Papers from the Sixth Nordic Conference on Bilingualism

cover

title: author: publisher: isbn10 | asin: print isbn13: ebook isbn13: language: subject publication date: lcc: ddc: subject:

next page >

Papers From the Sixth Nordic Conference On Bilingualism Multilingual Matters (Series) ; 80 Herberts, Kjell Multilingual Matters 1853591378 9781853591372 9780585253930 English Bilingualism--Congresses. 1991 P115.N67 1990eb 404/.2 Bilingualism--Congresses.

cover

next page >

< previous page

page_i

next page > Page i

Papers from the Sixth Nordic Conference on Bilingualism MULTILINGUAL MATTERS 80 Series Editor: Derrick Sharp Edited by Kjell Herberts and Christer Laurén

< previous page

page_i

next page >

< previous page

page_ii

next page >

< previous page

page_iii

next page > Page iii

Multilingual Matters The Acquisition of Irish MAIRE OWENS Age in Second Language Acquisition BIRGIT HARLEY Attitudes and Language COLIN BAKER Bicultural and Trilingual Education MICHAEL BYRAM and JOHAN LEHMAN (eds) Breaking the Boundaries EUAN REID and HANS H. REICH (eds) Citizens of This Country: The Asian British MARY STOPES-ROE and RAYMOND COCHRANE Continuing to Think BARRIE WADE and PAMELA SOUTER Education of Chinese Children in Britain: A Comparative Study with the United States of America LORNITA YUEN-FAN WONG English in Wales: Diversity, Conflict and Change N. COUPLAND in association with A. THOMAS (eds) Fluency and Accuracy HECTOR HAMMERLY Fourth International Conference on Minority Languages, Vols I & II D. GORTER, J.F. HOEKSTRA, L.G. JANSMA and J. YTSMA From Office to School: Special Language and Internationalization C. LAURÉN and M. NORDMAN Key Issues in Bilingualism and Bilingual Education COLIN BAKER Language, Culture and Cognition LILLIAM MALAVÉ and GEORGES DUQUETTE (eds) Language in Education in Africa CASMIR M. RUBAGUMYA (ed.) Linguistic Minorities, Society and Territory COLIN H. WILLIAMS (ed.) Migration and Intercultural Education in Europe U. PÖRNBACHER (ed.) Minority Education: From Shame to Struggle T. SKUTNABB-KANGAS and J. CUMMINS (eds) Multilingualism in India D. P. PATTANAYAK (ed.) One Europe - 100 Nations ROY N. PEDERSEN Our Own Language GABRIELLE MAGUIRE Reversing Language Shift JOSHUA A. FISHMAN Special Language: From Humans Thinking to Thinking Machines C. LAUREN and M. NORDMAN Please contact us for the latest book information: Multilingual Matters Ltd, Bank House, 8a Hill Road Clevedon, Avon BS21 7HH England

< previous page

page_iii

next page >

< previous page

page_iv

next page > Page iv

Disclaimer: This book contains characters with diacritics. When the characters can be represented using the ISO 8859-1 character set (http://www.w3.org/TR/images/latin1.gif), netLibrary will represent them as they appear in the original text, and most computers will be able to show the full characters correctly. In order to keep the text searchable and readable on most computers, characters with diacritics that are not part of the ISO 8859-1 list will be represented without their diacritical marks. British Library Cataloguing in Publication Data A CIP catalogue record for this book is available from the British Library. ISBN 1-85359-137-8 Multilingual Matters Ltd UK: Bank House, 8a Hill Road, Clevedon, Avon BS21 7HH, England. USA: 1900 Frost Road, Suite 101, Bristol, PA 19007, USA. Australia: PO Box 6025, 83 Gilles Street, Adelaide, SA 5000, Australia. Copyright © 1991 Kjell Herberts, Christer Laurén and the authors of individual chapters. This book is also available as a special issue of the Journal of Multilingual and Multicultural Development Vol. 12, Nos 1&2, 1991. All rights reserved. No part of this book may be reproduced in any form or by any means without permission in writing from the publisher. Typeset by Photographics, Honiton, Devon. Printed and bound in Great Britain by Short Run Press, Exeter.

< previous page

page_iv

next page >

< previous page

next page >

page_v

Page v

Contents Introduction Kjell Herberts and Christer Laurén

vii

Comments on the Topics of the Sixth Nordic Conference on Bilingualism Bengt Sigurd

17

LANGUAGE IMMERSION PROGRAMS AND THEIR IMPLICATIONS Directions in Immersion Research Birgit Harley

919

The Catalan Immersion Program: The Joint Creation of Shared Indexical Territory Josep Maria Artigal

2133

Ritualised Routines and L2 Acquisition Siv Vesterbacka

3543

Children in Welsh-Medium Education: Semilinguals or Innovators? Peter Wynn Thomas

4553

Immersion Principles in Second Language Programs: Research and Policy in Multicultural Australia Michael Clyne

5565

A Two-Phase-Didactics for School Christer Laurén

6772

MULTILINGUAL INDIVIDULS IN MULTILINGUAL SETTINGS Triangulation and Trilingualism Reitze J. Jonkman

7383

Switching Between Seven Codes Within One Family: A Linguistic Resource Bent Søndergaard

8592

ABORIGINAL LANGUAGES The Sámi Language: Pressure of Change and Reification Marjut Aikio

93103

SOVIET LANGUAGE POLICY AND ITS CONSEQUENCES Influence of Ideology in the Linguistic Policy of the Soviet Union 105110 Mart Rannut

< previous page

page_v

next page >

< previous page

page_vi

next page > Page vi

The Changing Language Situation: Russian Influences on Contemporary Estonian Mati Hint

111116

A Sociolinguistic Comparison Between Québec's Charter of the French Language and the 1989 Language Laws of Five Soviet Republics Jacques Maurais

117126

THEORY OF LANGUAGE PLANNING Language Planning as Discipline Björn H. Jernudd

< previous page

127134

page_vi

next page >

< previous page

page_vii

next page > Page vii

Introduction The papers included in this volume represent a selection of those given at the Sixth Nordic Conference on Bilingualism held in the bilingual city of Vaasa/Vasa in Finland, June 4 to 6, 1990. Another selection including papers in English as well as in Scandinavian languages is being published by Institutet för finlandssvensk samhällsforskning, P.B. 311, SF-65101, Vasa. The comments on the conference as a whole and on the Nordic activities in the field made by Bengt Sigurd are included in this volume in order to provide a Scandinavian analysis for an international audience. KJELL HERBERTS, INSTITUTET FÖR FINLANDSSVENSK SAMHÄLLSFORSKNING (VASA), ÅBO AKADEMI UNIVERSITY. CHRISTER LAURÉN, SCHOOL OF MODERN LANGUAGES, UNIVERSITY OF VAASA.

< previous page

page_vii

next page >

< previous page

page_1

next page > Page 1

Comments on the Topics of the Sixth Nordic Conference on Bilingualism Bengt Sigurd Department of Linguistics and Phonetics, Lund University, Lund, Sweden The organisers have kindly asked me to comment on the topics of the conference and to speculate on the future of the small languages in Europe, in particular the Nordic languages. The word bilingual can be used in two different ways. It can be used 'regionally' about an area, as in: Finland is a bilingual country, or 'individually' as in Kjell Herberts is a bilingual person. All men and women in a country may speak two languages each, but the country may be termed bilingual even if half the population speaks one language and the other half another language. This is a rare case, however. Regional bilingualism is generally combined with frequent individual bilingualism. This situation is often dangerous for the minority language, however, if it does not have good support from multilingual speakers. This was pointed out by several speakers at this conference, thereby challenging its motto : Bilingualism is a natural resource. The paper by Robert Comet I Codina attacked the motto openly when commenting on the situation in Spain. It was also maintained by a local politician that a bilingual policy in our nice host city Vaasa/Vasa was, in fact, a threat to the bilingual character of the city. But I have not heard anybody claim that bilingualism is not a resource from the point of view of the individual. It is a general opinion among the participants of conferences like this that two (or more) languages make it possible to live in several worlds and to see things from different angles. But 'out there' it is a widespread opinion that mastering two languages may be a burden, and unfortunately linguists have contributed to the dispersion of this myth by inventing and using the word semilingualism (Swedish: halvspråkighet). There was an interesting paper by Wingstedt & Stroud at this conference pointing to the dangers of linguistic terminology and metaphors. The term halvspråkighet is extremely well known in Sweden, maybe supported by a vague underlying idea that the space available for language is fixed and if it has to be divided by two languages each of them can only

< previous page

page_1

next page >

< previous page

page_2

next page > Page 2

be half. We who believe that competence in two or more languages is generally good for you must always be prepared to fight for this idea 'out there'. The Sixth Nordic Conference on Bilingualism included papers on both meanings of the word bilingual. But considering the dramatic developments in Central and East Europe and the whole of the Soviet Union I think the reports from the Estonian scholars about the situation in Estonia and about the Soviet language policy attracted special interest. I was shocked to learn to what extent the language policy, the russification, of the Soviet Union has destroyed and was intended to destroy the ethnic identities of the different regions. Roughly speaking, about half of the contributions to this conference dealt with regional bilingualism including the problems of minority languages, and half with the psychological, educational and social problems of individual bilingualism. But many papers dealt with both topics and a number of papers dealt with general methodological or theoretical problems related to the area. I congratulate the organisers on having managed to attract such a wide range of interesting papers from all over the world. Languages Treated and Languages Used at the Conference The number of languages treated at this conference is impressive. I note Australian languages (several papers), Catalan, Spanish, Turkish, French, Welsh, English, German, Dutch, Frisian, Estonian (several papers), Ingermanland Finnish, Karelian, Sami (Lappish), Danish, Norwegian and Swedish. I note two particularly exotic languages, which I had not heard of before, namely: Montagnais, an Algonquian language in Canada and Misiones-Swedishthe remnants of a little Swedish settlement in South America. This set of languages indicates that the problems of bilingualism are universal, but it is also clear that certain countries have invested more in bilingualism research than others. Seemingly, countries have the research they deserve. It is no coincidence that Canada, Catalonia and Finland play an important role in the study of bilingualism (cf. relevant journals). It is also to be noted that Finland has much more research into translation and training of translators than the other Nordic countries. Swedish research into bilingualism is fairly new and came about as a result of the recent immigration. Until recently Sweden was a very homogeneous and monolingual nation, and Swedes were not aware of the fact that bilingualism is very common in the world. The languages of the immigrantsand the immigrantshave had a most stimulating effect on the field of linguistics in Sweden. Language typology was a fairly esoteric field until the immigrants arrived. Typological differences became most interesting to those who were to teach Swedish to these immigrants and wanted to compare their languages with Swedish. There is an interesting survey and history of the research into bilingualism

< previous page

page_2

next page >

< previous page

page_3

next page > Page 3

among the papers at this conference by Kuure & Siponen. They follow the two conflicting ideas: bilingualism is good/bad for you from Grammont (1902: good) and Epstein (1915: bad). The paper was given in Swedish, however, and listeners had to know Swedish or at least a Nordic language and this leads me to comment on certain signs of changes in the language situation in Northern Europe. Judging from the abstracts there has been a great number, some 20, presentations in Swedish out of a total of about 60 presentations. I do not take this as an indication that these 20 speakers believe that Swedish is a world language, a competitor of English (or Esperanto). But I take it as a sign of the belief in the Nordic languages as a means of communication, when the audience is mainly Nordic, a kind of Neo-nordism. I have noted a similar trend at other conferences during the last few years. A New Nordism? I have participated in a couple of Nordic conferences in recent years, two in Iceland (Nordiska Datalingvistdagarna, Nordisk forskarkurs i översättning) and one on the Faroe Islands (Nordic and General Linguistics). Although these conferences had nonNordic participants and the topics were international many of the speakers chose to present their papers in a Nordic language. Swedes use Swedish, Danes Danish, Norwegians Norwegian (a variant) and scholars from Iceland and the Faroe Islands use Danish. All Icelanders and Faroe Islanders are (at least) bilingual and understand and talk Danishsome say they speak the clearest and best Danish. All of them learn Danish as their first second language at school. As I remember similar conferences some ten years ago all papers were given in English and the participants seemed eager to show their command of English, the international language of the sciences, including linguistics. This was partly an effect of the dominance of the American school of linguistics, generative transformational grammar, at the time. The will to use a Nordic language at this and other recent conferences may have meant taking stand for the Nordic languages and the Nordic unit or community. Maybe English is seen as a threat to Swedish, Danish and Norwegianalthough everyone must agree that the acceptance of English as an international medium has had great advantages and made possible the modern kind of scientific international conferences. I think many people are beginning to think that it might be wise to hold on to the Nordic neighbours in the present world of unrest and new political constellations. The Nordic languages symbolise a common heritage and ethnicity. The mutual Nordic comprehensibility may turn out to be of importance if the situation in Europe develops in certain ways. I think, of course, of the emerging Great German nation, which may make the Nordic countries feel smaller and bring them closer together. It is also clear that the Baltic countries have oriented themselves towards the Nordic countries recently. The contacts

< previous page

page_3

next page >

< previous page

page_4

next page > Page 4

with Sweden are now very frequent, and I noted the Estonian demand for Swedish and Swedish teachers mentioned by the Estonians speakers at this conference. TV3a Multinordic TV Channel Speaking about the language situation in Scandinavia I want to tell you about an interesting new experiment in bilingualism or rather quatrolingualism carried out by a new commercial channelTV3. The channel TV3 is sent via satellite (Scansat) from London, but the presentators, the men and women reading the news, reading the weather reports, leading the games, etc. all speak a Nordic language: Danish, Norwegian (a variant) or Swedish. And there is no translation (no subtitles). The commercials may be read in Danish, Norwegian, Swedish or English and the channel may therefore be said to be quatrolingual. The English moments are, however, marginal and the channel is clearly Nordic. The owners of TV3 have started a Nordic channeland managed to do something which the politicians and adminstrators of the Nordic Council have not. The Nordic Council has spent years investigating and discussing whether it would be possible to have a Nordic channel and came to the conclusion that it is unfeasible. One main reason is that the council does not believe in the mutual comprehensibility of the Nordic languages. The experts suggested that a staff of translators is needed in order to translate the texts and to carry out simultaneous interpretation between Danish and Swedish, Danish and Norwegian, etc. The experts proposed that a staff of translators and interpretors be established, but the cost of such an arrangement would, of course, be forbidding. And the decision where to place such an organisation for translation could certainly cause a minor Nordic war. The commercial channel TV3 does not owe its existence to sheer idealism, nordism, of course. Rather, the owners are interested in the bigger Nordic market of some 20 million people. But, as a matter of fact, TV3 is an experiment in Nordic multilingualism. Its development is worth following closely from many points of view and for many reasons. The Situation of Minority Languages The signs of a new nordism mentioned are, however, a mild wind compared to the endeavours to reinstate the regional ethnic languages in various corners of Europe and the Soviet Union in recent years. Peeters gave an interesting survey of the language situation in Europe and the work with a charter to protect the rights of minority languages in the Council of Europe. The political situation in Europe and the Soviet Union is changing rapidly, and so is the language situation. Various scenarios can be imagined for the future.

< previous page

page_4

next page >

< previous page

page_5

next page > Page 5

Peeters offered criteria which may be combined into a kind of ladder or scale which allows one to see how high a country has climbed in its treatment of minority languages (or how low it has sunk). I suggest the following ladder or scale with steps for the evaluation of countries with several languagesthe bottom being the most deplorable: (1) Several co-official languages (2) One official language, but individual language rights granted (3) One official language, ignorance of minority languages (4) One official language, repression of minority languages It is interesting to see how countries are placed or move up or down the ladder, but it is necessary to go into more detail and specify different kinds of language use (domains or niches). One may suggest the following important domains of language use: 1. Legal procedures and documents 2. Administration and bureaucracy 3. Teaching and education 4. Business life 5. Entertainment 6. Family life The Estonian papers at this conference witnessed how the Estonian language was completely dominant during the time of the Estonian Republic 19181940. With the occupation by the Red Army in 1940, the deportation or killing of Estonians and importation of Russians, Russian was made the only official language. Russian was made to occupy the top niches: legal procedures, administration, education, business. Estonian was repressed and only accepted in the lowest spheres: entertainment (songs, folklore, games) and family life. Estonian was hardly taught and at a very poor level. With the liberation movement in the Baltic countries the people have tried to move the language up the ladder. The situation in Georgia, which I have tried to follow in some detail, is similar. Russian has been the dominating language for a long time since the annexation, but Georgian was accepted as a co-official language (Stalin was a Georgian). The priority of Russian can be seen in many details, however. One could write a doctorial thesis in Georgian or Russian, but if it was written in Georgian it had to be translated into Russian and the evaluation and grading was to be done in Moscow. Now Georgian theses are written in Georgian and graded in Georgia (Tbilisi). Administration and business is now carried out in Georgian and presently the language of transportation, e.g. railways and air-flights is being changed from Russian into Georgian. Courses in Georgian are offered to those who want to work in Georgia as knowledge in Georgian is now required in many work positions. I just learnt, however, that this proposal has been countered by a central Soviet (Moscow) law stating that nobody should be denied work because of lacking

< previous page

page_5

next page >

< previous page

page_6

next page > Page 6

knowledge of a language. The language controversies go on. It must be noted that the Russians in Estonia now complain about their minority situation and that the Georgians have minority peoples and languages within their country as well. The situation in the Caucasus region is and has always been very tense. Many regions think of their own minority situation, but are less inclined to appreciate the minority situation of still smaller languages within their territory. There are several papers at this conference which deal with the development of minority languages, e.g. the Aboriginal languages in Australia, which I think have been secretely ignored by the authorites in the hope that they will decay and die awaya hope which is being fulfilled with increasing speed. There are also interesting papers on the status of Frisian, Catalan, Karelian, Sami, Misiones-Swedish, Montagnais and the Swedish dialect of Dalecarliaa dialect seemingly decaying because of the new mobility of the Swedes which creates bidialectal marriages, according to Helgander. He does not subscribe to the common opinion that mass media (TV, Radio) is the main threat to the dialects. Trends in Bilingualism Research Many papers deal with the immersion programs, which have spread after the Canadian models. The people involved have much to talk about at conferences like this. I also note a number of big, often longitudinal, projects, e.g. Bilingualism at school. They will offer interesting corpus data for various studies and exchange among scholars. They too indicate the extension and expansion of the field of bilingualism research. There are several departments or academic programs in the world geared towards bilingualism studies, which also indicates the maturity of the field. The evaluation of linguistic competence is an eternal problem in bilingualism studies and testing methods are discussed in several papers. I note that different indicators are useful in different languages. In Welsh prepositions inflect for person and number and according to Wynn's paper children's mastery of these subtleties is a good indication of their competence in Welsh. At the same time, the children's reductions indicate the way Welsh may change. On the whole, linguistic competence is measured on several parameters these days taking into account communicative (pragmatic) behaviour, non-verbal signals, etc. in addition to the traditional examination of phonology, writing, morphology, lexicon, syntax and semantics. The development of a scientific field is generally characterised by the development of theories, laws and models to explain data. Some papers included such matters. I did not, however, note many papers on the psychological or neurophysiological basis of bilingualism at this conference. There are interesting problems concerning the storage and availability of the lexical and grammatical knowledge in bilinguals. There has also been interesting research into the neurophysiological basis of bilingualism and the loss of

< previous page

page_6

next page >

< previous page

page_7

next page > Page 7

competence in bilinguals suffering from aphasia or Alzheimer's dementia. Artigal outlined an interesting topological theory of comprehension and syntax in his lecture on the immersion program in Catalonia. It was based on practical teaching experience but can certainly be associated with current topological theories in cognitive science and linguistics. Conclusion Research in bilingualism may often be called 'action research'. It is often prompted by a desire to change something, e.g. to change the attitudes towards bilingual children, to prove that bilingualism is good for you, to improve methods in language teaching or testing, etc. The research may also have come about by an interest in a (one's own) minority language or a desire to prohibit the language from being marginalised or even killed by the dominant language. The researcher may want to show the risks a certain language or dialect runs if things are allowed to go on as they do. Many bilingualism researchers fight for threatened languages as biology researchers fight for threatened species of animals or plants. A conference like this is often filled with political ambitions in addition to scholarly ambitions. I think this makes conferences on bilingualismin all senses of the wordparticularly interesting.

< previous page

page_7

next page >

< previous page

page_ii

next page >

< previous page

page_9

next page > Page 9

Directions in Immersion Research Birgit Harley Ontario Institute for Studies in Education, 252 Bloor Street W., Toronto, Ontario, Canada M55 1V6 Abstract. This paper begins with an overview of the French immersion programs offered in Canada, and then focuses on some current issues and directions in French immersion research. The monitoring of learning outcomesin English mother tongue, French second language, and other school subjects taught in Frenchis seen as one major area of continuing importance for research. A second area of study is the diagnostic analysis of the learners' second language development. This kind of analysis is viewed as going hand in hand with the investigation of teaching procedures in immersion classrooms, a key topic of increasing research activity. Finally, teacher education is seen as a priority area for further research related to immersion programs. This paper focuses on the Canadian form of bilingual education known as French immersion. My purpose is to briefly highlight some existing lines of immersion research and to consider some further issues that need to be addressed. (For a more detailed discussion of current research issues, see Lapkin, Swain & Shapson, 1990). Definition and Documentation French immersion in the Canadian context is a program in which majority English-speaking students receive a substantial portion of their schooling in French, their second language. It is based on the principle that in order to learn a second language well it is essential to have plenty of opportunity to use it for non-trivial communicative purposes. French immersion is an optional kind of program offered in the context of otherwise English-speaking school systems. Generally speaking, a program is considered to be immersion if 50% or more of each school day takes place in French: that is, students take at least half of their curriculum (such as history, maths, music, physical education) in the second language. The goals of French immersion are to produce much improved levels of French proficiency, without detriment to students' English language development, to their academic achievement in other subjects, or to their social-psychological well-

< previous page

page_9

next page >

< previous page

page_10

next page > Page 10

being, and at the same time, to develop positive attitudes and cultural sensitivity towards French-speaking Canadians. French immersion is now a quarter century old, having begun as a small parent-initiated experiment in 1965, involving 26 kindergarten children in a Montreal school. In response to public demand, immersion programs have spread to every province in Canada and according to the latest statistics, there were over 256,000 students enrolled in immersion programs in the 1989/90 school year, representing a 7% rise over the previous year (Commissioner of Official Languages, 1990). The original immersion program in Montreal set the pattern for what is still the most widespread form of French immersion; it is known as 'early total' French immersion. Children in this type of program receive 100% of their education in French for the first few years in school, usually starting in kindergarten. Around their third or fourth school year, English language arts is introduced for about an hour a day, and by the time they are 1011 years old, the children are getting about half their education in English and half still in French. At the secondary school level, the students are offered a post-immersion follow-up program, with a period of French language arts along with one or two subjects taught in French each year. Among other formats for French immersion there is: 'middle' immersion, starting with from 50% to 100% in grade 4 or 5, continuing with at least 50% up to grade 8, and with a secondary school follow-up program as for early immersion; then there is 'late' immersion, starting at grade 6, 7 or 8 with approximately 80% in French for one or two years, again with follow-up at the secondary level. Middle and/or late immersion programs may be offered side by side with early immersion in the same school jurisdictions. A distinction between 'partial' and 'total' immersion is also sometimes made, to reflect the proportion of time devoted to French in the initial years of a program. Note that program documentation is an apparently mundane, but vitally important, continuing aspect of immersion research that is essential to the interpretation of learning outcomes and to educational policy and planning. This includes the recording of enrolment patterns and program dropout rates, tracking of the educational paths of students from elementary to secondary school, basic program information (e.g. proportion of time in the second language for each year of the program, subjects taught in French and English), staffing patterns, characteristics of the student population, and so on. The task of recording this kind of information has fallen mainly on local school boards, although a valuable role undertaken by a national organisation of parents (Canadian Parents for French) has been to collate information on French second language programs across Canada into an annual registry. In addition to the recording of the hard facts, a certain amount of research has probed the perceptions of those whose lives are directly or indirectly affected by immersion programswhether school board personnel, teachers,

< previous page

page_10

next page >

< previous page

page_11

next page > Page 11

students, or parents (e.g. Burns, 1987). These questionnaire data provide a valuable complement to the facts and figures. Learning outcomes The measurement of learning outcomes in French immersion has long been a major focus of the research (e.g. Swain & Lapkin, 1982; Genesee, 1987). Learning outcomesEnglish skills In addition to the obvious question of how much French immersion students learn, parents and educators have right from the beginning been concerned with monitoring the development of students' English skills, and with ensuring that they keep up in subject matter taught to them in French. The issue of English skills has naturally received most attention in the context of early total immersion where literacy skills are introduced via the second languagei.e. French, and where English instruction does not occur for the first two or more years. The research has shown quite consistently that, in the first few grades, immersion classes receive lower average scores than English comparison groups on some standardised measures of English literacy skills. Soon after the introduction of English language arts, however, the immersion classes obtain equivalent results and in some instances have outperformed the comparison groups at grades 5 and 6 (Swain & Lapkin, 1982; Harley, Hart & Lapkin, 1986). Possible benefits to first language skills is one area that merits further study, including probing to determine what kind of cognitive advantages may be promoted by participation in an immersion program. Notwithstanding the favourable results for students who stay in the early immersion program, questions have been raised about the incidence of reading problems among children who transfer out in the early grades. In a study of grade 13 children who transferred out of immersion programs in two Ottawa-region boards (Morrison et al., 1986), teachers identified reading skills as a factor in 64% and 80% of cases respectively. There is currently no evidence that the reading problems of these children were caused by the fact that initial literacy skills were introduced in the second language. Had they been in the regular program they might have had the same kinds of problems. And in some cases, the unavailability of remedial help appears to have contributed to parents' decisions to withdraw their child. The issue is one that requires more research, however. The relative merits of alternative ways of providing early reading instruction in immersion need to be examined, including the option of beginning reading instruction in Englishan option which has in fact been adopted in one province (Manitoba).

< previous page

page_11

next page >

< previous page

page_12

next page > Page 12

The suitability of immersion for all children The issue of beginning literacy leads to the question of whether French immersion is an experience that is equally suitable for all children. We know that, generally speaking, immersion programs draw considerably more enrolment from higher socioeconomic groups than is representative of the school population as a whole. What are the learning outcomes for children with low socioeconomic backgrounds, for children who are of below average academic ability, or for children who have specific learning disabilities? Based on a detailed examination of the available data on these issues, Genesee (1987) concludes that such students have reached the same levels of first language development and academic achievement as similar students in the regular English program while at the same time learning more French. A key priority for further research is to determine what kinds of remedial services are most helpful for those immersion students that need them. Subject matter taught in French With respect to the broader issue of academic achievement, the research indicates that students taught subjects in French have generally done as well as students taught the same subjects in English, with the most consistent results coming from early total immersion students (Swain & Lapkin, 1982; Genesee, 1987). So far these findings relate mainly to the elementary school level, and much less is known about outcomes of subject-matter courses in the secondary school follow-up program. Here it seems that the question of learning outcomes is blurred by issues that have emerged with respect to program delivery. At the secondary level, teachers in the follow-up program need to be both specialists in the relevant subject and proficient in French, a combination that is not readily available for a wide range of subject matter options within a single school. A recent report by one Ottawa-area school board with a substantial immersion student population, estimates that between 50% and 75% of students have transferred out of its immersion follow-up program by the final year of school (Carleton Board of Education, 1989). As contributing factors in this phenomenon the report cites program delivery features such as the dispersion of students across secondary schools rather than immersion centres, as well as pressure on senior students to get high marks for university entrance and the perception that this would be more difficult in a French course. While this board's transfer figures may be unusually high, different models of program delivery at the secondary level clearly need to be examined. If it is the case that in voting with their feet students are prejudging the question of learning outcomes, it may be that by the senior grades at least, subject matter options in French should not compete directly with courses available in English. One possibility to be explored is that the program at this level should deliberately concentrate on special interest topics in French literature and culture, drawing mainly on teachers who are specialists in

< previous page

page_12

next page >

< previous page

page_13

next page > Page 13

teaching French as a subject. According to one report (Morrison & Bonyun, 1984), for example, French literature and reading was most likely to be cited by senior secondary students in the Ottawa area as a particularly good aspect of the immersion follow-up program they had taken. Learning outcomesFrench skills Research directed to the issue of how much French the immersion students learn has naturally been a major preoccupation of second language educators. It has involved not just making the necessary assessments but also the intensive development of appropriate second language tests to reflect the communicative goals of immersion programs. As in the investigation of other kinds of learning outcomes, the research on the French language outcomes of immersion students has been largely comparative in nature. In comparisons with native speakers, for example, we learn that early immersion students develop excellent listening and reading skills, in some cases achieving scores on global listening comprehension tasks that have been equivalent to those of French speaking comparison groups (Swain & Lapkin, 1982, 1986). As early as grade 6, early immersion students have also manifested strong discourse skills in French, obtaining coherence ratings on open-ended production tasks that approach those of native French speakers at the same age (Harley et al., 1990). On grammatical and sociolinguistic measures of their productive skills, however, these immersion students have clearly not approached native speaker levels (Harley et al., 1990) and this finding has continued to apply at the secondary school level (Swain & Lapkin, 1986, 1990). Comparisons of the French proficiency of early and late immersion students in secondary level follow-up programs reveal a general tendency for late immersion students to catch up in reading and writing skills, with early immersion students usually maintaining some advantage in listening and in spoken French (Swain & Lapkin, 1986; Wesche et al., 1990). With respect to middle immersion, a more recently established option in most locations, there are so far no comparative data beyond grade 8. At this grade level a recent study (Hart, Lapkin & Swain, 1988) shows mixed results, with early immersion students outperforming those in a 50% middle immersion program that had started in grade 5 but not doing better than a group of students who had started with 100% in grade 4. Comparisons between early and middle immersion will continue to be needed at higher grade levels. Although there is a consensus that native-like competence is an unrealistic goal for students whose exposure to the second language is restricted mainly to the classroom context, most educators seem to agree that, particularly in early immersion which has the greatest exposure to French, the grammatical and sociolinguistic skills of students leave room for improvement. This then raises the question of what adjustments to the immersion program proper and/or the follow-up program might lead to the enhancement of such

< previous page

page_13

next page >

< previous page

page_14

next page > Page 14

competencies. It is a question which has already led in two complementary research directions: first, the diagnostic analysis of immersion students' second language development, and secondly the investigation of classroom processes in immersion. I will return to these topics shortly, but first I want briefly to raise the issue of the long-term learning outcomes of the various programsthat is, what happens to students' French once they graduate from secondary school? One line of research being pursued (Hart et al., 1989; Wesche et al., 1990) is the tracking of immersion graduates, partly to determine what use they make of their French in the wider environment, and also to continue monitoring the language skills of those with different immersion program backgrounds. Further research has investigated the relevance of immersion graduates' bilingual skills to jobs available in the workplace (Lapkin & Swain, 1988). Another prospective area of research is the investigation of good language retention strategies that immersion graduates can draw on to sustain this 'natural resource' of bilingualism that has been so enthusiastically nurtured at school. Second Language Development Turning now from outcomes to the question of the process of second language acquisition in immersion, this section provides a brief overview of the research and the kinds of conclusions that have come out of it. The main focus has been on the diagnostic examination of students' oral interlanguage productiontypically in an interview setting (e.g. Harley & Swain, 1984), but there have also been a number of studies involving written compositions (Harley et al., 1990; Swain & Lapkin, 1990). Topics of investigation in this type of research have included, for example: the development of the verb system, the acquisition of noun gender, the use of tu and vous as sociolinguistic markers, and the development of verb vocabulary. An important aspect of the research has been the use of native-speaker comparison data, which serves as a necessary check on the analyst's intuitions about what French native speakers of the same age would characteristically do under similar task conditions. In various areas of their language production, differences between the performance of the learners and the native speakers have been found, but there have also been some surprises: for example, that native French speakers in grade 6 have just as much trouble with the correct written forms of some homophonous verb inflections as do early immersion students at this grade level (Harley et al., 1990). What are some of the diagnostically relevant conclusions that can be derived from this research? Those familiar with the second language acquisition literature will note some obvious parallels with findings from non-classroom settings. 1. First, it is clear that although immersion students demonstrate excellent

< previous page

page_14

next page >

< previous page

page_15

next page > Page 15

understanding of language in context, this cannot be taken as firm evidence that they have correctly identified all the formmeaning connections involved. They become adept at inferring global meaning, using clues in the surrounding discourse or in the external situation. One example of the discrepancy between global comprehension and oral production is their use of conditional verb forms. Grade 1 early immersion students are readily able to comprehend conditional sentences and can translate them into English, but years later in grade 10, we find some students still have trouble with conditionals in their oral production (Harley & Swain, 1984). 2. A communicatively useful strategy that at least some beginning learners in immersion adopt is to extract some globally meaningful phrases or chunks of the second language from the stream of oral language they receive and go on to produce them in appropriate circumstances, without necessarily being able to segment them into individual words. Such unsegmented chunks may go unnoticed by teachers and can persist for several years. 3. In the area of vocabulary development, immersion students tend to favour high coverage items of general utility and only gradually incorporate more specific terms into their production. A general verb such as aller or faire, for example, may be stretched to cover various contexts where native speakers regularly use more specific terms. 4. Immersion students very soon become experts at other kinds of communication strategies. When at a loss for a word, instead of abandoning their message, they tend typically to find their way around the problem, whether by mime, gesture, paraphrase, subsitution of an English word, or a direct request for help (Comment tu dis 'look around'?). 5. In any given area of the French language system, learners in an immersion context do not normally take one giant step from no knowledge to target native-like knowledge of French. The fact that they can produce a particular form in French does not necessarily imply that they have mastered its distribution in the target language nor that they have fully comprehended its functional range. They do not first fully comprehend and only then start producing. Both aspects are developing simultaneously. 6. The learners' mother tongue and complexities within the second language are two sources which together appear to be the combined basis for some typical interlanguage patterns that persist in the classroom setting. It is important to note, however, that mother tongue influence is not always negative (Ringbom, 1987). In the Hart et al. (1988) study of early and middle immersion, for example, it was found that for students who used another Romance language in the home there was an advantage in French proficiency. With respect to future research needs in the area of L2 development,

< previous page

page_15

next page >

< previous page

page_16

next page > Page 16

there are certainly gaps in the research that need to be filled (for example in the area of phonological development and fluency), but a priority at this point is to disseminate existing findings in a digestible form to teachers and curriculum developers. This is a project that I am currently working on with a colleague, Raymond Mougeon. Classroom Processes Another line of research, spurred on by the findings in the area of learning outcomes and second language development, is the investigation of classroom processes in French immersion. Observations conducted in early immersion classrooms at grades 3 and 6 suggest that the general discrepancy found between the receptive and productive skills of immersion students reflects the typical demands of a classroom setting where the main emphasis is on comprehension of content and where individual opportunities for oral production tend to take the form of short answers requiring little syntactic elaboration (Swain & Carroll, 1987). Classroom research has led to the conclusion that there is a complementary relationship between an experiential approach to language teaching with a focus on the message and an analytic approach with a focus on the language code (Allen et al., 1990). This complementary role does not mean simply offering experiential and analytic activities side by side, however. Classroom observation in immersion suggests that there are dangers in isolating one from the other. Grammar lessons focused mainly on the teaching of grammatical form may fail to promote meaningful use of the relevant distinctions in message-oriented communication, while use of the language as a vehicle for subject matter content does not guarantee that students will have much opportunity to observe or express specific form-function relationships. The conclusion that analytic and experiential approaches are complementary still leaves open the crucial question of how these two approaches are most effectively combined in the immersion context, so that content does not lose out to an overemphasis on language development, and vice versa, that language development is not subordinated to an exclusively content-oriented emphasis. Two experimental studies have already begun to address the issue of how an analytic approach can be effectively integrated with content in grades 6 and 7 of early immersion programs (Harley, 1989; Day & Shapson, 1990). These studies focus on developing students' functional command of grammatical distinctions that previous acquisition research had diagnosed as persistent problem areas and that classroom observations had shown were rarely occurring in the classroom context. Building on these initial experiments, more longterm studies of the integration of experiential and analytic approaches, conducted in close co-operation with teachers, are clearly indicated for all levels of immersion, including the follow-up program in secondary school.

< previous page

page_16

next page >

< previous page

page_17

next page > Page 17

Teacher Education Research in the domain of teacher education has so far consisted mainly of surveys designed to assess the existing situation and elicit the views of practising immersion teachers, teacher educators and provincial ministry staff (e.g. Tardif, 1984; Wilton et al., 1984; Frisson-Rickson & Rebuffot, 1986). The various surveys show a need not only for more preservice programs geared to the special needs of immersion teachers, but also for more systematic longterm inservice support. Given the rapid expansion of immersion programs, a major current issue is how to keep up with the need for qualified teachers who have adequate linguistic skills in French as well as appropriate pedagogical trainingi.e. training which is focused on students' linguistic and academic development and on their individual learning needs. The findings from surveys which have revealed teachers' concerns about how to balance the different needs of their students, serve as important input into the further development of immersion teacher education programs. Numerous other issues arise concerning the content of preservice education, its mode of delivery, what the linguistic qualifications of an immersion teacher should be, what kinds of continuing professional development are most effective, and what the relationship should be between teacher education for immersion teachers and for those who teach in the regular core French program. (For further discussion, see Lapkin et al., 1990.) As new developments take place it will be important for them to be associated with adequate documentation and research as to their utility and practicality. One development of interest in the area of preservice education is a proposed 3-year B.A. program in second language teaching at the University of Ottawa (Wesche, personal communication) which will include courses in language, literature, linguistics, psychology, and second language teaching. Prospective immersion teachers would also take courses in subject matter areas. Another development is the production of video packages which have the advantage in a country the size of Canada of being easily transported to where teachers are. One of these consists of unrehearsed, real-life examples of immersion classroom teaching in grades 6 and 8 (Argue et al., 1990). Designed for either preservice or inservice education, its purpose is to stimulate discussion and promote reflection by teachers on their own practices. Earlier, the dissemination of diagnostic information about immersion students' language development was mentioned. This kind of information is currently being prepared in the form of a handbook which could be used as resource material in a preservice course or for reference by individual teachers. These provide simply a few examples of components that will have to prove their efficacy in a broader framework for immersion teacher education which has yet to be generally agreed upon.

< previous page

page_17

next page >

< previous page

page_18

next page > Page 18

Conclusion This overview of some directions in French immersion research has ranged from outcomes to classroom processes and finally to teacher education, though this should not be regarded as a linear progression. It is important to recognise the interlocking nature of the issues that arise in each of these domains. Researchers need to keep working on many fronts to provide information that is helpful to teachers, policy makers and of course to students themselves. References Allen, P., Swain, M., Harley, B. and Cummins, J. (1990) Aspects of classroom treatment: Toward a more comprehensive view of second language education. In B. Harley et al. (eds). Argue, V., Howard, J., Lapkin, S. and Swain, M. (1990) Teaching French in French Immersion: An Overview at Grades 6 to 8. Toronto: Modern Language Centre, OISE. Burns, G.E. (1987) Planning, doing and coping with FSL change. The Canadian Modern Language Review 44, 4766. Carleton Board of Education (1989) Immersion/Regular Program Study: Research and Development. Nepean, Ontario: Carleton Board of Education. Commissioner of Official Languages (1990) Annual Report 1989. Ottawa: Office of the Commissioner of Official Languages. Day, E.M. and Shapson, S.M. (1990) Integrating formal and functional approaches in language teaching in French immersion: An experimental study. Paper presented at AILA 90, Cassandra-Halkidiki, Greece, April 1990. Frisson-Rickson, F. and Rebuffot, J. (1986) La formation et le perfectionnement des professeurs en immersion: pour des critèes nationaux. Ottawa: Association canadienne des professeurs d'immersion. Genesee, F. (1987) Learning Through Two Languages. Cambridge, MA: Newbury House. Harley, B. (1989) Functional grammar in French immersion: A classroom experiment. Applied Linguistics 10, 33159. Harley, B., Allen, P., Cummins, J. and Swain, M. (eds) (1990) The Development of Second Language Proficiency. New York: Cambridge University Press. Harley, B., Hart, D. and Lapkin, S. (1986) The effects of early bilingual schooling on first language skills. Applied Psycholinguistics 7, 295322. Harley, B. and Swain, M. (1984) The interlanguage of immersion students and its implications for second language teaching. In A. Davies, C. Criper and A.P.R. Howatt (eds) Interlanguage. Edinburgh: Edinburgh University Press. Hart, D., Lapkin, S., Howard, J. and King, M.L. (1989) French immersion at the secondary/postsecondary interface: Final report of the follow-up survey of the 1988 secondary school graduates. Toronto: Modern Language Centre, OISE (mimeo). Hart, D., Lapkin, S. and Swain, M. (1988) Early and Middle Immersion Programs: Linguistic Outcomes and Social Character. Toronto: Modern Language Centre, OISE. Lapkin, S. and Swain, M. (1988) Bilingual job vacancies survey: An exploratory study. Toronto: Modern Language Centre, OISE. Lapkin, S. and Swain, M. with Shapson, S. (1990) French immersion research agenda for the 90s. The Canadian Modern Language Review 46, 63874. Morrison, F. and Bonyun, R. (1984) Evaluation of the Second Language Learning (French) Programs in the Schools of the Ottawa and Carleton Boards of Education, Volume 3: Surveys of Bilingual-Program Graduates, 1984. Toronto: Ministry of Education, Ontario. Morrison, F., Pawley, C., Bonyun, R. and Unitt, J. (1986) Aspects of French Immersion at the Primary and Secondary Levels: Evaluation of the Second Language Learning (French) Pro-

< previous page

page_18

next page >

< previous page

page_19

next page > Page 19

grams in the Schools of the Ottawa and Carleton Boards of Education. Toronto: Ministry of Education, Ontario. Ringbom, H. (1987) The Role of the First Language in Foreign Language Learning. Clevedon, England: Multilingual Matters. Swain, M. and Carroll, S. (1987) The immersion observation study. In B. Harley, P. Allen, J. Cummins and M. Swain. The Development of Bilingual Proficiency, Vol. 2. Toronto: Modern Language Centre, OISE. Swain, M. and Lapkin, S. (1982) Evaluating Bilingual Education: A Canadian Case Study. Clevedon, England: Multilingual Matters. (1986) Immersion French in secondary schools; 'The goods' and 'the bads'. Contact 5/3, 29. (1990) Aspects of the sociolinguistic performance of early and late immersion students. In R. Scarcella, E. Andersen and S.D. Krashen (eds) On the Development of Communicative Competence in a Second Language. New York: Newbury House. Tardif, C. (1984) La formation des enseignants en situation d'immersion. The Canadian Modern Language Review 41, 36575. Wesche, M.B., Morrison, F., Ready, D. and Pawley, C. (1990) French immersion: Post-secondary consequences for individuals and universities. The Canadian Modern Language Review 46, 43051. Wilton, F., Obadia, A., Roy, R., Saunders, A.B. and Tafler, R. (1984) National Study of French Immersion Teacher Training. Ottawa: Canadian Association of Immersion Teachers.

< previous page

page_19

next page >

< previous page

page_ii

next page >

< previous page

page_21

next page > Page 21

The Catalan Immersion Program: the Joint Creation of Shared Indexical Territory Josep Maria Artigal Rambla del Prat, 5, 08012 Barcelona, Spain Abstract. This paper begins with a brief overview of the sociolinguistic situation in which the Catalan immersion program is being implemented. It then goes on to focus on certain issues related to the main feature that distinguishes this program from othersthe low socioeconomic background of the childrenand reflects on current theories about discontinuity between the communicative context-bound language used in the home and the decontextualised, formal language of school. The remainder of the text is devoted to an attempt to analyse the first uses of the new language (Ln) as what will be termed a 'shared indexical territory', and to hypothesise that this analytical option enables us to postulate the existence of a semiotic continuity between these first uses of the Ln and subsequent, linguistically more highly developed texts in the same language. Finally it is argued that this continuity operates as a recognition mechanism within the process of acquisition of the new, decontextualised school language. The Context of the Catalan Immersion Program The Catalan language is spoken by close to 7 million people and is thus the largest stateless language in western Europe. It is the native language of an area known as the Catalan countries, comprising the northeastern corner of the Iberian peninsula (Catalonia, Valencia and the Balearic Islands) and a small part of the Mediterranean side of the French Pyrenees (the Roussillon). Catalonia gradually emerged as a nation with its political, cultural and linguistic personality in the course of a historical process which began a thousand years ago. Though in 1659 the Catalan countries were divided up politically between Spain and France, and in 1714 the Spanish part was deprived by military force of its national rights and political institutions, the social status of the Catalan language within its own territory continued to be high until 50 years ago. The brutal repression of the Catalan

< previous page

page_21

next page >

< previous page

page_22

next page > Page 22

signs of identity, notably the language, perpetrated by the dictatorship of General Franco (19391975), and the immigration of large numbers of monolingual Spanish speakers from other parts of Spain in the 1950s and 1960s, caused Catalan to be reduced to a position of inferiority while Spanish became the dominant language. The approval of the new Spanish democratic Constitution in 1978, and the later subsequent passing of Statutes of Self-Government for Catalonia, Valencia and the Balearic Islands, marked the beginning of a new and still unfinished process aimed at restoring the social status of Catalan as the country's own language. In this context, the immersion program is playing a decisive role. The Catalan immersion program for children whose mother tongue is Spanish was first offered as a legal option in the 1983/84 school year, though the earliest experiences date back to 1978/79. Nearly 100,000 children between the ages of 3 and 14 are presently taking part in it. The immersion program in Catalonia fulfils the requisites for a successful home-school language switch: the family language and culture of the children attending the program are dominant; their attitudes and motivation towards the school language, and the optional character of the program, are guaranteed; and the pedagogical quality of the teaching in immersion classes is generally acknowledged. One feature, however, sets this immersion programe apart from others: since the majority of the pupils are children of immigrants referred to above, who moved to Catalonia in the 1950s and 1960s, they usually come from low socioeconomic backgrounds and consequently many enter school with a poor command of their home language. In the rest of the present paper I will be concentrating on certain issues related to this aspect of the immersion program in Catalonia. Some preliminary considerations will be necessary to enable me to do this. First, I will discuss language acquisition as a recognition process. Then I will refer to linguistic signs as a vehicle of multifunctionality and especially to the indexical nature of some of these functions. Thirdly, I will argue that any type of text can be viewed as an indexical territory shared by the different speakers. Finally on the basis of this reasoning, I will suggest that there is semiotic continuity between the first uses of the new language, built up through context-bound routines, and subsequent, linguistically more highly developed texts, and that this continuity operates as a recognition mechanism in the acquisition process of the new language. I am aware that the procedure used is a lengthy one, owing to the large amount of preliminary information to be introduced, but it is my hope that it may prove useful. Language Acquisition as a Recognition Process During the 1970s and 1980s, various authors maintained that the language acquisition process cannot be explained without the existence of some type of background affording access to linguistic input, that is, of a set of procedures enabling learners to recognise this input as such. Thus most

< previous page

page_22

next page >

< previous page

page_23

next page > Page 23

psycholinguists now agree that, when children come into contact with language, they do not arrive 'with empty hands'. Moreover, despite the attempt to demonstrate that this previous knowledge is socially built up and not innate, there has been basic agreement about the need for some type of point of departure enabling the learner to recognise meaningfully the language offered by the environment. Initially, during the 1970s and 1980s, this background was considered to be mainly semantic and pragmatic (Brown, 1973; Bruner, 1975, 1982; Greenfield & Smith, 1976; Bates, 1976). Those who defended this position believed that language had a semantic and/or pragmatic base and hypothesised that the superficial structure of the utterances exchanged by the speakers must be connected in some way to the topics, tasks or goals the speakers convey through the aforementioned forms. Accordingly, children are believed to acquire language by recognising in the superficial structure of the input they are offered the way in which their language identifies different semantic and pragmatic aspects of reality which they have previously learned to master. In fact, several of the theoretical arguments behind communicative and functional approaches to second language acquisition are partially founded on this psycholinguistic explanation of language acquisition as a recognition process. However, as various authors later pointed out (Hoff-Ginsberg & Shatz, 1982; Gleitman & Wanner, 1982), the theoretical models of the 1970s and early 1980s were built on the partly fallacious assumption that there is a complete and univocal relationship between the pragmatic and semantic knowledge the learner possesses at a given time and the syntactical procedures he or she will later have to acquire, i.e. that children are able to relate the underlying syntactical procedures in the input, one by one, with the set of pragmatic and semantic procedures which the same environment has previously helped them to build up. This assumption is incompatible with a notion that is gradually gaining acceptance: that language is built up in way that is not totally isomorphic in terms of the meaning and functionality of what it conveys, that, in some ways at least, it operates by means of a set of regularities that are independent of the regularities that occur in the world to which it refers (Silverstein, 1976, 1985; Gleitman & Wanner, 1982). Without denying the validity of this pragmatic and semantic background, the current view is that it is insufficient, that some aspects of the input offered to learners would be unrecognisable solely on the basis of the type of 'previous background' that was formerly proposed. This lack of isomorphism between the key used for recognition purposes and the thing to be recognised has certain implications for immersion programs. Cummins' threshold level hypothesis (1979) is a clear example. When it comes to explaining access to formal, decontextualised school language from the use of a home language that is still deeply enbedded in a communicative context, a qualitative gap appears that is neither more nor less than a form of the discontinuity already defined. In other words, the

< previous page

page_23

next page >

< previous page

page_24

next page > Page 24

threshold hypothesis can partly be explained by the impossibility of using the 'Basic Interpersonal Communicative Skills' (Cummins, 1980) developed in the family language as a key to the formal, decontextualised procedures of the 'Cognitive Academic Language Proficiency' (Cummins, 1980) that is indispensable at school. As pointed out at the beginning, this discontinuity takes on considerable proportions in an immersion program like Catalan, one in which the majority of the children are from low socioeconomic backgrounds and in which it does not seem possible to guarantee that they enter school with the linguistic-cognitive skills required by the formal, decontextualised use of language in which instruction will be conveyed. The Indexical Value of Linguistic Signs The coming paragraphs are an attempt to consider this last problem. I will be endeavouring to establish ways of creating a 'fil conducteur', or connecting thread, which is neither semantic nor pragmatic but semiotic in nature, and is capable of operating as a recognition mechanism in the new school language acquisition process. The first stage in the development of my argument will consist of discussing certain reflections made by Peirce on the subject of signs. For Peirce, signs are not stable forms but functions conveyed through forms. The functions that can be expressed by a sign are iconical, indexical and symbolic, though for my present purposes I will mainly be concerned with the second. 'An index' writes Peirce (1987a: 104), 'is a sign which represents the corresponding object by being truly connected to it'. Thus an index is a sign whose existence depends on a real relationship of co-presence with its object. Using an example given by Lyons (1980: 102), it can be said that the relationship between 'smoke' and 'fire' is in some way indexical. Certainly the connection between 'smoke' and 'fire' is such that 'smoke' would cease to exist as soon as the 'fire' was put out. The indexical aspects of 'smoke' depend on a contiguous relationship with its object, and would immediately stop pointing to, or indicating, its object on elimination of the latter. Thus the value of an index depends on a set of co-presences. An index, therefore, can introduce nothing that is not already given in the context which it is pointing to. An index is like an arrow calling attention to something that is already present. An index remains connected to the precise, specific, unrepeatable facets of its object, focuses attention on what is already given. As Peirce suggests (1987b: C. P. 2318) 'the index is the common environment between interlocutors'. Now, as Silverstein (1976, 1985, 1987) has shown, linguistics has traditionally treated linguistic signs as symbols, that is, as arbitrary, discrete elements capable of operating independently from the set of (con)textual co-presences, as elements governed by some type of law or generalisation. But not all linguistic signs fulfil a symbolic or purely symbolic function. When we use a language in order to do things with other people, we also use many

< previous page

page_24

next page >

< previous page

page_25

next page > Page 25

indexes, i.e. signs which perform multiple functions, some of which are indexical. In other words, we use linguistic forms to indicate elements or aspects which can be seen as already shared by the speakers at the same time the signs are made explicit, we use forms which, as in the case of smoke, require the co-presence of their object in order to operate as signs. Examples including Jakobson's studies (1975) on shifters; Lyons' reflection (1980) on anaphora as intralinguistic indexical relationships; Benveniste's (1966) research into the indexical value of first and second person pronouns in enunciative situations; or Silverstein (1976) and Hickmann's (1985) hypotheses about the importance of indexicality as a metapragmatic function. But as Peirce has already observed, the notion of indexicality in language can be carried even further. Peirce considers that any use of a linguistic symbol depends on, presupposes, knowledge of some grammar, of some pattern of interpretation, which the speakers must share. Consequently, it seems possible to maintain that the use of any symbolic component of language will at the same time involve the use of some indexical component, at the least in so far as the use of the symbol depends on the copresence of someone who shares knowledge of the 'law' upon which the symbol rests, but also in so far as such use consists, at any rate with reference to certain procedures, of making places, spaces, or functions in a (con)text which must of necessity be shared with other speakers. In this second sense, any use of language will need indexical procedures in order to mark 'locations in a (mental) space' shared by interlocutors. The symbolic nature of this space, or (con)text, and the formal procedures for pointing things out within it, does not invalidate the fact that 'pointing out that which is already shared by the interlocutors' is an indexical function. I will give a very simple example of this. Linguistic studies have shown that the meaning of the sentence (1) (1) John loves Mary does not depend solely on the identity of 'John' and 'Mary'. It is obvious, in other words, that there is no reason why sentences (1) and (2) (1) John loves Mary (2) Mary loves John should mean the same, even if 'John' and 'Mary' are the same people in both cases, and the same co-ordinates of time and space apply. Or, to put it another way, the meaning of (1) and (2) is not merely the sum of [the referent of 'John'] + [the referent of 'Mary'] + [the referent of 'loves']; because if this were so, (1) and (2) would indeed mean the same. Let us now briefly examine the Finnish sentence (3) Juhania Maria rakastaa. To anyone unfamiliar with Finnish, it is not enough to know that 'rakastaa' means 'loves' and to know the referents of 'Juhania' and 'Maria' in order to understand whether (3) means the same as (1) or (2). The important point is that the segments of text

< previous page

page_25

next page >

< previous page

page_26

next page > Page 26

Mary/Maria()', and 'John/Juhani (a)', point in (2) and (3) to particular locations in a formal space which must necessarily be shared by English and Finnish speakers respectively in order to understand that (3) Juhania Maria rakastaa [means] (2) Mary loves John, whereas (4) Juhani Mariaa rakastaa [means] (1) John loves Mary. The fact that the formal procedure used in (1) and (2) is positional while in (3) and (4) it is inflectional is of little importance: in each case indexical procedures appear which consist of pointing arrows towards a mental space which is already co-present for the speakers before being 'filled in' and without which the meaning of (1) (2) (3) and (4) could not be built up. The Shared Indexical Territory To provide further support to my reasoning, I will now refer to the ideas of language as a game and as a shared territory as proposed respectively by Wittgenstein and Bakhtine. Let us think for a moment about the game of soccer. It is a well known fact that in order to score a goal in soccer you have to send the ball between the goalposts. But a little more thought will show us that this is not necessarily so. Like any other game, soccer is a territory with boundaries. The visible limits of the territory of soccer are the lines round the edge of the field but other limits are the times at which the referee blows his whistle to indicate the start and the finish of the match. The important point is that sending the ball between the goalposts outside the territory is notdoes not meana goal. If someone sends the ball into the goal before the match begins, or after it has ended, then this is notdoes not meana goal. In Wittgenstein's view, the pieces of a gameor those of a languagehave meaning as a function of the actual game in which they are used and of the place where they are put in the territory of this game. In other words, if a goal outside the territory is not a goal, then a piece (for example 'Johania') which cannot be put (for example by inflectional means) into a territory (such as (3)) will be void of meaning. The meaning of a texta word, a sentence, a discoursedoes not consist solely, or even fundamentally, of what is denoted in the text, nor does it derive from the existence of the respective entities represented by such denotations. A goal is not simply 'sending the ball between the goalposts' and nor, as I attempted to demonstrate above, does the sentence (1) mean only the sum of its referents. As Wittgenstein (1988: 1, 3031) argues, 'the ostensible definition explains the use (the meaning) of the word when it is already clear what role the word must generally play in the language (...)When someone is shown the king in chess and is told ''this is the king", he is being offered no explanation as to the use of the piece (...) We will only say that he is shown the use if the location is already prepared'. Obviously all text is 'anchored' in some way in the particular situation in

< previous page

page_26

next page >

< previous page

page_27

next page > Page 27

which it is enunciated and, in this sense, it seems possible to state that all text has an indexical relationship with its production context. Thus, coming back to sentence (1) John loves Mary it is clear that the meaning of (1) will be different according to whether it is a statement that Paul makes to Susan who is in love with John, or whether it is a simply sentence used as an example in the present text. But furthermore, and without contradicting what has been said so far, (1) also has meaning on account of the fact that the person who makes the statement and the person to whom it is made have a shared semiotic territory in which references to particular locations may be made, these locations having already existed and been known to the speakers before either opened his or her mouth. What I am suggesting is that showing the location to be occupied by a piece in a shared territory is a semiotical indexical procedure that consists of 'pointing an arrow towards a co-presence with the interlocutor'. And this holds true, regardless of the different formal ways in which each language attains the same objective. It is my belief, then, that the signs in a text require the existence of a territory shared by the speakers and that their semiotic relationship with this mutually recognised territory is indexical in nature. Whatever the level of symbolisation the sign has reached, it will never lose this indexical relationship with the territory in which it is used. Consequently, just as each variation in the fire brings about a change in the smoke, and each time the fire is put out the smoke vanishes, when the semiotic territory is changed, the value (meaning) of the sign will be modified, and when the semiotic territory is lacking, so is the sign (meaning). As Wittgenstein argues (1988), every time a text is 'said', the speaker 'shows' a territory shared with his or her interlocutor, since otherwise the text would be void of meaning and nothing would have been said. Consequently, all use of language requires, not only obvious semiotic en-and decodable relationships, capable of operating in a decontextualised manner and conveyed by means of discrete and arbitrary, i.e. symbolic, signs, but also indexical semiotic relationships, which are (con)textually bound, which depend on co-presences and which direct attention towards those non linguistic and/or linguistic[!] elements the speakers already share. The indexical nature of linguistic signs applies not only to their propositional roleas in the case of sentences (1), (2), (3) and (4) abovebut also to their value as occurrences in a discourse. A good example would be the various possible occurrences of (1) (1) John loves Mary in an imaginary novel. For instance, (1) will obviously have a different meaning according to whether it occurs at the beginning of the novelif 'John loves Mary' then problems may well lie in store for themor at the end of the same novelafter countless ups and downs, finally, 'John loves Mary'.

< previous page

page_27

next page >

< previous page

page_28

next page > Page 28

Now, what appeared to be the most symbolic aspect of language, its formal structure, may be understood, at least in part, as an indexical territory. Thus all use of language, whatever its level of formalisation and its degree of contextual 'anchorage', will always imply some indexical-type procedures. In other words, texts, whether oral or written, will themselves become contexts and as contexts shared between the speakers may be indexed. In short, the use of any symbolic component of language will always of necessity involve the use of some mechanism of internal re-indexicalisation (Peirce), will imply a territory shared with others (Bakhtine), will constitute itself into a social rule of game (Wittgenstein). The Joint Construction of a Semiotic Connecting Thread Returning now to Peirce's reflections about linguistic signs as 'vehicles of functions', it seems possible to claim two things: (a) that such signs will be liable to evolve during the language acquisition process in so far as their degree of conveyable (multi)functionality will gradually develop; and (b) that at any level of language use, some type of indexicality will always be necessary. Any use of a linguistic sign will require a shared context, not as something to denote but as a location in which to be placed. In order to 'say' something in a decontextualised way, a sign will inevitably have to 'show' a shared context, a territory on which the use of the sign will depend indexically. And this will be so from the very first uses of a new language. In order to clarify this argument, let me give an example of the 'use/acquisition' of a Ln in an immersion program. There is a typical kindergarten game called 'the photographer' which is used in many Catalan immersion classes. The basic structure of the game, notwithstanding certain minor variations, is as follows. First certain objects are drawn on the blackboard; then one of the children is chosen as the 'photographer' and has to pretend to photograph the drawings before leaving the classroom. While the 'photographer' is outside, the rest of the class wipes out one of the objects drawn on the board. Finally the child who has gone outside is asked to come back and guess what has been eliminated from the blackboard. Observations of this activity have shown that, once they are familiar with the structure of the game, not only do pupils in Catalan immersion classes experience no difficulty in understanding the explanations given by the teacher in the new school language, but they in turn use the Ln to take part and enable the game to proceed. My purpose in giving this example is to stress the importance of the gameas a structured and stable activityin building up the meaningfulness of the first (oral) texts in the Ln. Let us imagine, for example, that one of the drawings on the blackboard represents an apple ('poma' in Catalan); in this case the production 'poma' can be what the 'photographer' says when he or she goes up to the blackboard to photograph the objects drawn there; or alternatively 'poma' can be what the rest of the class says

< previous page

page_28

next page >

< previous page

page_29

next page > Page 29

when suggesting which drawing should be wiped out once the photographer is outside the room; or again 'poma' can be what the photographer says when he or she comes back into the classroom and guesses what has been wiped out. In any case, as the reader will already have realised, 'poma' has a different meaning and affects the development of the game in a different way. On one occasion it may mean something like, I [in my role of photographer and at the point in the game which precedes my leaving the classroom] must remember that one of the things that may be wiped off the board is a 'poma'; later it may amount to saying, we [who are not the photographer, while the photographer is outside the room and cannot hear what we are saying] propose that you, the teacher, wipe out the 'poma' on the board; or finally the word 'poma' may mean, I [in my role as the photographer, after returning to the classroom] know that the object you have wiped off the board is the 'poma' and so I have won the game. Now, this diversity and wealth of meaning in the first uses of the new language does not derive from the referent but the fact that all the participants are aware of 'who' has produced each utterance and in which 'internal location of the game' he or she has done so. Obviously references are important for the progress of the game and if the child who is the 'photographer' had said 'taronja' [orange] instead of 'poma', his or her performance would have been inefficient. But, as I have tried to argue, the references at their disposal of learners in the new language would not suffice to build up these first uses of the Ln. It is vital that the pupils should recognise the boundaries which define the game and the exact limits of its internal locations in order to be able to place 'the language they already know' into a structure that is known and recognised by all and which will therefore endow the things they 'say' with meaning and allow the game to work. Similar observations have been made, though not in terms of a 'territory' as proposed here, by Read & Scheiber (1982) and W.L. Fillmore (1985). More specifically, the latter author (1985: 24) states that 'the first type of structural characteristic of successful (L2) lessons is consistency in organisation: students knew what to expect and what to do procedurally, because the routine was well established'. It is in this sense that the game functions as a shared territory, and that the procedures for producing and/or understanding the first 'texts in the Ln'in so far as they consist of putting pieces into mutually agreed locations that are co-present and preestablished for all the participantscan be considered indexical. If there were no shared territory, or, rather, if this territory were not clearly established and recognisable for all the learners, then nothing 'said' could be indexed and the first productions would cease to be meaningful and efficient. Just as the smoke stops when we put out the fire, it seems possible to argue that if the speakers had no shared

< previous page

page_29

next page >

< previous page

page_30

next page > Page 30

territory, if the territory were not jointly recognisable by them, then nothing meaningful could be 'said'. In the same way as a goal outside the territory is not a goal, any element of the new language that cannot be used (understood and/or produced) in a jointly recognisable territory will cease to function as a sign. It seems especially important to underline that the possibility of meaningful and efficient use of a language one does not yet know well enough is something that goes beyond what each participant knows, something that inevitably implies the existence of skills which, if they were not collective, would disappear. Thus learning to use a new language depends to some extent at least on the 'things' the learner shares with his or her companions. As Ervin-Tripp (1987: 32754) points out 'play is possible with a minimum of language because children already know the schema for the type of play involved. (...) The foundation for the children's language learning in play is their prior knowledge of the activity structure'. Thus in the last resort it is the group that constructs social procedures for using the new language, in such a way that proficiency in the Ln, either the lack of it, is no longer a question of individuals levels and instead becomes a collectively constructed skill. As pointed out earlier, one of the consequences of the present argument is that a semiotic, jointly constructed, continuity appears in the language use/acquisition process (Vygotsky, 1973; Silverstein, 1976, 1987; Hickmann, 1987; Wertsch, 1985a, 1985b). I am referring, therefore, not to an order in the joint action which allows the learner later to recognise the syntactic order of the different elements in the sentenceas was proposed (1975) and subsequently challenged (1982) by Brunerbut simply to the possibility of recognition deriving from the fact that all use of a language necessarily needs a shared indexical territory, since anyone incapable of recognising territories shared by the interlocutor will be left without meaningful language. Taking up a similar hypothesis proposed by Gleitman (1989: 29), one could say that a child trained to parse sections of text as potential components of shared territories which, though non-linguistic in origin, are indexical, would find it easy to recognise other shared indexical territories in the syntax and discourse rules of his or her language. This would imply that the language acquisition process does not amount exclusively to recognising (phrasal) rules of Generative Grammar (GG) in the input offered by the environment but the rather involves recognising different social semiotic territories, syntax (of GG) being one of the possible options. I agree that without territory there is no meaningChomsky (1988: 155) expresses this by saying that 'the greater part of the theory of meaning is called syntax', but perhaps more than one territory is possible.

< previous page

page_30

next page >

< previous page

page_31

next page > Page 31

The New Language is Learned by Being Used To some extent the arguments put forward so far in the present paper can be summed up as follows. A language cannot first be learned and then used: a language is learned by using it. Hence the fact that various authors have insisted repeatedly in recent years on the importance of 'use situations' in the new language acquisition process, hypothesising, in one way or another, that language acquisition will depend largely on the opportunities available to learners of being users of the Ln (Hatch, 1978; Long, 1983; Artigal et al., 1984; Terrell, 1985; L.W. Fillmore, 1985; Swain, 1985; Taeschner, 1986; Di Pietro, 1987; Ervin-Tripp, 1987). This same argument has in turn provided the basis for many of the studies devoted to immersion programs, so that this type of education has been viewed as a large, natural 'new language use/acquisition context' (Cummins, 1979; Swain & Lapkin, 1983; Krashen, 1984; Artigal, 1989; Vila, 1990). From this point of view, what I have called the joint creation of a shared indexical territory is, in part at least, a requisite if the students are to use, and thus learn, a language they do not yet know, first of all because the things the speakers share are vital for their co-operation in using the language that is to be learnt, and secondly because this shared indexical use will enable the pupil to go on more easily to recognise further input. By organising time, space and their own joint activity, teacher and learners make games like that of the photographer into a mutually recognisable territory, so that the learners not only become collectively competent to use the Ln they are still in the process of acquiring, but at the same time gain a very useful point of departure from which to recognise the formal, decontextualised texts the school will offer later on. In short, as Bakhtine (1977: 124) states, 'the word is [always] the common territory between speaker and interlocutor'. It matters little whether the word is 'poma' as used by children in the game of 'the photographer', or as used in the present paper in describing this same game. A shared indexical territory will always be necessary. This is at least part of the connecting thread, of the process of semiotic continuity on which the methodological construction of language acquisition in an immersion program must be based. Especially if, as in Catalonia, the children come from a low socioeconomic background. Acknowledgements This paper was written thanks to a grant from the Department of Education of the Government of Catalonia. References Artigal, J.M. (1989) La immersió a Catalunya: consideracions psìcolingüístiques i sociolingüístiques. Vic: EUMO. [An English version of this work is currently in press: The Catalan Immersion Program: A European Point of View. Norwood, NJ: Ablex]

< previous page

page_31

next page >

< previous page

page_32

next page > Page 32

Artigal, J.M. et al. (1984) Com fer descobrir una nova llenqua. Vic: EUMO. Bakhtine, M. (1977) Le marxisme et la philosophie du langage. Paris: Les Edition de Minuit. Bates, E. (1976) Language and context. The acquisition of pragmatics. New York: Academic Press. Benveniste, E. (1966) Problèmes de linguistique générale. Paris: Gallimard. Brown, R. (1973) A First Language. Cambridge, Mass.: Harvard University Press. Bruner, J. (1975) From communication to language. Cognition 3, 25587. (1982) The formats of language acquisition. American Journal of Semiotics 13, 116. Chomsky, N. (1988) El lenguaje y los problemas del conocimiento: Conferencias de Managua 1. Madrid: Visor. Cummins, J. (1979) Linguistic interdependence and the educational development of bilingual children. Review of Educational Research 49 (2), 22251. (1980) The cross lingual dimensions of language proficiency: Implications for bilingual education and the optimal age issue. TESOL Quarterly 14, 2, 17587. Di Pietro, R. (1987) Strategic Interaction. Learning Languages Through Scenarios. London: Cambridge University Press. Ervin-Tripp, S. (1987) Activity structure as scaffolding for children's second language learning. In J. Cook-Gumperz, & W. Corsaro, (eds) Issues in Theory and Method of Studying Children's Worlds. The Hague: Mouton, pp. 32757. Fillmore, L.W. (1985) When does teacher's talk work as input. In M. Gass & C. Madden (eds) Input in Second Language Acquisition. Newbury House Publ, pp. 1750. Gleitman, L. (1989) The structural sources of verb meaning. Papers and Reports on Child Language Development. Vol. 28, August 1989, Stanford University, California. Gleitman, L. and Wanner, E. (1982) Language acquisition: The state of the state of the art. In L. Gleitman & E. Wanner (eds) Language Acquisition: The State of The Art. Cambridge University Press, pp. 348. Greenfield, P. and Smith, J. (1976) The Structure of Communication in Early Language Development. New York: Academic Press. Hatch, E. (1978) Discourse analysis and second language acquisition. In E. Hatch (ed.) Second Language Acquisition. Rowley, Ma.: Newbury House, pp. 40135. Hickmann, M. (1985) The implications of discourse skills in Vygotsky's development theory. In J. Wertsch (ed.) Culture, Communication, and Cognition: Vygotskian Perspectives. New York: Cambridge University Press, pp. 23657. (1987) The pragmatics of reference in child language: Some issues in developmental theory. In M. Hickmann (ed.) Social and Functional Approaches to Language and Thought. London: Academic Press, pp. 16484. Hoff-Ginsberg, E. and Shatz, M. (1982) Linguistic input and the child's acquisition of language. Psychological Bulletin 92 (1), 326. Jakobson, R. (1975) Los conmutadores, las categorías verbales y el verbo ruso. In Ensayos de lingüística general. Barcelona: Ariel, pp. 30732. Krashen, S. (1984) Le pourquoi de sa réussite. Langue et sociéte, num. 12: L'enseignement immersif en français, 6467. Long, M. (1983) Native speaker/non-native speaker conversation in the second language classroom. In M. Clarke & J. Handscombe (eds) TESOL '82: Pacific Perspectives on Language Learning and Teaching. Washington DC.: TESOL, pp. 20723. Lyons, J. (1990) Semántica. Barcelona: Teide. Peirce, C. (1987a) Selected writings. Obra lògico semiótica. Madrid: Taurus, pp. 37105. (1987b) Collected papers. Obra lògico semiótica. Madrid: Taurus, pp. 159429. Read, C. and Schreiber, P. (1982) Why short subjects are harder to find than long ones. In L. Gleitman and E. Wanner (eds) Language Acquisition: The State of the Art (pp. 78100). Cambridge: Cambridge University Press. Silverstein, M. (1976) Shifters, linguistics categories, and cultural description. In K.H. Basso & H.A. Selby (eds) Meaning in Antropology. Albuquerque: Univ. of New Mexico Press. (1985) The functional stratification of language and ontogenesis. In J. Wertsch (ed.)

< previous page

page_32

next page >

< previous page

page_33

next page > Page 33

Culture, Communication, and Cognition: Vygotskian Perspectives. New York: Cambridge University Press. (1987) The three faces of 'Function': Preliminaries to a psychology of language. In Maya Hickmann (ed.) Social and Functional Approaches to Language and Thought. London: Academic Press, pp. 1738. Swain, M. (1985) Communicative competence: Some roles of comprehensible input and comprehensible output in its development. In M. Gass & C. Madden (eds) Input in Second Language Acquisition. Newbury House Publ, pp. 23553. Swain, M. and Lapkin, Sh. (1983) Evaluating Bilingual Education: A Canadian Case Study. The Ontario Institute for Studies in Education. Clevedon: Multilingual Matters. Taeschner, T. (1986) Ensegnare la lingua straniera. Bologna: Il Mulino. Terrell, T. (1985) The natural approach to language teaching: An update. In The Canadian Language Review 41 (3), 46179. Vila, I. (1990) Activitats lingüistiques en aules preescolers d'immersio. In II Congrés Internacional de la Llengua Catalana, Area 6, Ensenyament. Barcelona: II Congrés Internacional de la Llengua Catalana Ed, pp. 34347. Vygotsky, L. (1973) Pensamiento y lenguaje. Buenos Aires: La Pleyade. Wertsch, J. (1985a) Introduction. In J. Wertsch (ed.) Culture, Communication, and Cognition: Vygotskian Perspectives. New York: Cambridge University Press. (1985b) Vygotsky and the Social Formation of Mind. Cambridge, Mass.: Harvard University Press. Wittgenstein, L. (1988) Investigaciones filosóficas. Barcelona: Ed. Crítica, Grijalbo.

< previous page

page_33

next page >

< previous page

page_ii

next page >

< previous page

page_35

next page > Page 35

Ritualised Routines and l2 Acquisition: Acquisition Strategies in an Immersion Program Siv Vesterbacka University of Vaasa, Baastuvankatu 31, 65100 Vaasa, Finland Abstract. In 1987 an early total immersion program based on the Canadian model was established in Vaasa/Vasa, Finland. It is evident that the functional approach of the immersion program has consequences for the teacher, the test methods and the language usage of the children. This paper is mainly centred on an analysis of six-year-old Finnish-speaking children's second language acquisition during their first year of Swedish immersion in kindergarten. A conclusion to be drawn from the analysis is that a very important didactic task for the teacher is to provide the children with meaningful routines, in which the role of the second language, both quantitatively and qualitatively is gradually increased. The importance of context-bound situations is illustrated by the code-switching strategies in the children's language usage. In the second part of the paper further evidence of the importance of building up ritualised context-bound situations with standardised language usage is demonstrated on the basis of the children's first second language utterances. Early Total Immersion in Vaasa In 1987 an early total immersion program was established in Vaasa. The participating two groups of Finnish-speaking children (with 25 children in each group) began their Swedish immersion in a kindergarten when they were six years old. In accordance with the Canadian model the children were carefully chosen so that none of them had any knowledge of Swedish when entering the program. This was checked by an enquiry which was sent to the parents. All parents estimated that none of the children was able to understand or produce a whole simple sentence in Swedish at the beginning of the program. As the program is an early total one, the teachers used only Swedish in all their communication with the immersion children during the children's first year in kindergarten. After a year in kindergarten the

< previous page

page_35

next page >

< previous page

page_36

next page > Page 36

immersion children continued their immersion at school. When attending the primary school the children have 75% of the instruction time in Swedish and the rest in Finnish. The immersion program is being supervised by a group of linguists, psychologists, teachers and administrators. The intention is to follow these children at the primary and possibly also the secondary level. The Functional Approach of Immersion Immersion differs from traditional language teaching mainly because it uses a functional approach, which must have certain effects on the whole process, i.e. on the teaching strategies, the research methods and the second language acquisition of the children. I will here deal with the first year of immersion in kindergarten and the children's second language acquisition during that year. I will focus on the acquisition process and show how it indicates certain teaching strategies for immersion programs. This does not mean that I consider the choice of research methods less important. The functional approach to the immersion program causes serious validity problems for many standarised tests. During the first test periods for second language acquisition I noticed that a traditional vocabulary test (test results, Vesterbacka, 1990), where the child is expected to look at some pictures (e.g. a window) and explain what it is, does not suit immersion children at all. They were not able to produce isolated words, whereas in a given context (e.g. 'What does the teacher say when entering the classroom?') many children were able to produce a sentence containing the word (e.g. 'Open/ Close the window') During the children's first year of immersion I had no possibility of following the children's second language development every day. Therefore I used different kinds of standardised and unstandardised linguistic tests for measuring the language proficiency. With the help of the test results I was able to follow how the children's comprehension of Swedish developed in morphology and syntax. I was not, however, satisfied with only these conventional test methods. I wanted to get more information about the spontaneous second language usage of the children in everyday situations during their first immersion year. I felt that it was very likely that these first phrases in Swedish in many ways must reflect the on-going process of early second language acquisition as the children started with no knowledge of the second language. I asked the teachers to note the children's second language utterances and to write them down. The material I got is naturally not complete. The teachers had no possibility of following every Swedish utterance of each child in a group of 25 children. Instead the teachers have chosen to note utterances that represent a new phase in the observed second language development. Also the children's adding of new elements to old phrases was noted. Another problem for my linguistic analysis was that enough context for the utterances was not always given. Despite all these

< previous page

page_36

next page >

< previous page

page_37

next page > Page 37

problems I think it is possible to draw some general conclusions about the early process of second language acquisition in the immersion children and what consequences this process has for the teacher. The notions of the teachers include a variation of language mixing from basically Finnish utterances containing a word in Swedish with a Finnish ending (e.g. Siellä voi olla lingoneitakin) to phrases with two or three words in Swedish and the rest in Finnish (e.g. Ooksä kattonut Tv:tä hela dagen). Apart from the usage of mixed language the teachers have also noted grammatically correct utterances in Swedish. Firstly, I will look at the mixed language utterances and show under what conditions mixed language appears. Secondly I present the second language utterances and their importance for the teaching strategies. Favourable Conditions to Code-switching The utterances with mixed language can clearly not be regarded as an interference phenomenon as the children use them consciously in dialogues with other immersion children and the teachers. Also, there are no reports of the immersion children using mixed language utteranes with monolingual Finnish-speaking children in the school yard. Thus, the children seem to use code-switching strategies. Leonore Arnberg (1988: 423) has defined the following conditions as favourable for code-switching: (a) lack of one word in either language. (b) some activities have only been experienced in one of the languages. (c) some concepts are easier to express in one of the languages. (d) some words are easier, more distinguishable and easier to use in one of the languages. (e) a misunderstanding has to be clarified. (f) one wishes to create a certain communication effect. (g) one continues to speak the language latest used because of the trigger effect. (h) one wants to make a point. (i) one wishes to express group solidarity. (j) one wishes to exclude another person from the dialogue. According to this list I have found that immersion children make use of code-switching for situations a, b, g, i and j. I will briefly comment on these but I would also like to introduce another new situation for code-switching that is typical of immersion children. Code-switching because of lack of one word in either language is clearly distinguishable from other code-switching conditions as regards the immersion children. On the basis of the teachers' notions it is evident that the children's intention is to express themselves in Swedish but they fail because they have either forgotten the Swedish word or have not learnt it yet. As a consequence they temporarily switch to Finnish or finish their utterance in

< previous page

page_37

next page >

< previous page

page_38

next page > Page 38

Finnish. This is very common when a grammatically complex structure should be expressed as in examples 1 and 2. 1. Skriva ekana namn (Write firstly name) 2. X måste sätta rött tommonen rusetti (X has to put on red such a rosette) In all other conditions for code-switching the children start in Finnish and switch to Swedish. During the first year in kindergarten the children learnt activities that belonged specially to the kindergarten. An area of the kindgergarten has been separated for different kinds of games and there is, for example a waterpool which the teachers refer to as 'vattenlek' (waterplay). This activity has only been experienced in the kindergarten and the children use code-switching whenever they talk about the activities in the water pool (example 3). This is also the case with the drier, which many of the children had not experienced before kindergarten (example 4) 3. Saanko mennä vattenlekiin? (May I go to waterplay?) 4. Sitten mä pa ne torkskåpiin (Then I put 'em in drier) Among others, e.g. Saunders (1988: 879) mentions different kinds of trigger effects. He distinguishes consequential, anticipational and contextual triggering and he also mentions so called sandwich-words. In the case of the Swedish immersion children I would rather identify the trigger effect as person-related. It seems that the teachers are very strongly connected with the use of Swedish and therefore the occurrence of the names of the teachers is enough to cause a code-switching as in examples 5 and 6. 5. Pia ja Helena, nu är ni tysta (Pia and Helena, now you keep quiet) 6. Mutta Baabo, klä på dej (But Baabo, put your clothes on) To express group solidarity the context and the participating communication partners are very important. The typical situation for immersion children can be described as a dialogue between one child and another or a whole group of immersion children discussing with each other (examples 7, 8, 9, 10). The teachers are not involved. 7. Se on äppelpuu (It is appletree)to other immersion children 8. Slellä voi olla lingoneitakin (There may be lingonberries as well)to other immersion children. 9. Siuna maten (Bless the foodto another immersion child 10. Meillä on lelupäivä imorron (We have toyday tomorrow)to another immersion child I do not have examples of immersion children using code-switching to

< previous page

page_38

next page >

< previous page

page_39

next page > Page 39

exclude someone from the conversion, but many parents have reported situations where immersion children playing with monolingual children have used Swedish as some kind of 'secret language'. To the list given by Arnberg I would like to add another condition which is favourable to code-switching. This situation is typical of an immersion child, when it is engaged in a dialogue with the teacher. Very often the child is answering a direct question from the teacher. But code-switching can also occur when the teacher is conducting an activity in a smaller group of children. The child does not express the central part of the message in Finnish although the rest of the utterance is in Finnish. Thus these words are closely connected with the teaching situation. 11. Mjölkkiä vaan mulle (Only milk for me)to teacher 12. Mä otin kaks potatista, montako potatista sä otit? (I got two potatoes, how many potatoes did you get?)to the teacher 13. Pitäisikö laittaa regnjackan päälle? (Should I put on the raincoat?)to the teacher 14. Mössan on taskussa. (The cap is in the pocket)to the teacher Ritualised Routines in the Second Language During the Swedish immersion children's first year in kindergarten many visitors in the classroom commented on the children's fluent and correct Swedish. This observation can to a large extent be explained by what different linguists call 'prefabricated routines' (Brown, 1973) or 'prefabricated patterns' (Hakuta, 1974), 'formulaic speech' (Ellis, 1984) or 'conventionalised language' (Yorio, 1989). I will here use the terms 'routines' for unanalysed utterances with no changes within them and 'patterns' for utterances where one or two elements can be changed. Routines have been observed in many studies of successive acquisition of two languages. Among others, Huang & Hatch (1978) and Wong Fillmore (1979) have shown that especially children seem to make use of routines at the very beginning of a second language acquisition. This is true also of the immersion children, where the routines are easily recognised by the obvious influence of the teacher (Klä på dej (Put your clothes on), va tysta (be quiet), inte springa (do not run), ät nu (eat please) or closely connected with the daily program of a kindergarten (städa undan (clean up), vi ska gå ut/in (let's go outdoors/indoors), tack för maten (thank you for the meal), de va gott (that was tasty), ta för dej (please help yourself), vi ses imorron (see you tomorrow)). As the teachers serve as good language models of Swedish the children's routines are usually correctly repeated in Swedish. It is typical of the routines that the bilingual child has not separated the meaning of each word in the routine but is all the same able to use the routine properly in different

< previous page

page_39

next page >

< previous page

page_40

next page > Page 40

situations. So the routine in a way allows the child to express itself in its second language without fully understanding every linguistic element. The most important function of the routines seems to be to establish social interaction, whereas the linquistic/grammatical aspect has no major influence on the routine. In the following, the use of routines is very evident in the children's speech (examples 15, 16) where the routine functions as a label for a certain kind of activity: 15. Onko Vi-ska-gå-ut? (Is it Let's-go-outdoors?) 16. Teldänkin pitää städa undan. (You too have to clean up) For the teaching strategy of an early total immersion teacher, the children's use of routines must be the starting point for promoting the first step of second language acquisition in immersion. Routines give the children confidence to express themselves in their second language at the very early stage of second language acquisition and should be promoted by the teacher. Hence, the teacher's most important function in an early total immersion should be to build up situations or contexts which will become ritualised by being used every day. This context-bound teaching strategy is a teaching method for early immersion that Dr Josep Maria Artigal (an expert on Catalan immersion and didactics) has been working on for several years. On the basis of my examples I hope to give further evidence of the suitability of this method in immersion. When the teacher builds up a context, it is very important that he verbally expresses every nuance of an activity. Only with the help of the context can the child identify the various elements of the activity and be able to follow the activity from its beginning point through its middle to the end. In kindergarten the teacher will get useful help from the strict daily programme which gives him the opportunity of repeating approximately the same linguistic message within the same context almost every day until the verbal message becomes part of the context, i.e. has been ritualised. The teacher and the children are now not sharing only the situational context but also the linguistic code of the context. Once the child is able to recognise the linguistic basic information within a context, it will automatically be able to communicate in its second language by using ritualised routines. At the beginning of an early total immersion program all daily situations should be very context-bound, but by and by the child can make use of routines from previous contexts and adapt them to new similar contexts. In the following examples (1722) I will show how context and ritualised routines promote production of the second language In the immersion kindergarten the teachers always prayed a little prayer before letting the children sit down at the tables to eat. After one and a half months in immersion a boy suddently one day read the prayer before the techer got a chance to do it (example 17). 17. Gode Gud, välsigna maten, amen (Dear God, bless our food, amen)

< previous page

page_40

next page >

< previous page

page_41

next page > Page 41

In example 18 the child recognises the sewing instruction from previous, similar instruction times and completes the utterance of the teacher. 18. Lärare: En upp ... (Teacher: One up ... Elev: ... and one down Child ... and one down) In examples 1922 the children talk to themselves but in Swedish and thereby recall the context for learning and the ritualised routines within it. In all examples the children express themselves exactly as the teacher has previously done. In the first three examples the children are sewing and in the last example a boy commands himself to hurry up. 19. Inte här, men där, inte här, men dår.. (Not here, but there, not here, but there.... 20. Hoppa över en (Skip one) 21. ... o här upp och här her ... (... and here up and here down.... 22. Snabba på (Hurry up) From this first phase of second language acquisition with ritualised routines and shared context, it should be possible to enlarge the teaching of the second language by all the time providing the children with the linguistic 'tools' for producing their own productive and creative second language rules. This can be done either by enlarging the linguistic ritualisation with new linguistic elements within the same context or by gradually introducing new contexts, which are closely connected with the old context. The changed teaching conditions will emerge like routines with variations (patterns). Wong Fillmore (1979) shows in her study the development from routines to patterns in the speech of a little girl. In the beginning the girl uses only the question type: 'How do you do dese?' but a little later on she varies the phrase by adding an element ('How do you do dese little tortillas') and some time later on by varying the verb ('How do you make the flower?'). According to Wong Fillmore the early acquirer of a second language is able to analyse the existing routines in his speech and on the basis of the routines create grammatical rules for producing second language. Wong Fillmore regards routines as a necessary part of the second language acquisition and the only approach for the acquirer to later on be able to construct productive and creative rules in the second language. In the immersion children's second language the same development from routines to patterns can be observed. The children use phrases that are common in the kindergarten but they put some kind of variation into them. They can, for example, combine different routines as in example 23 or they enlarge the routine with a lexical or a syntactical element (example 24). Now it is also evident that the children's need to communicate efficiently and deliver a message is more important than to interact socially. 23. Pappa e här, pappa väntar (daddy is here, daddy is waiting)

< previous page

page_41

next page >

< previous page

page_42

next page > Page 42

24. X är borta ® X och Y är borta (X is gone ® X and Y are gone) In the next phase of the second language acquisition the immersion children make use of unique utterances, which cannot be directly connected with the daily programme of the kindergarten. Instead the child is guided by its need to give or require information. The correctness of the utterance is directly related to each acquirer's mastering of the productive rules of the second language. In examples 25 and 26 the children master the linguistic rules whereas examples 27 and 28 show that the acquisition of the preposition system is not complete yet. 25. Vart försvann cykeln? (Where did the bike disappear?) 26. Vi gillar inte sanå här små barn (We do not like such small children) 27. Jag går också simhallen (I go the swimming hall too) 28. Det huset bor X (That house X lives) The described acquisition process from routines to productive creative second language rules for second language learners has been questioned by Krashen & Scarcella (1978). They point out that automatic drill exercises are not enough to produce bilingualism. I agree with this statement, but I do not think that the shared context and the ritualised routines will lead to automatic drill exercises. Instead the teacher's focus will be on building up contexts with meaningful ritualised routines, which later on can be linguistically enlarged and thereby help the early acquirer both to use the second language in proper contexts and to be confident about doing so in the early phases of the second language acquisition. This method must be especially suitable for early total immersion children who start with no knowledge of the second language and whose first second language utterances are aimed at fulfilling their basic needs within the daily life of the kindergarten. References Arnberg, L. (1988) Så blir barn tvåspråkiga. Vägledning och råd under förskoleåldern. Helsingborg: Schmidts Boktryckeri ab. Brown, R. (1973) First Language: The Early Stages. Cambridge, MA: Harvard University Press. Ellis, R. (1984) Formulaic speech in early classroom second language development. In R. Handscombe, R. Orem & B. Taylor (eds) On TESOL '83: The Question of Control. Washington, D.C.: TESOL. Hakuta, K. (1974) Prefabricated patterns and the emergence of structure in second language acquisition. Language Learning 24, 28797. Huang, J. and Hatch, E. (1978) A Chinese child's acquisition of English. In E. Hatch (ed.) Second Language Acquisition: A Book of Readings. Rowley, Mass.: Newbury House. Krashen, S. and Scarcella, R. (1978) On routines and patterns in language acquisition and performance. Language Learning 28, 283300. Saunders, G. (1988) Bilingual Children: From Birth to Teens. Second Edition. Clevedon: Multilingual Matters. Vesterbacka, S. (1990) Ett språkbadsförsök i svenska i Finland: om didaktiska lösningar,

< previous page

page_42

next page >

< previous page

page_43

next page > Page 43

testmetodik samt elevernas svenskutveckling. In G. Tingbjörn (ed.) Andra symposiet om svenska som andraspråk. Stockholm: Scriptor Förlag. Wong Fillmore, L. (1979) Individual differences in second language learning. In C. Fillmore, D. Kempler and W. Wang (eds) Individual Differences in Language Ability and Language Behaviour. New York: Academic Press. Yorio, C. (1989) Idiomaticity as an indicator of second language proficiency. In K. Hyltenstam & L. Obler (eds) Bilingualism Across the Lifespan: Aspects of Acquisition, Maturity and Loss. Cambridge: Cambridge University Press.

< previous page

page_43

next page >

< previous page

page_ii

next page >

< previous page

page_45

next page > Page 45

Children in Welsh-medium Education: Semilinguals or Innovators? Peter Wynn Thomas Department of Welsh, University of Wales College of Cardiff, PO Box 910, Cardiff CFI 3XW Abstract. Whilst the success of Welsh-medium schools is often popularly appraised in numerical terms, this paper addresses the issue of the quality of the spoken Welsh of children attending such schools. Analysis of the morphology of the inflected pronouns produced by a random sample of children from such schools suggests that a sub-set of the apparent errors may in fact represent a stage in the process of remoulding traditional paradigms to produce elements of a code which is unique to these new speakers of Welsh and which may function as an identity marker. Introduction Until comparatively recently the processes whereby Wales was becoming increasingly anglicised generally progressed most rapidly in those areas bordering on England. The results of the phenomenon are particularly evident in south-east Wales, where the numbers and density of Welsh speakers have steadily declined since the beginning of the century (Aitchison & Carter, 1985). Further testimony to the general lack of vitality of traditional native varieties of south-eastern Welsh is borne by the fact that those varieties are almost extinct. Whilst we are, then, currently witnessing the attrition of native varieties of south-east Welsh, south-east Wales has gained a reputation as an area where Welsh is currently being revived. A prominent factor in the process is the development of education through the medium of Welsh and numerically, at least, we are now reaping the benefits of the continuing growth of Welshmedium schools, the first of which were established in the area some 40 years ago (Griffiths, 1986: 19). So successful has been the campaign for Welsh-medium education in south-east Wales that in an area just north of Cardiff some 20% of children attend Welsh-medium schools (Humphreys,

< previous page

page_45

next page >

< previous page

page_46

next page > Page 46

1988: 3) and the number of Welsh-speaking children in the region is now increasing (Aitchison & Carter, 1985). The number of the speakers of a language is, of course, one of the key determinants of its vitality. Numbers are also powerful indicators in public profiles. But while quantity is without doubt important, an equally salient ingredient in a revival process is the quality of the language spoken by its new speakers. There may, indeed, be considerable tension between the quantitative and qualitative aspects of a language, particularly if a certain variety is stigmatised. Stigmatisation is certainly a problem in southeast Wales, for the Welsh spoken by the new speakers is stereotypically marked by features which are not characteristic of other, traditional, varieties. Whilst some of the innovations are clearly attributable to the influence of English, we are not yet able to determine the ultimate source of the characteristics of our new varieties of south-eastern Welsh. It should, perhaps, be stressed that south-eastern children are likely to score well on any test devised to assess their communicative competence. My particular concern, however is with grammatical and sociolinguistic competence and performance. From these perspectives, it is significant that native speakers of traditional varieties have coined several (generally disparaging) labels for the new accents; the spoken language associated with a school called Ysgol Rhydfelen, for example, has been dubbed Rhydfelenese. Whilst there is, then, a general awareness of the existence of new southeastern varieties of Welsh, lay opinion is divided about the status and prestige which should be accorded to them. On the one hand, there are those who argue that we have witnessed the growth of a new regional dialect. Others are more emotive, and evaluate the new varieties as 'debased' and unacceptable as legitimate forms of Welsh. Such opinions apart, we know very little about the Welsh which the new speakers actually speak. Neither do we know much about the contexts in which, let us say, 'Rhydfelenese' is used. We do, however, have many first hand and anecdotal accounts of style-shifting which testify that at least some children shift in the direction of (a) a new variety when they are in the company of their peers, and (b) a more (non-south-eastern) regionally-biased variety in the company of adults, particularly members of their immediate family. As a result of the Education Reform Act of 1988, a 'national curriculum' is being implemented in England and Wales. Basic to the concept of this 'national curriculum' is the specification of 'appropriate attainment targets, programmes of study and assessment arrangements' for various subjects (The Welsh Office, 1990: x). In Wales, the vast majority of pupils will soon be taught the Welsh language (The Welsh Office, 1990: vivii). Since tests are currently being devised to test pupils' command of Welsh, it is especially important at this time that we should have detailed profiles of the characteristics of the hitherto uncodified varieties of Welsh which are known to be spoken by children already attending Welsh-medium schools. In particular, teachers and educationalists need to be able to differentiate between accept-

< previous page

page_46

next page >

< previous page

next page >

page_47

Page 47 able and unacceptable linguistic features. The investigation of linguistic acceptability is, of couse, a notoriously difficult area, and one which has only recently been probed for standard Welsh (Thomas, 1989). Before the issue of the acceptability of any new varieties of Welsh may be addressed, however, we need to ascertain the forms which the new speakers actually use. A Specific Problem One of the features which distinguish Welsh (and the other Celtic languages) from English is that the former has a sub-set of prepositions which inflect to show person and number. For the third person singular, inflection extends to distinguishing between masculine and feminine gender. Altogether, the simple prepositions may be divided into four basic sub-sets:

(i) Non-inflecting:

Prep.S1

S3m

vel vel vi

vel ve

(ii) Inflecting, with no stem- formative at

at·av (i) at·o (ve) ·ov

·ð·o (ve)

(iii) Inflecting, with 3rd person stemformative

(i)

Inflecting, with stem-formative for all persons ar

ar·n·av ar·n·o (i) (ve)

On the basis of their stem-formatives, then, we have three basic sub-sets of inflecting prepositions, those which have: (a) No stem-formatives at all (b) Stem-formatives for all inflected forms (c) Stem-formativess for third person forms only A further complication is that the various regional dialects and the standard may choose different or no stem-formatives. Thus, to cite two of the most striking examples: Preposition

Standard

South

North

·ð·i

·t·i

·q·i

·ð·i

·d·i

·i

A pilot study indicated that south-eastern children did not have native-speaker-like production of inflecting prepositions in impromptu speech. Since these forms do not feature prominently in spontaneous styles, it was decided to systematically elicit a large number of forms from the population in question by means of a test. The sample The sample is drawn from two types of Welsh-medium schools in southeast Wales. It consists of 60 randomly chosen 12 yearolds and is stratified according to sex and linguistic background. Particular importance is attached to this last variable because the nature of the contact which pupils have

< previous page

page_47

next page >

< previous page

page_48

next page > Page 48

with native varieties of Welsh varies considerably within the study area. Classes in the Cardiff region typically feature L1 speakers who are the offspring of professional immigrants to the capital city. The cultural awareness of such parents generally extends well beyond the school setting to ensuring that their children benefit from all the Welsh-medium activities and networks which Cardiff has to offer: youth groups, entertainment, frequent contact with other Welsh-speaking families, and attendance at Welsh-medium places of worship and their associated activities. These children partake, then, in the Cardiff Welsh 'package'. It is from amongst these that our L1 sample is drawn. Although L1 speakers are always in a minority in any particular school class in south-east Wales, it is a matter of debate as to whether they exert any linguistic influence on their L2 peers. Do L1 speakers, for example, provide linguistic models which their peers may emulate? Or is the pressure of the peer group such that new group-based norms are likely to predominate? In order to probe these possibilites, the sample includes L2 speakers from the Cardiff area whose parents are both non-Welsh speakers, and who do not partake of the Cardiff Welsh package. The only contact which our L2 children typically have with Welsh is in school, where they may hear native varieties spoken by their peers and by their teachers. In contrast to the Cardiff type of Welsh-medium school, schools in the south-eastern valleys typically include few if any native speakers of Welsh. From one such school we have chosen a third group, which we label L3. For the members of this group, then, contact with L1 Welsh speakers is typically confined to the school setting and particularly to their teachers. The test The prepositional forms were elicited by means of a substitution test. This was in the form of two short stories which contained examples of simple prepositional phrases, i.e. structures comprising a noun phrase dominated by a preposition. The noun phrase could consist of either a proper noun or a definite article and a common noun. Each child would be given a copy of the stories in which the prepositional phrases were underlined. The story would then be read aloud to the child, who would be able to follow the written version during the reading. After having the story read to him, the child would then be required to read the story aloud himself and to substitute inflected prepositional forms for the underlined prepositional phrases. For example: Allwn ni ddim mynd heb y gwningen. Can we not go without the rabbit. 'We can't go without the rabbit.' Written trigger: /at n ni ðim mind hebði/ Can we not go without-her. 'We can't go without it.' Spoken response:

< previous page

page_48

next page >

< previous page

page_49

next page > Page 49

Three phrases were added at the beginning as a run-in. The children were, however, clearly familiar with tasks of this nature and soon realised what was required of them. Now in order to produce the correct form/hebði/in the above example, the speaker must have knowledge of a number of features, none of which can be taken as given: 1. that Welsh has a sub-set of inflectional prepositions; 2. that HEB is a member of that sub-set; 3. that the preposition must agree with the gender and number of the noun which it substitutes for; 4. that CWNINGEN is a feminine singular noun; 5. that HEB requires a stem-formative; 6. that the appropriate stem-formative is /ð/; and 7. that the S3f ending is /i/. The tests were designed to attempt to discover to what extent the children in the sample operationalised these concepts in spoken forms. Since the matter of the concord between the form of the inflected preposition and the noun phrase for which it substitutes is complex, this paper will be restricted to the forms of the inflected prepositions. Results Sex There were no significant differences in the repsonse patterns of the two sexes in each linguistic group. The responses of the two sexes have therefore been combined. Standard forms Although the responses featured examples of all the standard forms, these were consistently produced by only one of the 60 children, an L2 girl. Across the three linguistic groups, standard forms featured as follows:

Group

% Standard

L1

74

L2

61

L3

62

The above scores might well be cited as evidence that overall the three linguistic groups favoured standard forms. However, whilst there is little difference between the two second language groups, these clearly perform differently from the L1 children. More particularly, the second language groups produced standard forms in somewhat fewer than two thirds of their

< previous page

page_49

next page >

< previous page

page_50

next page > Page 50

responses; the L1 group, on the other hand, produced standard forms in almost three out of every four instances. It was not surprising to find that the second language groups had produced non-standard forms: it was after all, because the pilot study had clearly indicated that this was the case in impromptu speech that the feature was being studied more systematically. What was unexpected was that the L1 children should also be prone to producing non-standard forms. What, then, were the characteristics of these non-standard responses? Uninflected forms All three groups of children could produce uninflected forms. Thus, for example, instead of /am·dan·o ve/, we could have /am ve/; instead of /gan·ð·o ve/, we could have /gan ve/. The percentage occurrences of uninflected forms were: Group

%

L1

4

L2

12

L3

14

Clearly, uninflected forms were not particularly prominent among the answers of any group, and were more characteristic of the L2 and L3 groups than L1; the L1 speakers did, none the less, produce uninflected forms. Although we have no comparable data from L1 speakers from high density Welsh speaking communities in west Wales, our own native speaker intuition tells us that we would expect 12 year-olds to have completely mastered this piece of inflectional morphology. That the south-eastern L1 children produced uninflected forms at all suggests that, at least as far as this feature is concerned, their linguistic competence is not on a par with that of the typical traditional native speaker. Forms which feature no stem-formatives Treating inflected prepositions as though they were non-inflectable might be construed as the ultimate form of inflectional simplification. Levelling morphological processes could also be less severe and take the form of omitting stem-formatives. Thus instead of / ·t·o, heb·ð·o, ·ð·o/, children from all three groups could produce / ·o, heb·o, ·o/, Superficially, absence of stem-formatives might be attributed to the influence of forms such as northern / ·o/, with the feature spreading to other contexts. However, since no example was found of northern / ·o/ (as opposed to standard /gan·ð·o/, we must be very cautious about appealing to northern varieties in this context. Although forms which lack stemformatives are comparatively infrequent, it is of interest that they are most prominent with the L2 group:

< previous page

page_50

next page >

< previous page

page_51

next page > Page 51

Group

% Formative-Less

L1

7

L2

12

L3

7

This pattern, where a non-standard feature peaks with the L2 group is not uncommon in our data, and suggests that this group may be leading the development of certain innovations. Forms which featured in the L2 and L3 groups only The second language groups produced many permutations and combinations of inflected forms and prepositional objects. We need not detail all of these here, but as examples may be cited: Masculine inflection + feminine pronoun Expected

Produced

S3f prep + S3f pronoun

S3m prep + S3f pronoun

at·i hi:

at·o hi:

i·ð·i hi:

i·ð·o hi:

L2but not L3children could extend this pattern to include nominal objets to the prepositions as well as pronominal ones, e.g. Expected

Produced

prep + noun

S3m prep + noun

rq davið

rq·o davið

mali

·t·o mali

Incorrect person Example

Actual meaning

Intended meaning

am·dan·at

'about you'

'about it'

o ni

'of us'

'of it'

o nu:

'of them'

'of it'

Forms which include non-standard stem-formatives The features which have been outlined above are all indicative of one underlying phenomenon: incomplete native-speaker-like production of the morphology of inflected prepositions, with innovations being particularly characteristic of L2 and L3 speakers. Of the remaining patterns, a particularly striking phenomenon is the use of the stem-formative /ð/. Whilst all groups could produce forms containing appropriate stem-formatives, they could all also introduce /ð/ to all the tested prepositions, i.e. to forms which do not feature it in any native variety of the language. To illustrate using S3m forms, we have:

< previous page

page_51

next page >

< previous page

page_52

next page > Page 52

Preposition

Stnadard S3m

Children's S3m

am

am·dan·o

am·ð·o

ar

ar·n·o

ar·ð·o

dros

·t·o

·ð·o

o

o·hon·o

o·ð·o

at

at·o

at·ð·o

o'dan

odan·o

odan·ð·o

These examples illustrate that the children may regularise standard forms by: (i) substituting /ð/ for four standard stem-formatives, and (ii) introducing /ð/ as a stem-formative to two other prepositions which have no stem-formative in any other variety of the language. What is immediately evident about the children's forms cited above is that they have been produced by drawing on the stock of native morphological processes. For these are not simply 'incorrect' forms which are at variance with the whole history of the language, but are, rather, entirely credible creations. Indeed, from an historical point of view, it is almost surprising that such forms did not arise centuries ago. 1 Morphologically, then, the children's /ð/ forms parallel those of other varieties of Welsh in that they comprise Stem + Stem formative. + Ending, e.g. Stem

+ Stem-formative

+ Ending

Standard:

am

+ dan

+o

Children:

am

+ð

+o

Apart from being morphologically striking, the children's /ð/ forms are intriguing also from the point of view of their inter-group distribution. For not only did children from all three linguistic groups produce these forms, but there was little or no inter-group difference in their frequency of occurrence: Group

% non-standard /ð/

L1

15

L2

15

L3

17

The groups are, then, characterised by a measure of uniformity with respect to this feature. And this in spite of the fact that there has been no direct contact between the L3 children and those in the other two groups. Whilst the distribution of the new /ð/ forms may simply be fortuitous and yet another manifestation of incomplete mastery of Welsh morphology, it may also be the case that they represent a stage in the process of remoulding traditional paradigms in order to produce elements of a code which is unique to these new speakers of Welsh.

< previous page

page_52

next page >

< previous page

page_53

next page > Page 53

Conclusions The data elicited by the preposition test are in several respects typical of other data-sets. In particular, the aspects discussed in this paper suggest that: 1. So called L1 Welsh speakers in the anglicised south-east Wales do not necessarily have native-speaker-like control of all aspects of the language. 2. L2 children may be leading the development of certain innovations in the spoken language. 3. Careful consideration must be given to the possible sociolinguistic significance of 'errors'; at least some of them may be manifestations of the linguistic vitality of the children as they draw on the stock of native morphological processes to create their own linguistic identity. Acknowledgements The research on which this paper is based was supported by a grant from the Welsh Office. A fuller account of the work reported upon here may be found in Thomas & Powell (1989). The data for this paper were elicited by my colleague, Mrs Rebecca Mary Powell. I am grateful to Mrs Powell for her untiring enthusiasm and assistance during the project. Note 1. The standard /ð/ stem-formative of /gan/ and /

/, for example, is a modern development.

References Aitchison, J. and Carter, H. (1985) The Welsh Language 19611981. Cardiff: University of Wales Press. Griffiths, M. (1986) Twf yr Ysgolion Cymraeg. In M. Griffiths (ed.) Addysg Gymraeg: Casgliad o Ysgrifau. Caerdydd: Cydbwyllgor Addysg Cymru. Humphreys, G. E. (1988) Addysg Ddwyieithog yng Nghymru: Camu 'mlaen yn hyderus. Eisteddfod Genedlaethol Cymru. Thomas, P. W. (1989) Derbynioldeb Cymraeg Ysgrifenedig. Cardiff Working Papers in Welsh Linguistics 6. Thomas P. W. and Powell, R. M. (1989) Morffoleg arddodiaid syml. Unpublished report presented to the Welsh Office. The Welsh Office (1990) Welsh in the National Curriculum. Cardiff: The Welsh Office.

< previous page

page_53

next page >

< previous page

page_ii

next page >

< previous page

page_55

next page > Page 55

Immersion Principles in Second Language ProgramsResearch and Policy in Multicultural Australia Michael Clyne Department of Linguistics, Monash University, Clayton, Vic., Australia Abstract. This paper deals with the setting-up and evaluation of a primary school German program in Melbourne using modified immersion principles. 5 to 5 1/2 hours of instruction are given through the medium of German. During the evaluation, this program was compared to a more traditional second language program and one in which immersion principles operated to a lesser extent. The paper shows how the evaluation was able to contribute to policy. The German program has acted as a model for the more general introduction of languages other than English in primary schools in the state of Victoria and contributed a 'bilingual education' perspective to the development of Australia's National Policy on Languages. Over the past 20 years, Australia's self-concept has changed from a British outpost in the Pacific pursuing a rigid assimilation policy on its large migrant population to a multicultural nation with close links to its Asian neighbours. Part of the change has been the replacement of a policy imposing mono-lingualism on immigrants, with school foreign language programs being an unrealistic and unpopular academic exercise, by a widespread official and societal acceptance of bilingualism as something legitimate, desirable, and advantageous to the Australian nation as well as to individuals. Language policies were first ad hoc, but linguists and educationists, ethnic groups, teachers, and trades unions joined together to push to a co-ordinated and comprehensive National Policy on Languages (Lo Bianco, 1987) which is currently being implemented. The guiding principles of the policy are (Senate, 1984): (i) Competence in English for all, (ii) Maintenance and development of languages other than English, (iii) Provision of services in languages other than English, (iv) Opportunities for learning second languages.

< previous page

page_55

next page >

< previous page

page_56

next page > Page 56

Because primary and secondary education are prerogatives of the state governments, most states have developed their own language in education policies to complement the national policy. In this paper, I will demonstrate the interaction between linguistic research and language policy in Australia by focusing on the introduction of second language programs in primary schools. I will attempt to show how the setting-up and evaluation of some pilot German programs has contributed to model-building and how policies arising from this have necessitated further research. In the early 1970s, four main sets of arguments promoted experimentation with the early introduction of second languages: (i) There was a general disenchantment with secondary school language programs (generally French programs) which were often not achieving their goals and which were being marginalised by school principles (and other decision-makers) with a monolingual/monocultural ideology. (ii) The reported success of FLES programs in the US and the primary school English programs in Scandinavia motivated some schools to experiment in an ad hoc way, for there were no structural possibilities (at least in state schools) of introducing a second language prior to Grade 7. It was widely believed at the time that younger children could learn a second language faster than older children. This was supported by observation of the children's motivation and pronunciation. (iii) Existing language resources were being wasted by language shift because of the pressure to assimilate exerted by schools and their inability to offer programs that even went some of the way to enable children's home language to keep pace with their English and their cognitive development. This detracted from the children's self-esteem and from family communication and cohesion. It was felt by advocates of early introduction of a second language that even a limited program in the language would give recognition to the child's bilingualism and promote a 'sharing of languages' in the Australian community. This view was concomitant with a general extension of the range of languages taught in schools. (iv) A second language should be an integral part of every child's total education at primary school as well as secondary school. The ensuing programs were generally one hour or two half hours per week. Some were taught by mothers as part of an electives session (as an alternative to cooking, photography, pottery or floral-arrangement). Others were offered by a class teacher with a background in a particular language, and in non-government schools which cover all thirteen years of education by a secondary teacher of the language. Senior university students collecting data for theses or term papers also provided tuition. There were periodic workshops at universities and colleges for people teaching languages at primary schools. In one state, South Australia, primary schools introduced a second language

< previous page

page_56

next page >

< previous page

page_57

next page > Page 57

as part of their normal curriculum, usually starting in Grade 3 or 4, with a brief weekly time allotment, and an emphasis on cultural awareness rather than language proficiency. Most of the teachers were itinerant (Murray & Lundberg, 1977). The shortcomings of most of the programs mentioned were: (i) Lack of continuity, (ii) the absence of suitably trained teachers, and (iii) the absence of systematic evaluation which would ascertain objectively why some programs were more successful than others. By about 1980, there was a strong desire among proponents of language study and multicultural education to institutionalise primary school second language teaching in a serious way. Applied linguists were impressed by the immersion programs that had developed in Canada (as reported in Lambert & Tucker, 1972) but there was scepticism among educational authorities. The argument 'Canada is different' represented the viewpoint that since no one language other than English had the same significance for Australia that French has for Canada, no such language could claim the place on an Australian curriculum that French had attained in English-speaking Canada. At the time, funds were being made available for innovative projects in Multicultural Education, and a successful submission was made by the Association of German-speaking Communities, representing the German-speaking people of Victoria and two Melbourne state primary schools who had previously taught German but had had to drop their programs due to the lack of a teacher. The objectives of the project were: (i) To offer a second language program with continuity which does not have to rely on unqualified voluntary teachers. (ii) To teach a language which is spoken by families of some children in the school and is also of interest to pupils from monolingual English-speaking families. (iii) To try out models including ones incorporating some bilingual education. (iv) To conduct a progressive evaluation of the program to assess the (relative) suitability of the models for different starting age groups as well as the children's language proficiency. In this way, intervention can take place to rectify problems. The evaluation would be carried out by staff and graduate students of Monash University. It was hoped that the programs would have model-building value beyond the particular schools and the language taught. Initially one teacher was shared between the two schools, A and B.

< previous page

page_57

next page >

< previous page

page_58

next page > Page 58

Model A Model A (the model employed in school A) was based on the immersion principle that children acquire L2 by being taught (part of) the curriculum in that language. As in Canada, the teacher talks to the children exclusively in L2, both inside and outside the classroom. This model differs from the Canadian programs in its limited time allotment (5 to 5 1/2 hours per week) and in that it is intended for children from German-speaking as well as English-speaking backgrounds. (Although the area of the school has a large number of German-speaking residents, there are few schoolchildren of this background. However, there are opportunities for children to receive German input from German-speaking neighbours, shops, and community organisations. Older children at the school participate in local German activities and visit one of the local German old people's homes and conduct conversations with the residents as part of their social studies.) The subjects chosen to be taught in German were selected for their contribution to L2 development: Science to provide a close link between linguistic and cognitive development and a basis for frequent interaction between teacher and class; Art to introduce basic notions (e.g. colour, size), emotional expressions (of liking, disliking) and requests (e.g. asking for art and craft utensils) and to provide a basis for interaction between individual pupils and the teacher: Physical Education to introduce functions relating to human activity and to develop listening skills. These subjects have a strong non-verbal component to assist the pupils through the early stages of second language acquisition. Two other subjects, Music and Social Studies, were divided into a German and an English cultural segment, each taught in the appropriate language by the respective teacher. Thus the teaching of language and culture could be integrated. With the integration of Science, Social Studies, and other subjects in later years, this link could be intensified. The use of L2 in Art and Physical Education was eventually discontinued because of the passivity and repetitiveness of the classroom discourse in these subjects. Initially, School A introduced German only in Grades 1 and 3. Model B As School B wanted to have one class at each level participating in the program, each group received only two half hours per week of German. Thus, model and exposure time cannot be separated. For comparative purposes, only the children starting in Grades 1 and 3 were evaluated in this model. Though the teacher spoke German virtually all the time, the emphasis was entirely on the language, including the grammar (albeit in an informal way), within the framework of a functional-notional syllabus. As

< previous page

page_58

next page >

< previous page

page_59

next page > Page 59

in the other schools involved in this project, children's ungrammatical forms are corrected only indirectly with the teacher employing the grammatical form in the response. There is some presence of German speakers in the area of this school although not as much as in that of School A. And as elsewhere in Melbourne, there are German children's books in the local public library and German programmes (some for children) on radio and TV. Extension to other classes and to Model C The apparent success of the programs led to an extension of funding after two years. This facilitated the employment of additional teachers, the introduction of two new cohorts into the German program at School A (starting in Grades 2 and 4 respectively) and the development of a compromise model (C) trialled on children starting in Grades 1, 2, 3 and 4 at a school neighbouring School A. Here children received language instruction (though the medium of L2) for one hour per week, and two hours per week of German-medium curriculum segments from Social Studies, Science and Music/Movement/Drama. Care was taken, as in the other schools, not to teach the same content in the two languages as then children would be tempted to 'switch off' in their weaker language. A similar program is offered in Italian, and all children at School C are required to take one of these languages. The evaluation German development and proficiency, English and subject skills imparted in German, attitudes, and language maintenance of children from German-speaking families were all included in the evaluation. 49 tests were administered in the first four years, and lessons were observed and recorded with a radio microphone. Diaries and letters written by the children were also studied. The tests included listening-comprehension, sound discrimination, elicited sentence repetition (Adams, 1978), conversation (including role-play, telling and retelling of narratives, giving directions, and other functional exercises), grammar (contexualised transformations), Cloze tests, dictation, written expression, and reading-comprehension. In this paper, I shall concentrate on the German, suffice it to say that testing confirmed evaluations in Canada that children in immersion-type second language programs performed at least as well in English and subject skills imparted in L2 as pupils in control groups receiving all instruction in English (Eckstein, 1986). The evaluation was able to identify the features of second language programs most likely to lead to successful outcomes: (i) Functional specialisation between the languages on the basis of 'one person one language' similar to that often practised in bilingual language

< previous page

page_59

next page >

< previous page

page_60

next page > Page 60

Table 1 Mean scores for all skills over the fist two years: Comparison of models according to entry age Start Year Model Grade 1 Grade 2 Grade 3 Grade 4 BS (A) 66.47 63.95 69.95 69.44 BW (C) 56.41 63.96 63.23 65.44 SN (B) 50.50 59.91 acquisition (cf. Saunders, 1988) with different subjects or different content being taught in the two languages. (ii) The stimulation of communicative need by making L2 the sole language of some 'mainstream' subjects. (iii) Comprehensible input (Krashen, 1981) provided in class and through school- and community-based extra-curricular activities. To an increasing extent, a need for grammatical awareness was detected. The evaluation indicates the following advantages of Model A over Model C and especially over Model B (Clyne, 1986): (i) A higher level of performance in all test measures (with the exception of Grade 2 starters in School C whose aggregate scores were marginally better than that of their School A counterparts (see Table 1). (ii) Less range of scores indicated by a lower coefficient of variability, the difference being greatest among Grade 1 starters (see Table 2). This suggests that immersion experience can cater for children from a wider range of abilities and backgrounds. The superiority of Model A for children from German-speaking families has been discussed by Imberger (1986). (iii) A greater ability of the children to function effectively in German, often with the use of neologisms and integrated transfers from English. (iv) On the whole, a more rapid movement through four phases of interlanguage development, subject to largescale individual differences: Table 2 Average coefficient of variability (over six measures) by model for different start years, first two years of program Model Grade 1 Grade 2 Grade 3 Grade 4 BS (A) 0.22 0.20 0.23 0.23 BW (C) (0.25) 0.23 0.26 0.23 SW (B) 0.40 0.27 -

< previous page

page_60

next page >

< previous page

page_61

next page > Page 61

Phase I. English responses to German utterances, one and two word German utterances or unanalysed formulaic utterances, e.g. Zu Fub. Ich bin sechs Jahre alt. Phase II. The matrix language (usually determined by the language of the verb) is English, but individual German nouns, noun phrases or uninflected adjectives are transferred, with occasional code-switching, e.g. Mein Grobvater und mein Grobmutter have come to stay at our place because mein Mutter is in the Krankenhaus and she's had a baby. Phase III. An atempt to speak German with the support of code-switching to English and creative integration of English items into the German phonological and grammatical systems, e.g. Heidi macht das Vater helfen und Vater helf Heidi, then Heidi gets the paper und Vater macht Breakfast. Phase IV. German utterances with the integration of transfers common in German-English bilinguals: Heidi bringt das Essen zu Mutti. Heidi find' die Mutter. Die Mutter ist im Bett. Heidi macht ein Trink for die Mutter. Heidi ist auf die Toilette. Heidi zieht da pyjama auf (= off). (v) Generally, Model A children also experienced a swifter phased development of verb morphology, from the overgeneralisation of the -st form via unsystematic, then systematic variation and finally consistently grammatical forms. Alongide variation is continued -st overgeneralisation and a progressive form based on the infinitive (Er essen/ist essen) and -st with variation inversion (e.g. hast er?) (see Table 3). (vi) More uninhibited speech and a greater degree of linguistic creativity. (vii) A wider spread of subjects for acquisition of different functions and notions. (viii) The integration of language and culture. (ix) Because well established primary school subjects are taught in L2, that language is, in effect, 'mainstreamed'. (x) In Models A and C it is not necessary to make a special case for a more than tokenistic allotment for the second language to be introduced as Table 3 Verb morphology acquisition phases Perfective (non-progressive) 1. Formulaic 2. -st 3. Unsystematic variation -st in inversion 4. Systematic variation, but -st in inversion 5. Grammatical

< previous page

Progressive -en ist + -en/ -ing

page_61

next page >

< previous page

page_62

next page > Page 62

the L2-medium subjects are an essential component of the curriculum anyway, and the language is an additional bonus. In comparison with Models B and C, the possible disadvantages are: (i) The danger that answering in English may become habitual. (ii) The social upheavals generated by having to keep children in a bilingual program in a separate stream in each year if the bilingual program is optional. One of the advantages of the progressive evaluation is that it has been possible to rectify problems during the development of the program. It has also been possible to 'borrow' features of other programssuch as the teaching of some curriculum content in Model B and the introduction of half an hour of language arts in German in the second year of the program in Model A. As to the 'age question', the older starters in each school have generally tended to score better than the younger starters. This difference is significant only in Model B (the 'straight second language program'). The younger children have more time in which to improve their competence. The maximum benefits of language learning are experienced in Grade 3, pupils in that grade in all schools attaining the highest or almost the highest overall results in practically all skills (Reich, 1986). Spelling was substantially better in written work of the older starters than that of their younger counterparts (Clyne, 1986: 7381) supporting the argument in favour of literacy skills imparted in Ll and later transferred to L2. Research and Policy Let me return to the interaction between research and policy. The progress results of the project and the apparent success of the programs encouraged the Victorian Ministry of Education to yield to pressure from sections of the public to institutionalise the teaching of languages other than English in state primary schools. In 1983, fifty language teachers were appointed to schools presenting a successful submission for a bilingual or second language program and the number of such teachers would be increased annually. Budget cuts slowed down this process, but there are at present 174 equivalent full-time supernumerary teachers teaching 17 languages in state primary schools, including Italian, Greek, German, Chinese, Turkish, Japanese, Macedonian, and French. In 1984, a Ministerial Policy Statement on curriculum gave school councils the responsibility to ensure that progressively all students gain proficiency 'in a second language used in the Australian community' (Victoria, 1984: 17). In the meantime, the federal parliamentary inquiry on National Language Policy (Australia, 1984) designated maintenance and development of languages other than English and opportunities for learning second languages as two of its four guiding principles. Model

< previous page

page_62

next page >

< previous page

page_63

next page > Page 63

A was discussed in some of the hearings and in submissions from ethnic and professional associations. The report of the inquiry formed the basis for the Victorian policy statement The Place of Languages other than English in Victorian Schools (1985) which featured 'Bilingual Education for all Children' as one of three models recommended to schools. It clearly set out the continuum of backgrounds from 'native' to 'non-native speaker' that pupils are likely to come from. The report also stressed the need for ongoing research and evaluation. The subsequent Victorian Languages Action Plan (Lo Bianco, 1989), which includes a case study of the Model A program, recommends the doubling of bilingual and immersion programs in the state every three years for the next decade, specific encouragement of such programs, and the training of bilingual teachers who can teach other subjects through the medium of a language other than English. Models A and C (the former by then in its sixth year) inform some of the descriptive sections of the final National Policy on Languages (Lo Bianco, 1987) and the recommendation of 'where possible immersion bilingual teaching' as a method of second language teaching (p. 143). The importance of 'continuing and detailed evaluation' to monitor programs is expressed (Lo Bianco, 1987: 155). Most states now have a languages in education policy statement. They differ in their recommendations depending on their demography and their achievements to date. All of them recommend the expansion of language study at the primary school. The minimum number of hours prescribed or recommended for such study variesthree in Victoria, two in South Australia, one in Western Australia. South Australian programs generally start in Grade 4, those in New South Wales tend to concentrate on the later years of primary school, and the Victorian ones (in state schools) usually start in Grade 1 or the Kindergarten class. None of the other states are as strongly in favour of immersion principles as Victoria, where the experiments described took place. The Western Australian language policy recommends the setting up of a limited number of partial immersion programs based on Model A (W.A., 1988), whereas its New South Wales counterpart favours bilingual programs only for children from appropriate ethnolinguistic backgrounds. Two challenges remainthe demand for suitably trained teachers and continuity at secondary school. Although some primary teacher training institutions are now offering advanced programs in languages and methodology of bilingual education, these courses are not yet sufficient in many languages, and there is a shortfall in numbers of teachers required and the quality, considering the combination of skills needed. This issue is being considered at present by two inquiriesone on the teaching of modern languages in higher education, the other on language teacher need and supply. A follow-on program was developed at a local high school for the children from schools A and C and another primary school which had adopted a

< previous page

page_63

next page >

< previous page

page_64

next page > Page 64

Model B program for two hours per week. The high school gave the children who had taken German or Italian the opportunity to continue at their own level. It taught L2, Geography, and History through the medium of that L2. Theses describing the second language development of the children (Csipek, 1986; Doblin, 1986) indicate a continuing improvement of German proficiency and a gradual levelling of ex-Model A and C children with pupils from the Model B program obtaining either the best or the worst scores. For the latter group, which is very small, the secondary school program is like a late partial immersion program. The secondary school program, compared to the feeder primary school programs, is relatively marginal to the school. In some districts, children are offered continuity from the language level they attained at primary school but not within a bilingual program. In some others, the unsatisfactory situation persists where children have to 'relearn' from scratch in a different way a language which they have been exposed to for several years. This is inevitable for some children whose family moves house when they complete primary school or who are sent to a non-government secondary school and they do not benefit from geographical school cluster arrangements. A small contribution to the alleviation of the problem is student profiles written by the primary school language teacher. Where children transfer to a secondary school not teaching their primary school language at all, it is possible for them to continue that language on Saturdays at a regional centre of the Victorian School of Languages or its equivalent in New South Wales or South Australia. The expansion of primary school language teaching in all states has generated a great deal of experimentation with different models, different starting ages, different time allocations and distribution through the week for instance, four hours of subjects in L2, one day's immersion, late immersion. There is a need to research the effectiveness of these models and practices in various sociolinguistic contexts and for different languages. Much of this will be facilitated by the recent establishment of the Languages Institute of Australia to inform the National Policy on Languages. Its nationwide network of research and development centres is enabling various universities to concentrate on their areas of strength in applied linguistic research as part of a plan of national co-operation in which policy formulation and implementation are served by research. The study of the effectiveness of particular models for the introduction of primary second language programs in given environments will, in fact, be one of the projects of the Institute's Language and Society Centre at Monash University. In this paper, I have attempted to show briefly how immersion principles have been introduced into programs that are less ambitious than the Canadian immersion ones but which address the needs of our multicultural society. The programs I have described have played a role in language policy and enabled me to demonstrate the necessity for a close link between research and policy.

< previous page

page_64

next page >

< previous page

page_65

next page > Page 65

References Adams, M. A. (1978) Methodology for examining second language acquisition. In E. Hatch (ed.) Second Language Acquisition (pp. 22796). Rowley: Newbury House. Australia (1984) National Language Policy. Report of the Senate Committee on Education and the Arts. Canberra: AGPS. Clyne, M. (ed.) (1986) An Early Start. Melbourne: River Seine. Csipek, C. (1986) A Study of Year 7 students learning German by immersion. B.A. (Hons.) thesis, Monash university. Doblin, J.C. (1986) The second language development of Year 8 students in a German partial immersion program. B.A. (Hons.) thesis, Monash University. Eckstein, A.L. (1986) Effect of the bilingual program on English language and cognitive development. In Clyne (1986): 8299. Imberger, B.H. (1986) Children from German-speaking families. In Clyne (1986): 11227. Krashen, S.D. (1981) Second Language Acquisition and Second Language Learning. Oxford: Pergamon. Lambert, W.E. and Tucker, C.R. (1972) Bilingual Education of Children. Rowley: Newbury House. Lo Bianco, J. (1987) National Policy on Languages. Canberra: AGPS. (1989) VictoriaLanguages Action Plan. Melbourne: Ministry of Education. Murray, D. and Lundberg, K. (1977) A Register of Language Teaching in South Australia, 19756. Adelaide: South Australian Department of Education. Reich, H. (1986) The effect of age on performance in different language skills. In Clyne (1986): 5770. Saunders, G.W. (1988) Bilingual Children: From Birth to Teens. Clevedon: Multilingual Matters. Victoria (1984) Ministerial Policy Statement 6. Curriculum Development and Planning in Victoria. Melbourne: Ministry of Education. (1985) The Place of Languages other than English in Victorian Schools. Melbourne: State Board of Education and Ministerial Advisory Committee on Multicultural and Migrant Education. Western Australia (1988) Languages for Western Australians. Perth: Ministerial Working Party on the Development of Policy for the Teaching of Languages other than English in Western Australian Schools.

< previous page

page_65

next page >

< previous page

page_ii

next page >

< previous page

page_67

next page > Page 67

A Two-Phase-Didactics for School Christer Laurén University of Vaasa, Raastuvankatu 31, Vaasa/Vasa, Finland Abstract. Results gained by traditional language teaching at school are discussed. Far too often they manifest themselves in less linguistic ability than the underlying competence should allow. Taking into account what we know about language acquisition, different theories on so-called critical periods for language acquisition and about the immersion schools, it would be natural to propose a very early start of language studies at school. This would obviously give more fluency to the basic skills. Language should also be used deliberately as an instrument as is done in the Canadian immersion schools. Because of differing acquisition abilities at different ages, the students could have a clearly differentiated input during two main periods at school. Excellent pronunciation is, for instance, easily acquired very early. Studies in bilingualism generally give useful insights into language acquisition. They function as an efficient heuristic process for didactics of language. Many new findings could be immediately applicable in language teaching at school level. There are, however, some problems. The Scandinavian school systems tend to be rather rigid because of their uniformity. It is difficult and risky to introduce new thinking into a huge and well organised machine. Nevertheless, in Vasa/Vaasa parents, politicians and school authorities have been keen on changing things (Laurén & Vesterbacka, 1990), so we should not be pessimistic. My considerations are mainly based on experiences from the immersion schools, and they will be supported by other findings with regard to language acquisition. My aim is to draw some conclusions for the language programme of primary and secondary schools, especially for the bilingual areas of Finland but also, I think, applicable elsewhere. School Results vs. Expectations First, I will cite from a report on the language used by a learner: Her 'language performance often does not reflect her underlying linguistic ability. In production she rarely makes use of certain rules (pluralization, for example) or produces certain structures she is capable of producing

< previous page

page_67

next page >

< previous page

page_68

next page > Page 68

(e.g. complements, embeddings). In comprehension, she manifests waxing and waning performance for some aspects of language (e.g. me versus you, future with going to). Whereas normal children appear to continually make use of the range and variety of linguistic elements and rules they have acquired in both comprehension and production, Genie does not. Especially in production, but in comprehension as well, Genie often manifests less linguistic ability than her underlying competence would allow.' (Curtiss, 1977:203) I think the report cited could be a good description of the linguistic competence achieved by most students in the languages they studied at school. We devote a lot of time in the schools of the Scandinavian countries to teaching languages, for we cannot manage without a knowledge of many languages. Our economy as well as our culture demand that most of us master several languages at a high level. It is, however, a commonly known fact that the results are weak in relation to the time used at school and in relation to the expectations of employers and even to the expectations of the students themselves. The problem is not only a Finnish or a Nordic one, it is a European and an international one, too. During the last few decades research in bilingualism has taught us a great deal about successful development of individual bilingualism and the stages of language acquisition. The Canadian early total immersion school (Cf. 'Ni imitació ni deducció: reconeixement', Artigal, 1989:41; Swain & Lapkin, 1982) has been able to produce highly bilingual students with most language skills far beyond those achieved by students in old-fashioned second-language programmes. It is time to draw some radical conclusions for those parts of the school system which, for very practical reasons, cannot be transformed into early total immersion schools. We know that the normal programme usually does not develop the pronunciation and the fluency we would like to reach. The interlanguage produced as a result of the school programme marks its speakers as nonnative speakers in a way that may lead to social and psychological problems. We also know that even superficial deficiencies in the pronunciation may lead to misunderstanding or total lack of understanding in communication. As a result a student may suffer from lacking confidence in his own ability to use a language he may have studied for almost a decade at school. Not even his teacher is always able to manage a telephone conversation with a native speaker without problems. Of course, every one of us is eager to inform about new trends in teaching and learning. We also know that the teachers of today are well trained for their tasks. The basic form, however, of language teaching has not been changed. Students begin their language studies at about the same age as they did for a whole century before us, even if we already know enough for a radical change of system.

< previous page

page_68

next page >

< previous page

page_69

next page > Page 69

The Critical Period and Individual Differences For children the period up to the age of puberty is immensely effective for the acquisition of linguistic fluency, including pronunciation, basic syntax and communication ability. This aptitude for language acquisition is greater the earlier the child can be exposed to language, weaker the closer we come to puberty. In the school today languages are introduced into the programme too late; the ability to acquire a new language is not at its optimum any more. Such perfect imitation of pronunciation as can be managed by children at kindergarten and primary school is not possible any more. There are many theories, with different empirical backgrounds, formulated in order to explain the easily observed phenomenon that small children may acquire some levels of their second language very successfully. There is a brain plasticity hypothesis (Penfield, 1964: 80, according to Harley, 1986: 4) maintaining that the child is a specialist in learning to speak before the age of nine to 12. After that age the human brain becomes progressively too stiff for learning languages. There is also a theory of a biologically based critical period for language acquisition between the ages of two and about 13 (Lenneberg, 1967; Harley, 1986: 58). The period is limited by cerebral immaturity in its beginning, and it is closed by the brain's incapacity for reorganisation. In her work Age in Second Language Acquisition (1986) Birgit Harley points out that this theory is very close to the transformational-generative theory of language acquisition. A third theory is the cognitively based critical period hypothesis (Harley, 1986: 813). It has been proposed that the ability of formal operational thinking at about the age of puberty may inhibit the individual's natural language acquisition ability. Since Lenneberg's hypothesis was put forward in 1967, the elaborations and numerous new theories indicate that the problem is a very multifaceted one. In a critical discussion of age and language acquisition Jürgen Quetz ends up by reformulating the question: age should perhaps be considered as a co-variable of other, more influential factors, mainly socio-psychological and biographical ones, which seem to cause precisely those obstacles to adult second language learning which are most difficult to overcome (Quetz, 1982: 1187). So, irrespective of the hypothesis we put forward about the reasons behind children's very efficient acquisition of a second language, we tend to agree that children for some reason or other do acquire some levels of language more easily than older children and grown-ups. One common characteristic of these theories and their varieties and elaborations is also that they do not tend to take into account individual differences between those beyond puberty. This should of course not be forgotten when planning programmes. At the University of Technology of Compiègne in France, language courses are designed in accordance with the tendency of learners to work

< previous page

page_69

next page >

< previous page

page_70

next page > Page 70

as serialists or, alternatively, as holists (Narcy & Rebelo, 1985; Narcy, 1990). Those students who are working holistically have tendencies in common with children before puberty: a) observation through experience, b) the rule: one should speak to learn, c) communicative activities are regarded as enjoyable. There are, as we know, other tendencies in the didactics of language to work holistically. One of these is represented by the SGAV society 1. The didactic means involve activation of many senses of the learner. Also for adult learners the results are reported to be at least partly efficient where we would otherwise not succeed because we introduce language acquisition too late. We have some empirical knowledge of children acquiring their first language too late, after their puberty, the so called 'Wild Children'. A very tragic case was that of the American girl Genie, who was not given the possibility to acquire a language until the age of 13. She had been isolated by her father, who never uttered a word to her. He beat her when she made some noise and he barked at her to frighten her and in that way to hinder her from producing any noise at all. The paragraph I cited earlier was a report on her language ability, on her mastering her first language which she began to acquire at 13. Her father did not accept any noise; therefore there was no television or radio set at home and Genie was alone and isolated in her room. According to reports from the end of the 1970s, her language acquisition suffered from bad articulation, difficulties with pronunciation and word order. She also had great problems in keeping up any normal conversation. Her vocabulary grew fairly rapidly, possibly due to the fact that her conceptual development was a little more developed than for children acquiring their first language (Curtiss, 1977). Children obviously have a perceptive language ability, the more favourable the earlier it is used. This ability is also linked to a very flexible productive ability, a combination we should use more efficiently. The perception does not only work with the surface structures of language but also with communication strategies. The students at the Canadian immersion schools are able to take into account the communicative needs of those they are talking to more than unilinguals do. We have some experiences from our immersion school in Vasa/Vaasa pointing in the same direction. It would be worth while to invest in early second language acquisition because of favourable results in pronunciation, in some syntactic matters (basic patterns and word order) and in communication strategies. It would also be important to encourage the development of positive attitudes towards language learning in general, towards those using other languages and towards the learner's own potentialities. We know that older learners easily expand their vocabulary. Their cognitive abilities make it possible for them to make use of conscious strategies for efficient learning of new words. Generally speaking, those strategies do not function in the same way for younger children (Schneider & Uhl, 1990; Le Boudec, 1989). For children there is, of course, a successive growth

< previous page

page_70

next page >

< previous page

page_71

next page > Page 71

with increasing age found in the usage of semantic inter-item-relations (Knopf & Waldman, 1990). But the big lexical expansions have to take place when they have the cognitive level necessary for such a task. Additive Language Acquisition When acquiring a second language, it is always important that there is an additive language acquisition, not a subtractive one, i.e. the mother tongue, the first language must have a safe and strong support in the surroundings. This means that linguistic minorities and majorities cannot always use the same strategies for second language acquisition. Finland-Swedish children in Karleby/Kokkola and Helsingfors/Helsinki, to mention just two bilingual cities in Finland, mainly need additive support for their Swedish. A Two-Phase-Didactics for School Taking into account what we know about language acquisition, an ideal programme for second language at school would entail two phases: Phase A (a) very early: at kindergarten and primary school, (b) using language almost exclusively as an instrument, (c) goals: especially for reaching native-like pronunciation, automatised basic syntax, efficient communicative strategies and positive attitudes towards language learning. Phase B (a) following Phase A, (b) using language mainly as an instrument, (c) language teaching emphasised, (d) goals: especially for developing lexicon and syntax according to cognitive level. As a consequence, during Phase A one should work as they used to do at an immersion kindergarten, e.g. two days a week. If there is an obvious switch to the immersion programme, by use of ritualised routines, the drawbacks of using the same teacher for the immersion programme as the children have the other days of the week may be minimised. There would be no possibilities to realise the programme on a large scale in Finland, if native speakers of the second language were required. There are some results of experiments indicating that an efficient way of working may be to use good recordings and minimising teacher talk, at least during some circumstances. The results are positive even for the active oral use of language. This could be one of the strategies used for introducing second language acquisition at an early stage if we do not have access to the ordinary immersion resources.

< previous page

page_71

next page >

< previous page

page_72

next page > Page 72

We would also need to have kindergarten teachers and first and second grade teachers specialised in children's language, language development and literature for children. Of course, every teacher at those levels should be specialised in the language development of a growing child. The Two-Phase-Model would, roughly speaking, create a basic linguistic fluency at a very early stage, when the prerequisites for it are optimal. Later on there is the optimal period for expanding the linguistic repertoire. The bilingual districts of Finland can without difficulty realise the Two-Phase-Model for Swedish in Finnish schools and for Finnish in Swedish schools. The results will be interesting indeed. We can possibly gain better results with fewer teaching hours, especially in the grades with the greatest burden of teaching demands from many subject fields. The Two-Phase-Model can also, I think, with acceptable results be used for other languages as well. The quantitative need of language teachers will surely not decrease but the roles will change. Note 1. La Société de la methode structuro-globale audio-visuelle. References Artigal, J. M. (1989) La immersió a Catalunya. Barcelona: Eumo Editorial. Curtiss, S. (1977) Genie. A Psycholinguistic Study of a Modern-day 'Wild Child'. New York: Academic Press. Harley, B. (1986) Age in Second Language Acquisition. Clevedon, Avon: Multilingual Matters. Knopf, M. and Waldman, M.R. (1990) Die Entwicklung itemspezifischer und relationaler Geächtnisprozesse bei 4- und 6jährigen Kindern. Zeitschrift für Entwicklungspsychologie und Pädagogische Psychologie 22:1. 4253. Laurßn, Ch. & Vesterbacka, S. (eds), (1990) Language Immersion School of Vaasa/Vasa. Vaasa/Vasa: University of Vaasa. Le Boudec, B. (1989) Quelle représentation pour la mémorisation d'informations chez l'enfant de 4 et 5 ans? L'Année Psychologique 89, 3748. Lenneberg, E. H. (1967) Biological Foundations of Language. New York: Wiley & Sons. Narcy, J. P. (1990) Apprendre une langue etrangère. Paris: Les Éditions d'organisation. Narcy, J. P. and Rebelo, M. (1985) What happens to L.S.P. when students take the initiative? In M. P. Perrin (ed.), Pratiques d'aujourd' hui et besoins de demain. Bordeaux: Université de Bordeaux 11, 8194. Penfield, W. (1964) The uncommitted cortex: the child's changing brain. Atlantic Monthly 214, 7781. Quetz, J. (1987) Child-adult differences in second language learning: Reformulating the approach. In W. Lörscher and R. Schulze (eds) Perspectives on Language in Performance. Tübingen: Gunter Narr Verlag, 11741189. Schneider W. and Uhl, Ch. (1990) Metagedächtnis, Strategienutzung und Gedächtnisleistung: Vergleichende Analysen bei Kindern, Jüngeren Erwachsenen und alten Menschen. Zeitschrift für Entwicklungspsychologie und Pädagogische Psychologie 22(1), 2241. Swain, M. and Lapkin, S. (1982) Evaluating Bilingual Education. A Canadian Case Study. Clevedon: Multilingual Matters.

< previous page

page_72

next page >

< previous page

page_73

next page > Page 73

Triangulation and Trilingualism Reitze J. Jonkman Fryske Akademy, Doelestrjitte 8, 8911 DX Leeuwarden/Ljouwert, The Netherlands Abstract. My study concerns the urban vernacular 'Leewarders' and the two other languages spoken in Leeuwarden Dutch and Frisian. The description of the communicative function of the urban vernacular in relation to those languages is intended to be a contribution to the sociology of language in the tradition of Multilingualism in the Barrio by Joshua Fishman et al. (1971). My study is executed by means of several methods and techniques: interviewing, a survey, a variant of the matched-guise experiment and participant observation. The approach of investigating the same phenomenon from different methodical angles is known as triangulation. The idea behind this approach is that every single method has got its own merits and demerits. The outcome from dissimilar measurements of the object can be mutually compared and together they lead to more reliable results. I will try to illustrate this by comparing a selection of a preliminary outcome obtained by four different methods. An additional goal of my paper is to give a description of the social position of the urban vernacular linked to the variable 'education'. Before I start with the display of this part I think some background information concerning the language situation in a historical perspective may be in order. Historical and Linguistic Introduction Friesland is one of the 12 provinces in the Netherlands. It is the only officially recognised bilingual province. About 75% of 600,000 people living there are able to speak the minority language Frisian next to the state language Dutch (Gorter et al., 1988). In the capital counting 80,000 inhabitants, Leeuwarden (Dutch) or Ljouwert (Frisian), the urban vernacular 'Leewarders' is spoken next to the two languages. Half of the city population can speak it. Until the middle of the 16th century Frisian was in use alongside Dutch as a language for official affairs. Thereafter Frisian was abolished as such and Dutch was used as the sole language for officialdom. The main language of the Low Countries also became the language of the urban highest class during that period. Before the end of 16th century a mixture of Dutch

< previous page

page_73

next page >

< previous page

page_74

next page > Page 74

dialect and Frisian came into being, called 'Leewarders'. It was spoken by the middle and lower classes. The syntax, morphology and phonology of this language are not Dutch, its lexicon is mainly not Frisian. It is considered a Dutch dialect. Until World War II Dutch remained the only high status variety, Frisian and the urban vernacular being the relatively lower status varieties. After the war the revival of Frisian as a high(er) variety started. Today it is used as an official language once again (though in a limited way). The use of the Ljouwert vernacular is restricted to the lowest domains. In Table 1 a few examples are given to give an impression of the linguistic identity of Leewarders compared to Frisian and Dutch. Introduction on Triangulation After this historical and linguistic introduction, I now return to the methods I used and the results they yielded. In most scientific research only one or two methods are used to study a Table 1 Contrastive linguistic comparison of English, Frisian, Leewarders and Dutch English:

The little

Frisian:

It

Leewarders:

Ut kleine

Dutch:

Het kleine

lytse

man mantsje mantsje

said to

the

girl,

sei

tsjin

it

famke,

sei

teugen ut

meiske,

zei tegen het meisje, mannetje English: he would see what his father had Frisian: dat er sjen soe wat syn heit dien Leewarders: dat er sjen sú wat syn fader deen Dutch: dat hij zou kijken wat zijn vader had Frisian/Leewarders < > Dutch Syntax: sjen soe / sien sú < > zou kijken dien hie / deen had < > had gedaan Morphology: man-tsje < > mann-etje fam/meis-ke < > meis-je #deen / #dien < > ge-daan Phonology: sei < > zei syn < > zijn fader < > vader Frisian < > Leewarders/Dutch Lexicon: lytse < > kleine famke < > meisk(/j)e heit < > f(/v)ader

< previous page

done. hie. had. gedaan.

page_74

next page >

< previous page

page_75

next page > Page 75

certain phenomenon. The researcher tries to choose the method which is best fitted to the nature of the problem under investigation, but even if the method chosen is adequate, it still has its own bias. If various methods are used we become aware of the bias of every single method. The totally different kinds of information obtained can supplement each other, because we do not know how to interpret some of the information unless we can couple it with other information, or because we need a crosscheck to verify the validity of the observations. The dissimilar measurements can also lead to more reliable results. Such an approach from different methodical angles is known as triangulation (Denzin, 1970). Other names are 'multi' or 'multiplemethod' or 'multiple-instrument research'. I studied my object in a big random sample for the survey and matched-guise, selected individuals for the interviews and several small selected groups for the participant observation. The participants are all part of the total population of the Frisian capital; the people who live or are in Leeuwarden. In the following part of my paper I will present: firstly each of the four methods used with a short explication of its merits and demerits. I will not give this topic an exhaustive treatment; it is meant as a reminder of some aspects of the method; secondly some aspects of the actual situation of research, like the number and characterisation of informants; thirdly the results obtained. After having presented the four methods in this way, I will come to a short discussion of the overall outcome. Presentation of the Methods and Techniques Used, and the Results Obtained Interviewing On the one hand, the interview is a special social situation; it is not the situation you want to investigate. The effect of social desirability on the answering by the respondents is a special problem in connection with negative social issues. There is a danger for the reliability. On the other hand, by means of the interview with open, non-directive questions the respondents have got the possibilities to give their opinions from their points of view, which can be clarified or deepened by new questions. So the validity can be heightened, though the mutual comparison with other respondents is more difficult than mutual comparison of survey-respondents. The open interview is a good instrument to explore what aspects are important to the issue. I have interviewed 30 respondents, most of them having been born in the 1920s or earlier. They were selected on their social backgrounds and occupations, for instance journalists, and knowledge on the pre-war period

< previous page

page_75

next page >

< previous page

page_76

next page > Page 76

in Leeuwarden. I used a cassette-recorder for taping the interview. During the interviews they told me about the language use in the school yard. Some general conclusions can be derived: With regard to the primary schools: people told me about differences between schools. In the inner-city there were two special schools for high or middle-class children (of doctors, lawyers, high shopkeepers). No vernacular and no Frisian was spoken there. At the edge of the centre there was a working-class school, and almost all other schools were working class as well. The big majority used vernacular and hardly any Frisian or Dutch. In those days some secondary schools were only meant for the high and middle classes (Gymnasium/Highschool(HBS)). So only Dutch was spoken there. I will give one example of more detailed information. This was told to me by a teacher at a Roman-Catholic school in the innercity, born in 1901. RJJ: What was the language use in the school yard in those days? The teacher: At the Vitusschool there we had two kinds of public; the one kind of the well-to-do shopkeepers, Brenninkmeijers (C&A) and that sort of people, the other kind of the lowest strata of the community. There was a special cause for that. In 1924 a special school was founded for the lower class by the Catholic schoolboard. In 1934 this school was closed down again. Some of the pupils were sent to the Vitus-school. Then we received (he was hesitating when formulating, RJJ) the lot. So we had two kinds of public. Let us call them, just for convenience sake, the poor half and the rich half. The poor half spoke as good as only the urban vernacular. RJJ: Also in the classroom? The teacher: Well no, well not if they were answering a question, not then of course. One tried to discourage the use of the vernacular as much as possible. Actually it was not allowed, you know, because they should learn proper Dutch as a matter of fact. (...) About the poor half again, actually only the vernacular was spoken in the school yard, as good as only. It was a rather coarse language, you know. That's quite obvious. The other half was trying, now and then, to speak in a proper way, but were not successful all the time. As matter of fact a lot of vernacular words were mixed into it, but it was a bit more decent. RJJ: You would not call that 'Leewarders', the way of speaking of the rich half? The teacher: Well, let's call it 'mixed language'. Overall conclusion of the interviews: Dutch has got a high (but small) position, because it was only spoken in the school yard of the élite, the

< previous page

page_76

next page >

< previous page

page_77

next page > Page 77

vernacular a very low (but big) one, because it was spoken only by the lower class schoolchildren and Frisian has no position at all because there was no report on the use of it in the school yard. The survey (closed questionnaire) On the one hand, the survey, as a standardised, scheduled interview, has the same disadvantage of the effect of social desirability on the respondent's answering. By means of presenting a limited number of reply alternatives it is made easy for respondents to choose. They are directed by the alternatives. On the other hand, because all respondents have had the same alternatives to choose from, the outcome of different groups can be compared. It is also made easier to interview a big sample by the standardised questionnaire. When the respondents of that sample are selected at random the outcome is representative for the population as a whole. I started with a random sample out of the city registration of January 1, 1988. The number of sample units was 500 inhabitants from the age of 12. About 20 interviewers, mostly students, ended up with a 75% response, so 375 respondents. 333 of them had Dutch, Frisian or Leewarders as their only first language. People with more than one first language and other varieties are excluded in the next presentation to get a clear understanding of the mutual relations. As I focus on the variable 'education' I show the results of the question on education: 'What is the highest level of education you finished?' (See Table 2) In Table 3 the high position of the Dutch group is confirmed. More than half of the Dutch speakers, 54%, have a high education, a much higher percentage than the speakers of Frisian, 32, and the speakers of Leewarders, 22. The proportion of Dutch speakers with only primary school is very small, 9%, in contrast to the two other groups: Frisian 35% and vernacular 26%. The Dutch speaking group differs clearly from these two groups. The difference between the Frisian and the vernacular group is less pronounced. Overall conclusion of the survey: The highest educational level is attained by the Dutch speaking group, the lower positions are held by the Frisian and the Ljouwert vernacular group, though the former group has got a somewhat higher proportion with a high education. A variant of the matched-guise experiment ('matched-label') The experimental procedure most widely utilised for investigating speakers' views on speech is known as the 'matched-guise technique' (Lambert et al., 1960). The procedure is as follows: Judges are told that they are going to hear voices of different speakers on tape, reading or telling the same passage of verbal material, and are asked to evaluate the speakers on a rating scale or bipolar adjective scale (the so-called 'semantic differential').

< previous page

page_77

next page >

< previous page

next page >

page_78

Page 78 Table 2 Classification level of education into LOW, MIDDLE and HIGH only primary education (1) }LOW technical and vocational training for 1216 years old } 'LBO' (2) }MIDDLE advanced primary education 'ULO' (3) } lower general secondary education 'MAVO' (3) } technical and vocational training for 1618 years old } 'MBO' (4) } higher general secondary education 'HAVO' (4) } pre-university education 'VWO' (4) }HIGH } technical and vocational training for 18+ years old } 'HBO' (5) } university (5) } The figures refer to the presented reply alternatives which I reduced later to the three categories LOW, MIDDLE and HIGH. This categorisation is used to display the differences between the groups. Table 3 Educational level of linguistic groups in percentages L1 Level LOW MIDDLE HIGH Dutch 9 37 54 Frisian 35 34 32 Leewarders 26 52 22

total 100 100 100

n (175) (89) (69)

The speech is actually produced by one person using realistic guises of different languages. So the judgements are about the languages spoken (or better, speech groups) and not about the presented speakers, and are therefore an indication of the attitudes towards those languages. In my study I did not present, for practical reasons, a tape with three voices to the judges, but three labels on paper; the names of the three varieties written in that variety: 'Nederlands' (Dutch), 'Frysk' (Frisian) and 'Leewarders' (the vernacular), so I used what might be called 'matched labels'. I asked the judges to evaluate these varieties on fifteen linguistic characteristics by means of bipolar scales (seven-point), for instance: 'educated |-|-|-|-|-|-|-| not educated'. The word 'experiment' refers to a laboratory situation, and in a way, this

< previous page

page_78

next page >

< previous page

page_79

next page > Page 79

is one. The matched-guise technique isolates spoken language from its social context and my version of the experiment, presenting just a label, even isolates it from spoken language. The evaluation on separate linguistic traits is not natural either. Only stereotypes are elicitated. On the other hand, like the standard questionnaire, the scores can be compared very easily, which is useful for the mutual comparison of the three varieties. Factor-analysing makes it possible to reduce separate traits to a few dimensions, which are more realistic. And furthermore, stereotypes do play a role in contacts between people. Now I am going to present the evaluation of the three varieties on the linguistic characteristics: 'educated', 'distinguished', 'businesslike', 'civilised' and 'elaborate'. Together they are part of the so-called status dimension (Ryan, 1979). Statistical validation is found by a factor analysis (Varimax Rotation-Kaiser Normalisation). The judges were the same as the respondents of the survey, so a random sample and representative for the population. If we watch the mean scores of the Dutch speakers in Table 4, you can see Dutch is evaluated higher than Frisian and Frisian higher than the vernacular on the trait 'educated'; the mean score 5.43 for Dutch is significantly (decided by the student's t-test) higher than 4.40 for Frisian and this score is again significantly higher than 2.78 for Leewarders. The scores for the other associated traits 'distinguished', 'businesslike', 'civilised' and 'elaborate' have the same order: Dutch >> Frisian >> Leewarders. If we look at the scores of the Frisian speakers in Table 5 we see the same order for all traits as in Table 4, only the differences made by the Frisian speakers between Dutch and Frisian on the traits 'civilised' and 'elaborate' are not significant. Table 6, the mean scores of the vernacular speakers, is a copy of Table 4. Table 4 Mean scores and t-values for the varieties evaluated by DUTCH speakers (N = 157) Trait Mean score t-value 'Nederlands' 'Frysk' 'Leewarders' N > F > (Dutch) (Frisian) (vernaculers) F L 6.69** 10.51** educated 5.43 4.40 2.78 6.44** 11.20** distinguished 4.51 3.55 2.04 11.35** 6.17** businesslike 5.27 3.38 2.64 6.63** 11.79** civilised 5.66 4.65 2.97 7.47** 6.69** elaborate 5.61 4.30 3.21 N > F means Dutch ('Nederlands') compared to Frisian ('Frysk') F > L means Frisian ('Frysk') compared to the vernacular ('Leewarders') degrees of freedom 156 **p < 0.001

< previous page

page_79

next page >

< previous page

page_80

next page > Page 80

Table 5 Mean scores and t-values for the varieties evaluated by FRISIAN speakers (N = 76) Trait Mean score t-value 'Nederlands' 'Frysk' 'Leewarders' N > F > (Dutch) (Frisian) (vernacular) F L 4.97** 8.08** educated 5.83 4.66 2.53 4.87** 7.00** distinguished 4.74 3.63 1.88 4.91** 6.19** businesslike 5.57 4.42 2.86 1.47 7.65** civilised 5.63 5.29 3.63 1.63 6.97** elaborate 5.32 4.87 2.93 degrees of freedom 75 **p < 0.001 Table 6 Mean scores and t-values for the varieties evaluated by LEEWARDERS speakers (N = 61) Trait Mean score 'Nederlands' 'Frysk' 'Leewarders' (Dutch) (Frisian) (vernacular) educated 5.90 4.66 2.72 distinguished 5.08 3.43 2.07 businesslike 5.63 4.10 2.95 civilised 6.26 5.00 3.50 elaborate 5.63 4.68 3.32 degrees of freedom 60 *p < 0.01 **p < 0.001

t-value N > F > F L 5.31** 5.80** 6.42** 5.40** 5.28** 3.99** 5.22** 7.09** 2.81* 4.66**

For all traits there is the same order again: Dutch >> Frisian >> Leewarders. Overall conclusion of the matched-label experiment: the relative evaluation on the status dimension by all three linguistic groups is positive for Dutch, in between for Frisian and negative for the urban vernacular. Participant observation On the one hand, the situation cannot be manipulated as in interviews and experiments. Aspects which are important for the researcher may not show up. In the 'natural' setting a lot of variables cannot be isolated. Behaviour can be seen, but not the motives which are behind it. So if you do not have the opportunity to ask for an explanation, the behaviour can be hard to interpret. It is also possible that the presence of the researcher may have unintended effect on the behaviour of the one observed. Seen from a quantitative point of view the biggest disadvantage is that the selected

< previous page

page_80

next page >

< previous page

next page >

page_81

Page 81 sample is not representative for the population. On the other hand, when playing an accepted role inside a group the researcher has access to the non-reduced reality. Things are said and done in a way which are needed in that setting and are not reports on what is said and done. The validity is high. I visited several types of schools: 1. 'LTS'; technical training for 1216 years old (lowest level) only boys 2. 'MAVO'; lower general secondary education 3. 'HAVO'; higher general secondary education 4. 'VWO'; pre-university education The types 2, 3 and 4 were in one so-called school community (1,700 pupils). I spent a month in the classes of these types. Type 1 was a separate school with only 160 pupils. In this school I spent one week for observation. I especially joined classes in which pupils were allowed to talk to each other, like gymnastics and handicraft. I was introduced to the pupils as a student teacher, a very acceptable role for someone walking around in a school building. So the pupils did not know my real intentions until the last part of the observation, when I started to ask them questions about their language behaviour. The results of my participant observation are summarised in Table 7 in which vernacular use in a type of school is crosstabbed with sex, appearance, or non-appearance of urban vernacular use in classroom, age and intensity (i.e. degree of linguistic purity). In the column of the variable 'sex' on the line of the LTS, the school for technical training, the 'o' means that there are no girls in this type of school. The boys do speak the vernacular. For the other types of schools it is very clear that sex is a very determinant variable. It is only boys who are vernacular speaking, girls are not. It is a very well known contrast in sociolinguistics. The use of Leewarders in the classroom, even to Dutch speaking teachers, shows a difference between the school for technical training and the other levels of education. In the former school they are trained for low status manual labour for which no special linguistic capacities Table 7 Educational level and urban vernacular use Level Category Sex In Age male fem. Classroom 13 1416 LTS + 0 + + + MAVO + + HAVO + + VWO + +

< previous page

+ x x

page_81

Intensity + x x

next page >

< previous page

page_82

next page > Page 82

are needed. The boys are not motivated to learn Dutch. I think this also offers an explanation for the difference on the variable 'age' between the former and the latter schools. The boys at the latter schools just start talking the vernacular when they are in the second form. For them there is no social problem but a personal problem: the onset of puberty. They want to diverge from their parents who have forbidden them to speak the vernacular and they want to show their toughness to each other. The vernacular is an obvious vehicle for the expression of both feelings. All the same the intensity of the vernacular, this means the degree of linguistic distance from Dutch, is an indicator of some social dimension for the types of general education, but it is more subtle than speaking or not speaking Leewarders in the classroom. The MAVO pupils use the same pure forms of the vernacular as those at the LTS and in a dialogue Leewarders is used consistently. The HAVO pupils mix more Dutch forms in their speech and do not use it consistently; one sentence is in the vernacular, the next ones are not, then again some sentences vernacular, and so on. The VWO pupils speak Dutch mixed with an occasional sentence, a catchphrase or a filler in the urban dialect. Overall conclusion of the participant observation: The use and intensity of the urban vernacular is influenced by the variable educational level, age, sex and setting: The higher the educational level, the later and less intensive boys speak the vernacular in fewer settings. Summary, Discussion and Conclusions In this paper I have discussed various methods and techniques I used for my study of the multilingual situation in the capital of Friesland; triangulation and trilingualism as I called it in short. Interviewing, a survey, the 'matched-label' and participant observation have provided me with many results on the social positionindicated by the variable 'education'of the urban vernacular 'Leewarders' in relation to Dutch and Frisian. The combined results of the various methods all point in the same direction: the vernacular is in the lowest social position compared to the other two varieties. In the pre-war period Leewarders was the most spoken variety in the lower class school yards and absent on the high and middle class ones. The general education level of the vernacular group is much lower than the Dutch and almost the same as the Frisian group. This position, lower than Dutch and Frisian, was also expressed in the evaluation on the status dimension, so in the language attitude, and in the more frequent and consistent use in the lower types of schools, so in language behaviour. My goal was not only to give this description of the social position of the urban vernacular but I hope I have shown that the use of dissimilar measurements can lead to more reliable results. By pointing in the same direction of the various results I can be more sure about the validity of the overall outcome than it would be in case of conflicting results. Therefore I refer to two detailed examples

< previous page

page_82

next page >

< previous page

page_83

next page > Page 83

of strengthening results. In the interview the Roman-Catholic teacher used the label 'the lot', referring to children who spoke the vernacular and called Leewarders 'a rather coarse language'. The very low matched-label evaluation on the characteristics 'distinguished' and 'civilised' validated this report. The same teacher reported on the character of the language use by the 'rich half' children: 'The other half was trying, now and then, to speak in a proper way, but were not successful all the time. As a matter of fact a lot of vernacular words were mixed into it, but it was a bit more decent'.RJJ: 'You should not call that ''Leewarders", the way of speaking of the rich half?'The teacher: 'Well, let's call it "mixed language". I did not know what he meant by 'mixed language' exactly, because he could not give me any linguistic details. I think I was confronted with the same phenomenon at the secondary schools for general education during my participant observation: the intensity of vernacular use. My conclusions now after having executed the study of trilingualism by triangulation are positive. It is not only that the use of various methods had the effect of reinforcement of the overall outcome, which heightens its reliability, but it is also the possibility of mutual comparison, which clarified some unclear issues. I hope I have inspired you to triangulation, though I presented just a small selection of my studies on trilingualism in the Frisian capital, but that is to arouse your curiosity about it. References Denzin, N. K. (ed.) (1970) Triangulation: A case for methodological evaluation and combination. In Sociological Methods: A Sourcebook. Chicago: Aldine. Fishman, J. A., Cooper, R. L. and Ma, R. (1971) Bilingualism in the Barrio. The Hague: Mouton. Gorter, D., Jelsma, G. H., van der Plank, P. H. and de Vos, K. (1988) Language in Friesland. Ljouwert: Fryske Akademy. Lambert, W. E., Hodsgon, R., Gardner, R. C. and Fillenbaum, S. (1960) Evaluational reactions to spoken languages. Journal of Abnormal and Social Psychology, 60, 4451. Ryan, E. (1979) Why do low-prestige language varieties persist? In H. Giles and R. N. St Clair (eds) Language and Social Psychology. Oxford: Basil Blackwell and Baltimore: University Park Press, 14557.

< previous page

page_83

next page >

< previous page

page_ii

next page >

< previous page

page_85

next page > Page 85

Switching Between Seven Codes Within One Familya Linguistic Resource Bent Søndergaard Pädagogische Hochschule Flensburg, Mürwiker Strasse 77, D-2390 Flensburg, Denmark Abstract. Until now the extensive exploration of code-switching within multilingual research has focused on this phenomenon in idiolects or sociolects. In this paper, however, it is analysed as part of a 'familylect', e.g. code-switching is described as it occurs in the spoken language of a multilingual family with four members. Seven different language codes are involved, namely three languages in Scandinavia (Danish, Finnish, Swedish), two dialectal Danish sub-codes, and two European main languages (German, English). This complicated linguistic pattern leads to frequent shifts between all the seven codes involved: code-switching occurs daily. Through a linguistic and extra-linguistic analysis of the context in a great material I will tryfrom my point of viewto answer the three classical questions connected with code switching: when? how? why? I. The Seven Language Codes Code-switching is a feature which is easy to observe in multilingual speech, and is therefore one of the traditional objects of research within the extensive exploration of multilingualism. It has already been analysed in one of the classics of bilingualism, namely Weinreich (1953) and the interest in it has been on the increase since the 1950s, McClure (1981: 69). Normally this complicated linguistic phenomenon is described as part of an idiolect (Saunders, 1982: 11ff., 92ff. is a typical example) or as part of a sociolect (Poplack, 1987 may be mentioned as an example). In this paper, however, I will try to look upon code-switching from another point of view, namely as part of a 'familylect'. This means I shall describe it as it occurs in a multilingual family consisting of four members, each with a different linguistic background, namely informant V (linguist and language teacher),

< previous page

page_85

next page >

< previous page

page_86

next page > Page 86

X (also linguist and language teacher), Y (student), and Z (a 13-year-old pupil). The seven language codes involved are: A1: Rigsdansk, the official language of Denmark, which is the common code of the family and Ll for Y and Z, whereas V speaks it as a foreign language; A2: Sønderjysk, the local dialect of South Jutland, which must be considered a language in its own right because A1 and A2 are not mutually comprehensible. This code is widespread in the area where the family lives. Only Y is able to speak it fluently. To X and Z it is comprehensible, while V has some difficulties; A3: Morsingsk, a dialect in the north-western part of Jutland, the mother tongue of X. Apart from him, only Y has an (elementary) command of this code, which is not understandable to the speakers of Al and A2; B: Finnish, the mother tongue of V. Apart from her, only Y speaks this language (with a communicative competence), while X has an elementary receptive command of it; C: Swedish, Originally V and X were not able to understand each other's standard languages (A1 and B), and therefore they used C as a lingua franca (cf. Søndergaard, 1986: 123). Both of them have a good command of this code. It is also the case with Y; D: German, which to a certain extent is the working language of X. Y and Z, but not V, speak it as a foreign language; E: English, which is partially the working language of V. X, Y, and Z speak it as a foreign language. II. Code-switching This complicated linguistic pattern leads to frequent shift from the main code to the six other codes in the speech of all four informants (except for switching to A3, which nearly without exception is used only by X). To put it another way: code switching is 'infectious'. This forms a unique 'familylect', which often surprises others. In the surrounding sociolects it is only possible to find switching to A2, D (and E). III. When does Code-switching Occur? A prerequisite for such extensive switching between so many codes is that the speakers are very conscious of the boundaries between the different codes: they have to know exactly where switching from the main code to one of the others occurs. Otherwise this language situation would create an unconscious mixed code, a sort of 'pidginised' language. This is especially important between A1, 2, 3. It might therefore be necessary to 'signalise' if one supposes that one's partner is switching unconsciously, as the following example 1 shows:

< previous page

page_86

next page >

< previous page

page_87

next page > Page 87

1. (A2).Z: Der er længe nok julelys. X: Nå, men ved du for resten, at udtrykket længe nok ikke er rigsdansk? Z/forbavset/: Jamen, er det så sønderjysk? .. Næ, det vidste jeg ikke. Jeg bruger det da selv. X: Det er i orden, men hva hedder det på rigsdansk? Z/usikkert/: Nok, vel? X: Ja, for længst nok. In other cases the speaker himself marks the switching by means of signals such as: som vi sier, som det hedder, det såkaldte etc., as the following examples indicate: 2. (A3).X/forarget/: Er det ikke, hva vi på morsingsk kalder lampanniværk. V: Hva er det? X: Noget usolidt noget, noget sjuskeri. 3. (A3)V: Er hun sådan én, der tiltrækker sig alle slags sygdomme? X: Næ, hun er, hva der på morsingsk hedder en syllekjæsten, én der altid klager sig. 4. (B).V: Jeg prøver at finde en anden ... nå ja, vi sier munalukko .. jeg ved i grunden ikke, hvorfor vi sier 'æglås'/ for: hængelås/. 5. (B). V: Det er da vældigt, som deres morfar er ketterä, som vi sier på finsk, rask til bens. 6 (D).Z: Hvorfor laver man overhovedet pulverkaffe? X: Fordi det er, hva tyskerne kalder magenfreundlich/ for: god mod maven/.Cf: IV 1, V 6, VII 4,6,9, VIII 1,4,5. It is difficult to see exactly from the context what the speaker's aim is when he marks the switching in this way. Is he apologising for the switching (= impure code)? Is he warning his partner to be conscious of the switching? Or might there be other reasons?Hasselmo (1979: 180) uses the term 'markers' for pauses, inflections of the voice, and phrases such as 'as we say in English', which are used by the speaker to underscore that he is switching code, but I interpret pauses another way, see below. IV. Alternating Codes Another important prerequisite is that the family members have so good a command of all the alternating codes that they are able to understand the new context or to infer the meaning. Otherwise switching is dangerous, but if a speaker is in doubt, he may put a question (or his partner may do so), or he may translate the switched word(s) into the common code (or into another code, cf. example 4), as the following examples show: 1. (A3).X: Så tar du de nye bukser på straks. Jeg hader at se dig gå rundt og pløse i tøjet på den måde.... Ved du for resten, hva det morsingske ord pløse betyder? Z/usikkert/: Nå, drive? X: Nej, at ha tøjet i uorden. 2. (A3).Z: Thyge har brugt 200 kroner på én dag. X: Ja, han er rust med penge. Z: Hva mener du med rust med? X: Nå, ja, at man strør om sig med penge.

< previous page

page_87

next page >

< previous page

page_88

next page > Page 88

3. (B).V: Til påske henter vi i Finland altid pajunkissat, som vi sier, 'pilemisser', ind. 4. (B).V: Hun sae, at selv om manden er syg, så er han gudskelov kiltti 'snäll'. 5. (C).Z: Hva er Blå Kors for noget? X: En nykterhetsrörelse. Z: Hva er det? X: En afholdsbevægelse. 6. (D).X: Det er vanskeligt at komme dertil, fordi stedet ligger ... ja, forstår du udtrykket jenseits von Gut und Böse? Z: Nej, hva betyder det? X: Egentlig hinsides godt og ondt, altså meget afsides.Cf: 111 2,3,5, VI 1,3,6, VII 4, VIII 6. V. Patterns in the Use of Codes After analysing a comprehensive amount of materialespecially on the lexical levelover the past two years, I can state that codeswitching does not occur haphazardly in this family's speech. I shall now try to find some of the main patterns in the use of the codes, but I wish to point out beforehand thatat the present stage of researchit is not always possible to find out why a person switches in some situations but not in others, nor can one always predict the switching. There is still something 'mysterious' about this linguistic phenomenon. As it has often been pointed out in previous research, however, very much depends on the subject, the situation, and the interlocutor. I will give some examples of this: 1. (B).V/taler om finske forhold/: Den ene hed Rauha, men hva hed den anden? .. Nå det må stå i sukukiria/ for: slægtsbog/. 2. (B). Z: Jeg tror nok, Timo blev overrasket over, at joulupukki kom/ for: julemand/. 3. (C).X/om en svensker/: Du kan tro, han blev förbannad, da han hørte det/ for: gal/. 4. (C).V/svensksproget kontekst/: Da han så hade dumpet mig, sae han, jeg ikke sku bli ledsen, men det du han jo sagtens sie/ for: ked af det/. 5. (D).X: fortæller om en oplevelse på sin arbejdsplads/ Z: Ja, disse tyskere: Leistung, Leistung! /for: 'prïstation'/. 6. (D).X: Det er vigtigt, at disputatsforsvaret finder sted før det såkaldte Referendariat/ for: 'pædagogicum'/. 7. (E).V/i forbindelse med engelske goester/: Hvor meget the ska der i, hvis vi vil ha en hel pot/ for: kande/? These examples show that if a person is relating something which he has experienced in a context where another code was used, it facilitates the switching to this code, perhaps because he does not always remember the equivalent in the main code.

< previous page

page_88

next page >

< previous page

page_89

next page > Page 89

VI. Switching Due to Lack of Vocabulary The subject, the situation, and the interlocutor cannot of course automatically explain every switching. I have also observed that the speaker's mood plays a great role, both negatively and positively. For instance, if he is tired, he cannot find the right word in the 'right'code, and if he is 'fit', he is linguistically creative. This means he is playing with his whole linguistic repertoire, codeswitching included. This observation leads me to the following consideration. In Søndergaard (1984) I tried to put forward a typology of code-switching which I still think is valid. In my opinion, the most important point in this typology is the fundamental differentiation between code-switching as a psycholinguistic 'deficit phenomenon' and a ditto 'surplus phenomenon'. Let me explain this in detail. The first category is an enforced switching due to a lack of vocabulary. That means that at a particular moment the speaker cannot remember the word(s) in the code being used for the conversation in progress, but in one of the other codes, and he then transfers directly from the second code instead of trying to paraphrase in the main code. The context is characterised by pauses, especially before the switching. The linguistic analysis shows that the speakers themselves often remark that they cannot remember the right word(s) in the 'right' code, or they ask a question, appealing for assistance. A rich non-verbal communication is also taking place, for instance the speakers are trying to obtain eye-contact, asking for help. This form of switching (most typical for V to code B and E) can be illustrated with the following examples: 1. (A2).Y: Og så spjarrer bilens hjul, som de sier på sønderjysk. Hva hedder det på rigsdansk? /d.v.s. køre rundt uden vejgreb/. 2. (B).V: Og så har jeg hende, jeg kalder 'sorgens barn', hva kalder manegentlig en sådan surunlapsi på dansk? 3. (B).N: Da begge mine forældre har ... nå, hva sier man på dansk? .. maahenki 'jordånd' sier vi på finsker det mærkeligt, jeg ingen har. 4. (B).V/om en finsk dame/: Hun er nu stor ... ja, hva sier man på dansk? .. på finsk sier vi suurellinen, og det er typisk karelsk. X: Vi sier vel: stor i slaget. 5. (C). V: Tænk, hun har købt en regnfrakke, der ... nå, hva sier man? .. glänsar/ for: skinner/. 6. (D).X: Dette publikum er langt mere ... anspruchsvoll .. krævende. 7. (E).V: Jeg må prøve ... hva hedder det? .. at nedtrykke min tilbøjelighed. X: Undertrykke. V: Repress hrdder det pä engelsk. Jeg sier det bare på engelsk, når jeg synes, det er bedst.Cf: III 4, VIII 6.

< previous page

page_89

next page >

< previous page

page_90

next page > Page 90

VII. Voluntary Switching The other main type is a voluntary switching, used to 'colour' the language. This means that the speaker prefers 'to use words from other codes, because he finds them more suitable, confidential, full of connotations and the like, even if he would easily be able to avoid switching. Most of the examples in the material belong to this sophisticated type of switching. (There are no specific linguistic or extra-linguistic characteristics, for instance no pauses in connection with the switching. The words from the other codes are totally integrated in the main code.) Examples: 1. (A2).V: Kommer I hjem allerede nu? X: Ja, det har regnet hele tiden. V: Nå, her har det kun smusket/ for: støvregnet/. 2. (A2).X: Niels, luk æ pesgaf i!/ for: knap bukserne/. 3. (A3).X: Pas nu på, du ikke blir pus-i-mog!/ for: taber/. 4. (A3).V: Jamen, er Inger da ikke meget yngre end brødrene? X: Jo, hun er en klat-ibåg-æter, som vi sier på morsingsk. V: Hva er det? ... nå, en efternøler. 5. (B).Z: Det er da helt hullut at lave en omfartsvej omkring Tønder/ af finsk hullu 'skør' med tilføjet dansk neutrums-t/. 6. (B).V: Det hele var jo blevet en susi, som vi sier på finsk/ om noget mislykket/. 7. (C).V: Og da jeg så endelig fik øje på ham, si hev jeg ham til side, og vi gik i armkrok hen ad gangen/ for: arm i arm/. 8. (C).Y: Det er da mærkeligt, de har ændret vejen mellem Tønder og Abild, den var da så rak/ for: lige/. 9. (D).X: Det hele er én stor Teufelskreis, som det så malerisk hedder på tysk. V: Nå, hedder det sådan/for: en ond cirkel/. 010. (D).X: Hvorfor råbte han sådan? Z: Nå, han var vel besoffen/ for: fuld/. 11. (E).Y/meget højt/: Det gælder kun om én ting, og det er survival/ for: at overleve/. 12. (E).Z/spøgende/: Der er ædelse on the table/ for: på bordet/. I do think that this type of code-switching can be looked upon as a kind of linguistic creativeness: the speaker is trying to find le mot propre, not only with regard to the denotations, but also to the connotations of the words, and the latter is the more important. Therefore he uses his whole linguistic repertoire. In this connection code-switching is one of the resourcesas indicated in the subtitle of this paper. The great amount of code-switching in this family may perhaps be due to the fact that linguistic creativeness plays an important role in the family's conversation, for instance playing on words, as shown in the following example: 13. (B).V: Nu husker jeg, hva den ungarske pølse hedder: sigøjnerpølse. X/spøgende/: Det er måske, fordi der er mustalaiset i den. V: Ha, ha, mustalaisen makkara, nej, sådan en pølse ville ingen i Finland spise.

< previous page

page_90

next page >

< previous page

page_91

next page > Page 91

As part of the play on words it is typical that in some examples the switching which was initiated by one speaker goes on to the next: 14. (B).V: Der står i avisen, der har været meget is i Salon-joki. Bare sommerhuset ikke har taget skade. X: Jamen, det ligger ikke ved nogen joki/ for: å/. 15. (D).X: Vil du ikke cykle en lang tur på din nye cykel i morgen? Y: Da ikke, når der er festival i byen! X: Jamen, er dét da noget for gebildete Leute? Y: Jeg er da ligeglad med, om det er gebildet eller ej/ for: dannet/.Cf. III 2,3 VIII. Spontaneous Switching Of course the above mentioned dichotomy (linguistic surplus versus deficit) cannot exlain all the examples in the material. I should therefore like to mention some other observations from the material to supplement this. In my first analysis of code-switching (Sondergaard, 1984: 7f). I had already found out that spontaneous switching is often connected with emotions, especially in the case of lively informants. I do not know how this can be explained in linguistic terms, but perhaps the following rather simple common sense explanation may be sufficient. When one is angry, upset, surprised, shocked etc. one does not have much time to code one's language. Therefore one just takes the first word(s) one finds, no matter in which code. Examples: 1. (A3).V: Der er én portion fromage til i morgen. X/hånligt/: Det er jo kun en niesful, som det hedder på morsingsk/ for: en 'klat'/. 2. (A3).V: La være med at trykke på bordpladen. Den kan ikke holde til det. X/irritereth Mæ om det/ for: Det ska jeg nok selv bestemme/. 3. (B).X/overrasket/: Herra jumala! I dag har vi fået to aviser i stedet for én/ for: Herregud!/ 4. (B).V/opbragt/: Ja, la os bare sie, det er min skyld, så kan I to føle mielihyvä, som vi sier på finsk/ for: tilfredshed/. 5. (D).X/vredt/: Jeg kender kun ét ord for noget sådant, og det er det gode tyske ord Schlamperei/ for: sjuskeri/. 6. (E).V/forarget/: Vidste du, hun var sådan en turncoat... hva hedder det på dansk? .. vi har ikke ordet på finsk/ for: vendekåbe/.Cf. III 2, IV 1, VII 11 IX. Conclusion Summing up, I wish to point out two things about code switching in this particular 'familylect':

< previous page

page_91

next page >

< previous page

page_92

next page > Page 92

1. Some of the switchings are linguistic 'snapshots' which are recorded only once in the large amount of material, whereas others are well established (cf. Hasselsmo, 1970: 179: integrated forms). 2. In most of the examples the switching only consists of single words. I am aware that not all researchers would call this code-switching, but borrowing. Gumperz (1982: 66), for instance, defines borrowing as 'the introduction of single words or short, frozen, idiomatic phrases from one variety into the other. The items in question are incorporated into the grammatical system of the borrowing language', and code switching as 'the meaningful juxtaposition of what speakers must consciously or subconsciously process as strings formed according to the internal rules of two distinct grammatical systems.' It is, however, generally admitted that it is very difficult to distinguish between code switching and borrowing, as stated by for instance Poplack (1987: 55). See also the arguments pro and con this distinction by McClure (1981: 70f). To me, it does not seem sensible to divide the same linguistic phenomenon into two parts on the basis of the degree of integration, and personally I prefer a rather simple definition, put forward by Valdés (1981: 95) 'In essence, codeswitching may be defined as the alternating use of two languages at the word, phrase, clause, or sentence level'. Note 1. In the examples code-switching has been marked, but the sentences have not been translated into English, because from experience I know that translation does not help much with understanding code-switching. You have to understand the codes involved if you require more than a superficial understanding. References Durán, R. P. (ed.) (1981) Latino Language and Communicative Behavior. Norwood, New Jersey: Ablex. Gumperz, J. J. (1982) Discourse Strategies. Cambridge: Cambridge University Press. Hasselmo, N. (1970) Code-switching and modes of speaking. In G. G. Gilbert (ed.) Texas Studies in Bilingualism. Berlin: de Gruyter. McClure, E. (1981) Formal and functional aspects of the codeswitched discourse of bilingual children. In R. P. Durán (ed.). Poplack, S. (1987) Contrasting patterns of code-switching in two communities. In E. Wande et al. (eds) Aspects of Multilingualism. Uppsala: Acta Universitatis Upsaliensis, Studia Multiethnica Upsaliensia 2. Saunders, G. (1982) Bilingual Children. Clevedon, Avon: Multilingual Matters. Søndergaard, B. (1984) Code switching in bilingual speech. An interpretation within a linguistic and extralinguistic framework. Nordic Linguistic Bulletin 8(12), 59. (1986) Code switchingilmiö suomenkielisissä olosuhteissa. Fenno-Ugrica Suecana 8, 12334. Valdés, G. (1981) Codeswitching as deliberate verbal strategy. In R. P. Durán (ed.). Weinreich, U. (1953) Languages in Contact (9th Printing 1979). The Hague: Mouton.

< previous page

page_92

next page >

< previous page

page_93

next page > Page 93

The Sámi Language: Pressure of Change and Reification Marjut Aikio Department of Social Sciences, University of Lapland, School of Language and Literature, University of Tromsf Abstract. The author analyses dominant trends in the linguistic situation of Finnish Sámi. Her aim is to discuss these trends as a component of indigenous peoples' problems with linguistic inequality. The crucial question is whether the changing international status of indigenous people will lead to concrete political decisions in Finland. Three factors contribute to Sámi language shift: an overall decline in the status of native language, an assimilationist school system, and a lack of functional contexts to use the Sámi language. Present school instruction in Sámi is too insignificant to counteract the pressure of the mainstream society. Parallel to this is the process of language reification: the written part of the language becomes dehumanised and moves beyond the reach of its users. The author emphasises that the language must not be seen in a position too detached from the Same culture. The close relationship between linguistic, political, social, economic, and territorial rights must be recognised. Introduction Two years ago I published the results of a lengthy research project that analysed language shift and language death in Sámi language buffer zones (Aikio, M., 1988). The subject is sensitive. Some Sámi teachers branded me as a promulgator of destruction and death. I was given the nickname of 'the ambassador of a dying language'. This means that now, when the preservation of the language is being attempted, such depressing research results should not be published. I have also experienced the cold shoulder of the local Department of Education: my research results apparently do not fit with the official picture. But another reaction has come: many Sámi who were born in the 1940s and 1950s and who have lost either partially or entirely their language have told me that they feel that they have been absolved from the guilt of letting

< previous page

page_93

next page >

< previous page

page_94

next page > Page 94

their language die. They understand themselves better and do not feel such a great burden of guilt for the abandonment of their mother tongue. All this encourages me to proceed in my study of the present situation of the Sámi language in Finland. I am especially interested in this subject as a part of the problem of Sámi people as an indigenous people, ethnocentrism, and linguistic inequality. In conclusion, I discuss one of the more rarer forms of linguistic inequality, which I call language reification. According to the current international definition, an indigenous people is characterised by pre-existence, non-dominance, dissimilarity of culture, and self-identification (Aikio, P., 1990). The view of western people and industrial cultures is generally dominated by a dualistic dogma in which the indigenous people are viewed with the eyes of colonial conquerors who are seen as conquering a dying culture (Rivas-Rivas, 1987). This evolutionist bias suggests that native languages, one of the core values (see Smolicz, 1981) of cultures, are also likely to disappear. According to the ethnocentric way of thought, language is only useful as an object for research. Ethnic cultures and their products are seen as interesting museum relics and research objects. Languages, for example, must be recorded before they die. Sámi culture has been treated in this fashion. In recent times, the status of indigenous peoples and their cultures and languages has changed remarkably. A concrete indication of this is the recently revised indigenous peoples' section of the ILO convention. Its goal is to preserve indigenous cultures, including languages. In an earlier version dating from 1957, the assimilation of indigenous cultures was seen as inevitable and the goal was set to soften its impact. The crucial question is whether this international agreement will lead to concrete political decisions in the Sámi area. The Sámi Language and its Speakers The Sámi language in Finland is sliding towards extinction. To prove this trend we can take research on any aspect or area of the language. The conclusion does not change even if we go outside the actual buffer areas of strong acculturation pressure (Aikio, M., 1976; Guttorm, 1986; Hokka, 1972; Magga & Syväjärvi, 1990; Müller-Wille, 1974). In estimates of the number of Sámi people, the conclusive factors are not demographic facts but the ethnic definition of Sámi people (see Aikio, M., 1988: 547). On the basis of the linguistically oriented definition used in Finland until 1991, the number of Sámi people has been inevitably decreasing because of language shift (Aikio, M., 1984; 1986). If the definition included all the recognised elements of ethnicity and their indicators the statistics would look quite different. Such a definition would show that there are people of Sámi origin outside the Sámi Home Area, whose language is Finnish due to their environment (Aikio, M. & Aikio, P., 1985; 1986). The established statistical figures of the Sámi population in Northern

< previous page

page_94

next page >

< previous page

page_95

next page > Page 95

Fenno-Scandia and USSR can be seen in Table 1 (Aikio, P., 1990). Considering all the uncertainties underlying these evaluations, the total number of ethnic Sámi can be estimated to 100,000. According to present estimates, less than half of the Sámi speak their native language. Language does not symbolise group identity in all cases. Although proficiency and the use of a language are fundamental elements of ethnic identity, in practice a natural coupling of these factors is not possible in a situation in which an indigenous minority lacks linguistic human rights, as Sámi people do in Finland. This is a case of unrealised human rights, which cannot be used to prove or disprove that language is a symbol of Sámi identity. Table 1 Sámi population in Northern Fenno-Scandia USSR Traditionally Minimum Actual Accepted Figures Figures Figures 1970s 1980s Finland 3,000 4,500 5,700 Norway 20,000 27,500 40,000 Sweden 10,000 17,000 20,000 USSR 1,500 1,500 1,800 Total 34,500 50,500 67,500

and Author's estimate

100,000

The Mechanism of Language Shift To date, language shift has been studied extensively in different parts of the world, although many of these studies deal with migrant minorities. The situation of indigenous minorities is less satisfactory, since they lack a linguistic home country. Language shift involves mechanism that occur in the same form all over the world and have a cumulative effect. Roughly speaking, language shift has two phases. In the first phase, the contexts of language use disappear; in the second one, the language itself disappears. Compulsory bilingualism is an intermediate phase in the transition to monolingualism in the dominant language. The process takes two to four generations. Three factors contribute to language shift. The key factor is the decline in the status of the minority language directly caused by the pressure from the majority culture. (For a detailed discussion of this pressure and the concept of linguicism, see Phillipson & Skutnabb-Kangas, 1986: 4323.) In the pressure of linguistic change, this has definite consequences for primary socialisation in the early phases of education within the family. Research from all over the world has shown that, instead of transferring the indigenous

< previous page

page_95

next page >

< previous page

page_96

next page > Page 96

language and cultural values to the next generation, the family tends to adopt a socialisation strategy which steers children towards the mainstream language. If parents are bilingual, the language shift may take the minimum period, two generations. Studies of language shift have repeatedly shown that the schools are a central assimilation factor. From a historical perspective, an overt assimilation policy seems to involve the same patterns in the schools everywhere: instruction in the dominant language, sanctions against pupils who speak their native language even in their leisure time, and physical separation of the pupils from their original community. In the phase of covert assimilation policy or indifferent attitudes towards indigenous people, the small positive steps taken by the school system often do more harm than good. On the other hand, the factor launching language shift at the family level is the parents' anxiety over the difficulties and labelling to which the child might otherwise be subjected at school. At its worst, such a situation has led to a tabooan irrational belief whereby people feel it is forbidden and, in fact, impossible to speak Sámi. Experience has shown that the Sámi find it extremely difficult to violate this taboo. Most older Sámi refuse to speak the language to children because they have been indoctrinated by the school system into believing that the teaching of Sámi weakens the children's knowledge of Finnish. The third set of factors furthering language shift is the lack of functional contexts for the use of the primary language. This chiefly concerns public media, administration and public services. In Finland history is repeating itself: the Sámi language is in the same situation in which the Finnish language found itself in the 19th century. With the strengthening of the Sámi identity as a result of revitalisation efforts, the symbolic value of the Same language has grown, but its real value in communication still remains insignificant (e.g. Asp, 1965: 241). In my own research (Aikio, M., 1988), I measured the language shift process for both speaker and listener by village, individual, social network and interaction, sex and age. This study exemplified how language ceased to be transferred from generation to generation. Figure 1 shows the dramatic decline in the use of the Sámi language over time. The biggest gap between children and youth occurs in the 1940s: people still spoke with youth using both languages while they spoke almost exclusively Finnish with children. From the 1950s on, no one uses Sámi with children. This was caused primarily by the established school environment: the language was Finnish and it was more or less overtly recommended for exclusive use when speaking to children. A Specific Case: Sámi Schooling The key role of schools as an assimilation factor has been gradually recognised. This has taken a concrete form in efforts to develop teaching

< previous page

page_96

next page >

< previous page

page_97

next page > Page 97

Figure 1 Language choices in discourse situation according to the age of the listener from the 1910s to the 1970s as percentages of all material Symbols in the columns: black = Sámi, diagonal lines = Sámi and Finnish, dotted = Finnish and Sámi, white = Finnish (Aikio, M., 1988: 203)

< previous page

page_97

next page >

< previous page

page_98

next page > Page 98

of and in Sámi since the 1970s. In its present form the Sámi teaching has stabilised at a level which may seem sufficient to guarantee its future in the short term. Its present extent does not, however, suffice to counteract linguistic Finnicization for several reasons. Increase in Sámi instruction from 19751984 seems to be an explosion of activity, but such a conclusion would be misleading (Aikio, M., 1990: 37980). The beginning started at zero, and Sámi instruction as compared to total instruction is of little significance. Teaching of and in Sámi represents a modest proportion of the weekly lesson hours. This is mainly blamed on the lack of adequate teaching materials, plus a shortage of Sámi-speaking teachers, although the teacher situation is improving. The true causes can, however, be found in the depths of dominant monism (see Smolicz, 1981: 19) and reluctance to undermine the apparent homogeneity of Finnish society. The new school acts promulgated in 1985 make it possible for comprehensive and secondary schools to use Sámi as a language of instruction. The persisting gap between Sámi teaching and the number of Sámi students is shown in the maps in Figure 2, compiled on the basis of data from the school year 1988/89, obtained from school authorities and supplemented and revised from data obtained directly from the schools. The left-hand maps reveal the number of Sámi students in the lower and upper stages of comprehensive and upper secondary schools, and the right-hand maps show teaching in (or of) Sámi in each school as a ratio to the weekly number of lesson hours. The set of maps shows clearly that instruction in the mother tongue decreases radically at the upper stage of the comprehensive and upper secondary school. The ethnic classification is open to interpretation and subject to errors but may be considered indicative. On the other hand, it must be borne in mind that some schools may teach Sámi to Finnish speakers, too. The different statistics available on lesson hours contain some inaccuracies and contradictions. It proved surprisingly difficult to get precise data, although the schools evidently had confirmed their annual work programmes and this information also goes to the Provincial School Department. Nevertheless, the general conclusion that can be drawn from the data is the small extent of teaching of and in Sámi. There are very few subjects that are taught in Sámi. The provision of teaching still has some significant gaps, especially as regards the upper stages and upper secondary schools. On the whole, authorities must invest more resources in Sámi teaching to meet the expectations and hopes directed at it. To date, no Sámi-speaker has majored in the Sámi language at a university in Finland although Sámi has been a university subject for more than ten years. Their proficiency in Sámias defined by outsidershas not been sufficient, and this made Sámi students afraid to complete a university degree in their own language. Here we can also see the effect of the low status of Sámi: a significant proportion of the students ethnically defined as Sámi do not avail themselves

< previous page

page_98

next page >

< previous page

page_99

next page > Page 99

Figure 2 Number of Sámi students and weekly lesson hours of Sámi teaching in the Sámi Home Area, 1988/1989.

< previous page

page_99

next page >

< previous page

page_100

next page > Page 100

of the teaching because of their families' or of their own wishes. The status of Sámi in school is thus largely dependent on its status elsewhere in society. The Right and the Possibility to use Language The status of Sámi in society may be examined from two different perspectives. On the one hand, the right to use it is a question of defending human rights. On the other, it is a question of preserving the language by ensuring functional contexts for its use. The Sámi language is being used more frequently in different contexts. In the same way, cultural provision in the Sámi language is considerable compared to the population. Language undeniably has a strong symbolic value in this connection, but it still has only marginal significance, compared to the dominant language. There are some unexpected gaps in linguistic rights: the new Ecclesiastical Bill does not mention the Sámi language, Finnish legislation governing organisational activities does not recognise the right to set up associations using the Sámi language, and so on. A Sámi language act is under preparation at present. It is expected to improve the language situation among the Sámi in Finland and to prevent or at least curb linguistic assimilation, even though the act in the proposed form will not give the Sámi the same linguistic rights as Swedish speakers have in Finland or the Inuit have in Greenland. The draft act does not take into consideration that the Same are an ethnic indigenous people who are in a minority and who have no state of their own. The Sámi are seen as a linguistic minority. Even as such, the Sámi would be given only limited linguistic rights. The proposed bill must be characterised as a compromise, but one which constitutes a necessary minimum transitional solution. According to the current view, a culture without writing is unprotected. The Sámi orthography is somewhat problematic. Safeguarding the status of a language presumes that it exists in written form. The principal orthography of the North Sámi has been changing, and it further contains phonetic conventions alien to the surrounding written linguistic environment (Swedish, Norwegian, Finnish). It is obvious that the difficulty of the written form of the language, differing greatly from that of the dominant languages, is exaggerated. The language then becomes the object of exaggerated demands for quality, which in itself impedes the use of Sámi in a written form. On the other hand, it must be emphatically underlined in this context that the majority of the adult Sámi population has never been taught to read and write in Sámi. Most adult Sámi are illiterate in their mother tongue. This should be corrected though efficient adult education. The Reification of Language One of the rarer appearances of linguistic inequality in Finland is the phenomenon which I call language reification. Language reification is the

< previous page

page_100

next page >

< previous page

page_101

next page > Page 101

process where part of the language, especially the written part, becomes dehumanised and moves beyond the reach of its users. Berger & Luckmann (1980: 82) define reification thus: Reification is the apprehension of the products of human activity as if they were something else than human productssuch as facts of nature, results of cosmic laws, or manifestations of divine will. Reification implies that man is capable of forgetting his own authorship of the human world, and further, that the dialectic between man, the producer, and his products is lost to consciousness. When reified the language moves out of the reach of its users. The language loses its essential human charm; it becomes dehumanised. The language users do not feel that it is their own language any longer. At the same time, the right to make decisions on linguistic variations shifts to the linguistic experts. Also connected with this phenomenon is the exaggerated importance of orthography. It is fought over, and complex phonetic symbols are taken into use. The language is 'preserved' as an ancient artifact. What is to be feared is that the reified language will not produce living traditions: it is in a frozen, static state. This is tragic, especially in a language belonging to people with a minority status whose existence and continuation is already under threat. We must ask whether reification is a prelude to the disappearance of native language. As a consequence, people are afraid to use their own language, even the spoken variant of it turns into a taboo. The language is employed to stratify by ability the native population and finally only few are declared qualified. Even outsiders can attain qualifications easier than native Sámi, a result of reification indicated by university graduations (see above). Further, the distortion of reification seems to go hand in hand with the academic researchers' mania for recording and documenting. Some Further Observations An examination of the Sámi language and its future should not raise the language into a position too detached from Sámi identity and culture. The Sámi language has been brought to the forefront as an indicator of acculturation, and as a symbol of the efforts to resist assimilation to the mainstream culture. The possibility of treating language as a core value of the Sámi culture is a field of further inquiry. This must not prevent us from seeing the importance of the material basis of cultural and the interrelationship between the political, social, economic and territorial rights of the Sámi people. The language relates to all these dimensions. As for the measures taken to revive the Sámi language, we must bear in mind the time lag in their implementation (Aikio, M.,1984). For instance, the most recent statistical data available to decision-makers dates from the late 1980s. The administrative procedures take time, and by the time the

< previous page

page_101

next page >

< previous page

page_102

next page > Page 102

eventual measures are implemented, the situation has changed for the worse. The key is to 'overquantify' the measures taken to improve the situation of the Sámi. At the general level of decision-making, the Sámi situation requires a strategy of positive discrimination: Sámi culture can be promoted by according special advantages or ethnic privileges to it over the interests of the dominant population. This is, of course, a joint Nordic task touching upon Sweden and Norway as well. These measures could be geared to increase resources earmarked for culture, which would also create conditions for safeguarding the language. This requires political will: the desire to save Sámi culture and language entails material sacrifice from mainstream society. To deny this or to avoid taking a position in the question of Sámi culture would constitute a conscious effort to stifle the northern minority culture, which would not pass unnoticed in the international community. The official recognition of the Sámi language would only relate to a restricted area. It would not involve the same kind of expenditure as protecting the rights of the Swedishspeaking minority. It cannot be denied that Finland has recently taken some steps in the right direction. Proposals for strengthening the material basis of Sámi culture have been put forward in regional planning. Furthermore, a proposal for a comprehensive Sámi Act was published in June 1990, as co-ordinated with corresponding projects in Sweden and Norway. These aspirations have met with a passionate reaction from the local Finnish population because of the threat it poses to their economic interests. It is symptomatic that the term 'Cultural Sámi' is used as an insult in letters to the editor in the northern press. This indicates the strength of modern forms of racism in Finnish society. The point of departure in political decision-making geared to the preservation of Sámi culture must be that the Sámi are recognised as an indigenous people, which some statements, such as the one delivered by the Ministry of Agriculture and Forestry on the ILO Convention, have endeavoured to deny (Kotkasaari & Havu, 1990). On the other hand, such traditional 'lappological' ethnocentric approaches, which build on the past and contain downright untruths, as characterised the preamble to the State Committee report on Sámi culture (KM, 1985: 66), also lead astray. The Sámi, their culture and language live in the present and have vested interests to protect in present day society. Acknowledgement The author wishes to thank Timo Hirvelä for preparing the set of maps. References Aikio, M. (1976) Saamen ja suomen kielen suhteesta Angelissa, Lismassa ja Nunnasessa. Esitutkimus valtion humanistisen toimikunnan projektiin 'Kulttuurin sopeutuminen arktiseen ekologiaan'.

< previous page

page_102

next page >

< previous page

page_103

next page > Page 103

(1984) The position and use of Sámi language: Historical, contemporary and future perspectives. Journal of Multilingual and Multicultural Development 5 (3 & 4), 27791. (1986) Some issues in the study of language shift in the Northern Calotte. Journal of Multilingual and Multicultural Development 7(5), 36178. (1988) Saamelaiset kielenvaihdon kierteesä: Kielisosiologinen tutkimus viiden saamelaiskylän kielenvaihdosta 19101980. Suomalaisen kirjallisuuden seura. Mänttä. (1990) The Finnish perspective: Language and ethnicity. In Dirmid R.F. Collis (ed.) Arctic Languages: An Awakening. France: Unesco. Aikio, M. and Aikio, P. (1985) Sámi and ethnicity problems in Finland. Pletsch, A. (Hrsg.) Ethnicity in Canada: International Examples and Perspectives. Wenzel. Marburg. (1986) Saamelaisten etnisiteettiongelmista. Lapin tutkimusseuran vuosikirja XXVII. Aikio, P. (1990) The circumpolar peoples and the protection of Arctic environment. Manuscript submitted at the preparatory meeting in Yellowknife, NWT. Asp, E. (1965) Lappalaiset ja lappalaisuus. Turun yliopiston julkaisuja. Forssa. Berger, P. and Luckmann, T. (1980) The Social Construction of Reality. Irvington. New York. Guttorm, J. (1986) Alle kouluikäisten saamelaislasten kasvuolosuhteet ja niiden kehittämismahdollisuudet Utsjoen kunnassa. Sosiaalihallituksen julkaisuja 2. Helsinki. Hokka, H. (1972) Inarin saamelaisten kieliolot. Pro gradu -työ, Turun yliopiston sosiologian laitos. Turku. ILO (1989) Partial revision of the Indigenous and Tribal Populations Convention 1957 (No. 107). 76th Session, Geneva 1989. Komiteamietintö (1985) Saamelaiskulttuuritoimikunnan mietintö (1985: 66). State Committee. Kotkasaari, T. and Havu, S. (1989) Maa- ja metsätalousministeriön lausunto Kansainvälisen työjärjestön 76. Kansainvälisessä työkonferenssissa hyväksymästä yleissopimuksesta. Maaja metsätalousministeriö. 12.12.1989. Magga, M. and Syväjärvi, T. (1990) 'Ei sitä aina jaksa olla niin saamelainen -sitä vässyy.' Enontekiön porosaamelaisten kotikieli 19201980. Pro gradu, Lapin korkeakoulu, kasvatustieteiden osasto. Müller-Wille, L. (1974) Lappen und Finnen in Utsjoki (Ohcijohka) Finnland, Eine Studie zur Identität ethnischer Gruppen im Kulturkontakt. Wästfälische Geographische Studien. Münster. Phillipson, R. and Skutnabb-Kangas, T. (1986) Linguicism Rules in Education. Part 2. Roskilde University Centre. Rivas-Rivas, S. (1987) Racismo lucha de clases y dominacion sociocultural. World Council of Indigenous Peoples. V Assemblea General, Lima, Peru. Julio 1116, 1987 (Unpublished). Smolicz, J.J. (1981) Culture, ethnicity and education: Multiculturalism in a plural society. In Megarry, Nisbet & Hoyle (eds) World Yearbook of Education 1981: Education of Minorities. Kogan Page - Nichols. London - New York.

< previous page

page_103

next page >

< previous page

page_ii

next page >

< previous page

page_105

next page > Page 105

Influence of Ideology in the Linguistic Policy of the Soviet Union Mart Rannut Institute of Language and Literature, Roosikrantsi 6, Tallinn 200106, Estonia/USSR Abstract. In this article a historical overview of the linguistic policy of the Soviet Union is given. Attention is paid to the ideological goals influencing linguistic policy, their realisation and the results. Though the linguistic policy of the USSR is an implicit phenomenon, 3 main periods may be observed, differing in the methods used for achieving ideological goals. These were preceded by an indistinct period without a unified complex of means, due to the struggle for power in the leadership of the USSR. The first period during the Stalin regime may be characterised by urgent and violent measures taken to bring Communist ideals to life by eliminating whole ethnic groups according to the class-based approach. During the second periodpost-Stalinismless violence was used, while the main emphasis was laid on rebuilding human nature and eliminating signs of ethnic origin. A special type of education was worked out called international education; for extreme cases psychiatric hospitals and prisons were used. This period, now called 'stagnation', was mainly connected with the name of Brezhnev. The third period began when Gorbachev came to power. To get out of the impasse 'perestroika' was introduced. Nevertheless, the ideal of the USSR, the Communist empire, was maintained, though Glasnost was allowed. One of the widespread postulates of Soviet ideology has been the claim that the national question was solved in the Soviet Union long ago and there are and will never be any ethnic or linguistic conflicts between nations and ethnic groups and the central government. Recent developments have shown it to be false. There are several reasons for this, among them the structure of the Soviet Uniona multinational empire, formed by force on the ruins of tsarist Russia and maintained by violence, but also the unnatural political boundaries between ethnic groups, uniting some nations and ethnic groups compulsorily and separating others (Central Asia, Karabakh). So mention must be made of at least two reasons why linguistic policy could have been

< previous page

page_105

next page >

< previous page

page_106

next page > Page 106

only implicit: first, there has never existed any Soviet linguistic policy per se as it has been just a part of the greater system called the Soviet Communist ideology, and second, there has never been any stable policy as the official ideology has been propagating the views of the current General (or First) Secretary of the CPSU as well as fresh CP documents backed by convenient quotations from Marx, Engels and Lenin. Nevertheless, some general principles influencing linguistic policy may be mentioned: (1)Primacy of class over nation, where according to the logical end of the class struggle (serving as the motor of historical development) the most progressive classthe proletariatmust dominate society. At the time of the October Revolution such a class-based approach meant in Russia the primacy of a 1 m. proletariate over a 140 m. non-proletariate and further on the reign of the Communist party apparatus'the leading and directing force of the society and the vanguard of the proletariate'over all others. To keep this structure in balance, the dictatorship of the proletariate was carried from theory into practice as a euphemism for terror and totalitarian rule. Such a violent hierarchical structure made it possible to keep citizens under control and surveillance. As the primacy of class over nation is the necessary as well as the sufficient condition for the happy society, manifold hierarchies (e.g. those based on ethnic, linguistic or cultural markers) as well as foreign bodies (e.g. church, remants of other classes, etc) and survivals of the pluralist society were eliminated. (2)The approach of Communism as an objective inevitability and historically progressive process. Communism as the most perfect order of society has several features including general happiness and freedom, the absence of the state, police and prisons as well as nations and ethnic groups, and monolingualism. As speeding while aiming at positive goals is a universal human vice, it is understandable that with the goal of reaching Communism within the period of one generation the policy carried out during the rule of Stalin (19241953) used severe measures to achieve it, including shooting and the Gulag facilities (cf. Solzhenitsyn, 1990). Thus according to the first principle of the class-based approach, it turned against the enemy of the working classin reality it meant the persecution of intellectuals, including the Russian ones. According to the second principle, it handicapped and eliminated all sorts of nationalists who tried to maintain their mother tongue even in the dawn of Communism. Whole nations and ethnic groups were deported and persecuted, including the Kalmyk, the Ingrians, the Cherkess, the Chechens, the Ingush, the Balkars, the Meshi, the Crimean Tartars, the Volga Germans etc. The smaller the number of the languages left, the nearer to Communism. Therefore the solution could not be the substitution of one language by another, but eliminating the ethnic structure as a wholewith the help of migration flows, deportations, mechanical mixing of people in the work places and even through mixed marriages. After the death of Stalin less violence was used in ethnic and linguistic policies (prisons and psychiatric hospitals were used in extreme cases); instead some new theoreti-

< previous page

page_106

next page >

< previous page

page_107

next page > Page 107

cal concepts were introduced. As the ethnic element restricts the free development of a person, one may exterminate the factors maintaining a person's nation and language instead of the person. Two phases of languages operated during socialismthe first phase of Communism: sblizheniyeconvergence and sliyaniyeassimilation. These things happen particularly due to the advantages of socialism in linguistic policy, which are as follows (Medvedyev, 1986): 1. convergence of languages and mutual enrichment of cultures, 2. languages become stylistically diverse (especially when comparing them to the period of illiteracy), 3. several writers write their production in Russian instead of their mother tongue, 4. the phoneme inventory is enriched by new phonemes (e.g. the Kirgiz language was enriched by /f/ when the capital of Kirgizia was named Frunze after one famous Russian revolutionary). Of course, there are some constraints in applying the advantages of socialism to language, as Desheriyev (1982: 95) has pointed out: 'Though language in Soviet society carries an important ideological function, in Soviet philosophical and linguistic literature it has been proved that language by its nature is a non-class-based phenomenon'. To strengthen the positive tendencies in society, the Communist-Party-based Leninist ethnolinguistic policy was proposed, presupposing (Desheriyev, 1982: 12): 1. the absolute equality of languages, 2. creation of the necessary conditions for the evolution and mutual enrichment of languages, 3. unlimited usage of national languages in all spheres, 4. guaranteeing the bilingualism of mother tongue and Russian as an instrument of brotherly co-operation between the nations of the USSR, 5. consistent implementation of national-international flourishing, enrichment and convergence, development of a common Soviet culture, 6. the replenishment of the lexical fund of the languages of the USSR nations (cf. academician Y. Trubatchov's theory in M. Hint, forthcoming). All this requires internationalisation of all spheres of society. Internationalisation (internatsionalizatsiya) has been a directed and controlled process, influenced by policitical, economic, ideological as well as purely propagandist means. The goal, according to the programme of the CPSU from 1961, has been the unity of nationsa common economy, common communist features, common international culture. This has been carried out by means of so-called international education (internatsionalnoye vospitaniye) including the following topics: theory of the national question and propaganda of the co-operation of the USSR nations, inculcations of new international traditions, explanations of the nature of proletarian internationalism, propa-

< previous page

page_107

next page >

< previous page

page_108

next page > Page 108

ganda of the importance of the Russian language, distribution of the (positive) experiences of multinational working parties, creation of the feeling of belonging to the family of the Soviet peoples, struggle against the phenomena of chauvinism and nationalism (Metelitsa, 1982). Other researchers have added the following topics: strictly scientific and consistently class-based interpretation of the history of nations, the advantages of socialism in solving national problems, exposition of the bourgeois falsifications of the national policy of the CPSU and history of the USSR. The aim of such education was to produce brainwashed and trustworthy people with no roots and ethnic preferences. At the 24th Communist Party USSR Congress they were given a beautiful name: the Soviet people, having 3 characteristic features: carrying the Marxist-Leninist ideology, having the goal of building Communism, and non-ethnicity. In the singular this new man born in the Soviet totalitarian regime is called homo soveticus. He has got even more markers: his address is the Soviet Union, he is free of nation and culture; instead of democracy he struggles for socialist democracy, always thinking like his tovarishch and speaking the language of the future Communist society called the international languge (yazyk mezhnatsionalnovo obshcheniya). In the 1920s the chief propagandists spread the theory of the export of Communism, according to which all the countries by a chain reaction would be socialist in the near future. In this case, according to some theoreticians, the language would be English. As a result during the 1920s Latin alphabets were introduced to the languages with a non-cyrillic alphabet or wholly illiterate (the whole campaign against literacy aimed at spreading Communist ideology and converting people to Communism). But as this export theory failed in practice, in the 1930s Russian was chosen to fulfil this historic task. From this period on begins the rise of the status of Russian as primus inter pares to guarantee effective communication and cooperation in the political, cultural and economic spheres. The reasons for this were the following: Russian is spoken by Russians who form about half of the population, brotherly co-operation takes place via this language; the Russian people have liberated others and given them brotherly help, historical traditions and the predominant role of Russian culture and science in developing the culture of other nations; Russian belongs to the most developed world languages as being the language of current scientific, cultural, technological co-operation and international communication (Desheriyev, 1982: 44). In addition, Russian has several unique functions as being the mother tongue of the Russians: it is the freely chosen international language for the peoples of the USSR as well as for the countries of the socialist friendship union. This language contains the richest literature on Marxism-Leninism as well as on socialist economic, cultural and linguistic construction experiences in the USSR, highly necessary for the countries in the Third World (Desheriyev, 1982: 47). To reach the inevitable stage of Communism, Russians must keep their language alive and prosperous. For non-Russians the only option is the exchange of the mother tongue for

< previous page

page_108

next page >

< previous page

page_109

next page > Page 109

Russian through voluntary self-assimilation (Khanazarov, 1982). The temporary solution proposed during Brezhnev's period of rule now called stagnation (zastoy), was named Russiannational bilingualism (natsionalnoruskoye dvuyazychiye). According to this doctrine, Russian operates as 'the second mother tongue' for those Soviet citizens not speaking it as the first one. The 'national' mother tongue, unsuitable for scientific and technological purposes and an obstacle to the scientific progress of the USSR (Kulichenko, 1981: 425), was provided with the functions Russian could not fulfil (e.g. national folklore). As a result, a two-level linguistic empire was formed, where all languages were equal and knowledge of Russian as lingua franca obligatory. In reality it meant diglossia, as the Soviet top administration and diplomacy, as well as compulsory army service and optional prison detainment, required the usage of Russian only. Such well-organised russification resulted in language deaths during the existence of the USSR. About 70 languages have perished as their speakers 'surpassed national isolation' and switched to Russian. While according to the census of 1926 194 various ethnic units were counted, the number according to the census of 1989 is 128. Secondary education (at least the mother tongue lesson) is provided in 40 languages, higher education in 5 languages in the USSR. The whole education system is orientated to an empire-minded Moscow-centred view (e.g. history, literature); mother tongue lessons are displaced by Russian ones, and 'national' schools closed down. Some improvement in linguistic policy has taken place during the years of perestroika and glasnost, initiated by the current Soviet leader Gorbachev. The reason for the new political course lay in the catastrophic economic situation, driven there by the CPSU. Being a dire necessity, perestroika reached other spheres also, though the aim was to improve the economy only and keep the status quo elsewhere. Due to liberalisation tendencies, political repressions were stopped, enabling local nations to take action to save their language and culture, and even dream about their independence. Facing an entirely new situation the official imperial ideology stopped preaching the most unpopular slogan, thus influencing linguistic policy too. However, the main postulates remained the same, getting some cosmetic help only (Guboglo, 1989): 1. languages will die out in the future, 2. Stalin spoilt the correct national policy proposed by Lenin; thus Lenin's heritage in the national policy must be restored, 3. the Russian language as the international language must have the highest status in the Soviet Union, though local languages may be legally protected, 4. the USSR represents a happy multinational society with minor ethnic and linguistic conflicts that are easily solved by the USSR leadership (instead of the previous statement of the USSR as a multinational country without any ethnic and linguistic conflicts).

< previous page

page_109

next page >

< previous page

page_110

next page > Page 110

The last task is the only one in which Soviet ideological workers have succeeded (cf. Guy, 1989). As the main postulate of keeping the Soviet empire together with the legislative and executive power in the hands of Communist party officials is highly unpopular, several legal and political conflicts have appeared. Most of the republics have taken steps to preserve their language by regulating language use on their territory. The central government responded by declaring Russian the international language with the highest status in the USSR in the 'Law of the Languages of the Soviet Union'. According to this every child in the Soviet Union must speak enough Russian to make a Russian-speaking doctor understand their complaints. Such legal conflicts have introduced a new type of ethno-linguistic conflict between Russian-speaking homines sovetici and 'nationalists' loyal to their own language and culture. P. Dostal (1989) has proposed various alternatives for the future of the USSR according to the level of russification. In Rannut (1989) only one alternative was predicted: the splitting up of the Soviet Union. Everybody will soon see which answer is the right one. References Denison, N. (1977) Language death or language suicide. International Journal of the Sociology of Language 12: 1322. Desheriyev, Y (ed.) (1982) Yazyk v razvitom sotsialisticheskom obshchestve. Yazykovye problemy razvitiya sistemy massovoy kommunikatsii v SSSR. 'Nauka' Moskva. Dostal, P. (1989) Regional interests and the National Question under Gorbachev. Nationalism in the USSR. Problems of Nationalities. Amsterdam: Second World Center. Grin, F. (forthcoming) 'The new Estonian language law: Presentation with comments.' To appear in Language Problems and Language Planning. Guboglo, M. (ed.) (1989) Chto delat? V poiskach idei sovershenstvo-vaniya mezhnatsionalnych otnoshenii v SSSR. Moskva. Guy, G. (1989) International perspectives on linguistic diversity and language rights. Language Problems and Language Planning, 13(1), 4553. Hint, M. (forthcoming) 'The changing language situation: Russian influences on contemporary Estonian'. To appear in Journal of Multilingual and Multicultural Development. Khanazarov, K. (1982) Resheniye natsionalno-yazykovoi problemy v SSSR, Moskva. Kulichenko, M. (1981) Rastsvet i sblizheniye natsii v SSSR, Moskva. Maurais, J. (forthcoming) 'A sociolinguistic comparison between Quebec's Charter of the French Language and the 1989 Language Laws of 5 Soviet Republics', to appear in Journal of Multilingual and Multicultural Development. Medvedyev, V. (1986) 'Osobennosti funktsionirovaniya yazykov v usloviyach sotsializma', Materyaly pyatoi respublikanskoi nauchnoi konferentsii molodych lingvistov. Yerevan: 17172. Metelitsa, L. (1982) Yedinstvo internatsionalnogo i patrioticheskogo vospi-taniya, Moskva. Rannut, M. (1989) Sovinismilainetest *. Looming 5: 67880. Viikberg, J. (1990) The Siberian Estonians and language policy. Journal of Multilingual and Multicultural Development. Note added in proof This paper was written in the first half of 1990. Subsequent and ongoing changes in the USSR have already made much of it out-of-date.

< previous page

page_110

next page >

< previous page

page_111

next page > Page 111

The Changing Language Situation: Russian Influences on Contemporary Estonian Mati Hint Trummi 23a-21, Tallinn, Estonia, USSR Abstract. The paper gives a short outline of the Soviet language policy, i.e. of how linguistic structures of the Russian language should be imposed upon the other languages of the empire. This development is illustrated by tendencies in the structure of Estonian today. Introduction Up to now the Soviet language and national policies have been overtly directed towards the assimilation of non-Russian nations and cultures of the Soviet Union. The final aim of these policies has been to mould a new peoplethe Soviet peoplewhich would dissolve all the contemporary Soviet nations into a new Russian-speaking nation with a denationalised cultural identity and with a new national selfawareness. This conception and the policies following the aims of assimilation are to be blamed for national conflicts in the Soviet Union. Three main types of tactics used to achieve these goals may be illustrated by the leading principles of Soviet scholars who have been in charge of Soviet language policies. Academician Yulian Bromley has underlined the following factors and means of promoting the cause: territorial mingling of nationalities; large-scale migrations and ethnic deconcentration; broad international (between Union Republics) exchange of personnel and creation of multinational work teams; nationally mixed marriages; bilingualism (of non-Russian peoples); development of all-Soviet culture which functions in the Russian language;

< previous page

page_111

next page >

< previous page

page_112

next page > Page 112

struggle against various forms of nationalism and against idealisation of the past of individual peoples; cultural integration of Soviet peoples on the basis of Marxist-Leninist ideology; acknowledgement of the leading role of the Communist Party in managing national processes. Mikhail Guboglo has suggested this repertoire of measures in order to ensure at the first stage of development the bilingualism of non-Russian peoples, and at the second stage an overall happiness in the Russian-speaking, highly developed socialist community (society). Guboglo adds some more active measures: as he is especially critical of the 'artificial' expansion of social functions of the languages of the Union Republics, and of 'violating' the rights of the Russian language, he suggests the restriction of the use of these languages in education, mass communication, and publications. According to Guboglo, the less the Soviet Union publishes in non-Russian languages, the better for all nations of the Soviet Union, provided that the reduction in the volume of non-Russian languages is compensated for by growth in the volume of Russian. Nowadays in the Soviet Union such a model of forcible Russification is losing ground. Because of this the great cause is to be taken over by more sophisticated and more intellectual representatives of Pan-Slavism. Oleg Trubatchov, a well-known linguist, has proposed a 'scientific theory' of Soviet language Union. According to this 'theory', all languages in the Soviet Union belong to a language union with the Russian language having a leading role. In order to facilitate the cultural exchange between nations in the Soviet Union, we should help forward the structural assimilation of these languages. The structure of all non-Russian languages, their grammars as well as word usage and word meanings, should become similar to the Russian grammar and the Russian word usage. Thus linguistic structures become isomorphic while language substance may remain different for some time. Russification in Estonia In Estonia the forcible Russification has failed in its purpose. Although the Soviet state has achieved the external presuppositions for Russification in Estonia (migration, territorial mingling of nationalities, privileges of the Russian language), the Estonians have not become bilinguals, the percentage of mixed marriages is low (and children in these families are more often Estonian-speaking than Russian-speaking), the Estonians have not become accustomed to reading in Russian, the prestige of the Estonian language among Estonians is higher than the prestige of Russian. The external measures of Russification have only created national tensions without any positive consequences, and without too much success for the cruel aims of Soviet language policies.

< previous page

page_112

next page >

< previous page

page_113

next page > Page 113

The structural Russification (Trubatchov's idea) is much more dangerous than it seems at first sight. Since real glasnost began in Estonia, new parallel types of linguistic constructions have emerged. Very often these innovations show the influence of the Russian language. The main source of new expressions is the Estonian-language press (including other mass media, such as TV and radio). Without any doubt, Estonian journalists make use of age materials very intensively, and the Estonian newspapers are read more zealously than ever. This is a paradox of the recently achieved freedom of the press. In Estonian the linguistic situation is complicated, because this language has during its long history lived under the influence of Indo-European (Germanic, Baltic, Slavic) languages. Nowadays, the Russian influence revives some old German influences which for some generations have been considered non-standard. The following is an illustration of some types of newly expanding expressions in Estonian. Under the influence of Russian the use of reflexive verbs has increased markedly. In some semantic fields the derivation of reflexive verbs is becoming almost paradigmatic, the grammatical expression of reflexiveness being increasingly iconic (only the u-suffix instead of various suffixes used previously, such as -i- and -ne-). The reflexive verbs are used to replace the impersonal forms (passive) of transitive verbs, whereby the active semantics of the transitive verb is changed into the passive (although the verb forms are formally active). Meie veekogud

on saastatud (transitive, impersonal mood). on saastunud (reflexive, personal mood).

'Our water supply is being polluted.' Iseseisev Eesti riik

taastatakse (transitive, impersonal mood). taastub (reflexive, personal mood).

'The independent Estonian state will be restored.' Sikud

eraldatakse (transitive) lammastest. eralduvad (reflexive)

'The goats will be separated from the sheep.' Ebameeldivused

seletatakse (transitive) imperialistide sepitsustega. seletuvad (reflexive)

'Unpleasantries are blamed on imperialist intrigue.' With reflexiveness the very ideology of grammar changes. The reflexive verbs convey the passive semantics: the water supply becomes polluted as if of its own will, the Estonian state will be restored as if without human effort, and in the case of unpleasantries, we get the impression that the imperialists are actually responsible for all the ills. This kind of linguistic thinking has been rather alien to Estonians. There are other statistical shifts in verb morphology which have some influence upon the typological structure of verb forms. In newspaper style,

< previous page

page_113

next page >

< previous page

page_114

next page > Page 114

the traditional agglutinative verb forms are often replaced by analytical forms where grammatical meaning (present or past tense) is expressed by the auxiliary verb olema 'to be'.

Traditional:

Ilmateade luba/b vihma (object). 'The weather forecast promise/s rain.' Ilmateade luba/s vihma. 'The weather forecast promised rain'. Ilmateade on vihmalubav (aux. + present participle). 'The weather forecast is rain-promising.'

Innovating:

Ilmateade oli vihmalubav 'The weather forecast was rain-promising'. In Fenno-Ugric languages negation is expressed by a special negation verb. In Estonian this verb is analogised to a negation word ei which is used in all persons (ei + verb stem without personal ending, e.g. mina ei. jää, sina ei jää, tema ei jää, meie ei Jää... 'I, you, he/she, we... don't remain...'). Kôige ohtlikum on alkohol, sellest ei jää kaugele ravimid (pl.) 'Most Traditional:dangerous is alcohol, the drugs don't remain far behind.' The Russian-influenced innovating construction uses another negation word mitte together with a verb form which has the personal ending agreeing with the subject: Kôige ohtlikum on alkohol, mitte kaugele sellest jää/vad ravim/d (pl.) There are new types of agreement which may be called semantic congruence. Traditional:

Innovating:

osa töötaja/te/st

osa/d töötaja/d

(-te- 'pl', -st 'elative ending')

(-d 'pl. nominative')

'part of workers' enamiku/le osavôtja/te/st (-le 'allative ending, attached tosing. stem'; -te- 'pl', -st 'elative ending')

enamike/le osavôtja/te/le (-le 'allative ending, attached to pl.stem', -te'pl')

'for the majority of/from participants' Some construction types in Indo-European and Fenno-Ugric languages are rather different:

< previous page

page_114

next page >

< previous page

page_115

next page > Page 115

Fenno-Ugric main word + postposition

Indo-European preposition + main word

teed mööda 'along the way' = (tee/d 'part sing.', mööda 'along') mööda teed (prep + noun) raskuste kaudu = (raskus/te 'pl läbi raskuste 'through the gen', kaudu 'via, postposition') difficulties' (prep + noun) genitive (or nominative) attribute + main word

main word + attributive construction

Institut für Keeleteaduse instituut (attr. in Sprachwissenschaft (noun + gen sing. + noun) attr. construction) The Indo-European patterns are used more and more, however the traditional means are not forgotten yet: Traditional:

Innovating:

Eesti annekteeriti vägivalla abil Eesti annekteeriti läbi (main word + postposition) vägivalle (prep + main 'Estonia was annexed by violence.' word). küsimus Balti riikidest Balti riikide küsimus (attr. in gen (main word + attr. in sing. + attr. in gen pl + main gen. sing. + attr. in elat. word) 'the question of Baltic states'pl) ooperi peaosa (attr. in gen. sing. + main word) 'the main rôle of/in an peaosa ooperis (main opera' word + attr. In inessive) teise küsimuse vastuseks (attr. in gen. sing. + main word in translative case) 'as an answer to the second question'

vastuseks teisele küsimusele (main word in translative + attr. in allative case)

parteide seadus (attr. in gen. pl + main word) 'the law of/about seadus parteidest (main parties' word + attr. in elat. pl.) Sometimes the new constructions are grammatically ambiguous as their interpretation by means of traditional and innovating rules gives different meanings: Uutes demokraatiates on oht totalitarismile (totalitarismi oht). 'In new democracies there is a danger of totalitarianism.' There are other examples of the invasion of internal strucutre of Russian into Estonian. One striking case is the adverb öö/päev/ringse/It '24 hours' 'night + day + circle + adverbial suffix'

< previous page

page_115

next page >

< previous page

page_116

next page > Page 116

cf. Russian kruglo/sutoc */no 'circle + 24 hours + adverbial suffix' There is no need in this Estonian adverb for a circlethe adverbial phrase used to be ööpäev ringi ööpäev läbi 'all 24 hours' (expressed by means of postpositions). The new Estonian adverb is derived under the influence of the Russian adverb (and there is a small subtype of these adverbs). Despite all these influences the Estonian language continues to be highly regarded in Estonia. It is still the main means of communication, education, culture, religion, and administration in Estonia. References Bromley, Yu. (1983) Ethnic processes. Soviet Ethnographic Studies 3. USSR Academy of Sciences. Guboglo, M. (1985) Na dvuedinoi osnove. Sovetskaya Estonia, Febr. 26. (1988) Kak dva kryla v polëte. Pravda, August 25. Trubatchov, O. N. (1988) Slavyane: jazyk i istoriya. Drushba Narodov 5, pp. 24349. The Estonian-language newspapers, magazines, TV and radio programmes 19881990.

< previous page

page_116

next page >

< previous page

page_117

next page > Page 117

A Sociolinguistic Comparison Between Québec's Charter of the French Language and the 1989 Language Laws of Five Soviet Republics Jacques Maurais Conseil de la langue française, 800 place d'Youville, Québec (Qc), G18 1C4, Canada Abstract. This paper will compare the language laws of 5 Soviet Republics (Estonia, Latvia, Lithuania, Moldavia and Kazakhstan) with Québec's Charter of the French Language mainly along the following lines: 1. proclamation of an official language 2. the common language question 3. language of communication with customers and citizens 4. language of education 5. linguistic aspects of immigration. The paper comments on bilingualism as viewed in Quebec's Charter of French Language and in the language laws of the aforementioned Soviet Republics. In 1989 nine republics of the Soviet Union passed language laws similar in scope to Québec's Charter of the French Language. In this paper, I have chosen to make a comparison between Québec's language laws and the laws of five Soviet Republics: the three Baltic States (Estonia, Latvia and Lithuania) along with Moldavia and Kazakhstan. My paper will take up the following points: the proclamation of official languages; the question of a common language within Québec and within the borders of the Soviet Republics; the language of education; the language of service to customers and citizens; finally, the question of immigration and the linguistic integration of immigrants. I will comment on those various language laws from a sociolinguistic point of view, not from a purely legal or even from a legalistic perspective. My aim will be to try and define some basic sociolinguistic principles underlying those legal texts.

< previous page

page_117

next page >

< previous page

page_118

next page > Page 118

Proclamation of an Official Language Section 1 of the Charter of the French Language makes French the official language of Québec. It should be noted that the proclamation of French as the official language is not inscribed in our Constitution since Québec has no written constitution. The very name Charter of the French Language seems to imply that it is in a certain way a constitutional law or at least a law that has a certain preponderance over ordinary laws, like our Charter of Rights, but, as a matter of fact, the Charter of the French Language is just an ordinary law. On the other hand, it should be mentioned that certain provisions of the 1867 Canadian Constitution refer directly to Québec, as for instance section 133 which specifies that both French and English shall be used in Québec's National Assembly and in Québec's courts of justice. Thus, whereas the Charter states that French is the official language, the Canadian Constitution adds English in the aforementioned domains. The Soviet Republics have adopted constitutional amendments in order to proclaim an official language. The phrasing of such amendments is quite simple; the shortest is section 73 of the Constitution of Latvia: 'The Latvian language is the official language of the Soviet Socialist Republic of Latvia. The use of Latvian and other languages will be defined in the Law on Languages.' 1. The Lithuanian text is longer because it specifies the domains where the use of Lithuanian will be guaranteed: 'The SSR of Lithuania guarantees the use of Lithuanian in the activities of government and public agencies, in the domains of public instruction, culture, science, production, and in institutions, firms, and organisations'. The Lithuanian Constitution also mentions that the government has to create the conditions for the development of the other languages spoken in Lithuania. In this respect, the Constitution of Moldavia specifically refers to the protection and development of Gagauz, a Turkic language spoken mainly in Moldavia and adjacent parts of the Ukraine. (As for Russian, its rôle will be dealt with in the next section.) A last point is worth mentioning in connection with the question of official languages. In the law on languages passed by Kazakhstan, section 4 provides for the opportunity for local soviets to declare local official languages; and the text goes on: 'The local official language is used on an equal footing with the official language [Kazakh] and the language of interethnic communications [Russian]'. My understanding is that there are three official languages, which may be cumbersome and expensive. But this may set an example of how to improve the status of Québec's nine native languages (one of which is on the verge of disappearing); having them declared local official languages could indeed enhance their chances of survival, as it would open up to them domains of formal use; thus, their use would no longer be restricted to colloquial speech.

< previous page

page_118

next page >

< previous page

page_119

next page > Page 119

The Common Language Question In the case of Québec, the preamble of the Charter of the French Language states that, '[...] the National Assembly of Québec [...] is resolved [...] to make of French the language of Government and the Law, as well as the normal and everyday language of work, instruction, communication, commerce and business'. The will to see French become the common language of Québec is also seen in the provision requiring that all children of new immigrants should attend French schools. It should be pointed out that, according to official statistics, more than 90% of Québec's population speaks French (with varying degress of fluency, 83% as native speakers). If we now examine the question of a common language in the Soviet Union (jazyk mezhnacional'nogo obshchenija), at least two levels of analysis can be singled out: (1) the language of exchanges with the central government and between the governments of the various Republics; (2) the language of interethnic communications within the Republics. For lack of space, I will only comment on the second point since the question of the language used for interethnic communications within a given Republic has a greater political and social importance; in this respect, the most interesting case was offered by Moldavia whose disputes over the designation of the language of interethnic communications were even reported in Western newspapers. The first draft of the Moldavian language bill had a provision declaring that Moldavian should become the language of interethnic communication. This was strongly opposed by more than 80,000 Russian-speaking workers who reportedly went on strike at 200 businesses and industries. A demonstration in support of the language bill drew more than 300,000 ethnic Moldavians. And an obscurely worded compromise was finally reached after Moldavia's Communist Party chief had a telephone conversation with President Mikhail Gorbachev (The Globe and Mail, 30 August 1989 and 1st September 1989; The Gazette, 31 August 1989; Le Soleil, 1st September 1989; Le Devoir, 1st September 1989). The actual text of the law passed by Moldavia's Supreme Soviet is a clear reflection of those social tensions: section 1 unequivocally lays down that Moldavian shall be the language of interethnic communication within Moldavia ('On Moldavian territory, the Moldavian language is used as the language of interethnic communication') but this is contradicted by section 3, which provides for the use of both Russian and Moldavian ('As the language of interethnic communication in the USSR, the Russian language is used on the territory of the Republic along with the Moldavian language as the language of interethnic communication, which guarantees the accomplishment of a real national language-Russian and Russian-national language bilingualism'). The Moldavian example shows that the language question conceals power struggles in a given society; as it has been noted repeatedly, extralinguistic factors play their part in language planning (see references given by Daoust & Maurais, 1987: 26).

< previous page

page_119

next page >

< previous page

page_120

next page > Page 120

The question of a common national language is a very important one and is clearly reminiscent of the Jacobinist' ideology stemming from the French Revolution. Though Jacobinism' is a convenient wordand is after all but old language imposition writ shortit should be borne in mind that it is not an invention of the French Revolution since it had already been applied before and elsewhere (for an historical sketch on linguistic repression in the United Kingdom, see Maurais, 1990; for Spain, see Ferrer, 1985). The sociolinguistic problem set by Jacobinism is how to reconcile the imposition of a sole official language with respect for minorities without imposing systematic bilingualism on regional linguistic majorities (by which term I refer to groups like the French-speakers in Québec, the Moldavian-speaking population in Moldavia, etc.). The way out of such contradictions definitely lies in compromises or, more precisely said, in various forms and degrees of bilingualism; there is however no model which would apply to all particular situations; instead, solutions have to be worked out for every particular situation in order to take into account historical, social, demographic, constitutional, etc., peculiarities. Québec's answer was to declare one official languagethus removing the burden of systematic bilingualism from the French-speaking majoritybut to allow for the use of English in certain circumstances, especially for external commercial communications (with the rest of Canada and the United States) provided that the positions requiring the knowledge of English would be negotiated between business firms and a specially commissioned government agency (Office de la langue française); these bilingual positions have been informally called 'linguistic bridges' and their very existence is intended to allow French to be the working language in business firms (Corbeil, 1975 [1974]: 25; see also Corbeil, 1980). Language Used in Communications with Clients Section 5 of the Charter of the French Language states that 'Consumers of goods and services have a right to be informed and served in French'. It should be noted that there is no such provision for English and this is quite understandable since, historically, the problem was to be served in French, especially in some Montreal districts. However, the English minority has legally been guaranteed social and health services in its language (Bill 142) and some ethnic groups enjoy services in their language (for instance, there is a bilingual Italian-French hospital in Montreal and a bilingual Polish-French home for the elderly). But I will restrict discussion in this section to the question of the language used in communications with clients. It is noteworthy that the aforequoted section 5 of the Charter of the French Language is only declaratory, i.e. the law provides for no penalty in case of infringement (for a legal discussion, see Commission de protection de la langue française, 19841985). However, a survey made in 1988 in the Montreal area showed that, on the whole, the interviewers were accosted in French in 90% of the cases; yet, in some areas of the West Island the

< previous page

page_120

next page >

< previous page

page_121

next page > Page 121

interviewers were accosted in English in 40% or even 50% of the cases but only in 7% to 13% of those cases was it impossible for them to receive information in French (see Monnier, 1989). The Soviet legal texts we are discussing all impose bilingualism in services, for instance section 3 of the 20 February 1989 Lithuanian Order in Council ('Heads of ministries, agencies, executive committees of local soviets, enterprises, institutions and organisations shall guarantee that citizens when settling their personal matters may apply and receive information and needed documents in Lithuanian or Russian, according to their wishes'); in Moldavia and Kazakhstan, the law even provides for services in a third language in designated local communities. The Estonian law mentions that linguistic proficiency requirements will be defined for all State employeeswhich in a socialist economy means almost every workerif their position makes it necessary for them to communicate with the general public; similar requirements for employees of co-operatives are said to be contained in regulations passed in April 1989 (Grin, forthcoming). Section 38 of the Estonian law vaguely hints at sanctions for those not complying with section 4 but no further detail is given (though presumably that will be dealt with in forthcoming regulations). However, the Latvian law is very clear in this respect: two sections deal with penalties. Section 22 stipulates that citizens shall be reimbursed for any losses due to a failure of employees to speak Latvian or Russian and that this reimbursement may even be demanded from the guilty employee. And according to section 23, infringement on a citizen's freedom of language choice shall bring the guilty party before court. The bilingual requirements of the Soviet Republics must be evaluated in reference to the demolinguistic composition of their populations. The Russian population is a force to be reckoned with in all those five Republics: ethnic Russians count as 9.4% of the population of Lithuania, 12.8% of the population of Moldavia, 27.9% of the population of Estonia, 32.8% of the population of Latvia and in Kazakhstan they even outnumber (40.8%) native Kazakhs (36%). And ethnic Russians tend to be monolingual: only 3.1% of ethnic Russians are said to be bilingual (Comrie, 1981: 28) whereas a majority of non-Russians are able to speak Russian with varying degrees of fluency (the overall figure for the Union was 62.1% in 1979, according to Solchanyk, 1982 2); bilingual requirements are clearly set in order to force the Russian minority learn the majority language of the individual Republics. However, from a sociolinguistic point of view it seems questionable whether imposing penalties as is intended in Latvia would bring about greater bilingualism among the Russian population; one may think that it will only bring about frustration and resentment. I think the solution lies clearly in declaring certain positions both in the civil service and in sales and services as requiring bilingualism when there is contact with the general public. The rule should be that one could have the opportunity to be served in one's language without imposing the burden of bilingualism on every employee and, more specifically, the general rule ought to be that one should be

< previous page

page_121

next page >

< previous page

page_122

next page > Page 122

served in the official language in any given situation. This is, for instance, the solution that has been chosen for institutions under federal jurisdiction in Canada. My own conclusion would be that it is feasible to impose the use of a given language in certain circumstances, especially in domains of formal use, but that it is unadvisable to impose systematic bilingualism on the workforce. Language of Education This is an instance where there is a marked difference between Québec and the situation extant in the Soviet Union. In 1959 the Union Republics and Autonomous Republics of the Soviet Union passed laws on educational reforms which, among other things, gave parents the right to send their children to a school where the language of their choice is used; moreover parents who had decided to send their children to Russian schools were also given the right to choose whether or not they wanted their children to be taught the indigenous language of their Republic as a subject (Bilinsky, 1962: 150). The result of the language provisions contained in the education laws of 1959 has been an overall gradual shift to Russian as the medium of education and, according to Brian D. Silver (1974: 29), 'highly reliable and convincing data have now accumulated indicating that enrolment in non-Russian schools has [...] significantly declined during the 1960s.' The language laws passed in 1989 do not substantially alter the situation that has been described so far, since they maintain the right for parents to choose to send their children either to a school where the basic language of education is the native language or to one where this language is Russian: for instance, the first paragraph of section 18 of the Kazakh law states that 'The Kazakh SSR guarantees the right of every citizen to choose the language of education'. This right does not apply only to Russian and the languages of the Union Republics, but usually to minority languages as well. However, the languages of the Union Republics are made a required subject in all schools in Lithuania (Decree, section 7) and Latvia (section 13); in Moldavia (section 21) and Kazakhstan (section 19), both the Union Republic language and Russian are declared required subjects, but in the case of Kazakhstan a 10-year delay is allowed for this provision to come into force (section 2 of the Order in Council published in Kazakhstankaja Pravda, 28 September 1989). As for Estonia, freedom of choice is not mentioned but the right is granted to every Estonian citizen to receive education in his native language. If we now examine the question of the language of education in Québec, freedom of choice used to be the rule though it was not mentioned in the Constitution nor in any act of Parliament prior to 1969, when Bill 63 granted parents the formal right to choose the language of schooling for their children while requiring in English-medium schools the learning of French. Bill 63 then closely resembles the legal situation now extant in the Soviet Union

< previous page

page_122

next page >

< previous page

page_123

next page > Page 123

and its effects were detrimental to French: in 19711972, 85% of immigrant youngsters 3 attended English schools, while only 15% attended French schools (see St-Germain, 1980; Paillé, 1981) and in 1973, 25,000 French-speaking pupils were enrolled in English schools (Duchesne, 1973). This situation was seen as alarming because 'immigrants were becoming assimilated into the English-speaking group, while French-speakers, because of their declining birthrate, could not offset the number of immigrants that were swelling English-speakers ranks. This was then one of the factors that motivated in 1974 the enactment of Bill 22 which, in its section 43, required from children to pass language tests in order to be admitted to English schools. This provision was deemed as inhuman by immigrant and English-speaking parents and as an inadequate step to promote French by the French population. The rule was again changed in 1977 when the Charter recognised the right to education in English of a child whose father or mother had received elementary education in English in Québec (section 73a), a child whose father or mother was domiciled in Québec when the Charter came into force and had received his or her elementary education in English outside Québec (section 73b), and a child, and his or her younger brothers and sisters, who, when the Charter was adopted, were already receiving education in English in Québec in a kindergarten, elementary or secondary school (section 73c and d). These provisions were compatible with the Constitution of 1867, that guaranteed Protestant education in Québec (which at the time meant education in English, for all practical purposes). Certain aspects of the situation were modified rather substantially when the Supreme Court noted, in July of 1984, that new constitutional rules had changed Québec's capacity to enact or enforce its own conditions for access to education in English. The Supreme Court recognised that a part of the 1982 Constitutional Act of Canada (particularly section 23) was designed to establish a different system of access to education in English. This judgement opened the door to English education in Québec to two new categories of children: a child whose mother or father had received elementary education in English anywhere in Canada, and the brothers and sisters of a child of a Canadian citizen, who had received or was receiving elementary or secondary education in English in Canada. The effects of the Charter of the French language on enrolment in French schools have been very apparent: from 197677 to 198788, the proportion of immigrant children enrolled in French schools rose from some 20% to slightly more than 67% (see Paillé, 1989: 70). As can be judged from both the Québec and the Soviet Union experiences (for data on this latter case, see Bilinsky, 1962; Silver, 1974; Solchanyk, 1982), freedom of choice in the matter of language of education paves the way to assimilation to the dominant language. But this topic also raises the question of immigration, which will be the last point discussed in this paper.

< previous page

page_123

next page >

< previous page

page_124

next page > Page 124

Immigration Though immigration comes under federal jurisdiction, Québec has more control over it than the Republics of the Soviet Union. Only linguistic consequences of immigration will be considered here. The Soviet Republics impose their language on the children of immigrants at least as a required subject in all schools and they try and force the employees who have contact with the general public to learn the Republican language; in at least the case of Latvia, but presumably also of Lithuania and Estonia if my understanding is right, such a provision, owing to the low level of bilingualism on the part of the Russian-speaking population, is aimed at offsetting russification. These are two means taken by the Soviet Republics in order to counterbalance the linguistic effects of massive immigration. The situation in Québec is very different: new immigrants are not only required to take French as a subject, they also have to attend schools where the teaching medium is French. This can be done because we deal here with international immigrants, for the rule set out in the Charter of the French Language does not apply to interprovincial immigrants, i.e. to immigrants coming from the English-speaking provinces, since, as has already been mentioned, the Supreme Court of Canada has declared that section 23 of the 1982 Constitutional Act of Canada has been devised to change Québec's capacity to enact its own conditions to education in English. So, were immigration to return to its pre-1976 level, at a time when half of the immigrant children in the schools came from other Canadian provinces, there could be, especially in the Ottawa River border region, a decrease in the relative importance of French schools (Paillé, 1985, 1986). But, as far as immigration is concerned, the major point to be mentioned is the Cullen-Couture Agreement of 1978, by which Québec obtained from the federal government a right to participate in the selection of its immigrants; this right would even have become a part of the Canadian Constitution if the so-called Meech Lake Constitutional Accord had been ratified by all provincial legislatures before 23 June 1990 (Hogg, 1988). According to the Cullen-Couture Agreement, Québec may set out its own criteria for the selection of its immigrants and one such criterion is the knowledge of French. However, it should be pointed out that this right of inspection does not apply to all persons coming from abroad but only to immigrants as defined in the federal law; therefore, refugees do not come under the Cullen-Couture Agreement nor do foreigners coming to Canada within the framework of the families reunification programme. It is estimated that, on average, Québec supervises only about 50% of its total immigrant population: for instance, in 1987 a total number of 26,806 immigrants came to Québec, 14,781 of which (55.1%) were independent immigrants and accordingly were selected by Québec's officials (Ministère des Communautés culturelles et de l'Immigration, 1989: 16).

< previous page

page_124

next page >

< previous page

page_125

next page > Page 125

Conclusion At first sight, a rapid comparison between Québec's Charter of the French Language and the language laws passed in 1989 by five Soviet Republics (the three Baltic States plus Moldavia and Kazakhstan) leads to the conclusion that Québec's law, with more than 200 sections whereas the longest Soviet textthe Estonia lawhas only 39 sections, is much more comprehensive. But such a superficial comparison leaves out the fact that almost the same fields are covered by all these laws. However, Québec's Charter gives more details on administrative implementation and many sections are devoted to the francisation of business firms and Civil Service. Even if we accept the basic tenet that no two linguistic situations are alike and if we exclude the case of Kazakhstan, which is very peculiar from a demographic point of view 4, there remains a basis for interesting comparative studies, viz. the fact that in all these cases a local linguistic majority tries to keep in check the language of a minority which is actually the majority language of a federation. Notes 1. Translations from Russian are my own. 2. A word of caution is needed here: Comrie's figure is not based of the data of the 1979 census, whereas Solchanyk's is. 3. Immigrants whose mother tongue is neither French nor English are now referred to as Allophones, but this term has not gained popularity yet outside Canada. For the sake of clarity and international understanding, the word immigrant will be used in this paper. 4. For, as it has already been shown, Kazakh-speakers are outnumbered by Russian-speakers. References Bilinsky, Yaroslav (1962) The Soviet Education Laws of 19589 and Soviet Nationality Policy. Soviet Studies 14/2, 13857. Commission de Protection de la Langue Française (19841985) Rapport d' activité 19841985. Québec: Éditeur officiel. Comrie, Bernard (1981) The Languages of the Soviet Union. Cambridge: Cambridge University Press, Cambridge Language Surveys. Corbeil, Jean-Claude (1975[1974]) Description des options linguistiques de l'Office de la langue française. Montreal: Régie de la langue française, coil. Études, recherches et documentation No. 2. (1980) L'aménagement linguistique du Québec. Montreal: Guérin. Daoust, Denise and Maurais, Jacques (1987) L'aménagement linguistique. In J. Maurais (ed.) Politique et aménagement linguistiques. Québec and Paris: Conseil de la langue française and Éditions Le Robert, 546. Duchesne, Louis (1973) La situation des langues dans les écoles du Québec et de ses régions administratives (196970 à 197273). Québec: Ministère de l'Éducation. Ferrer i Girons, Francesc (1985) La persecució política de la lengua catalana. Barcelona: Edicions 62. Grin, François (forthcoming) The new Estionian language law: Presentation with comments. To appear in Language Problems & Language Planning. Hogg, Peter W. (1988) Meech Lake Constitutional Accord Annotated. Toronto, Calgary & Vancouver: Carswell.

< previous page

page_125

next page >

< previous page

page_126

next page > Page 126

Maurais, Jacques (1990) Petite histoire des législations linguistiques au Royaume-Uni. L'Action nationale 80/1, 3543. Ministère des communautés culturelles et de l'immigration (1989) Immigration au Québec, bulletin statistique annuel vol. 13 (198687). Monnier, Daniel (1989) Langue d'accueil et langue de service dans les commerces à Montréal. Québec: Conseil de la langue française, coll. Notes et Documents No. 70. Paillé, Michel (1981) The impact of language policies on enrollment in public schools in Québec. In Paillé 1985: 13349. (1985) Contribution à la démolinguistique du Québec. Québec: Conseil de la langue française, coil. Notes et Documents No. 48. (1986) Conséquences des politiques québécoises sur les effectifs scolaires selon la langue d'enseignement. In Aspects de l'évolution de la situation linguistique du Québec. Québec: Conseil de la langue française, coll. Notes et documents, No. 52, 3951. (1989) Nouvelles tendances démolinguistiques dans l'île de Montréal. Québec: Conseil de la langue française, coll. Notes et Documents No. 71. Saint-Germain, Claude (1980) La situation linguistique dans les écoles primaires et secondaires, 197172 à 197879. Québec: Conseil de la langue française. Silver, Brian D. (1974) The status of national minority languages in Soviet Education: An assessment of recent changes. Soviet Studies 26/1, 2840. Solchanyk, Roman (1982) Russian language and Soviet politics. Soviet Studies 34/1, 2342.

< previous page

page_126

next page >

< previous page

page_127

next page > Page 127

Language Planning as Discipline 1 Björn H. Jernudd Department of Humanities, Footscray Institute, Victoria University of Technology, Footscray, Australia 3011 Abstract. A social science approach to language planning is contrasted with a language management approach to language planning. The language management approach connects language planning to discourse and requires a separation of linguistic and non-linguistic motivations for language planning actions. Language planning is the kind of language management that social and political scientists are the most interested in because it is typically motivated by non-linguistic interests. Social Science Approaches to Language Planning In the concluding paragraph to his very recent book on Language Planning and Social Change, Cooper (1989: 182) answers the question whether a theory of language planning is possible by first quoting and thereby supporting Weston (1977) in favour of a 'single general theory of social change', then stating what a theory would enable us to explain. It would explain: the motivation for setting particular status, corpus, and acquisition goals and for choosing particular means and the reasons that the means do or do not effect the goals within a given social context. Such a theory is unattainable, writes Cooper, because the language planning goals serve a diversity of latent goals such as economic modernization, national integration, national liberation, imperial hegemony, racial, sexual, and economic equality, the maintenance of elites, and their replacement by new elites. What's needed, resolves Cooper, is a theory of social change. Thus, we observe, his interest is focused on the social and not on the linguistic. Because acquisition of social knowledge is such a vast enterprise, he temporarily suspends theorising in favour of description2 (1989: 9798). He develops an accounting scheme to gather at least some of the information one would need to understand language planning. His accounting scheme is hinged on first identifying the actors in a decision-making process. What

< previous page

page_127

next page >

< previous page

page_128

next page > Page 128

language problems are involved in the process are accounted for in terms of those behaviours that these actors' have chosen to attempt to influence. Thus, Cooper's approach to language planning is sociological. So is Joshua Fishman's, to whom Cooper pays tribute in his introduction. Fishman (1987) assesses the state of the art in language planning study and in so doing presents the social science approach to language planning. One of his definitions even opposes a societal direction of study to a linguistic one: For me, language planning remains the authoritative allocation of resources to the attainment of language status and language corpus goals, whether in connection with new functions that are aspired to, or in connection with old functions that need to be discharged more adequately. This definition [ ... ] leads in societal directions more than it does in linguistic ones. (Fishman, 1987: 409) Fishman (1987: 423) subordinates corpus planning to status planning: status planning is the real engine of the language planning train. Only when status planning is seriously enforced does corpus planning really take root [ ... ] the products of corpus planning [ ) ] have no dynamic of their own. Many languages will never get much corpus planning codification or elaboration, and even less implementation [ ... ] Fishman (1987: 411) compassionately constructs language planning around such issues as ethnic identity, nationism and nationalism, functional inequality and undercut pluralism: it [language planning] is primarily the means whereby less fortunate language communities (i.e. those less powerful in their particular confrontation with another ethno-linguistic aggregate) organize their self-defense, as well as their intertranslatability-at-least-to-some-extent-and-in-some-functions vis-à-vis one or another 'international language'. [Thus] They [language planners] are (or should be) issue definers and consciousness raisers vis-à-vis the goal of ethno-cultural pluralism and ethnocultural democracy. (p. 413) He regards the study of language planning as one area of study to contribute to theory of social change, social planning, and at its best, sociological theory in general. People claim allegiance to languages as symbols of the nation, they may fear the loss of income or (is it worse?) suffer alienation from their true origins, indeed face dissolution as a separate and distinct community because of loss of their language. These very real claims are typically produced by and evaluated in terms of goals and ideologies that are not rooted in problems of communication although the use of the language is trivially necessary for the goals and ideologies to apply. Rather, group interests for reasons of

< previous page

page_128

next page >

< previous page

page_129

next page > Page 129

control of institutions, and management of identity, prescribe the endorsement of a particular language. There is no question but that processes and problems of this kind are best studied according to the rules of disciplines that are equipped to dissect relationships of ideology, power, and identity. Works in history, political science, ethnography, literacy, geography, literature, linguistics, social psychology, and so on, hold crucial insights into these aspects of language planning processes and offer methods. If political science accommodates institutional and rhetorical approaches, then the study of the politics of language should accordingly be informed by these institutional and rhetorical approaches (cf. Jernudd & Shapiro, 1989). When theories of development replace each other, so could their projections on language behaviour be re-evaluated on the basis of appropriately enriched and enlarged data bases that correspond to the analytical requirements of the adopted theoretical perspective. In commenting on the status of particularly 'status planning' in his state of the art paper, Joshua Fishman says that [those interested in language planning] have been making up social science theory far too long and, as a result, have benefited far too little from the theory that has been elaborated by specialists working in other areas of social change and social planning. But sociolinguistics is also concerned with the exhaustive, multidimensional depiction of the present, with attitude studies, with usage studies, with criterion evaluation studies, in short, with quantitative studies of various kinds. We ultimately want to know more about what kinds of populations are more likely, and what kinds are less likely, to adopt the status planning and corpus planning products of language planning authorities and why these differentials exist. (1987: 410) For an example of this kind of work, Fishman refers to the International Research Project on Language Planning Processes (1987: 426) 3. Led by Fishman, the project's research team formulated their questions in 196869. The questions then, as Fishman's questions also now, focused on the authorisation, policy-processes, agency operations, and products of language planning agencies and associations, and on responses among the 'target populations'4. These were the kinds of questions that interested the research group, and these are the kinds of concerns that accompany langage planning processes which implement policies that are driven by élite, nationalistic, partisan ethnic, developmentally ideological, or such-like motivations. The project did not connect with individual management of language in discourse because impact of agency work on discourse was the main issue. It gathered language data relevant to evaluating agency influence on language use.

< previous page

page_129

next page >

< previous page

page_130

next page > Page 130

A Language Management Approach to Language Planning In the 1990s, language planning is also motivated by migrations and ethnic relations, by new nationalisms, by the maintenance of state power by the one interest group (even self-identified by 'race' in some corners of our world), by consequences of economic globalisation of the economy, by the growth and increased intensity in use of global, regional, and local networks of communications. Naturally therefore, researchers engage with language planning and from just as many different perspectives of study. However, there is one perspective that can be regarded as language planning's own perspective. This perspective arises out of a theory of language problems (Neustupny, 1968, 1978: 24357). A theory of language problems is explicit about relationships between discourse (communication) and peoples' behaviour towards language in that it must reveal whether and how language problems occur in communicative acts, i.e. in discourse. If participants in language planning processes claim that certain user groups' language use, in terms of specific features of language, or in terms of repertoire and distribution in use by domain or network, are inadequate, how do these claims arise? Do they arise out of linguistic interest or out of non-linguistic interest? With what differential consequences? Language problems that arise out of linguistic interest form a direct part of the communication process, while the latter have to be introduced into discourse in order to become problems of language (Jernudd & Neustupny, 1987: 77). Scientifically minded linguists, if they were to pay attention to language management planning processes, would most likely opt to inspect the sources in discourse for language planners' and managers' claims about problems in need of correction, and solution in need of implementation, just as political scientists and sociologists and social psychologists now opt to inspect sources in ideologies, power conflicts, migrations, nationalisms, ethnic aspirations, etc. for the authorisation and operation and implementation processes in language planning. A discipline of language management which includes language planning as one type of language management activity organises its work with focus on language. Relative to Cooper's and Fishman's approaches to language planning processes, three questions articulate the linguistic interest in the language management approach quite clearly: where is the language problem? whose is the problem? is there a problem in discourse? In seeking answers to these questions, language planning researchers give themselves the opportunity to relate individuals' management of language in discourse, on the one hand, to institutional, ideological, attitudinal and survey-of-language-use findings, on the other. The model for language management in discourse (Jernudd & Neustupny, 1987: 7576) holds that a person

< previous page

page_130

next page >

< previous page

page_131

next page > Page 131

(1) produces messages (2) monitors the language that constitutes these messages, and notes (or not) a difference from norm by monitoring language production, thus identifying a product-item; (3) evaluates (or not) the kind and degree of inadequacy of the product-item; (4) selects (or not) an adjustment strategy or at least ad hoc means of adjustment for the inadequacy; (5) acts (or not) to pre-, in-, or post-correct self or to react to the other's speech, to implement adjustment. It is difficult to recover notings, evaluations, and adjustment strategies from recordings, even when there is reason to think that a participant took overt action to implement adjustment. Conversation analysis assumes that participants' communicative actions are both context shaped and context renewing. It shares with language management an interest in trouble and repair in conversation; and trouble and repair are as central to ethnomethodologists' reconstruction of language processes as they are to the language managers' search for solutions. Conversation analysts' untiring attention to details constitute their (ethno)methodology; language managers make accessible evaluation, adjustment, and implementation processes in discourse and situate these in institutional context. Tapping participants' reports on own behaviours also offers a promising methodology (Ericsson & Simon, 1984). Since individual discourse stands at the centre of the language management discipline, application of methods that rely on participant reports is obviously extremely important. Language management is not alone in sharing this methodological interest. Students of language teaching are returning to self-report and even stream-of-consciousness methods to explore the language acquisition process (see Faerch & Kasper, 1987; Cohen, 1987a, b). Neustupny (1986a, b) discusses the application and interpretation of interaction and follow-up interviews in language management (cf. also Clyne, 1975). For written language, while principles guide editing and evaluation of congruent parameters of a style, of document formulation and so on, language management also needs data from the writing process itself. A researcher can arrange to register all input a writer makes at a computer keyboard and build up a detailed record of rewrite which can be analysed, and then, with questions guided by the analysis, presented to the writer for recollected comment and for re-evaluation (Severinson Eklundh, 1988). There are communications about language inadequacies that can be studied directly. In language learning, the researcher may experience overt, consultative management as a participant either in the role of teacher or learner (Schmidt & Frota, 1986). In language cultivation, the researcher may take note of queries directed over the telephone or by mail to language cultivation agencies to help solve language problems, or collect problems in

< previous page

page_131

next page >

< previous page

page_132

next page > Page 132

any situation of another seeking help with adjustment of inadequacies (cf. Nordisk Språksekretariat, 1988). The study of language management critically depends on an explicit understanding of some discourse events. An interest in discourse is very much a matter of the climate of times. The discursive interest in anthropology 5, practical philosophy6, literary criticism7, political science8, and history9, now percolating in all the human and social sciences, and interest in discourse in branches of language study, are not accidental and not accidentally related. One motivating factor for the shared foregrounding of the discursive in the human and social sciences is an interest in the individual, and perhaps also therefore interest in the ordinary, us-all-encompassing, in contemporary endeavours of any kind. The study of language management focuses on trouble in discourse because processes of overcoming trouble validate practices of language cultivation and language planning. But Fishman's assertion is also true, that basic issues impinging upon language planning not only go beyond language planning (substantially involving, as they do, culture planning and identity planning, i.e., some of the most sensitive and value-encumbered aspects of human society), but go beyond the social sciences themselves. (1987: 411) Indeed, students of language planning need to go beyond both discourse management and the social sciences if our task is to explain, as ends Cooper his book and I with him this paper, that Language is the fundamental institution of society [therefore] To plan language is to plan society. (1989: 182) Notes 1. This conference presentation borrows from a paper just written together with Professor J. V. Neustupny (Jernudd & Neustupny, forthcoming), because this content is what is on my mind. 2. 'I offer the [descriptively-adequate] framework as a guide to future investigators in the hope that it will improve our ability to describe, predict, and explain language planning.' 3. Rubin et al. (1977) report on this project; and for its comparative methodology, note especially Fishman (1977) in the report. 4. The questions are published in an appendix to the book Can Language Be Planned? (Rubin & Jernudd, 1971: 293305). The team relied mainly on questionnaires, and intensive interviewing of well-informed participants in policy and work processes. 5. Cf. Geertz, 1988. 6. 'Heidegger displaces the ego subject, the subject of consciousness from the centre of knowledge and puts in its place an historical, changing subject constituted as a set of skills and/or practices, including (and especially) linguistic practices which ''house" human existence.' (Shapiro, 1984: 216) 7. In the same article, Shapiro lets Beckett speak for the literarily discursive (1984: 239): 'Beckett places the "I" in a place where it receives the action. It is in a head, but the kind where it gets pissed on.' 8. For applications, see Shapiro and Henningsen on language purism in Jernudd & Shapiro,

< previous page

page_132

next page >

< previous page

page_133

next page > Page 133

1989. For an overview, see Shapiro (1987) with an annotated bibliography or his edited collection (1984). Philosophers and political theorists who work in this mode grapple with the problem of our free will in context of peoples' production of relationships of power and authority. 9. Muecke (1983: 71): 'Events in history exist only insofar as they exist in discourse.' References Clyne, M. (1975) A non-verbal check on acceptability of transference among bilinguals. ITL 30, 5564. Cohen, A. D. (1987a) Studying learner strategies: how we get the information. In Wenden & Rubin (1987), 3140. (1987b) Student processing of feedback on their compositions. In Wenden & Rubin (1987), 5769. Cooper, R. L. (1989) Language Planning and Social Change. Cambridge: Cambridge University Press. Ericsson, K. A. and Simon, H. A. (1984) Protocol Analysis. Verbal Reports as Data. Cambridge, Mass. and London: MIT Press. Faerch, C. and Kasper, G. (eds) (1987) Introspection in Second Language Research. Clevedon, England: Multilingual Matters. Fishman, J. A. (1977) Selected dimensions of language planning: a comparative analysis. In Rubin et al. (1977), 195214. (1987) Conference comments: reflections on the current state of language planning. In Laforge (1987), 40528. Geertz, C. (1988) Works and Lives: The Anthropologist as Author. Stanford: Stanford University Press. Hermingsen, M. (1989) The politics of purity and exclusion: literary and linguistic movements of political empowerment in America, Africa, the South Pacific, and Europe. In Jernudd & Shapiro (1989), 3152. Jernudd, B. H. and Neustupny, J. V. (1987) Language planning: for whom? In Laforge (1987), 6984. (forthcoming) Multi-disciplined language planning. In D. F. Marshall (ed.), in the volume on language planning in a series of three focusschriften. John Benjamins. Jernudd, B. H. and Shapiro, M. J. (eds) (1989) The Politics of Language Purism (=Contributions to the Sociology of Language, 54). Berlin and New York: Mouton de Gruyter. Laforge, L. (ed.) (1987) Proceedings of the International Colloquium on Language Planning, May 2529, 1986, Ottawa (=Publications of the International Center for Research on Bilingualism, A-21). Québec: Les Presses de l'Université Laval. Muecke, S. (1983) Discourse, history, fiction: language and Aboriginal history. Australian Journal of Cultural Studies 1, 7180. Neustupny, J. V. (1968) Some general aspects of 'language problems' and 'language' policy in developing societies. In J. A. Fishman, C. A. Ferguson & J. Das Gupta (eds) Language Problems of Developing Nations, 28594. New York: John Wiley & Sons. (1978) Post-Structural Approaches to Language. Language Theory in a Japanese Context. Tokyo: University of Tokyo Press. (1986a) Inquiry for language management. Conference on Language Policy and Language Planning in Hong Kong, City Polytechnic of Hong Kong, December 1921, 1986. (1986b) The follow-up interview. Ms. (Originally written as an appendix for the author's paper 'Language norms in AustralianJapanese contact situations'. In M. Clyne (ed.) (1985) Australia. Meeting Place of Languages (pp. 16170). Canberra: Pacific Linguistics. Nordisk Språksekretariat (1988) Språknemndenes telefon-rådgivning (=Nordisk språksekretariats rapporter, 9). Oslo. Rubin, J. and Jernudd, B. H. (eds) (1971) Can Language be Planned? Honolulu: The University Press of Hawaii.

< previous page

page_133

next page >

< previous page

page_134

next page > Page 134

Rubin, J., Jernudd, B. H. et al. (eds) (1977) Language Planning Processes (=Contributions to the Sociology of Language, 21). The Hague: Mouton Publishers. Schmidt, R. W. and Frota, S. N. (1986) Developing basic conversational ability in a second language. In R. R. Day (ed.) Talking to Learn: Conversation in Second Language Acquisition. Rowley, Mass.: Newbury House. Severinson Eklundh, K. (1988) Datorloggning som metod att studera skrivprocessen. ASLA's Symposium on the Writing Process, November (=ILPLab, 16). Stockholm: Royal Institute of Technology. Shapiro, M. J. (ed.) (1984) Language and Politics. New York: New York University Press. (1987) Language and politics. In Robert B. Kaplan (ed.) Annual Review of Applied Linguistics Vol. 7 (1986). Cambridge, New York, Melbourne: Cambridge University Press, pp. 7485. (1989) A political approach to language purism. In Jernudd and Shapiro (1989), 2129. Weston, L. (1977) The Study of Society. Second edition. Guilford: The Dushkin Publishing Group. Wenden, A. & Rubin, J. (eds) (1987) Learner Strategies in Language Learning: Process and Approaches for Foreign Language, English as a Second Language, and Bilingual Educators. Englewood Cliffs, NY: Prentice Hall International.

< previous page

page_134

next page >

< previous page

page_135 Page 135

Index A Australia 55-65 B Bilingual education 1-7, 55-65 Bilingualism 35-43, 111-116 C Catalan immersion program 21-33 Code-switching 85-92 Context-bound language use 21-33 D Danish 85-92 Didactics 67-72 Directions in immersion research 9-19 E English 85-92 Errors 45-53 Estonian 1-7, 105-110, 111-116 F Finnish 85-92 French immersion programs 9-19 Frisian 73-83 G Georgian 1-7 German 85-92 I Ideology 105-110 Immersion 35-43, 45-53, 55-65 L Language acquisition 21-33, 67-72 Language immersion 67-72 Language laws 117-126 Language management 127-134 Language planning 117-126, 127-134 Language policy 117-126 Language reification 93-102 Language shift 93-102 N Nordic 1-7 Nordism 1-7 Q

Quebec 117-126 R Research methods 127-134 Russification 1-7 S Sami language 93-102 Second language acquisition 35-43 Second language development 9-19 Sociolinguistic theory 127-134 Soviet language policy 105-110, 111-116 Soviet Union 117-126 Swedish 85-92 T Trilingualism 73-83 U Urban vernacular 73-83 W Welsh 45-53

< previous page

page_135