( CentralDaily News ,9May,1983). In April 1981,this list of standardorthography of commonly used characterswas authorised and published by the MOEatthe recommendationof CPPNL ( CentralDaily News , 27 April, 1981). Any accountof NLM inTaiwanwould be incomplete withoutmentioning the roleplayed by the armyand the NationalLanguage Daily .In the 1940sand 1950s, allyoung men planning toenter the armywere encouraged toenrol in a Mandarincourse first since Mandarinwas (and still is) the language of the Taiwanmilitary forces. Later on, the army,in cooperationwith the Taiwan ProvincialCPPNL, published specialmanuals for teaching Mandarin as part of itsbasic training. Since twoto three years’military service is required ofevery young man,many who had not had a chanceto learnMandarin picked up the language during their compulsorymilitary service. This has not only helped propagatethe nationallanguage, but ithas helped promoteliteracy in the country as well (Tsao, 1998). Anotherimportant factor that has contributed to the successof NLM isthe NationalLanguage Daily ,anewspaper using transcribedcharacters, which has been popular forthe pastfifty yearsespecially amonglower-grade students. Since itsfounding in 1948,it has madea greatcontribution to the standardisation andpropagation of the nationallanguage. Itssignificance in the promotionof the nationallanguage wasnever morekeenly felt thanafter the abolitionof the TaiwanProvincial CPPNL. The NationalLanguage Daily PressService Committeefor national language educationwas organised in 1960to provide serviceto education in the nationallanguage. Totallysupported by the National Language Daily , it provides the following services: · tocompile andpublish textbooksand teaching materials in the national language; · toassist in the trainingof teachersand promoters of the nationallanguage; · toanswer, research, and experiment withproblems relatedto the national language; · toprovide otherservices related to the educationin the nationallanguage. 346 Journal of Multilingual and Multicultural Development

An evaluation Like manyother language planning programmes,evaluation, no doubt,is the weakestaspect of language planning endeavoursin Taiwan.In fact,little is knownabout whether there hasbeen anyprovision made forcontinuing evaluationof the NLMatthe nationalor locallevel. Todate,to the bestof my knowl- edge, noofficial assessment has ever been attempted.Those essays that appeared in the anthologyof papers collectedby The Executive Yuanin 1982were all impressionisticand, generally speaking, devoidof useful information.This lack ofinformationmakes this present attemptat evaluationa difficult, but aworth- while task.Fortunately for our present endeavour,many evaluative reports aboutwhat has been going onin language planning inMainlandChina since 1949arereadily available.In the evaluationwhich follows, these reportsare cited for comparison whenever feasible. Taiwan’s successfulpropagation of Mandarin Chinese asthe national language hasbeen well documented(Tse, 1987; Zhou, 1992; Tsao, 1997a). While itis certain that Tse (1987) wasover-optimisticin hisestimate of the percentage of people unable tospeak the nationallanguage (5%),as Huang (1993)and Tsao (1997a)have pointed out,he wasnottoo far off the mark.A morerealistic figure hasbeen given by Ke (1990),who based his estimate on the schoolenrolment and people’s educationalattainment figures providedby the Executive Yuan,as displayed in Table 1.

Table 1 Percentages ofeducational attainment forpeople above six inTaiwan in1987 Elementary School 37.54% Junior High 19.12% Senior High 7.40% Vocational School 16.17% Junior College 5.15% College 4.37% Graduate School 0.22% Self-study 1.65% Illiterate 7.79% Total 100.00%

Hiscalculation is as follows. The totalpercentage ofpeople witha middle schooleducation or higher is52.43%. If wethen add37.54%, the percentage of people whohad only elementary educationor whowere atthattime enrolled in anelementary school,then the totalcomes up to89.97,roughly 90%of the popu- lationaged 7orabove. In otherwords, it is safe to estimatefor those over six years oldin Taiwanin 1987,roughly 10%of themwere notable tospeak Mandarin. Thisfigure, though notas high asTse’s estimate,is actually quite remarkable. ZhouYou-Guang, who is a very seniorscholar and researcher in the propagation anddevelopment ofMandarinin MainlandChina, is of the sameopinion. In his recent book(1992), he comparedthe speed withwhich MandarinChinese has been propagatedin Taiwanand Singapore witha dragonflying andthat in Main- land China as a turtle’s crawling. 8 The Language Planning Situation in Taiwan 347

Twopoints should be madein connectionwith this phenomenal successof Mandarinpromotion in Taiwan.First, being able tospeak MandarinChinese meansthat speakers with this ability are able tomake themselves understood in the language when calledupon todo so.It does not include the abilityto carry on asustainedconversation in the language, nordoes it imply thatthey areable to speak like aPeking resident,as the normsuggests people shouldbe able todo. Quite the contrary,after fifty yearsof strenuous propagation, a number of discrepanciesin allaspects of the grammar,but especially noticeablein phonology (pronunciation),have been found (Cheng, 1985;Li, 1983; Kubler, 1985;Tsao, 1987). Many ofthe featureshave been fossilisedto the extent thatthis varietyhas come to be knownas ‘ TaiwanMandarin’ . Such discrepancies between the normand the actualspeech arenot unexpected. Itis this expected discrepancybetween the normand the actualperformance in the speech communitythat has prompted both Rubin (1971)and Karam (1974) to stress the need forevaluation in language planning andof using the evaluationresults to adjustthe normafter the language hasbeen propagatedfor a certainextended period oftime.However, in the caseof the propagationof the nationallanguage in Taiwan,either the authoritiesconcerned arenot aware of the need fordoing an evaluationor they arereluctant to dosobecause they think thatthe issueis still politicallysensitive. No evaluation of this kind hasbeen done since the normwas setup someseventy yearsago, and the failure toadjust the normhas caused language teachersa lotof problems. To begin with,they aretorn between what they areexpected toteachand what they feel they shouldteach. If they teach accordingto anunrealisticnorm, they will be teachingtheir studentsa language thatnot many people use in Taiwan,but ifthey teachwhat they think they should teach,they arenot doing the thing thatthey havebeen taughtto do. To complicatethe mattereven more,they actuallycannot teach what they areexpected to teachin anyreal sense, since manyof them are ‘ TaiwanMandarin’ speakers, and therefore they cannot serve as models for their students. Secondly, the phenomenal successhas been achieved atthe expense ofthe indigenous languages,i.e. Southern Min, Hakkaand the aboriginal Austro-Polynesianlanguages. In otherwords, while Mandarinhas been gaining in popularity,the indigenous languageshave been fasteroding. Many scholars (Huang, 1993;Li, 1994; Tsao, 1994, 1996a, 1996b, 1997a, 1997b among others) haveseen thisas anaturalconsequence ofthe government’s policy ofpromoting Mandarin,the nationallanguage, while neglecting andat timessuppressing the indigenous languages, a subject which is examined in the next section.

Language policy effects on other indigenous languages The policy and its implementation Romaine(1995: 242), after an extensive studyof the language policiesof many countries has come to the following conclusion: The traditionalpolicy, either implicitly assumedor explicitly stated,which mostnations have pursued withregard to various minority groups, who speak adifferent language, hasbeen eradicationof the native language/culture and assimilation into the majority one. 348 Journal of Multilingual and Multicultural Development

Taiwan’s pastexperience showsclearly that it is no exception. Although this policy in anovertwritten form is nowhere tobe found, allthe indicationsare it hasbeen the covertpolicy allalong. Evidence forthis is that when some open-minded scholarssuch as HongYen-chiu spoke up forthe minoritypeoples’ language rightsand argued againsta hard-line approach,he wasimmediately attackedby manyhard-liners who criticised his views as notleading tonational unity (Hong, 1978). In fact,this hard-line, high-handed propagationof Mandarinwas very preva- lent up toten yearsago. Romaine (1995: 242) reported that in Australia,the United States,Britain and Scandinavia, minority children were until recently still subject tophysical punishment in schoolfor speaking their homelanguage. In Turkey, where Kurdish isa minoritylanguage whoseexistence isnot recognised, the situationwas even worse.Thus one Kurdish womanwho attended a special boardingschool provided forKurdish children described her heartbreaking experience vividly (Clason& Baksi,1979: 79, 867, translated by Skutnabb-Kangas, 1984: 311–12): Iwasseven when Istartedthe firstgrade in 1962.My sister,who was a year older,started school at the sametime. We didn’t knowa wordof Turkish when we started,so wefelt totallymute during the firstfew years.We were notallowed to speak Kurdish during the breaks,either, but hadto play silentgames with stones and things like that.Anyone whospoke Kurdish waspunished. The teachershit us on the fingertips oron ourheads with a ruler. Ithurt terribly. That’s why wewere alwaysfrightened atschooland didn’t want to go. Many shortarticles, appearing inLin’s (1983)collection of essays,described similarexperiences thatmany indigenous language speakershad in their early yearsof schooling.My ownexperience in learning the nationallanguage in a suburban primaryschool in Taipei alsobore thisout. Even though punishment wasnot as severe asthe one the Kurdish sistersunderwent, there were several waysof punishing astudentwhen he orshe wascaught speaking Taiwanese Southern Min inschool.However, as wewere allindoctrinated with the ideaof the imminent threatof communismas well asthe importanceof national unity, andtherefore the necessityto speak the nationallanguage, these unnecessarily severe punishments were notthought to be very drasticat the time.Furthermore, manyindigenous language speakerswere informed by their teachersthat their languageswere baseand vulgar and that they shouldfeel ashamedfor being speakers of such languages. Controlof newspapersand electronic media was equally oppressive.News- papers were exclusively in Mandarin,with one ortwoEnglish papers being the exceptions.In the fifties, soonafter the Nationalistgovernment moved to Taiwan,it was stipulatedthat, in view ofthe factthat most people did notknow Mandarin,Taiwanese programmes in electronicmedia would be allowedon conditionthat they be graduallyreplaced by Mandarinprogrammes. In the seventies,it was further stipulatedthat programmes in the ‘dialects’, meaning Taiwaneseand Hakka, would be airedfor only one houra day.The ban wasin effect forabout ten yearsbefore itwas finally lifted togetherwith the lifting of martial law (Huang, 1993; Tsao, 1997a). The Language Planning Situation in Taiwan 349

Under the double oppressionof schooleducation and the massmedia, it would indeed be odd if indigenous languages did not begin to die out. The effect of the policy and its implementation Huang andChang (1995), in arecent paper onthe sociolinguistichistory of the GavalandPingpu tribe,report that in I-lan areaaround 1650 there were nearly ten thousandGavaland speakers, but by the timeProfessor Ruan did his field workin 1969,only about800 speakersremained there. If weinclude the number ofpeople whomigrated to Hua-lian,the totalwould not exceed twothousand. Butless than thirty years later even those800 speakershave disappeared, leaving the I-lan area with no Gavaland speakers. The Gaoshangroup, though luckier thanthe Pingpu tribe asitis protectedby the mountains,is actually not doing toowell. Accordingto statistics released by the government in 1989, the population of the nine Goashan tribes was: · Amis 129,220 · Atayal 78,957 · Paiwan 60,434 · Bunun 38,627 · Puyuma 8,132 · Rukai 8,007 · Tsou 5,797 · Saisiyat 4,194 · Yami 4,335. Accordingto Huang’ s (1991,1993) calculations, based on a questionnaire survey of AboriginalCollege Students,the attritionrate was estimated to be 15.8%between twogenerations and 31% between three generations.If Huang’s estimatewas correct, almost half ofthe existingaboriginal languages are going to disappear from Taiwan in another two generations. Similarresults also were obtainedin Lin’s (1995)survey report.After surveying one thousandjunior high schoolstudents studying in 25schools,Lin found that,for the aborigine students,only 37%claimed that the aboriginal language wasthe one mostfrequently used athome.Only 68%claimed that they couldspeak their parents’language andamong the lattergroup only 16% claimed to be fluent. The Hakkastudents’ performance was only slightly better thanthat of the aborigines;40% of the studentssurveyed saidthat Hakka was the most frequently used language athome. Elsewhere, accordingto Huang’ s (1993) survey of327Hakka students in the Taipei areaand 404 Hakka Taipei citizens, only 70%ofthosepeople whoseparents were bothHakka speakers claimed that they could speak Hakka. AsforTaiwanese, both Huang’ s andLin’ s survey resultsindicate that it too showssigns of erosion,although the rateis relatively slow. Furthermore, Chan’ s study(1994) shows that the domainstraditionally attributed to Taiwanese,such asthe homeand the marketplace,are shrinking, indicatingthat the dominant language, Mandarin, has made inroads upon it as well. Basedon an island-wide telephone survey of934subjects conductedby the FormosaCultural and Educational Foundation, the relationshipbetween the 350 Journal of Multilingual and Multicultural Development Table 2 Relationship ofethnic groupidentity and mothertongue identity inthree age groups for Hakka I II III Ethnic group Mother tongue II/I (%) identity(%) identity (%) L (18-30) 7.4 5.7 77 M (31-40) 13.5 12.2 90 H (41-50) 11.9 10.6 89 N.B. Figures in Columns I and II of the table refer to the percentages of subjects in that age group who claimed that identity out of total survey population, and those in Column III are the percentages obtained by dividing the figure in Column II by that in Column I.

Table 3 Relationship between ethnic groupidentity and mothertongue identity in three age groups for Taiwanese I II III Ethnic group Mother tongue II/I (%) identity(%) identity (%) L (18-30) 80.2 70.8 88 M (31-40) 76.9 74.2 96 H (41-50) 76.9 79.1 99.4 proportionof subjects claimingto be ofHakka ethnic descent,and that of subjects claimingto have Hakka as their mothertongue, forthree age groups,is shown in Table 2.Table 2clearlyindicates that the erosionof the Hakkalanguage has intensified amongyounger people (thoseaged below 30),with the erosionrate reaching a dramatic 13% decline between younger and middle-aged Hakka. Forcomparison, consider the correspondingfigures fromthe Taiwanese group shownin Table 3.FromTable 3itis quite clearthat the Taiwanesegroup showssigns of erosionas well, althoughthe rateis slower, being 8%between the mid- and low-age groups, as compared to 13% for the Hakka group. My ownlarge-scalesurvey (Tsao,1997a) also yields basicallythe sameresult, i.e.while Mandarinwas gaining popularity,all the indigenous languageswere rapidly fading.These twotendencies areclearly demonstrated in Figures 2 (Mandarinproficiency) and3 (mothertongue proficiency) respectively. From the abovestatistics it is clear that the indigenous languagesin Taiwanare disap- pearing withthe aboriginallanguages declining the fastest,Hakka close behind andTaiwanese less markedly. This shows unmistakably the effect ofthe government’s policy onthe indigenous languagesother than Mandarin, the national language.

Teaching English and other foreign languages in Taiwan Like manydeveloping countriesin the world,Taiwan’ s past language-in-education policy hasbeen toalargeextent determined by twomain factors:(1) nationalismand national unification and (2) modernisationand economicgrowth. These twofactors are not in agreement atall times. The language-in-education systemcan be seen asa resultantstate of the interaction The Language Planning Situation in Taiwan 351

Figure 2 Comparison of Mandarin proficiency in the three ethnic groups

Figure 3 Comparison of mother tongue proficiency in the three ethnic groups between these twofactors, but tosee thisclearly, we need totake a lookat the system first. In describing alanguage-in-education system,Bamgbose’ s (1991:62) charac- terisationis very useful. He suggestscharacterising a language-in-education systemby seeking answersto the followingthree questions:(1) Whatlanguage? (2) Forwhat purpose? and(3) atwhat level? Forour present purpose, the languagesinvolved can be classifiedinto three types:the mothertongue, the nationallanguage, andother languages used forwider communication.As has been shownin the previoussections, in Taiwanthe mothertongue maybe Southern Min, Hakka,Mandarin or one ofthe aboriginalAustro-Polynesian languages.The majorlanguage ofwider communication(LWC) taughtin Taiwan is English, but French, German, Spanish and Japanese are also taught. 352 Journal of Multilingual and Multicultural Development Table 4 Language type and function in education in Taiwan Literacy Subject Medium Mother tongue O O O National language X X X LWC O X O

If we nowconcentrate on the firsttwo major questions, namely, ‘ what language?’and ‘ forwhat purpose?’ we mayarrive at nine possibilities,each represented by acell in the matrixshown in Table 4.By filling in the possibilities thatare actually realised with ‘ X’,andthose unrealised as‘O’,weget the tableas shown. Educationalpractice as represented by Table 4hasexisted almost unchallenged forfifty years. 9 Asis apparent in the table,Mandarin, the national language, hasplayed avery importantrole in the system.It is taught to every- body,regardless of their mothertongue orhome language, asa language for literacy.It is also a subject takingup atleastfive hoursof instructionevery week fromthe firstgrade up tocollege freshmanlevel, andis by farthe mostimportant subject inallelementary andsecondary courses. Finally, itisthe solemedium of instructionin the schoolsystem. Competency in itplays a decisive rolein a student’s scholastic achievement. In distinctcontrast is the roleof English in the system.It is required ofevery studentin the secondaryschool and the firstyear of college. Itused totakeup five hoursof instructionper week in allyears of secondaryeducation. However, when compulsoryeducation was extended fromsix years to nine in 1970,the hoursof instructionwere cutto twoor three inthe firsttwo years of junior high school. Despite itswidely recognisedimportance in literacy,the mothertongue, except where itis also Mandarin, played absolutelyno role in the systemoffi- ciallyuntil the 1997school year, when mothertongue educationbegan tobe allottedone hourper week in the elementary schoolprogramme. I will have moreto sayabout this change in the final sectionof thismonograph when some recent developments are examined. The mothertongue wasexcluded fromthe systemon the groundsthat it has been seen asanimpediment tonationalunification. However English, as arepre- sentativeof the so-calledlanguages of wider communication,has been included forthe purpose ofproviding informationaccess to the worldof technologyand science whichTaiwan needs forsocial modernisation and economic growth. This rolefor English, however, has never been emphasisedbecause itis, at the same time,perceived asa potentialthreat to nationalism. The equilibrium was achieved by assigning English the role as set out in Table 4. Thisbeing the case,it came as nosurprisewhen anationalsurvey ofEnglish teachingin Taiwanwas conductedin 1974–1976 anda number ofproblems were found, these problemsdid notreceive much attentionfrom the mediaor the authoritiesconcerned. The survey waspartof acooperativeproject between the Departmentof Linguistics,University of Southern California(USC) andthe English ResearchInstitute, National Taiwan Normal University (NTNU). Itwas The Language Planning Situation in Taiwan 353 jointly conceived,planned andimplemented by Prof.Robert B Kaplanof USC andProf. C. M.Yangof the NTNU.Dr Philip Sedlak, whospent abouttwo years in Taiwanimplementing the plan,conducted the actualsurvey. The survey was able togathera wealthof data about English teachingin the secondaryschool in Taiwan,but forsome reason, the firstreport, which was published in June 1976, wasso hastily composed that it left much tobe desired.As pointed outby Tsao (1982) and Tse (1987), there are specifically three important shortcomings: · much of the data collected in the survey was left unanalysed; · the statistical analysis was not very revealing; and · some of the recommendations proposed were highly impractical. Fortunately,these shortcomingswere correctedby Tse,who reanalysed much of the rawdata gathered in the survey.The resultswere reportedin Tse’s PhD dissertationwritten at USC in 1979andlater published inaslightly revised form in Taiwan (Tse, 1987). The reanalysedsurvey presented alargenumber ofsignificant findings, whichtell usmuch aboutEnglish teachingin the secondaryschools in Taiwanat that time. The important findings include the following: (1) MostEnglish teacherswere inadequately trained,both in English and educational methodology. (2) Learning and writing had been emphasised in their training. (3) Despite being highly desirable,the availabilityof in-service trainingwas low. (4) The weekly hours of English instruction were inadequate. (5) Grammar and translation were given disproportionate emphasis. (6) Most tests focused on reading and writing. (7) Audiovisual aids were woefully inadequate and underutilised. (8) English contact outside school was very limited. Thisamended survey gavea true picture ofEnglish teachingat secondary level, andyielded anumber ofimportantshortcomings that called for immediate attention.However, for some reason, it did nothave as much impacton the English teachingprofession as hadoriginally been planned. Asaresult,English teachingremained pretty much the sameforsome twenty years after the survey wasconducted with only twopossible exceptions.First, in-service trainingis nowmuch moreavailable to teachers,although what effect sucha change has brought toactual teaching remains to be examined.Second, English teachers’ educationaltraining does seem tohaveimproved over the years.This is clearly revealed in Table 5,where the English teachers’educational training in 1976,the timewhen the survey wasconducted,is comparedwith that in 1996. 10 However, thisupgrading ofteachers’educational training has been due in largepart to the general expansionof universitiesand graduate schools in Taiwan.The effect of this upgrading of teachers’ qualifications remains to be determined. These shortcomingsnotwithstanding, in allfairness it might be saidthat judging fromthe economicsuccess of Taiwan in the past30 years, some credit hasto be given tothe successfulimplementation of thisparticular area of educational policy. 354 Journal of Multilingual and Multicultural Development

Table 5 Secondary school English teacher’s educational training in 1976 and 1996 Total Master and Normal Graduates of Graduates of Doctorate university other universities junior college graduates or colleges and others 1976 100% 0.4% 28.30% 61.30% 10.0% 1996 100% 9.8% 44.64% 35.66% 9.9%

Priorto the reanalysisof the survey data,Tse conducted a small-scalesurvey oflanguage use in Taiwan.Among its many interesting findings, the following are particularly notable (Tse, 1987: Chapter 4): (1) English isthe foreign language mostoften used atwork.After English,Japa- nese is most often used. German and French are rarely used. (2) Even English is not frequently used at work. (3) When English isused atwork, reading andwriting skills are most often required. (4) Cramschools and English programmessponsored by employers generally are not considered helpful. (5) Although significantlymore respondents’ attitudes towards English were positive,over one-third ofthemhave an unfavourable attitudefor national- istic reasons. Points1 and5 deserve somefurther comment.The respondents’attitudes as revealed in the survey were basicallyin line withthe government’s language policy,but even asearlyas twentyyears ago, there were signsthat nationalism wasslowly giving wayto pragmatic considerations. This comment applies to English aswell astoJapanese.Recall that in the mid-1940s,when Taiwanhad just been restoredto the Republic ofChina, the language policy wasto wipe outthe Japanese influence in the indigenous languagesand culture. However,by the late1970s, because ofTaiwan’ s heavy tradewith Japan, Japanese hadalready replaced French andGerman as the secondmost frequently used foreign language in Taiwan.This change wasto be reflected in the educationallanguage policy in the 1980sand 90s when enrolmentin Japanese classesoffered by variousuniversities showed a rapidincrease and several Japanese departments were establishedin nationalas well asprivateuniversities. At the sametime the enrolmentfigures inGermanand French classesoffered atuniversities dropped considerably. Asfar as English isconcerned,the pragmaticattitude of the people hasactu- allymade English become increasingly popular.This general popularity coupled withthe general affluence of the populace andtraditional Chinese people’s emphasison children’s educationhas induced manyparents to send their young children toEnglish language classes,which have mushroomed in the pastdecade. As this trend has grown rapidly, it has recently pushed the governmentinto changing itsearlier policy ofbeginning English educationin secondaryschool, a matterthat I will takeup in somedetail in the next section. The Language Planning Situation in Taiwan 355

Language planning activities connected with language modernisation and development Similarto what happened in MainlandChina under the Nationalistadminis- tration,language planning activitiesin Taiwanhave been centred onthe problem oflanguage unification,especially pronunciation.Comparatively little hasbeen done in the areaof language development andmodernisation. However,as the nationallanguage hasbeen increasingly used in various domainsand in the educationalsystem in the Republic, anumber ofproblems occurred that called for solutions. One ofthese problems hasto do with whether in printing Chinese horisontally,the directionshould be fromleft torightor fromright to left. Tradi- tionally,Chinese textsare printed verticallyand arereadfrom right to left. This wasfine until Chinese wasused formaterials in science andtechnology, which often consistof quantitative data and sometimes include specialterminology in Romanletters which are read from left toright.This kind ofprinted material, therefore, often containsmatters printed in different directions,causing great confusion in reading. Arelateddebate broke outin the seventies aboutthe directionof printing Chinese horizontallybetween the traditionalpurists and the modernists,and the MOEwascalled upon toprovide asetof guidelines governing the printing of Chinese. Finally, aftermuch deliberationand discussion, a compromisedsolu- tionwas reached.When printing vertically,the directionshould be fromright to left, but in orderto accommodate scientific exposition, when printing horizon- tallythe directionfrom left toright is allowed. Such acompromise,which might seem tohave pleased the contendersof bothcamps at the time,actually ended up notsolving all the problems itwas intended tosolve. In fact,it has created a number ofothers. Let us takea concreteexample. In Taiwanthere aretwo major newspapers,the China Times and the LibertyTimes .While bothfollow the guidelines propagatedby the MOEin printing vertically,i.e. from right to left, they do itdifferently when printing horizontally.The China Times printsfrom right to left, thuscreating confusion when numeralsand roman letters appear, whereas the LibertyTimes printsfrom left toright, thus requiring readersto adopt different strategies when reading vertically and horizontally. Standardisationof orthographyof personalnames and place nameshas been promotedto facilitate the use of the Chinese language withcomputer technology.Specifically, the issueis that some characters used in personalnames, bothsurnames and given names,and place names,are very rarelyused itemsor in someextreme cases,are the idiosyncraticinventions of the individuals concerned.As such, they cancomplicate computer processing and have to be standardised ( Central Daily News , 29 March, 1983). Amajorcontribution of the CPPNL afterits re-establishment in 1980has been the re-compiling andupdating ofthe Dictionaryof the National Language (hence- forththe Dictionary),which wasfirst published in 1936in MainlandChina and enlarged andbrought up todatein 1981.The recompilationwork began in June 1988and was completed in January1994. The Dictionarywith its 160,000 entries boaststhe largestlist of Chinese charactersin existence.It has been postedon the computernetwork since 1993by the MOEandrecently aCDversionof the 356 Journal of Multilingual and Multicultural Development

Dictionaryhas been madeavailable to the general public foronly the costof mailing.A conciseversion is being compiled andis expected tobe completed soon (The Sixth Educational Yearbook of ROC 1996 : 1853). Developing efficient computerinput systemsfor Chinese charactershas been acommonconcern of the CPPNL andthe Institutefor Information Industry. The latteris a non-profit organisationfounded in 1979with the followingmain functions (The Institute for Information Industry, 1998): (1) toassist the governmentin informationindustry planning andin promoting national information construction; (2) tointroduce information and communicationtechniques andconcepts and to promote information industry development; (3) topropagate information science andto train information science profes- sionals; (4) tocreate a milieu anda conditionfavourable to the development ofthe information industry and to assist its development; and (5) toassist the governmentin settingup informationand communication standards. Since itsestablishment the Institutehas taken a very activepart in these functions.However, in the areaof developing new waysof processing Chinese characters,it is fair to saythat private companies have done the lion’s shareof the workwith the Instituteplaying the roleof the coordinator.The Instituteand the CPPNL havecooperated over the yearsto complete successfully registration withthe InternationalStandardisation Office (ISO) forthe standardisedcoding ofallthe standardisedChinese characters( TheSixthEducational Yearbook of ROC, 1996:1850). Finally, there isthe perennial problem ofthe unificationof technicalterminology.In Taiwan,as in MainlandChina under the rule ofthe Nationalistgovern- ment,the agency responsible forthis has been the Instituteof Compilationand Translationwhose roles, since itsinception in 1932,have been stipulatedas: (1) the reviewing andcompiling ofallbooks on Chinese culture, the humanities, social sciences, and natural sciences and of all textbooks for all levels; (2) the translatingand reviewing ofallthe translatedworks on worldliterature, humanities, social sciences and natural sciences; (3) the translatingand compiling of technicalterminologies for different sciences; and (4) the compilationof textbooksfor all levels ( TheFifth Educational Yearbook of ROC, 1974: 895–913, Taipei: MOE). In recent years,however, the Institutehas been chiefly concerned withthe compilationof textbooksat alllevels andthe compilationof bookson Chinese culture. Even though there isstill a sectionon Natural Sciences, ithas been understaffed andvery few advanceshave been madein the areaof the unificationof technicalterminology. From 1932 to 1974,66 specific scientificand technicalterminologies had been authorisedby the MOEthroughthe Instituteof Compilationand Translation. However, since 1974only twomore have been addedand, to date,no effort hasbeen madeto evaluate how well these terms havebeen acceptedand put touse (The NationalInstitute for Compilation and The Language Planning Situation in Taiwan 357

Translation,1997: 13). There isevery indicationthat they havebeen neglected. There areprobably tworeasons for this. First, according to Liu (1970),a former Directorof the Institute,in anumber ofcasesin the wordlistsprepared by the Instituteand authorised by the MOE,morethan one translationequivalent occurs.While thispractice may actually reflect the currentsituation, it is defi- nitely againstthe principle ofstandardisation and is a reflection onthe ineffi- ciency ofthe Institute.Second, andmore importantly, authors and teachers who arefamiliar with the foreign language in whichthe loanterms originate, tend to use the original form rather than the loan (Barnes, 1974: 473).

Recent Developments in Language Planning The language planning scene in Taiwanas depicted in the previoussection mayseem ratherunremarkable tosomepeople, andIcanreadily agree withthat observationif ourfocus is on whathad been done before 1987,the yearin which martiallaw was lifted. However,since itslifting, anumber ofexciting things havebeen happening. Asmost of them lie in the areaof language-in-education planning, it is in this area that I will begin my discussion.

Recent changes in language-in-education policies Shifting of emphasis to Baihuawen in Chinese courses in secondary education Asdiscussedin the previoussection, in Chinafor more than a thousandyears before the founding ofthe Republic, the extreme linguistic diversitymeant that ClassicalChinese hadalways served as awrittenlingua franca,much like the rolethat Latin played in medieval Europe. Thus,Classical Chinese wasthe schoollanguage aswell asthe language used in the governmentand in the civil serviceexaminations. As a consequence, ahuge volumeof fine literaturewas produced in the language. Therefore, in the earlier yearsof ROC,when the school language waschanged fromClassical Chinese toMandarin,it took effect quietly andonly in the elementary school.This difference in contentwas reflected inthe namesused toidentify the programmes.In elementary schoolit was called Guoyu ‘NationalLanguage’ and in the secondaryschool Guowen ‘NationalLiter- ature’. When the Nationalistgovernment came to rule Taiwan,this distinction waspreserved. Even though the Baihua (vernacular)literature flourished inthe 1930sand 40s in MainlandChina, especially forfiction, much ofitwas tinged withpro-communist ideology andwas consequently banned in Taiwan.In practice,this strongly biased the contentsof the Chinese coursesin the secondary schooland the curriculum ofthe Chinese departmentsin the universities, including the normaluniversity, towards Classical Chinese literatureand againstthe modernlanguage andthe vernacularliterature. Let us take for example the Chinese Departmentof the NationalTaiwan Normal University, the leading departmentwhere thousandsof Chinese teacherswere trained.In the Chinese Departmentat the university,there areabout70 facultymembers, about 20ofwhomare listed under the linguisticssection. However, in actuality,half of thosetwenty were teachersof Mandarinpronunciation, a coursewhich up tofive yearsago used tobe required ofevery normaluniversity student.Nine outof the remaining ten professorsspecialise in philology ratherthan modern linguistics. 358 Journal of Multilingual and Multicultural Development

Anotherindication that modern linguistics and the methodologyof language teachinghave been flagrantlyneglected canbe clearlyshown by examination topicsfor the masters’theses and doctoral dissertations that have been writtenat the department.From 1961 to 1995, a totalof 158 doctoral dissertations were written falling under the following categories: 11 Chinese classics 64 Chinese philology 24 Chinese history 2 Chinese literature: 61 (classical, 57; modern, 4 others 7 The 531masters’ theses produced from1958 to 1995 fall intothe following categories: Chinese classics 180 Chinese philology 74 modern Chinese studies7 Chinese history 21 Chinese literature 231 (classical, 224; modern, 7) Chinese arts 7 others 11 Lookingat these dissertationsand theses from another perspective reveals thatthere isnot a single thesisor dissertationon anything remotelyrelated to language teaching.Another point that cannot be missedis the extremely skewed distributiontowards classical as opposed to modern Chinese; in the caseof dissertations153 relate to classical topics, 5 tomodern, and in the caseof theses 517 are classical, 14 are modern. Itdoes not take much imaginationto see howeffective aMandarinteaching programmecan be, when itis taught by teacherstrained in aprogrammewith sucha strongbias towards Classical Chinese literature.In fact,Taiwan’ s college-bound studentsare found tobe quite lowin their writtenChinese proficiency. Asa language professorat aleading Taiwanuniversity, I frequently have been surprisedby the number ofcomplaints I havereceived frommy colleagues in science andtechnology about the poorcommand of Chinese thattheir graduate advisees have as reflected in their reports, papers, and theses. The students’low proficiency in Chinese musthave been the reasonthat promptedthe MOEtore-examine itsearlier policy ofplacing somuch emphasis onthe teachingof ClassicalChinese inthe secondaryschool curriculum. In the end, asensible decisionwas made in the curriculum standard;as of the 1997 schoolyear, the ratioof modernChinese toClassical Chinese in the firstyear of the junior high (equivalent tothe 7thgradein the United States)was tobe 8to2, but the ClassicalChinese proportionwill be graduallyincreased as the students progress through their secondary education. Thisis in facta long overdue change in the rightdirection. However, like so manyother changes that will be discussed,there isa serioushiatus in the deci- sionmaking. The change wasmade without taking the teachers’prior training The Language Planning Situation in Taiwan 359 intoconsideration and hence noprovision has been madeto retrainor re-educate them.12 Consequently, itsexpected effect onpromotingstudents’ Chinese proficiency remains to be seen.

Issues concerning the national phonetic symbols Aheated debate hasbeen ragingin Taiwanfor the pastthree yearshaving to dowith the NationalPhonetic Symbols (NPS). Recallthat there areactually two setsof NPS. NPS1,which employs componentsof traditional Chinese characters assymbols,was promulgated by the MOEon23November 1918and has played avery importantrole in promotingMandarin, the nationallanguage, in Taiwan. NPS2,which employs romanisedletters as symbols, was first devised by the famouslinguist, Y.R. Chaoand his colleagues, and was promulgated in 1926. NPS2 waslater revised inTaiwanon the ground ofitsextreme complexityand promulgatedby the MOEin 1986.The revisionof NPS2 wasevidently prompted atleastin partby the factthat the setof phonetic symbolspromoted by Mainland Chinasince 1956,officially knownas HanyuPinyin Fangan ‘Chinese Phonetic Scheme’(henceforth CPS), hasbecome sowidely acceptedthat the very existence of the original NPS2, and even NPS1, was threatened. In Taiwanthis has led toadebate,on-going nowfor some time, that has to do with two closely related issues: (1) In teachingMandarin Chinese toChinese people andto speakersof other languages,is NPS1 abetter scheme thanone employing romanised symbols? (2) If the answerto (1) isno, then whichof the three currently available schemes,i.e. NPS2, CPS orTong-Yong Phonetic Scheme (Yu &Xu,1998) (henceforth, TYPSforshort), a newly devised phonetic scheme designed by ananthropologist working at Academia Sinica, is the mostsuitable one? These questions are addressed in the following two sections. NPS1 vs. a romanised phonetic scheme Opinionswith regard to the firstissue have been divided.On the one hand,we havethe traditionalistswho argue thatNPS1 shouldcontinue tobe used, pointing fortheir supportto the following twoobservations. First, since Taiwan’s policy isto continue toteachChinese characters,the NPS1,being made up ofsymbolsderived fromcomponents of traditionalcharacters, inevitably is morecompatible with writing and printing ofChinese characters.The factthat itssymbols are derived fromChinese characterswill alsoenable itto provide a better transitionfrom learning the phonetic symbolsto learning Chinese characters.Second, andperhaps moreimportantly, NPS1, as has been repeatedly pointed out,has played avery importantrole in the promotionof Mandarin Chinese in Taiwan. The modernists,on the otherhand, have argued thata romanisedphonetic scheme shouldbe employed in the teachingof Mandarin,at homeand abroad. They havetwo arguments in supportof their view.First, the continued use of NPS1 isan impediment tothe modernisationof the Chinese language sinceit failsto provide either auniversally availableway of indexing oraneasyinput systemto the computer,as the systemis only knownin Taiwanand some restrictedareas of the world.Second, NPS1’s presumed advantageof being a 360 Journal of Multilingual and Multicultural Development better instrumentthan a romanisedphonetic scheme iscalledinto doubt since a romanisedphonetic scheme hasnever been triedin Taiwanand since a romanisedsystem has been inuse in MainlandChina for more than forty years with no reported undesirable effects. Currently (in 1999),the issue isstill unresolved, and it is difficult tosee what the outcomeof this policy deliberationwill be. However,I amofthe opinion that since Taiwanhas internationalisation as amajorobjective, easyaccess to infor- mationexchange andcommunication through the computerwill be afactorthat will only gainin importancewith time. This consideration, coupled withthe fact thatmany children actuallyrecognise lettersin the English alphabeteven before they enter elementary school,will eventually tip the scalein favourof using the romanised phonetic scheme instead of NPS1.

Which of the three romanised phonetic schemes is the most suitable? There arethree romanisedphonetic schemescurrently in use in Taiwan (namely, the NPS2,the TYPS, andthe CPS), but whichone isthe best?To answer thisquestion, let usfirst set up somecriteria for comparison. In devising a phonetic scheme ormore commonly a writingsystem for a language, there are three importantconsiderations: economy, consistency and convenience (Fishman,1968). Economy usually meansthat a phonetic scheme isprimarily basedon the phonemic systemof the language, i.e.there isa symbolfor each phoneme andwhere there isno phonemic contrastno additional symbols need be provided.Consistency means that one symbolstands for one soundand there areno other symbols that stand for the samesound. Conversely, one soundis represented by only one symboland there areno other sounds represented by the samesymbol. There area number ofnotions subsumed under the general rubric convenience .First,a setof phonetic symbolsis regarded asconvenient ifit canbe easilylearned. For example, the symbolsare so devised thatthe letterscan be easilyassociated with the soundsthey represent. Secondly, asetof phonetic symbolsis also convenient if itis easily processed in writingand in printing, which in ourpresent daytechnology means easily processed by using a computer.Finally, asetof symbols is taken to be convenient if itcan be generally used, allowingfor slight modifications, in anumber ofclosely associated languages. Since the CPSpromotedin MainlandChina is the oldestsystem, I will begin mydiscussionwith the CPSandthen compareit with the othertwo schemes. The CPS, which wasapproved by the Congressof the PRCin February 1958,can be briefly summarised in Table 6. In examining the scheme,the followingpoints may be noted(De Francis, 1967): (1) The symbol u represents ahigh backrounded vowelexcept after y and the palatalinitials j, q, x,when itrepresents ahigh frontrounded vowel.This is a fairly ingenious solutionto the problem ofthe symbol?, the use ofwhich can now be confined to combinations with l and n. (2) The symbolshave been sochosenthat there arebut rareoccasions to use the juncture symbol. The Language Planning Situation in Taiwan 361 Table 6 Initials Unaspirate Aspirated Nasals FricativesVoiced d stops stops continuants Labials b p m f Alveolars d t n l Alveolar sibi- z c s lants Retroflexes zh ch sh r Palatals j q x Gutterals g k h

Finals (Rimes) 1. simplex rimes: a, e, i, u, u 2. duplex rimes : ai, ei , ao, ou, an, en, ang, eng 3. semivowels (as initials): y, w Tones 1. ma-; 2. ma/; 3. mav3; 4. ma\; 5.ma (neutral tone) Juncture pi’ao

(3) The symbol i represents ahigh frontunrounded vowelexcept after retroflexesand alveolar sibilants when itrepresents the twodistinctive vocalisations of these two sets of initials. (4) The schemais in general quite economicalin thatit is basedon the phonemic systemof Mandarin.Take i mentioned in point3 forinstance. Even though phonetically atleastthree different pronunciationscan be found depending onthe kind ofconsonantthat precedes it,only one symbol i isused since these different pronunciations are non-contrastive. (5) Itis also in general quite consistent.The only pointat whichthe questionof consistencymay be raisedis in the factthat the symbol h isused bothfor the glottalfricative and for retroflexion as in the caseof zh,but since in the latter caseit is the secondpart of adigraph andin the formerit occurs independ- ently, the possibility of causing ambiguity is very small. Withrespect to convenience, ithas the followingmerits. (1) Digraphsare rarelyused, there being only three in the initials.(2) All the 26letters in the English alphabetare put touse in representing one soundor another, thus makingcertain that the English keyboard isfully utilised.(3) Diacriticmarks havebeen reduced tothe minimum,there being four forthe five tonesand the umlauting markfor ü,whichwe justobserved has been reduced totwocases, i.e. after l and n where minimalpairs between ü and u canbe found. Onthe other hand,some symbols used havebeen found tobe notso easilyassociated with the 362 Journal of Multilingual and Multicultural Development

Table 7 Differences between CPS and NPS2 Initials Finals (Rimes) Dental Palatals Retro- Semivowels Simplex Complex sibilants flexes (as initials) vowels vowels CPS z c q x zh y w i ü ao NPS2tz ts ch sh j i u Ø iu au (after r,z) soundsthey represent, namely, c,analveolar sibilant, q,anaspirated palatal, and x,apalatalfricative. In thisway, they complicatethe acquisitionof these symbols. Withthis brief descriptionand evaluation of the CPS asbackdrop,we cannow proceed tocompare CPS withNPS2 andTYPS. Table 7summarisesthe differences between CPSandNPS2. From Table 7,itis clear that, as faras initialsare concerned,the majordifferences lie in howthe alveolarsibilant series and the palatalseries are treated.While NPS2 stressesthe virtue ofsound– symbol association,thus choosing digraphs to represent affricates,CPS attemptsto find some lettersin the English alphabetas yet unoccupied by otherChinese soundswith a view to fully utilising the English keyboard. In the areaof vowels,the differences arefew andof minorimportance. Take the caseof high frontrounded vowelfor example. CPSchooses ü,aletterfound in French andGerman, but notin English,thus requiring adiacriticmark if English keyboard isadopted. NPS2, on the otherhand, selects a digraph iu torepresent the sound,obliterating the need fora diacriticmark but atthe sametime running up the cost in typing as two keys have to be pushed instead of one. Overall,however, it seems to manythat the twoschemes are actually quite similar.But since NPS2 wasdevised aboutthirty years later than CPS, ithas to be better in somewayto justify itsexistence andthis consideration is fully reflected in the final choices made. Thisgeneral attitudeis alsoreflected in TYPS, the mostrecent invention,as canbe clearlyseen in Table 8,which showsthat TYPS iseven moresimilar to CPS thanNPS2. Actually, the inventor,Mr Yu,has made a virtue ofthe factthat the scheme canbe easilyconverted to either NPSorCPS andcanalsobe easilymodi- fied torepresent otherHan dialectssuch as Southern Min andHakka, as well as the aboriginalAustro-Polynesian languages spoken in Taiwan.This is the reason

Table 8 Differences between CPS and TYPS Initials Finals (Rimes) Dental sibilants Simple Complex CPS j q x I ü en eng TYPS z c s ii* yu un** ong*** * only after dental sibilants and retroflexes ** only after w *** only after w & f The Language Planning Situation in Taiwan 363 why the inventorcalls his scheme Tong-YongPinyin ‘literally,general-use spelling’and has actually written a few pamphletsdemonstrating how this can be done. Returning toour comparison between TYPS andCPS, we find thatour previousremarks on the differences between CPSandNPS2 canbe applied here, i.e.the twosystems are basicallythe samewith some adjustments made by TYPS as‘improvements’over the CPS. These ‘improvements’are necessary, as in the previouscase of NPS2,to justify itsexistence. However,there isanother more importantfactor that should be takeninto consideration in the final decision regarding these modifications,and this factor has to do with how Taiwan perceives itself in relationto the MainlandChina, a topicthat I will discussin the nextsection. Here, in orderto facilitate discussion, I will assumethat Taiwan tendsto perceive itself asaseparateentity, independent ofMainland China, even though atthe sametime it admits that culturally it isclosely related to the latter andthere isa very strongneed forcommunication. This self-perception and general attitudetowards Mainland China are fully reflected in the designs of NPS2 andTYPS, i.e.against the general backdropof similarity, there aresome differences tokeep themdistinct. Since NSP2 isofficialwhile TYPS isnot, the former tends to be more conservative than the latter. Itis still too earlyto sayforsure howthis issue will be decided, but the whole issuehas aroused the attentionof the Executive Yuanand the MOE.Asthis writerwas working on this monograph, the Ministerof Education, Dr Lin Qing-Jiang announcedon 11February, 1999thata meeting will be held in March todeliberate onthe issue.He hasproposed that the scheme tobe used be determined according to the following three principles: (1) The scheme shouldbe instrumentalin helping the nationto promoteinter- nationalisation. (2) Itshould be easilylearnable given the present language-education situation in Taiwan. (3) The selectionshould take into consideration the factthat the MOEdecided in 1996that, as of that year, the streetnames and road signs should use NPS2;some counties and cities have already allotted some money forthe change (China Times , 11 February, 1999: 9).

An arising new supra-ethnic identity Asindicated in ourdiscussion of the socio-historicalcontext, Taiwan is, amongother things, an immigrantsociety and like manyimmigrant societies, it isbeset withproblems ofethnicity, language loyaltyand group (i.e. supra-ethnic) identity.In fact,there isperhaps noother place in the modern worldwhere people areas divided in their opinionswith respect totheir group identity asin Taiwan.Fortunately, as aresultof somerecent socio-politicaldevel- opments,a new group identity seemsto be emerging, indicatingthat ethnic harmonycould be achieved if the trendcontinues. There havebeen clearindica- tionsthat more and more Taiwan residents have come to identify themselves withthe place in whichthey live andcall themselves ‘ Taiwanese’. In ordernot to be confused with‘ Taiwanese’in itsold sense ofreferring tothe indigenous people ofTaiwanin contradistinctionwith ‘ Mainlanders’, anew termhas been 364 Journal of Multilingual and Multicultural Development coined– ‘New Taiwanese’– torefer tothis rising new supra-ethnic identity.In thissection, in additionto identifying whatthis new identity is,I explain why it hastaken so long forthis new group identity toemerge andwhat role language plays in the process. Socio-political context: A brief recapitulation Recallthat earlier wesaidthat during the Qing dynastyas moreand more Chinese immigrantsfrom Kuangtong and Fujian cameto settle on the island, four ethnic groupswere graduallyformed: the Zhangzhoupeople, the Quanzhoupeople, the Hakkaand the aboriginalAustro-Polynesian, the firsttwo groupsbeing Southern Min speakerswith different accents.Small-scale ethnic conflictswere almostdaily occurrences during thatperiod, and large-scale bloodshedand feuds were notuncommon (Lamley, 1981).However, when it wasannouncedat the conclusionof the FirstSino-Japanese WarthatTaiwan was tobe ceded toJapan, people in Taiwanfelt frustratedand humiliated because they hadbeen deserted by their motherland,and because they felt anurgent need todosomethingto protect themselves. It was this sense ofhumiliationand the fear ofbeing ruled byaforeign people thatstirred them into action. The First TaiwanRepublic washastily founded on16 May, 1895. Unfortunately, the Republic, lastingonly 148days,was soon defeated andoverthrown by the Japa- nese army.This incident, together with other signs, was interpreted by Huang (1993)as the beginning ofthe processof transforming Taiwan society from a purely immigrantone intoone ofmore-or-lessnative ethnicity (different from their ethnic Chinese origins). When the Japanese cameto rule the island,they ofcoursedid everything they couldto prevent thisgroup identity fromcoming into being, asthistrend was diametricallyopposed to their interestin JapanisingTaiwan. This being the case, noprogressin the formationof group identity wasmade during thosefifty years ofJapanese administration.However, being put under arepressive foreign rule evidently createda feeling ofbeing ‘related’or being ‘in the sameboat’ . In addition,during thathalf century the Chinese immigrantswere largely cutoff from their ancestralhome in MainlandChina and ethnic Chinese onthe islandgradu- allycame to identify withthe placesin whichthey resided (e.g. Zhanghua or Tainan)ratherthan their placesof originon the Mainland.Partly because ofthis changing conceptof their place of originand partly because of the Japanese government’s policy forbidding fighting between ethnic groups,ethnic conflicts gradually died down. When Taiwanwas returned toChinain 1945,people in Taiwan,having been placed under oppressive,discriminatory foreign rule forhalf acentury, warmly welcomedthe opportunityto become citizensof the Republic ofChina, expecting tobe treatedas equalsunder the new government.This high expecta- tion,as I indicatedearlier, never materialised.Misunderstandings abounded during thosefew yearswhen the rulersand their followersfrom the Mainland cameinto contact with the localpeople because they lackeda commonlanguage andthey didnot share a collectivememory, having been completely cutoff from eachother for fifty years.However as it turned out,the misunderstandings, insteadof being removedthrough patient and careful explanation,were actually The Language Planning Situation in Taiwan 365 increasedby the inefficient administrationof the ProvisionalGovernment headed by Chen Yi. On28February, 1947,the tragic228 incident broke out,killing thousandsof people, Mainlandersas well asthe indigenous inhabitants;the relationsbetween the indigenous groupsof the people andthe Mainlanderswere severely strained. Even though thisaccident was soon put downby troopsfrom Mainland China andeven though Chen Yi,the administrativehead, was sentenced todeath,the relationshipbetween the indigenous people andthe Mainlandnewcomers was nottruly restoreduntil very recently. Formany people the tragicevent wasthe seed ofseparation that later developed intothe TaiwanIndependence Move- ment. Twoyearslater, the Nationalists,having lost their warwith the Communists, retreatedto Taiwan, bringing abouta millionfollowers with them. Even though thisgroup wascomposed of people fromvirtually every province ofMainland China,speaking allthe majorHan dialectsand a number ofminoritylanguages, they andtheir children bornlocally after 1949 were perceived as Waishenren, ‘Mainlanders’by the indigenous people, whoreferred tothemselves as Benshenren,‘Taiwanese’, orto use the morecolourful metaphor of the local people, the distinction is between yam ‘Taiwanese’ and taro ‘Mainlanders’. 13 The Mainlanders(partly because manyof themwere associatedwith the Nationalistgovernment in one wayor another, partly because they were, on average,better educatedthan the localpeople atthattime, and partly because mostof them could manage to communicate in Mandarin,having learned Mandarinin schoolor picked itup in the military)occupied mostof the impor- tantpositions in the government(see Huang,1993, and the references cited therein). These discriminatoryofficial hiring practicescontinued forabout thirty years,even though the educationallevel ofthe localpeople sooncaught up with thatof the Mainlanders.That hiring practicebegan tochange in 1972when Chiang Ching-Kuo, the formerROC president whowas then serving asthe Premier, began to introduce young local talent into his cabinet. Onthe language front,things were nobetter. The language policy ofthe Nationalistgovernment can be briefly described asuni-directionalbilingualism (Chan,1994), i.e. while allspeakers of alocallanguage haveto learn to speak Mandarin,the nationallanguage, the Mainlanders,most of whom could speak someform of Mandarinwere notrequired tostudy a locallanguage. Thispolicy wasimplemented moreeffectively in schoolswhere thosewho spoke their mothertongue were punished. Students were taughtthat it was unethical and unpatrioticto speak their mothertongue if itwas a language otherthan Mandarin.In the massmedia, the use ofindigenous languageswas, for a long time,severely restricted.With all these repressive measuresin effect formore thanthirty years, it is little wonder that the use ofthe indigenous languageshas declined significantlyand that some of them face extinctionin agenerationor two. The adoptionof these relativelydiscriminatory policies in language andin the appointmentof public officers,especially athigher levels, wasconducive to conflictsbetween ethnic groups.These ethnic conflicts,in their turn,worked againsta genuine group identity.As aconsequence, people were splitin their 366 Journal of Multilingual and Multicultural Development viewson identity, with most Mainlanders considering themselves Chinese and most local people considering themselves Taiwanese. In addition,the government’s policy towardsMainland China had been ultra-conservative.For about forty years, the Nationalistgovernment considered Taiwana temporaryresidence, their final goalbeing torecoverMainland China.While mostMainlanders understandably honoured and cherished this policy,most local people hadlong agorealised how unrealistic and wishful such apolicy was.This split of opinionsalso contributed to widening the ethnic divide and was again not conducive to the emergence of a supra-ethnic identity. Signs of ethnic reconciliation and the emergence of the new identity The firstsign ofethnic reconciliationappeared when indigenisationof the island’s politiciansoccurred in 1972,under the processbegun by Chiang Ching-kuo. Soonafter that, in orderto makehis intention clear, Chiang, who was bornand raised in MainlandChina, proclaimed, ‘ I’maTaiwanese,too’ (Chung, 1999).Later, he choseLee Teng-hui tobe hissuccessor, and when Chiang died in 1988the latterwent on to become the firstTaiwan-born Hakka to govern Taiwan.14 Democratisationin politicsbegan in 1986when the firstopposition party, the DemocraticProgressive Party (henceforth DPP) (whosemembers were chiefly Taiwanesenatives advocating separatism) was founded andwastolerated even while the repressive martiallaw was still in effect. The DPP gained strength when martiallaw was revoked in July 1987.In 1992,the partymade an impres- siveshowing in the firstmajor democratic legislative election basedon universal suffrage, winning aboutone-third ofthe seatsin the Legislature,which had been occupied by Mainlanderssince the Nationalistgovernment retreated to Taiwan in 1949.Meanwhile, the ruling partyalso underwent democratisationas more andmore Taiwan-born Kuomingtang legislators appeared. Asthe voiceof the localmajority began tobe heard,relations between different ethnic groups improved. In 1993,a secondopposition party, the New Party(which wasmainly composedof Mainlanderswho advocated unification with China) wasfounded andTaiwan was onitsway toexperiencing preliminary multi-partydemocracy. In March1996, the people of Taiwanelected Lee Teng-hui astheir president throughdirect general election forthe firsttime in the historyof the island.As politicalresources have become increasingly proportionallydistributed among the ethnic groups,ethnic disharmonybegan tothaw, giving the new supra-ethnic group identity a chance to emerge. Noteverything, ofcourse,has gone well. Onthe diplomaticfront, setbacks havecome one afteranother since 1970when Taiwan(known asthe Republic of China) left the United Nationsin anticipationof the passageof a resolution admittingthe People’s Republic ofChina,and expelling Taiwanfrom the world organisation.This traumatic event starteda three-decade-long processof diplo- maticsetbacks with the United Statestaking the leadin severing diplomaticrela- tions with the Republic of China (ROC). Butfortunately, as Taiwan experienced the setbackswhich greatlyreduced its internationaldiplomatic space, a new sense ofgroup identity began toemerge – moreand more people in Taiwancame to identify withthe islandinstead of The Language Planning Situation in Taiwan 367

MainlandChina, where they ortheir ancestorsoriginated. The changeswere acceleratedin 1996when, during the height of the presidential election,the MainlandChinese conducteda seriesof military exercises and missile tests with the obviousintention of intimidatingTaiwan people intovoting against separatism.However, this sabre rattling backfired andhelped manypeople in Taiwandecide toidentify themselveswith the place in whichthey lived. This change canbe clearlyseen by the following comparisonof survey results.In a 1992survey, 26.9% of the respondentsidentified themselvesas Taiwanese (Wang,1993). When the samecategories were calculatedin a1996survey taken immediatelyafter the missilethreat (Sun andMa, 1996), the figure was46%,indi- catinga strongshift towards the Taiwaneseand away from the Chinese end of the identity spectrum within the short time span of four years. Meanwhile, governmentpolicies alsoshowed changes in the samedirection. Aspreviously mentioned,since 1997Taiwanese Southern Min aswell asHakka andthe aboriginallanguages have been promotedon the islandto the extent that they arenow taught in elementary schools.When comparedto the national language, Mandarin,all the indigenous languages,especially Southern Min, are getting moreand moreattention and now are spoken in public by manygovern- ment officials in the hope of relating better to the general public. Atthe sametime, some of the villagesfor military dependants aroundthe islandhave been reconstructedinto new communitieswhere Taiwaneseresi- dents mingle with the original residents of Mainland origin. Anothermajor measure that reflects the government’s desire toblur the line between ethnic groupswas taken when the Ministryof Internal Affairs decided tochange the formatof ROCidentificationcards. For cards issued before 1992 there wasa smallbox printed onthe backof the cardthat provided aspacefor the identificationof the cardholder’s ‘nativeplace’ , whichmeant the place from whichhis or her fatheroriginated on the Chinamainland (if the fatherimmi- gratedto Taiwan around 1949), or the place inTaiwanwhich was considered the cardholder’s hometown (if hisor her fatherhad arrived prior to 1945or wasa Taiwannative). This box was removed from cards issued after 1992. Such a change signalled ade-emphasis onone’s connectionwith Mainland China and an emphasis on personal identity with Taiwan. In line withthis governmental attitudinal change wasa view expressed by Chen Shui-bian, Taipei’s formermayor and a presidential hopeful likely torepre- sentDPP inthe upcoming presidentialcampaign in 2000.Chen dismissesthe metaphorof ‘yam’and ‘ taro’, preferring the image‘ peanut’. ‘In fact,we areall peanuts’, he saidat apublic gathering in late1998, addingthat, just like peanuts, people in Taiwanshould take root easily and be able toidentify withthe landon whichthey dwell andwhich they think ofas their permanent home(Chung, 1999:8). Such amessagecoming from a high-profile figure in the DPPstrongly indicatesthat the DPP istaking a pragmaticapproach and will likely embrace ethnic reconciliation in its platform in the forthcoming presidential election. The role language has played in the process Mandarinchauvinism, as I reportedearlier, used tobe quite commonin Taiwan.But with the conceptof ‘New Taiwanese’becoming prominent,will Southern Min, alanguage spoken by about70% of the people in Taiwan,take the 368 Journal of Multilingual and Multicultural Development place ofMandarin?While someobservers express apprehension aboutthis (see Chung, 1999),others have reported that the roleplayed by language in character- ising thisnew supra-ethnic identity doesnot seem particularlysalient. Accordingto Sun andMa’ s (1996)survey, only 22%of the young people considered aTaiwaneseto be apersonwho could speak Southern Min, andonly 9% defined aMainlanderas a personwho could not speak Southern Min. These resultsindicate that the abilityto speak Southern Min isnot an importantcrite- rionin distinguishing aTaiwanesefrom a Mainlander.The moresalient factors, accordingto the samesurvey, are being bornin Taiwan(55%); living in Taiwan (49%);regarding oneself tobe Taiwanese(39%); and having Taiwan listed as ‘native place’ (38%) (items allowed for multiple choices). Fromthese findings, itdoes seem thatlanguage playsa salientrole in charac- terisingthe emergent conceptof ‘New Taiwanese’. However,if weexamine the rolethat language hasplayed while the new identity concepthas been developing, we will find anotherstory. The strenuouspromotion of the national language in the firstthirty years after Taiwan’ s Retrocessionin 1945resulted in the functionalallocation of the four majorlanguages in Taiwan.Mandarin servedas the high language andSouthern Min, Hakkaand the indigenous aboriginallanguages served as the lowlanguages, forming adiglossicsociety withsocietal bilingualism where Mandarinalso served as the effective lingua franca.If the promotionof Mandarin had been kept atthislevel, the resultant statewould have been alotmore acceptable to all ethnic groupsconcerned. However,as wehaverepeatedly pointed out,the repressive policieswere kept fortoo long, and as a resultall languages except Mandarinare either quickly diminishing in use orare on the verge ofextinction. The sadstate that many indigenous languageswere in causedgreat resentment among the people. Once martiallaw was revoked, the resentmentthat had been suppressed forso long broke outin strongprotests in somecases, or manifested itself in the increasing use ofthe localmother tongue asa symbolof defiance againstthe government authority or simply as an expression of ethnic identity. Fora while itseemed thatlanguage wouldserve asa dividing forcerather thanas aunifying one in Taiwan.But as the conceptof ‘New Taiwanese’began to takeshape, a change ofattitude with regard to the use oflanguage tookplace – the use oflanguage began tobe ‘lessemotionally loaded and more pragmatically oriented’(Tse, to appear). Thistrend towards pragmatism can be detected in the use oflanguage in the televised campaignspeeches given by the four setsof pres- identialand vice-presidential candidatesin 1996(Tse, to appear) aswell asin the morerecent Taipei andKaohsiung mayoral elections and the legislativeelection (Chung, 1999;Kuo, 1998). To be morespecific, acommoncharacteristic in all these campaignspeeches wasthe use oflanguage towin votesrather than to rally forethnic identificationand division. In otherareas involving the use of language, the sameattitude is also found. Aradiostation run by the New Party, whichis largely composedof Mainlandersand sympathisers for reunification with China, has certain call-in programmes conducted in Southern Min. Asthe attitudeof treatinglanguage moreas ameansof communicationand lessas a markerof ethnicity gainsground, Mandarin, which has developed intoa lingua francain Taiwan,is likely tobe used moreby people whoused toemploy Southern Min orHakkaas agesture ofdefiance. Indeed, ithas been observedthat The Language Planning Situation in Taiwan 369 somepoliticians of the DPPshowless apologetic attitudes when they use Mandarinin public domains,especially in the massmedia (Tse, to appear; Kuo, 1998).Today, the only group ofpeople whoinsist on using Taiwaneseas a symbolof group solidarityare the members ofthe TaiwanIndependence Party (TIP), anewly establishedbreakaway group fromthe DPP, dissatisfiedwith the lessradical and more practical stance on the separatistissue. This view onthe use oflanguage isnot even sharedby DPPLegislatorYe Ju-lan, whootherwise is a greatsympathiser with the TIP. Instead,Ye advocatesthe ideathat the term‘ Tai- waneselanguage’ should be used torefer toall languages used onthe island ratherthan for Southern Min alone( GlobalViews Monthly 121,June 15,1996. pp. 79–80). Summary Tosummarise,in Taiwanin the pastfew yearswe canclearlysee anew sense ofgroup identity emerging. Thisemerging sense ofnew identity,which hasbeen termed‘ New Taiwanese’, hasmore to dowith the sharedfeelings amongpeople ofdifferent ethnic groups,towards the landin whichthey live, Taiwan,towards modernityand democracy, and towards the uncertaintyof their future relations withMainland China. In short,it is an identity built onasharedway of life and the commonfate of living onthe sameisland under the militarythreat of the People’s Republic ofChina.While thisnewly arisinggroup identity isnot tied to anylanguage atthismoment, language hasbeen observedto play animportant role in its development.

A Look at the Future In thismonograph, after a brief accountof the socio-historicalcontext, I have ventured toexamine criticallythe language planning situationin bothMainland Chinabefore the Nationalistgovernment moved to Taiwan and in Taiwanafter itsRetrocession, paying specialattention to somerecent developments occurring aftermartial law was lifted in 1987.While thisexamination seems to havefound moreinadequacies than strengths, I am,nonetheless, quite optimisticwhen thinking abouthow future language policy andplanning mightdevelop. This optimism stems from the following observations. (1) Tobegin with,language policy-making before martiallaw was lifted in 1987 hadalways been aone-way affair,i.e. it had always been top-down, allowingvery littleinput fromthe general public andfrom experts. Now there areclear indications that, as Taiwan moves towards democracy, public opinion andexpert adviceare playing anincreasingly importantrole in the processof language policy making.This is something that people in Taiwancould hardly have imagined even ten yearsago. As far as policy implementationis concerned, we find thatthe legislativebody of the governmentis paying moreattention to itso thatits practice may be sound. (2) In the closelyrelated area of language education,the oldsystem (as represented in Table 4),that existed for about forty years unchallenged and unchanged, hasbeen closelyexamined since the lifting ofmartiallaw. Some changes,such as adding one hourof mother-tongue instructionto the elementary schoolcurriculum, have been implemented. Otherchanges, like 370 Journal of Multilingual and Multicultural Development

beginning English educationin the fifth gradeand startingthe instructionof other LWCs in secondary school, will be implemented in 2001. (3) Plansare now being madeto reduce the classsizes in primaryand secondaryschools from the present 45to50 per classto about30 per class. When implemented, these planswill certainlygreatly facilitate language teaching and learning. (4) An importantimprovement in teachertraining has been implemented beginning in 1997.Previous to that, the three normaluniversities and nine teachers’colleges were the principal sourcesof primaryand secondary schoolteachers. It is easy to imagine thatproblems mighthave emerged out ofsucha closedsystem over the years.Realising the ills,the Ministryof Educationdecreed twoyears ago that universities and colleges otherthan thosenormal universities and teachers’ colleges will be allowedto train primaryand secondary school teachers. It is hoped thatthis new additionto the teachersupply canprovide the additionalteachers needed asaresultof the planned class-sizereductions. At the sametime, by makingthe job marketfor teachers more competitive, the development shouldinduce changesin the oldteacher training institutions, which had become stagnant due to lack of competition. (5) Finally, andmost importantly, as the use of language isbecoming less emotionallycharged and more pragmatically oriented, inter-ethnic rela- tionsin Taiwanbetween the four majorethnic groupshave shown signs of improvement.There ishope thatTaiwan will be able toemerge healthy fromits bitter past, which was full ofethnic conflictsand tensions. While ethnic harmonymay notbe easilyachieved, ethnic reconciliationmay well be in sight.Rather than ‘ yams’or ‘taros’, people maychooseto be ‘peanuts’.

Correspondence Any correspondenceshould be directedto ProfessorFeng-fu Tsao,Institute of Linguistics,National Tsing-Hua University, Hsinchu, Taiwan, China ([email protected]).

Notes 1.Bamgbose (1991)in his chapteron language planning alsomaintains this distinction although he does mention some overlapping cases.In this monograph, whenever such acaseoccurs, an arbitrary decision will bemade astowhich category it belongs to. 2.No census dataare availablebecause questions concerning people’s ethnolinguistic background were considered toosensitive tobe included in the previous census ques- tionnaires. 3.See Li (1992)and the references cited there fora summary ofarguments in supportof this view. 4. TaiwanStatistical Data Book ,various issues, Councilfor Economic Planning and Devel- opment as cited in Shieh (1998). 5.For ageneral discussion ofthe drawbacksin selecting anartificial or made-up composite language as the national language see Bamgbose (1991). 6.In retrospect, it is certainly agreat pity thatthe extension wasstopped atthe elementary school level, forlanguage educationin aliving language in allfour skills cannot becompleted in six years. The effectsof this oversight have been keenly felt in Taiwan today.Please refer tothe section on ‘Recent changes in language-in-education policies’ for more discussion. The Language Planning Situation in Taiwan 371

7.The February 28incident, commonly known in Taiwan asthe 228incident, occurred on28February, 1947.Although there have been anumber oftheories astowhy it occurred,no consensus has been reached. Itis doubtless, however, the most serious ethnic conflictin Taiwan during this century. Among its many far-reaching effectsis the spawning ofthe Taiwanese Independence Movement (TIM)that has been in existence formore than forty years. Officialstudies ofthe incident were forbidden under martial lawand reconciliatory measures such asmaking 28February apublicholiday were taken only aftermartial lawwas lifted in 1987.For further discussion relating to the incident see the section on the ‘Socio-political context’ of this monograph. 8.Knowledgeable asZhou was,he wasnot ableto give anestimate ofhow many people in Mainland China in about1990 were ableto speak Mandarin. Asurvey conducted by Kuang MingDaily in 1998indicated thatonly 22%of the respondents claimed tobe native speakers ofMandarin. The report, however, did not mention how many are able to speak it as a second language ( China Times , 14 December, 1998, p. 14). 9.The only minor change in the system in its fiftyyears ofexistence is the addition offour hours ofmother-tongue instruction in the third tothe sixth grade in elementary schools, starting in the 1997 school year. 10.Data for the year 1996were adaptedfrom Table3-5 ofthe StatisticalAbstract of Educa- tionin theRepublic of China,1997 ,published bythe MOE. The figures in the original tablerefer tosecondary school teachers asawhole butsince we have no reason to expectthat English teachers asa group will bedifferent from secondary school teachers, we have used the figures for comparison. 11.The dataare adaptedfrom the appendix to Papersby the Faculty and Graduate Students at the Graduate Institute of Chinese, National Taiwan Normal University, Vol. 39 . 12.The lackof training in teaching modern Chinese alsoexplains why, soon afterthe MOEmade the announcement ofthe change in 1994,many Chinese teachers went into a panic and protested. 13.The localpeople callthemselves ‘yam’because on the map,Taiwan looks like ayam. 14.Lee Teng-hui is ethnically a Hakka, but his first language is Southern Min.

References Bamgbose, A.(1991) Languageand the Nation: The Language Question in Sub-SaharanAfrica . Edinburgh: Edinburgh University Press. Barnes, D.(1974)Language planning in mainland China: Standardization.In J.A.Fishman (ed.) Advances in Language Planning (pp. 457–80). The Hague: Mouton. Chan,H.C. (1994)Language shift in Taiwan:Social and politicaldeterminants . Unpublished PhD dissertation. Georgetown University. Chang, K.Z.(1995) An examination ofsome approachesto the study ofindigenous Taiwan from the standpoint ofarcheology. In Y.H.Shi et al. (eds) Papersfrom the First Conferencein theIndigenous Culture of Taiwan (pp.1– 8). Taipei: College ofHumanities, Taiwan Normal University. Chao, Y.R. (1943) Language and dialects in China. The Geographical Journal 102, 63–66. Chao,Y.R. (1976b)What is correctChinese? In A.S.Dil (ed.) Aspectsof Chinese Sociolinguistics (pp. 72–83). Stanford, CA: Stanford University Press. Chen, M.R.(1996)A study oflanguage educationalpolicies in Taiwan afterits retrocession. Unpublished MA thesis, National Taiwan University [in Chinese]. Cheng, R.L. (1985)A comparison ofTaiwanese, Taiwan Mandarin, and Peking Mandarin. Language 61 (2), 352–377. Chung, O. (1999) Neither yam nor taro. Free China Review (February), 6–13 De Francis, J.(1967)Language and script reform. InJ.A.Fishman (ed.) Advancesin the Sociology of Language II (pp. 450–75). The Hague: Mouton. Fang, S.D.(1965) FiftyYears of theChinese National Language Movement .Taipei: Mandarin Daily Press. [in Chinese] Figueroa, E.(1988)Evaluating language policy in Taiwan –some questions. InR.L. Cheng and S.F.Huang (eds) TheStructure of Taiwanese: A ModernSynthesis (pp.285– 299). Taipei: Crane. 372 Journal of Multilingual and Multicultural Development

Fishman, J.A.(ed.) (1967) Bilingualism with and without diglossia: Diglossia with and without bilingualism. Journal of Social Issues 32: 29–38. Fishman, J.A.(1974) Language planning and language planning research: The stateof the art. In J.A. Fishman (ed.) Advances in Language Planning . The Hague: Mouton. Formosa Culturaland EducationalFoundation (1997)A survey on people’s language ability and their attitudetoward language policy.In F.F. Tsao(ed.) Ethnolinguistic Policy:A Comparisonof the Two Sidesof the Taiwan Straits. Appendix 7 (pp.181– 188) Taipei: Crane [in Chinese]. Hong, Y. C. (1978) Miscellaneous Talks on Language . Taipei: Mandarin Daily Press. Hsu,S. J.(1991)The language problems in the early years ofTaiwan retrocession. Si-Yu Yuan (Thought and Language) 29, 4–23. Huang, C.C.(1998)Historical reflections on the postwarTaiwan experience from an agrarian perspective. InC.C.Huang and F.F. Tsao(eds) PostwarTaiwan in Historical Perspective (pp. 17–35). Bethesda, MD: University of Maryland Press. Huang, S.F.(1991) Some sociolinguistic observations on Taiwan. TheWorld of Chinese Language & Literature 7 (6), 16–22. [in Chinese] Huang, S.F.(1993) Language,Society, and Ethnicity: A Studyof theSociology of Language in Taiwan. Taipei: Crane [in Chinese]. Huang, S.F.and Chang, Z.C.(1995) A sociolinguistics investigation ofKavaland: An Aboriginal language. In P.R.Li and Y.Lin (eds) Collectionof Paperson theAustronesian Languagesin Taiwan (pp.241– 256). Taipei: Councilon EducationalResearch, Ministry of Education. Jernudd, B.H. (1973)Language planning asatype oflanguage treatment. In J.Rubin and R.Shuy (eds) LanguagePlanning: Current Issues and Research (pp.11– 23). Washington, DC: Georgetown University Press. Karam,F.X. (1974)Towards adefinition oflanguage planning. In J.A.Fishman (ed.) Advances in Language Planning (pp. 103–24). The Hague: Mouton. Ke, J.-X.(1990) The national language educationin the Taiwan Province: Past,present, and future.Paper presented atthe Annual Conference ofthe Chinese Language Teachers Association, Nashville, Tennessee, USA. Kubler, C.C.(1985) TheDevelopment of Mandarin in Taiwan:A CaseStudy of Language Context. Taipei: Student Book. Kuo,S.H. (1998)Language has adividing aswell asaunifying function.In the opinion forum section, United Daily (Dec. 15), Taipei [in Chinese]. Lamley, H.J.(1981) Subethnic rivalry in the Ching Period. InE.M.Ahern and W.Gates (eds) The Anthropology of Taiwanese Society . Stanford, CA: Stanford University Press. Li, David C.C.(1983) The sociolinguistic contextin Taiwan:Trends and developments. In C.Chu,S. Coblin and F.F.Taso(eds) Papersfrom the Fourteenth Conference on Sino-Tibetan Linguistics and Languages (pp. 257–78). Taipei: Student Book. Li, PaulJ.-K. (1979) The origin ofthe Taiwan Aboriginal peoples: Linguistic evidence. Dalu Journal 59 (1), 1–14 [in Chinese]. Li, PaulJ.-K. (1990) Classification of Formosan languages: Lexical evidence. Bulletinof the Institute of History and Philosophy 64 (4), 813–848. Li, PaulJ.-K. (1992) TheInternational and External Relationships of the Formosan Languages . National Museum of Prehistory Planning Bureau [in Chinese]. Li, PaulJ.-K. (1995) The distribution ofthe Austronesian languages in Taiwan and the migration oftheir speakers. In F.F. Tsaoand M.Tsai (eds) Papersfrom the First InternationalSymposium on Languagesin Taiwan (pp.1– 16). Taipei: Crane [in Chinese]. Li, P.O.(1994)An evaluationof the current language policy on Mandarin and the indigenous dialects and languages. In P.R.Huang (ed.) Paperson CurrentLanguage ProblemsTaiwan (pp.147– 164). Taipei: Chinese Department, NationalTaiwan University. Lin, J.(ed.)(1983) CollectedPapers on LanguageProblems in Taiwan .Taipei: Taiwan Wenyi Zazhishe [in Chinese]. Lin, J.P.(1995) Mother tongue and culturaltransmission. In R.K.Li and Y.Lin (eds) Collectionof Paperson theAustronesian Languages in Taiwan (pp.203– 222). Council on Educational Research, Ministry of Education. The Language Planning Situation in Taiwan 373

Liu, T. (1970) An interview by Dayle Barnes, as cited in Barnes (1974). Ministry ofEducation(1974) TheFifth Educational Yearbook of theROC .Taipei: MOE[in Chinese]. Ministry ofEducation(1996) TheSixth Educational Yearbook of the ROC .Taipei: MOE[in Chinese]. Ministry ofEducation(1997) StatisticalAbstract of Educationin theRepublic of China 1997 . Taipei: MOE [in Chinese]. NeustupnýJ.V. (1970)Basic types oftreatment oflanguage problems. Linguistic Communications 1,77– 99. Reprinted in J.A.Fishman (ed.)(1974) Advancesin Language Planning (pp. 37–48). The Hague: Mouton. Paulston,C.B. (1984) Language planning in education.In C.Kennedy (ed.) Language Planning and Language Education . London: George Allen and Unwin. Romaine, S. (1995) Bilingualism (2nd edn). Cambridge, MA: Blackwell. Rubin,J. (1971)Evaluation and language planning. InJ.Rubin and B.H. Jernudd (eds) Can Languagebe Planned? Sociolinguistic Theory and Practice for Developing Nations (pp. 218–52) Honolulu, HI : University of Hawaii Press. Shi, T.F. (1987) TheGeographical Distribution of the Han EthnicGroups During the Qing Dynastyand Their Ways of Livingin thePlace of Origin. Taipei: Department ofGeography, National Taiwan Normal University [in Chinese]. Shieh, S.C.(1998)Unique features ofTaiwan’ s economic development (1955–1995). In C.C.Huang and F.F. Tsao(eds) PostwarTaiwan in HistoricalPerspective (pp.36– 54). Bethesda, MD: University of Maryland Press. Skutnabb-Kangas,T. (1984) Bilingualismor Not: The Education of Minorities . Clevedon: Multilingual Matters. Sun,S.H. and Ma,S.R. (1996) Who are the new Taiwanese? GlobalViews Monthly 121, June 15, 1996: 46–50. The Executive Yuan(1982) Reviewand Criticism of the NL: Policiesand Implementation . Taipei: The Executive Yuan, ROC [in Chinese]. The Institute forInformation Industry (1998) Profileof theInstitutefor Information Industry . Taipei. The NationalInstitute forCompilation and Translation (NICT)(1997) Profileof theNICT . Taipei: NICT [in English and Chinese]. Tsao,F.F. (1982)The NationalTaiwan Normal University–University ofSouthern CaliforniaSurvey ofEnglish Teaching in the Republicof China: An evaluationreport. EnglishLiterature and Linguistics 8,128– 134 [English Department, NationalTaiwan Normal University]. Tsao,F.F. (1987)Re-identifying the norm forGuoyu. In A.Brankamp et al. (eds) Chinese–Western Encounter: Studies in Linguisticsand Literature: Festechrift for Franz Giet, SVD on theOccasion of his85th Birthday (pp.385– 409). Taipei :Chinese Materials Center Publications. Tsao,F.F. (1994)Social changes and linguistic adaptations.In P.-R.Huang (ed.) Papers on CurrentLanguage Problems (pp.27– 52). Taipei: Chinese Department NationalTaiwan University [in Chinese]. Tsao,F.F. (1996a)Language educationin Taiwan:A review in sociolinguistic perspective. In P. Storey et al. (eds) Issuesin Languagein Education (pp.153– 166). Hong Kong: The Hong Kong Institute of Education. Tsao,F.F. (1996b)The teaching ofthe national language and the mother-tongue in Taiwan. In S.Y. You et al. (eds) Papersfrom the Conference on theFuture ofthe Chinese Language in HongKong after 1997 (pp.15– 31). Hong Kong: The Chinese Language Societyof Hong Kong [in Chinese]. Tsao,F.F. (1997a) EthnicLanguage Policy: A Comparisonof the Two Sidesof theTaiwan Straits . Taipei: Crane [in Chinese]. Tsao,F.F. (1997b)Preserving Taiwan’s indigenous languages and cultures:A discussion in sociolinguistic perspective. InN.Inoue(ed.) Globalizationand Indigenous Culture (pp. 97–112). Tokyo: Institute for Japanese Culture and Classics, Kokugakuin University. 374 Journal of Multilingual and Multicultural Development

Tsao,F.F. (1998)Postwar literacy programs in Taiwan:A criticalreview in sociolinguistic perspective. InC.C.Huang and F.F. Tsao(eds) PostwarTaiwan in HistoricalPerspective (pp. 158–180). Bethesda, MD: University of Maryland Press. Tsao,F.F. (in press) Diglossia, bilingualism and Taiwan’s language education.In C.S. Tung (ed.) Papersfrom the International Symposium on Languagesin Taiwanand their Teaching (Vol. 2). Taipei: The Association of Languages in Taiwan. Tse, K.-P(1986) Standardization in Chinese in Taiwan. InternationalJournal of Sociologyof Language 59, 25–32. Tse, K.-P(1987) LanguagePlanning andEnglish as a ForeignLanguage in MiddleSchool Education in the Republic of China . Taipei: Crane. Tse, K.-P(to appear) Language and arising new identity in Taiwan. InternationalJournal of the Sociology of Language . Tsuchida, S.(1983)Austronesian languages in Taiwan (Formosa).In S.A.Wurm and S. Hattori (eds) LanguageAtlas of the Pacific Area .Canberra:The Australian National University. Wang, F.C.(1993) The nature ofethnic assimilation. In M.K.Chang (ed.) Ethnicityand National Identity (pp. 53–100). Taipei: Ye-Chiang Publishing Co. [in Chinese]. Wen, C.Y.(1985) The industrialization and the socialchanges in Taiwan.The China Forum editorial committee (ed.) SocialChange and Culture Development in Taiwan (pp. 1–40). Taipei: Lianging Publishing [in Chinese]. Yu,B.-Q. and Xu,Z.-Q. (1998) Tong-YongMandarin Phonetic Scheme .Taipei: Lan-tian Book Co. [in Chinese]. Yuan, J-H. et al. (1960) ASurveyof Han Dialects .Peking: Wenzi Gaige Chubanshe [in Chinese]. Zhang, B.-Y.(1974) HistoricalData of theNational Language Movement in Taiwan . Taipei: Commercial Press [in Chinese]. Zhou, Y.-G.(1992) ASynchronicand Diachronic Account oftheLanguages in Chinaand Their Writing Systems . Peking: Renmin Jiaoyu Chubanshe.

Further Reading Chang, K.Z.(1987) The archaeologicalstudy ofthe southern coastof China and the origin of the Austronesian People. The Archeology of Southern People 1, 1–14. Chao,Y.R. (1976a)Some contrastive aspectsof the Chinese national language movement. InA.S.Dil (ed.) Aspectsof ChineseSociolinguistics (pp.97– 105). Stanford, CA: Stanford University Press. Chen, E.S.H.(1988) Functional perspective on the modernization ofthe Chinese language. Journal of Chinese Linguistics 16 (1), 145–50. Chen, S.H. (1979) Population Changes and Social Changes in Taiwan . Taipei: Lianjing. Cheng, C.C.(1975)Directions ofChinese charactersimplification. Journalof Chinese Linguistics 3 (2/3), 213–220. Cheng, C.C.(1976)Chauvinism, egaliterianism, and multilingualism: China’s linguistic experience. Studies in Language Learning 1 (2), 45–58. Cheng, C.C. (1979) Language reform in China in the seventies. Word 30 (1/2), 45–58. Fang, S.D.,Zhang, F.P.and Zhang, X.Y. (1972)A brief history ofthe national language movement in the pastsixty years. InF.R. Chang (ed.) TheChinese Studies in thePastSixty Years (pp. 461–554). Taipei: Zhengzhong. Hsiau,A.-C. (1998) Language ideology in Taiwan:The KMT’s language policy,the Tai-yu language movement and ethnic politics. Journalof Multilingual and Multicultural Development 18, 302–315. Li, X.(1998)A critique ofTong-Yong romanized symbols. TheWord of Chinese Language 90, 27–35 [in Chinese]. Lu,L-j. (1988)A survey oflanguage attitudes,language use and ethnic identity in Taiwan. Unpublished MA Thesis, Fu Jen Catholic University. Sedlak, P.A.S.(1976) Reportof theNationalTaiwan Normal University– University of Southern California Survey of English Teaching in the Republic of China . Taipei: Wan Pang Press. Sung,Wen-hsiun (1980)The prehistory ofTaiwan: An archeological perspective. InChen, C.L. (ed.) Taiwan:A Partof China .Taipei :Zhongyang Wenwu Gongyigshe [in Chinese]. The Language Planning Situation in Taiwan 375

Tse, K.-P.(1982) Language policy in the Republicof China. In R.B. Kaplan(ed.) Annual Review of Applied Linguistics (pp. 33–47). Rowley. MA: Newbury House. Tse, K.-P(1995)Bilingual educationand language planning. TheWorld Chinese Language 75, 32–36 [in Chinese]. van den Berg, M.E. (1986) Language Planning and Language Use in Taiwan . Taipei: Crane. van den Berg, M.E.(1988) Taiwan’ s sociolinguistic setting. InR.L. Cheng and S.F.Huang (eds) The Structure of Taiwanese: A Modern Synthesis (pp. 243–261). Taipei: Crane. Yang, B.C.(1987) AConiesHistory of Taiwan .Kaohsiung: The First Publishing [in Chinese]. Yang, R.(1989) LanguageMaintenance and Language Shift among the Chinese on Taiwan . Taipei: Crane.