Fourth International Conference on Minority Languages: Volume 1: General Papers

cover

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

next page >

Fourth International Conference On Minority Languages Multilingual Matters (Series) ; 70-71 Gorter, D. Multilingual Matters 1853591041 9781853591044 9780585071206 English Linguistic minorities--Congresses. 1990 P119.315.I5 1989eb 305.7 Linguistic minorities--Congresses.

cover

next page >

< previous page

page_i

next page > Page i

Fourth International Conference on Minority Languages Vol. I: General Papers

< previous page

page_i

next page >

< previous page

page_ii

next page > Page ii

Multilingual Matters Aspects of Bilingualism in Wales COLIN BAKER Bicultural and Trilingual Education MICHAEL BYRAM and JOHAN LEMAN (eds) Bilingualism and the Individual A. HOLMEN, E. HANSEN, J. GIMBEL and J. JØRGENSEN (eds) Bilingualism in Society and School J. JØRGENSEN, E. HANSEN, A. HOLMEN and J. GIMBEL (eds) Bilingualism or Not: The Education of Minorities TOVE SKUTNABB-KANGAS Citizens of This Country: The Asian-British MARY STOPES-ROE and RAYMOND COCHRANE Community Languages: A Handbook BARBARA M. HORVATH and PAUL VAUGHAN Key Issues in Bilingualism and Bilingual Education COLIN BAKER Language Distribution Issues in Bilingual Schooling R. JACOBSON and C. FALTIS (eds) Language and Ethnicity in Minority Sociolinguistic Perspective JOSHUA FISHMAN Mediating Languages and Cultures D. BUTTJES and M. BYRAM (eds) Migration and Intercultural Education in Europe U. PÖRNBACHER (ed.) Minority Education: From Shame to Struggle T. SKUTNABB-KANGAS and J. CUMMINS (eds) Minority Education and Ethnic Survival MICHAEL BYRAM Minority Language Conference: Celtic Papers G. MacEOIN, A. AHLQVIST, D. O'hAODHA (eds) Minority Language Conference: General Papers G: MacEOIN, A. AHLQVIST, D. O'hAODHA (eds) Our Own Language GABRIELLE MAGUIRE Papers from the Fifth Nordic Conference on Bilingualism J. GIMBEL, E. HANSEN, A. HOLMEN and J. JØRGENSEN (eds) The Use of Welsh: A Contribution to Sociolinguistics MARTIN J. BALL (ed.) Please contact us for the latest book information: Multilingual Matters, Bank House, 8a Hill Road, Clevedon, Avon BS21 7HH, England

< previous page

page_ii

next page >

< previous page

page_iii

next page > Page iii

Fourth International Conference on Minority Languages Vol. 1: General Papers Edited by Durk Gorter, Jarich F. Hoekstra, Lammert G. Jansma and Jehannes Ytsma MULTILINGUAL MATTERS 70 Series Editor: Derrick Sharp MULTILINGUAL MATTERS LTD Clevedon Philadelphia

< previous page

page_iii

next page >

< previous page

page_iv

next page > Page iv

Library of Congress Cataloging in Publication Data International Conference on Minority Languages (4th, 1989, Leeuwarden, Netherlands). Fourth international Conference on Minority Languages/edited by Durk Gorter et al. p. cm (Multilingual Matters: 70) Includes bibliographical references. 1. Linguistic minorities - Congresses. I. Gorter, D. (Durk). II. Title. III. Series: Multilingual Matters (Series): 70. P119.315.15 1989 305.7 dc. British Library Cataloguing in Publication Data International Conference on Minority Languages (4th, 1989, Leeuwarden, Netherlands). Fourth International Conference on Minority Languages. 1. Minority languages. I. Title. 400 ISBN 1-85359-104-1 Multilingual Matters Ltd Bank House, 8a Hill Road, Clevedon, Avon BS21 7HH, England & 1900 Frost Road, Suite 101 Bristol, PA 19007 USA Copyright © 1990 D. Gorter, J.F. Hoekstra, L.G. Jansma, J. Ytsma and the authors of individual chapters. This book is also available as a special issue of the Journal of Multilingual and Multicultural Development Vol. 11, Nos 1 & 2, 1990. All rights reserved. No part of this work may be reproduced in any form or by any means without permission in writing from the publisher. Typeset by Photographics, Honiton. 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 Preface Durk Gorter, J. F. Hoekstra, L. G. Jansma and J. Ytsma

1

What is Reversing Language Shift (RLS) and How Can It Succeed? Joshua A. Fishman

5

Minority Language Group Status: A Theoretical Conspexus Howard Giles, Laura Leets and Nikolas Coupland

37

On Coming to Our Census: A Layman's Guide to Demolinguistics John de Vries

57

Legitimating or Delegitimating New Forms of RacismThe Role of Researchers Tove Skutnabb-Kangas

77

Grammatical Borrowing and Language Change: The Dutchification of Frisian 101 Germen de Haan Comparative Analysis of Language Minorities: A Sociopolitical Framework A. B. Anderson

119

Notes for a Minority-Language Typology: Procedures and Justification John Edwards

137

The Economic Approach to Minority Languages François Grin

153

Summing Up Anders Ahlqvist

175

List of Participants

181

Contents of Volume II

184

Index

186

< previous page

page_v

next page >

< previous page

page_1

next page > Page 1

Preface The Fourth International Conference on Minority Languages (4.ICML) was held from Monday 19 to Saturday 24 June 1989 in Ljouwert/Leeuwarden. This conference was organised by the Fryske Akademy. The First International Conference on Minority Languages took place in Scotland (Glasgow) in 1980. It turned out to be the beginning of a successful series of conferences. At the time there was some discussion about the location: should it be held in Scotland or in Friesland? After having gone further north to Scandinavia (Abo/Turku 1983) and three years later to the most western part of Europe in Ireland (Galway 1986), the European continent was to have its turn in 1989 in Friesland (Ljouwert/Leeuwarden). The conference theme was 'Comparative research on minority languages and development of theories'. The three previous conferences focused mainly on problems of definition, on language in society and on the linguistics of minority languages. In this 4.ICML we have tried to go forward by concentrating, on the one hand, on comparative research regarding minority languages and on the other hand on the development of theories in this field. We have welcomed a confrontation of different emerging theoretical perspectives. The group of people involved in studies of linguistic minorities began, so to speak, in 1980 by getting to know each other and the languages their studies were about. Many papers provided basic factual information on a single language situation. As could be expected in such a heterogeneous field with such widely varying language situations a recurring question was how to define a minority language. It will be clear that all those variants and different aspects of the phenomenon are not easy to summarise in a few sentences. Thus the question of definition stood central again in the second conference in the opening key-note address by Allardt (in Molde & Sharp, 1984: 195205). Although the definition issue will continue to pop up time and again, by the time of the third conference Reuter (in Mac Eoin et al., 1987: 214) concluded in his summing up that 'In any case, we all know, approximately, at any rate, what the term "minority language" means'. The organisers of 4.ICML thought, if we ought to know by now what one such language looks like, it may be time to begin to compare two (or more) of them on certain aspects in order to classify the similarities and the differences.

< previous page

page_1

next page >

< previous page

page_2

next page > Page 2

A typology can be a handsome tool for categorisation, as a first step in comparison. In this volume in particular Anderson and Edwards provide us with examples of such typology-construction. From the papers contributed to this conference we conclude that the state of the art of our field is such from a theoretical point of view, that it needs new typologies. The inductive method seems to be implied by many researchers as a possibility for making progress. This can be concluded from the contributions (plenary lectures as well as session papers) because they are all either essayistic or an attempt at model building or descriptive-comparative or merely descriptive. According to us it is time that a systematic way be found in this ever growing amount of material which has been collected for one reason or other, if only by formulating tentative interpretative schemata or descriptive concepts, in order to make some generalisations possible. Seen from this state of affairs, induction is not at all without importance in the field of the sociology of language. We hope that the debate on this topic will be stimulated by publishing these contributions. From the above it should be clear that there is a need for theoretical work in the heterogeneous field of minority language studies. For that reason we invited plenary speakers to represent different theoretical perspectives currently present in the discipline. Some of them have given a very general approach, others turn to more special theoretical dimensions. Their contributions are all published here in this volume, as well as some of the session papers. The theoretical diversity shows clearly in the papers. There are some important contributions towards building more integral theories. Thus, in Fishman's paper we find an 8-stage model on Reversing Language Shift, where the factors in one stage are decisive for a next stage. In the contribution by Giles, Leets & Coupland an extensive discussion is found of several models of cognitive and structural perspectives which is a sort of 'complete' overview in a nutshell. More specific issues are at stake in the following papers. De Vries presents a methodological-conceptual skeleton of the perspective of demolinguistics. Skutnabb-Kangas underlines in her contribution the importance of a critical review of the conceptual apparatus used by researchers. Edwards' and Anderson's contributions were included for their relevance for typology construction. Grin's paper contains a promising and innovating approach by taking the discipline of economics as its perspective on minority language problems. All those papers lean strongly on the social scientific side: just one paper, by De Haan on syntactical interference in Frisian, has a strictly linguistic point of departure. The productivity in our field seems to grow exponentially. The first two conferences could be encompassed within one volume of papers (Haugen et al., 1981 and Molde & Sharp, 1984). The papers of 2.ICML were the first to be published as a special issue of the Journal of Multilingual and Multicultural Development. With the 3.ICML this was again the case, but the organisers had so much material that they also published a second separate volume of

< previous page

page_2

next page >

< previous page

page_3

next page > Page 3

(Celtic) papers. With 4.ICML we would have had enough material for three volumes, perhaps even more, if we had gone ahead and published all submitted papers, as happens in other quarters. We decided on a selection, and quality was our prime, though not only, arbiter. This special number of the Journal of Multilingual and Multicultural Development contains the first selection of papers, in particular the plenary and typology-construction papers. A second set of selected papers, is being published as a separate volume, subtitled 'Western and Eastern European Papers'. A table of contents of the second volume is also included. That second volume is not just limited to 'the old Viking areas' as Mikael Reuter has named them, but has papers from all over Europe, as well as Israel. It contains new, hitherto unknown, material on relatively less well known cases and adds to other empirically based, statistically sophisticated research on the better known cases. We want to thank all 137 active participants from 22 different states, who have been an obvious ingredient for the success of the conference. The list of participants is also included. We are grateful to those that were behind the previous conference for asking us to host this one. We are proud that during this conference at least one person from the organising committee of each of the former three conferences was present as well as the organiser of 5.ICML to be held in Ottawa in 1992. This is proof of the continuity of these conferences and of our work in the field of minority language community studies. This field of research is prospering and also has attracted many new persons over the years. The Fryske Akademy has already for over half a century occupied itself with research relating to the province of Friesland, its inhabitants and their language. Its three main disciplines are linguistics, social sciences and history. The Fryske Akademy realises the special relevance of activities for autochthnous lesser used languages in Europe. Publications in international journals are stimulated. Morever, relevant meetings are being organised, such as this Fourth International Conference on Minority Languages. The Board of the Fryske Akademy is fully in support of initiatives in the field of the sociology of language and comparative studies of lesser used languages. As a research institute the Fryske Akademy wants to give its international activities a more enduring base by participating in the recently established MERCATOR-network for information, documentation and research of lesser used languages in Europe. The conference organisers gratefully acknowledge the financial assistance from the Commission of the European Communities, the Task Force on Human Resources, Education, Training and Youth and from the Fryske Akademy. We are indebted to the Hotel Management School for providing us with the premises for the conference and to its director, staff and students for the unsurpassed hospitality. For their enjoyable reception of the participants we also thank the Municipality of Ljouwert/Leeuwarden, the Municipality of Dokkum and the Prov-

< previous page

page_3

next page >

< previous page

page_4

next page > Page 4

incial Government of Friesland. Assistance in kind was also received from Multilingual Matters, and moreover the stimulus from Mike Grover and Derrick Sharp was invaluable. We, as researchers in the field of linguistic minorities, have many different special interests, but we also share a common goal. We are not just neutral, indifferent spectators merely watching the processes and developments that go on in those language communities. All of us are engaged in one way or the other, many of us being members of the very communities we study. In other words, we are involved in the future of these languages and language communities. And we hope that the work we do can and does make a difference. DURK GORTER, JARICH F. HOEKSTRA, LAMMERT G. JANSMA AND JEHANNES YTSMA FRYSKE AKADEMY DOELESTRJITTE 8, 8911 DX LJOUWERT/LEEUWARDEN, THE NETHERLANDS References Haugen, E., McClure, J.D. and D. Thompson (eds) (1981) Minority Languages Today. Edinburgh: Edinburgh University Press (Updated edition reprinted in 1990). MacEoin, G., Alqvist, A. and O'hAodha, D. (eds) (1987) Third International Conference on Minority Languages: General Papers. Special issue of: Journal of Multilingual and Multicultural Development 8 (1 & 2), 1229. MacEoin, G., Alqvist, A. and O'hAodha, D. (eds) (1987) Third International Conference on Minority Languages: Celtic Papers. Clevedon: Multilingual Matters. Molde, B. and Sharp, D. (eds) (1984) Second International Conference on Minority Languages, June 1983, Turku/Abo, Finland. Journal of Multilingual and Multicultural Development, 5 (34), 193349.

< previous page

page_4

next page >

< previous page

page_5

next page > Page 5

What is Reversing Language Shift (RLS) and How Can it Succeed? 1 Joshua A. Fishman Distinguished University Research Professor, Social Sciences, Emeritus, Ferkauf Graduate School of Psychology, Yeshiva University, Bronx, N. Y. 10461, USA Abstract Several societal and social biases have conspired to make the study of language maintenance and language shift more advanced than the study of reversing language shift (RLS). RLS efforts have been confused with messianic (i.e. irrational) and past-oriented (i.e. nativistic) movements, overlooking their rational, priority setting and modernising thrust. Even those engaged in the study or practice of RLS, however, have tended to lack theoretical coherence and to be mesmerised by 'activism' rather than by the empirical relationship between any particular RLS efforts and the demonstrable intergenerational transmissibility of language-imbedded behaviours, attitudes and beliefs. Where bilingualism with diglossia is all that can be realistically attained, RLS emphases must concentrate on family-neighbourhood-community building boundary-setting efforts. Where largely monolingual cultural autonomy becomes realistically possible, more inter-group confrontation RLS efforts should be undertaken, but their link to intergenerational transmissibility still requires explicit attention. (Post)modernisation poses serious new RLS problems for the family-neighbourhood-community, and the school as well, making it imperative for RLS efforts to be incentive-related far above and beyond their initial language-insociety (Xmen-with-Xish) ideals. It is no exaggeration to say that millions of people throughout the world are consciously engaged in efforts to reverse language shift and that many hundreds of thousands do so as members of movements whose explicit goal is RLS. Yet the efforts of these millions and the goals of these hundreds of thousands have been relatively little mentioned in the social science literature and have remained only infrequently referred to even in the sociolinguistic literature. Part of the reason for this ethically unjustified and intellectually as well as practically disappointing state of affairs, it seems to me, is that both the social sciences as a whole and sociolinguistics in its own right have sliced up their treatments of social movements in general, and reformatory

< previous page

page_5

next page >

< previous page

page_6

next page > Page 6

or protest social movements in particular, in such a way that RLS never clearly appears as the distinctive phenomenon that it is. This is not terribly surprising. The modern, Western social sciences have only very recently come to recognise socially patterned language use and socially manifested behaviour towards language as topics to be reckoned with. Sociolinguistics itself, on the other hand, even in its more RLS-sympathetic (although less intensively cultivated) 'macro' or 'sociology of language' pursuits, has not yet arrived at a sufficiently refined taxonomy of language status planning to explicitly provide for the consideration of RLS activity. As a result, very refined terminological and conceptual distinctions are made with respect to the 'minus' side of the ledger (we speak of language attrition-shiftendangerment-loss-death and can itemise many studies of each way-station along this increasingly negative progression), while the 'plus' side remains rather gross and undifferentiated and studies of revival, restoration, revitalisation and restabilisation remain proportionately few and far between. At the same time, language 'status planning', of which RLS is a sub-category, is overly identified with central governmental efforts, hardly the most likely or the most sympathetic auspices for minority RLS efforts. The most general reason for the neglect of RLS is probably the fact that RLS is an activity of minorities, frequently powerless, unpopular with outsiders and querulous amongst themselves; it is an activity that is very often unsuccessful and that strikes many intelligent laymen and otherwise intelligent social scientists as 'unnatural', i.e. as counter to some supposedly natural drift of historical events or the obvious direction of social change. It is hard for self-serving mainstream intellectual spokesmen and institutions to be sympathetic to the lingering, cantankerous, neither fully alive nor fully dead quality of many (perhaps most) efforts on behalf of receding minority languages (and the majority of sidestream scholars, too, are ultimately dependent on the mainstream for their perspectives if not for their very livelihoods). Indeed, RLS efforts are often like the 'gomers' or 'crocks' that constantly reappear in the emergency rooms of major metropolitan hospitalselderly, complaining individuals who neither die not get better and for whom nothing effective can seemingly ever be done. Most young doctors, like most majority spokesman in other fields, learn to 'meet 'em, greet'em and street 'em', i.e. to make light of the complaints of these embarrassing unfortunates and to turn to other, more tractable cases as expeditiously as possible. Crocks take up scarce resources (staff time, energy, funds, supplies, equipment) and contribute disproportionately to staff burn-out. They are no more than obviously 'suspect' and unpopular reminders of the failure of modern medicine to be able to cope with chronic social and individual health problems, particularly those that are characterised by a goodly overlay of social pathology, on the one hand, and that are seemingly irreversible, on the other. Minorities that are struggling for their very lives, for dignity, attention and affirmative action are inevitably suspect and unpopular. Both RLS efforts and gomers are unwelcome testimony to

< previous page

page_6

next page >

< previous page

page_7

next page > Page 7

shortcomings of the mainstream and to the tremendous will of the neglected and the 'different' to lead their own lives and to find their own satisfactions, regardless of outside pronouncements that nothing can or should be done for them. RLS Among the 'Social Movements' RLS efforts may very well be an individual activity, even the activity of an isolated individual, but they are much more characteristically a socially patterned and organised activity of the type that sociologists refer to as 'social movements'. But where, exactly, do RLS efforts belong in the long array of types of social movements studied by social scientists? RLS movements not only differ in many respects from the 'collective behaviour' phenomena (crowds, mobs, panic scenes, riots, etc.) that sociologists of an earlier generation so frequently studied, but they also differ interestingly from the types of bona fide social movements that have elicited more recent sociological attention. Like the latter, RLS efforts have definite goals, they are enduring and organised, and like some of them, too, they are commonly enacted outside of 'normal' institutional channels (e.g. mainsteam political parties, voluntary organisations, schools, media, etc.) and, indeed, are often oppositional to such institutions and tend to set up alternative social institutions, organisations and structures of their own. However, there is often about RLS efforts a very palpable degree of affect, a sentimental (rather than merely an instrumental) bonding, a stress on real or putative ethnokinship, an aspiration towards consciousness and identity (re)formation, a heightened degree of altruistic self-sacrifice and a disregard for 'least effort' advantages, to the degree that RLS behaviours often impress outsiders as bordering on the 'irrational' and the 'mystic'. It is perhaps the latter characteristics that have tended to elicit fear, suspicion and rejection in the mainstream and that have led to frequent charges that RLS efforts are a species of 'collective behaviour' after all, rather than manifestations of the 'social movement behaviours' of the comfortably rational and familiarly materialistic mainstream or mainstream-proximate types. These fears and suspicions have raised a series of road-blocks to the appreciation of RLS and other ethnic 'behaviour-and-identity' movements during the past century and a half. 'Irrational' Ethnocultural Behaviour-and-Identity Movements There is, of course, a long history to the charges of 'irrationality' by those who are in control of secure ethnocultural establishments of their own against those who are without such control and seeking to attain it. Lord Acton criticised the 'preposterous' ethnocultural claims aroused by Herder

< previous page

page_7

next page >

< previous page

page_8

next page > Page 8

and fostered by the French Revolution among the peoples without states, and 'therefore without histories' throughout Europe. Subsequently, many of these claims were sufficiently powerful to be attacked from below (rather than from above as was the case with Lord Acton), by Marx and Engels, in their efforts to foster a new and presumably more rational proletarian identity. Again and again, ethnicity has been delegitimised in the West (by both secular and Church spokesmen) as anti-modern, anti-intellectual, irrational, anti-progressive and anti-civil. Little wonder, then, that modern social science should also be heir to this tradition, 2 particularly given the fact that so much of modern social science is American and America views itself (and its identity) as universalistically supra-ethnic rather than parochially ethnic in the deeply traditional (and self-styled deeply historical) Old World sense. One of the earliest social science attempts to understand the seemingly non-modern and anti-modern identity reestablishment efforts of small peoples was Ralph Linton's (1943) analysis of 'nativistic movements' among small, overrun, indigenous peoples in the colonial empires established by the modern West. Linton described as 'nativistic' any conscious, organised attempt on the part of a society's members to revive or perpetuate selected aspects of their severely dislocated culture. Although Linton differentiated between 'revivalistic' and 'perpetuative' nativism, as well as between 'magical' and 'rational' efforts on behalf of either, he devoted major attention to the revivalistic-magical quadrant in his four-fold table and considered nativism as a whole to be a reaction to the unbearable oppression, dislocation and domination of Western rule. Implicitly, therefore, there is a non-Western flavour to the entire phenomenon, such that the cultural self-protective efforts of small Western populations at the mercy of non-Westerners would not only be deemed 'perpetuative' and 'rational' but might be said not to be nativistic at all, but, rather, protective of Western civilisation. Clearly, Linton did not anticipate either 'revivalistic' or 'perpetuative' RLS efforts within the very West itself (where, indeed, the lion's share of such efforts have occurred) and his conceptualisation of cultural behaviours is excessively dichotomous; it is either X or Y, Western or non-Western, whereas most RLS efforts envisage more complex, more contextual and situational repertoires composed of ingredients of X and Xishness and Y and Yishness. Anthony Wallace (1956), writing a dozen years after Linton, as well as Bernard Barber (1941), writing just a few years before Linton, contributed an interest in 'Messianism' to the discussion of nativistic efforts. For Barber, such Messianism is the mystic solution to a cultural impasse; it is an attempt to find supernaturally derived stability and hope in a culture which is otherwise in realistic shreds and tatters. For Wallace, Messianism, nativism and millenarism are all types of 'revitalisation', i.e. movements which emphasise the elimination of alien persons, customs, values and/or artifacts. Thus, both Barber and Wallace focus essentially on irrational and backward-

< previous page

page_8

next page >

< previous page

page_9

next page > Page 9

looking solutions, and, as a result, would not have much to say about the bulk of RLS efforts which are really attempts to arrive at self-regulated modernisation, i.e. at Xish modernisation that is in the spirit of and under the aegis of Xishness as defined by 'Xmen-with-Xish'. Western social science seems to be primarily telling us that modernisation and authenticity preoccupations cannot go together, just as authenticity preoccupations and rationality cannot go together. The seeming anomaly of modernisation, rationality, affectivity and authenticity, the cornerstones of RLS, are theoretically unprovided for because such a complex combination strikes modern social science as perversely contradictory, as 'so near and yet so far' from the mainstream ethos. The true complexity of modern minority movements is elusive, doubly so since such movements are contraindicated given the simplistic theories according to which mainstream processes and virtues are considered simultaneously prototypical, normal and inescapable. In many ways, Russell Thornton's (1986) retrospective analysis of the Amerindian Ghost Dance movements of 1870 and 1890 hews close to the irrational and backward-looking characterisations encountered in Wallace's discussion of revitalisation movements, as one might expect from a consideration of manifestations that predicted demographic/cultural recovery as a byproduct of dances that would bring the dead back to life. Nevertheless, Thornton does add one crucial new twist to the discussion thus far: a realisation that such movements are at least phenomenologically rational and that they may be sufficiently motivating to be productive 'resource mobilisation' as well. However, Thornton's magnanimous admission that the ghost dances 'were deliberate responses ... that probably made sense to the Indians involved ... (and) in terms of their culture ... (were) essentially rational acts' is not only a condescending tautology but implies that no such rationality would obtain from the point of view of modern, Western culture. Setting aside the fact that Thornton overlooks the frequency with which modern Western populations engage in efforts to protect their sanctities (sanctities that are obviously above and beyond the rational), the rationality of goals and the rationality of means must always be analytically separated. RLS is usually a thoroughly modern enterprise in terms of the rationality of its means (any implied equation between RLS and ghost dances being totally out of order), while its goals admittedly partake of the rationality of modern religious and ideological verities that a major portion of mankind considers to be worth struggling toward, regardless of price in time, effort and resources. Still provocatively interesting for us today, however, more so than any of Thornton's judgemental comments, are his conclusions that the Amerindian efforts to which he refers were most extensively and rewardingly engaged in by those very cultures that had been most dislocated and that, in toto, these efforts were basically attempts to re-establish group boundaries. Neither of these considerations requires notions of 'irrationality' but rather, as Thornton fully realised, are fully consistent with 'resource mobilisation' and 'relative deprivation' theories within the social sciences.

< previous page

page_9

next page >

< previous page

page_10

next page > Page 10

A further step along the path of re-rationalising and de-mystifying revitatisation movements was recently taken by Duane Champagne (1983). He realises that there are basic similarities between the revitalisation movements that occur in structurally less differentiated societies and the reform or guided cultural change movements that occur in structurally more differentiated (= more modern) ones. Both types of efforts utilise the most effective means available to their societies, those of the latter societies being organisationally, institutionally, materially and conceptually more advanced and, therefore, more capable of accepting the inevitability of cultural change and able to influence its outcome via political and economic means. Champagne's analysis should once and for all remove the penumbra of backwardness and irrationality from efforts of the RLS types. Such efforts may fail or prevail, depending on rationally analysable factors (i.e. they do so on the basis of means and circumstances totally like those that govern the success and failure of the other social movements with which they must compete), differing from their contemporary competitors more with respect to ends than to means. Indeed, given the scarcity of means that most RLS efforts have at their disposition, a good case can be made that they often attempt to be more rational with respect to their deployment than is frequently the case for movements that are socially ascendent. 3 All in all, attempts to convince the modern mind of the rationality of ethnocultural behaviour-and-identity re-intensification movements have experienced some success during the past half century, but much movement in this direction is still necessary before RLS efforts will be commonly viewed as the natural, thoughtful and constructive undertakings that their participants take them to be. RLS, 'Backward Looking' Resistance to Change, and Cultural Conservatism Another oft-repeated stereotypic charge is that RLS and related re-ethnification or ethnic re-intensification movements are backward-looking ('past oriented'), conservative, change resistant dinosaurs. Of course, most basic philosophical values tend to have their origins in the past and small cultures that are now in particular danger of erosion naturally recognise a past when that was not (or not as much) the case. RLS efforts are very sensitive, due to their very goalconsciousness per se, to the constant diminution in the numbers or proportions of speakers/users of the language-inculture on behalf of which they struggle, to the incursions of time, to the fact that things were better 'then' than they are 'now'. But this does not need to make them more 'backward looking' (if by that we mean: seriously pursuing a return to and a preservation of the past) than are most other opponents of present evils, injustices and dislocations. Shall we designate as 'backward looking' all those who remember when urban neighbourhoods were much

< previous page

page_10

next page >

< previous page

page_11

next page > Page 11

safer cleaner and far less polluted than they are today, merely because they strive towards a closer approximation towards past superior standards in these respects? Many others, besides RLSers, yearn for social and cultural arrangements that will foster stronger family bonds of affection, mutual care and concern, respect, ethical behaviour and commitment. Many others, outside the RLSers, realise that local communities must be more fully involved in their schools, health agencies, playgrounds, zoning regulations, and child care services, if they are to overcome the problems and inefficiencies of the growing massification and bureaucratisation of modern society. One does not need to be a member of either the Old Order Amish or the Institute for Cultural Conservatism to bemoan the general lack of intellectual concern for the moral and spiritual dmensions of modern life and, accordingly, it is unfounded to accuse those RLSers who have any or all of the above concerns of displaying a stultifying opposition to modernity. As some social scientists have already recognised, 'defenders' of the core values of modern democratic systems are apt to perceive and point out real threats to the well-being of that system. Such defence should really be viewed as part of the process of change, part of the direction-finding or direction-setting field of forces that we call change, otherwise the empty cycle of 'change for the sake of change alone' will be upon us. 4 RLSers have sometimes been accused of goal fixation to the point of forgetting why the goal is being pursued to begin with. But certainly this charge also applies to those who are 'modernists in principle' while forgetting the humanistic religious, philosophical and ideological goals that make life human, purposeful and worthwhile. But in reality, RLSers are not merely not defenders of some mystical, mythical and bygone past; they are actually 'change-agents on behalf of persistence'. Very few social scientists indeed have been inclined to conceptualise and analyse the relationship between change and persistence, probably because of our modern fascination with the dynamics of change per se. But all change is interspersed with persistence as well, just as all persistence is interspersed with ongoing change. Persistence no more means equilibrium than change means chaos. The forces and processes of change coexist, in a single process, with the forces and processes of persistence, and what most social scientists mistakenly call 'change' is really the by-product of the interaction of persistence and change. Actually, the power of persistence helps provide the direction and generate the resultant of the total dynamics that are operative at any time and place. Ethnolinguistic persistence involves a basic continuity in the meaning of symbols. For RLSers a given language is the first and foremost of these symbols, as is their interpretation of that language as being truly fundamental to identity and continuity. Although RLS is rendered difficult by the values, movements and processes opposed to it, it is also rendered difficult by its inevitable interweaving with ongoing sociocultural change. Experienced RLSers realise that all cultures are constantly changing and that their goal is merely to regulate and direct this

< previous page

page_11

next page >

< previous page

page_12

next page > Page 12

change, so that it will not contradict or overpower the core of their cultural system, rather than legislate change out of existence. Identity persistence and ethnocultural persistence are not synonymous, of course. The former is purely phenomenological and is no basic criterion of RLS, i.e. it can be attained without RLS, while the latter involves a studied persistence of behaviours, i.e. commitments to implemented interpretations (of life, of relationships, of history, of symbols). RLS requires societal boundary maintenance, rather than merely being the result of such boundary maintenance. Language is a prime boundary-marker and protector, because it not only implies and reflects core boundaries but because it constantly creates and legitimises them as well. RLS seeks to avoid the dislocations that inevitably result from the destruction and substitution of core symbols, behaviours, boundaries and values, possible though it may be to come through such destruction and substitution with one's phenomenological social identity intact. It is not change per se that is opposed by RLSers but changes in a core behavioural complex in which the language is generatively and regeneratively linked to the protected cultural core. For the persisters, language is both corpus and message and the authentic message without the authentic corpus is as empty as the authentic corpus without the authentic message. Nevertheless, a corpus can and must generate endless novel 'messages too, since the novel ones not only enable the authentic ones to achieve their contrastive sanctity but assure that sanctity of a new life, timeliness and vigour as well. As Spicer (1980) recognised ever so long ago, 'a people that endures ... embodies the most important kind of social unit which men can create: a living, cumulative interpretation (and here I would add: enactment) of human life ... collective purpose and destiny'. Only by persisting in the midst of change, only by indigenising change, only by taming and refashioning change (thereby taming and refashioning persistence too) does RLS reflect a creative guarantee as to its living potential, rather than degenerate into some totally lifeless, antiquarian oddity. It is not the return of the past that RLS seeks, but the mining of the past so that the core that animated it can continue to be implemented. 5 For all of its fascination with change, much of the thoughtful West is also 'past appreciative'. For all of their use of the past, most RLS movements and efforts are futureoriented. The Oppressiveness of Tradition and the Pursuit of 'Authenticity' Although RLS is a type of sociocultural change, it is not of the type that the social science literature has classically and generally attended to. There is, of course, some recognition of a 'cultural paradigm' in mainstream social science discussion of social change, a paradigm that views innovation as an ongoing process that is literally impossible without continuity of beliefs,

< previous page

page_12

next page >

< previous page

page_13

next page > Page 13

values, behaviours and symbol systems. However, social theory knows much more about fostering change (the 'structural paradigm') and advancing individual authenticity via liberation from repressive societal regulation than it knows about fostering continuity. Indeed, change for change's sake has come to occupy a somewhat hallowed position in social theory and has influenced the social sciences as well. John Stuart Mill, well before the midnineteenth century, equated tradition with unhappiness. Thereafter, the view that humans can be themselves ('authentic') only if they live without any imposed social structure at all has occupied a central niche among Western social thinkers. Even earlier, Edmund Burke had proclaimed that repression was a basic fact of social life (power for some and obedience for the rest) and Montesquieu and Rousseau had anticipated him by declaring that to maximise one's happiness each individual required freedom to shape a happiness that would be completely of his own design. If the need for group liberation was acknowledged at all (as it was by Montesquieu, who, nevertheless, concentrated on sexual, religious and political groups but was strangely oblivious to ethnocultural groups) such liberation was justified on the basis of the happiness that would be derived from being freed from ties to groupness, i.e. authenticity was viewed as liberation from groupness rather than as liberation of groupness or for groupness. Only Rousseau vacillated, late in life, recognising that the emptiness and alienation of the totally free required for their alleviation those very communal bonds that decreased individual freedom. He reconciled the two, individual happiness and communal affiliation, only in utopian terms and recognised, even so, that utopias themselves were, therefore, at odds with the openness and the constant changeability of urban modernity. Among the more or less systematic schools of social thought only the Herderians, the nationalists, the racists and the Marxists defended carefully selected (and very different) affiliative ideals for any length of time after the flawed and failed spring of '48, and none of these were ever really taken seriously by the mainstream of Western social science. Furthermore, before the recent appearance of the new, post-industrial left, the Marxists themselves were as classically anti-ethnic as were the bourgeois thinkers. 6 Perhaps of greater import for recent theoretical opposition to planned ethnicity-fostering culture change, such as RLS, is the 'alternative society' or 'commune movement' of the 1960s and 1970s. It harks back to Montesquieu in its abhorrence of mechanistic, dehumanising, competitive and materialistic societal ties and by emphasising individualistic and counterculture, rather than mainstream or traditional cultural definitions of 'success'. The new 'life-style', studiously supra-ethnic in nature, stressed openness, intimacy, flexibility, co-operation and altruistic sacrifice, Gemeinschaft features that theoreticians had once ascribed to the ethnically integrated, small community but which the communes derived independently from their anti-establishment vantage point.7 Many RL efforts reveal a close similarity to the anti-establishment and anti-materialistic Gemeinschaft stri-

< previous page

page_13

next page >

< previous page

page_14

next page > Page 14

vings of the 'communes', thereby underscoring even more the void that so often (and so needlessly) separates the ethnic dimension and the radical dimensions of modern social criticism. Clearly, 'traditionless authenticity' and 're-ethnifying authenticity' are poles apart, regardless of the term 'authenticity' that they share. Their common stress on achieving selfregulatory status is often overlooked, primarily because the former is individualistic and the latter is group-cultural in orientation vis-à-vis the attainment of happiness. The perfect combination of both might well be optimal, but requires as much acceptance of the claims of (minority) ethnoculture as' of the claims of individuality. Modern Western thought has generally been more willing to suffer the pains of the latter rather than grant the legitimacy of the former. Growth in Theory Pertaining to Ethnicity A final area of attempted rapprochement and accommodation between Western social theory and ethnic reintensification efforts pertains more directly to an understanding of ethnicity per se. The ethnic revival of the mid1960s to the mid-1970s roused both bourgeois and leftist thinkers from the stupor that had clouded their thinking about ethnicity for over a century. It came to be belatedly recognised by those who sought to grapple with the far-flung (although not overly deep) ethnic stirrings in most parts of the First and Second Worlds, that many of the myths and biases that had previously coloured their views of ethnicity (irrationality, backward-looking focus, conservatism, oppression of individual 'authenticity' culminating in sociocultural/political oppression) were substantially erroneous or unfounded. Ethnicity did not involve attempts to preserve the traits of either static or pre-modern cultures. The transitions between and the combinations of traditional and modern behaviours and sentiments were actually exceedingly varied and constantly ongoing in all populations, including those that were non-state forming. States and non-states differed not only in market articulation and in power but in social organisation and in historically deep cultural manifestations. The greater power of the state ('the most successful predatory form of social organisation') might force irreversible changes upon non-state societies, but this was not as predictable an outcome towards the end of the twentieth century as it had been in the nineteenth, given the relative increase in non-state resources and the use of modern methods in all spheres of life (including non-state spheres), while the finite nature of state resources became ever more apparent. Indeed, the influences of non-state entities on their surrounding 'host' states came to be increasingly evident. Resource mobilisation theory, important though it continued to be for the study of social movements, seemed to provide less new insight into the persistence of indigenous sidestream ('peripheral') ethnicity than had initially

< previous page

page_14

next page >

< previous page

page_15

next page > Page 15

been hoped. Sidestream ethnicity resists reduction to the level of grievances (and manipulable, largely specious grievances at that) and the ethnic revival requires for its explanation additional factors, above and beyond the combination of minority grievances (old or new) and increased minority resources. Declining status inequalities between 'central' and indigenous 'peripheral' populations became quite noticeabledue both to the growing ethnic middle class which, nevertheless, did not forego its cultural identity as a means of social mobility, and the growing competitiveness of the latter vis-à-vis social rewards (including the reward of acknowledged cultural legitimacy). Indeed, the unexpected versatility of re-invigorated ethnic movements clearly highlighted their ability not only to maintain an indigenous leadership but to tap and activate indigenous resources of dedication as well. The ability of universality and particularism to develop and coexist simultaneously within the very same populations was a rude awakening for both Marxist and non-Marxist theorists who had assumed that industrialisation, urbanisation, modernisation and the spread of education would inevitably reduce ethnic consciousness and lead to the demise of narrower loyalties in favour of broader ones. This prediction was not confirmed, narrower and broader loyalties being far more syncretistic than theory-conscious intellectuals had imagined, and the post-industrial Left was forced to identify with rather than continue to reject the cultural self-regulatory aspirations of ethnic minorities in the West itself. Bourgeois thinkers too went through much soul-searching and reformulation in the light of the evident significance of ethnicity within mainstream academia itself. Ethnicity efforts came to be viewed as reformist of mainstream insensitivity, much like pro-environmental, anti-sexist and anti-bigness or anti-industrial-growth efforts. However, even this is not the entire story, since ethnicity efforts are sometimes allied with conservative political, religious and moral reform efforts as well. Ultimately, however, although much has changed in mainstream thought concerning ethnicity since Lord Acton, most recent developments in ethnicity theory still treat ethnicity as reactive to or as transformational of other, 'more basic' material circumstances and aspirations. This evident disinclination (even after 150 years of painful and reluctant theoretical change) towards accepting ethnicity in its own right, as a permeable, changeable but ultimately also quite robust and recurring identity-values-behaviour complex, a complex that situationally influences aspects of mainstream as well as sidestream life, even under the most modern circumstances, remains a blind spot in social theory that only further 'rethinking' can overcome. The ultimate theoretical contribution of RLS research to general social science theory is its rich potential for providing further empirical and theoretical perspective contributory towards exactly such rethinking. 8 Rather than being viewed as threats to the state or as byproducts of split labour markets or even of boundary maintenance processes per se9 (none of which, by the way, give signs of disappearing from the horizon in the foreseeable future), ethnicity and ethnicity movements must

< previous page

page_15

next page >

< previous page

page_16

next page > Page 16

come to be appreciated more ethnographically and phenomenologically. i.e. more from the point of view of the insider who experiences them rather than from the point of view of the outsider who views them, telescopically or microscopically, from afar without appreciating, therefore, their affective significance. So terrified are most Western thinkers of the charge of 'primordialism' that they refuse to understand the recurring appeal of primordialism to common folk the world over. Like physicians who refuse to appreciate the common man's dread of cancer or AIDS, they therefore, unknowingly but intrapunitively limit their own ability to understand the condition they are presumably treating. The ethnic rejection of the mainstream-identity as representing one's sole and complete identity (and, indeed, the growing search for sidestream roots) implies an acceptance of self and one's origins that is also rich in potential for better understanding of others, and of inevitable links to others, as well. If the simplistic ethnic myth of fixed, homogeneous and completely bounded cultures must give way to a more realistic sense of the changeability and intersectedness of all cultures, the awareness of this myth, on the one hand, and, on the other hand, the absolute necessity of undertaking attempts to cultivate the threads of intimacy, involvement and historical relevance, so that meaningful, unalienated social existence remains possible, are often better realised by minorities than by the majorities that smugly disregard, abuse or regulate them. The real question for modern life and for RLS is not whether this is a discrete or interacting world, but, rather, given an incredibly complex field of interacting forces, how one (not just minority ethnics but any social movement) can build a home that one can still call one's own and, by cultivating it, find community, comfort, companionship and meaning in a world whose mainstreams are increasingly unable to provide these basic ingredients for their own members. RLS Theory: A Perspective for Rational Effort to Build and Safeguard Gemeinschaft Aspirations So That They Will Be In Touch With But Not Inundated by the World at Large The eight stage analysis of and prescription for RLS that is presented here is an alternative planning theory in the sense that it attempts to bridge the gap between social science and societal reform. Planning scientists and planning practitioners tend to look down on one another, trading charges of conceptual poverty and lack of realism (and, therefore, lack of validity). Most of the explanatory theories advanced in language planning for example, do not reveal, and, therefore, cannot provide, insight into the struggle of some societies towards intergenerational linguistic continuity. Appropriate RLS-status planning can only occur if the societal link between generations is constantly kept in mind and if every putative RLS effort is tested by the

< previous page

page_16

next page >

< previous page

page_17

next page > Page 17

question 'how will this effort reach into and reinforce the intergenerational link?'a link that must take place early, affectively and verbally if RLS is to come about. This does, not mean that well-grounded RLS theory can provide us with a blueprint for the future. The future cannot be reduced to a series of technical applications of theoretically formulated steps, neither in economic planning, in agricultural planning, in educational planning, in family planning nor in RLS planning. The best that RLS theory can do is to provide greater societal perspective for negotiating the difficult priorities that any RLS effort inevitably involves. As with all other types of social planning, RLS planning will inevitably be accompanied by unexpected side-effects and even negative consequences. However, there is no alternative modern route to social problem-solving than the route via planning. For the advocates of RLS there is, therefore, no dilemma as to whether RLS planning should occur. If there is a dilemma, it deals with the how of RLS, i.e. with a systematic overall approach that can guide the efforts that must be undertaken. To begin with, even before concrete efforts are undertaken, RLS involves 'consciousness heightening and reformation'. The importance of ideological clarification and awareness for the process of directed cultural change can easily be exaggerated but it cannot be denied. RLS behaviours cannot challenge conventional institutions and mainstreamderived ideas as to the role of Xish without fully clarifying the ideal of 'Xmen-with-Xish'. It is hard enough to row against the current; it is virtually impossible to do so without knowing where one would like to get to. Any organised activity, particularly ethnically related organised activity, immediately raises questions of right and wrong, desirability and undesirability, legitimacy and illegitimacy, possibility and impossibility. RLS advocates must explore these issues as frankly and as openly as their surrounding political culture permits, i.e. not only with each other, but also with those who are RLS-uninterested and with those who are RLS-opposed. It is always easier to communicate only with those who are already converted, however, those who do so inevitably face the danger of ceasing to explain basic premises even to themselves. The premises that Xmen are not Ymen and that Xish culture (daily and life-cycle traditional observances, distinctive artifacts, beliefs and values, exemplary literature, art, music, dance, etc.) is not Yish culture must not be skipped over, no more than the premises that Xish culture is worth maintaining, that it can be maintained only if it becomes more self-regulatory, that one of its main props and creations is its own language and that the latter must be fostered in as many domains of individual and social life as are intra-culturally acceptable and feasible. Any such exploration will inevitably be difficult and initially 'touchy'. It forces to the confrontational surface hitherto quiescent assumptions of what being a good Xman entails. Unless these assumptions arc clarified and consensualised, at or soon after the outset, all RLS efforts coming thereafter will be conflicted and contested

< previous page

page_17

next page >

< previous page

page_18

next page > Page 18

from within. The goal of fostering commitment to and implementation of a society of authentic 'Xmen-with-Xish' is a difficult one to attain, all the more so if language shift is already far along and the phenomenon of 'Xmen-via-Yish' has already spread, taken root and proved itself to be rewarding. The hoped-for benefits of RLS must be clearly spelled out and its implied, suspected or alleged debits must be openly faced. People cannot be tricked into supporting RLS. They must be convinced to accept a definition of their 'best interest' and 'most positive future' that depends upon and derives from RLS and from the rewards and self-regulatory capacity of the 'Xmen-with-Xish' stance. The first ones to do so will obviously be pioneers and must be particularly ready to work hard in order to attain very sparse results. All this becomes possible only when the RLS enterprise can count on the participation of maximally dedicated and ideologically oriented individuals. The crucial importance of self-aware ideological communities for the process of cultural change is well-documented 10 ('solidarity' is the currently fashionable codeword for this state of affairs, but it is much better that this be spelled out in terms of social domains and directed efforts than that another codeword be bandied about), even though it is quite clear, and will become even clearer in what I still have to say, below, that consciousness and ideology are not enough. They are merely the first of many concerns, all of which, taken together, constitute a theory and a model of the intergenerational transmission of language, culture, society and identity. Stages 8 to 5: RLS on the 'Weak Side' The notion of a graded series of RLS priorities is offered here as a heuristic device more than as a proven fact. Real life is always full of more complexities and irregularities than theory can provide for. As a result, there may be less implicationality or reproduceability in real life than the theory implies.11 Nevertheless, the notion of graded priorities in RLS efforts, even if it is less than perfectly validated, has two virtues: (a) the virtue of more parsimoniously and forcefully directing attention to crucial issues or 'first things first', and (b) the virtue of constantly directing attention to the absolutely crucial question of the link to intergenerational continuity. The first virtue is a significant one, because RLS, like all minority-based efforts, is more likely than not to be characterised by a serious shortage of resources. Accordingly, it is important to focus the meagre resources that are available in as judicious a way as possible. The second virtue constitutes a reminder that RLS must not be carried away by the most fashionable technologies or the most glamorous institutions that are so very much 'in the public eye'. When all is said and done, any and all seriously intended RLS effort must still stand the acid test of fostering demonstrable transmissibility across the intergenerational link. It is the achievement of that transmissibility, rather than the modernity and glamour of the means employed, that characterises a good investment of RLS time and effort. Of the eight post-ideological

< previous page

page_18

next page >

< previous page

page_19

next page > Page 19

steps that I have in mind, four are particularly urgent and germane to RLS efforts at their earliest and weakest stages, when political conflict and power goals cannot be afforded, allies are few and far between, and self-help is, therefore, the only dependable approach. Stage 8 The reassembly of the Xish language-model per se, is an obviously rock-bottom stage at which RLS can begin, once adequate ideological clarification has been attained. This stage applies not only for total language communities, as Australian Aboriginal examples indicate, but also for particularly dislocated networks of languages that are possibly still in good repair elsewhere, although often inaccessible; indeed, wherever fluent native speakers are no longer available and where even second language speakers command dubious fluency and correctness, a prior stage of reestablishing community norms of Xish grammar, phonology, intonation and prosody, ideomaticity and semantic typologies is highly desirable. This may call for the importation of outside specialists and teaching-learning materials that can provide models of the variety or varieties of Xish that are to be 'oralised' and/or 'litericised'. The alternative is to indigenise a new, non-native local variety of Xish. This is no sin, of course, and has occurred in many places (e.g. in Ireland and among Amerindian as well as Aboriginal language advocates), but it obviously exposes an ethnolinguistic authenticity movement, such as RLS, to the particularly difficult-to-rebut or embarrassing charge of inauthenticity. Although such charges are ultimately answerable (precisely because mainstream authenticity also has about it a goodly proportion of conscious and unconscious innovation), it is probably better if one can avoid this issue to begin with and, thereby, to be free to turn RLS-attention to code implementation rather than to remain preoccupied with issues of code definiton and specification. The role of linguists is most obvious at this stage, although not linguists alone. Linguists are notoriously poor at motivating and organising the societal devotion that is required if stage 8 is to be transcended and if RLS is to become a social movement rather than merely a monograph, a textbook or an X-as-a-second-language-course. Stage 7 This is a remarkably gratifying stage on the one hand, and a remarkably misleading one on the other, in so far as the true stage of RLS affairs is concerned. The fact that there is a large, still active, elderly population ('elderly' being defined as 'past childbearing age') that organises and partakes of endless Xish public events, rituals, ceremonies, concerts, lectures, courses, contests, readings, songfests, theatrical presentations, radio and television programmes and publications is of course, a tremendous societal

< previous page

page_19

next page >

< previous page

page_20

next page > Page 20

achievement and a great joy to those individuals who are personally involved in and enriched by these activities. However, from the point of view of RLS, all of these activities are merely rallies of the 'last Mohicans'. They serve to enthuse the already enthusiastic, to convince the already convinced. At best, they may be said to keep motivation high among the already committed and, in that way, to keep open the possibility that other means of RLS, via efforts that are linked to and involve the younger generations, may still be devised, adopted, implemented and emphasised. It is hard for a thousand 'old-timers' who attend an absolutely first rate 'pageant for Xish' to believe that on the morning after Xish is still no better off than it was on the night before. However, that is really the case, quite regardless of how many 'young guests' were also present on any one such particular occasion or another, because 'special events' of this kind are just that; as such they are simply not linked into the ongoing, normal, daily family-socialisation pattern. Their audiences disperse and there is no carryover from the ideological and aesthetic highs that these events often attain, to the concrete rounds of daily life and, most particularly, to the child socialisation nexus on which RLS ultimately really depends. This is not to say that stage 7 is useless; it is merely to say that it is ultimately useless if it too, like stage 8, cannot be transcended. Stage 6 This stage consists of family-, neighbourhood-, community-reinforcement (and of organised RLS actvity squarely aimed at each of the foregoing) and constitutes the heart of the entire RLS venture. It may be merely metaphorically enlightening to believe, as Kenneth Burke claimed (1954: 136), that 'men build their cultures by huddling together, nervously loquacious, at the edge of an abyss', but it is inescapably true that the bulk of language socialisation, identity socialisation and commitment socialisation generally takes place through intergenerationally proximate, face-to-face interaction and generally takes place relatively early at that. Spicer (1980), one of the few social scientists to have invested a professional lifetime in the study of 'persistent (though stateless) peoples', put it this way in his final summary: ... [T]he persistence of configurations of identity symbols depends on the kind of communication possible in local community organizations, uniting household groups. It is in the milieu of the effective local community ... that the basis for choosing to identify with an enduring people becomes established. (Spicer, 1980: 358) Unfortunately, knowing that RLS must always feed into and connect up with real, natural, daily community life if intergenerational transmissibility is to be attained, is no guarantee at all that such linkages can be brought

< previous page

page_20

next page >

< previous page

page_21

next page > Page 21

about. In modern, democratic contexts it is not at all easy to plan or engineer RLS efforts that focus directly on family-, neighbourhood-, community-building. Although many Jewish settlements and kibbutzim in Ottoman Palestine became bastions of Hebrew language revernacularisation, they functioned in this fashion at the tremendous price of tearing themselves away both from the Jewish life round about them and from the Jewish life, primarily Yiddish-speaking Eastern European, from which most of the RLSers and their followers had come. That they could make a virtue out of this double alienation is merely a testimony to the self-sufficient life-style that they were able to establish and the robustness of the sociocultural boundaries that they long maintained between themselves and other Jews (whom they obviously viewed as 'Xmen-via-Yish'). Even so, many of the Hebrew-revernacularising townlets and kibbutzim failed, as have almost all of the experimental communities established too expressly for language maintenance or RLS purposes. We will examine some of the difficulties and solutions that pertain to this stage below, after we have reviewed all of the stages. At this point, it must suffice to say that if this stage is not satisfied, all else can amount to little more than biding time, at best generation by generation, without a natural, self-priming social mechanism having been engendered thereby. For a language that has shrunk to 10% of its former 'realm', remaining at 10% may seem like an accomplishment of sorts, but it is also a confirmation to 90% of the population that the ideal of 'Xmen-via-Yish' is really the more viable alternative. This is why simply maintaining the stage 7 status-quo-ante is an undesirable longterm 'solution' for endangered languages. Attaining stage 6 is a necessary, even if not a sufficient, desideratum of RLS. Unlike other stages, when stage 6 is transcended it is not merely 'left behind'; quite the contrary: all subsequent stages must be diligently tied back to and connected with stage 6 if they are to contribute to the living reality of RLS rather than merely to its propogandistic hoopla or one-upmanship. Stage 5 This stage entails formal linguistic socialisation. Although such socialisation does not need to be restricted to literacy and literacy alone, that indeed is by far the lion's share of what this stage entails in modern settings. Whether restricted to literacy or not, this stage adds additional varieties to the learner's repertoire, above and beyond those that can be acquired in the largely oral and familiar interaction within most family-, neighbourhood-and communityintergenerational situations. The availability of more formal varieties (and, in modern life, reading/writing essentially involves more formal varieties than does most of speech) gives Xish a range which enables it to be more comfortable vis-à-vis the greater range that is normally available in Yish and Zish, due to the governmental and econotechnical functions of

< previous page

page_21

next page >

< previous page

page_22

next page > Page 22

the latter. Xish religious, legal and oral traditions can all be tapped for the elaboration of such more formal varieties and the attainment of literacy in one or another of these varieties contributes to the solidification of wider intercommunal bonds and the cultivation of additional support opportunities for RLS. Clearly, stage 5 entails a kind of schooling, one that may initially be open and attractive to adults as well as to children. However, it does not involve schooling in lieu of compulsory Yish administered education, and, as a result, can avoid many of the expenses and most of the Yish control-and-approval requirements that such education generally entails. Stages 8 to 5 constitute the 'programme minimum' of RLS. These stages do not involve major costs and they do not crucially depend on Yish cooperation. They are generally of the 'do-it-yourself' variety and, as such, can be approximated in most types of political and economic climates. They are particularly appropriate for numerically and politically weak language-in-culture settings and are not restricted in applicability to permissive democratic settings, although the latter are always more facilitative in so far as overt organisation efforts are concerned. RLS concentration on these four initial steps, particularly on any subgrouping of them that also includes step 6, generally presumes a bilingual model of Xish society in which Y/X diglossia is attained and maintained by surrendering to Yish all effective control over the more modern and interactive media and pursuits. Such diglossia is not a rare or impossible goal, nor is it a goal which inevitably consigns its adherents or practitioners to poverty, backwardness, non-participationism or isolation from the mainstream. Compulsory and higher education, economic opportunity and governmental service may still be entirely open to those who espouse and maintain the lower level 'Xmen-via-Xish' position, but, except for presumably minor and voluntary Yish accommodations via translation, usually in connection with absolutely crucial public welfare services and whatever media and public visibility Xish can obtain by means of the numbers and funds at its own disposal, the bulk of such opportunities will clearly be available to them only in Yish. That being the case, the future of Xish rests squarely on the relative impermeability of the intergroup boundaries and on the non-negotiability of the 'Xmen-via-Xish' position at stages 6 and 5. This being the case, a case which is often described and experienced, justly or unjustly, as 'second-class citizenship', it is clear why RLS movements often seek to push on beyond these stages into the upper reaches of sociosymbolic life. However, it should be clear that just as not to do so constitutes an unjustified foreclosing of RLS opportunity and sociolinguistic potential, so the premature crossing over of RLS efforts into the arena of the second four stages runs the risk of burdens and challenges that may be excessive, non-productive and even dangerous for the entire RLS enterprise. The choice between these two types of risk is a fateful one, indeed.

< previous page

page_22

next page >

< previous page

page_23

next page > Page 23

Stages 4 to 1: RLS on the 'Strong Side' Education in lieu of compulsory schooling involves an intrusion of the state into the life of the family, neighbourhood and community. Most democratic states provide for the possibility of substantially curtailing or substantively modifying (or pro-RLS orienting) this intrusion by means of private, parochial or proprietary schools. Such schools must still follow the minimal essentials of the approved general curriculum, employ state certified teachers in conjunction with teaching those minimal essentials and maintain facilities and schedules that meet state specifications pertaining to fire, safety, health and attendance standards. Otherwise, however, they are free to lengthen the school day and the school year to facilitate the addition of courses and experiences that are particularly desirable to them. The maintenance of such relatively independent schools (type 4a schools in our complete typology) obviously entails major costs for RLS-advocates and their supporters. These costs can sometimes be avoided if Xish speaking parents and actually or potentially Xish speaking children are sufficiently concentrated and if the Yish authorities are sufficiently co-operatively inclined to justify and to permit special RLS public school programmes for minority language children. Generally, such programmes are reluctantly and unreliably offered, are really compensatory in nature and orientation and inferior in educational quality. They do not foster either the image or the reality of Xish cultural self-regulation. They are particularly unsuited to the attainment of the goals of weak and inexperienced RLS movements, and the more centralised the Yish educational establishment is, the more unsatisfactory such schools are if the local attainment of RLS success per se is utilised as the criterion of 'success'. Even where schools of the latter type (type 4b schools we have called them in our case by case reviews) do not suffer from outright or hidden sabotage by the governmental authorities on whose personnel, funds and approval they depend, they present the danger of leading away from the Xish community of orientation and can yield positive results only if sufficiently surrounded by and embedded within an RLS oriented family-neighbourhood-community field of forces. Several localised examples of RLS-oriented type 4b schools exist to indicate that this goal is not impossible of attainment, but, on the whole, it is vey rarely attained and its attainment often depends on the prior or concurrent establishment in nearby areas of cultural autonomy arrangements that transcend schooling per se. In essence, they represent the conversion of compensatory 4b programmes into self-regulatory 4a programmes via the attainment of a political accommodation at a governmental level higher than the local school authority. In the absence of such higher and transcendent considerations, which obviously aim at dovetailing schooling with stage 6 institutions and processes, there is absolutely no reason to assume that schooling (even type 4a schooling) is either a guarantee of or even a prop for successful RLS. We must guard

< previous page

page_23

next page >

< previous page

page_24

next page > Page 24

against allowing our academic affiliations and general biases (which tend to make us view education as the universal panacea for any and all problems) to lead us prematurely to assume that schooling is 'the solution' to RLS problems more specifically. The Irish experience alone should disabuse us of that fallacy. Unreconstructed schools of type 4b, even more so than schools of type 4a, are a bridge between the immediate and the larger, less Xish, less controllable environment. Only the demographically and economically strong can cross this bridge with relative safety by providing the societal support that schools themselves need in order to successfully extend RLS efforts outwards into the larger community. No such extension can succeed before the basic familyneighbourhood-community support of Xish is in place. We will return to this point below. Stage 3 This stage pertains to the worksphere in general, but it is at its most powerful in connection with the higher, more influential worksphere which cannot be contained within Xish neighbourhood/community limits. This is a tremendously influential setting for RLS efforts and, indeed, one which pervades and colours all of social life, particularly in modern, secular contexts. With its necessary implications for social status and mobility the worksphere has become the most fully rationalised and cross-nationally cross-ethnically, and cross-linguistically connected domain of modern functioning. As such, it is a particularly difficult area for RLS to penetrate, influence and control. The growing predominance of multinational firms and the frequent rotation of their office incumbants, plus the fact that services or products are provided to an ethnolinguistically very heterogeneous clientele and that there is a constant growth and change of technology, products and services, are all features that prove to be linguistically problematic for all but the largest 'establishment languages'. The insuperable road-blocks which they frequently represent in so far as RLS is concerned must be realistically viewed in that perspective, although opportunities for RLS will necessarily vary depending on the overall degree of modernisation that the Xish and Yish economies have attained. 12 However, even when aspects of everyday higher-worksphere operations can be altered to accommodate Local RLS pressures, the link between the worksphere (higher or lower) and intergenerational language transmission is far from direct or obvious. Individual or group economic circumstances certainly influence such RLS concerns as the rate of selection of own-group and own-language marriage partners, the rate of childbearing, the neighbourhood of residence, the language of family-life, of child-socialisation and of the medium of instruction selected for one's children. These are all 'lower order' (i.e. more fundamental) concerns than the worksphere itself and must also be tackled directly, as indicated earlier, rather than only

< previous page

page_24

next page >

< previous page

page_25

next page > Page 25

indirectly via RLS efforts pertaining to the worksphere per se. Thus, while it is true that the worksphere must be 'captured', particularly for those seeking a maximum of cultural autonomy (rather than a diglossic H/L arrangement), it is also true that it is quite difficult to do so and that in doing so the forging of links to stage 6 must be constantly kept in mind. The formation of 'contract groups' that work together in Xish (in the lower worksphere) and of service centres or counters operating in Xish, or of Xish 'cells', 'branches', 'floors' or 'networks' in large firms, these are all possible tactics for introducing and maintaining Xish at work, but it is the indirectness of any positive link between work and stage 6 that must constantly be kept in mind and innovatively tackled and reinforced in a pro-RLS fashion. Stage 2 Stage 2 is concerned with lower governmental services i.e. those that have direct contact with the citizenry, including the local mass media. As with the worksphere, these must be viewed as more than merely factors in the 'creation of a climate' for RLS (many language movements pay far too much attention to such symbolic goals) and even as more than the creation of contexts for Xish use. The services and media entailed at this stage reach into the very neighbourhoods and, indeed, into the very homes that constitute the nuclei of RLS and of Xish life itself. The importance of Xishising these services and influences is beyond question, but only to the extent that their links to stage 6 are focused upon. Without such links the Xishisation of these services and influences constitute no more than a holding pattern; they buy time but they do not become self-priming RLS devices. The location and staffing of these services and the content and orientation of their programmes cannot themselves function as intergenerational transmission linkages. Such linkages must be there, maintained by far more direct family-, neighbourhood- and community-building processes, before governmental services and lower mass media can make transmissible contributions to RLS. It is doubtlessly harder to build Xish neighbourhoods and to assist Xish-speaking families than to broadcast in Xish on the television. The former, however, are immediate building-blocks of intergenerational transmission whereas the latter obviously is not. Stage 1 Finally, stage 1 may be reached, the stage at which cultural autonomy is recognised and implemented, even in the upper reaches of education, media and government operations, and particularly within the region (or regions) of Xish concentration. Once again, it must be clear that it is not some very general, unfocused, amorphous 'atmosphere effect' or process of osmosis that is of primary RLS concern in this connection, but, rather, what it is that stage 1 can do for stage 6 that really counts. The communications,

< previous page

page_25

next page >

< previous page

page_26

next page > Page 26

rewards and opportunity structures emanating from 'on high' influence family, neighbourhood and community processes via a very long, involved and impersonal chain of indirect influences and, therefore, their contributions to RLS are, ultimately, equally indirect and uncertain. While stage 1 can make a definite contribution to RLS, this contribution must be successfully translated into intergenerationally transmissible stage 6 processes and interactions before it will have more than bureaucracy-building and élite-building effects. Stages 3 to 1 are not only difficult for Xish to penetrate (due to their Yish locus of control and their Yish-dominated, even if heterogeneous clienteles), but they all link back to mother tongue acquisition only in a roundabout fashion and with considerable time lapse, if at all. They may help shape adult identity and language use, much more so than they help implant basic identity and mother tongue use in the young. Were this not so, much of the indigenous non-English mother tongue world, so dependent on English-speaking jobs, English-speaking media and English-speaking governmental functioning, would be of English mother tongue and of Anglo-American ethnicity by now, whereas this is obviously not the case in any part of that world. Problems of Focusing on 'Lower Order' Neighbourhood and Community Organisation The stagewise discussion, above, clearly implies that there are weaknesses to RLS efforts on the 'strong side', just as there are undeniable strengths to RLS efforts on the 'weak side'. Clearly, the 'strong side', with its stress on the institutions of modernity and on the structures of cultural autonomy (control of education, the worksphere, media, governmental services) is more than most RLS movements can realistically aspire to in the foreseeable future. Equally clear is the fact that even when such props for RLS are attained, they must still be translated into the lower order processes of stage 6 if a self-priming intergenerational transmissibility system is to be constituted and set into motion. The 'strong side' itself is not such a system, although it may adequately trickle down to that system once that system is in operation. Thus, the key to RLS is stage 6 and it is to some of the problems of stage 6-building that we now turn. Stage 6 may be viewed as an arena for 'collective action' and, therefore, as politically encumbered and, in accord with the theory of such action, as subdivisible into interest articulation, organisation, mobilisation and opportunity utilisation. Adopting the 'collective action' approach to step 6 would help alert us to the fact that beliefs, resources and actions do not always come together in unproblematic ways. Particularly when it is at its early stages, RLS efforts find it difficult to convince others that sociocultural change is needed and that established power and interest can be influenced and modified by minorities with clearly focused views and a stagewise

< previous page

page_26

next page >

< previous page

page_27

next page > Page 27

programme of goals and priorities. Another early difficulty is the over-reliance on voluntary and part-time leadership and on part-time leisuretime activists. One route that has been much used to temper such over-reliance is the involvement of individuals whose normal work-responsibilities include participation in or attention to social action, e.g. teachers, professors, social workers, lawyers, etc. However, these too can rarely offer full-time commitment and are usually establishment regulated and establishment dependent in many ways. Moreover, much of the literature on collective action and social action is too general in its orientation to be of direct relevance to RLS efforts or to early RLS efforts in particular. In the latter connection, the accumulated body of theory and data pertaining to organisational functioning, to neighbourhood organisation and to interest-group processes may be more helpful. Those who believe that all significant political action occurs at the national level need to be disabused of that perspective, particularly at early stages of RLS activity. Many of the most salient and explosive domestic political issues are fully recognisable as struggles between residents and local public authorities or interests. Furthermore, the question of how powerless persons can gain power in local affairs is of the greatest importance to a real understanding of democratic politics. The fundamental task of the RLS neighbourhood organiser is to find RLS incentives that will induce self-interested local residents to support and become active on behalf of an RLS nursery, RLS day-care centre, RLS housing-cluster, RLS co-operative market, RLS employment centre, RLS recreational centre, RLS homework/ tutoring group, RLS work-transportation service, legal aid service, credit union, etc. Incentives are crucial in understanding the difficulties which RLS spokesmen and ideologists face in their efforts to attract and organise Xmen on behalf of RLS. The success of interest-group organisationand that is what RLS efforts aredepends on much more than ideological appeals and all such appeals must be heavily intermingled with concrete inducements. As modern Gesellschaft forces continue to expand, the incentives of RLS-sponsored neighbourhood care, companionship and assistance in managing one's work, health and family problems become ever more meaningful, particularly in poorer neighbourhoods. An RLS stress on human relationships in modern life, on local accountability, on neighbourhood collective responsibility for all who live there, on self-help activities focused upon priority concerns of the local residents, becomes ever more meaningful if both parents are in the workforce, if public education deteriorates, and if individual family means cannot provide for the old, the sick, the newborn or the young with no place to play or study. Voluntary organisations serving such needs are not merely wishy-washy expressions of piety and goodintentions, but vital experimenters and innovators on the way to social and cultural reform. They help people to find identity and purpose, self-realisation and fulfilment, the very things that RLS must stand for both ideolog-

< previous page

page_27

next page >

< previous page

page_28

next page > Page 28

ically and practically. Particularly in the current climate of conservative politics and budget cutting, more peopleoriented help and self-help becomes both a particularly important and a particularly effective context for introducing and fostering RLS. RLS neighbourhood building efforts must obviously relate local spiritual beliefs, family values, informal neighbouring and self-help notions to Xish and to the ideal of 'Xmen via Xish'. In response to Durkheim's dictum that 'it is impossible to artificially resuscitate a particularistic spirit which no longer has any function', RLS must stress the functionality of Xish in terms of satisfying the deeply cherished values and needs of all ordinary, rank-and-file Xmen. But, obviously, neighbourhoodism has a strong political component and potential and the energy and effectiveness of the networks generated by RLS neighbourhoodism will be tied to the fact that they satisfy needs, make demands, have goals, address purposes. RLS neighbourhoods must be battling neighbourhoods, struggling for social, cultural, economic, political and personal dignity. Action needs an image of community that local residents can identify with and that fosters commitment as a type of functional equivalence to kinship among non-kin. Among ordinary folk, Xish cannot be pursued in and of itself, for its own sake. It must be part of the warp and woof of social life and make a meaningful difference in the neighbourhood Gemeinschaft-life and in the Gemeinschaft-strivings of ordinary people. All the social movements try, in their own ways, to fulful needs and strivings. RLS efforts must be able to do so even better than others, because kinlike affect and mutual support are part of the basic promise of the ethnicity message and of the authenticity message and, therefore, are part of the distinctiveness of RLS efforts. Neighbourhoods built on such distinctiveness and affect can go on to build Xish schools (at stages 5, 4a and even neighbourhood controlled public schools of type 4b) and Xish worksites, and move towards other selected goals on the 'strong side'. 13 Sometimes the above advice, and the entire approach on which it is based, strikes RLS activists as contra-intuitive and as self-limiting. It is obviously harder to build Xish families, neighbourhoods and communities than to broadcast for a few hours a week (or even a day) in Xish. However, the former immediately provides a base for intergenerational continuity and a point of departure for stages that can come after it and can be supported by it, whereas the latter do not because they have no daily, intimate, socialisation foundation underlying them. At best, they can contribute to the 'spirit' necessary for such a foundation to be laid, but they do not lay it themselves. Thus, the approach being advanced here does not counsel ignoring the higher order domains, but, rather, counsels the necessity of a prior, firm childsocialisation base to which they can contribute and from which they themselves, in turn, can derive the political support which they require, given that they will always be outweighed by their Yish counterparts.

< previous page

page_28

next page >

< previous page

page_29

next page > Page 29

Problems of Focusing on RLS via 'Lower Order' Family Processes The affective and affiliative emphases that derive from the kinship-ethnicity dimension of RLS-efforts not only directly imply the neighbourhood community, as the fundamental link to the intergenerational transmission of Xish, but they also and equally directly imply the family, as the very building block of such transmission. It is in the family that social support and transactions with the community have traditionally been initiated and nurtured. It is also in the family that social commitments have traditionally been nurtured. Above all, it is in the family that a peculiar bond with language and language activities (conversation, games, stories, songs, proverbs and felicitous expressions, verbalised emotion, verbal ritual and verbal play) is fostered, shared and fashioned into personal and social identity. Unfortunately, the 'traditional' family has become harder and harder to find and to maintain. It has been eroded by the very same universalising macro-forces that erode small languages and caring neighbourhoods. Good RLS neighbourhood organisation, therefore, must include programmes designed to provide social support for families (particularly young families) through: the provision of home visits by RLS-oriented social workers; the organisation of parent groups for a variety of purposes, but for RLS-parenting as well; the organisation of drop-in centres for assistance with any and all family problems but for assistance with RLS as well. Such initially 'formal support systems' should aim at generating informal or internal support systems within the families they serve so that the latter will have less and less need for the formal supports. RLS family support efforts cannot be oblivious to the fact that all competent parenting (not to mention RLS-oriented patenting) must be grounded in the norms, values and behaviours of a particular culture. Nowadays, parents must often be taught not only parenting but also their own culture, and this inevitably becomes part of the task of RLS. Of course, all of this assumes that it is possible to break out of the impersonality and uniformity of Gesellschaft and to ameliorate the loss of Gemeinschaft without thereby creating more bureaucratic problems than one solves. RLSinspired neighbourhood and family services must constantly be founded on consultation and co-operation, on selfliquidation of formal structures as genuine participation increases, and, above all, on the realisation that it is unrealistic to expect reversals within a few years of the myriad of neighbourhood and family problems that have evolved over the past century or more. Naturally, all planning entails some unexpected negative side-effects as well. Nevertheless, these can be minimised if RLS-efforts are conducted in an experimental, consultative, self-correcting and self-liquidating mood. The Yish mainstream itself has failed at correcting the very neighbourhood and family problems that RLS faces if it is to succeed. Certainly RLS should

< previous page

page_29

next page >

< previous page

page_30

next page > Page 30

distance itself from the Orwellian vision of a war-ravaged world in which the functions of the family have been taken over by stultifying and impersonal child-rearing institutions. Quite the contrary! RLS should be at the forefront of returning neighbourhoods, schools and families to the values, norms and behaviours that have preferential and historical validity for them. Many of the problems of Yish society itself are due to the very fact that most larger societal systems will not work without important imput from the family. Although it is far from clear that RLS-efforts will be more successful than others that have attempted to infuence family patterns, its inevitably smaller scale and its greater focus on the young, the old, the sick, the poor and the neglected are its greatest safeguards and its basic strengths. The small successes that it can attain (small in the light of the enormity of the problems that it must tackle) will be beacons of promise for others who are willing to be involved in an ennobling struggle even though it is a struggle that they cannot fully win. 14 The basic dilemma of RLS efforts everywhere is that their success requires overcoming the very problems of modern life that the strongest societies and cultures have not been able to overcome. The basic strength of RLS efforts is that they can afford to take a less ponderous, more grass-roots approach to these problems and, thereby, seek to come to the attention of and become identified with those whose lives they aim to change. The School: The Bridge Between the 'Weak Side' and the 'Strong Side' Approaches to RLS For RLS success the school must be an integral part of the family-neighbourhood axis of child socialisation and identity-commitment formation. Schools cannot succeed, whether their goal be RLS or merely history or mathematics instruction, if the relationship between teachers, parents and students is such that they are estranged from each other and from the curriculum. 'Schools are the children of the community', it has been wisely said, but this adage is little more than novel verbiage, 'educationese', because the state of affairs that it describes has become an impossible dream, purely 'pie in the sky' as far as most education is concerned. On the one hand, mainstream parents and communities have little to say about what goes on in the schools that their children attend. On the other hand, mainstream schools themselves do not seem to comprehend the extent to which the school has been challenged, every bit as much as has the family, as the major force in providing children with the skills, attitudes and behaviours upon which success in modern life increasingly depends. Precisely because such a high proportion of families are either of the 'single parent' variety or of the 'both parents working' variety (only some 7% of American families currently consist of two parents, one of whom stays at home during the years in which the children are at school), student

< previous page

page_30

next page >

< previous page

page_31

next page > Page 31

participation in neighbourhood and community activities has become much more crucial for later success than ever before. Childcare/playgroup arrangements that involve contact with adult specialists in tutoring, computers, dance, drama, writing, library research, athletics, scouting and after-school jobs contribute to an amazing widening of perspective and learning experiences, as well as to a diversification of interests and interactive competence, that have great significance not only for academic success but for success in the larger society thereafter. The above out-of-school agencies and activities have substantially usurped the traditional role of the family as the major partner of the school, and, in addition, they are all neighbourhood/community related arenas of crucial language use, language views and language competencies. The type 5 school can attempt to be the institutional resource that corresponds to Xish in this connection. So can type 4a schools to some extent, although they probably require an auxilliary agency, such as the type 5 school, in order to divide up the day and the onerous responsibilities of constantly relating Xish to childlife, to local needs and resources, to changing demands of the worksphere, to the entertainment sphere and the knowledge sphere. The RLS school must no longer be concerned only with ethnically encumbered attitudes, knowledge, skills and beliefs, but also with relating Xish to the rapidly changing world in which the language must constantly scramble in order to find a place for itself, most particularly so if diglossic arrangements with the world at large are ideologically unacceptable. 15 Additional Concerns and Perspectives The foregoing remarks are not intended to make it seem that RLS is impossible or doomed to failure. They are intended to make it clear why it is so hard to succeed at it. Successful RLS implies remaking social reality and that is very hard for minorities to do. The social meaning of being a minority is that one is forced to spend almost all of one's resources on damage control, i.e. on merely staying alive within a reality that is not of one's own making and not even under one's own substantial ability to influence. If we add to these widely generalisable difficulties those of also engaging in and achieving acceptance of a modicum of corpus planning, at least in order to stay abreast of 'popular modernity', those of traditional second languages which cannot count on the usual home and neighbourhood process for their intergenerational transmissibility, those of purported national and official languages which cannot come close to fully controlling the 'high side' even within their own territorial borders, also the particular problems of 'expected shortgevity' that beset urban, immigrant minorities, then one may well conclude that the task is not only a formidable one but a hopeless one as well. Such a conclusion is belied by the success cases we have glanced at, Hebrew in Ottoman and PreMandate Palestine, Catalan

< previous page

page_31

next page >

< previous page

page_32

next page > Page 32

in Spain, and French in Quebec. None of these successes was pre-ordained or occurred without struggle and reverses and all of them qualified as 'successes', whether or not their leaders admitted it to the rank and file, well before cultural autonomy was attained. In addition, the status of Basque seems salvageable and that of Irish has stabilised, something which appeared quite impossible only some two generations ago, albeit at a much lower level of utilisation than can be considered optimal. Rebuilding society is very hard, even for those who are its masters. The eight stage model, with its stress on the foundational nature (not the exclusive nature) of stage 6, must not be 'fluffed off' by the conundrum that 'all stages are more fully interdependent than the model maintains'. Such a view provides no insight into intergenerational mother tongue transmission and how it differs from social influence processes more generally. Such a view disregards the very limited successes of countless RLS efforts (Frisian, Irish, Maori and American/Australian immigrational being foremost among them) that have been guided by it. Such an analysis provides no guide to action because it has no approach as to priorities and no evaluative prespective as to the differences between long-range and short-range 'success'. Such a comment merely promotes busywork in all directions, thereby foreclosing all prospects of success in connection with solving what is a most difficult problem at any rate. It is similarly not advisable to point to the myriad of 'additional factors' that can be appended to or derived from the eight stage RLS model. Untold learning, interactional, communicational, attitudinal and other psychosocial, intercultural and interlinguistic dimensions can be proliferated ad infinitum. But to do so is to ultimately obfuscate the intergenerational transmission issue rather than to clarify it, by endlessly delaying the realisation that RLS efforts must attempt to do a few crucial things well and early, rather than be delayed until academics can shed light on all possible interpersonal, intergroup and interlinguistic issues related to it. RLS entails sociocultural reforms on behalf of an ethnic ideal that are already so fargoing, even if they remain on the 'weak side', as to be extremely dubious of attainment. Nevertheless, such efforts go on and on, paying homage to human persistence in building the kind of life that is more consistent with deeply held societal values. Indeed, RLS is a peculiarly and admirably human endeavour, after all is said and done, an endeavour to rebuild life, to attain an ideal of cultural democracy and justice, to meet felt responsibilities vis-à-vis one's identity, to behaviourally implement the traditions to which one subscribes, to safeguard and activate perceived cultural imperatives and sanctities or, at least, to make supreme efforts on their behalf and to 'go down trying', if necessary. The methods and priorities utilised in this quest can often be improved upon, they can be rendered more rational with respect to priorities, more knowledgeable with respect to precedents the world over, more informed vis-à-vis the social science theories and findings that might bear upon them. This is what I have tried

< previous page

page_32

next page >

< previous page

page_33

next page > Page 33

to provide in my treatment in this paper. The sanctity of the task however, cannot be improved upon. Like all sanctities, it is an absolute for those who see it and hear it and savour it with inner commitment and faith. Notes 1. Prepared for the Fourth International Conference on Minority Languages, Ljouwert (Netherlands), 2024 June 1989, and constituting Chapter 10 of my Reversing Language Shift (in press). 2. For some of the earliest but still timely Western objections to the legitimacy claims of minority indigenous ethnic groups see Lord Acton (1862) [1907] and Engels (1866). The more general and enduring roadblocks to an appreciation of ethnicity in human behaviour are discussed in my Language, Ethnicity and Racism (1977) (reprinted: 1989). 3. On the rationality/irrationality and modernity/non-modernity of revitalisation efforts, in comparison to current cultural planning of our own day and age, see Barber (1941), Champagne (1983), Kehoe (1989), Linton (1943), Wallace (1956), Thornton (1986), and Traugott (1978). 4. On the dilemmas of forgotten values and those who pursue them, see Institute of Cultural Conservatism (1987) and Klein (1976). 5. The major work on cultural persistence is still that of Spicer (1980) and his pupils, e.g. Castile (1981) and Moore (1981). Although pioneering and stimulating to this day, this work requires updating and, above all, a more systematic approach to the theory and practice of cultural persistence. 6. For useful introductions to eighteenth and nineteenth century British and French theories of social change and, therefore, to social theory more generally, see Berman (1970), Janos (1986), and Ryan (1969). 7. Sperber (1976), and Borowski (1984) provide thoughtful reviews of the communes and other alternative life-styles of the 1960s and 1970s and of the various factors which led to the variance in their success and durability. 8. Ample evidence of recent Marxist and non-Marxist rethinking of ethnicity can be gleaned from Baumgarten (1982), Fishman (1983[1985]), Hall (1983), Hechter & Appelbaum (1982), Lipset (1985), McCarthy & Zahn (1973), Nielsen (1980), Olzak (1983), Ragin (1979), and Wax (1974). A rather full review of the literature and a critique of its minor innovations and major inadequacies can be found in Fishman op. cit., Olzak op. cit., and Yinger (1985). 9. See Bonacich et al. (1972), De Marchi & Boileau (1982), Lieberson (1963), Said (1977), Suhrke & Noble (1973), and Yinger (1985) for indications of the ubiquitous, relatively permanent and non-theatening nature of ethnicity in many modern contexts. 10. The importance of the initial and ongoing clarification of goals and beliefs for effective social action is discussed by Aidala (1984), Tilly (1978), and Wuthnow (1976). 11. High implicationality has proven to be attainable only in relatively few social behavioural domains (including a few sociolinguistic areas among them, see Rickford 1987; in press). Even there scalability may be more attractive because of its apparent conceptual elegance than because of any enhanced predictive or explanatory validity that it attains or provides. 12. Geertz clearly sketches the role of economic factors in fostering more pervasive cultural change (see, for example, his 1963 volume), but he is particularly stimulating for our own concerns in connection with his stress on gradualism and transitionalism with respect to most traditional/Gemeinschaft and modern/Gesellschaft characterisation and the large variety of combinations between modernist and tradition even when economic modernisation is ongoing. Corners and nooks of worksphere RLS may well present themselves even under conditions of ongoing worksphere modernisation under Yish impetus. 13. For examples of research and theory on the ability of minorities to influence and wrest concessions from majorities see Hirsch & Gutierrez (1972) and Moscovici et al. (1985). Collective action and social action are reviewed by Tilly (1978) and Crowell (1968). The

< previous page

page_33

next page >

< previous page

page_34

next page > Page 34

huge amount of literature on neighbourhood organisation can be sampled via works such as Arzac (1982), Burton (1986), Hallman (1977), O'Brien (1975), and Scott (1981). Burton's book constitutes a review and critique of the fruitful pioneering work in this area by Philip Abrams. For a compendium of the most recent work on organisation theory more generally, see Morgan (1989). 14. Methods, difficulties and doubts re planned efforts to strengthen modern family dynamics are provocatively reviewed in Burton (1986), Curran (1983), McCubbin et al. (1985), Moynihan (1986), and Ziegler and Weiss (1985). 15. For a fine discussion of the school's changed relationship to family and school see Heath and McLaughlin (1987). References Aidala, Angela A. (1984) Worldviews, ideologies and social experimentation: Clarification and replication of 'The Consciouness Reformation'. Journal for the Scientific Study of Religion 23, 4459. Albrecht, Johann and Gill-Chin, Lim (1986) A search of alternative planning theory: Use of critical theory. Journal of Architectural and Planning Research 3, 11731. Arzac, Adriana A. (1982) The development of community competence through a neighborhood organization. Dr. P.H. dissertation, University of Texas, Houston, School of Public Health. Barber, Bernard (1941) Acculturation and Messianic movements. American Sociological Review 6, 6639. Baumgarten, Murray (1982) City Scriptures. Cambridge: Harvard University Press. Berman, M. (1970) The Politics of Authenticity: Radical Individualism and the Emergence of Modern Society. New York, Atheneum. Bonacich, E. et al. (1972) A theory of ethnic antagonism: The split labor market. American Sociological Review 37, 5479. Borowski, Karol (1984) Attempting an Alternative Society. Norwood: Norwood Editions. Burton, Martin (1986) Neighbors; The Work of Philip Abrams. Cambridge: Cambridge University Press. Burke, Kenneth (1954) Permanence and Change (revised edition). Los Altos: Hermes. Castile, George P. (1981) Issues in the analysis of enduring cultural systems. In G.P. Castile and Gilbert Kusher (eds) Persistent Peoples: Cultural Enclaves in Perspective (pp. xvxxii). Tucson: University of Arizona Press. Champagne, Duane (1983) Social structure, revitalization movements and state building: Social change in four native American societies. American Sociological Review 48, 75463. Crowell, George (1968) Society Against Itself. Philadelphia: Westminster Press. Curran, D. (1983) Traits of Healthy Family. Minneapolis: Winston. De Marchi, Bruna and Boileau, Anna Maria (eds) (1982) Boundaries and Minorities in Western Europe. Milan: Franco Angeli. Dolitsky, A. and Kuzmina, L. (1986) Cultural change vs. persistence: A case from Old Believer settlements. Arctic 39, 22331. Fishman, Joshua A. (1983) Epilogue: The rise and fall of the ethnic revival in the United States. Journal of Intercultural Studies 4(3), 546. Reprinted in Fishman J.A. et al. (1985) Rise and Fall of the Ethnic Revival. Berlin: Mouton de Gruyter. Geertz, Clifford (1963) Peddlers and Princes; Social Change and Economic Modernization in Two Indonesian Towns. Chicago: University of Chicago Press. Hall, Thomas, D. (1983) Peripheries, regions of refuge and non-state societies: Toward a theory of reactive social change. Social Quarterly 64, 58297. Hallman, Howard W. (1977) The Organization and Operation of Neighborhood Councils; A Practical Guide. New York: Praeger. Heath, Shirley B. and McLaughlin, Milbrey Wallin (1987) A child resource policy: Moving beyond dependence on school and family. Phi Delta Kappan 68, 576580.

< previous page

page_34

next page >

< previous page

page_35

next page > Page 35

Hechter, Michael and Appelbaum, Malka (1982) A theory of ethnic collective action. International Migration Review 16, 41234. Hirsch, Herbert and Gutierrez, Armando (1972) Learning to be Militant; Ethnic Identity and the Development of Political Militance in a Chicano Community. San Francisco: R. & E. Research Associates. Institute for Cultural Conservatism (1987) Cultural Conservatism: Towards a New Agenda. Lanham: University Press of America. Janos, Andrew C. (1986) Politics and Paradigms: Changing Theories of Change in Social Science. Stanford: Stanford University Press. Jenkins, J. Craig (1983) Resource mobilization theory and the study of social movements. Annual Review of Sociology 9, 52753. Juliani, R.N. (1982) Ethnicity: Myth, social reality and ideology. Contemporary Sociology 11, 36870. Kehoe, Alice Beck (1989) Ghost Dance Religion: Ethnohistory and Revitalization. Fort Worth: Holt, Rinehart & Winston. Kent, Susan (1983) The differential acceptance of cultural change: An archaeological test-case. Historical Archaeology 17, 5663. Klein, Donald (1976) Dynamics of resistance to change: The defenders. In Bennis, Warren et al. (eds) The Planning of Change (3rd edn) (pp. 11726). New York: Holt, Rinehart & Winston. Lieberson, Stanley (1963) Ethnic Patterns in American Cities. New York: Free Press. Linton, Ralph (1943) Nativistic movements. American Anthropologist 45, 23140. Lipset, Seymour M. (1985) Consensus and Conflict: Essays in Political Sociology. New Brunswick: Transaction Books. McCarthy, J. and Zahn, M.N. (1973) The Trend of Social Movements. Morristown: General Learning. McCubbin, Hamilton I. et al. (1985) Family dynamics: Strengthening families through action-research. In Robert N. Rapoport (ed.) Children, Youth and Families: The Action Research Relationship (pp. 12665). Cambridge: Cambridge University Press. Mindick, Burton (1986) Social Engineering in Family Matters. New York: Praeger. Moore, Janet R. (1981) Persistence with change: A property of sociocultural dynamics. In G.P. Castile and Gilbert Kusher (eds) Persistent Peoples: Cultural Enclaves in Perspective (pp. 22842). Tucson: University of Arizona Press. Morgan, Gareth (1986) Images of Organization. Newbury Park: Sage. (1989) Creative Organization Theory. Newbury Park: Sage. Moscovici, Serge, et al. (eds) (1985) Perspectives on Minority Influence. Cambridge: Cambridge University Press. Moynihan, Daniel P. (1986) Family and Nation. New York: Harcourt Brace, Jovanovich. Nielsen, F. (1980) The Flemish movement in Belgium after World War II. American Sociological Review 45, 7694. O'Brien, David J. (1975) Neighborhood Organization and Interest-Group Processes. Princeton: Princeton University Press. Olzark, Susan (1983) Contemporary ethnicity mobilization. Annual Review of Sociology 9, 35574. Ragin, C.C. (1979) Ethnic political mobilization. American Sociological Review 44, 61934. Rao, M.S. Changing moral values in the context of social-cultural movements. In Adrian C. Maywe, (ed.) Culture and Morality: Essays in Honor of Christopher von Furer-Haimendorf pp. 191208. Delhi: Oxford University Press. Rickford, John R. (1987) The haves and have nots; sociolinguistic surveys and the assessment of speaker competence. Language in Society 16, 14978. (in press). Implicational scaling and critical age limits in models of linguistic variation, acquisition and change. In C.A. Ferguson and T. Huebner (eds) Second Language Acquisition and Linguistic Theory. Amsterdam: John Benjamins. Ryan, Bryce (1969) Social and Cultural Change. New York: Ronald. Said, A.A. (ed.) (1977) Ethnicity and United States Foreign Policy. New York: Praeger.

< previous page

page_35

next page >

< previous page

page_36

next page > Page 36

Sathyamurthy, T.V. (1983) Nationalism in the Contemporary World. London: Pinter. Scott, David (1981) ''Don't Mourn for Me, Organize ..." The Social and Political Uses of Voluntary Organization. Sydney: Allen and Unwin. Smith, Elsie and Clement, B.G. (1981) A union of school, community and family. Urban Education 16, 24760. Sperber, Mae T. (1976) Search for Utopia. Middleboro: Country Press. Spicer, Edward H. (1980) The Yaquis: A Cultural History. Tucson: Arizona University Press. Suhrke, R.P. and Noble, L.G. (eds) (1973) Ethnic Conflict in International Relations. New York: Praeger. Thornton, Russell (1986) We Shall Live Again; The 1870 and 1890 Ghost Dance Movements as Demographic Revitalization. Cambridge: Cambridge University Press. Tilly, Charles (1978) From Mobilization to Revolution. Reading: Addison-Wesley. Traugott, Mark (1978) Reconsidering social movements. Social Problems 26, 3849. Wallace, Anthony (1956) Revitalization movements. American Anthropologist 58, 26481. Wax, Murray, L. (1974) Cultural pluralism, political power and ethnic studies. In Wilton S. Dillon, (ed) The Cultural Drama: Modern Identities and Social Ferment (pp. 10720). Washington, D.C.: Smithsonian Institution. Wolf, Eric R. (1982) Europe and the People Without History. Berkeley: University of California Press. Wuthnow, Robert (1976) The Conscious Reformation. Berkeley: University of California Press. Yinger, J. Milton (1985) Ethnicity. Annual Review of Sociology 11, 15180. Ziegler, Edward and Weiss, Heather (1985) Family support systems: An ecological approach to child development. In Robert N. Rapoport (ed.) Children, Youth and Families: The Action Research Relationship (pp. 166205). Cambridge: Cambridge University Press.

< previous page

page_36

next page >

< previous page

page_37

next page > Page 37

Minority Language Group Status: A Theoretical Conspexus Howard Giles, Laura Leets and Nikolas Coupland Communication Studies, University of CaliforniaSanta Barbara, CA 93106, USA Abstract This paper is an attempt to articulate basic questions in this fragmented field in pursuit of a more interdisciplinary framework, one which recognises the important mediating roles of individual cognitions as a complement to sociostructural analyses. As an untapped resource, there is an everincreasing number of, admittedly diverse, theoretical models in the sociopsychological areas of ethnolinguistic differentiation, ethnic language attitudes, second language learning, intercultural accommodation, communication breakdown, ethnic values, beliefs about talk, and so forth which are relevant for a more rounded understanding of minority language situations. These will be integrated into a framework leading to a set of predictive propositions which, it is argued, will elaborate the psychosocial climate under which different degrees of group level maintenance on the one hand, and the processing of majority-minority encounters and the kinds of interactive strategies within them on the other, will prevail. Models are formulated which lead to the meshing and enriching of the sociological with the psychological which, at the very least, provide a coherent agenda for assessing, comparing and contrasting a wide variety of minority language processes at the individual and collective levels. Members of ... [language-related disciplines] ... are often deeply saddened to learn of mother-tongue loss and cultural assimilation on the part of small and powerless ethnolinguistic entities. Indeed, in deeply unconscious and prescientific ways, convictions such as these are among the very ones that brought many of us to linguistics, to anthropology, to bilingual education, and to a variety of ethnic studies. (Fishman, 1982: 7) This short paper has an ambitious remit. To begin, we would like to assess the value of a recent critique of the field of minority languages, then move on to consider an existing valued model which overviews it. We wish to call attention to various lacunae in this model to the extent that important cognitive and interactional considerations are omitted which may serve very

< previous page

page_37

next page >

< previous page

page_38

next page > Page 38

important functions in language survival and non-survival. Hence, we would like to introduce the flavour of a very diverse set of research efforts and theories in the social psychology of language and intergroup relations on the one hand, and intercultural communication studies on the other. Our assumption is that the formal recognition of these broad literatures may well be useful in devising a more comprehensive, interdisciplinary framework. These literatures tackle a plethora of different issues and constitute for the most part a set of independent areas of enquiry. Often, they pay no more than lip service to each other in passing. Here, we shall at least attempt to organise these areas into an integrative framework which has some predictive value. Finally, we shall then address the formal modelling of sociological and cognitive forces, ending up (conventionally enough, but we think necessarily here) with an agenda for the future. Now whether or not we endorse the debatable view that Proceedings are an adequate reflection of the state of an art, Williams' (1988) scathing critique of the Proceedings of the Galway Conference demands consideration. His opening salvo was: 'These volumes all too clearly indicate that there is something drastically wrong with the study of minority languages' (p. 169). This is a disturbing overview, particularly by those of us who feel those Proceedings were indeed a good barometer of current work, but particularly since it is in some ways correct. We certainly do not share Williams' pessimism or despair, and we would question whether the blunt and dismissive manner of his appraisals is ultimately the most productive way to confront these issues individually let alone collectively. Yet we do feel that confront them we must. One of Williams' main points was that the field has no integrative theme. This is, we think, an entirely valid criticism, though we are not alone in this admissionit was certainly anticipated by the organisers of the 4th International Conference on Minority Languages in setting up their very conference theme. Many of us are indeed describing, often historically and dynamically, different minority language situationshere and there (Aikio, 1983; Gillies, 1987; Neville, 1986; Shield, 1983). Despite valiant attempts by some at cross-societal comparisons and sterling endeavours at transcending national boundaries in educational matters (Hansen, 1986) and linguicism (Skutnabb-Kangas, 1988), we are probably too locked into local peculiarities for our own theoretical, and perhaps even practical, good if this is to be our mainstay diet for the foreseeable future. That said, this phase of investigation has been a natural one and essential in providing us with fundamental data to work on; indeed, obviously there is still more yet to be accomplished in this vein. But now we need a consensually agreed set of ground rules which establish the parameters of our field so that we can all address our local concerns within the context of the larger-scale enterprise. Some of these integrative themes are plain for all to see. It seems that we are all oriented towards at least two clusters of issues. First, the explanatory paradigm: what conditions give rise to language maintenance, survival,

< previous page

page_38

next page >

< previous page

page_39

next page > Page 39

preservation, revival, and death, when and why? Second, the interventionist paradigm: given emotional identification and often pragmatic involvement with minority language survival in the context of widespread linguicism, what are the most appropriate means (educational, institutional support, conflictual) of protecting which minority language rights, and how? As Fishman (1982: 5) has noted, these two paradigms are often intertwined, the second predisposing many of us to work in the field and, defensibly, conditioning our moral positions on 'little peoples' and (hence) 'little languages'. What relevant theory is available to model these integrating themes? One which is fairly prototypical, yet more inclusive and parsimonious than others is de Vries' (1983) model (see Figure 1), which neatly illustrates the direct and indirect routes demographic factors can have on language survival. Language planners then would be drawn towards particular kinds of sociostructural and political-legal enabling forces as a means of facilitating language maintenance. First, from a comprehensive, interdisciplinary perspective, we would have to note a few caveats. First, the model beams into 'collective survival' as the ultimate dependent variable. It seemsand in fact this is acknowledged in other aspects of de Vries' (1987) conceptual workthat we also need to focus explicitly upon related phenomena including 'promotion', 'revitalisation', 'reintroduction', 'conservation', 'maintenance', 'revival', 'expansion', and 'preservation'. These are different constructs to the extent that they presuppose different survival baselines. Contrary to general use, we need to operationalise each of these terms more precisely (see Bourhis, 1984a). Certainly, the processes underscoring these outcomes (assuming we do specify them as unitary phenomena) are likely to be qualitatively different and therefore the interventions deemed appropriate will be necessarily various. Moreover, the focus on 'survival' needs to be balanced with schematic attention to 'non-survival' processes such as 'assimilation', 'atrophy', 'deterioration', 'decline', 'loss', 'contraction', 'death' and 'language suicide'.

Figure 1 de Vries' (1983) model.

< previous page

page_39

next page >

< previous page

page_40

next page > Page 40

Again, it is likely that processes enabling non-survival are not simply the converse of those facilitating survival (cf. Giles & Byrne, 1982). Second, minority group survival is conceived of here in an intergroup vacuum (see Tajfel, 1978). In other words, an analysis of the dominant group's ideologies, strategies, and language status is an important and dynamic set of contributing forces here, as discussed by Coupland & Thomas (1989) and Nelde (1986) amongst others. If language is an important aspect of dominant group identity, changes in the status of a minority language (be they growth or decline) are likely to have direct implications for dominant language identity in ways that social identity theory and ethnolinguistic identity theory (Giles & Johnson, 1987) can articulate, and in ways that will, transactively, then affect minority language-related activities. Third, the focus here is on group products and sociological antecedents. Some of us concerned with the development of the notion of 'vitality' (i.e. those status, demographic, and institutional factors working in favour or not of a group's survival; see Giles, Bourhis & Taylor, 1977) would argue that sometimes the potency of sociostructural and politicallegal factors may be mediated by individuals' subjective interpretations of languages and statuses in an intergroup context. In other words, the cognitive representation of these sociological forces is open to differential interpretation at different historical times and with regard to different ideologies even amongst members of the same group. There are other cognitive processes at work mediating group level maintenance and death (Giles & Johnson, 1987) and are what we shall shortly refer to as 'intergroup cognitions'. Fourth, the processes of influence as specified are abstracted out in an interactional vacuum. Yet the fate of languages and other social phenomena are in crucial respects worked out, defined and redefined in everyday discourse between members of the minority and majority language communities, as well as in intergroup encounters. An insufficient number of studies have been conducted (see, however, Gorter, 1986; Ytsma, 1988) into how the minority language is introduced as a topic and talked about in context, managed and reacted to by ingroup and outgroup collectives. Such interactional experiences are compelling day-to-day fodder for sustaining or inhibiting language use at the group level over the longer term, and could usefully be set alongside sociolinguistic paradigms which investigate code-switching and interlingual processes. In sum, and labelling the independent variables of de Vries' model as the 'sociological climate', the following elaborated model is a blueprint which recognises outcomes and processes beyond survival, captures the intergroup arena of minority language situations by attending also to dominant group dynamics, acknowledges that outcomes are not frozen endstates but rather part and parcel of ongoing changes, attends to the 'cognitive climate' as a potentially important mediator of minority language status (see also Foster, 1980; Bourhis, 1984b), and grounds the whole in interactional contexts.

< previous page

page_40

next page >

< previous page

page_41

next page > Page 41

Figure 2 An elaborated componential model of minority language group status. Now, as Williams rightly warns, there are potential epistemological dilemmas involved in blithely wedding sociological and sociopsychological (let alone textual and contextual) levels of analysis together. However, that does not mean we should abandon the exercise as hopeless and that interdisciplinary meshes are of necessity doomed to failure (see for example, Edwards, 1988; Haarmann, 1986; Haugen, 1972). The models unfolded during the course of this paper are an attempt to scratch the surface of productive engagement. Much effort has, rightly of course, been devoted to defining the sociological climates of language maintenance and shift (e.g. Akutagawa, 1987; Wande, 1983). Let us now see what is on tap to explore the sociopsychological, intergroup, and interactional climates. There is much work in sociolinguistics, social psychology, and communication studies which bears directly on the integrative themes we drew out as of concern to us. Not only is little of this explicitly directed towards minority language issues, the body of work itself is oriented towards different problems set at different levels of analysis. There have been attempts at rapprochement, but usually limited and no more than citational cross-references (see, however, Gudykunst, 1986). We shall attempt to introduce these diverse frameworks systematically (see Figure 3), our goal being to highlight general traditions of research, to point to key existing studies, and to draw out implications for future developments.

< previous page

page_41

next page >

< previous page

page_42

next page > Page 42

Commencing at the top of Figure 3 and working down, let us consider briefly some of the sociostructural perceptions possible for analysis (see Giles & Franklyn-Stokes, 1989). As mentioned already, different people have different fluctuating views about the vitality of their group and its language vis-à-vis relevant others. The stronger the perceived sociostructural support for the language amongst the group and beyond it, the more it may seem worthwhile investing energy in supporting it (see Figure 3). Hence, Ryan, Giles & Sebastian (1982) have formulated a model in which they speculatively place language attitude situations (including language minorities) in two-dimensional space of standardness-non-standardness and increasing-decreasing vitality. Allard & Landry (1986) have pointed to the fact that vitality can be usefully unpacked in terms of separate belief systems, namely the extent to which you believe the vitality situation should prevail (normative beliefs), your own beliefs about your ability to contribute to it (self beliefs), and your aspirations towards these ends (goal beliefs). Interestingly, Allard & Landry find the separate belief systems to be predictive of group linguistic assimilation and not the overall vitality measureat least as measured by them. Finally here, vitality is constituted by status/power and demographic factors (as well as institutional support). Sachdev & Bourhis (1990) overview a series of studies they have conducted under laboratory conditions looking at the interactive effects of status/power and numbers on intergroup discrimination. They find, for instance, that experimentally-constructed, dominant minorities show considerable ingroup bias whereas subordinate minorities show considerable outgroup favouritism; and whereas the perception of group numbers is important to attitude expression, power perceptions influence actual intergroup behaviour. Hence, the kinds of criteria we invoke concerning what constitutes 'minority language situations' ought to be sensitive to demographics on the one hand and power variables on the other, certainly until we have explored much further the important differential effects perceptions of these variables can have. Moving now to intergroup cognitions in Figure 3, much of the development of ethnolinguistic identity theory has been concerned with specifying the sociopsychological climates which promote ethnolinguistic differentiation. Most recently, Gallois, Franklyn-Stokes, Giles & Coupland (1988) have encapsulated our thinking by recourse to two continuous variables: dependency and solidarity. Dependency relates to the extent to which you are dependent on your ethnic ingroup for identity definitions and so refers to the number of social group options available to you. Solidarity relates to the degree of identification and affect subjectively associated with membership of one's ingroup. Another important distinction made here is in the way participants themselves construe interactions between majority and minority group speakers as either 'interpersonal' (i.e. dependent on the moods, temperaments, and personalities of those involved) or 'intergroup' (i.e. based entirely on the interactants dealing with each other as representatives of different social categories). Indeed, rather than consider interpersonal

< previous page

page_42

next page >

< previous page

page_43

next page > Page 43

Figure 3 An integrative model specifying predictive interrelationships.

< previous page

page_43

next page >

< previous page

page_44

next page > Page 44

and intergroup constructs as opposites along a single, bipolar continuum, a number of us feel it more productive to consider them as two separate continua where, for instance, one could define an interaction as low or high on both interpersonal and intergroup dimensions (Giles & Hewstone, 1982; Gudykunst & Ting-Toomey, 1990). There is a whole body of work (e.g. Hammer, 1987; 1989) which looks at individual characteristics that are predictive of effective intercultural communication (again see Figure 3). On the basis of this work we would expect that maximally-effective majority-minority relations, to the extent that these can be adequately modelled and assessed, would accrue when interactants from both sides were able to satisfy three criteria: to be able to cope effectively with psychological stress; to establish satisfying relationships quickly and understand the feelings of others; and to be able to communicate effectively with people from different backgrounds and deal adequately with miscommunications. The study of language attitudes has a long history (for review, see Giles, Hewstone, Ryan & Johnson, 1987) and is beginning to make its mark on the European scene in quite sophisticated ways, as seen in van Hout and Knops' (1988) important book, Language Attitudes in the Dutch Language Area. Until the early 1970s, much language attitude research was arguably descriptive and atheoretical but now we have a plethora of models on which to draw (see Giles & Coupland, 1990). While Lambert (1967) contended that language cues (as heard, say, in a matched-guise situation) triggered social categorisations which in turn unleashed social inferences, Berger & Bradac (1982) have suggested that other, more or less complex generative models could have explanatory power. They claim that an important goal of initial interaction is to reduce uncertainty about the other, and also how to respond appropriately. Hence, for these scholars, language cues are important to the extent they achieve this goal with judgments of cultural similarity as important mediators. Gudykunst & Ting-Toomey (1990) have speculated that the four Berger and Bradac models can be located differentially in the quadrants of the two-dimensional space of interpersonal and intergroup mentioned above. For example, under conditions of high inter-group and low interpersonal interactional perceptions, it is proposed that a speaker's language is a cue for in/outgroup membership which enables inferences concerning similarity between speaker and his or her listeners which, in turn, has an effect on the reduction of uncertainty; if similarity is high, the reduction is large (hence Language Group Similarity UCR, see Table 1 subsequently). Theories have also attended to other facets of processing too. Gallois, Callan & Johnstone (1984), for instance, have attempted to model the processes determining how linguistic cues (out of the many on offer, of course) are given evaluative significance in voice judgement studies. They claim that ethnicity is usually the prime one in intercultural settings with 'social distance' between listener-judge and target speaker (high-low) dictating when, and which other variables (e.g. gender,

< previous page

page_44

next page >

< previous page

page_45

next page > Page 45

context) assume salience. Other models point to the nature of the context as determining language attitudes with particular emphasis being placed on the nature of the scales used. That is, whether the dependent variables and or the context itself is group- versus personalistic-stressing (Giles & Ryan, 1982; see also Bradac, 1990), the extent to which the topic of conversation is involving on core identity domains or not (Giles & Johnson, 1986), and also the kind of stance your interlocutor, sociolinguistically, takes towards you can all have predictable effects on language attitudes expressed. For instance, Bradac (1990) contends that positive outgroup attitudes emerge if another converges towards you in a normatively-valued way, but one is more negatively-inclined if the person diverges in a non-valued direction. Finally, Ryan, Giles & Sebastian (1982) have suggested there are at least four kinds of language attitude profiles. For instance, 'Type A' is where the minority accedes favourable evaluations to majority group speakers and 'Type C' is where ingroup favouritism across evaluations abounds. In more recent work (Ryan, Hewstone & Giles, 1984), these Types have been conceptualised as emerging at different stages in the development of relations and conflict between groups in contact. Examining actual and perceived communication strategies in majority-minority encounters, accommodation theory has recently been elaborated sociolinguistically to take into account not only the productive performances of others in context, but also their interpretive competences (Gallois, Franklyn-Stokes, Giles & Coupland, 1988). This theory concerns itself with the conditions under which speakers may attune or contra-attune discoursally, paralinguistically and non-verbally to at least two features of their addressee: to the other's presumed facility to comprehend on the one hand, and their heard and anticipated performances on the other (Coupland, Coupland, Giles & Henwood, 1988). The theory also concerns itself with the evaluative, attributional and behavioural consequences of such attuning. Majority-minority interaction can of course involve more outright conflictual matters than just symbolic ethnolinguistic differentiations. Work by Ting-Toomey (1988) and others (e.g. Shockley-Zalabak & Morley, 1984) is then important for us as they have identified a whole range of verbal conflict stylesincluding direct versus avoidance, confrontational versus solutionoriented, and collaborative versus accommodative stylesfor which cultural groups have differential preferences. The language variables can of course be seen as ones which are constructive as well as reflective of the experiential climate of majority-minority encounters. Four very separate bodies of work relate to what we shall call 'interactional cognitions'. There is a literature which examines not only cross-national and cross-cultural value systems, but also highlights the role of the language of testing within this. Many studies have shown that when bilinguals use their second language (L2) they also shift their values in the direction of the L2 community. Bond (1983) has shown in the context of Hong Kong that this so-called 'cross-cultural accommodation' of values can

< previous page

page_45

next page >

< previous page

page_46

next page > Page 46

sometimes, in complete contrast, be supplanted by what he terms, 'ethnic affirmation'. More specifically, he shows that when Cantonese-English bilinguals are forced to complete a values questionnaire in English, they will become more Chinese in certain of their values than when they are administered the same questionnaire in Chinese. Thus, in terms of ethnic boundary theory (Giles, 1979), if a positive social identity cannot be maintained by linguistic means, then individuals who identify strongly with their group will resort to pertinent non-linguistic means at their disposal to rescue a positive group identity. It is curious that language and values research is a traditionally quite distinct area from language attitudes, though both can affect, and can be affected by, actual language strategies used in context. Another feature of intercultural encounters, according to Gudykunst (1988), is the amount of anxiety experienced, and doubtless other emotions too; the more anxiety, the less effective the encounter. Inevitably related are dimensions of 'satisfaction'. Hecht & Ribeau (1987) have studied the experiences of satisfaction-dissatisfaction in inter-ethnic encounters and pointed to an array of these such as 'negative stereotyping', 'acceptance', 'powerlessness', to name but a few (see also, Hecht, Ribeau & Alberts, 1989). Interestingly, their work highlights the fact that different groups are aware of and experience different levels and kinds of 'dissatisfaction'. Earlier, Taylor & Simard (1975) in Quebec and the Philippines demonstrated that even when inter-ethnic communication in formal and informal encounters was as objectively efficient as intragroup communication, the participants themselves in between-group encounters believed their interactions to have been far less efficient relative to those in within-group encounters. Finally here, models are beginning to emerge in the area of miscommunication (Coupland, Giles & Wiemann, in press). Constructs such as threat to identity, clear social categories, perceptions of social injustice, and sociolinguistic stereotyping are being linked to so-called 'intergroup communication breakdown' where fault is attributed to social group memberships (DubéSimard, 1983; Hewstone & Giles, 1986). The last component in Figure 3 is 'language learning inclinations'. Current theoretical work by Gardner (1985) and by Clément and Kruidenier (1985) looking at the sociopsychological factors responsible for L2 proficiency has highlighted two affective clusters, namely attitudinal/motivational, and situational anxiety. We, on the other hand, have preferred to see L2 learning as explicitly more of an intergroup process, claiming that intergroup cognitions such as dependency and solidarity in the context of particular kinds of sociostructural perceptions determine levels of motivation, the role of situational anxiety, and the kinds of proficiency outcomes possible (Giles & Byrne, 1982). Moreover, we can specify the sociopsychological climate which facilitates proficiency on the one hand, and lack of it on the other arguing also that the processes fostering a low motivation for learning a majority L2 are precisely those enabling a high motivation for learning one's own minority language where their proficiency is lacking (Garrett, Giles & Coupland, 1989).

< previous page

page_46

next page >

< previous page

page_47

next page > Page 47

We hope there is some perceptible logic to the sequencing of this top-down menu, as described in Figure 3. The principal merit of this overview is to encapsulate some of the key interrelationships between these separately studied research domains. Hence, the recurring schematic diamonds are intended to suggest that processes and phenomena are mutually influential and interdependent. A more precise theoretical sequencing of these elementswhich attends to their contextual and cultural dependenceshould be a priority for future conceptual developments. Though there is of course much fuzziness to the interrelationships, let us see if we can take this schema one step further, and specify a more predictive framework with respect to individuals who belong to a minority language group, whether they be functionally competent in the minority language or not. Table 1 represents a series of particular, though idealised, hypotheses deriving more or less directly from the intergroup cognitions of solidarity/ dependency and the brief overview above. Minority group members are often heterogeneous in their construals of their social identity and here we have identified two polar opposites: high dependency and solidarity contrasted with low dependency and solidarity. We are proposing that these cognitions lead to predictably contrastive ways of perceiving intercultural encounters that are mediated by general vitality and vitality belief systems which also are associated with differential motivations to become proficient in the majority and minority languages respectively. As Table 1 indicates, perceptions of minority and majority group speakers (as realised, for example, in a matched-guise task) would lead to certain kinds of generative mechanisms where the salience of other non-ethnic cues varies, and where the attitude profile is of a particular kind. Again, various situational factors are highlighted as mediating these attitude expressions and processes. The controversy concerning the relationships between language attitudes and interactional behaviours notwithstanding, we are suggesting that the aforementioned ingroup cognitions, situational dispositions, and intergroup perceptions will all be enabling factors for certain kinds of accommodative communications and conflict styles (see Table 1). Finally, interactional experiences are likely to promote the expression of certain kinds of values, and levels of anxiety, felt satisfaction and efficiency, as well as the potential for communication breakdown to occur or not. As Figure 3 attested, these experiences and outcomes have reciprocal influences such that sufficient instances where a high dependency and solidarity person is the recipient of positive-attributed attuning and experiences collaborative interactions with a majority group speaker and where satisfaction is high and anxiety low is likely ultimately to have predictable effects on their intergroup cognitions and the ways they are disposed towards defining intercultural encounters. To facilitate clarity, let us briefly describe the sequence of hypotheses outlined in one of the columns of Table 1. We argue that those minority group members who have low dependency on and low solidarity for their minority group will likely see interactions with a majority outgroup member as low in intergroup but high in interpersonal terms. They will be highly

< previous page

page_47

next page >

< previous page

page_48

next page > Page 48

Table 1 An integrative 'experiential-outcome' model.

< previous page

page_48

next page >

< previous page

page_49

next page > Page 49

motivationally-inclined to acquire the majority group language but far less motivated to acquire, or use extensively, their own minority group language. This will be the case especially when, after Allard & Landry (1986), they perceive the three vitality belief systems as being in favour of the majority language. When evaluating minority-majority language speakers (say, in a voice evaluation study) then vocal cues will trigger certain personality traits which will determine the degree of similarity felt between listeners and their speakers which will, in turn, influence the amount uncertainty reduced. In this situation, the evaluative salience of non-ethnic cues (e.g. gender) will be quite high (cf. Gallois et al., 1984). The actual attitude profile emerging in such a context would be very favourably inclined towards majority outgroup speakers (i.e. 'Type A' after Ryan, Giles & Sebastian, 1982) and especially on traits of status which are personalistic (e.g. intelligence, confidence; see Giles & Ryan, 1982) and when the context relates little to minority identity concerns. Such a pro-majority group predisposition would, however, be tempered if the outgroup speaker had diverged (and particularly in a normatively non-valued directionsee Bradac, 1990) from the minority listeners. In interactive situations, our low dependency/low solidarity minority speakers would attune their language, discourse and non-verbals towards majority group speakers (Gallois et al., 1988), and in the event of interpersonal conflict would adopt more indirect and accommodative styles of engagement; the particular communicative competences of the minority group speaker (see Hammer, 1987) would be particularly important in accessing effective attuning and conflict tactics. Finally, in such intergroup situations, these individuals would likely accommodate towards the values of the majority community, experience little anxiety, high satisfaction and communicational efficiency when talking with members of the latter, thereby creating the potential for communication breakdown to be low. Let us now return to our interdisciplinary brief and unpack the implications for a more comprehensive model of minority language dynamics. Rather than return to our Figure 2, let us reconsider de Vries' (1983) analysis of minority language survival. Therein, he discussed the sociological climate associated with non-survival in terms of a 'diffusion' model (see Figure 4), whereas the supposed sociological climate attending survival was said to follow an 'internal colonialism' model. Although we would acknowledge the possibilities of these distinct sociological climates operating in the direct ways specified (particularly if their dynamics were compellingly specified further), we would argue that they have their effects because of the manner in which they mould the sociopsychological, and within them interactional, climates. Put another way, group level survival/non-survival is, none the less, often effected through the minds and acts of individuals. We contend then that the kinds of intergroup cognitions introduced above are potent subjective states to the extent that diffusion forces are likely associated with low dependency/low solidarity and internal colonialism forces associated with high dependency/high solidarity. Sociopsychological climates should

< previous page

page_49

next page >

< previous page

page_50

next page > Page 50

Figure 4 An interdisciplinary predictive model of minority language survival/non-survival. therefore be preciser predictors of survival-non-survival than the larger-scale sociological climates which, at least arguably, gave rise to them. In addition, it is also possible that (as the question-marks in Figure 4 are meant to indicate) particular sociopsychological/interactional climateslow dependency and solidarity in the case of non-survival, and high dependency and solidarity in the case of survivalmay, at least in some contexts, be determined by other sociological forces thus far unspecified. Certainly, we need to acknowledge explicitly that the sociopsychological and interactional climates are influenced in part by perceptions of the dominant group's states, beliefs and behaviours, and also what other individuals within one's own minority ingroup appear to have adopted as sociolinguistic norms and strategies (again see Figure 4). Finally, let us repeat that we ought to be wary of dichotomising survival from non-survival, though also sociological from non-sociological. It is evident that there are other sociopsychological climates than the two completely opposing ones of Table 1; others would include low dependency/high solidarity and high dependency/low solidarity. And on the basis of the kinds of predictions we could make for these intergroup cognitive types (also along the lines of Table 1), we would speculate that they would be differentially inclined towards points along the survival-non-survival continuum in its simplistic form as represented in Figure 4. At the very least, we are opening up the inherent variability endemic in any minority language situation, and

< previous page

page_50

next page >

< previous page

page_51

next page > Page 51

in a way that should assist in handling the complexity apparent outside the realm of dichotomies. Caveats as we have identified in previous models could easily be levied against the frameworks introduced here. It is possible that attempts to weld approaches involving different levels of analysis will be hampered by revealing fundamentally-different ideological assumptions. The constructs in our menu require further conceptual refinement and operationalisation and are obviously not as static as we have perhaps conveyed. For instance, miscommunication is a multidimensional process and not simply an all-or-none phenomenon (Hewstone & Giles, 1986). We need to specify more clearly the sociolinguistics of attuning, and identify the discursive domains in which it operates with symbolic effect. There are doubtless richer ways of specifying situational dispositions beyond the intergroup and interpersonal, and richer ways of specifying intergroup cognitions beyond dependency and solidarity. We need to articulate better the sociological and sociopsychological climates underscoring the different outcomes of more or less survival, revival, protection, promotion, and so on. Whether the predictions formulated here will be more or less confirmedand doubtless many will not beis not really at issue given the larger goals of this paper. While the kind of analysis we have suggested here has seemingly more direct relevance to what we called, at the outset, the explanatory paradigmthe who, when and why of 'survival' in its qualitatively different formsit also has significance for the second, interventionist themehow we intervene effectively in the promotion of minority linguistic causes. Bourhis' (1984a) model of language planning is undoubtedly one of the few frameworks which acknowledges, even implicitly, sociopsychological climates in the pursuit of governmental and community goals. In reconstructing our analyses of sociological climates towards successful minority language planning in educational contexts and elsewhere, we also then need to tackle the sociopsychological and interactional environments in which different planners themselves operate, evaluating and implementing policy (see chapters in Bourhis, 1984b). However, we should not limit our consideration to language planning for language situations only. We need to plan our own academic and interventionist policies in a way that Williams (1988) may find constructive, perhaps by establishing a working party briefed to negotiate ground rules, integrative academic themes and objectives, and practical policies to be reported back to a definable, public assembly of us. Certainly the world climate is more amenable to minority language survival than ever before. The 'green movement' of pro-environmentalists has been, with varying degrees of commitment, embraced by political parties and ideologies in ways that would have been quite unthinkable just a few years ago. People are likely to be more susceptible now to the message of the ecological human value of maintaining valued cultural institutions and recycling languages along the same lines as they would condone preserving endangered species,

< previous page

page_51

next page >

< previous page

page_52

next page > Page 52

raw and precious materials and localities for the common global good. Indeed, there might be some merit in consulting with environmentalist-activist groups, as well as exploring related academic literatures, in order to glean lessons and garner useful strategies here. More and more one detects, perhaps optimistically, the hint of a new spirit of worldiness evolving that is cutting across narrow, parochial boundaries. It may well be that we need to harness such feelingsas an action-oriented academic bodyif many of our linguistic minorities are to survive in what Cummins (1988) and many others rightly refer to as 'the global village'. We say this because recent research is showing that positive attitudes to nuclear disarmament internationally are highly related to individuals' subscribing to a notion of 'worldliness' (Rigby, Metzer & Dietz, in press). Hence, these scholars are arguing for an about-turn in terms of attitude change interventions. They propose that in order to modify peoples' particularistic beliefs about say nuclear disarmament we may need to tackleat least as wellthe root-level social assumptions that likely underlie these if the changes advocated are ever to be realised. For our purposes then, if the ecological message of language survival is transmitted widelyfor example, we carry 'the world's little peoples and little languages ... on our sleeves in our country rather than merely in someone else's' (Fishman, 1982: 11)and we can engender widespread support for it as well as promoting sentiments of 'worldiness' (which incidentally can be and has been measured, see Sampson & Smith, 1957), we may have a better chance of supporting each of our own localised interests in the face of those authorities who have often to bow now to wider international pressures. That said, it would seem critical to work for minority language survival being endorsed more generally as an appropriate local facet of internationalism, and thereby counteracting any fallacy that 'internationalism requires international languages and no other'. Whatever, we may well need to rethink our explicit goals and the seemingly vacuous way our (limiting) dependent outcomes are schematised in theoretics (see Figure 1, and even Figure 4). References Aikio, M. (1983) The position and use of the same language: Historical, contemporary and future perspective. Journal of Multilingual and Multicultural Development 5, 27791. Akutagawa, M. (1987) A linguistic minority under the protection of its own ethnic state: a case study in an Irish Gaeltacht. In G. MacEoin, A. Ahlquist, and D. O'hAodha (eds) Third International Conference on Minority Languages: Celtic Papers. Clevedon: Multilingual Matters. Allard, R. and Landry, R. (1986) Subjective ethnolinguistic vitality viewed as a belief system. Journal of Multilingual and Multicultural Development 7, 112. Bradac, J. (1990) Language attitudes and impression formation. In H. Giles and W.P. Robinson (eds), Handbook of Language and Social Psychology. Chichester: Wiley. Berger, C. and Bradac, J. (1982) Language and Social Knowledge: Uncertainty in Interpersonal Relations. London: Edward Arnold. Bond, M. (1983) How language variation affects inter-cultural differentiation of values by Hong Kong bilinguals. Journal of Language and Social Psychology 2, 5766.

< previous page

page_52

next page >

< previous page

page_53

next page > Page 53

Bourhis, R. (1984a) Introduction: Language policies in multilingual settings. In R. Bourhis (ed.) Conflict and Language Planning in Quebec. Clevedon: Multilingual Matters. (ed.) (1984b) Conflict and Language Planning in Quebec. Clevedon: Multilingual Matters. Clément, R. and Kruidenier, B.G. (1985) Aptitude, attitude and motivation in second language proficiency: A test of Clement's model. Journal of Language and Social Psychology 4, 2137. Coupland, N., Coupland, J., Giles, H. and Henwood, K. (1988) Accommodating the elderly: Invoking and extending a theory. Language in Society 17, 141. Coupland, N., Giles, H. and Wiemann, J.M. (eds) (in press) Miscommunication: Problem Talk and Problem Contacts. Newbury Park, CA: Sage. Coupland, N. and Thomas A.R. (1989) Introduction: Sociolinguistic perspectives on English in Wales. In N. Coupland (ed.) English in Wales: Diversity, Conflict and Change. Clevedon: Multilingual Matters. Cummins, J. (1988) From the inner city to the global village: The microcomputer as a catalyst for collaborative learning and cultural interchange. Language Culture and Curriculum 1, 113. de Vries, J. (1983) Factors affecting the survival of linguistic minorities: A preliminary comparative analysis of data for western Europe. Journal of Multilingual and Multicultural Development 5, 20716. (1987) Problems of measurement in the study of linguistic minorities. Journal of Multilingual and Multicultural Development 8, 2331. Dubé-Simard, J. (1983) Genesis of social categorisation, threat to identity and perception of social injustice: Their role in intergroup communication. Journal of Language and Social Psychology 2, 183206. Edwards, J. (1988) Sociopolitical aspects of language maintenance and loss: Towards a typology of ethnic minority language maintenance and loss of ethnic minority languages. International Conference on Language Maintenance, Noordwijderhout, The Netherlands, AugustSeptember. Fishman, J.A. (1982) Whorfianism of the kind kind: Ethnolinguistic diversity as a worldwide societal asset (The Whorfian Hypothesis: Varieties of validation, confirmation, and disconfirmation II). Language In Society 11, 114. Foster, C (1980) Agenda for research. In C. Foster (ed.) Nations Without A State (pp. 20910). Gallois, C., Callan, V. and Johnstone, M. (1984) Personality judgements of Australian aborigine and white speakers: Ethnicity, sex and context. Journal of Language and Social Psychology 3, 3957. Gallois, C., Franklyn-Stokes, A., Giles, H. and Coupland, N. (1988) Communication accommodation in intercultural encounters. In Y. Kim and W. Gudykunst (eds) Theories in Intercultural Communication. Newbury Park, CA: Sage. Gardner, R.C. (1985) Social Psychology and Second Language Learning: The Role of Attitudes and Motivation. London: Edward Arnold. Garrett, P., Giles, H. and Coupland, N. (1989) The contexts of language learning: Extending the intergroup model of second language acquisition. In S. Ting-Toomey, and F. Korzenny (eds) Language, Communication, and Culture. Newbury Park, CA: Sage. Giles, H. (1979) Ethnicity markers in speech. In K.R. Scherer and H. Giles (eds) Social Markers in Speech. Cambridge: Cambridge University Press. Giles, H., Bourhis, R. and Taylor, D. (1977) Towards a theory of language in ethnic group relations. In H. Giles (ed.) Language, Ethnicity, and Intergroup Relations. London: Academic Press. Giles, H. and Byrne, J. (1982) An intergroup approach to second language acquisition. Journal of Multilingual and Multicultural Development 3, 1740. Giles, H. and Coupland, N. (1990) Language attitudes: Discursive, contextual and gerontological considerations. In A.G. Reynolds (ed.) Bilingualism, Multiculturalism, and Second Language Learning; A Tribute to Wallace E. Lambert. Hillsdale, NJ: Erlbaum. Giles, H. and Franklyn-Stokes, A. (1989) Communicator characteristics. In M.K. Asante and W.B. Gudykunst (eds) Handbook of International and Intercultural Communication. Newbury Park, CA: Sage.

< previous page

page_53

next page >

< previous page

page_54

next page > Page 54

Giles, H. and Hewstone, M. (1982) Cognitive structures, speech, and social situations. Language Sciences 4, 187219. Giles, H., Hewstone, M., Ryan, E.B. and Johnston, P. (1987) Research on language attitudes. In U. Ammons, N. Dittmar and K.J. Mattheier (eds) Sociolinguistics: An International Handbook of the Science of Language and Society: Vol. 1. Berlin: de Gruyter. Giles, H. and Johnson, P. (1986) Perceived threat, ethnic commitment, and interethnic language behavior. In Y. Kim (ed.) Current Studies In Interethnic Communication: 10th International and Intercultural Communication Annual. Beverly Hills: Sage. (1987) Ethnolinguistic identity theory: A social psychological approach to language maintenance. International Journal of the Sociology of Language 68, 6999. Giles, H. and Ryan, E. (1982) Prologomena for developing a social psychological theory of language attitudes. In E.B. Ryan and H. Giles (eds) Attitudes Towards Language Variation: Social and Applied Contexts. London: Edward Arnold. Gillies, W. (1987) Scottish gaelic the present situation. In G. MacEoin, A. Ahlquist and D. O'hAodha (eds), Third International Conference on Minority Languages: Celtic Papers. Clevedon: Multilingual Matters. Gorter, D. (1986) Aspects of language choice in the Frisian-Dutch bilingual context: Neutrality and asymmetry. Journal of Multilingual and Multicultural Development 8, 12132. Gudykunst, W. (1986) Towards a theory of intergroup communication. In W.B. Gudykunst (ed.) Intergroup Communication. London: Edward Arnold. (1988) Uncertainty and anxiety. In Y. Kim and W.B. Gudykunst (eds) Theories in Intercultural Communication. Newbury Park, CA: Sage. Gudykunst, W. and Ting-Toomey, S. (1990) Ethnic identity, language, and communication breakdowns. In H. Giles and W.P. Robinson (eds) Handbook of Language and Social Psychology. Chichester: Wiley. Haarmann, H. (1986) Language in Ethnicity: A View of Basic Ecological Relations. Berlin: Bouton de Gruyter. Hammer, M.R. (1987) Behavioural dimensions of intercultural effectiveness: A replication and extension. International Journal of Intercultural Relations 11, 6588. (1989) Intercultural communication competence. In M.K. Asante and W.B. Gudykunst (eds) Handbook of International and Intercultural Communication. Newbury Park, CA: Sage. Hansen, S. (1986) Mother-tongue teaching and identity: The case of Finland-Swedes. Journal of Multilingual and Multicultural Development 8, 7582. Haugen, E. (1972) The Ecology of Language. Stanford: Stanford University Press. Hecht, M. and Ribeau, S. (1987) Afro-American identity labels and communication effectiveness. Journal of Language and Social Psychology 6, 31926. Hecht, M., Ribeau, S. and Alberts, J.K. (1989) An Afro-American perspective on interethnic communication. Communication Monographs 56, 385410. Hewstone, M. and Giles, H. (1986) Social groups and social stereotypes in intergroup communication. In W. B. Gudykunst (ed.) Intergroup Communication. London: Edward Arnold. Lambert, W.E. (1967) The social psychology of bilingualism. Journal of Social Issues 23, 91109. Nelde, P. (1986) Language contact means language conflict. Journal of Multilingual and Multicultural Development 8, 3342. Neville, G. (1986) Minority language in contemporary France. Journal of Multilingual and Multicultural Development 8, 14758. Rigby, K., Metzer, J. and Dietz, B. (in press) Factors predisposing individuals to support nuclear disarmament: An international perspective. Journal of Peace Research. Ryan, E. and Giles, H., (eds) (1982) Attitudes Towards Language Variation: Social and Applied Contexts. London: Edward Arnold. Ryan, E., Giles, H. and Sebastian, R. (1982) An integrative perspective for the study of attitudes towards language variation. In E.B. Ryan and H. Giles (eds) Attitudes Towards Language Variation: Social and Applied Contexts. London: Edward Arnold. Ryan, E., Hewstone, M. and Giles, H. (1984) Language and intergroup attitudes. In J.R. Eiser (ed.) Attitudinal Judgment. New York: Springer-Verlag.

< previous page

page_54

next page >

< previous page

page_55

next page > Page 55

Sachdev, I. and Bourhis, R.Y. (1990) Bi/Multilinguality. In H. Giles and W.P. Robinson (eds) Handbook of Language and Social Psychology. Chichester: Wiley. Sampson, D.L. and Smith, H.P. (1957) A scale to assess world-minded attitudes. Journal of Social Psychology 45, 99106. Shield, L. (1983) Unified Cornishfiction of fact? An examination of the death and resurrection of the Cornish language. Journal of Multilingual and Multicultural Development 5, 32937. Shockley-Zalabak, P. and Morley, D. (1984) An exploratory study of relationships between preferences for conflict styles and communication apprehension. Journal of Language and Social Psychology 3, 21318. Skutnabb-Kangas, T. (1988) Multilingualism and education of minority children. In T. Skutnabb-Kangas and J. Cummins (eds), Minority Education. Clevedon: Multilingual Matters. Tajfel, H. (ed.) (1978) Differentiation Between Social Groups. London: Academic Press. Taylor, D. and Simard, L. (1975) Social interaction in a bilingual setting. Canadian Psychological Review 16, 24054. Ting-Toomey, S. (1988) Intercultural conflict styles: A face-negotiation theory. In Y. Kim, and W.B. Gudykunst (eds) Theories in Intercultural Communication. Newbury Park, CA: Sage. van Hout, R. and Knops, U. (eds) (1988) Language Attitudes in the Dutch Language Area. Dordrecht: Foris. Wande, E. (1983) Two Finnish Minorities in Sweden. Journal of Multilingual and Multicultural Development 5, 22541. Williams, G. (1988) Reviews of Third International Conference on Minority Languages: General Papers/Celtic Papers. In G. MacEoin, A. Ahlquist and D. O'hAodha (eds). Clevedon: Multilingual Matters. Language Culture and Curriculum 1, 16978. Ytsma, J. (1988) Bilingual classroom interaction in Friesland. In A. Holmen, E. Hansen, J. Gimbel and J.N. Jorgensen (eds) Bilingualism and the Individual. Clevedon: Multilingual Matters.

< previous page

page_55

next page >

< previous page

page_i

next page > Page i

Fourth International Conference on Minority Languages Vol. I: General Papers

< previous page

page_i

next page >

< previous page

page_57

next page > Page 57

On Coming to Our Census: A Layman's Guide to Demolinguistics 1 John de Vries Department of Sociology and Anthropology, Carleton University, Ottawa, Ontario K1S 5B6, Canada Abstract In this paper, I will provide a layout of a demographic approach to the study of linguistic minorities (now in somewhat esoteric circles known as demolinguistics). After the specification of a conceptual framework, I will discuss the main analytical issues, particular data needs and research possibilities. On the basis of empirical results from Canada and Finland, I will illustrate the potential of the demolinguistic approach for the study of second-language acquisition, language maintenance and shift pertaining to linguistic minorities. Finally, I will show how this approach can assist us in assessing the relative contributions of fertility, mortality, nuptiality, migration and language shift to the survival or decline of minority language communities. Introduction Intellectual developments in the past few decades have helped us to realise that 'minority languages', at the conceptual level, refer to two different aspects. We deal, on the one hand, with a social group, a minority, marked by language. On the other hand, the language used by members of such a group may be characterised by the fact that the group is in a minority position. The former aspect usually referred to by the term 'language community', tends to be the object of study for social scientists, while the latter one is more commonly studied by linguists. My paper will deal exclusively with the former aspect. Given that position, it should not be surprising that I refer to languages such as French and Swedish as the languages of minorities (which they are, in Canada and Finland respectively). So we are dealing with linguistic minorities. A few very basic points need to be made first. The use of the term 'minority' implies that, within the same society, there is at least one other group, the linguistic majority. The same society could contain additional minorities, or even two or more 'majorities',

< previous page

page_57

next page >

< previous page

page_58

next page > Page 58

or no majority at all (as an illustration of this last case, take as an example Cameroon, where the largest indigenous language community has been reported to be Fang, used by 19% of the population). All these configurations have in common that two or more language communities share the territory of a state, that these groups are distinguished from each other by a criterion related to language, and that at least one of them has been designated as a minority. We will skip over a discussion of who initiated this designation, or whether everyone agrees with it. The research tradition of the last twenty years has developed two approaches to the designation of groups as minorities. The simpler, strictly demographic one, is based on counts of group members. In this approach, the numerically smaller of two groups, or smallest (ones) among three or more groups, are labelled minorities. Conversely, the larger of two, or the largest of three or more, are labelled majorities. Based on this approach, the language communities served by the European Bureau for the Lesser Used Languages would virtually all be classified as minorities, but so would the Afrikaans and English language communities in the Republic of South Africa. The scond approach involves not only the partitioning of a population into language communities, but also the measurement of each group's access to resourcespolitical, economic, social, cultural. We tend to agree that a group which has a comparatively low level of access to these resources is in a minority position, regardless of its demographic position. With this approach, we do not find it difficult to view all non-whites in South Africa as minorities. In contrast, French speakers in Switzerlanda demographic minorityare generally not regarded as a minority in the latter sense. What unites these two intellectual approaches to the definition of 'minority' is the fact that they both require a form of counting. In the use of the demographic definition, such counting is (relatively) simple, in that we need 'only' to count the population and partition it (as we shall see, life is in fact much more complex than that). If we use the 'resources' approach, we will also need to determine the extent to which each group in a society has access to the various resources. This approach therefore also involves the demographic head count, but goes well beyond it. In many linguistically plural societies, this demographic 'count-plus-partition' has been, or is becoming, an essential prerequisite for the granting of language rights. While 'collective rights' (such as the right to use a minority language in Parliament, or to have official government documents published in several languages) probably do not require exact counts, the case is different when we deal with 'individual rights'. For example, the Canadian Charter of Human Rights and Freedoms gives, in Section 23, members of official language minorities (English in Quebec, French in the rest of the country) the right to have their children receive public education in the official language of their choice where numbers warrant. This type of right involves costs, to be borne by society, which are almost linearly related to

< previous page

page_58

next page >

< previous page

page_59

next page > Page 59

the number of persons entitled to such rights. Thus, minority groups themselves, as well as the state, will want to have access to exact counts, to substantiate the claims for individual rights, to plan rationally for the granting of such rights, or for the denial of claims. Of course, the agreements to collect such data and to accept their validity already assume a fair amount of co-operation between the state (generally with a monopoly over the collection of administrative and statistical data) and the minority groups. History abounds with cases where official data collection either never took place, or yielded unacceptable results as a consequence of mutual distrust, the deliberate use of political propaganda and/or widespread sabotage. However, many bilingual or multilingual states have developed longstanding methods of collecting official data on language (or language-related characteristics). In many cases, these counts go back a century or more. For example, Finland has collected data on main language in its parish records (which were combined into the equivalent of a population census) since 1880; Canadian censuses have contained data on ethnic origin (which, until 1951, was very strongly correlated with language) since 1871; the British census of 1891 was the first one to contain a question on Welsh (Pryce & Williams, 1988). Many other illustrations may be found. In contrast, some societies never introduced questions on language in their population census (e.g. The Netherlands), while others include language questions only for selected regions (e.g. France, Italy, the United Kingdom) and yet others have abolished language questions in censuses for political reasons (e.g. Belgium, 1960). Convenient inventories of language data in national censuses are provided periodically by the Population Division of the United Nations (see U.N. Demographic Yearbooks for 1956, 1963, 1972, 1973, 1979 and 1983). Basic Demolinguistics The rather broad and varied collection of census data on language was, initially, used for very little other than some basic administrative decisions. For example, Finnish language laws divided municipalities into linguistic categories as a function of the proportion of the population having Swedish as their main language, or the number of Swedish speakers in the municipality. Residents of the municipalities were (and are) then given local rights as a direct consequence of the municipality's classification (unilingual Finnish, unilingual Swedish, or bilingual). There were discussions and proposals for the academic use of these data in the early part of the twentieth century (e.g. Kloss, 1929; Winkler, 1923), but there is little of lasting scientific value. The major exception here is the work by the late Heinz Kloss, who dedicated much of his time during the 1970s to the building of research programmes at the International Center for Research on Bilingualism (CIRB) at Laval University in Quebec City.

< previous page

page_59

next page >

< previous page

page_60

next page > Page 60

The first harvest of scientifically valuable analyses of these data occurs shortly after World War II. The Canadian censuses of 1921, 1931 and 1941 yielded monographs on the 'nativity and origins' of Canadians. Especially the 1941 monograph (written in the early 1950s and released as 'restricted' in 1955) contained interesting analyses of the data on language and ethnic origin, with separate chapters on residential segregation and ethnic intermarriage. This is also the first thorough discussion of data quality. During this same decade, we also see the appearance of Gunnar Fougstedt's (1951) doctoral dissertation on the Swedish-speaking population of Finland between 1936 and 1945. This pioneering application of demography to the study of a linguistic minority regrettably never appeared in English and thus remained virtually unknown outside a small circle of Scandinavian demographers. It may be said that some of the research conducted in the 1950s and 1960s heralded the beginnings of demolinguistics as an organised approach to the study of linguistic minorities. The term itself originated among Quebec demographers, probably during the 1970s, as demolinguistique (see Lachapelle & Henripin, 1980). It was later translated as demolinguistics (see Lachapelle & Henripin 1982, a straight translation of the French text). Now the term is commonly understood by Canadian demographers working on language data. In an analogy to other hybrid terms (such a sociolinguistics, geolinguistics, psycholinguistics), the term demolinguistics indicates that we are dealing with the application of one discipline (in this case, demography) to a specific object (in this case, language). Thus, we take an approach usually applied to total populations (e.g. those of Norway, Friesland or London) and adapt it to what has been called a quasi-population. A quasi-population may be defined as a proper subset of an overall population, which is 'self-reproducing'. In other words, a quasi-population must have a structure by age and sex of such a nature that (in principle, anyway) the group could reproduce itself by the normal demographic processes. Thus, many meaningful social groups do not qualify as quasi-populations (e.g. PhD holders, the unemployed, teenagers), but language communities do. If we take this quasi-population approach to the study of language communities, we need to settle the following analytic issues. Definition of Membership A fundamental issue is, of course, that we need an unambiguous definition of group membership. In addition, the membership criterion should be such that any designation (of the form 'person A belongs to group X') is relatively stable, though not necessarily constant. That is, a designation based on the language used by a person answering the telephone would not be permanent enough, while one based on the majority language in the person's place of birth would be too permanent. A demolinguistic approach deals with the size, structure and geographic concentration or dispersion of language com-

< previous page

page_60

next page >

< previous page

page_61

next page > Page 61

munities. As such, we need to allocate every individual in a society to one and only one language community; in addition, we will want to collect some other basic data, such as age, sex, marital status, usual place of residence. In general, membership in language communities has been measured by means of three different criteria: Mother Tongue In general, questions on mother tongue have dealt with a person's first language (alternative question phrasings have used such references as 'language first learned in early childhood', etc.). There are many obvious advantages to this approach: (1) In principle, this provides a measure of an ascribed or inherited characteristic which a person cannot lose (similar to such characteristics as place of birth, year of birth, gender). (2) In addition, mother tongue is generally a good indicator of one's cultural orientation, since one's language origin is usually highly correlated with other cultural characteristics. (3) As an ascribed characteristic, it serves as a good basis for the measurement of various processes of change, such as language shift. (For such processes, we will need additional measures of language, which are not as permanent as 'mother tongue'). There are, of course, also some disadvantages, e.g.: (1) Not everyone has just one mother tongue. Many children in immigrant or minority language communities grow up with two or more languages. For such cases, we will have to apply somewhat arbitrary measures to deal with persons who report that they learned or spoke more than one language in early childhood. (2) Some people forget their mother tongue. Thus, based on such measures, declining or widely dispersed language communities may end up with many 'members' who are unable to communicate in the language. (3) Second-language learners would, by strict application of the definition, not be counted as members of a language community. This leads to potentially odd consequences, e.g. Joseph Conrad would not be included in the English language community, although he wrote all of his major works in English. NB The Canadian census has used a variant on this question, in which a respondent is asked to indicate the language first learned in childhood and still understood. This addition moves the mother tongue characteristic from being an ascribed one to an achieved one, with the obvious possibility of respondents forgetting (i.e. no longer understanding) the language first learned in early childhood. In that case, respondents are asked to report the next language learned (provided that it is still understood). In other words,

< previous page

page_61

next page >

< previous page

page_62

next page > Page 62

the Canadian census data on mother tongue provide information on the language(s) learned earliest and still understood. There is evidence that several linguistic minorities in Canada have indeed declined through this process of mother tongue loss. Main Language and Its Variants A more current measure of language behaviour, and thus of language group membership, is given by data on a respondent's main language. Many variants have been used, e.g. the language of inner thought, the language spoken most often, the language in which the respondent feels most comfortable, the language spoken (most often) in particular domains. Among the latter, 'home' appears to be the most commonly used one. This type of measure has the obvious advantage that it tells us something about current language use. If we are interested in measuring the size of language communities at present, such measures are among the more pertinent ones. They, of course, have their own shortcomings: (1) Whatever variant one uses, some respondents will report two or more languages. In some cases, this reflects 'domain segregation', e.g. situations in which someone uses different languages at home and at work. In other cases, the languages used are a function of language characteristics of the interlocutor (e.g. different family members). (2) In many cases, bilingual or multilingual individuals are constrained with respect to language choice due to the lack of other speakers of the language. Members of a linguistic diaspora are often affected by this factor. In such cases, the language spoken most often may not be the individual's main or preferred language; it may even be a language he or she speaks poorly. (3) When the question deals with 'home language', respondents are often constrained by the presence of one unilingual member of an otherwise multilingual household. In one-person households, such data reflect the language composition of visitors to the household. (4) Relative instability: many of these current measures suffer from high degrees of variability over time, as a respondent's social and/or economic circumstances change. Ability to Speak Designated Languages In many societies there is an interest, for policy reasons, in the number of persons able to speak specific languages. Such languages are usually official languages of the country (e.g. French and English in Canada; Swedish and Finnish in Finland) or regional languages (e.g. Welsh and Scots Gaelic in the. United Kingdom). Such measures normally do not deal with actual use of the language, but only with a respondents' ability to use the language.

< previous page

page_62

next page >

< previous page

page_63

next page > Page 63

Throughout the paper, we will refer to these three types of measurement. It is the case that they provide different kinds of indication. For example, data for Canada show that, during the period 19711986, the proportion of the population with French as mother tongue declined, and French as language spoken most often at home declined even faster; however, the proportion able to speak French (with or without being able to speak English also) increased. Given that different measures may tell different stories (or that the story may be a function of the measurement instrument), it is imperative that we know the exact question wording, and the exact circumstances under which such data were collected. Modes of Entry and Exit A typical demographic approach is to decompose population change (growth or decline) into the various modes of entry and exit, i.e. the processes by which a population gains new members and loses existing ones. When we apply this approach to the study of language communities as quasi-populations, we get the following: Entry fertility in-migration immigration language shift (in)

Exit mortality out-migration emigration language shift (out)

While most of the terms in this little array are not hard to understand, a few comments will be necessary. (1) Fertility and mortality represent the biological processes of birth and death. In the former case, we have a slight problem: while newborn infants do belong to households, neighbourhoods, regional and national populations, they do not at birth belong to a linguistic quasi-population. There are several ways to deal with the problem: (i) infants could be (arbitrarily in some cases) allocated to a language community on the basis of the language characteristics of parents, e.g. mothers main language or language spoken most often at home; (ii) one could delay the allocation of infants and very young children to a language community until they have reached a specified minimum age, say three years. In this latter approach, the delay also adjusts for differences in infant mortality. Such adjustments are not trivial in cases where language communities differ drastically in socioeconomic status. Consider, for example, North American Indians and Inuit, where infant mortality has been, and is, substantially higher than it is for the white populations in the same regions.

< previous page

page_63

next page >

< previous page

page_64

next page > Page 64

(2) In-migration and out-migration represent the social processes of residential mobility across 'internal' administrative boundaries, e.g. from one province to another. These processes have an impact on language communities if we define them in terms of regional residence as well as the language criterion. Such regional specifications are essential wherever language rights are allocated differently to different regions. For example, speakers of Frisian in the province of Friesland have reasonable access to public services in Frisian in the province, but not in the other eleven provinces in the country. Similar contrasts may be seen in Wales, Scotland, Belgium, Switzerland, Finland (and more!). (3) Immigration and emigration represent the social processes of residential mobility across 'external' administrative boundaries, i.e. from one country to another. In many instances, linguistic minorities have been created by these processes (e.g. the various Francophone communities in North America), while in other cases, minorities maintain themselves primarily by means of the arrival of new immigrants (e.g. most of the immigrant communities in Canada, Australia, the United States). (4) Language shift (yielding losses or gains, depending on the direction) represents the socio-cultural process of individuals changing from '... the habitual use of one language to that of another' (the definition first given by Weinreich, 1953). This is the process for which clearly defined membership criteria are important. In an analogy to the study of internal migration, where we use 'legal residence' or 'usual residence' as indicators of mobility or stability, we use a membership criterion as the basis of the analysis of language shift. As we shall see further on, this conceptual analogy has led to an analytical analogy with regards to measurements and estimation. These four paired processes provide 'logical closure'. That is, language communities can gain or lose members only by means of the four types mentioned above. I omit the case where minorities are created or abolished by means of changes in administrative boundaries (i.e. secession, occupation, annexation, and the like). The next step is to combine them all into what in demography has been called 'balancing equations', of the following kind: P2 = P1 + [Entries] - [Exits] + [Errors], where P2 is a population count at time t2, P1 is a population count at time t1, [Entries] and [Exits] are the effects of the processes, outlined above, between t1 and t2, [Errors] are the net effects of errors in data collection in all of the measurements taken, as well as errors introduced in the manipulation of the data (such as: coding, data entry, computer edits and imputations). Such balancing equations may be used on raw frequencies, but also on rates. That is, we may decompose an annual or decennial rate of change

< previous page

page_64

next page >

< previous page

page_65

next page > Page 65

into various components. This is, of course, analogous to better-known demographic equivalents, such as rates of natural increase (equal to the difference between crude birth rates and crude death rates). It should be evident that such balancing equations require consistent measures of group membership. Any changes in measurement procedure will yield apparent changes in group size. This should not be hard to appreciate if we just reorder the terms of the balancing equation to: P2 - P1 = [Entries] - [Exits] + [Errors]. Some perusing of this formula should make it clear that changing definitions, procedures for data collection or data manipulation will affect the [Errors] term in the equation, a clearly undesirable result. Such differences occurred in Finland when language group membership was defined through 'mother tongue' instead of 'main language' in 1977. Finnäs (1986: 51-55) discusses the effects of this change. A special survey, conducted in 1978 by the Finnish Statistical Office, showed that of all respondents with Swedish as mother tongue, 7% reported having Finnish as main language (Finnäs, 1986: 55; see also Allardt and Starck, 1981). It should not be difficult to see that these balancing equations can also be applied to distinct subgroups in the population. Thus, one can subdivide any quasi-population by one or more variables such as: gender, period of birth, place of birth. Given proper constraints (e.g. by limiting ourselves to appropriate age-groups), we can also use such variables as: marital status (subdividing the population into 'never married' and 'ever married'); highest level of education attained (assuming that negligible proportions of adults return to school or university after having entered the labour force); period of immigration. In many of these examples, some of the processes of entry become logically empty, e.g. if we deal with birth cohorts (groups of inividuals born during a specified period), no entries can occur between t1 and t2 due to births; if we deal with immigration cohorts (i.e. groups of persons who immigrated during a specified period), neither births nor immigration can serve as modes of entry. By using many balancing equations, we can often get a clear perspective on the specific nature of these social and cultural processes. A second, not yet fully exploited, feature of this demolinguistic approach is that we can adopt the entire 'tool kit' of demographic estimation procedures. Demographers often work with incomplete and/or faulty data. Since such data are in many cases needed for social planning purposes, much effort has been spent on the development of estimation methods and procedures for quantitative inference. Especially demographers working on non-industrial societies (where official data collection tends not to be well established) have access to a formidable 'bag of tricks' to make sense of very deficient data. It should not be surprising that demolinguistics, even in the most developed societies, often face comparable problems.

< previous page

page_65

next page >

< previous page

page_66

next page > Page 66

Data Needs From the preceding discussion, it should be clear that we have rather well-defined data needs. The most useful data source is a periodic count of the population, partitioned into linguistic quasi-populations by consistent criteria. Ideally, such counts should be obtained in a population census taken at regular intervals, say every five or ten years (such as has been done in, among others, Canada,' Finland, the United Kingdom, and Switzerland). While censuses at irregular intervals are not as desirable, we have a good supply of adjustment procedures to deal with these situations (as occur, for example, in the USSR with censuses in 1959, 1970 and 1979). As an alternative to population censuses, population registers may serve as useful data sources, provided that there are possibilities for the 'updating' of language characteristics (analogous to the requirement that changes of address be entered in the population register). The absence of periodic, or at least repeated, census data does not necessarily obviate the research possibilities in demolinguistics. Even one population census with good tabulations of language characteristics by age and sex will yield good analytical possibilities. Alternatively, one or more large surveys may provide a good starting point for analysis. The obvious problem with surveys is that we confound issues of statistical inference (from sample to population) with those of analytical inference (e.g. that of decomposing overall change into its logical components of natural increase, net migration and net language shift). Moreover, given that self-report data of this kind are never perfectly reliable, we would need at least three consistent measures if we wish to separate 'real changes' in behaviour from 'artificial changes' resulting from imperfect measurement procedures (see Heise, 1969, for a discussion of this problem). In all of these situations, it is important that we know exactly how group membership was determined. This is not to stipulate that there is only one 'perfect way' to do this, nor that some methods are inherently better than others. The main point is that it is generally preferable to have consistent (ideally: totally identical) measures, even if imperfect, than to have inconsistent measures (even if this involves 'improvements' in the data). To appreciate the consequences of varying membership criteria (or ambiguous, or entirely unreliable ones) on estimated group size, consider the claims shown in Table 1 for the sizes of selected European linguistic minorities. In no case is it clear how authors arrived at these estimates. Unfortunately, few if any raw data are available for secondary analysis. A recently produced study on Albanians in Southern Italy supplies another example of poor or deficient specification of group membership (De Bartolo, 1987). The author uses an earlier study (Rother, 1968) in which estimates are made, by community, of the percentage of the population speaking Albanian in 1966. We do not know by which criterion Rother decided on the percentages speaking Albanian. We do know that De Bartolo used these same percentages, of unknown quality, to estimate the number of Albanian

< previous page

page_66

next page >

< previous page

page_67

next page > Page 67

Table 1 Claims for the sizes of selected European linguistic minorities Minority Reference Alsatians: 1,300,000 (Stephens, 1976) 1,400,000 (Kremnitz, 1977) 1,200,000 (Haarmann, 1,550,000 1975) (Petrella, 1978) Basques 80,000 (Kremnitz, 1977) (France): 90,000 (Stephens, 1976) 100,000 (Krejci, 1981) 130,000 (Haarmann, 1975) Bretons: 600,000 (Krejci, 1981) 1,100,000 (Haarmann, 3,200,000 1975) (Alcock, 1979) Occitans: 950,000 (Salvi, 1973) 1,150,000 (Alcock, 1979) 78 million (Kremnitz, 1977) 1014 (Krejci, 1981) million (Haarmann, 1020 1975) million (from de Vries, 1985b: 214215) speakers in 1981. Worst of all, we do not have any way of validating any of these estimates. In addition to the need for good (i.e. reliable and consistent) 'stock data', there is a need for good 'flow data'. Ideally, we would like to have data on: (i) births by mother tongue/main language of the mother (it would be nice to have data on the father as well); (ii) deaths by language of the deceased; (iii) migrants classified by language characteristics; (iv) changes in language use. Information on births and deaths by language is rare, but not entirely nonexistent. The Canadian province of Quebec has collected vital statistics (births, deaths, marriages) classified by mother tongue since 1975 (see Bourbeau, 1988: 2628 for a discussion). In a similar fashion, some of the vital statistics of Finland have been classified by language since 1975 (see Finnäs, 1986: 712, 24, 41). Data on marriages classified by the language of bride and groom have been collected since 1951 (Finnäs, 1986: 149). Unfortunately, these illustrations seem to represent the only cases where vital statistics are available. In other cases, we need to resort to estimation techniques.

< previous page

page_67

next page >

< previous page

page_68

next page > Page 68

This emphasis on census data may appear somewhat esoteric and unrealistic to those of you who work in countries which either never collected census data on language or stopped doing so. I should point out that the supply of basic data has been relatively good, and is improving. As I mentioned, the Canadian censuses have contained a wealth of data for over a century, with especially rich opportunities for research in the censuses of 1971, 1981 and 1986. Within Europe, the census takers in the United Kingdom have asked a question on the Welsh language in Wales since 1891, and on Scots Gaelic in Scotland since 1881. The census of Ireland has contained a language question since 1881. Finland has had language questions in its population censuses and their predecessors since 1880. Many Eastern European censuses have contained questions on language (e.g. the USSR, Yugoslavia). France and Italy have collected census data on some of their minorities (e.g. Alsacians in France, German speakers in Italy). The newest arrival is Spain, where data on Catalan, Basque and Galician were collected in the census of 1986. There are, of course, exceptions. Noteworthy in this category is Belgium, where Parliament decided in 1960 to delete the language question from the census. I gather that collecting any data on the use of Dutch and French in Belgium is likely to invoke official sanctions. European societies also appear to be ill-equipped to handle the collection of data on the 'new immigrants', be they classified as guestworkers, refugees or immigrants. These groups are faced with tremendous problems of integration; most industrial and post-industrial societies have had to grapple with the issues of health care, education, unemployment and underemployment, and so forth. Most of those issues are associated with an individual's ability to communicate in the language of the host society. It would appear that rational strategies for planning and implementing social policies in these areas require some fundamental knowledge about the ability of various immigrant groups to use the majority language. Yet, no European census has, to my knowledge, attempted to gather data on these groups (in contrast to the 'traditional' immigration countriesCanada, Australia, the United States). Research Possibilities So far I have laid out a general framework for demolinguistic analysis of linguistic minorities. In doing so, I have referred at many points to the use of census data in a fairly general way. It is time to discuss some of the research possibilities of this approach in a somewhat more specific fashion. Fertility There are many ways in which a consistent set of censuses, or even one census, can give us reliable estimates of a group's fertility. Depending on

< previous page

page_68

next page >

< previous page

page_69

next page > Page 69

the timing of the data, and the details of the age distribution, these estimates usually deal with the preceding five or ten years. Shryock & Siegel (1976: 297312) have an entire chapter dealing with fertility estimation on the basis of censuses and surveys. The most common approaches are: (1) The child-woman ratio, normally calculated as the ratio of children under five years of age to the number of women between the ages of 15 and 49. This ratio obviously provides an estimate of the average fertility of a group over the preceding five-year period. It has several known flaws, but does provide a convenient 'short-cut' measure when no better data are available. The measure only requires a cross-classification of the population by age-group and sex. When we apply this measure to the study of language-group fertility, we run into a problem. If we were to subdivide a population by, for example, mother tongue and calculated child-woman ratios for each mother tongue category, we would confound two processes in this final measure: differential fertility (i.e. different levels of fertility for different mother tongue categories) and net intergenerational language shift (i.e. children having a different mother tongue from that of their mother's). It should be clear that the child-woman ratio is only a valid indicator of a group's fertility if intergenerational language shift is negligible. As an illustration: Lieberson (1970: 5155) uses data on agestructure by language for Canada from 1931 to 1961 to make statements about fertility differences between francophone and anglophone women. Such statements are not tenable, given the rather pronounced shift from French to English among the Francophone minorities outside Quebec. (2) Measures of 'children ever born'. A question asked in many censuses deals with the number of children ever born to women (the question is usually restricted to women above a certain age, and in many instances to ever-married women). These questions are restricted to live births and, of course, exclude adopted children, foster children and stepchildren. Such data have been used by Lachapelle & Henripin (1982) to estimate fertility differences of Canadian women by age-group, residence, and home language for 1971, and by mother tongue for 1976. It should be noted that this type of measure does not confound differential fertility with language shift, since it uses a direct relation between mothers and their own children. Mortality Without vital statistics data in which deaths are classified by language characteristics of the deceased, the estimation of mortality by language group is a very difficult process. Here is the area where the estimation techniques developed by demographers for work in 'statistically underdeveloped areas' may prove to be suitable. A common problem in such societies is the

< previous page

page_69

next page >

< previous page

page_70

next page > Page 70

estimation of mortality without the availability of reliable registration data on deaths. One approach is the use of theoretical population models, usually developed on the basis of rather stable age-specific rates of fertility and mortality. As an illustration of this approach, I estimated mortality patterns for Swedish speakers in Finland during the period 19651970 by using the data on age-structure in the 1970 census. By selecting the appropriate table from the system of United Nations Model Life Tables, I found a close match between the age structure for Swedish-speaking Finns in 1970 and a Model population with a Gross Reproduction Rate of 1.0 and an expectation of life at birth of 70.2 (de Vries, unpublished manuscript). For comparison, consider the data in Table 2. The index of dissimilarity between these two distributions is a puny 1.1. The selected Model population had a crude birth rate of 13.3 per 1,000, a crude death rate of 15.1 per 1,000 and a rate of natural increase of -1.8 per 1,000 per year. It is interesting to note that Finnäs (1986: 47), on the basis of much more detailed data and much more refined analyses, gives an expectation of life at birth of 70.9 for Swedish-speaking males, and of 78.2 for females, for the period 19761980. In later work (1988: 911), dealing with the period 19811985, he gives crude birth rates of 10.9, crude death rates of 12.7 and a rate of natural increase of -1.8 for Swedish-speaking Finns. He also reports a rate of natural increase of -1.7 for the period 19781980. Other approaches to the estimation of mortality are based on having access to at least two consecutive censuses, preferably five or ten years apart. As I already mentioned, suitable tabulations allow us to follow birth cohorts from one census to the next. Shryock & Siegel (1976: 497507) suggest possible approaches to the estimation of mortality from data in several censuses, a single census or even a single survey. Analytically, these estimations of differential mortality run into intractable problems. While the estimation of mortality for national populations is, to some extent, confounded by international migration (we cannot distinguish between someone emigrating or dying, unless we have direct information), the estimation of differential mortality by language is, in addition, confounded by language shift. In other words, we cannot distinguish between a person dying and a person changing his habitual language use. Table 2 Percentage distribution by age Model Swedish Finns, 1970 014 19.5 19.1 1544 39.5 38.8 4559 19.1 20.0 60 and over 21.9 22.1

< previous page

page_70

next page >

< previous page

page_71

next page > Page 71

Lachapelle and Henripin (1980: 8998) work through a variety of estimates of mortality differences between francophones and anglophones in Canada. Their conclusion is that, in advanced industrial societies, mortality differences standardised by age and sex are so insignificant that they can safely be ignored. The losses to a group due to relatively higher age-specific mortality are trivial in comparison to the losses due to differential migration or net language shift. For example, their analyses suggest that Quebec's fertility would have to be only 0.4% higher than that of Ontario to compensate for Quebec's higher mortality (Lachapelle & Henripin, 1980: 95). It is likely that similar arguments can be advanced with regards to differential mortality between Swedish-speaking and Finnish-speaking Finns. Language Shift As we shall see later, the assertion that we can ignore the effects of differential age-specific mortality, at little risk, comes in handy when we deal with the analysis of language shift. The argument developed so far asserts that language shift may be regarded as a demolinguistic process, by means of which language communities gain or lose members. The analytical structure for the study of this process requires that we measure the language characteristics of individuals at least twice, at times t1 and t2. Weinreich's (1953) definition suggests that we attempt to measure 'habitual language use' at t1 and t2 and that different responses from the same individual indicate' language shift. Identical responses would, of course, indicate language maintenance. Alert observers will have noted that this analytical structure is analogous to the ones we use for the study of social and economic mobility and for the analysis of internal migration. In both of those cases, we want to relate the respondents' status of origin (be it social, occupational or residential) to her/his status of destination. Such analyses are based on broad sociological or demographic theories of status attainment, or models of occupational opportunities. While the theoretical linkages to the study of language shift are difficult to establish, the methodological ones are rather easily shown. Thus, much of the demolinguistic work on language shift has been based on the availability of two census questions for the same respondent: mother tongue and language spoken most often at home. The Canadian census introduced the latter question in 1971 and repeated it in 1981 and 1986, thus enabling researchers to do sustained quantitative work on the subject. Given the definitions, we can agree that 'mother tongue' may be taken as a reliable and valid indicator of an individual's habitual language use in early childhood, and that 'home language' is an acceptable indicator of current language use (recall possible problems of domain segregation, which would render the data less than perfectly valid). Then the cross-tabulation of mother tongue with home language for specific age-groups gives us a good description of language

< previous page

page_71

next page >

< previous page

page_72

next page > Page 72

shift. As an illustration of what such cross-tabulations may reveal, consider the data on language shift in Table 3, from French mother tongue to English home language, for Canada. Comparable analyses were carried out by Lachapelle & Henripin (1980: 125), and by Bourbeau (1988). In many cases, we do not have access to two language questions for the same individuals. Thus, we would have to approximate the origin and destination measures by other means. A powerful approach to the problem, based on the availability of repeated aggregate measures, is to use cohort analysis. By defining a cohort in terms of period of birth, say 19411946, we can locate its members as the group aged 2529 in 1971, aged 3539 in 1981 and so on. By using suitable crosstabulations of language characteristics by age in different censuses, we may be able to construct the cohort's intercensal 'history'. (Thus, the initial census would serve as a measure at t1, the later census as a measure at t2). We normally have to make several simplifying assumptions, for example, that minority and majority have the same age-specific mortality rates (which we can then calculate from an appropriate life table). On the basis of these assumptions and the available data, we can then estimate net language shift, by means of a 'residual analysis'. That is, we set up a balancing equation for each cohort, looking like this: P2 = P1 + [net migration] - [deaths] - [net language shift]. We estimate [net migration] and [deaths] from external data sources (and/ or simplifying assumptions) and then rearrange the equation to look like: [net language shift] = [net migration] - [deaths] + P1 - P2. Table 3 Percentage of French mother tongue group shifting to English, by age Age Canada, 1971 04 2.5 59 3.1 1014 3.3 1519 4.1 2024 6.6 2529 7.7 3039 8.3 4064 8.3 65+ 6.7 (1971 data from de Vries & Vallee, 1980: 122)

< previous page

page_72

next page >

< previous page

page_73

next page > Page 73

Usually, net language shift is then converted into a percentage of P1, the cohort's initial size. Note the analogy is a well known approach to the study of internal migration. The cohort survival approach has been used in a variety of settings, e.g.: (1) Anderson & Silver (1983) studied the language maintenance or loss among ethnic minorities in the Soviet Union. (2) Henripin (1974) studied the 'survival' of Canadian immigrant ethnic groups. In particular, his analysis suggests that long-established ethnic groups show a drift towards British ethnic origin. (3) Lallukka (1987) analysed the ethnic identification (through language maintenance) of the Mordvins in the USSR between 1959 and 1970. (4) I used the approach to estimate net language shift from Swedish to Finnish in 19501960 and in 19601970 (de Vries, 1974; 1977). I have concentrated in this paper on work done on data from Finland and from Canada. These appear to be the only two societies in which demographic approaches to the study of language communities have developed. The absence of such research approaches in other countries is not necessarily explained by the lack of suitable data. For example: Wales has had a long series of census data on the Welsh language. Ravenstein's (1879) address to the Royal Society did result in data collection, but not in any interest on the part of British demographers. The most sustained quantitative research tradition is maintained by geographers (e.g. Williams, 1982; Carter, 1980). As a consequence, we have a beautiful set of maps displaying the geographic concentrations of Welsh speakers (as well as many other interesting spatial displays), but no estimates of language shift, differential fertility or mortality. It may be that the Longitudinal Data Base, maintained at the University of London, will eventually yield some fruitful analyses. For the remainder of Western Europe, even the availability of census data on language appears to have produced little attraction for researchers. Although the censuses of Switzerland have provided information on the size of the major groups (German, French, Italian and Romansch) since 1880, no detailed demographic work has come to my attention. Kurt Mayer (1952) wrote an early book describing the broad development of these groups, followed more recently by another descriptive article (1977). The absence of more analytical work cannot be explained by the availability of only one question, as the Finnish examples demonstrate. Other sketchy remarks may be addressed to some of the other settings where data are available. For Alsace-Lorraine, data on language affiliation are available, but to my knowledge they have only been analysed by two Canadian demographers (Veltman, 1982; McQuillan, 1988). Few publications have so far appeared. Data on selected regions of Italy (South Tyrol, Trieste/Venezia Giulia), on Irish Gaelic in Ireland and on Gaelic in Scotland appear to have escaped the interests of demographers. The paper on Basques

< previous page

page_73

next page >

< previous page

page_74

next page > Page 74

at the Fourth International Conference on Minority Languages appears to be among the very first demolinguistic analyses to have been inspired by Spain's 1986 census data. Conclusions I have tried to lay out the basic framework for demolinguistic analyses of language communities. For obvious reasons, much of the existing work has been based on available census data. For those same reasons, the most sustained research traditions have been established in Canada and in Finland. However, many societies do already have usable census data, while many others may follow. The 'new immigrants', guest workers and refugees in the industrially developed societies of Western and Northern Europe are likely to be permanent fixtures of those populations. The social problems associated with these relatively recent arrivals will probably persist for a long time. Sooner or later, demographers will be called upon to help in estimating the numbers of people in these societies who are unable to communicate effectively in the majority language. The likely developments after '1992' can only aggravate the problems. A recent illustration of what may happen if the quest for meaningful public data is not handled carefully and intelligently is provided by the sequence of attempts by the British authorities to include a question on ethnicity to the population census. Aside from the fact that ethnicity is among the least useful measures in this field (see de Vries, 1985a, for discussion), it appears that the first attempts in this direction encountered opposition from several sides. On the one hand, many groups were apparently offended by what they regarded as pejorative labelling (this could have been avoided by a more general question about ethnic origin or ancestry, and a more comprehensive listing of response categories). On the other hand, well-intending researchers seem to have interpreted the proposed question as an invasion of privacy. As a consequence of this opposition, the Office of Population and Census Statistics now has to use complex estimation procedures to estimate the size of various minority groups on the basis of survey data. Reliability, validity and accuracy of such estimates are questionable. It would seem that the British census takers would have been better served with a sequence of questions on country of birth, mother tongue, main language. Such a sequence could have avoided accusations of 'racism' but could have yielded usable data. References Alcock, A.E. et al. (eds) (1979) The Future of Cultural Minorities. London: The MacMillan Press. Allardt, Erik and Starck, Christian (1981) Sprakgränser och Samhällsstruktur. Stockholm: Almquist and Wicksell.

< previous page

page_74

next page >

< previous page

page_75

next page > Page 75

Anderson, Barbara A. and Silver, Brian D. (1983) Estimating russification of ethnic identity among non-russians in the U.S.S.R. Demography 20, 461489. Bourbeau, Robert (ed.) (1988) Analyse démographique de la mobilité linguistique au Canada et au Québec. Montreal: Université de Montréal (Démographie). Carter, Harold (1980) The Welsh language in 1971. In Harold Carter and H.M. Griffiths (eds) The National Atlas of Wales. Cardiff. University of Wales Press. De Bartolo, Giuseppe (1987) La démographie des minorités Albanaises de Calabre. Premiers resultats. Paper presented at the European Population Conference, Jyväskylä, Finland, 11th16th June. Finnäs, Fjalar (1986) Den finlandssvenska befolkningsutvecklingen 19501980. Helsinki: Svenska litteratursällskapet i Finland. (1988) Aldersstrukturen i de svenska och tvasprakiga kommunerna och dess inverkan pa den nuvarande och framtide befolkningsutvecklingen. Abo: Institutet for Finlandssvensk Samhällsforskning (Research report #6). Fougstedt, Gunnar (1951) Finlands svenska befolkning 19361945. Helsingfors: Akademiska Bokhandeln. Haarmann, Harold (1975) Soziologie und Politik der Sprachen Europas. Munich: Deutscher Taschenbuch Verlag. Heise, David R. (1969) Separating reliability from stability in test-retest correlation. American Sociological Review 54, 93101. Henripin, Jacques (1974) L'immigration et le déséquilibre linguistique. Ottawa: Information Canada. Kloss, Heinz (1929) Sprachtabellen als Grundlage für Sprachstatistik, Sprachenkarten und für eine allgemeine Soziologie der Sprachgemeinschaften. Vierteljahrsschrift für Politik und Geschichte 1, 10317. Krejci, J. and Velimsky, V. (1981) Ethnic and Political Nations in Europe. London: Croom. Helm. Kremnitz, Georg (1977) Die ethnischen Minderheiten Frankreichs. Tübingen: Günther Narr Verlag. Lachapelle, Rejean and Henripin, Jacques (1980) La situation démolinguistique au Canada. Evolution passée et prospective. Montréal: The Institute for Research on Public Policy. (1982) The Demolinguistic Situation in Canada. Montréal: Institute for Research on Public Policy. Lallukka, Seppo (1987) Changing age-sex compositions as an indication of ethnic reidentification: the Mordvins. Paper presented at the European Population Conference, Jyväskylä, 11th16th June. Lieberson, Stanley (1970) Language and Ethnic Relations in Canada. New York: John Wiley and Sons. Mayer, Kurt B. (1952) The Population of Switzerland. New York: Columbia University Press. (1977) Groupes linguistiques en Suisse. Recherches Sociologiques 8, 7594. McQuillan, Kevin (1988) Protoindustry and Marriage: Some Evidence from Alsace. London, Ontario: University of Western Ontario. (Population Studies Center, Discussion Paper 888) Petrella, Ricardo (1978) La Renaissance des cultures régionales en Europe. Paris: Editions Entente. Pryce, W.T.R. and Williams, C.H. (1988) Sources and methods in the study of language areas: A case study of Wales. In Colin H. Williams (ed.) Language in Geographic Context. Clevedon: Multilingual Matters Ltd. Ravenstein, E.G. (1879) On the Celtic languages in the British Isles: a statistical survey. Journal of the Royal Statistical Society 42, 579636. Rother, K. (1968) Die Albaner in Suditalien. Mitteilungen der Osterreichischen Geographischen Gesellschaft 110. Salvi, Sergio (1973) Le nazioni proibite. Guida a dieci colonie 'interne' dell Europa occidentale. Firenze: Valecchi. Shryock, Henry H.S. and Siegel, Jacob S. (1976) The Methods and Materials of Demography. New York: Academic Press.

< previous page

page_75

next page >

< previous page

page_76

next page > Page 76

Stephens, Meic (1976) Linguistic Minorities in Western Europe. Llandysul, Wales: Gomer Press. United Nations Annual Demographic Yearbooks. New York: United Nations. Veltman, Calvin (1982) La régression du dialecte. Chiffres pour l'Alsace 3, 3942. Vries, John de (1974) Net effects of language shift in Finland, 19511960: A demographic analysis. Acta Sociologica 17, 14049. (1977) Explorations in the demography of language: Estimation of net language shift in Finland, 19611970. Acta Sociologica 20, 14553. (1985a) Some methodological aspects of self-report questions on language and ethnicity. Journal of Multilingual and Multicultural Development 6, 34768. (1985b) Secondary data analysis on ethnolinguistic minorities: Problems and pitfalls. In Peter H. Nelde (ed.) Plurilingua V: Methoden der Kontaktlinguistik. Bonn: Dümmler Verlag. (unpublished) Notes for demographic analysis of Swedish Finns. Vries, John de and Vallee, Frank G. (1980) Language Use in Canada. Ottawa: Statistics Canada. Weinreich, Uriel (1953) Languages in Contact. New York: Linguistic Circle of New York. Williams, Colin H. (1982) The spatial analysis of Welsh culture. Etudes Celtiques 19, 283322. Winkler, Wilhelm (1923) Die Bedeutung der Statistik für den Schutz der nationalen Minderheiten. Leipzig/Wien: F. Deuticke.

< previous page

page_76

next page >

< previous page

page_77

next page > Page 77

Legitimating or Delegitimating New Forms of RacismThe Role of Researchers 1 Tove Skutnabb-Kangas Roskilde University Centre, DK-4000 Roskilde, Denmark Abstract When 'races' were defined on the basis of purportedly biological criteria and when various psychological characteristics were linked with the resulting 'races' (which were thus hierarchised), researchers participated in this definition process. 'Races' were socially constructed, and the constructions were used to legitimate an unequal division of power and resources between these different 'races'. Now biologically based racism has in most European and Europeanised countries been replaced by more sophisticated forms of racism, ethnicism and linguicism, which use the ethnicity, culture and languages of different groups as the basis for hierarchisation. Ethnicism/linguicism are defined as 'ideologies, structures and practices which are used to legitimate, effectuate and reproduce an unequal division of power and resources (both material and non-material) between groups defined on the basis of ethnicity/culture/language' (Skutnabb-Kangas, 1986: 45). In this new phase of racism the role of researchers in socially constructing ethnic groups and producing arguments for the hierarchisation of languages becomes vital. According to a human rights oriented argumentation, it should be the right of every individual and group to have their own definition of their ethnicity/mother tongue accepted and respected by others. In this sense both ethnicity and the mother tongue can be seen as reified, as characteristics of an individual or a group, who (should) themselves have the right to decide about them. Integration into a 'mainstream' (as a prerequisite for forming a national ethnic minority) has often also been seen in this way as a characteristic of a minority group, where the group itself is studied and its 'degree of integration' evaluated. Even if integration is studied as a dynamic process, the outcome of this process is still treated as a characteristic of the individual/group to be integrated. If, by contrast, ethnicity and integration are treated as socially constructed relations, rather than inherent or acquired characteristics, the power relationships between the parties in the definition process become a primary object of studyas opposed to just one of the parties, mostly the dominated one, being studied. The human right to selfdefinition makes sense only when the parties are equal. If minorities are defined

< previous page

page_77

next page >

< previous page

page_78

next page > Page 78

on the basis of power, not numbers, minorities are per definition not equal parties in the negotiation processes about their mother tongues, ethnicities and integration. This paper will discuss some of the ways in which researchers participate in constructing and legitimating the new forms of racism, through the ways we construct and treat concepts like mother tongue, ethnicity and integration. Introduction It is a tragic academic bon mot that 'an Aboriginal/Inuit/Same family consists of a mother, a father, three children and an anthropologist'. The phrenological anthropologist of race hygiene days has been replaced. Nowadays 'a minority family (migrant or indigenous) consists of a mother, a father, three children and a linguist/sociolinguist. A minority group is a group followed by several sociolinguists, anthropologists and sociologists, who are outsiders. In many countries they are also followed by a horde of fixers, mediators and understanders.' Researchers have been and are collectors of information and producers of meanings. These can be used and have more often than not been used for controlling minorities. The control of some groups by others ensures an unequal division of the power and the resources which our planet has. For this control to take place, structures, ideologies and practices are constantly being constructed and, after contestation and struggle, reconstructed in ways typical of hegemonies. These ideologies, structures and practices make the control work, and even make it seem fair. One of the most powerful tools for control has been and is racism. It is not the only oneit interacts in complex, multilayered ways with classism, nationalism, sexism and othersbut it is racism in some of its contemporary forms that I will concentrate on in this paper. Minorities and colonies were once controlled by the gun and the Bible. The Bible was of course provided by linguists. Those linguists/missionaries, who were outsiders, worked in an anthropological way, living with the group while learning its language in order to be able to reduce this to writing. This type of control still exists, and the Summer Institute of Linguistics is a good example of how to support it. Their role in introducing and maintaining control has been analysed to some extent (Calvet, 1987; Hvalkof & Aaby, 1981; see also Vollmer & Knödler, 1988). But on a global scale the control is now effected more through technology and textbooks. Both are provided by researchers. In the underdeveloped countries multinational or transnational companies, international organisations and 'aid' organisations are in charge, together with native élites. In the case of indigenous and immigrated minorities in Europe and Europeanised countries, the whole state apparatus, including researchers, is involved. When control moves from physical to psychological, when the direct brutal

< previous page

page_78

next page >

< previous page

page_79

next page > Page 79

control of the land, the raw materials and the body as a means of production are being replaced or have been replaced by control of the mind through 'information', the role of researchersin providing the means for this control and the legitimatory arguments for itgrows. When people are controlled through verbal and visual messages, through significations and representations, the experts on messages, among them linguists, come to play a crucial role. We can help to construct and legitimate the messages used for controllingor to deconstruct and delegitimate them. The defining criteria for dividing people into groups with more and groups with less access to power and resources are now in the process of a paradigmatic shift on a broad basis. This change started after World War II. It was prepared by researchers through Unesco in a series of four conferences which culminated in a Unesco statement on racism in 1967, which speeded up the change. Now the change of paradigm is almost complete, and the consequences start to show. According to Robert Miles' (1989) analysis of racism, which I will follow, the criteria for how Others were differentiated from SelfSelf being 'white' Europeanswere initially during a millenium both biological and cultural, and were not seen as hereditary and fixed. Then for a couple of centuries mainly biological criteria were used. These were thought of as hereditary. They were linked to cultural characteristics which were also seen as hereditary, but the cultural criteria were during the main phase of biologically based racism not part of the defining criteria. When, for several reasons, biological criteria became untenable, cultural criteria again started to gain prominence (after World War II). Now they are being used more or less exclusively to divide people into groups. In this paper I am not going to present any definitive theories about the new forms of racism, but only outline some recent actual and possible developments and our role as researchers in producing and legitimating or delegitimating them. As I see it, control legitimation is supported by researchers who use concepts which socially construct differences so that they are seen and treated as characteristics in individuals and groups, characteristics which human beings somehow possess, characteristics like white or black earlier, or ethnic this or that now. If you are 'different', you have to be different from something. It is only comparison which brings out a difference. Therefore differences should rather be conceptually constructed and treated as mutual relations between the definer and the defined, not as characteristics of the defined. This way of conceptualising can also help the delegitimation of control. Therefore we have to redefine what is now seen as cultural characteristics of individuals and groups, as relations, instead of characteristics. I will show in the rest of the paper some of the consequences of defining a few concepts either as characteristics or as relations, and use ethnicity and integration as examples. I will also discuss intercultural education and the role of language in it as an example of how the earlier racist theories are

< previous page

page_79

next page >

< previous page

page_80

next page > Page 80

perpetuated in a more covert form. The goal of the exercise is to show how. the way we construct our concepts can legitimate or delegitimate a racist discourse. I will start with a short description of the development of racism, mainly based on Miles (1989). The Development of Racism Racism presupposes a discourse of 'race' (however we define it), and a racialisation of groups, i.e. attributing groups a 'race'. In the racialisation of groups, 'race' is treated as a characteristic of a person or a group with a certain 'race'. Being seen as 'white' or 'black', or 'ethnic this or that', is a result of attributing a person or a group with a characteristic, which the person or group is then seen and treated as possessing. In differentiating between groups we start from ourselves, and compare with others, i.e. differentiation presupposes images about the Others that are possible only when we also have images about Self. According to Miles (1989), the earliest 'white' European images of the Others, at least from the eighth century onwards and up to the late eighteenth century, saw these Others as both biologically and culturally different, but not necessarily sub-human. They were God's creations, as were white Christian Europeans, and thus human beings. But since God had organised all living beings in hierarchies, where for instance animals were lower than humans, this permitted hierarchisation of human beings too, with the help of either purely religious or environmentalist arguments. Even if 'races' were discussed in terms of common lineage and described mainly with the help of biological criteria like skin colour, 'lineage' was not seen as strictly biological but also as cultural (cf. Miles (1989): 'the Anglo-Saxon race'). Other 'races' were seen as capable of developing to the same standard (of civilisation) as the Europeans. Thus the psychological (i.e. cultural) characteristics of the Others were not seen as inherent and hereditary, but as capable of change. When slavery was initially legitimated, it was mainly done by reference to the presumed 'fact' that 'black' people were environmentally better suited than Europeans to work in the tropical climates on plantations. It was also seen as giving 'black' people a chance to develop away from savagery, when they came into contact with 'white' achievements and ways of life. In this way we 'helped' and 'civilised' them, so that they at some point might reach the same level of civilisation which 'we' had, i.e. acquire new cultural characteristics. Then they might be mature enough to govern themselves. But towards the end of the eighteenth century 'a major transformation occurred in European representations of the Other as a result of the secularisation of culture and the growth and increased hegemony of science' (Miles, 1989: 3031). The phenotypically observed differences of 'race' started to be' interpreted from the late eighteenth century onwards as biological and unalterable. Thus 'distinct ''races" of human beings had always existed' and

< previous page

page_80

next page >

< previous page

page_81

next page > Page 81

'the hierarchy of inferiority and superiority was therefore natural, inevitable and unalterable' (Miles, 1989: 33). This biologically based racism of course also noted different psychological characteristics, but saw them, not as defining, but as biologically determined or mediated and therefore also as more or less natural, inherited, fixed, and therefore unalterable. Purportedly biological characteristics were thus bunched together to function as defining criteria for 'races', and various psychological characteristics were then linked with or attributed to the resulting 'races'. In this way the 'races' were hierarchised, on the basis of an evaluation of 'their' unalterable psychological characteristics, and some 'races' were then seen as more fit to rule than others. 'Races' and 'their' characteristics were thus socially constructed. Researchers were central in this definition process all the way through. Historians now agree, according to Miles, about the role of researchers as decisive in this paradigm shift. The ideological racial constructions were used to legitimate an unequal division of power and resources between these different 'races'. The ideology of biological racism legitimated the control and exploitation by the 'white' 'race' of other 'races'. Now the racist ideology defined the Others as 'sub-human' or 'non-human', or at least 'less-human-than-us', claiming that they also had inherited psychological characteristics which made it necessary and fair for 'us' to control 'them'. Now we know that from a scientific point of view, 'races' do not exist as biologically based categories. Likewise, it cannot be shown that there would be causal correlations between biological and psychological 'characteristics' in humans. Therefore racism based on biologically based criteria makes no sense, regardless of whether 'races' are defined on the basis of visible biological criteria like skin colour, or less visible or 'invisible' biological criteria of the kind that genetics use, like blood groups, or measures of brains or skull form. Of course, 'race' still exists as a sociological category, in both scientific and lay use, even if biologically based 'races' are scientifically dead. But racism has not died with the death of biologically based 'races'. It is still needed for the unequal division of power and resources (there are other reasons, too, less functionalistic). Present day racism has changed shape in order to seem more plausible. Cultural differences have again become important, but this time around not together with biological differences but rather to the virtual exclusion of biological differences. Biologically based racism has in most European and Europeanised countries been replaced by more sophisticated forms of racism, ethnicism and linguicism. These use the ethnicities, cultures and languages of different groups as defining criteria and as the basis for hierarchisation. It is not being claimed any more that certain 'races' are more fit to rule. Now it is certain ethnoses, cultures and languages which are claimed to be more fit to rule, expand and be learned by others. In this cultural, ethnic and linguistic Darwinism the fittest cultures, ethnoses and languages survive.

< previous page

page_81

next page >

< previous page

page_82

next page > Page 82

Or put in sociobiological language: those cultures, ethnoses and languages that are best equipped to adjust to forms of life which guarantee the continuation of their representatives will survive. In this new phase of racism the role of researchers in socially constructing ethnic, cultural and linguistic groups and producing arguments for the hierarchisation of ethnoses, cultures and languages may again be as vital as it was in constructing concepts and categories for and thus legitimating biologically based 'race' and racism. First a clarification. We have to distinguish the new forms of racism from those where the same old ones are just euphemistically called something else. 'Race' as a biological concept still prevails in lay use, but it has not completely disappeared as a scientific concept either. There are researchers who use the word 'ethnicity', when they in fact mean biological 'race' (and admit it, at least indirectly) in order to avoid the nasty connotations of 'race'. In the following quote only the word has been abandoned, not the concept: The term ethnicity is used throughout this chapter whenever possible. The use of ethnicity, rather than race, is based on Montagu's (1972) argument that potentially negative connotations associated with racial terms may be lessened by substituting the word ethnicity for race. (Hammer, 1986: 235) In talking about the new racism, I do not include this type, where biologically based 'races' are still taken as a reality. The new forms of racism, ethnicism and linguicism, give immigration countries a new opportunity for concealing even a soft, non-brutal assimilation policy. These do not say, like the old social Darwinists, that there is a struggle for survival amongst the different human 'races', in the course of which those with lesser intelligence or capacity for 'civilisation' would eventually disappear, their elimination being evidence of their natural inability to evolve. (Miles, 1989: 3637) Now they first say that the cultures and languages which are going to survive will do so because they are more modern, more adapted to modern life, more developed or have more potential for developing than others. Then they see to it, through schools and other institutionally controllable measures, that languages and cultures other than their own do not get a real chance to survive. The majority language and culture always somehow turn out to be the strongest survivors. Then it is easy to be nice and liberal: 'Isn't it a pity that the second/third generation migrant youngsters did not want to maintain their parents' cultures and languages, they would have been such an enrichment to our country. It is, of course, understandable, that they wanted to learn our language and culture; after all they saw how important they are, but still, it is a real pity that they didn't want to learn Kurdish/Finnish/ Vietnamese/Turkish/Urdu etc.'

< previous page

page_82

next page >

< previous page

page_83

next page > Page 83

When different 'races' were defined and hierarchised, it was even then already partly a question of symbolic power: who had the right to decide on an individual's or a group's 'race', and perform the hierarchisation. To racialise people, to divide them into 'races', meaning the right to define which characteristics were bunched together and attributed to a group to form a 'race', is a prerequisite for hierarchising the resulting 'races'. That is, of course, why South Africa clings to the ultimate power of the state to accept the racial identification of its people. In ethnicism and linguicism it is in principle a similar question. In order to be able to hierarchise ethnoses, cultures and languages, they have to be delineated first. The crucial question then is who has the ultimate right to define them. Do every individual and group have the right to define their own ethnicity and linguicity? Is it sufficient that they themselves do it, or does the definition need a validation by an outsider? How often is this outsider a researcher, directly or indirectly (so that the researcher's concepts and arguments are used)? Defining ethnicity and linguicity, and then ethnicising and linguicising groups, attributing them these predefined ethnoses and languages as characteristics which they are then seen as possessing, is an important symbolic act, because it forms the basis for present day division and redivision of power and resources. In the remainder of this paper I will illustrate some of the ways in which we researchers participate in constructing and legitimating the new forms of racism, through the ways we construct and treat concepts like racism, ethnicity, and integration (or accept their treatment in official papers). I will obviously have to touch upon only a few aspects of the complex story. Defining Racism, Ethnicism and Linguicism Racism can be defined in a number of different ways (see Miles, 1989, and references in it). One of the basic debates in racism research is about the scope of the concept. Some researchers define racism as an ideology only. Some want it to include exclusionary practices and consequences of these also (see Miles, 1989, especially chapter 2, for a review). Basically I agree with Robert Miles' criticism of the shortcomings of many of the definitions which have included practices and consequences. At the same time I see it as necessary to include structures and practices, in order to be able to discuss the newer forms of racism. This is so partly because many of those definitions which see racism as an ideology only, intentionally or unintentionally come to treat racism as an information problem and at an individual level only. To me, basic requirements that any usable definitions of racism in all its forms need to fulfil are that they are multidisciplinary and multicausal, and look at both actors and structures. A racism definition is from my point of view pretty useless, if it treats racism only as a problem which exists as

< previous page

page_83

next page >

< previous page

page_84

next page > Page 84

images, attitudes and opinions in people's heads (like some psychologists or social psychologists tend to do) (cf. Lange & Westin, 1981; Liebkind, 1984), or if it treats racism only as a societal structural problem (like some sociologists and political scientists tend to do). To me, racism is not a bunch of nasty illwilled ignorant people. Racism is not only the attitudes of individuals towards other individuals, or the prejudices of individuals towards other individuals. Racism is not an information problem alone, where people would have prejudices and negative attitudes because they don't know enough about other people, and where racism would disappear once they receive enough relevant information about each other. But neither is racism only a structural labour market or housing market problem, or a political power structure which would continue to exist regardless of what thoughts, prejudices and attitudes individuals and groups have about each other. Both these ways of seeing or treating racism are simplistic or result in a simplistic treatment of racism. Both reflect those monolithic, monocultural either/or ideologies which are so typical of Europe and especially Europeanised countries. We can as linguists see these ideologies clearly in relation to language, as the use of norms according to which it is difficult to understand or accept bi- or multilingualism as ecologically sound and feasible, and which often assume either/or solutions, demanding a focus on the majority language only, and seeing an emphasis on the minority language as excluding the majority language. In the same vein, racism can be seen in this monocultural way, as being about either attitudes or structures, but not both. Theories which see racism only as something in people's heads contribute to empty idealistic liberalism: let's be nice to each other, and racism disappears. Theories which see racism only in terms of social structure contribute at best to radical pessimism that ignores people and their possible contributions as a factor for change, or at worst to apathy: since you cannot change the world's power structure anyway, you might just as well forget about it and garden or ski slalom or exploit the system yourself and invest in South Africa or Ireland ... I think it is more profitable to see racism as a relation. It is a relation between groups of individuals (groups of actors), a relation between different societal structures, and a relation between these actors and societal structures. Measures to fight racism which are geared only towards individuals (either 'the racists' or 'the victims of racists') are often wellmeaning, but naîve and not worth much to minorities. Measures geared towards societal structures only may achieve more for minorities, at least initially, but they run a risk of being too deterministic or based on theories which seem to be vulgar Marxist, ignoring the complexity and dialectic nature of the issue. It is the relations between groups and relations between groups and structures that need to be addressed.

< previous page

page_84

next page >

< previous page

page_85

next page > Page 85

Sweden has repeatedly been singled out by international organisations and researchers from other countries as The Model when it comes to the treatment of immigrants and refugees. Sweden even has an ombudsman against discrimination. The Swedish State Committee Against Racism and Xenophobia has just published its main report. In its discussion about ethnic conflict, the Committee mentions several times with approval Lange & Westin's multicausal theory with its three basic elements: 'unequal division of power and resources, lack of contact between ethnic groups, and lack of self-confidence in individuals' (SOU, 1989: 13, 20). Here they thus approve of a model which sees causes for ethnic conflict at three levels, at a structural societal level, at a group level and at an individual level. But when the Committee then set out to give their own definition of racism, which of course is the key concept in the report, the societal structural level disappears completely, and racism exists only in people's heads as images, views and opinions. This is their definition: Racism can be described as a view of the superiority of one's own group of people and an opinion about the existence of biological differences between groups of people which motivate a division of them into those who are more worthy and those who are less worthy. In addition, it means that a group of people which sees itself as a more worthy 'race' holds the opinion that they have the right to oppress, exploit or control the others or force them to live separated from other groups of people. (SOU, 1989: 13, 19, my translation and underlining.) In addition to seeing racism in people's heads only, this definition sees only the 'old' primitive biologically based racism as racism. Newer, more sophisticated forms of racism, forms which Sweden uses in an expert way, are not encompassed by this definition. As I see it, this type of definition makes it impossible to analyse present day racism in Sweden or in Scandinavia in any meaningful way. It is sad to have to add that the Committee has had researchers as consultants in its work. In our definition, Robert Phillipson and I combine structures and actors. We define racism/ethnicism/linguicism as 'ideologies, structures and practices which are used to legitimate, effectuate and reproduce an unequal division of power and resources (both material and non-material) between groups defined on the basis of race/ethnicity/language' (Skutnabb-Kangas, 1986: 45). With this definition in mind, we will proceed to look at some of the other concepts which we use as our professional tools, in order to see whether they can (be used to) legitimate the newer forms of racism, where cultural characteristics are used as defining criteria to hierarchise groups and what the consequences are of conceptualising them as characteristics or as relations. The first concept to look at is integration.

< previous page

page_85

next page >

< previous page

page_86

next page > Page 86

Integration In many countries integration is still used as a euphemism for assimilation. Just as 'ethnic' is used because it does not have the same nasty connotations as 'race', without questioning the validity of what the orginal concept refers to, saying integration when meaning assimilation may sound nicer. The aforementioned Swedish State Committee also does this. It is clear from the quote that follows that they in fact mean assimilation when saying integration: There is reason to believe that some groups of immigrants will successively integrate into the Swedish society. But a probable development is also that other groups will maintain and specifically emphasize their ethnic distinctiveness. (SOU 1989: 13, 115) If immigrants maintain their ethnic distinctiveness, they thus cannot integrate, according to this official Swedish view. Not surprisingly, integration has often in this direct assimilation phase been discussed, as in the Swedish quote, as a static and non-dialectic, non-reciprocal concept, which is treated as a final characteristic of the individual or group to be integrated. It has been discussed as a final result of a process which might start with segregation, and go via functional adaptation and acculturation to either integration, assimilation or marginalisation. In this view, the immigrant or the immigrant group is seen as the one who has to be integrated into a majority mainstream society, which exists an sich, without being influenced to any noticeable extent by the immigrant. There is something static in the concepts used even when the dynamic process aspect has been emphasised, i.e. when arriving at integration has been described as a process rather than describing only the product of the integration process, or despite images like 'melting pot', 'salad bowl' or 'cultural mosaic', images which imply some sort of change in the majority society, or at least an addition to it. Once you are integrated, that's it. Moreover, it is only the immigrant or the immigrant group which is to be integrated into something that is more or less there, waiting. It is the immigrant's degree of integration which is discussed. The Turks are said not to have integrated into Danish society. And the Turks in the Federal Republic of Germany are said not to be willing to integrate into German society. Integration is not discussed and evaluated in terms which would see to what extent both parties involved in the integration process have integrated, namely both the immigrants and the rest of the society. Integration is thus treated as a characteristic of a minority individual or group. Not 'being integrated' is seen as a negative cultural characteristic, which can then be used and has been used to legitimate unequal access to power and resources.

< previous page

page_86

next page >

< previous page

page_87

next page > Page 87

When integration is understood as the final result of a one-way process, this also has consequences for how the majority society is treated. When those who come have to change and to embrace the goals, ideals and identity of the new country, in order to fit into a virtually unchanged society, this society can continue to regard itself as the mainstream, and the mainstream is always seen as integrated into 'its own' society. 'Being in the mainstream' is thus also treated as a characteristic, a characteristic of the majority, which they somehow possess in a natural way. Even in avowedly immigration countries like Australia, dominant group policy wants 'to develop ... mechanisms that would bring nonEnglish speaking minority groups into recognised relationship with the "main-stream" Australian group life' as Jerzy Zubrzycki (1988), following Jean Martin (1971; 1978) presents this 'emasculated pluralism'. Minority groups have to participate 'in the shared and "universalistic" structures of the wider society' (Australia, 1975: 4851, quoted in Zubrzycki, 1988: 12). This static and ethnocentric view where the whole burden of integration is on the incomer alone, and where the dominant group's values are presented as somehow 'shared' and 'universal', rather than particularistic and changing, like all values are, still prevails in many countries. When the majority population is presented in this way as an integrated mainstream, homogeneously sharing universal cultural values, this characteristic legitimates its access to power and resources (which are, of course, shared unevenly on a class and gender basis within the majority population, but this is often not mentioned in the integration discussion). Part of this view of integration is also that the minority member is seen mainly as A Problem both for herself and for the 'host' society, until s/he is integrated/assimilated. This problem-orientation shows nicely even in the programme for AILA's world congress in Thessaloniki in 1990: the only section (11) which explicitly has to do with migrants is symptomatically called 'Problems in Migrant Education', analogous with sections like 'Language Disorders and Communication Problems', rather than, simply, 'Migrant Education', analogous with other sections such as 'Language Policy and Planning', or 'Child Language Development'. Minorities are seen within a deficiency-based framework, where the problems they face are not analysed in terms of the racism, ethnocentrism and discriminatory practices of the majority society (i.e. in terms of the relation between the minority and the majority), but are seen as resulting from handicaps and deficiencies in the minorities themselves. Minority groups are seen as possessing integration-preventing characteristics which cause their problems. They are defined negatively, in terms of what they are not, do not have or do not represent (majority-language-speaking middleclass majority members). Terms like LEP- or NEP-children in the United States (Limited English Proficiency, or No English Proficiency), or fremmedsprogede børn in Denmark reflect this. I will give you a tiny but symptomatic anecdotal

< previous page

page_87

next page >

< previous page

page_88

next page > Page 88

example of such unconscious ethnocentricity. My husband Robert Phillipson and I have both had letters from our children's schools, saying (in 'English', and this is a direct quotation): 'Dear parent. You have a foreign speaking child.' One of the children reacted like this: 'What do they mean, foreign. I don't speak foreign. I speak my own language.' The same English-language letter, sent in Denmark to many Muslim parents, wanted to have 'the Christian name of the child' identified ... Much of the research into minority education and about minority women is still based on different types of deficiencybased frameworks, some more patronising than others, some using more sophisticated pitying approaches (SkutnabbKangas & Leporanta-Morley, 1988). Minorities should of course be defined positively, in terms of what they are, have or represent, and their cultural and linguistic inheritance should be seen as assets in a mutual exchange relationship with the majority. I will return to the deficiency-oriented frameworks in connection with discussing intercultural education. The non-dialectic view of integration also has consequences for what is seen as ethnic. Even a lot of researchers see only minorities as possessing ethnicity, whereas majorities are not ethnic. Even if it is acknowledged that the migrant may have something to contribute, the host society decides what of it is accepted as enriching, and thus allowed to exist as part of the 'mainstream': ethnic food, dances, clothes, etc. The rest, meaning a lot of the more vital parts of minority cultures, are seen as belonging to the domain of the private sphere. They are parts of the minorities' 'private ethnicity', and 'public institutions' such as schools should not 'promote private ethnicity', because 'matters of ethnicity are best left to those directly concerned', as John Edwards (1984b: 299300) puts it. But the same attributes are not seen as part of any private ethnicity when it comes to the majority population. Their 'private ethnicity' is of course being supported by public institutions such as schools. Majority language schools for majority children 'naturally' maintain their language, which should logically also be seen as a private aspect of their ethnicity. But Edwards does not seem to be worried because he obviously does not see majority members as having ethnicity. Otherwise the logical conclusion of the view he takes would be to abolish the use of Danish in Danish schools, the use of English in British schools, etc., to judge by the following quote: Educational programmes aimed at sustaining ethnic identity through communicative language maintenance are misguided and may even damage those private aspects of ethnicity (i.e. language) which are essentially out of the reach of external intervention. (Edwards, 1984a: 14) Perhaps a challenge to the educational language planners of integrated Europe? Abolish majority languages and cultures for majority populations

< previous page

page_88

next page >

< previous page

page_89

next page > Page 89

completely from all European schools, because they are part of a private ethnicity and programmes for their maintenance are misguided ... Even in a more progressive atmosphere, in an intercultural phase, when integration is defined in a way which allows the minority to retain what it wants of its own cultural heritage, in addition to being allowed to learn and use cultural and structural features of the host society on an equal basis, i.e. is allowed to become structurally incorporated and bicultural, this may still mean that the host society does not really change. It just lets a bilingual, bicultural migrant (group) incorporate structurally into an unchanged structure. All these ways to construct the concept of integration can be used to legitimate less access to power and resources to minority groups, which are seen as possessing the characteristic of being unintegrated. At the same time integration is in them defined in a discriminatory way which forces minorities to assimilate, because no change is required in the majority population or society. Nor do these deficit theories challenge the static, ethnocentric view of majority cultures as more universalistic, and majorities as devoid of ethnicity. In thus making invisible those 'characteristics', on the basis of which the majority group gets access to more power and resources, these definitions implicitly legitimate and reproduce this unequal access, meaning that they function in a racist/ethnicist way. As an alternative to this view, I want to consider the definition of integration used by Drobizheva & Gouboglo (1986): 'Integration: formation of a series of common features in an ethnically heterogenous group.' This definition of integration has, as I see it, a few extremely important implications for theorising about both integration and ethnicity, also when thinking of the European integration project. If integrated Europe is to be 'an ethnically heterogenous group', that means that all groups living in Europe must become conscious of their ethnicity and its implications, also the fact that most majority ethnicities have elements of dominance. Minorities are mostly aware of it, possibly with the exception of migrants who come from former imperialist countries. (In Canada 'the British are not considered immigrants', according to Helen Ralston, 1988). The first implication of the definition is that majority ethnicities have to be acknowledged and made visible. Unconscious invisible ethnicity is one of the sources of ethnocentrism. Secondly, all groups have to change, and everybody should have the right to contribute on an equal basis. There cannot be any kind of Europeanness, that we 'in the fringe areas' are invited to join. If, like many marine historians have shown (e.g. Matti Klinge (1983) for the Baltic area in Merivaltiot ('The Sea Nations')), seas have connected people and made communication easy while land, woods, and mountains have divided them, our mental pictures when looking at maps where we place the centre of Europe somewhere in the middle of the land mass are incorrect. Maybe those with access to seas are in fact the centre, and those in the middle of the land the fringes.

< previous page

page_89

next page >

< previous page

page_90

next page > Page 90

Scandinavia is pretty well placed, then, compared to France or the Germanies. Sjaellands Odde is more central than Strasbourg. Thirdly, and most importantly for our present discussion, integration should be regarded as a process and a socially constructed relation, not a product or a characteristic in a person or a group. There is, or should be, a methodological development in relation to what is being studied in integration research, which in a way parallels the theoretical and political development. When the migrant has to do the integrating into an unchanged 'host' society, and has done her or his job properly, then the migrant is at some point described as integrated. 'To be integrated' is in this phase of direct assimilation seen as a more or less stable characteristic of the migrant, or a migrant group. In this phase researchers can just study the migrant, or the migrant group, and compare with some type of statistical mean for the host population, in order to judge to what extent the migrant is integrated. Does the Turkish man drink as many beers as the Danish man, and does the Turkish woman wear 'Western' clothes? Integration is often in this phase discussed in terms of the migrant's voluntary wish to integrate or not, without structural and ideological hindrances in the majority society being discussed (as mentioned earlier, the Turks in all Scandinavian countries and in FRG are often accused of lacking a wish to integrate). In the intercultural phase, when it starts to be acknowledged that the host society has to change also, integration is seen as a characteristic of both the migrant (individual and/or group) and the host society (mostly at a statistical, not individual level). The view may not be as static as in the first phase, i.e. the characteristic is not seen as stable but as changing, but integration is still seen as a characteristic of individuals/groups. Thus it is these individuals and groups (both minority and majority) that are seen as the ones to be studied. When we come to the radical-utopian phase, which the Drobizheva & Gouboglo, (1986) definition implies, integration must be seen as a socially constructed relation, in the same way as gender, class and ethnicity. 'Formation of a series of common features in a ethnically heterogeneous group' implies that integration is seen as a process of negotiation about the legitimacy of accepted features, which may originally come from any of the groups involved, and where compromises must be sought about them. Since the integration relation is socially created and constructed through interaction between the parties to be integrated, the conditions for this interaction become equally important as objects of study as the actors. If there is unequal access to power and resources in the negotiation process, there is a risk that some features are being forcibly imposed as 'universal' and 'common' to alland these are invariably features of the powerful groups. If a minority group persists, and wants the majority group to accept some of its features too (in the same way as the minority accepts some of the majority's features), an unwilling majority can either dictate terms or cease negotiating. It can revert back to (or never stop) seeing integration as a

< previous page

page_90

next page >

< previous page

page_91

next page > Page 91

characteristic, and claim that the majority is integrated anyway, because it represents the mainstream. Thus a powerful majority may subvert the integration process, and simultaneously arrogate to itself the power to define the minority as non-integrated and itself as integrated. Even in this situation it is the process of negotiation about integration that should be the major focus of study, in addition to a study of the actors involved in the negotiation process. It thus becomes a study of the structural and ideological conditions for the negotiation, including the power relationships between the actors. Integration has barely yet been studied in this way. By concentrating the study of integration on researching the migrants/minorities and 'their' characteristics, the way the initial examples of Same, Aboriginal, Inuit or migrant families described, the conditions for real integration, most importantly the power of the majority to prevent integration, are neglected. Thus the concept of integration is constructed in a way which permits it to be used in legitimating new forms of racism. Ethnicity An ethnic group can be defined in many ways too. One of the more precise definitions used in the literature is that by Erik Allardt (Allardt & Starck 1981: 43) who uses the following four criteria: (1) self-categorization (self-identification) (2) descent (3) specific cultural traits, e.g. the capacity to speak a specific language (4) a social organisation for interaction both within the group and with people outside the group. According to Allardt, there are no criteria for inclusion in an ethnic group that all the members should fulfil. But it is necessary that some members fulfil all the criteria before one can speak of an ethnic group, and every member must fulfil at least one criterion. Often most members fulfil all the criteria. But there are also some ethnic lukewarms and ethnic self-haters who do not categorise themselves as members despite fulfilling all the other criteria except selfcategorisation and despite being categorised as members by others. One problem in such situations with cognitive dissonance is that 'forced other-categorisations' are seen by many researchers (e.g. Liebkind, 1984: 19) as violations of basic human rights. According to a human rights oriented argumentation, it should be the right of every individual and group to have their own definition of their ethnic group membership (or mother tongue) accepted and respected by others. On the other hand, according to Allardt, self-categorisations and other-categorisations have different logical structures. Other-categorisations imply reference to other superficial criteria (like cultural traits, language or organisation), which are part of the definition anyway, whereas self-categoris-

< previous page

page_91

next page >

< previous page

page_92

next page > Page 92

ation only presupposes a wish to identify or categorise as a member of the group. Therefore Allardt sees it as unnecessary to have exo-categorisations (Bromley, 1984) as a part of the definition, but on different grounds from the human rights-oriented argument of Liebkind and others. In a human rights sense ethnicity (or the mother tongue) can be seen as reified, as a characteristic of an individual or a group, who (should) themselves have the right to decide about it. But even without this orientation, the criteria for belonging to an ethnic group as outlined by Allardt are all seen as characteristics in individuals making up that group or in the group itself. All the characteristics can be seen as cultural. By contrast, I think ethnicity can also more profitably be treated as a socially constructed relation, in the same way as gender or class. I do not see it as theoretically profitable to treat ethnicity as an inherent or acquired characteristic of an individual or a group who is said to 'have' a certain ethnicity. Ethnicity is also a relation. And relations cannot be decided by one party in the relation alone, relations are to be negotiated. Ethnicity is not a characteristic that a member of an ethnic group 'possesses', it is something that the individual or group must negotiate about with significant others. If ethnicity is seen in this way, as something to be negotiated, something that needs validation from both parties in order to exist, all the parties involved in the negotiation process must be taken into account when defining somebody's ethnicity. If this is so, then the power relationships between the parties in the definition process become a primary object of studyas opposed to just one of the parties, mostly the dominated one, being studied. One consequence of this is, in contrast to the human rights oriented demand for self-categorisation, that othercategorisations have to be seen as a part of somebody's ethnic identity and thus ethnicity. Self-categorisation is not enough. We can draw a parallel to a state. A state is also a relation, not a characteristic of the supposed state. It is not enough for PLO to proclaim a state (which is an act of self-categorisation). The state does not 'exist' before (at least some) other states have accepted its existence, meaning exo-categorisations are needed as a validation. We can see the development of the definitions in constructing ethnicity from a dialectic point of view. The thesis is the phase (which still seems to be the most common one) where the majority defines (and names) the ethnic identity and the ethnicity of the minorities, meaning only other-categorisations, exo-ethnonyms, are valid. The progressive antithesis, represented by both Allardt and Liebkind (but with different arguments) is that the minority itself defines or has the right to define (and name) itself, i.e. only self-categorisations, endo-ethnonyms, are valid. During both the thesis and the anti-thesis phase cognitive dissonance, conflict between endo-ethnonyms and exo-ethnonyms, is probable. Both treat ethnicity as a characteristic of the group to be characterised, and both concentrate, accordingly, on a study of the group or groups involved.

< previous page

page_92

next page >

< previous page

page_93

next page > Page 93

The synthesis which I try to develop here, sees the ethnonym, the name by which the group is characterised, as a signifier of the power relation between the groups. Cognitive consonance can be reached and the endo-and exoethnonyms agree only in a situation of balanced power between the parties. In this phase it is again, as in the study of integration, the negotiation process between the parties, the conditions for negotiation, and thus the power relations between them which should become the main object of study. Ethnicity has not commonly been studied in this way either. It has mostly been studied as a characteristic of the ethnic group, with the same implications as studying integration similarly. The human right to self-definition thus makes sense only when the parties are equal. If minorities are defined on the basis of power, not numbers, minorities are per definition not equal parties in the negotiation processes about 'their' ethnicities or 'their' degree of integration. In the same way as in discussing integration, we can also see the possibility here of the powerful party being able to prevent a migrant minority group from constructing itself as a national ethnic minority, by refusing to do 'its' (i.e. the majority's) part in the validation process. When a migrant minority group feels ready to integrate and wants to change its status from a migrant group to a national ethnic minority group, with the legal rights which this entails, the party to negotiate with about the validity of this minority ethnicity is the state. We can exemplify the possibilities of agreement/ disagreement as follows. If both the state and a migrant group itself sees the group as temporary migrants, there is no conflict. This was the situation of Italian migrants into Sweden in the 1950seven if most of them in fact have stayed. If both see the group as a national ethnic minority, there is no conflict either. In Sweden this is the situation for the indigenous Same minority. If the group itself sees itself as temporary whereas the state sees them as immigrants, there may be a conflictthis is the situation with some refugee groups in Sweden. And if a group wants to have a national ethnic minority status whereas the state sees it as an immigrant group, there is also a conflict. The options left open by the state to this type of a migrant minority group are thus either to assimilate or to remain immigrants for ever. This is the situation of Sweden Finns today (see Skutnabb-Kangas, 1987; in press). Several Swedish researchers study the Finnish group, and come up with judgements about its characteristics: to what extent is it integrated? To what extent do members see themselves as a minority? In contrast, other researchers, mostly minority members themselves, study the situation as a power conflict. The first group of researchers can, knowingly or unintentionally, come to function as parts of the repressive state apparatus, and can ideologically contribute to a support of racist ideologies, structures and practices, by defining integration and ethnicity as characteristics of the minority group,

< previous page

page_93

next page >

< previous page

page_94

next page > Page 94

characteristics, the lack of which is then used to deny the group access to certain resources which the majority as a group has and which a group defined as a national ethnic minority group might get better access to than a group defined as an immigrant group (not to speak of a guest worker group). The latter group may be able to participate in delegitimating racist ideologies and structures, by supporting the group when it tries to empower itself so that it can negotiate about integration and ethnicity from a position of the power which a theoretically sound analysis may give. Intercultural Education In the last section of the paper I will discuss intercultural education and the role of language in it, in order to show how the workings of linguicism are made easier by the way intercultural education has been conceptualised and developed, partly as a result of the type of concept construction discussed earlier. As Stacy Churchill (1986) has shown in his large OECD study, interculturalism belongs to measures used in a cultural handicap or cultural deficiency phase, where majority cultures are seen as the natural norms, the shared, universal mainstreams, and where minority cultures are seen as handicaps, preventing the acquisition of this universal mainstream culture. Even in later phases of interculturalism where it is accepted that the majority culture can be enriched to some extent by those aspects of minority cultures which the majority chooses to include as exotic and colourful, lacking competence in majority culture is still seen as a handicap in minorities (e.g. Skutnabb-Kangas, 1986). Interculturalism should denote a relation and exchange between cultures. Partly as a reflection of the reified and superficial nature of culture, seen as a characteristic of a group, and a reflection of the lack of studies of power relationships in constructing relational concepts, intercultural education implicitly assumes a balance of power between cultures in a naîve, liberalistic way, and fails to acknowledge the unbalance. In that way it contributes to the unequal access to power and resources in a more sophisticated way than earlier openly racist education based on physical segregation. I will discuss a couple of concrete examples of how intercultural education is conceptualised. The dominant groups in Western European countries have been forced to recognise the existence of minorities, indigenous and immigrant, in their midst. A substantial amount of educational activityand here 'multicultural' or 'intercultural' education are just now the most prominent examplesrepresent a response to the perceived needs of a more variegated clientele. But when such efforts are analysed, one can see that the celebration of ethnicity in intercultural education can (but need notsee the discussions of this in Jayasuriya, 1986; Skutnabb-Kangas, 1988) in fact function both

< previous page

page_94

next page >

< previous page

page_95

next page > Page 95

as a new more sophisticated type of control mechanism and as a pacifier, to divert attention from social and economic inequality (panem et circenses-type). An example of how these control mechanisms operate is a major Council of Europe project, 'The education and cultural development of migrants' (Council of Europe, 1986). The project report, summing up five years' work involving all Council of Europe countries (often represented by senior civil sevants), is methodologically explicit and attempts to make a solid bridge between multi-disciplinary theory and action. It is conceived as explicitly pursuing anti-racist goals. It recognises that European societies are increasingly multicultural and can benefit from this experience, and that each culture, including those of migrants, has its own specific features, which should be respected as such (ibid., 7). The report also recognises that contexts vary according to a large range of social, economic and historical factors, and that innovation and change are invariably resisted by the dominant group (ibid., 12). So far so good. Yet scant attention is paid to language in the report. The few references focus essentially on the language of the country of residence, and confine the mother tongue to the home (ibid., 51) and parents' associations (ibid., 52). In other words, despite lip-service to interactional interculturalism, there is no strategy for how an awareness of language issues and the securing of language rights for dominated groups can contribute to respect for other cultures. Bilingualism does not appear to have been considered, whether for majority or minority group representatives, and the everyday reality of the bilingualism of minority groups is ignored. The affirmation that interculturalism should contribute to 'the defence of human rights as agreed by all the member States of the Council of Europe' therefore smacks of empty liberal rhetoric. Although the document is well-meaning (a stand is taken against racism and the need for the dominant' group to change is accepted), it totally neglects one of the most fundamental issues in the schooling of migrants, namely both the instrumental role of the mother tongue for educational achievement and for high levels of bilingualism and biculturalism and the human right to identify with, learn fully and use one's mother tongue. The same report holds up Sweden as being the ideal country in relation to the pursuit of 'intercultural' education (ibid., 57). Sweden is, in fact, an excellent example of the most sophisticated forms of linguicism. The statements of goals in all official Swedish documents about minority education during the last decade without exception refer to the importance of the minority mother tongues, and proclaims active bilingualism as a goal, and intercultural education as the means to reach the goal. Even so, minority mother tongues are forced out of the pre-school and the school, in the sections on action, where most of the syllabus time and the funds are allocated to Swedish only. As Kenneth Hyltenstam (1986) puts it, the language-in-school debate in Sweden started with the harsh assimilation phase, went rapidly through an 'investment-in-minority-mother-tongues'

< previous page

page_95

next page >

< previous page

page_96

next page > Page 96

phase, and is now in the 'do-not-forget-Swedish' phase, which is again assimilationist, but now in a more covert and sophisticated way than earlier. Likewise, new Swedish plans for a 'language programme' for pre-schools and schools in Rinkeby, the Stockholm suburb with the highest percentage of foreign nationals in Scandinavia, have recently (December 1987) (Benckert et al., 1987) been proposed in a report which recommends massive teaching of Swedish in daycare centres and pre-schools. This is in conflict with the clearly articulated views of many parents. No proposals are made for improving mother tongue teaching, even though bilingualism is a declared goal. Large sums are spent on propaganda for Swedish, whereas the minority members of course have to do all the information work for the importance of the mother tongue teaching during their free time, without any financial support. As demonstrated elsewhere (Skutnabb-Kangas, 1988), research imperialism even in Sweden leads to ethnic Swedes travelling around the world as 'experts' on minorities in Sweden, while minority members themselves have great difficulties in getting research funds or publishing (as Hujanen & Peura, 1983, have shown in their review of Swedish migrant research). Our own definitions and descriptions of our own realities and wishes are invalidated. The same societal tendency, namely a new acceptance of assimilationist goals and linguicist ideologies, can also be seen in studies of the attitudes towards minorities, where representative samples of Swedes have been asked the same questions in 1969, 1981 and 1987. Between 1969 and 1981 there was a development towards less assimilationist demands. But now that has reversed. There are more Swedes now than in 1981 agreeing to the following two statements: Immigrants who intend to stay in Sweden should in their own interest try to become as much like the Swedes as possible (63% agree, of whom 35% completely and 28% with reservations), and All immigrant children in Sweden should from the beginning be taught that Swedish is their mother tongue (57% agree, 34% completely, 23% with reservations). Likewise, fewer now agree with the following statement: Society should create opportunities for those immigrants who so wish, to maintain their language and their cultural traditions (36% disagree, completely or with reservations, as compared to 27% in 1981). The figures come from Charles Westin's (1988) book, a bit ironically called Den toleranta opinionen (The Tolerant Opinion) written for the Diskrimine-ringsutredningen.

< previous page

page_96

next page >

< previous page

page_97

next page > Page 97

What, then, is intercultural about an education or a society where what minorities have is not allowed to be maintained, even for and by themselves? How can they trade something in an intercultural exchange, if they are not allowed to have it? Public opinion in Sweden with, relatively speaking, a supportive policy vis-à-vis immigrated minorities, is clearly way behind the ideals presented by the Council of Europe academics and bureaucrats in the report (1986) mentioned above. If even what these Eurocrats and Euroresearchers say can be analysed as ethnocentric and linguicist, despite their good intentions, interculturalism is still very much an empty phrase. To Conclude Policy is being reformulated constantly at the official level, with researchers playing one of the key roles. But it does not help much, with more progressive reports being formulatedand the newest Council of Europe report (1988) has moved a little bit forward since 1986if the gap between the nice formulations and social realities is widening and if our research does not reflect these realities of power conflicts. Assuming harmony and balanced power and therefore not studying power conflicts and conditions for negotiating relations, is one way of hiding the reproduction of unequal accessand thus legitimating it. Not studying power conflicts and condition for negotiating relations is a logical result of not seeing concepts (like integration or ethnicity) as relations but as characteristics which can be freely traded. It is liberal and 'objective', free from researcher bias, then, to say that the market decidesand on the market presumably only the best products survive. An ideology of the survival of the fittest ethnoses, languages and cultures thus fits well in our competitive world. So well, that it may be difficult for us to see the ideology in our own work. At a recent conference of OMEP (Organisation Mondiale de l'Education Prescolaire) on 'Intercultural Education' in Yugoslavia, attended by representatives of 15 nations, the following recommendations were included in the working group reports (Murdzeva-Skaric, 1987: 178). It is instructive for all of us to think over to what extent research that we are involved in lives up to the criteria specified in the recommendations: A precondition for majority group researchers to work in the interests of migrant minorities and escape from ethnocentricity is that they have first-hand experience, affectively and cognitively, of using the language and living in the culture of the minorities. A precondition for more adequate descriptive, evaluative and comparative work is clarification of key concepts (e.g. minority, intercultural, anti-racist) across languages and disciplines. The way minorities theorize about their own lives should be given due prominence. This will promote greater cross-cultural understanding between different

< previous page

page_97

next page >

< previous page

page_98

next page > Page 98

dominated groups, and between practitioners, bureaucrats and the scientific community. There is a need for research to be done for and by the minorities themselves ... There is a risk that the growing national and international management apparatus in migrant education will monopolise the available funding, and that the migrants themselves will not benefit ... The ethical responsibility of the researcher is primarily to the minority groups themselves, at each phase of the researchfrom identification of needs to interpretation of results. Underlying these concerns is a basic question about the nature of the scientific community of which we form part. When we know that researchers have played a key historical role in legitimating biological racism through the way they constructed scientific concepts, this awareness should present a challenge to us to critically examine our own present concepts too, so that we know who stands to benefit from our theoretical work in constructing tools for our own trade and from our empirical work in using these tools. Notes 1. Many of the thoughts in this paper have been developed together with Robert Phillipsonsee Phillipson 1986, 1988, in press, and our joint publications. References Allardt, Erik and Starck, Christian (1981) Språkgränser ock samhällsstruktur. Finlandssvenskarna i ett jämförande perspektiv. Stockholm: Almqvist & Wiksell. Australian Committee on Community Relations Final Report (Chairman: W.M. Lippman) (1975). Canberra: AGPS. Benckert, Susanne, Goldstein-Kyaga, Katrin and Persson, Solweig (1987) Gemensamt språkprogram för förskola-skola i Rinkeby. Stockholm: Stockholms skolförvaltning och Stockholms socialförvaltning. Bromley, Yu. V. (1984) Theoretical Ethnography. Moscow: Nauka Publishers. Calvet, Louis-Jean (1987) La guerre des langues et les politiques linguistiques. Paris: Payot. Churchill, Stacy (1986) The Education of Linguistic and Cultural Minorities in the OECD Countries. Clevedon: Multilingual Matters. Council of Europe (1986) The education and cultural development of migrants. Final report of CDCC project no. 7, DECS/EGT (86) 6. Strasbourg: Council for Cultural Co-operation, Council of Europe. Language learning in Europe. CDCC (88) 21/CC-GP12 (88) 7. Strasbourg: Council of Europe. Drobizheva, L. and Gouboglo, M. (1986) Definitions. Appendix to papers given by the authors at the symposium Multilingualism: Aspects of Interpersonal and Intergroup Communication in Plurilingual Societies, Brussels, 1315 March 1986. Edwards, John (1984a) Introduction. In Edwards (ed.) (1984c) op. cit., (pp. 116). (1984b) Language, Diversity and Identity. In Edwards (ed.) (1984c) op. cit., (pp. 277310). (ed.) (1984c) Linguistic Minorities. Policies and Pluralism. London: Academic Press. Hammer, Mitchell R. (1986) The influence of ethnic and attitude similarity on initial social penetration. In Kim (ed.) (1986) op. cit., (pp. 225237). Hujanen, Taisto and Peura, Markku (1983) Rapport om nordisk invandrarforskning. In Peura (ed.) (1983) op. cit., (pp. 148215).

< previous page

page_98

next page >

< previous page

page_99

next page > Page 99

Hvalkof, Sören and Aaby, Peter (1981) Is God an American? An Anthropological Perspective on the Missionary Work of the Summer Institute of Linguistics. Copenhagen: IWGIA. Hyltenstam, Kenneth (1986) Invandrarspråkenen ratad resurs? Stockholm: Forskningsrådsnämnden. Jayasuriya, D.L. (1986) Ethnic minorities and issues of social justice in contemporary Australian society. Keynote address at Australian Adult Education Conference 'Learning for Social Justice', Australian National University, Canberra, 79 December 1986. Kim, Young Yun (ed.) (1986) Interethnic Communication. Current Research. Newbury Park: Sage. Klinge, Matti (1983) Muinaisuutemme merivallat. Helsinki: Otava. Lange, Anders and Westin, Charles (1981) Etnisk diskriminering och social identitet. Stockholm: Liber Förlag. Liebkind, Karmela (1984) Minority Identity and Identification Processes: A Social Psychological Study. Commentationes Scientiarum Socialium 22, 1984. Helsinki: Societas Scientiarum Fennica. Miles, Robert (1989) Racism. London and New York: Routledge. Murdzeva-Skaric, Olga (ed.) (1987) Intercultural Education Interculturelle. Ohrid/Skopje: OMEP. Peura, Markku (ed.) (1983) Invandrarminoriteter och demokratisk forskning. Stockholm: Riksförbundet Finska Föreningar i Sverige. Peura, Markku and Skutnabb-Kangas, Tove (eds) (in press) Man kan väl vara tvåländare ocksåden sverigefinska minoritetens väg från tystnad till kamp. Stockholm: Carlssons Förlag. Phillipson, Robert (1986) English rules: A study of language pedagogy and imperalism. In Phillipson & SkutnabbKangas (1986) pp. 124343. (1988) Linguicism: Structures and ideologies in linguistic imperialism. In Skutnabb-Kangas & Cummins (eds) (1988), pp. 33958. (in press) Some items on the hidden agenda of second/foreign language acquisition. In Phillipson et al., in press. Phillipson, Robert, Kellerman, Eric, Selinker, Larry, Sharwood-Smith, Mike and Swain, Merrill (eds) (in press) Second/Foreign Language Pedagogy Research: A Commemorative Volume for Claus Faerch. Clevedon: Multilingual Matters. Phillipson, Robert and Skutnabb-Kangas, Tove (1986) Linguicism Rules in Education 3 volumes. Roskilde: Roskilde University Centre. Ralston, Helen (1988) The lived experience of South Asian immigrant women in Atlantic Canada. In New Frontiers in Social Research: Conference Session IV, The (Re)presentation of Ethnic Identity. CRES Publication Series, Occasional Paper No. 11. Amsterdam: CRES (Centre for Race and Ethnic Studies), Universiteit van Amsterdam. Skutnabb-Kangas, Tove (1986) Multilingualism and the education of minority children. In Phillipson & SkutnabbKangas (1986) op. cit., (pp. 4272). Also in Skutnabb-Kangas and Cummins (eds) (1988) op. cit., (pp. 1044). (1987) Are the Finns in Sweden an ethnic minorityFinnish parents talk about Finland and Sweden. Research project: The education of the Finnish minority in Sweden, Working Paper 1. Roskilde: Roskilde University Centre. (1988) Minority research between social technology and self-determination. Research Project: The Education of the Finnish minority in Sweden, Working Paper 2. Roskilde: Roskilde University Centre. (in press) Sverigefinnar förhandlar om etnisk identitet. In Peura and Skutnabb-Kangas (eds) (in press) op. cit. Skutnabb-Kangas, Tove and Cummins Jim (eds) (1988) Minority Education. From Shame to Struggle. Clevedon: Multilingual Matters. Skutnabb-Kangas, Tove and Leporanta-Morley, Pirkko (1987) Migrant women and education. Scandinavian Journal of Development Alternatives VII: 1, 1988: 83112. SOU 1989: 13 Mångfald mot enfald. Slutrapport från kommissionen mot rasism och främlings- fientlighet, Del I. Stockholm, Statens offentliga utredningar.

< previous page

page_99

next page >

< previous page

page_100

next page > Page 100

Vollmer, Ralf and Knödler, Uwe (1988) Nachruf auf Leonidas N. Proano. Pogrom. Zeitschrift für bedrohte Völker 144: 66. Westin, Charles (1988) Den toleranta opinionen. Inställningen till invandrare 1987. DEIFO-rapport 8. Stockholm: DEIFO. Zubrzycki, Jerzy (1988) Australia as a multicultural society. Siirtolaisuus/Migration 4: 916.

< previous page

page_100

next page >

< previous page

page_101

next page > Page 101

Grammatical Borrowing and Language Change: The Dutchification of Frisian Germen de Haan Research Institute for Language and Speech OTS, University of Utrecht, Trans 10, 3512 JK Utrecht, The Netherlands Abstract A scientific discipline is defined by a set of research objectives, and a set of methodological principles, the latter in part derived from a metatheoretical worldview. From this perspective the study of minority languages as a coherent domain of research hardly exists. It seems to be more accurate to say that minority languages are approached from (a combination of) traditional disciplines such as sociology, psychology and linguistics. What these approaches share, is, may be, no more than their research objects. This makes it unavoidable, in my opinion, to discuss the theme of 'comparative research and development of theories' from the perspective of one or more of the disciplines just mentioned. My starting point will be linguistics, and more specifically, generative grammar. From a grammatical point of view, minority languages have, as natural languages, no special properties. Nevertheless these languages are of particular interest for the grammarian; more specifically, for the grammarian who is interested in the dynamics of natural language. The speakers of minority languages find themselves in a situation which puts all kinds of pressure on the use of its minority mother tongue. Changes occur within a reasonably short period of time. Minority languages are therefore well qualified for studying factors that are involved in language change. In this paper I would like to address the very old problem of grammatical borrowing. I will discuss this problem within the context of minority language studies. Is it possible that the grammatical system of a minority language changes by borrowing grammatical elements and/or grammatical principles from the dominant language? It seems to be natural to take the Dutchification of Frisian as an example. Three cases of grammatical borrowing of Frisian from Dutch that have been proposed in the literature will be critically discussed: 1. the borrowing of verbal endings that is supposed to be responsible for transition of verbs from the je class to the e class; 2. the transition of the diminutive suffix ke to tsje;

< previous page

page_101

next page >

< previous page

page_102

next page > Page 102

3. changes in the word order of the Frisian verbal complex. I will conclude that the grammatical system of Frisian does not Dutchify. Furthermore this examination gives rise to some general conclusions concerning the ways in which minority languages can and cannot be influenced by dominant languages. Introduction Let me start by bringing up a point that most researchers in the field of minority languages should agree upon. The point is: there is no principled difference between the grammatical systems of minority languages and non-minority languages. This implies that the study of both types of languages is equally interesting from a grammatical point of view. In my opinion, the study of both minority languages and non-minority languages can contribute equally well to the goals of grammatical research. This general statement appears not to be so obvious if one is interested in the change of grammatical systems. Grammatical change begins if an adult modifies his grammar, or if a child acquires a grammar that differs from the grammar of the older generation. Grammatical change is completed if that modified grammar is accepted socially. In this perspective grammatical change is determined by the interaction of a range of synchronic factors. Alongside the innate language capacity we find production and perception strategies, and pragmatic, functional and sociological factors all contributing in one way or another to grammatical change. Traditionally the study of the change of grammatical systems belongs to the domain of historical linguistics. This field has considerably enlarged our knowledge of what changes took place, but less so why this has happened. This is not surprising. If one wants to account for why a particular change took place, one needs to have insight into all the synchronic factors that are potentially relevant. It is, of course, very difficult to reconstruct these synchronic factors for language stages of centuries ago. Consequently, it is not to be expected from historical linguistics to provide insight into the question why grammatical change occurs (see Weerman, 1989 for discussion of this point). It is at least conceivable that we can do better in this respect if we look into the situation of minority languages still in existence, and in particular those minority languages that appear to undergo grammatical restructuring. I believe that the study of existing minority languages that are in some sense unstable, is extremely interesting for the study of grammatical change. Frisian is an excellent example of such a minority language. In this century, the situation of language contact between Dutch and Frisian has changed dramatically from stable diglossia to unstable bilingualism. In their important monography Taal yn Fryslân, Gorter et al. (1984) show that there is a strong process of Dutchification going on in the province of

< previous page

page_102

next page >

< previous page

page_103

next page > Page 103

Fryslân. Traditionally, Frisian was the language of the lower and middle classes of the countryside, used mainly in informal domains. Nowadays, although the social stratification of the language situation is more or less maintained, the different language groups are completely mixed geographically. Moreover, Dutch is no longer the language of the more formal domains: it has intruded in domains such as family and neighbourhood to a considerable degree. Students of Frisian agree that 'Frisian is a minority language in Friesland (= Fryslân, GdH) due to its social position and prestige in comparison to Dutch' (Gorter et al., 1988: 25Gorter et al., 1988, is an English summary of Gorter et al., 1984). It is not surprising that this Dutchification process in the geographical and sociological sense is regarded as a big threat for the survival of the Frisian language. Nowadays nearly all native speakers of Frisian are raised as bilingualsnative speakers of both Frisian and Dutch. DutchFrisian bilinguals have different attitudes towards their mother tongues, and do not regard these languages as equivalent for use in domains such as home, school, business dealing, politics, science etc. It is not the case that there exists a variety of well-defined, non-overlapping domains for both Dutch and Frisian. This situation gives rise to language shift from Frisian to Dutch, Dutch being the dominant language. The social and cultural situation in Fryslân has changed in such a way that it not only leads to language shift, but it also affects the Frisian language itself. This is particularly clear in the domain of the lexicon. In order to fulfil the requirements of discourse in a modern society, Frisian needs words for new concepts. These needs are satisfied to a considerable extent by means of extensive lexical borrowing from Dutch. There are all kinds of other factors, which I will not go into here, that strengthen this borrowing process (see for example Sjölin, 1976 for discussion). In short, the bilingual situation in Fryslân is so intensive that words are introduced almost unconsciously into the Frisian lexicon. Lexical borrowing can take place with and without adaptation to the phonological and/or morphological structure of the borrowing language, in this case Frisian. Lexical borrowing with adaptation can lead to language change that does not affect the grammatical system. Lexical borrowing without adaptation raises the question whether such a phenomenon can give rise to systematic changes in the grammar of Frisian. Change in formal properties of grammatical systems is generally possible. So grammatical properties of Frisian can be affected by grammatical properties of Dutch. The central question that will be addressed here is whether this influence can have the specific form of grammatical borrowing. Does language A influence language B in such a way that grammatical elements and/or (part of) rules of grammatical system A are incorporated into grammatical system B? Is it the case that the grammar of Frisian borrows in this sense from the grammar of Dutch? Does the Frisian grammar gradually evolve into the grammar of Dutch? In other words, do we have a Dutch-

< previous page

page_103

next page >

< previous page

page_104

next page > Page 104

ification process of Frisian in the linguistic sense in addition to a Dutchification process in the geographical and sociological sense? Before we try to answer that question we have to discuss briefly the concept of grammatical borrowing. Grammatical borrowing is a specific kind of language change. Here I would like to point out that the term 'grammatical borrowing' is misleading as a description of a process of language change. As noted, language change is determined by quite a range of factors. Languages cannot change beyond the limits set by the innate language acquisition capacity of human beings. The set of possible language changes is defined, therefore, by Universal Grammar, the theory of this innate language acquisition capacity. Which changes actually occur depends on the interaction between Universal Grammar, the language data of the environment, production and perception strategies, and pragmatic, functional and sociological factors. The point I would like to make here is that the linguistic factor that determines change is not the grammatical system (or grammatical systems) of the environment, but it is the output of such a systemin other words, the unanalysed sentences of the environment. So if grammatical borrowing exists, it is not because grammatical principles of grammar B are borrowed directly from grammar A, but because the grammar B, constructed in part on the basis of the output of the grammar A, resembles in part the grammar A. In short, grammatical borrowing is not a process of change, but the result of such a process. Since grammatical systems do not belong to the factors that determine language change in any direct sense, grammatical borrowing as a result of change would be a rather unexpected outcome. Nevertheless, it seems to be possible, given appropriate conditions. Let me sketch some of these conditions. A grammatical system of the environment can influence a co-occurring grammatical system indirectly through its output. This influence through the output can result, for example, in the borrowing of specific items in the speech signal, if the items to be borrowed have hardly any systematic import. As a case in point, it is relatively easy to borrow lexical items due to their lack of systematic connections to the rest of the grammar. On the other hand, it is very unlikely that bound morphemes will be borrowed due to their systematic import. Rules of grammar are not directly reflected in the speech signal. In this case, borrowing can only be the result if the following two conditions are fulfilled simultaneously: one, there are such frequency differences between the contrasting output data that the borrowing language is obscured almost totally by the dominant language; two, the rules of grammar underlying such output data can be considered a kind of separate component of the grammar, that is, they can be reconstructed in isolation. Since these conditions will not be satisified easily, let alone at the same time, borrowing of grammatical rules is a very rare phenomenon, although not excluded in

< previous page

page_104

next page >

< previous page

page_105

next page > Page 105

principle (see Apel & Muysken, 1987, for a general discussion of grammatical borrowing). Against this background, it is rather remarkable that it is claimed recently that the grammar of Frisian Dutchifies in the sense that it borrows not only lexical items from Dutch, but also bound morphemes and even grammatical rules (compare Breuker, 1979; Breuker, 1982; Breuker, 1984; Eising, Koopmans and Wieberdink, 1981; for the view that grammatical borrowing is not easy, compare Feitsma, 1971). Here I would like to examine the three most important examples of alleged grammatical borrowing of Frisian from Dutch, in order to see whether this claim can be maintained. I will discuss here subsequently: the transition of verbs from the -je class to the -e class; the replacement in the diminutive system of -ke with tsje; inversion in the verbal complex. The Transition of Verbs of the -je Class to the -e Class Frisian verbs are divided into two classes: a class of weak verbs, and a class of strong verbs. The distinction between these two classes is that the conjugation of weak verbs is systematic, whereas that of the strong verbs is not. The class of weak verbs is subdivided into two categories: a subcategory that is characterised by an infinitive which ends in -e /@/; and a subcategory with infinitives ending in -je /j@/. Both subcategories have systematically different conjugations. We refer to these classes as -e verbs, and -je verbs, respectively. Examples that are illustrative for these classes are shown in (1). (1) The -e class Tense infinitive present tense 1 sing. ik 2 sing. do 3 sing. hij plural wij/jimme/sij past tense 1 sing. ik 2 sing. do 3 sing. hij plural wij/jimme/sij past participle imperative

< previous page

ending ending mean-@ 'mow' stap -@ 'step' mean-ø mean-st mean-t mean-@

stap -ø stap -st stap -t stap -t@

mean-d@ mean-d@st mean-d@ mean-d@n mean-d mean

stap -t@ stap -t@st stap -t@ stap -t@n stap -t stap

page_105

next page >

< previous page

page_106

next page > Page 106

Note that we do not use the written forms, but an approximation of the phonological structure. The difference between meane and stappe lies in the realisation of the past tense morpheme. It is argued that this morpheme is underlying /d@/ which undergoes assimilation to /t@/, see Tiersma (1985: 69). (2) The -je class Tense infinitive

ending klop -j@ 'knock'

present tense 1 sing. ik klop -j@ timmer -j@ 2 sing. do klop -@st

ending timmer -j@ 'carpenter'

3 sing.

hij

plural past tense 1 sing. 2 sing. 3 sing. plural past participle imperative

wij/jimme/sijklop -j@

timmer -@st timmer -st timmer -@t -t timmer -j@

ik klop -@ do klop -@st hij klop -@ wij/jimme/sijklop -@n klop -@

timmer -@ timmer -@st timmer -@ timmer -@n timmer -@

klop -@t

klop -j@

timmer -j@

It has been observed that some verbs that have originally endings of the -je class adopt new verbal endings (see, for example, Breuker, 1979; Eising, Koopmans & Wieberdink, 1981; most of the examples used below have been taken from this literature). It is suggested that this transition is a matter of borrowing bound morphemes from Dutch, since these new endings are claimed to be Dutch ('Hollandic') verbal endings. Let us see whether this claim can be maintained. Let us assume that klopje ('knock') is a verb that undergoes Dutchification. (3) compares the paradigm in (2) with its Dutch counterpart and the 'Dutchified' Frisian forms.

< previous page

page_106

next page >

< previous page

page_107

next page > Page 107

Observe with respect to the verbal endings that it is not true that 'Dutchified' Frisian conforms systematically to Dutch: the endings only match in the first person singular present and past, the third person singular present and past, and the imperative. So if one looks carefully, it is not at all obvious that Frisian borrows Dutch verbal inflection. If we compare 'Dutchified' Frisian with the paradigm of the -e class of Frisian, it can be demonstrated, convincingly to my mind, that we do not have a case of grammatical borrowing here (see (4)). (4) Tense

'Dutchified' Frisian klop -@

infinitive present tense 1 sing. klop -ø 2 sing. klop -st 3 sing. klop -t plural klop -@ past tense 1 sing. klop -t@ 2 sing. klop -t@st 3 sing. klop -t@ plural klop -t@n past klop -t participle imperative klop -ø

Frisian -e class stap -@ stap -ø stap -st stap -t stap -@ stap -t@ stap -t@st stap -t@ stap -t@n stap -t stap -ø

Note that there is a perfect match between the verbal endings. So what is going on here is not Dutchification in the grammatical sense, but a grammar-internal transition of verbs of the -je class to the -e class. There is, in my opinion, no evidence for the assumption that Dutch verbal endings undergo transfer to Frisian. This is not to deny that the change in Frisian may be partly due to the specific language contact situation. It is still possible that under the influence of Dutch the bilingual speaker assigns certain verbs to the 'wrong' verb class, a verb class that is closer to Dutch than the alternative class superficially. What is important is that these verbs receive the Frisian verbal endings; what changes is their lexical specification for verb class membership. One could argue that Dutchification is going onnot in the strict grammatical sense, but in the sense that a specific verb loses its membership of a verb class that distinguishes Frisian from Dutch. But this argument is clearly false, since this verb joins a verb class, the -e class, that is distinct from the Dutch counterpart. Moreover, the transition from members of the -je class to the -e class does not void the properties of the conjugation system that are characteristically Frisian. That is, it is not the case that the

< previous page

page_107

next page >

< previous page

page_108

next page > Page 108

-je class of verbs is emptied; on the contrary, this class, is productive in Modern Frisian. Diminutive Formation: -ke > -tsje A second example of alleged borrowing of bound morphemes in Frisian concerns diminutives (see in particular Breuker, 1982; Breuker, 1984). In Frisian diminutives are formed by adding -ke /k@/, -tsje /tsj@/ or -je /j@/ to the noun stem. The descriptions that can be found in traditional grammar can be translated into the rules under (5).

< previous page

page_108

next page >

< previous page

page_109

next page > Page 109

In the rules (5ab) I assume that Frisian has two underlying diminutive suffixes: -ke, and -tje. (5ce) give some surface variants of the latter. To the right of the slashes in these rules the relevant environment is specified. (5fh) represent some additional rules. (6) gives some sample derivations. There is a change in the diminutive system of Frisian, more specifically in the distribution of the -ke suffix. This change results in a diminutive system that is described by replacing (5ae) with (7af).

Rule (7b) presents a choice between suffixes. Nouns have to be specified for this choice. Some nouns select -ke; some -tsje; some both. This selection is subject to idiolectic variation. This change in the system appears to be triggered by Dutch: -ke is replaced gradually with -tsje in those contexts where Dutch has -tje. Again we refer to (7) as opposed to (5) as 'Dutchified' Frisian. Granted the influence of Dutch, do we have a case of grammatical borrowing here? In order to answer this question, we have to see whether Frisian borrows diminutive suffixes from Dutch, or (part of) a rule or rules of diminutive formation. This boils down to writing and comparing partial grammars of the different stages involved. In addition to (5) and (7) we therefore need a grammar of Dutch diminutive formation. According to Trommelen (1983), Dutch has one underlying suffix with five surface realisationssee (8).

There appears to be a development in Frisian to the effect that the diminutive suffix -ke is replaced in those environments where Dutch has -tje on the surface. In fact, it is suggested implicitly that the optional rule (7b) will altogether disappear from the Frisian grammar, and that (9) will be the final stage of that development:

< previous page

page_109

next page >

< previous page

page_110

next page > Page 110

In this final stage, we observe that the domain of the rule that spells out the suffix as tje is enlarged at the cost of ke. But note that this enlargement does not involve borrowing of a bound morpheme from Dutch, since the morpheme tje is already part of the Frisian system. Furthermore, the development of the rule system from (5) via (7) to (9) in no way shows incorporation of (part of) the Dutch diminutive rule system. So as far as I can see there is no evidence whatsoever that the transition from Frisian, rule system (5), to 'Dutchified' Frisian, rule system (7)/(9), can be described in terms of borrowing from Dutch. I think that there are reasons to believe that the development in the Frisian diminutive system is to a great extent an internal affair. If we look at rule system (5), we note that the conditions on the rules are not particularly phonologically systematic; compare, for example, the environment of (5a), and the splitting up of the natural class /r, l, n/ and /s, z, t, d/. In addition, variation of the suffix is not generally homorganically conditioned; this is only the case with respect to nouns ending in a /ng/, /l/ and /n/. The transition of (5) to (7)/(9) can be seen as a simplification of the Frisian rule system. The extension of the domain of the -tje suffix makes the environment for this suffix [+ sonorant]: vowels, /j, w/ leave the -ke class and enter into the -tje type with /n, l/. In general terms, the system becomes more homorganic, but only to the extent that is possible within the system, that is without borrowing of a bound morpheme. Significantly, there is one sonorant that does not undergo the change from ke to tje, and that is /m/. If one disallows borrowing of a bound morpheme, this is understandable, since in this case a homorganic change would give us the suffix -pje. Since Frisian does not have this suffix, and can not easily borrow it from Dutch, the suffix after /m/ remains -ke. In a framework that allows for unconstrained borrowing, this remains unexplained. If one allows for unconstrained borrowing, it is not possible to explain why other possible Dutchifications do not occur. For example, why is it not the case that Frisian borrows -tje after /s/ from Dutch, under extension of a t-deletion rule, realising surface forms such as hûs-je. Why is it not possible for Frisian to adopt from the Dutch diminutive system vowel length as a distinctive property. Many other possibilities suggest themselves. I would like to claim that they do not occur, since borrowing of rules and bound morphemes is hardly possible.

< previous page

page_110

next page >

< previous page

page_111

next page > Page 111

Word Order in the Verbal Complex: Inversion As to word order, Frisian shares an interesting property with other West Germanic languages: main verbs accompanied by 'auxiliary' verbs and/or modals occur in one cluster in sentence final position (see den Besten & Edmondson, 1981). This so-called verbal complex shows up in main and dependent clauses, be it that in main clauses the finite verb is not part of the verbal complex, but is moved to sentence second (or first) position. In (10) and (11) I have given some Frisian examples in which the verbs are in italics. dependent clause (10)a. dat er it dwaan kind hawwe soe 'that he it do be able have should' that he should have been able to do it b. dat et it dwaan wolle soe 'that he it do want' should that he should want to do it independent clause (11)(a)hij soe it dwaan kind hawwe 'he should it do be able have' b. hij soe it dwaan wolle 'he should it do want' Both Dutch and Frisian possess the verbal complex together with the distinction noted between main and embedded clauses. Other properties of the verbal complexes in both languages differ in important ways. First, the order of the verbs in the verbal complex in Dutch is the mirror image of the Frisian ordersee (12) and (13). dependent clause (12)a.dat hij het zou hebben kunnen doen 'that he it should have be able do' b.dat hij het zou willen doen 'that he it should want do' independent clause (13)a.hij zou het hebben kunnen doen 'he should it have be able do' b.hij zou het willen doen 'he should it want do' Note that Frisian order is similar to the order of the German verbal complex. Second, the order in the Frisian verbal complex is fixed, whereas Dutch allows for several inversion possibilities. If the verbal complex consists of

< previous page

page_111

next page >

< previous page

page_112

next page > Page 112

an infinitival main verb and a finite modal, then the verbs may invert, as in (14). (14) a. dat hij het b. dat er it 'that he it

wil doen *wol dwaan will do

/doen wil / dwaan wol / do will'

(Dutch) (Frisian)

Inversion is also possible if the verbal complex consists of a past participle of a main verb and an auxiliarysee (15) and (16). (15) a. dat hij het b. dat er it 'that he it (16) a. dat hij b. dat er 'that he

heeft gedaan *hat dien has done is gevallen *is fallen is fallen

/gedaan heeft / dien hat / done has' /gevallen is / fallen is / fallen is'

(Dutch) (Frisian) (Dutch) (Frisian)

Third: in Dutch, aspectual auxiliaries such as hebben ('have') and zijn ('be') assign past participle morphology to the verbs they govern, compare (15a) and (16a). If these verbs are governed themselves, then the assignment of past participle morphology is blocked apparently, since the past participle morphology of their governees is ''replaced' by infinitival morphologycompare (12a). This phenomenon, the Infinitival-Pro-Participle, does not occur in Frisiancompare (15b) and (16b) with (10a). It is claimed that the Dutch inversion possibilities gain influence in Frisian as the consequence of syntactic borrowing; (17) shows some examples (see Breuker, 1979; Eising, Koopmans & Wieberdink 1981; Breuker et al., 1982; Tiersma, 1985: 123): (17)a.dat it fan 't simmer moai waar wurde sil/sil wurde 'that it this summer nice weather become shall/shall become' b.oft er it hynder in spuit jaan kin/kin jaan 'whether he the horse an injection give can/can give' c.dat er in ferhaaltsje fertelle soe/soe fertelle 'that he a story tell should/should tell' d.dat de kapper myn hier goed knipt hat/hat knipt 'that the haircutter my hair well cut has/has cut' e.dat it him wol slagje soe/soe slagje 'that it him well succeed should/should succeed' To the right of the slashes, the 'Dutchified' order is found. At first sight the claim that syntactic borrowing is at stake is rather implausible. Recall that grammatical borrowing with syntactic borrowing as a special case, is very unlikely to occur. Furthermore, there is no internal reason why the Frisian verbal complex system should change. On the contrary, the system appears to be quite balanced. Let us see then whether this claim can be maintained.

< previous page

page_112

next page >

< previous page

page_113

next page > Page 113

Syntactic borrowing is borrowing of (part of) syntactic rules. I would like to stress that the claim of syntactic borrowing in the literature has been made without reference to explicitly formulated syntactic rules of both Frisian and Dutch so this claim is in fact not argued for. If we want to know whether the inversion properties of (17) are a case of syntactic borrowing, we have to construct a grammatical system underlying the output data involved and see whether the development in the grammatical system of Frisian can be described in terms of the Dutch system. If we want to formulate rules that account for inversion in the verbal complex, it is important to know how the verbal complex is to be derived. Let us assume for the sake of concreteness that the dialects of Continental West Germanic all involve the same restructuring process leading to the verbal complex. This restructuring process is schematically represented in

Following an idea originally due to M.A.C. Huybregts (see also Haegeman & van Riemsdijk, 1986), we assume a clause union rule that raises verbs into clusters, preserving the linear order of the verbs. Inversion rules are formulated as rules operating on the output of clause union. Differences between the languages involved are expressed in terms of these inversion rules. As for Dutch, inversion in the verbal complex is obligatory with the exception of finite modals and past participles. 'Finite model inversion' and 'past participle inversion' are optional, be it with different applicational conditions. Compare the examples in (19) and (20).

< previous page

page_113

next page >

< previous page

page_114

next page > Page 114

The difference in grammaticality between (19) and (20) indicates that 'finite modal inversion' is optional only if the verbal complex contains no more than two verbs, that is, the verbs involved in inversion do not branch. Otherwise inversion is obligatory. This restriction does not hold for 'past participle inversion', compare (21) and (22):

Some speakers are even more permissive in allowing inversion possibilities. Everybody accepts (22a) and (22e), i.e. past participles can occur to the right or left periphery of the verbal complex. These examples show that the optionality of 'past participle inversion' does not depend on the presence of no more than two verbs. That 'finite modal inversion' does indeed have a finiteness restriction, is shown by the contrast in (23) and (24).

'Past participle inversion' lacks this finiteness restrictionsee (25) and (26).

< previous page

page_114

next page >

< previous page

page_115

next page > Page 115

The properties of inversion in the Dutch verbal complex can be summarised by (27). (27) Inversion in Dutch (informally)

a. optional, if Vd is a finite modal, and Vb and Vd are non-branching; b. optional, if Vb is a past participle; c. obligatory elsewhere. As noted, the Dutch verbal complex is the mirror image of Frisian. So the order of the Frisian verbal complex corresponds directly to the output of clause union in (18). Frisian does not have any of the rules informally represented in (27). The question now is whether the inversion possibilities of 'Dutchified' Frisian are the consequence of a grammar that reflects the rules of the grammar of Dutch. If the answer is yes, we have a good case for grammatical borrowing. Since 'Dutchified' Frisian does not have obligatory inversion, we have to assume that rule (27c) does not belong to this system. It is also not the case that 'Dutchified' Frisian has an optional variant of (27) without the restrictions mentioned. Such a rule would provide for the following not attested variants of (28a).

Consequently, the question is whether we can derive the inversion possibilities of 'Dutchified' Frisian with the rules (27bc). The answer turns out to be negative. Rule (27a) undergenerates: it precludes the generation of the attested inversion variants of both the type (299b) and (30b), since there is no finite modal present:

< previous page

page_115

next page >

< previous page

page_116

next page > Page 116

Rule (27b) cannot derive the (b)-examples either, since there is no past participle involved. Rule (27b) overgenerates also: it allows the generation of non-attested inversion variants of the type (31b):

As far as I can see, it is not possible to account for the characteristic inversion properties of 'Dutchified' Frisian on the basis of syntactic borrowing from Dutch. In fact I have not been able to discover a grammatical system that underlies the properties in the verbal complex of 'Dutchified' Frisian. These properties appear to be a direct translation of the Dutch system into a Frisian vocabulary. The properties of 'Dutchified' Frisian look like the use of the grammar of Frisian interfering with the grammar of Dutch in production/perception of Frisian. The speaker/hearer uses the Dutch grammatical system during the production/perception of the verbal complex. There is an independent argument for this assumption. The morphological properties of the verbs involved in inversion in 'Dutchified' Frisian point to mixing between two independent systems. In the Dutchified order of (29b), the Infinitival-Pro-Participle is present. In (30b) another property of Dutch verbal morphology crops up. Let me explain this briefly: Frisian has two complementary infinitival forms. One with verbal ending -e /@/, and one with verbal ending -en /@n/. In Frisian te + infinitive, we find systematically the en infinitive. In Dutch there are two forms in free variation, -en, /@n/, and -e /@/. Observe now that in (30b), we find Dutch infinitival morphology. If these morphological properties of the verbs in 'Dutchified' Frisian would not be an example of mixing, but really belong to Frisian grammar, we would have to assume that the presence of the infinitival-pro-participle and the morphological realisation of the infinitive depend on the application of a verbal inversion rule. This kind of interaction between syntactic and morphological rules is without supporting evidence. I think it is fair to conclude that no case can be made for syntactic borrowing here.

< previous page

page_116

next page >

< previous page

page_117

next page > Page 117

Concluding Remarks As I stated earlier, the language contact situation between Dutch and Frisian changes from stable diglossia to unstable bilingualism. One of the consequences of this change is a Dutchification process in the geographical and sociological sense. However, in this paper I have tried to argue that there is no Dutchification of the Frisian language. The discussion of the three alleged processes of grammatical borrowing leads to the conclusion that these processes are not what they seem to be. There simply is no Dutchification in the grammatical sense going on, contrary to what is assumed sometimes. Since grammatical change is determined by surface forms, not by grammatical rules underlying these surface forms, lexical borrowing is easy. On second thoughts, what looks superficially like grammatical borrowing is often a grammar-internal development that gives rise to surface forms that correspond more or less to forms in languages of the environment. In conclusion, Dutchification of Fryslân may be a threat for the survival of Frisian, but Dutchification of Frisian in the grammatical sense is not, simply because it does not exist. Acknowledgement *I would like to thank Jacqueline Frijn, Jacqueline van Kampen, Geert Koefoed, Arie Sturm, and Mieke Trommelen for comments on a preliminary version of this paper and helpful discussion. The usual disclaimers apply. References Appel, R. and Muysken, P (1987) Language Contact and Bilingualism, London: Edward Arnold. Besten, H. den, and Edmondson, J.A. (1981) The verbal complex in Continental West Germanic. Groninger Arbeiten zur Germanistischen Linguistik 19, 1161. Breuker, P (1979) De takomst fan it Frysk: standerdisearje of fierder forhollânskje. Us Wurk 28, 5164. (1982) Ta de oergong -ke > -tsje yn it ferlytsingswurd. It Beaken 44, 8595. (1984) Oer it lienen fan bûne morfemen út it Hollânsk yn it Westerlauwerske Frysk. In N.R. Arhammer et al. (eds) Miscellanea Frisica; in nije bondel Fryske stúdzjes. Assen, 219228. Breuker, P. et al. (1982) Bydragen ta de didaktyk fan it Frysk yn it fuortset ûnderwiis. Grins/ Ljouwert. Eising, G., Koopmans, W. & Wieberdink, M. (1981) Een onderzoek naar interferentie in de vervoeging van het zwakke werkwoord en de volgorde in de bijzin in het Fries van Abbega. Unpublished Masters thesis, Groningen. Feitsma, A (1971) Onderzoek naar tweetaligheid in Friesland. In A. Feitsma & M. van Overbeke (eds) Tweetaligheidsproblemen. Bijdragen en Mededelingen der Dialectcommissie van de Koninklijke Academie van Wetenschappen 41, 521. Gorter, D. et al. (1984) Taal yn Fryslân; ûndersyk nei taalgedrach en taalhâlding yn Fryslân. Ljouwert: Fryske Academy. (1988) Language in Friesland. Ljouwert: Fryske Academy.

< previous page

page_117

next page >

< previous page

page_118

next page > Page 118

Haegeman, L. and Riemsdijk, H. vana. (1986) Verb projection raising, scope, and the typology of rules affecting verbs. Linguistic Inquiry 17, 41766. Sjölin, B. (1976) 'Min Frysk'. Een onderzoek naar het ontstaan van transfer en 'code-switching' in het gesproken Fries. Bijdragen en Mededelingen der Dialectcommissie van de Koninklijke Academie van Wetenschappen. Amsterdam: Noord-Hollandse Uitgerers Maatschappen. Tiersma, P.M. (1985) Frisian Reference Grammar. Dordrecht: Foris. Trommelen, M.T.G. (1983) The Syllable in Dutch. Dordrecht: Foris. Weerman, F.P. (1989) The V2 Conspiracy. A Synchronic and Diachronic Analysis. Dordrecht: Foris.

< previous page

page_118

next page >

< previous page

page_119

next page > Page 119

Comparative Analysis of Language Minorities: A Sociopolitical Framework A.B. Anderson Department of Sociology, University of Saskatchewan, Saskatoon, Canada Abstract Much research has been devoted to comparative analysis of language minorities from the perspective of sociolinguistics. This research has focused primarily on trends in language contacts, including diglossia, bi- and multi-lingualism, goegraphical and circumstantial patterns of language use, the birth of hybrid sublanguage forms (argots, dialects, creolisation, etc.), and minority language loss or decline. While language has long been recognised by scholars as a possible key component of ethnic identification, hence of nationalism or its minority form, ethnonationalism, relatively limited attention has been paid directly to the political context of minority language trends. This paper represents an attempt at synthesis of theoretical typologies in the fields of ethnic relations, studies in ethnonationalism, and sociolinguistics, developed by the author over the past decade (Anderson, 1978, 1979, 1980, 1981, 1983, 1986, 1987, 1989; Anderson & Frideres, 1981). Based on this earlier work, the present paper provides a sociopolitical framework, in the form of several revised typologies for analysing various types of ethnolinguistic minority situations, of alternative state policies toward minorities, of minority responses to state policies, and finally of considerations relating specifically to minority languages. Throughout this theoretical discussion, reference is made to particular minority situations in Europe, North America, and the Third World. Introduction During recent decades many indigenous (non-immigrant) ethnolinguistic minorities, as well as immigrant minorities, have revealed an increasing politicisation concomitant with a renewed awareness of their sub-national identity. Throughout the world may be found remarkable examples of the longevity of indigenous ethnolinguistic minorities and of persistence of minority languages. Similarly, although in a more limited period of time, many immigrant ethnic minorities have effectively resisted assimilation

< previous page

page_119

next page >

< previous page

page_120

next page > Page 120

andtogether with the native peopleshave recently become more active in their demands for state recognition of their cultures and languages. Following earlier work, a primary concern in this paper, then, will be to discern contrasts or similarities in the reaction of central governments toward the attempt of ethnolinguistic minorities to gain political autonomy (or possibly even complete independence), or at least to reinforce their ethnic identity through further recognition of their language. In other words, how do national, provincial, or municipal governments respond to what they have tended to view as the 'problem' of minority cultures and languages? 'Official' recognition of minority cultures and languages at the national, regional, or local levelsor lack of itmay be instructively compared in various countries. In Western Europe, despite the continuing prevalence of the 'nation state' concept, most countries retain clearly defined indigenous ethnolinguistic minorities: more than fifty cases in Western Europe (that is, excluding the additional heterogeneity introduced by millions of labour migrants and other immigrants) (Ashworth, 1977; de Vries, 1984; Heraud, 1966; Stephens, 1976). Continental Western Europe (i.e. excluding Britain, Ireland, and Iceland) consists of thirteen countries, excepting six very small states (Luxembourg, Andorra, San Marino, Monaco, Liechtenstein, and the Vatican City). It is a striking fact that all but one of these countries (Portugal) include indigenous (non-immigrant) ethnolinguistic minorities. The number of such minorities within each country varies, from six (France) to four (Italy, Switzerland), three (Austria, Belgium, Spain), two (Sweden, Denmark, Finland, West Germany), or just one (Norway, Netherlands). There is an even greater number of indigenous ethnolinguistic minorities in Eastern Europe and the Soviet Union. It can be convincingly argued not only that the ethnonationalism of diverse ethnic minorities within Yugoslavia continues to counterbalance the nationalism or sub-nationalism of the dominant Serbs, Croats, and Slovenes, if not the emergence of an effective pan-Yugoslavian nationalism, but also that the failure of arbitrary political boundaries to coincide with ethnic territoriality has greatly complicated Yugoslavia's relations with its neighbours (Anderson, 1989). International boundaries ideally mark the delimitation of the nation state. But few, if any, countries in the world today actually conform to a definition of the nation state. Most, in fact probably all, independent countries possess ethnic minorities which form sizeable proportions within the total population. Some are confederations of ethnic nation states bound together into an overarching federal framework, a sort of super-state. Others are preeminently multicultural conglomerations. In either of these latter two instances, the ethnic confederation or the multicultural conglomeration, no one ethnic group within the total population is completely dominant (Anderson, 1989). In assessing the relative success or failure of state minority policies, and conversely of minority ethnonationalistic movements, these two forces must

< previous page

page_120

next page >

< previous page

page_121

next page > Page 121

be counterbalanced: state policy, on the one hand, and minority politicisation, on the other (Anderson, 1986). It would seem germane to ponder whether a successful relationship between state and minority, in other words beneficial to both national state and the minority, may always be dependent upon an open, benevolent state policy as well as a high degree of minority infrastructural organisation enhancing a potent ethnonationalism. The latter factor would seem crucial: while there are some cases of liberal state policies combined with minority organisation or politicisation, even in cases where the state policy has been very negative toward a minority, if the minority response has been persistent and vigorous (perhaps short of terrorism), eventually significant changes in policy have been realised. Thus there are indications that very recently the markedly centralist, even assimilationist policies offor exampleFrance, Spain and Italy toward indigenous minorities have been changing. To repeat our central point: Just as political states have varied considerably in their minority policies, so too have ethnic minorities varied in their degree of organisational strength and politicisation. In appraising the current situation in each case, it is essential to bear in mind that state-minority relations are always in flux, never static. A truly enlightened minority policy would seriously weigh minority rights and accept the fundamental right of indigenous ethnic minorities to preserve and enhance their historic cultures within their own territories, better yet would assist these minorities to accomplish these aims (Anderson, 1986). Types of Minority Situations Let us now examine first the types of situations confronting 'indigenous' (i.e. non-immigrant) ethnolinguistic minorities, then those confronting immigrant minorities. Typology of Indigenous Minority Situations Seven basic types of indigenous minority situations may be distinguished. First, the language minority may be situated in its own compact 'homeland' territory, within a specific country, but not where minority status is created by the drawing of international frontiers. In Canada a good example of this situation is provided by northern native people and by treaty Indians, or by the 'Québecois' (French Canadians within Quebec), who have tended to regard themselves as 'maitres chez nous' (masters in their own home, i.e. in Quebec); French Canadian nationalists and separatists have stressed that Quebec is their nation-state, whereas Canada is at best an artificial political entity. One Western European example would be the Mouvement Occitanien in France; while in Eastern Europe the major ethnic ('national') groups

< previous page

page_121

next page >

< previous page

page_122

next page > Page 122

enjoy full republic status within ethnic confederations (i.e. states-within-states) both in Yugoslavia and the USSR. There are myriad examples of compact ethnic territoriality in developing countriesin some countries the minority claim is disputed by the national government (e.g. Tamil claims for Tamil Eelam within Sri Lanka, or Sikh claims for Khalistan within India); but in at least one case, South Africa, minority 'homelands' may be imposed against the wishes of the minority. Second, a language minority may culturally (but not necessarily politically) represent the linguistic majority in the neighbouring country; in this sense the ethnolinguistic frontier does not precisely coincide with the international boundary. In Canada the French population has 'spilled over' the borders of Quebec into the adjacent provinces of Ontario and New Brunswick (although it should be noted that most French in the latter province are of Acadian rather than Québecois origin) as well as into adjacent American states. This type of situation is very common in Europe: for example, the Flemish, German dialect, and perhaps Italian or Italo-Provencal-speaking populations in France respectively adjacent to Belgium, Germany, and Italy; or the Gallego (Portuguese)-speaking people in north-western Spain; Flemish (Dutch), Walloon (French), and German-speakers in Belgium; German, French, and Italian-Swiss; French-speaking Savoyards (Aostans) and Vaudois (Waldenses) as well as German-speaking South Tyrolians in Italy; French-speaking Channel Islanders (who are British citizens); the Swedish-speaking population in Finland; ethnic Germans (Volksdeutsche) in Poland, Czechoslovakia, and Hungary; Albanians in the Kosovo autonomous region of Yugoslavia; Hungarians in the Transylvanian region of Rumania; Bulgars in Yugoslavia, Greece, and Turkey; and so forth (Ashworth, 1977; Heraud, 1966; Stephens, 1976). It should be stressed that international borders seldom coincide with the true extent of a relevant ethnic population. On the one hand, ethnic nationals may appear to have 'spilled over' the border into neighbouring countries, or the boundary apparently had not been drawn accurately in the first place to coincide with ethnic frontiers. In either case, ethnic group members form a minority population in areas on the opposite side of the international boundary, so that minority nationalism can assume the form of irredentist claims for the 'reunification' of these co-ethnics across the border with the 'true' homeland. In other words, in the view of minority nationalists the border should be corrected to re-incorporate them back into the 'mother country'. This situation could be termed unidirectional irredentism. Examples abound: In Yugoslavia perhaps the best defined and politically the most controversial, has been the case of the Carinthian Slovenes, who have been variously estimated at between 20,00070,000 on the Austrian side of the border between Yugoslavia and Austria, or an even larger number, 60,000160,000 Slovenes on the Italian side of the border between Yugoslavia and Italy. The government of Yugoslavia has not pressed for rectification

< previous page

page_122

next page >

< previous page

page_123

next page > Page 123

of the border as it now stands, although in the past there was bitter controversy, closely linked to ethnonationalistic aspirations and rivalries, over exactly where the boundary should be drawn. There used to be, during the initial quarter century of Yugoslavia's existence as a unified state, far more Slovenes living west of the Italian-Yugoslavian frontier, however this border was moved westward in 1945 so that most Slovenes in Italy were reincorporated into Yugoslavia. And results of a plebiscite conducted in 1920 (still hotly disputed by Slovenian nationalists) confirmed that in alpine southern Carinthia Austrians outnumbered Slovenians, with the result that this region remained in Austria (although certain areas and rural communities here are almost completely Slovenian) (Anderson, 1989). On the other hand, ethnic populations of a neighbouring country may form a minority on the opposite (e.g. Yugoslavian) side of the frontier, in which case their minority nationalism may press for rectification of the border, at the expense of national (Yugoslavian) territory (Anderson, 1989). Again, examples abound; particularly in wartime Europe when numerous boundary changes occurred. It seems obvious that reverse irridentism, confiscation of territory, has been politically most problematic; the continuing political fragility of some of these territories, cannot be overstated, such as Transylvania (currently in Rumania, but with substantial Hungarian and German minorities, and formerly in Hungary). Third, the situation of ethnic minorities across international frontiers becomes still more complicated when there are complimentary minorities in both directions at the same point across the boundary, i.e. when the predominant population of one country constitute a minority across the border in a neighbouring country, and vice versa. Thus irredentist claims would tend to cancel each other out, assuming that the minorities on either side of the frontier are of roughly equal strength. These minorities are in a unique position to press for state recognition of their cultural rights and even regional autonomy, on the basis of a 'trade off': one country, for example Yugoslavia, agrees to protect a minority, for example Italians, in exchange for the neighbouring country, in this case Italy, similarly protecting its Yugoslav (Slovene) minority. Yet even the most conservative estimates suggest that now there are far more Slovenes on the Italian side of the border than Italians on the Yugoslavian side, thereby creating a power differential in the bargaining process. A similar predicament probably exists across the Austrian border: Austrians claiming Slovenian origin far outnumber even the total estimated population claiming German ethnicity within all of Yugoslavia (Anderson, 1989). Other examples of this bipartite minority reciprocity abound: in Eastern Europe, Hungarians in Slovakia yet Slovaks in Hungary; in Western Europe, speakers of Italian or Italo-Provencal dialects in the southeasternmost area of France (within Italy till after World War II) yet speakers of French or Franco-Provencal dialects across the border in Italian Alpine valleys (Val d'Aoste and the Waldensian Valleys);

< previous page

page_123

next page >

< previous page

page_124

next page > Page 124

or Danish-speakers in West Germany yet a German minority in the adjacent border area of Denmark; in North America, French in Ontario yet Anglo-phones in Quebec. Other trans-border situations become even more complicated. For example, there are no less than six points along the Yugoslavian frontier where extremely, complex minority situations straddle the borders of not just two but three countries, in which case we have tripartite minority reciprocity (Anderson, 1989). Where Yugoslavia, Italy, and Austria meet, just to cite one example, on the Yugoslavian side of the frontiers Slovenian is the official language, while Italian is given limited recognition, and in a couple of communities closest to the Austrian border German may also be heard but is not given official recognition. The region across the border in Italy is the Val Canale (German das Kanaltal, Slovenian Kanalska Dolina, Friulian Chianal de Fele); here Italian is the official national language, Friulian the official language of the autonomous province of Friuli-Venetia Giulia, and both Slovenian and German are given limited local recognition. On the Austrian side of the frontier, German is the only official language, but the Slovenian minority is 'protected' and given limited right to use their language in the school system, mass media, and various institutions (Anderson, 1989). Fourth, a situation of imposed minority internationality is formed when a language minority is indigenous to a specific region yet divided between two or more states. For example, the native Amerindian peoples are divided between many countries constituting North, Central and South America, while the Inuit (Eskimo) people are divided between Alaska, Canada, Greenland, and the USSR. European examples might include the Frisians in the Netherlands and West Germany; the Sami (Lapps) in Norway, Sweden, Finland, and the USSR; the Basque and Catalan-speaking people divided between Spain and France. A classic case of imposed minority internationality is found on the Yugoslavian frontier: Greater Macedonia is divided between Yugoslavia (about 39%), Greece (51%), and Bulgaria (10%). The ethnic demography of Greater Macedonia has been greatly complicated by frequent largescale transfers of ethnic populations across changing international boundaries (Anderson, 1989). Cases of imposed minority internationality are particularly common in Africa, where colonial boundaries often were imposed upon ethnic territoriality. Thus the Sotho and Tswana straddle the South Africa-Botswana border; the Ndemba the borders of Angola, Zambia, and Zaire; the Fulani are found in at least half a dozen West African countries, the Tuareg and Berbers in even more North and West African states; the Somali is not only Somalia but also neighbouring Ethiopia, Djibouti, and Kenya; and so forth. Fifth, certain indigenous language minorities might be widely scattered, such as Armenians, Gypsies, Jews, Vlachs (speaking a Rumanian dialect), Pomaks (Moslemized Bulgars), Turks and Tatars in the Balkan countries,

< previous page

page_124

next page >

< previous page

page_125

next page > Page 125

the Rhaeto-Romansch (Ladin) dialect groupings in Switzerland and Italy, etc. Such minorities usually constitute relatively small proportions except in very localised situations, although collectively their numbers could add up into millions. Sixth, a special case might be made for interrelated language minorities enjoying ethnic and linguistic revival within separate countries. The 'Celtic Revival' is a good case in point, involving limited survival of the Scottish Gaelic language in Scotland (Alba) and perhaps Canada, revival of Irish Gaelic as an official national language in Ireland (Eire), occasional use of now virtually dead Manx in the Isle of Man (Mannin) and Cornish in Cornwall (Kernow), and nationalistic revival of Breton in French Brittany (Breizh) as well as Welsh in Wales (Cymru). Another example might be the teaching of modern Hebrew, the national language of Israel, in Jewish schools in many countries. Seventh, the most complicated situation is found when language minorities exist within language minorities. In a sense, the English in Quebec are in this predicament: they are a disadvantaged and thoroughly outnumbered minority in a predominantly French-speaking province (whose government insists on a policy of French language primacy), which in turn is situated within a predominantly English-speaking country (Canada). A more specific Canadian case might be, for example, a Volga German Baptist settlement within a largely Ukrainian Orthodox and Catholic region in the western province of Saskatchewan. Some intriguing examples of this degree of complexity are still found in Europe: The people inhabiting the upper portion of the Val Gressoney in the Val d'Aoste region in north-western Italy speak a German dialect, although this region is an autonomous French-speaking province within an Italian-speaking country. Again, the people in several valleys of the Dolomites speak Ladin dialects; some valleys are situated in the Province of Bozen (Italian Bolzano), or South Tyrol (German Sud Tirol), where German is recognised as an official language, within the autonomous region of Trentino-Alto Adige in Italy. In this respect Yugoslavia is particularly complex: The principal national groups of Yugoslavia, the charter members of the Yugoslavian confederation, may constitute minorities outside of their own republics, which are virtual nation states within the confederation as a whole, but inside neighbouring republics within Yugoslavia, thereby further increasing the ethnic complexity of Yugoslavia. In a sense, then, these are official, protected minorities possessing indigenous minority status. In turn, Serbia contains two self-governing autonomous regions: Kosovo, where most of Yugoslavia's Albanian Moslem population is concentrated, and Voivodina, where Hungarians are the largest minority, but where significant Rumanian, German, Czech, Slovak, Ruthenian, Ukrainian, and other minorities coexist with both Serbs and Croats. Other locally significant minorities within Yugoslavia are Italians, Turks, Bulgars, Greeks, Vlachs, Jews, and Gypsies (Anderson, 1989).

< previous page

page_125

next page >

< previous page

page_126

next page > Page 126

Each official national group in Yugoslavia is itself a minority within the total population, so Serbs in Croatia, for example, are essentially a minority within a minority. So, too, are Serbs in Kosovo. This latter case is intriguing, as Kosovo technically remains an autonomous region of Serbia, rather than being elevated to the status of a full republic. In other words, these particular Serbs actually find themselves a disadvantaged minority within their own republic (Serbia) because Kosovo is an Albanian Moslem state within a state (Anderson, 1989). Other similar examples could be cited. Immigrant Minorities The rapid increase in the proportion which non-indigenous minority populations form within total national populations, might seem more problematic. Yet again, immigrant ethnic minorities can no longerif everbe considered simply temporary sojourners, a transient exploitable labour market profitable to largescale capitalist economies. The rising tide of racism in West European countries does not bode well for the future of immigrant minorities (Anderson, 1986). It has never been more clear that the increasingly anachronistic ideology of the nation state must be superseded by a more rational contemporary policy constructively recognising the vast ethnocultural diversity of these countries and encompassing programmes of assistance to ethnic minoritiesmulticultural education and language proficiency programmes, anti-defamation, housing projects, employment programmes and skills upgrading, immigration and particularly refugee admission and settlement (Anderson, 1986). Among immigrant minorities, we may also distinguish between seven types: First, what could be called 'imperial relics', the demographic remnants of imperial settlement policies representing ethnic kinship with the colonising power, yet remaining behind after independence. Of course, this is most obvious in the case of European settlers in independent African countries, such as French in Senegal, Italians in Eritrea, English in Zimbabwe, Portuguese in Angola. Yet this is not purely a Third World phenomenon. In Europe, Yugoslavian unity has been further complicated by the prevalence of smaller, widely scattered ethnic minorities which are the remnants of settlement policies introduced by the Austrian and Turkish empires which long controlled most of the area of presentday Yugoslavia: these include ethnic Germans (Volksdeutsche), Turks and other Moslems. Second, imperial or colonial settlement policies might have created ethnic settlements not representative of the colonisers, but of the colonising powers' desire to import farmers or a rural labour force. The migration of Slavic minorities which were not South Slavs (i.e. Yugoslavs) into regions now comprising Yugoslavia also resulted from the imperial era. Particularly from the early eighteenth century until after the First World War, Czechs,

< previous page

page_126

next page >

< previous page

page_127

next page > Page 127

Slovaks, and Ruthenians migrated from other regions within the Austro-Hungarian Empire to resettle in Voivodina. Third, alternatively 'middleman minorities' have found a commercial niche in unnumerable cities, both in developed and less developed countries, such as Chinese in Malaysia or North America, Vietnamese in Paris, Syrians-Lebanese in West Africa or Trinidad, etc. Fourth, as mentioned above, cheap migrant labour may be imported to industrialised states to supplement a labour shortage and thereby maintain capitalist industrial expansion. The estimated 12 million 'gastarbeiter' (literally: 'guest workers') in north-western Europe provide a striking example. Fifth, vast numbers of refugees constitute a primarily urban immigration which effectively bypasses regular immigration procedures. They are found everywhere: Tamil Sri Lankans in West Germany, Sikhs in Canada, Vietnamese Boat People. Earlier refugees in Paris would include Eastern European Poles, Ukrainians, Russians, and Jews. Sixth, colonial migrants may arrive in former colonial countries (metropoles) from one-time, or remaining, colonies: Thus the French population includes francophone West, North and Central Africans, Indochinese, Lebanese, Pacific Islanders, Réunionais, Haitians and other Caribbean migrants, and French Guyanese; the British population now includes large numbers of East Indians and West Indians from former British colonies; the Dutch population Moluccans, Surinamese, and Dutch Antillean migrants. Seventh, finally there are, of course, the millions of 'permanent' immigrants who come primarily to developed countries from former or extant relatively less developed countries, or simply in search of 'new opportunities': Mediterranean migrants into France, Europeans and migrants from many Third World countries to Canada and the USA, both European and Asian migrants to Brazil, Argentina and other South American countries. Alternative State Policies Alternative state policies affecting ethnolinguistic minorities could be placed on a continuum ranging from negative through conservative and moderate to liberal treatment of language minorities (Anderson & Frideres, 1981; cf. Ashworth, 1977; Hunt & Walker, 1974; Stephan, 1987). Doubtless the most negative policy is for a state to eradicate a minority. This has been done by means of outright physical extermination (genocide) (exemplified by Hitler's 'Final Solution of the Jewish (and Gypsy) Question'). A minority may also be expelled. In fact, Turkey both exterminated and expelled its Armenians. At least fifteen million ethnic Germans (Volksdeutsche) were expelled or obliged to retreat from Eastern European countries, particularly Poland and Czechoslovakia, between 1944 and 1947. Many European countries have employed such policies in the past, and perhaps still do in the

< previous page

page_127

next page >

< previous page

page_128

next page > Page 128

sense of refusing would-be immigrants 'landed' or resident status (except on relatively short-term work contracts), implementing a 'politique de retour' (e.g. France, Switzerland), or refusing immigrants and refugees entry into the country. Canada has by no means been immune to this sort of policy: witness the extermination of the Beothuk Indians in Newfoundland (when that province was a British colony), or the Chinese head taxes at the turn of the century, or the Oriental exclusion act aimed at keeping out Chinese and Japanese (19231947), or the deportation and imprisonment of Japanese Canadians (many of them Canadian citizens) during and even following the Second World War. It should be added that another, perhaps less negative way in which a country could rid itself of a minority is to exchange this minority (counter-ethnics, i.e. different from the majority) for co-ethnics (i.e. people of the same ethnicity or language as the majority) who were themselves a minority in another (usually neighbouring) country. This was done most extensively, for example, between several Balkan countries before, during and after the First World War, and again during the Second World War. A second basic type of negative policy toward ethnic or language minorities is assimilation, in which case the state still attempts to get rid of a problematic minority, yet not by deportation or genocide, rather by means of an explicit policy of forced assimilation. The minority is not permitted to maintain any cultural distinctiveness apart from the majority, therefore no allowance is made for use of a minority language in schools. Such a policy, in the form of Angloconformity, prevailed for all ethnolinguistic minorities in Canada, with the possible exception of French Canadians within (but certainly not outside) Quebec and of native peoples, at least until the 1930s (and in some cases longer), although it was resisted in varying degrees by the minorities. Examples of assimilatory policies are still readily found in Eastern Europe: Sorbs in East Germany, ethnic German minorities remaining in Eastern Europe, Macedonians in Greece and Bulgaria, etc. Until recently, assimilation of language minorities was the prevalent policy in France, Spain and Italy. A number of policies may be termed conservative, yet are clearly not as negative as those which we have just discussed in so far as they are not necessarily designed to get rid of a minority (through genocide, deportation, or forced assimilation). Thus, in countries still exhibiting fairly strong centralism, assimilation may not be as explicit as it is implicit as a desired goal. Minority languages may be recognised yet restricted to private schools in lieu of public financial support. In fact, the national state may simply choose to largely ignore the minority. In moving progressively away from Anglo-conformity, some Canadian minorities have passed through this stage. Until quite recently certain European countries pursued this sort of policy: for example, German-speakers in Belgium, some of the language minorities in Austria, Lapps in northern Scandinavia, and still perhaps Frisians (i.e. defined as people of Frisian origin) in West Germany.

< previous page

page_128

next page >

< previous page

page_129

next page > Page 129

More moderate policies are found in states exhibiting limited or weak centralism. If the national government could be termed 'somewhat centralist', perhaps it would have abandoned assimilation as a general policy (although assimilation might still be advocated by conservative circles); and there may be evidence of limited public (i.e. government) support of minority languages in selected public schools (de Vries, 1987). Alternatively, if the government clearly exhibits weak centralism, there may be fairly generous encouragement (including financial support) of minority languages in schools, etc., and possibly an official national policy of ethnic pluralism (however defined), yet no (or very limited) recognition of minority territorial rights, i.e. regional autonomy. Quebec has been guided in recent years by provincial governments (particularly by the avowedly separatist Parti Québecois (19761985) toward increasingly conservative language policies: restricting funding to English-language universities, requiring French to be the dominant language in shop signs, advertisements, and industrial companies, and forcing new immigrants to send their children to French schools. In Europe, perhaps the situation of the German minority in Denmark and conversely the Danish one in Germany is representative of the 'limited centralism' type, while the current situation of the Slavic minorities in Austria, the Romansch in Switzerland, and the Lapps in Scandinavia and Finland represent the 'weak centralism' type. Rumania, while somewhat pluralistic, currently falls well short of recognising full regional autonomy for its language minorities; in fact the Hungarian minority in Transylvania claim discrimination verging on cultural genocide, and an increasing number are fleeing to Hungary. Both the policies of regional autonomy and of ethnic federalism are obviously quite liberal treatment of language minorities. The former policy (regional autonomy) implies explicit recognition of minority territorial rights: thus the minority forms a 'state within a state' and presumably is allowed to use its own language 'officially' within this territory (Hall, 1979; Jakobson, 1972). While not quite being granted this status by the national government of Canada (except in certain highly controversial clauses in the 'Meech Lake Agreement'), in effect the province of Quebec is regarded by its French population and the provincial government as a virtual 'state within a state'. On a far smaller, more localised scale, Canadian history has seen some development of such ethnic autonomy, very temporarily in the one-time creation of German Mennonite and Russian Doukhobor reserves and even a self-proclaimed Icelandic republic in Western Canada, more lastingly in the continuing system of Indian (and in a couple of provinces, Métis) reserves (currently being re-enforced by the success of native land claims). In Europe, limited regional autonomy is now enjoyed by Basques and Catalans in Spain (but not yet by their co-ethnics or other minorities in France), by French and German minorities in Italy, Frisians in the Netherlands, Swedes in Finland's Aland Islands, as well as by Greenlanders; while more far-reaching autonomy is enshrined in the Soviet constitution (the

< previous page

page_129

next page >

< previous page

page_130

next page > Page 130

Union of Soviet Socialist Republics consists of the Russian Soviet Federated Socialist Republic and fourteen member Soviet Socialist Republics, which in turn may include Autonomous Soviet Socialist Republics, Autonomous Oblasts, and Nationality Okrugs with varying degrees of autonomy along ethnic lines). Ethnic federalism may be defined as the situation when the national state as a whole is viewed essentially as a partnership between ethnolinguistic minorities; therefore, the principle minority languages are fully recognised, probably throughout the entire country. Not without some reservations, we could consider the English-French bilingualism and biculturalism (since 1971 redefined as multiculturalism) of Canada to be an illustration of this policy; yet better examples might be found in Switzerland, Belgium, Yugoslavia, the Soviet Union, and possibly Luxembourg (Deutsch, 1968, 1975; Mackey, 1967). Some qualifications merit mention (Anderson, 1989): the extent to which ethnic minorities in heterogeneous situations may have their rights respected, in other words the degree to which they may enjoy relative privilege or disadvantage, may be closely related to their relative demographic strength in the local region or even specific community, more than to their overall official status in an ethnic federation as a whole. State treatment of ethnic minorities does in fact change over the years, and may have ranged from extreme persecution, evidenced in eradication of the troublesome minority through physical and cultural genocide, deportation, or forced assimilation, to more moderate treatment, evidenced in token multicultural policies and limited concessions, to complete acceptance, exemplified first in official nationality designation, then regional autonomy (although in Yugoslavia the status of Kosovo, for example, may still be questionable to Albanian Moslem dissidents), and ultimately full republican status within an ethnic federation. Alternatively, in his Typology of Multi-Ethnic Societies, Marger (1985) has distinguished between three basic types: Colonial, Corporate Pluralistic, and Assimilationist, and has concisely described the salient features of each type. Minority Responses to State Policies It is important to emphasise that state policies may be only attempts to control ethnic minorities. The relative success or failure of these policies depend upon minority acquiescence, organisation, and politicisation. Minority responses range along a continuum from passive to active. Some minorities have seemed to be rather submissive or disorganised (e.g. Gypsies, certain German minorities, the Gallego-speakers of Spain, and numerous smaller, isolated groups such as Ladins or perhapscurrently but certainly not historicallyWaldenses in the Italian Alps). Other minorities, while somewhat docile, are better organized and may reveal latent minority nationalism

< previous page

page_130

next page >

< previous page

page_131

next page > Page 131

(e.g. Flemish-speakers, Alsatians, and the Mouvement Occitanien in France, and perhaps the majority of the Scottish and Welsh populations). Still others are both well-organised, actively mobilised, and may reveal overt or manifest nationalism (e.g. Catalans in France and particularly Spain, French Aostan and German Sud-Tiroler minorities in Italy, Flemings and Walloons in Belgium, Albanians in Yugoslavia, Franco-Québecois and increasingly native peoples in Canada) (Connor, 1976; Deutsch, 1968, 1975; Hall, 1979; Williams, 1980, 1982; Wood, 1981). Lastly, violent terrorism stands as the 'end of the line' in this continuum of minority reactions, utilised by such divergent radical nationalist movements as the Front de Liberation Québecois, IMRO in 'Greater Macedonia', ETA among Spanish and French Basques, the Breton Liberation Front, the Front National de Libération Corsicain, the Liberation Tigers of Tamil Eelam, the Sikh Dal Khalsa, and so forth (Connor, 1976; Deutsch, 1968, 1975; Esman, 1977). Language Typologies Three theoretical frameworks relating respectively to first, language policies, second, minority language use, and third, standardisation vs. localisation, would seem useful. First, countries may be classified as officially unilingual (monoglot), bilingual, or multilingual (N. Anderson, 1969; Lieberson & O'Connor, 1975; Lieberson et al., 1975; Mackey, 1967; Simon, 1969). However, it is essential to consider the level of official (i.e. government) recognition. For example, in Eastern Europe (possibly excluding the USSR), only Yugoslavia is clearly multilingual in an official sense. In Western Europe only Switzerland is officially multilingual as a whole. In Luxembourg, German and French are official while Letzeburgesch is the prevalent local dialect. Belgium is officially bilingual (Walloon/Flemish) yet grants some regional recognition to its German-speaking minority; similarly Finland is officially bilingual (Finnish/Swedish) yet grants some recognition to its Lapps. Several countries are officially unilingual as a whole, while they recognise bilingualism officially at an intra-national, regional level, such as in the Netherlands (Frisian/ Dutch in Friesland), Italy (French/Italian in Valle d'Aosta and German/ Italian in AltoAdige), and Spain (Basque/Spanish in the Basque Provinces and Catalan/Spanish in Catalonia). Other countries, officially unilingual at the national level, grant limited concessions to ethnolinguistic minorities short of regional bilingualism. Sri Lanka was declared a unilingual state in its 'Sinhala only' policy early in the post-colonial period; however it has been obliged to retreat to recognising English as a neutral de facto lingua franca (as has India), while recently granting concessions to the Tamil-speaking minority. Second, there could be considerable variability of language use and attitudes towards ethnic identification. Individuals may find it expedient to speak a particular language, or identify with a specific ethnicity, in one

< previous page

page_131

next page >

< previous page

page_132

next page > Page 132

circumstance, such as with family members in the privacy of the home, but identify differently and speak a different language in another setting, such as in the workplace or the local community, in a larger town or city, etc., Therefore frontier, and not infrequently internal, minority areas are characterised by functional diglossia. Moreover, it goes without saying that there can be considerable complexity in the degree of using minority vs. national languages in these various settings. For example, a francophone in Saskatchewan might talk only French at home with elderly parents, some French mixed with some English with his wife, mostly English to his children, who nonetheless go to a Frenchlanguage school but play with an ethnically mixed peer group in English; in the community French is spoken among the French-origin people, but English with people of other origins; English is the usual language at work and beyond the community. Third, a contrast should be drawn between 'standardised' and 'localised' languages: which type is used, in what circumstances, and officially recognised at which level? Two European cases illustrate this sort of complexity: In the Val Gressoney/Lystal, the easternmost alpine valley in the Province of Valle d'Aosta (French: Val d'Aoste), Italy, three communes are mainly ethnic German: Gressoney-la-Trinité (Greschonei-Dreifaltigheit in Walser German dialect), Gressoney-St-Jean (Greschonei-Sent Johann), and Issime (Eischime), and three are mainly ethnic French: Gaby, Lillianes, and Pont-St-Martin. In the 'German' communes, dating back to a 12th century migration of Walser people southward and eastward out of the Swiss Valais/ Wallis, a unique archaic Walser dialect ('Walser-Titsch') is spoken in many homes, whereas textbooks for German language classes are in standardised German. Their neighbours down the valley in the 'French' communes speak a Franco-Provencal dialect, whereas standardised French is the official language of the autonomous province. The ethnic Italians who migrate up into these valleys from the south tend to speak the Piemontese dialect, whereas standard Italian is the national language of citizens of Italy, whatever their ethnicity, taught as the principal language in the school system. In sum, a Walser child growing up in one of the 'German' communes might expect to become reasonably well acquainted with virtually six languages! For a second example, let us return to the setting for this conference: Friesland. The Frisian population, if defined as people of ethnic Frisian descent, though not necessarily Frisian-speaking, is found in three distinct regions along the southern North Sea Coast: West Friesland (W. Frisian: West Fryslân or Westerlauwersk Fryslân), consisting of the West Frisian Islands and adjacent mainland in the Netherlands, all of which territory is situated within the Province of Friesland except the island of Texel/Teksel and the north tip of mainland Noord-Holland (W. Frisian: Noard-Hollân); East Friesland (German: Ost Friesland, W. Frisian: East Fryslân), the East Frisian Islands, adjacent mainland, and Saterland, in West Germany; and North Friesland (German: Nord Friesland, W. Frisian: Noard Fryslân), the North Frisian Islands, the isolated island of Helgoland (W. Frisian: Helgo-

< previous page

page_132

next page >

< previous page

page_133

next page > Page 133

lân, but in local Frisian dialect: Hallilönj), and adjacent mainland, mostly in West Germany but the northernmost portion in Denmark. In the Netherlands, the West Lauwers variety of Frisian is the official language of the Province of Friesland, Dutch the national language (Gorter, 1987, 1988; Gorter et al., 1988; Pietersen, 1974; Ytsma, 1988; Ytsma et al., 1989). But there are several localised variants of West Frisian: on the mainland, Klaaifrysk to the west, Sudhoeksk to the southwest, Wâldfrysk to the east (and across the provincial border into Groningen, east of Surhuisterveen/ Surhústerfean), and Noardhoeksk to the northeast. Unique local dialects are spoken on the islands of Terschelling/Skylge, Ameland/It Amelân, and Schiermonnikoog/It Skiermuntseach, as well as in the town of Hindeloopen/ Hylpen. Dutch-Frisian admixture is discernable in Stadsfries ('city Frisian'), spoken in some of the larger towns, and in certain unique dialects restricted to small areas, such as It Bildt and Kollumerland, as well as so-called 'island Dutch' dialects (on Vlieland/Flylân, the Midslan district of Terschelling/ Skylge, Ameland/Amelân, and Texel/Teksel). The unique Oost/East- and West- Stellingwerfs dialects of West Germanic (Saxon), are spoken in the southeastern district of the province. The population of East Friesland, in West Germany, speaks German in standardised or Lower Saxon form; the East Frisian language is today reportedly found only in several villages of Saterland (Fort, 1988). Whereas in North Friesland remnants of North Frisian are still found in various localised forms (at least nine varieties) together with a German-Danish Jutish dialect, although German prevails (Steensen, 1988). And on the three northernmost islands and adjacent mainland, in Denmark, the Danish language and Jutish variation prevail, while there is a substantial German-speaking minority around Tonder (Svalastoga & Wolf, 1969). In sum, people of Frisian origin (immediate or distant) all speak, and are educated to be literate in, one standard 'national' languageDutch, German, or Danish (depending on their citizenship). Only in the Dutch Province of Friesland are they also given the opportunity to be educated in Frisian, although there is limited state recognition of the Frisian language in North Friesland (West Germany). Yet throughout the historic Frisian areas in the three countries localised dialects or variations are heard (Fort, 1988; Gorter, 1987, 1988; Gorter et al., 1988; Steensen, 1988; Ytsma, 1988; Ytsma et al., 1989). Hopefully we have succeeded, by briefly examining just these two European cases, in emphasising the great complexity of minority language use. Conclusion The several theoretical typologies outlined in this paper should prove useful to scholars of minority languages, particularly social scientists studying minority language situations. The first typology, focusing on 'indigenous' ethnolinguistic minority situations, particularly emphasised the failure of arbitrary political borders

< previous page

page_133

next page >

< previous page

page_134

next page > Page 134

to coincide with actual ethnic or linguistic frontiers; moreover this typology stressed the complexity of zones where ethnolinguistic groups meet. In fact, this typology explored the very nature of ethnic territoriality; whereas the second typology was concerned with non-indigenous, i.e. immigrant minoritieshere the emphasis, perhaps, was not as much on territoriality as utility of migrants and their integration (or lack of it) into the socioeconomic system. The next couple of typologies, concerned respectively with alternative state policies and minority responses, particularly emphasised that the politics of minority language use should be researched, not just the changing forms and varieties of minority languages; languages spoken by ethnic minorities should profitably be studied within the context of ethnonationalism, state policies toward ethnic minorities and the concomitant response of these minorities. Lastly, the paper commented on language policies, minority language use, and language standardisation versus localisation within minority areas. This discussion essentially provided a set of qualifications aimed at countering oversimplified sociolinguistic analyses of minority languages. Doubtless there are many other theoretical or methodological considerations which may occur to researchers involved in studies of ethnic' or linguistic minorities. Yet hopefully these various typologies which have been outlined within a general sociopolitical framework are sufficiently provocative guidelines for continuing research into specific minority situations in diverse locations. References Anderson, A.B. (1978) Language minorities and international frontiers: The contemporary situation in Western Europe. Paper presented at the meetings of the International Sociological AssociationResearch Committee on Sociolinguistics, Ninth World Congress of Sociology, Uppsala Universitet, Sweden. (1979) The survival of ethnolinguistic minorities: Canadian and comparative research. In H. Giles & B. St-Jacques (eds) Language and Ethnic Relations. Oxford: Pergamon Press. (1980) The problem of minority languages: Canadian and European contrasts. Paper presented at the First International Conference on Minority Languages, Glasgow University. (1981) The problem of minority languages: Reflections on the Glasgow Conference. Language Problems and Language Planning. 5, no. 3, Fall. (1983) Methodological problems in comparative analysis of language contacts. In P.H. Nelde (ed.) Theorie, Methoden und Modelle der Kontaklinguistik. Bonn: Dummler. (1986) Measuring the success and failure of state minority policies: Western Europe and Canada. Paper presented at the Seventh Annual International Human Rights Symposium and Research Conference, Columbia University, New York. (1987) Ethnonationalism and regional autonomy: Western Europe and Canada. Paper presented at the International Workshop on Comparative Secessionist Movements, Colombo, Sri Lanka. (1989) The changing situation of ethnolinguistic minorities along the Yugoslavian frontier. Canadian Review of Studies in Nationalism XVI. Anderson. A.B. and Frideres, J.S. (1981) Ethnicity in Canada: Theoretical Perspectives. Toronto: Butterworths.

< previous page

page_134

next page >

< previous page

page_135

next page > Page 135

Anderson, N. (1969) Some comparisons of bilingual communities. In N. Anderson (ed.) Studies in Multilingualism. Leiden, Netherlands: E.J. Brill. Ashworth, Georgina (1977, etc.) World Minorities, vols. I, II, and III. Sunbury, Middlesex, UK: Quartermain House. Connor, W. (1976) The political significance of ethnonationalism within Western Europe. In A. Said and L.R. Simmons (eds.) Ethnicity in an International Context. New Brunswick, New Jersey: Transaction Books. Deutch, K.W. (1968) The trend of European nationalismThe language aspect. In J.A. Fishman (ed.) Readings in the Sociology of Language. The Hague/Paris: Mouton. (1975) The political significance of linguistic conflicts. In J.-G. Savard and R. Vigneault (eds) Multilingual Political Systems: Problems and Solutions. Quebec: Centre International de Recherche sur le Bilingualisme & les Presses de l'Université Laval. de Vries, J. (1984) Factors affecting the survival of linguistic minorities: A preliminary comparative analysis of data for Western Europe. Journal of Multilingual and Multicultural Development 5 (34). (1987) Education and the survival of linguistic minorities. Centre for Research on Ethnic Minorities, Etc. series, no. 8. Esman, Milton J. (ed.) (1977) Ethnic Conflict in the Western World. Ithaca, N.Y.: Cornell University Press. Fort, M.C. (1988) Uber den gegenwartigen Stand der saterfriesischen Sprache. Pogrom 141, June. Gorter, D. (1987) Dutch state policy towards the Frisian language. Paper presented at the International Workshop on Comparative Secessionist Movements, Colombo, Sri Lanka. (1988) Das Friesische in den Niederlanden. Pogrom 143, Sept. Gorter, D. et al. (1988) Language in Friesland. Ljouwert: Fryske Akademy. Hall, Raymond L. (ed.) (1979) Ethnic AutonomyComparative Dynamics. New York: Pergamon Press. Heraud, G. (1966) Peuples et Langues d'Europe. Paris: Denoel. Hunt, C.L. and Walker, L. (1974) Ethnic Dynamics: Patterns of Intergroup Relations in Various Societies. Homewood, Ill.: Dorsey Press. Jakobson, R. (1972) The beginning of national self-determination in Europe. In J.A. Fishman (ed.) Readings in the Sociology of Language. The Hague/Paris: Mouton. Lieberson, S. and O'Connor, J.F. (1975) Language diversity in a nation and its regions. In J.-G. Savard and R. Vigneault (eds) Multilingual Political Systems: Problems and Solutions. Quebec: Centre International de Recherche sur le Bilingualisme & les Presses de l'Université Laval. Lieberson, S. et al. (1975) The course of mother tongue diversity in nations. American Journal of Sociology 81, 3461. Mackey, W.F. (1967) Bilingualism as a World Problem. Montreal: Harvest House. Marger, M.N. (1985) Race and Ethnic Relations: American and Global Perspectives. Belmont, California: Wadsworth. Pietersen, L. (1974) Language ideologynational ideologybilingualism, the Frisian case. Research paper presented at a session on 'Language and National Identity', I.S.A. Research Committee on Sociolinguistics, Eighth World Congress of Sociology, Toronto. Simon, W.B. (1969) Multilingualism: A comparative study. In N. Anderson (ed.) Studies in Multilingualism. Leiden: E.J. Brill. Steensen, T. (1988) Die Nordfriesen und ihre Sprache. Pogrom 144, December. Stephan, W. (1987) Ethnicity, multiculturalism and political socialisation. Paper presented at the International Workshop on Comparative Secessionist Movements, Colombo, Sri Lanka. Stephens, Meic (1976) Linguistic Minorities in Western Europe. Llandysul: Gomer Press. Williams, Colin H. (1980) Ethnic separatism in Western Europe. Journal of Economic and Social Geography lxxi, 3.

< previous page

page_135

next page >

< previous page

page_136

next page > Page 136

(ed.) (1982) National Separatism. University of Wales Press. Wood, J.R. (1981) Secession: A comparative analytical framework. Canadian Journal of Political Science XIV, 1. Ytsma, J. (1988) The Frisian Language in Primary Education in Friesland, The Netherlands. Ljouwert: Fryske Akademy. et al. (1989) Frysk: Taalbehald, Taalferoaring, Taaluntjouwing. It Beaken Jiergong li 1989 no. 2.

< previous page

page_136

next page >

< previous page

page_137

next page > Page 137

Notes for a Minority-Language Typology: Procedures and Justification John Edwards Department of Psychology, St Francis Xavier University, Antigonish, N.S., Canada B2G 1C0 Abstract In line with the stated themes of the Fourth International Conference on Minority Languages(a) comparative research and similarities/differences among minority-language communities, and (b) theory generationthis paper will present an approach to a typological outline with which minority-language situations can be categorised. It is the latest development in a line of work reported on most recently in Riis (Edwards, 1988a) and in Noordwijkerhout (Edwards, 1988b). Previous work has provided a geographical underpinning, has referred to the sociolinguisic profiles of Ferguson, and Stewart (dealing with language functions, degrees of use, standardisation and writing, etc.), has examined the language 'ecology' studies of Haugen, Haarmann, and Giles, and has produced a rudimentary typological framework in which the categories 'speaker', 'language' and 'setting' are juxtaposed with dimensions including 'sociology', 'linguistics', 'psychology', and so on. The present paper summarises work done to date, suggests further directions, and attempts a justification for the methodology outlined. The Fourth International Conference on Minority Languages held in June 1989 in Ljouwert had two broad themes: (a) comparative research on minority languages, with special reference to similarities and differences among contexts; (b) theoretical development. This article addresses both of these, for a typology is at once an attempt to impose theoretical/descriptive order, and a means of codifying and facilitating cross-community comparison. It is also worth pointing out, as Ó Riagáin (1989) has done, that the comparison of cases can often tell us what more we need to know about the individual cases themselves. Here, I intend to briefly outline some relevant background material and some existing typological efforts, and to sketch out a new model. This, as will be seen, is at quite a rudimentary level; it is in fact, a simple checklist of important variables. However, even a fairly

< previous page

page_137

next page >

< previous page

page_138

next page > Page 138

thorough-going checklist is not without value. Plans, however, include coding and weighting of variables with a view to the eventual production of an instrument which could be used to assess the weakness/strength of minority situationsboth in 'objective' terms and according to the 'subjective' perceptions of those involved (à la the transformation of Giles' 'ethnolinguistic vitality' model into one measuring people's perceived vitality; see below). We might first consider why a typology is a worthwhile exercise at all. Williams, who, in several recent reviews, has severely criticised the current state of the sociology of language in general and minority-language matters in particular, has questioned the utility of typologies (see, for example, his 1980 review of Allardt's comparative study of European minorities). More specifically, he has claimed (1986) that typologies reflect 'implicit theoretical assumptions' while perhaps having 'limited analytical usefulness' (p. 509). In 1988, he repeated these observations, adding: 'I fail to understand the preoccupation of students of language with typologies' (p. 171). I think Williams' points are of interest, but they do not necessarily sound a death knell for typologies. All endeavours proceed from implicit assumptions, but the constraints these imply can be greater or lesser depending, among other things, upon the comprehensiveness of the undertaking; a broader typology with many elements is thus more likely to be useful than a narrower approach. Also, whatever the verdict on the purely analytic utility of a typology, it would seem that simply having a broad listing of potentially important elements is worthwhile. Roberts (1987), a colleague of Williams, has also criticised the use of typologies in language policy/conflict situations. They 'are born out of static, descriptive accounts of situations, and imply permanent relationships' (p. 311). They can provide 'snapshot accounts of particular language situations, but the tendency to 'fit' the parameters of a given typology onto a language situation results in some serious limitations' (p. 312). They take no account of the 'historically specific dimensions of a language situation' and are constrained by their 'inability to pinpoint the dynamic (and frequently contradictory) interrelationships between different elements (p. 312). A specific difficulty is that their application 'forces discussion of societal bilingualism as a stable state' (p. 321). These are generally useful cautions, but it should be noted that a typology per se need not imply permanence (typological models could be re-administered from time to time, for example), that any account will necessarily be a 'snapshot' one, that it is only misuse of a model which would lead to a forcing of parameters, that there is no good reason why a typology cannot explore historical dimensions, that a good model could actually elucidate relationships among variables, and that a typological treatment of bilingualism which permitted discussion of it only as a stable state would be obviously flawed.

< previous page

page_138

next page >

< previous page

page_139

next page > Page 139

My point is simply this: since there is every reason to assume that people will continue to interest themselves in language situations, and wish to describe and account for their dynamics, since it makes no sense to assume that different contexts are unique in every element, and since we are inevitably and rightly drawn to the task of theory construction or, at least, classification/descriptionthen a comprehensive and well-specified typology may serve as a useful guide. Byram (1988) has observed that cross-context comparisons might well be facilitated, for example, if attention was given to the same variables in all settings; any student in the area will have experienced frustration in attempting comparisons and contrasts where this sort of attention has not been paid. Finally here, we can simply observe that many respected workers in the fieldHaugen, Stewart, Kloss and Ferguson among themhave felt it meaningful to employ a typological approach. Haugen's concern, for example, was that: most language descriptions are prefaced by a brief and perfunctory statement concerning the number and location of its speakers and something of their history. Rarely does such a description really tell the reader what he ought to know about the social status and function of the language in question. Linguists have generally been too eager to get on with the phonology, grammar, and lexicon to pay more than superficial attention to what I would like to call the 'ecology of language'. (Haugen, 1972: 325) We could expand on this by noting that, besides linguists, educationalists, sociologists, psychologists, historians and others have also often failed to give sufficient treatment to ecological variables. As well, it seems to me that useful sociolinguistic forays into minority-language mattersand many others, toomust be interdisciplinary; we can no longer afford the luxury of simply remaining within our own narrow boundaries, particularly since the location of these boundaries is very much open to debate. Previous Work General information has already been presented elsewhere (Edwards 1988a, 1988b), and so I shall be brief here. It will also be noted that I omit entirely here treatment of several important contributionsfrom Anderson, Moss, LePage, Paulston, Weinreich and others; further work incorporating their insights is ongoing. It might be useful, however, to provide here what I consider to be a practical geographical underpinning for more socially-orientated work. There are actually rather few geographic typologies (see Edwards, 1988a), although many writers have somewhat informally drawn attention to specific

< previous page

page_139

next page >

< previous page

page_140

next page > Page 140

points of similarity across contexts, and I have concluded that the most useful model to build upon is that proposed in 1987 by White. This model introduces three basic distinctions. The first is among minority languages which are unique to one state, those which are non-unique but which are still minorities in all contexts in which they occur, and those which are minorities in one setting but majority varieties elsewhere. This initial distinction gives rise to the terms unique, non-unique and local-only minority. The second distinction deals, with the type of connection between speakers of the same minority language in different states; are they adjoining or non-adjoining? Finally, what degree of spatial cohesion exists among speakers within a given state? Here, the terms cohesive and non-cohesive can be used. Given that the adjoining/non-adjoining distinction does not apply to unique minorities, it follows that a ten-cell model emerges. White's framework, as adapted by Edwards (1988a) dealt with indigenous minority-language settings. However, in Edwards (1988b), the model was extended to also treat immigrant situations. Table 1, in reflecting this extension, provides examples for each of the cells. There are, of course, problems with this framework (as with any other); some of these are noted in Edwards (1988b). However, it may possess some utility. Indeed, a geographical framework alone might be quite usefulfor example, other things being equal (which they seldom are) or, indeed, unequal, the strength of a minority language will vary along the three dimensions of the model. A purely geographical approach must, nevertheless, have a severely limited significance; in order to more fully apprehend the complexities of minority languages and their speakers, further information from a variety of sources is clearly required. Ferguson and Stewart, to cite two well-known examples, have presented classificatory information concerning language types, functions, status and degrees of use. Thus, Ferguson (1962; 1966) discusses patterns of language dominance, extent of standardisation and degree/type of language use; Stewart (1962; 1968) outlines language types (vernacular, standard, classical, etc.) and functions (in education, religion, officialdom, cross-group communication, and so on). These schemes, valuable as they areand, by the way, relatively neglected as they are, at least outside that group of 'typologically-inclined' scholarsclearly deal with an incomplete subset of the important features. There is little attention given to social-status elements for example, or to what Haugen would term 'ecological' variables (i.e. those describing and illuminating interactions among languages, speakers and environments). Haugen's (1972) own ecology-of-language approach embraces linguistic, sociological, sociolinguistic and other aspects, and directs attention to such matters as language domains, use, traditions, politics and attitudes. Haarman (1986) has built upon Haugen's perspective, expanding it considerably in terms of greater specificity of variables. However, while both Haugen and Haarmann move in the right directions, there remain some difficulties; these

< previous page

page_140

next page >

< previous page

page_141

next page > Page 141

are addressed in Edwards (1988b), but can be conveniently summarised here: (1) There is insufficient specificity of variablesboth schemes sketch out important areas, but lack of precision clearly detracts from typological utility. (2) Some important matters are almost entirely neglected; little is said, for example, about historical, psychological, educational and geographic dimensions. Haarmann and Haugen could respond to these points by noting that their models in no way restrict amplifications and expansions. Nevertheless, the fact remains that many points of detail are not explicitly presented, and I take this to be a failing in models supposed to facilitate cross-situational comparability. Also, the introduction of neologisms and the use of terms which show considerable overlap are unfavourable features (although I must admit that my own scheme, in its present incarnation at least, also fails somewhat on these grounds). Another point to be made here is that the ecological approach has not formally been taken up very much by other researchers, although Haarman (1986) rightly points out that scholars had been conducting 'ecological' investigations before the birth of Haugen's particular outline. But, given the aim of more rigorously formulating the necessary requirements for ecological understanding it is rather surprising that one does not see more direct acknowledgement; for example, two recent books on the sociology of language (Fasold, 1984 and Wardhaugh, 1986) do not refer at all to the ecology of language. A recent typological approach of particular psychological interest is found in the 'ethnolinguistic vitality' model. Giles, Bourhis & Taylor (1977) outlined three factorsdemography, status and institutional supportseen to contribute to group viability. In 1981, Bourhis, Giles & Rosenthal extended this 'objective' scheme to one of perceived or subjective vitality; they made the reasonable argument that groups' perceptions of vitality may be at variance with objective reality, and that such perceptions may be important determinants of individual and group behaviour. They thus presented a 22-item 'subjective vitality questionnaire' based upon the three factors just noted. Again, however, there are difficulties (see Edwards, 1985; 1988b, for details). As with the Haugen and Haarmann models, the vitality conception contains elements which are too general, and neglects altogether some vital features. In presenting the original, 'objective' format, Giles et al. (1977) do provide useful discussion of the three factors, and they acknowledge that their analysis is not exhaustive. Nonetheless, they point out that the three-factor scheme meaningfully deals with linguistic minorities, and the subsequent expansion into 'subjective' vitality assessment might be seen as a premature solidification of factors in the 22-item format. In this latter

< previous page

page_141

next page >

< previous page

page_142

next page > Page 142

Table 1 Some examples of minority language situations. Type Indigenous Immigrant minorities minorities 1.Unique Sardinian Perhaps dialect communities (often religiously Cohesive (Sardinia) organised), where variety is now quite divergent from Welsh (Wales) language in region of origin (e.g. Pennsylvania Friulian (Friuli- 'Dutch') Venezia-Giulia) 2.Unique Cornish As above, but where speakers are scattered Non(Cornwall) cohesive 3.NonOccitan Any enclaves of immigrants found in adjoining states unique (Piedmont and Adjoining Liguria/and in Cohesive France) Basque (France/and in Spain) Catalan (Spain/and in Andorra) 4.NonSaami (Finland, Scattered immigrants in adjoining states unique Norway, Adjoining Sweden and NonRussia) cohesive 5.NonCatalan Welsh (Patagonia) unique (Spain/and in Scots Gaelic (Nova Scotia) NonSardinia) adjoining Cohesive (Table continued on next page)

< previous page

page_142

next page >

< previous page

page_143

next page > Page 143

(Table continued from previous page) Table 1 Some examples of minority language situations. (continued) Type Indigenous minorities Immigrant minorities 6. NonRomany (throughout Europe) Scattered immigrants of European unique origin in 'new world' countries Nonadjoining Noncohesive 7. Local- French (Valle d'Aosta/and in French (in New England town only France) enclaves) Adjoining Spanish (South-West U.S.A.) Cohesive Italian gastarbeiter (Switzerland) 8. Local- German (Piedmont/and in French (scattered in New England) only Switzerland) Adjoining Noncohesive 9. Local- French (Apulia/and in France) Immigrant enclaves in 'new world' only countries NonItalian gastarbeiter (Germany) adjoining Cohesive 10.Local- Albanian (throughout the Scattered immigrants in 'new world' only Mezzogiorno/and in Albania) countries Nonadjoining Noncohesive

< previous page

page_143

next page >

< previous page

page_144

next page > Page 144

format, dimensions such as the historical, economic, religious, political and educational are assessed by only one question each; it is thus inevitable that only a very rudimentary overview can result. The subjective vitality questionnaire has been used by Giles and his associates in a number of settings in the last few years. For example Bourhis & Sachdev (1984) used it to compare the perceptions of Italian-Canadian and EnglishCanadian high-school students in Ontario; Pierson, Giles & Young (1987) compared the assessments made by Chinese and non-Chinese secondary-school students in Hong Kong. These are but two of a dozen or more empirical studies employing the instrument, and Young, Bell & Giles (1988: 288) note that: sufficient and necessary exploratory groundwork therefore has been laid down now for a more systematic programme of cross-national empirical research to be undertaken. Therein, a vitality theory can be formulated. Based upon points made above (see also Edwards, 1988b), it will be clear, however, that I have some reservations. These revolve around what I consider to be inadequate and incomplete delineation of variables. Perhaps the most important aspect of any typology is comprehensiveness. Without this, it may have superficial plausibility and it may, indeed, enable separation and categorisation of groups. It will, however, be limited in exactly the same way that a factor analytic programme is constrained by the appropriateness and breadth of input. An Expanded Typological Approach I have mentioned here the work of Haugen, Haarmann, Giles and others, and have tried to indicate that much further analysis of existing literature is needed before anything approaching a comprehensive typology is produced. Suggestions were also made in Edwards (1988b) about further developments which would logically follow upon the construction of a more complete 'checklist'; these include attempts to provide relative weightings for variables, probing for meanings attached to elements by respondents (i.e. going beyond measurement of belief to assessment of attitude, in the usual psychological meaning of these terms), factor-analytic reduction exercises, and the general formulation of an instrument which could be used for both 'objective' and 'subjective' purposes. The present stage of development is clearly quite elementaryexcept, perhaps, for the basic geographic frameworkbut several important elements have suggested themselves. For example, if we were to consider: (a) three basic categoriesspeakers, language, setting (these recall, of course, Haugen's ideas of ecological relationships); (b) a number of substantive perspectives (in Edwards, 1988b, 11 of these were discussed: demography,

< previous page

page_144

next page >

< previous page

page_145

next page > Page 145

sociology, linguistics, psychology, history, politics/law/government, geography, education, religion, economics/business/industry, media); (c) cross-tabulations of these categories and perspectivesthen a scaffolding of cells emerges which might serve as a useful starting point. Elsewhere (Edwards, 1988b), I outlined some of the areas which such a framework would focus attention on, and also pointed to some of the difficulties, overlaps and so on associated with it. The full elucidation of this model awaits further work. However it has already thrown light upon some interesting features, and I would like to conclude here by drawing attention to some of these; in particular, in what follows, I hope it can be seen that what may, at first glance, appear to be rather straightforward matters are somewhat more involved, and deserving of our closer scrutiny. At least one question, or area of interest, from each of the substantive perspectives noted above will be very briefly presented here. Demography The whole area involved in extracting information from basic statistics is a much more complicated one than might first be supposed. In discussing 'demolinguistics', de Vries (1989) has indicated the usefulness of this resource for the study of second-language acquisition, language maintenance and shift pertaining to linguistic minorities ... assessing the relative contributions of fertility, mortality, nuptiality, migration and language shift to the survival or decline of minority language communities. (de Vries, 1989: 18) The urban/rural nature of the minority-group situation is also of great importance. The very rurality and isolation which can constitute a stronghold for a language can also, over time, cause it to become associated with a backward or impoverished way of life fuelling out-migration. Sociology It is widely considered that majority-minority intermarriage is often detrimental to minority-language survival and transmission. Yet, two recent studies in Nova Scotia (Edwards & Doucette, 1987; Edwards & MacLellan, 1989) have suggested that even students who clearly see themselves as ethnic-group membersand who are, of course, of more or less marriageable age themselvesplace within-group marriage at the bottom of a list of factors seen to be important for ethnic-group identity and continuity. A second sociological factor of great importance is the degree of what Breton (1964) has termed 'institutional completeness'. High levels of such

< previous page

page_145

next page >

< previous page

page_146

next page > Page 146

completeness lead, among other things, to increased survivability of group languages. I would suggest that this factor has considerable illuminative valueit would be interesting, for example, to see what an institutional-completeness perspective would make of Smolicz's (1988) notion of 'core' values. Linguistics A matter not sufficiently discussed is the degree of dialectal variation found within a given minority-language community. In Nordfriesland, for example, in an area of some 800 square miles, there are five languages in regular use. One of these, Frisian, is divided into ten major dialects, not all of which are mutually intelligible, among a population of only 10,000. Psychology A major psychological thrust has always been the study of attitudes. With regard to language attitudes, important areas include differences between communicative and symbolic facets of language, and between group 'spokesmen' and more 'ordinary' members (see Edwards, 1985). We also need more information about the perceptions of majority-group members. Do their attitudes and actions indicate a desire to see the minority language actively promoted, or a more passive goodwill towards diversity? Can goodwill be translated into something more dynamic and positive for minority-group viability? A major study in Canada was that of Berry, Kalin & Taylor (1977) and plans are now underway for an updating national survey. History An historical dimension is essential in comprehensive study of minority-language situations. On the one hand, as Seton-Watson (1981: 2) has said, 'the history of language ... forms a very important part of social history, and one which seems to me to be relatively neglected by most historians'. On the other hand, as implied by Haugen (1972), most students of language have paid very little attention to history. Apart from the fact that the historical perspective has typically been given short shrift, examination of the historical record is sometimes downplayed for not producing 'data' of a sort immediately familiar to researchers in sociology or psychology; I think that this indicates a very restricted conception of data. Politics An interesting matter which sometimes arises in minority-language contexts is a potential clash between group and individual rights. In Quebec

< previous page

page_146

next page >

< previous page

page_147

next page > Page 147

recently, the passage of the so-called 'sign law' has been seen by some as a restriction of the rights of individual anglophones (in this case, to display commercial signs in English) in the cause of support of francophone language and culture in the province. Difficulties suggest themselves, particularly within cultures where rights are generally seen to inhere in the individual person rather than in groups or collectivities (see also Edwards, 1985). Geography A geographical framework has, of course, already been outlined. More attention could be given, as well, to the physical avenues of transportation and communication available to a language minority; the road desired for mobility may also be the road of group dilution. Education Perhaps the most important area here has to do with the type and extent of school support for minority languages. Only fairly fine-grained investigation will reveal what really goes on in classrooms as opposed to what official policy dictates should go on. Only careful study will tell us if the 15 hours per week given to a minority language in context 'A' is in any way comparable to the same time allotment given in context 'B'. Religion It is sometimes the case that a strong association exists between a minority language and religion; in the Irish situation, for example, much was made by revivalists of this connection. A related matter worthy of more study is the question of whether and/or when secularisation contributes to language shift. Economics While not wishing to argue for a simplistic economic 'reductionism' in language matters, it does seem difficult to deny that mundane economic facts have a great deal to do with minority-language viability. This is not, of course, a popular line among many of the more romantically-inclined apologists for language maintenance and, perhaps for that reason, has not received due attention (see Edwards, 1985). Media A useful perspective on the media is to view them as double-edged swords. On the one hand, it can be argued that minority-language presence, particularly on television, is of great importance for group solidarity and

< previous page

page_147

next page >

< previous page

page_148

next page > Page 148

legitimacy; indeed, it has been suggested that television has become a new language domain in its own right. On the other hand, the pervasiveness of satellite-transmitted television, coupled with the overwhelmingly American (or Americanised) content, may create real difficulties for minority-language maintenance efforts. Conclusion (With An Observation On Methodology) This brief paper has done no more than indicate a typological direction. It is regrettable that the work has not advanced as quickly as I had initially hoped, but this in itself confirms my belief that to do justice to the issue requires a considerable amount of synthesis and study. The further work still to be done is suggested above; perhaps the most fruitful direction now involves the refinement of typological elements, the provision of scalar measures, and the subsequent application to several real settings. The theoretical 'quality of the typological approachsuch as it isrests upon an inductive argument which, it has been suggested, may require some justification. Certainly, the exercise I outline here involves a move towards generality (or theory) from an assembly of data. It is thus prey to the difficulties which have, at least since the time of Hume, been well-understood. The chief problems with induction, at its most basic level at least, are that one will never collect all the necessary facts and, relatedly, even if one were to attempt to restrict oneself to an assemblage of the most relevant ones, there is clearly a difficulty in ascertaining a priori what is and is not germane. Building upon Hume's questioning of the possibility of reasoning from instances we know to those we don't, Popper (1972) notes that we are essentially asking if a belief that the future will be like the past can be justified. The central process of inductionreasoning from a sample to what Whitehead (1925) called the 'whole species'may thus seem fatally flawed; in such a process there can be no generally applicable 'inductive rules'. We might then agree with Popper that induction is a myth because its products are always invalid. We might also then wish to employ the hypothetico-deductive method, a commonly-observed procedure in social science: a 'known' law is tested by drawing conclusions from it (i.e. deriving hypotheses) and putting them through further observational trials. This, logically, is a tighter procedure. Is, then, a deductive exercise always to be preferred over an inductive one? What are the implications for, say, the typological work discussed in this paper? Current philosophy of science does not, of course, simply dichotomise induction and deduction as I've characterised them here. Forms of induction have been discussed, for example; indeed, we can go back to Aristotle for an outline of the differences between 'complete' or 'enumerative' induction (where all cases are considered) and 'ampliative' induction (involving a sample of cases). Modern views generally emphasise the probabilistic nature

< previous page

page_148

next page >

< previous page

page_149

next page > Page 149

of induction and, for example, Carnap's (1952) five varieties of inference build on this. Similarly, there has been a variety of approaches to deduction, which might be generally characterised as involving necessary inference (i.e. as opposed to induction's probabilistic nature) whether the form is the most familiar 'general-to-particular' one or whether the argument is 'general-to-general' or 'particular-to-particular'. I put 'known' in quotation marks above to signal a further issue and to point out that in ordinary scientific practice induction and deduction do not exist as neat opposites. First, we may ask how a law to be tested was itself arrived at. Second, we can see that, in hypothesis confirmation itselfi.e. in 'normal' scientific methodologywe are in effect using the confirmed predictions to bolster our faith in the hypothesis; this seems not unlike that reasoning from data to theory which denotes induction. The point here is simply that induction and deduction are linked. Thus Whitehead (1925: 23) noted that 'the theory of induction is the despair of philosophy, and yet all our activities are based upon it'. Another point of interrelatedness, incidentally, is that in inductively-based data gathering we are usually motivated, however illogically, by tentative hypotheses; this is the basis of Popper's rejection of the common belief that science proceeds from observation tout court to theory. The place to look for 'pure' deduction is within a more 'closed' system than science. Deese (1972) thus observes that induction is in fact the defining characteristic of empirical science, and distinguishes it from mathematics. He also points out, howeverin line with what has already been discussedthat induction means that science cannot be wholly rational. To elaborate here, we can turn to Popper (1963; 1972) who claims to have solved the problem of induction. His discussion hinges upon the fact that, while induction is logically insupportable, it is psychologically compelling; we form expectations, beliefs and even laws based on it. Popper's solution is essentially to remove induction from the realm of the purely rational, while at the same time emphasising that all our knowledge is tentative (or, as he would say, conjectural; Hempel, 1966, makes an apposite point here when he states that theories are not derived from facts so much as they are invented to account for themthey are 'guesses'). He does not, however, simply agree with Hume's own conjecture that the psychological appeal of induction rests upon custom or habit conditioned by repetition. Rather, Popper stresses a more deep-seated human desire/expectation to find regularity and order (most psychologists would surely agree). Our irrational reliance upon induction doesn't look so bad; while we should not, to be sure, delude ourselves that we have achieved certainty, still we can consider that our inductively-based beliefs represent a current 'best bet'. My suggestions here, then, are that induction is inevitable in science, that it is not an inferior quantity to hypotheticodeduction (from which, in any event, it cannot easily be dissociated), and that an examination of induction and its difficulties reminds us that, outside tautological systems, all know-

< previous page

page_149

next page >

< previous page

page_150

next page > Page 150

ledge is conjectural. All of this can be seen as a digression in support of a typological approach. I'd like to end though, with the further suggestion that a franker admission of both the inevitability and, indeed, the desirability of inductive methods might be salutary in social scientific fields where the brandishing of grand 'laws', 'theories' and 'hypotheses' is seen (mistakenly) to establish both that coveted linkage with 'real' science and the intellectual high ground. Acknowledgement I wish to thank my colleagues, Professors G. McGuire and E. Wright, for their helpful comments on this paper. References Berry, J., Kalin, R. and Taylor, D. (1977) Multiculturalism and Ethnic Attitudes in Canada. Ottawa: Supply & Services Canada. Bourhis, R., Giles, H. and Rosenthal, D. (1981) Notes on the construction of a 'Subjective Vitality Questionnaire' for ethnolinguistic groups. Journal of Multilingual and Multicultural Development 2, 14555. Bourhis, R. and Sachdev, I. (1984) Vitality perceptions and language attitudes: Some Canadian data. Journal of Language and Social Psychology 3, 97126. Breton, R. (1964) Institutional completeness of ethnic communities and the personal relations of immigrants. American Journal of Sociology 70, 193205. Byram, M. (1988) Comment made at International Conference on Maintenance & Loss of Ethnic Minority Languages, Noordwijkerhout, The Netherlands. Carnap, R. (1952) The Continuum of Inductive Methods. Chicago: University of Chicago Press. Deese, J. (1972) Psychology as Science and Art. New York: Harcourt Brace Jovanovich. de Vries, J. (1989) On coming to our census: A layman's guide to demolinguistics. Fourth International Conference on Minority Languages Programme, 18. Edwards, J. (1985) Language, Society and Identity. Oxford: Blackwell. (1988a) Socio-educational issues concerning indigenous minority languages: Terminology, geography and status. Paper to Colloquy on Lesser Used Languages in Primary Education, Riis, The Netherlands. (1988b) Sociopolitical aspects of language maintenance and loss: Towards a typology of ethnic minority language situations. Paper to International Conference on Maintenance & Loss of Ethnic Minority Languages, Noordwijkerhout, The Netherlands. Edwards, J. and Doucette, L. (1987) Ethnic salience, identity and symbolic ethnicity. Canadian Ethnic Studies 19, 5262. Edwards, J. and MacLellan, B. (1989) A sociolinguistic profile of Inverness County. Unpublished paper, St Francis Xavier University. Fasold, R. (1984) The Sociolinguistics of Society. Oxford: Blackwell. Ferguson, C. (1962) The language factor in national development. In F. Rice (ed.) Study of the Role of Second Languages in Asia, Africa, and Latin America. Washington: Center for Applied Linguistics. (1966) National sociolinguistic profile formulas. In W. Bright (ed.) Sociolinguistics. The Hague: Mouton. Giles, H., Bourhis, R. and Taylor, D. (1977) Towards a theory of language in ethnic group relations. In H. Giles (ed.) Language, Ethnicity and Intergroup Relations. London: Academic Press.

< previous page

page_150

next page >

< previous page

page_151

next page > Page 151

Haarman, H. (1986) Language in Ethnicity: A View of Basic Ecological Relations. Berlin: Mouton de Gruyter. Haugen, E. (1972) The Ecology of Language. Stanford: Stanford University Press. Hempel, C. (1966) Philosophy of Natural Science. Englewood Cliffs, N.J.: Prentice-Hall. Ó Riagáin, P. (1989) Comment made at Fourth International Conference on Minority Languages. Ljouwert, The Netherlands. Pierson, H., Giles, H. and Young, L. (1987) Intergroup vitality perceptions during a period of political uncertainty. The case of Hong Kong. Journal of Multilingual and Multicultural Development 8, 45160. Popper, K. (1963) Conjectures and Refutations. London: Routledge & Kegan Paul. (1972) Objective Knowledge. Oxford: Oxford University Press. Roberts, C. (1987) Political conflict over bilingual initiatives. Journal of Multilingual and Multicultural Development 8, 31122. Seton-Watson, H. (1981) Language and National Consciousness. London: British Academy. Smolicz, J. (1988) Language maintenance and educational policy: Language as a core value of ethnic cultures in Australia. Paper to International Conference on Maintenance & Loss of Ethnic Minority Languages, Noordwijkerhout, The Netherlands. Stewart, W. (1962) An outline of linguistic typology for describing multilingualism. In F. Rice (ed.) Study of the Role of Second Languages in Asia, Africa, and Latin America. Washington: Center for Applied Linguistics. (1968) A sociolinguistic typology for describing national multilingualism. In J. Fishman (ed.) Readings in the Sociology of Language. The Hague: Mouton. Wardhaugh, R. (1986) An Introduction to Sociolinguistics. Oxford: Blackwell. White, P. (1987) Geographical aspects of minority language situations in Italy. Paper to International Seminar on Geolinguistics, Stoke-on-Trent, England. Whitehead, A. (1925) Science and the Modern World. New York: Macmillan. Williams, G. (1980) Review of Implications of the Ethnic Revival in Modern, Industrialized Society (E. Allardt). Journal of Multilingual and Multicultural Development 1, 36370. (1986) Language planning or language expropriation? Journal of Multilingual and Multicultural Development 7, 50918. (1988) Review of Third International Conference on Minority Languages: General Papers; Third International Conference on Minority Languages: Celtic Papers (both volumes: G. MacEoin, A. Ahlquist and D. Ó hAodha). Language, Culture and Curriculum 1, 16978. Young, L., Bell, N. and Giles, H. (1988) Perceived vitality and context: A national majority in a minority setting. Journal of Multilingual and Multicultural Development 9, 28589.

< previous page

page_151

next page >

< previous page

page_i

next page > Page i

Fourth International Conference on Minority Languages Vol. I: General Papers

< previous page

page_i

next page >

< previous page

page_153

next page > Page 153

The Economic Approach to Minority Languages François Grin C.R.D.E., Université de Montréal, C.P. 6128, Succ. A, Montréal, Québec H3C 3J7, Canada Abstract This paper begins with a brief survey of existing contributions about language by economists. The relevance of these contributions to the specific case of minority languages in Europe is then critically assessed. After defining what is meant here by 'minority language', the paper explains the fundamental economic reasoning on the basis of which minority language use can be modelled. A simple model of minority language use is developed, providing some results that can be of help in evaluating the efficiency of proposed language policies. The impact of subsidising minority language goods, of increasing wage rates in minority language areas, and of offering better exposure to the minority language are shown to yield ambiguous results unless some conditions are met; these conditions regard the sensitivity of minority language activities to certain prices and to the wage rate offered on the labour market. The model also suggests that pouring money into minority language areas will yield disappointing results unless there is a firm commitment to improve the image of the language. Introduction At present (June 1989) there seems to be no organised body of economic theory specifically designed to study the case of minority languages. There are, however, a number of contributions which can be put under the general heading of the 'Economics of Language'. This body of literature derives its name from the fact that (a) it has been written by economiststhe vast majority of them from Quebecand also that (b) it relates various questions regarding language use to variables such as income, wage rates, savings, etc. which are traditional ingredients of economic theory. This paper provides a brief survey of this literature and then addresses the following two points: first, to what extent are these contributions relevant to minority languages? Second, what is the nature of an economic approach to language problemsnamely, can an economic approach dispense with

< previous page

page_153

next page >

< previous page

page_154

next page > Page 154

so-called economic variables, and apply economic reasoning to the non-material variables involved? Finally, this paper suggests a model of minority language use whereby the amount of time devoted to minority language use can be related to prices, average wage rates, and various forms of exposure to the minority language. The results obtained may be of help in assessing the efficiency of certain language policies. Economic Theories of Language: A Brief Survey Though economists' contributions to language matters still are relatively few, they are numerous enough to make some sort of classification useful. We shall apply the one suggested by Vaillancourt (1985), which contains five headings: (a) pure theory (b) language policies (c) socio-economic status (d) consumption-savings decisions (e) multivariate analyses. This paper being primarily concerned with the precise focus and relevance of theoretical texts, headings (b) to (e) will not be discussed here. In so far as language policies can be sensibly applied only after some generali.e. theoreticalview on language use is developed, it seems more appropriate to concentrate on theoretical approaches to begin with. For similar reasons, studies documenting inequalities linked with language proficiency, whether or not multivariate analyses are used, should be discussed only after some theoretical examination has been carried out. The same applies to contributions about the influence of linguistic parameters on consumption-savings decisions. The present section is therefore restricted to a discussion of the first of the above categories. Texts falling into this category are not very homogeneous, be it in terms of the specific topic addressed, the type of causation involved or the hypotheses made. The earliest contribution is probably the paper by Marschak (1965) entitled 'The economics of language'. In fact, Marschak raises a very central question, namely why languages change, and why some languages are more likely to survive than others. Marshak's answer is that the languages that are most efficient will survive, the efficiency of a language being defined as the ability to transmit a certain amount of information in relatively less time than another language. There has not been much follow-up on this definition of efficiency, if only because it does not seem to be firmly supported by evidence. However, the significant point here is Marschak's pioneering idea that language can be looked on as an object of 'choice' geared towards a certain goal, in the very same way as all economic decisions made by

< previous page

page_154

next page >

< previous page

page_155

next page > Page 155

individuals, according to standard micro-economic theory, reflect a constrained choice. A second group of theoretical texts attempts to model the impact of individuals' linguistic skills on their income. Obviously the topic is of particular interest to Canadian economists, and this line of study has been a frequent one. These texts draw a parallel between language group and ethnic group, and the central concept in these approaches is discrimination, which can be accommodated into economic modelling using Becker's (1976: 1730) economic theory of discrimination. Wage differentials can be explained either by a positive taste for discrimination by employers (Raynauld & Marion, 1972), or by lower productivity if minority language speakers are hired (Migué, 1970). Language can also be approached without making any analogy with ethnic or cultural groups, and viewed purely in terms of communication and exchange. For example Carr (1976) suggests that money and language share similar characteristics: using money as a common medium of exchange enables an economy to move beyond barter, which is cumbersomeand therefore costly. In the same way, when various language communities want to trade among themselves, they will lower transaction costs by using a common language. The overall cost of adapting to such a situation will be minimised if minority language speakers learn the language of the majority rather than the reverse. Breton (1964; 1978) and Breton & Mieszkowski (1979) also focus on language primarily as a means of communication with other linguistic communities. Breton examines the impact of language policies on the distribution of wealth within the community whose language is being protected by such policies. Breton & Mieszkowski apply the neo-classical international trade model and interpret the use of one common language as resource-saving technical progress. The rate of return on language learning will of course be higher for the language used as a lingua franca and will dwindle for the other language. The rather restrictive set of hypotheses used in many of the above articles has been criticised by Lavoie (1983), who advocates alternative approaches to the analysis of language. Lavoie stresses the fact that languages define not only what an individual has, but also what he is, and that a theoretical discussion should take proper account of this. Language can also be viewed in terms of human capital, and this hypothesis is certainly a very fruitful one. It draws on a well-established theoretical background in which professional experience, formal schooling, etc. are assets that enhance the individual's money-making ability. In the same way, language proficiency helps explain individual incomes and the choice of the language spoken at work. This framework, as used by Grenier (1982), also sheds light on second language learning by minority language speakers. The human capital approach as extended by Sabourin (1985) includes a phenomenon which is particularly relevant to the problems of minority

< previous page

page_155

next page >

< previous page

page_156

next page > Page 156

languages, i.e. the extent to which a language is used by a large number of other speakers. This also helps understand bilingual labour markets and has developed into a 'theory of language environment'. Another group of contributions defines language as both a form of human capital and a feature of ethnocultural identity. This idea is used for instance by Hocevar (1975) to discuss market equilibrium for goods' or factors displaying specific linguistic features (examples of such goods are: newspapers and various goods and services with instructions for use in a minority language, etc.). Studying Minority Languages versus Languages in General Though far from exhaustive, this quick glance at theoretical writings about language makes two points clear: first, the 'economics of language' as a branch of economics is an extremely wide and rather scattered area of study. Although most of the contributions mentioned here link language to some traditional economic variablemany of them stressing the problem of income differentialsthis range of topics need not be the core of the investigation. Explaining why language parameters give rise to wage differentials is only one of the many questions that arise. To mention only a few: Why do peope bother to learn a second or third language? Why and how do they split their time between two different languages over a normal working day? Are certain languages more closely linked with certain specific activities? To what extent is the use of one language dependent 'on its actual or perceived nature as 'religious/sacred', 'vernacular', 'vehicular', 'dominant', etc? Does the use of a language for communication within a linguistic community react to the same variables as a language used for communication outside it? Is diglossia stable? Does it offer greater stability than clear-cut linguistic divides? Do geographical factors systematically contribute to language shift? The list is endless, but the vast majority of the items do lie within the scope of economics. As I wish to show later on in this paper, there is no reason to restrict economic theory to material problems. Once it is acknowledged that economics is defined not by its subject matter, but by its approach, it may fruitfully be applied to a wide range of questions pertaining to linguistic behaviour. It is safe to assume that one of the most fundamental issues to solve is the following: 'Why is a language used at all (rather than another)?' Indeed, this covers most of the questions listed above. The second noteworthy fact about the survey above is that this problem has not been addressed as such, except by Marschak in very general terms. How can this fact be accounted for? Two answers can be suggested, the first of them methodological. It would indeed seem desirable to answer this

< previous page

page_156

next page >

< previous page

page_157

next page > Page 157

broad question by developing a general theory of language use in which changes in the latter as a response to a wide selection of variables would be explained on the basis of a few plausible axioms about human behaviour. However, the very generality of the approach may defeat its own purpose. Since a general theory of language use would need to assign a specific role to most of the numerous factors mentioned above, i.e. geography, historical status of the language, wage rate discrimination, time patterns of language use etc., the maze of causations would prove far too complex. If we want to retain a high level of generality, as Marschak does, we must sacrifice some of the detail; in other words, explaining language use generally would make it very difficult to include enough of the significant explanatory variables. The alternative is to focus on more specific points, for instance the decision to learn a second language for use outside one's traditional linguistic group. The absence of a general theory to answer our fundamental question (Why is a particular language used?) can also be explained by the fact that it may not, after all, be quite as important to everyone. Why a language is spokenor drops out of useis a particularly meaningful question when minority languages are studied, because it implicitly scrutinises the reasons for the very existence (or survival) of a language. This need not be a matter of major concern to North American economists: even if the use of French in Quebec may have been jeopardised, the existence of French as a language has never been threatened; language problems arose not so much because French-speakers were a minority in Quebec (they never were), but because both economic and political power were firmly in the hands of English speakers. In the same way, although Spanish and English are competing in some southern States where English still remains the majority language, Spanish can hardly be labelled an endangered language. It is therefore quite natural that Canadian economists should have paid such attention to income inequalities rather than to the question 'Why is a language spoken at all?' This latter question however, must be answered if we are to understand the challenge faced by most minority languages in Europe today. A lot of the existing contributions to the economics of language may no doubt prove very useful when analysing minority language survival or decline; but the fact remains that there seems to be no published economic literature specifically aimed at a theoretical analysis of situations such as those of Irish or Romanche. If attempts be made in this direction, how should we go about it? Bearing in mind the methodological warnings sounded above, what are the proper questions to ask? What is the appropriate level of generality? What exactly do we mean by 'minority language'? Which variables are relevant to most minority languages? Which can be dispensed with? Which contributions from the existing literature will provide the best stepping-stones? The following sections are intended as a blueprint for investigation of these issues.

< previous page

page_157

next page >

< previous page

page_158

next page > Page 158

A Definition of Minority Languages As stated earlier it would be too complicated to develop a general model of language use. Restricting ourselves to minority languages as we know them in Europe is a way to scale down the scope of the model, and also make it more tractable. Our central question narrows down to: Why is a minority language usedor not used?1 This question is here put in static terms, and can be seen as a pointer to the dynamic questions that are of major concern in Europe, namely: Why do minority languages die or survive (in the course of time)? What language policies will be able to keep them alive? Designing a dynamic model is usually more difficult, and as a first step it is reasonable to restrict our attention to a static model. To make our question perfectly clear, the sense in which 'minority language' is used here must first of all be clarified. 1. A minority language is a language spoken by less than 50% of the population in a given geographical area. This reference area is usually a nation-state, but although convenient, this criterion is not perfectly satisfactory. 2 2. The above definition includes languages facing quite different situations: it covers Welsh, Irish, Gaelic, Breton, Romanche, etc., as well as Hungarian in Romania, Spanish in the United States, French in Canada and Western Switzerland, or Chinese in Malaysia. The first restriction we will make is that we will consider only languages that are not a majority language in any other country. This excludes the five latter situations mentioned here. 3. Further restrictions are still needed, since most languages of India would still be defined as minority languages even though their speakers number millions. We shall therefore consider only those minority languages that are competing with one majority language, as is the case with most minorities in Europe. 4. We will also limit our attention to situations in which the minority and the majority language are sufficiently different from each other for them not to be mutually intelligible: in other words, speaking a minority language does not per se imply a working knowledge of the majority language and vice versa. This restriction will enable us to overlook the case of local dialects in Germany, for instance. 5. Extinct or near-extinct languages are a special case we will not consider either. This rules out Western Swiss dialects, Manx, Cornish or Walloon, etc.3 6. Furthermore, we will not consider languages spoken by elective minorities. This excludes the slang used by groups to which its members have decided to belong, whether it be a profession, a religious group or a street gang. Even though the deaf have not chosen their condition, sign language will not be considered here either. To put it another way, our attention will be restricted to minorities in an ethnic sense.

< previous page

page_158

next page >

< previous page

page_159

next page > Page 159

The definition suggested here does not provide a perfectly clear-cut standard according to which any language could be classified. Further and more refined criteria may still have to be introduced. 4 However, it provides a way to pick out among minority languages in Western Europe those that we may want to focus on. These are: Irish Gaelic Welsh Breton Basque Romanche Ladinian Friulian Occitanian Catalan Galician Corsican Sardinian Frisian Lappish. Clearly outside our definition are:5 Swedish in Finland Danish in Southern Sweden and Northern Germany Hungarian and Slovenian in Austria Flemish in Northern France French and Italian in Switzerland German, Greek, French and Albanian in Italy German in Belgium. Having defined what is meant here by a minority language, it is time to examine what tools economic theory can offer in order to help explain why such languages are used or not. A Possible Use of the Economic Approach Economic theory suffers from a major misunderstanding. Misperceptions on the part of other social scientists as well as a degree of blindness among economists themselves have blurred the fundamental nature of economics, to the point where collaboration between economics and say, sociology and anthropology, has become very difficult. I do not intend to make an in-depth analysis of this misunderstanding. Suffice it to say that economics is essentially concerned with human behaviour. Its central paradigm is best captured by Lionel Robbins' definition, who said that 'Economics is the science which studies human behaviour as a relationship between ends and scarce means which have alternative uses'. In other words, every time a choice between competing options must be made, we face an economic problem. The choice, of course, is a constrained one, because resources are limited; the central idea of the economic paradigm is that these resources will be used so that satisfaction ('utility') is maximised. Such optimal allocation of resources is what economists call 'rationality'. It is essential to note that rationality as defined here means nothing more than the decision-making principle applied by individuals. It does not in the least imply that some goal 'A' towards which resources are channelled

< previous page

page_159

next page >

< previous page

page_160

next page > Page 160

is in any normative sense more or less rational (or, for that matter, better or worse) than some other goal 'B'. The individual may pursue any ends he wishes, and these do not have to be materialistic; they do not even have to be linked with financial matters, and someone else's well-being may be among the ends being maximised. In the same way, the 'means' or resources used to achieve a goal may be quite immaterial. The central idea is therefore that economics is described not so much by its subject matter as by its approach. This idea, although it flows quite naturally from Robbins' definition (the most generally accepted definition of the discipline) has more often than not been neglected by economists themselves, who have devoted most of their attention to models in which individuals are assumed to maximise their personal well-being through strictly materialistic choices. The fact remains that the actual core of the discipline is the mechanism of constrained utility maximisation, as Becker (1965; 1976) forcefully explained. This approach has allowed considerable progress in human capital theory, which has been used by Grenier (1982) to explain language attitudes of Spanish-speaking Americans. The Beckerian framework might also help us to understand why a given minority language is used at all. I do not believe that economic reasoning can provide a complete and general theory of language use; however, it can shed light on certain influences on language use, and be combined with theories from other social sciences. A major advantage of the economic approach is that choices about language use, instead of being studied in isolation, can be analysed in relation to other choices made by the individual. The interdependence between various types of decision is built into the economic paradigm. At the same time, the entire theory rests upon a single general hypothesis about human behaviour, namely that everyone tries to make the most of the resources available. In the simplest version of the model, only two kinds of ingredients are needed, an objective (or 'utility') function, and a constraint. Such a simple model would state that minority language speaking itself is among the individual's goals, and is therefore one of the arguments of the utility function. At the same time, many costs may be associated with minority language use: sticking to one's minority language may imply the forgoing of earnings if minority language speakers are discriminated against; communication with majority language speakers may be seriously hampered; insisting on getting everything translated into the minority language also entails a money cost which will be borne indirectly by the individual. Thus a cost may theoretically be assigned to minority language use. The higher the cost, the fewer the resources left for other pursuits: this opportunity cost of speaking a minority language can turn out to be a key concept in explaining language use. Indeed, a rational individual will tend to decrease his or her use of the minority language if it proves too costly, unless he or she very definitely wants to retain it. (The degree to which they want to

< previous page

page_160

next page >

< previous page

page_161

next page > Page 161

make the effort reflects their individual tastes, and can be built into their utility function.) The above model is both simplistic and too abstract, and suggests that a third kind of ingredient should be added, namely 'technical relationships' which specify the way in which language use is related to ends and means or, to put it another way, how it shows up in goals and resources. In my view, these technical relationships offer two major advantages. The first one is realism: it helps us move down from very abstract generality closer to the individual's actual situation. Technical relationships allow the model to take into account various causal links, for instance the fact that minority language use may be hampered if individuals have to go out of their way to find video tapes in the language. This may be quite obvious in itself, but the aim of an economic model is to integrate such causal links into a broader and more general analysis of behaviour, and technical relationships allow us to do precisely that. However, for these technical relationships to describe such influences on language use adequately, the economist needs help from linguists, sociologists and geographers. This is therefore the second advantage of using technical relationships: they foster collaboration between economics and other social sciences. A Model of Minority Language Use I am not suggesting that the following model is the best possible contribution that the economic approach can make to answering the question 'Why is a language used at all?' My present goal is to exemplify what can be done using the traditional hypothesis of constrained utility maximisation. More specifically, the technical relationships suggested below are only an attempt to include some meaningful influences on minority language use. They do not purport to be exhaustive, and the actual relevance of the factors mentioned here is a matter open to discussion. Let me stress again that economic theory can be helpful in shedding light on some of the causations involved, provided they fit well into utility maximisation calculus. Comments and criticism are therefore welcome. To begin with, there are many ways in which language use can be fitted into an individual's utility function. As mentioned earlier, we may suppose that an individual derives direct satisfaction from minority language proficiency. We may also suppose that individuals are not interested in the language per se, but that they like a certain way of life of which their minority language is a necessary part. These particular options could be discussed at great length, but we may be guided in our choice by a definition of language use. Language use will therefore be measured here by the amount of time (per day, week, month, or any other appropriate period) in which the language is spoken. This suggests that language should be included in the individual's utility function in the following way.

< previous page

page_161

next page >

< previous page

next page >

page_162

Page 162 Assume that an individual maximises utility over a wide range of activities: the arguments of the utility function may be extremely disaggregated activities, or rather wide-ranging ones. For instance, eating can be treated as one very broad activity, or it can be broken down into a set of activities like eating at home, eating at friends' homes or eating in a smart restaurant. In the same way, 'transport' can be seen as one activity or as a class of distinct activities such as driving, using a public transportation system, or flying. Any level of generality may be adopted, depending on the set of activities analysed. As the point of the present paper is simply to describe the principles of such an approach, we will not go into detail, and we win assume that an individual engages in an unspecified number of activities every day. These n activities will be denoted by a vector [z1,z2,...,zn]. Assume that any of these activities can be pursued in either the majority language or the minority language. For the sake of the argument, let us refer to English and Welsh. Thus activity zj, for example 'reading', can be conducted in English or Welsh, and can therefore take one of two different forms which we will label zej or zwj. The individual will seek to maximise his or her utility function which can be written as: [1] Let us now focus on one of these activities, say . What is needed in order for an individual to read in Welsh? First of all, a book, magazine or newspaper in Welsh is needed. It is a material good used as an input in 'producing' the activity 'reading in Welsh'. Then, of couse, time is also needed. Some of the time used will be spent reading, but some time must also be used getting the reading material from a shop. (The amount of time actually spent reading in Welsh is part of the 'language use' defined here, since we have chosen to evaluate language use in time units). Time is therefore another input in the production process of activity . In its simplest form, the technical relationship we have can be written as: [2] This technical relationship is precisely what standard economic theory calls a 'production function', where xwj stands for the amount of goods (such as books, magazines or newspapers) while twj denotes the amount of time used. Some of it will be actual reading in Welsh, and some of it will be used to go out and buy reading material. If the latter can be bought from a Welsh-speaking shop, the proportion of twj corresponding to actual use of Welsh will be close to 100%. Time and goods are not the only factors that enable an individual to read in Welsh. Various other factors, which have been studied by sociologists and geographers, are also important: geographical isolation of speakers; language used in advertisements; frequency of minority language broadcasts on television or radio; government attitudes to minority language use; psychological pressure from family and friends one way or the other; distance

< previous page

page_162

next page >

< previous page

next page >

page_163

Page 163 to the nearest Welsh bookshop, etc. These factors are both very numerous and interdependent, and it would be a hopeless task to try and include them all in a production function. However, some of them can be made to fit neatly into the theoretical framework suggested here. Many of the above factors can be summarised by the generic term 'exposure'. Exposure to Welsh is higher if the individual hears Welsh at home, at school or in the streets, if Welsh is frequently used on television and radio, or if advertisements are in Welsh rather than in English. All these factors will increase the individual's proficiency, and will make reading in Welsh easier. To put it in more formal terms, the more an individual is exposed to Welsh, the more efficient that individual will be at reading it. The precise meaning of 'efficiency' in this case is that less time will be needed to get a certain amount of reading done in Welsh. We can also say that the higher the individual's efficiency (or exposure to Welsh), the higher the productivity of time used reading Welsh. The amount of reading that can actually be done in an hour also depends, of course, on how far it is to the nearest shop that sells Welsh reading matter. Here again the productivity of time is involved. Psychological pressure is likely to have a more direct effect not so much on the production function analysed here as on the shape of the utility function itself. This rather technical point will not be discussed here. However, psychological pressure may, to a certain extent, be interpreted in terms of time productivity and included in the production function. 6 Working out the precise impact of such factors is just where economists need help. For the time being, however, let us examine how such factors can be accommodated into our formal model. We have just seen that exposure to Welsh, availability of Welsh goods and, to a certain extent, the absence of psychological pressure against Welsh will increase the productivity of time devoted to 'reading in Welsh', in other words, the individual will be 'more efficient'. This generic efficiency (denoted by F) can be written into the production function as follows: [3] 7 The productivity of time in formal terms is defined as: [4] 8 The idea that productivity increases with exposure, availability or freedom from psychological pressure may be formally expressed as: [5] What has been said here about reading can be extended to all other activities, which will be to a greater or lesser degree influenced by exposure, availability or freedom from psychological pressure, and the efficiency indicator F can be included in all the production functions for activities in Welsh. To avoid unnecessary technicalities, it is helpful to reduce the

< previous page

page_163

next page >

< previous page

page_164

next page > Page 164

number of activities in the utility function. Instead of 'two times n' activities, we will keep only two general activities: 'doing things in English' and 'doing things in Welsh'. These activities will be pursued by using goods and time resources in English and Welsh (indicated below by capital letters). The utility function now reads: U = U{Ze(Xe,Te); Zw(Xw,Tw;F)}.

[6]

An individual will seek to maximise utility or satisfaction by choosing the optimal amounts of time and goods to be devoted to 'doing things in English' and to 'doing things in Welsh'. Obviously, the amounts of time and goods available to the individual are limited. First, the amount of waking time is clearly finite. It has to be apportioned between work, 'doing things in English' and 'doing things in Welsh'. Work is treated separately for two reasons. First of all, the total amount of time not devoted to work represents more than 75% of total waking life, even if the individual works 42 hours a week from age 20 to age 65 without ever taking holidays, and enjoys 5 years' retirement before dying. Besides, it is less relevant here to investigate which language will be used at work, because the individual has less influence on this. For our purposes, assume work hours are flexible. The time constraint reads: H = L + Te + Tw,

[7]

where H stands for 'total waking time available (per period)' and L for 'hours spent on the labour market (per period)'. The individual also faces a material constraint: money to buy goods and services, whether they are used to do things in English or in Welsh. These goods are bought at the going market price. The price of the aggregate 'English good' Xe is pe, while the price of the aggregate 'Welsh good' Xw is pw. The actual expenditure has to be covered by the individual's income, which is the number of hours worked (L) multiplied by the going wage rate (r). The goods constraint reads as follows: rL = peXe + pwXw.

[8]

However, since L equals total time minus the time spent 'doing things English or Welsh', constraints [7] and [8] can be combined into a single constraint and rearranged to read: rH = peXe + rTe + pwXw + rTw.

[9] 9

In other words, the total resource constraint is rH, which represents the amount of money the individual would earn if all his or her time was spent at work. This is actually not true, and the individual forgoes some income to 'do things' in English and in Welsh. The forgoing of earnings is one way to spend rH, which can be defined as 'full' or 'potential' income. The latter part of the individual's full 'income is spent on goods and services.10 We now have both an objective function and a constraint. From here on it is easy to derive 'demand functions' for the time spent doing things in

< previous page

page_164

next page >

< previous page

page_165

next page > Page 165

Welsh. These demand functions for language use have the same analytical status as traditional demand functions for television sets, clothes or food products: they define the optimal amount of time spent doing things in Welsh as a result of prices, wage rate, and efficiency, for an individual whose tastes are defined by the shape of the utility function. 11 Formally this gives us: Tw* = Tw(pe,pw,r;F).

[10]

This demand function may be used to derive comparative statics. In other words it can tell us how Tw* changes as a result of changes in pe, pw, r, and F. These results do not involve any particular hypothesis about the utility function; in other words, we do not suppose that the individual systematically prefers to do things in English or in Welsh. (If this were so, much of individual behaviour regarding language use could be attributed to tastes. This would leave us little room for explanation, since economics does not have much to say about tastes: these are mostly for sociologists and psychologists to explain.) As mentioned earlier, the only restriction imposed is that the production functions are of the Leontief type (that is, substitution between time and goods is not possible) and display constant returns. to scale. It is somewhat complicated to derive these results, and they will therefore be presented here without further demonstration.12 The Price of Welsh Goods To begin with, it is no surprise that the amount of time spent doing things in Welsh will decrease if the prices of the associated goods increase. On the other hand, if there is a decrease in the price of 'Welsh goods', such as books and newspapers, we can expect the amount of time spent doing things in Welsh to increase. However, the impact on the various activities done in Welsh will be different. If an activity involves the use of many goods and relatively little time, the amount of time used at this activity will respond relatively strongly, whereas it will be less affected in activities using less goods and more time. This can be explained by means of an example in which two activities are compared: reading in Welsh and going to a night-club in Welsh. It is clear that reading involves spending relatively little money but relatively much time; the reverse is true of going to a trendy night-club. Suppose the average price for books in Welsh goes up by 10%, and that the price of drinks in the smart night-clubwhere Welsh-speakers can be metalso goes up by 10%. Clearly the average cost of an evening out will have risen more than the average cost of an evening spent reading books in Welsh. It would not be surprising if the individual reacted as the model predicts, namely by reducing time spent going to night-clubs in Welsh more than time spent reading in Welsh.

< previous page

page_165

next page >

< previous page

page_166

next page > Page 166

This has implications for language policy. Suppose a given policy subsidises minority language goods. The price of these goods will be lower; language planners may expect minority language use to increase as a result, but the actual increase may be disappointing if subsidies have been channelled only to goods which are involved in time-intensive production processes. In other words, it may not be much use subsidising Welsh publishers since the goods they produce are used in an activity (reading in Welsh) that does not cost much money anyway; it may be more efficient to subsidise Welsh speaking night-clubs and car dealers. If working hours are not flexible, the amount of time spent on certain activities can be increased only if less time is spent on certain other activities. The effects of the subsidising policy on the use of Welsh will be positive only if the amount of things done in English actually decreases. This will be the case if doing things in English or doing things in Welsh are generally good substitutes, which implies that the individual can easily switch from English to Welsh in many activities. Alternatively, the individual may be unwilling or unable to reduce the amount of time spent doing things in English; the only way to use the increased spending power is to move resources from time-intensive to goodsintensive activities, keeping time expenditure constant. For the use of Welsh to increase, such goods-intensive activities must be available in Welsh. If this is not the case, then only activities in English will benefit from lower prices of Welsh goods. The Wage Rate A wage rate increase will raise 'full' or 'potential' income (rH) but also the amount of money forgone by doing anything other than paid work. 13 Assuming working hours are flexible, more time will be devoted to every activity, except for what economic theory calls 'inferior' activities. These are activities engaged in more when an individual is poorer than when richer. For example, buying and wearing second-hand clothes is probably an inferior activity. Among non-inferior activities, some will respond more strongly to the increase in wage rate. Higher increases will be recorded for activities that use relatively less time per unit of goods, because their opportunity cost rises by less than the opportunity cost of activities that use relatively more time. These results regarding the wage rate obtain if work hours are flexible. This hypothesis is fairly plausible in the case of women, who more frequently work part-time, and more easily leave or re-enter the labour market in response to changes in real wage rates. Suppose for a moment that the number of working hours is fixed. As before, there are two consequences: first, the amount of time devoted to doing things in Welsh can increase only if less time is spent doing things in English; second, the increased resources must be spent on material goods. The individual will therefore shift from

< previous page

page_166

next page >

< previous page

page_167

next page > Page 167

time-intensive to goods-intensive activities, engage in upper-class types of leisure, replace an old car by a more expensive one, and so on. For the use of Welsh to increase, it is necessary that goods-intensive activities in Welsh actually exist. If goods-intensive activities are predominantly conducted in English the wage rate increase will bring down the use of Welsh. This also has implications for language policy. Increasing the material standard of living in minority-language areas has often been seen as a way to keep minority languages alive. Our model shows that increasing incomes will not necessarily increase minority language use: this result is even ruled out if working hours are fixedwhich is often the caseand if no goods-intensive activities in the minority language are available. This may explain why some results of the Irish language policy have been disappointing: tax incentives have been used to attract foreign companies to the Gaeltacht areas, but there seems to be no conclusive evidence that the use of Irish has increased as a result. Goods-intensive activities are frequently, though not systematically, correlated to what economic theory calls 'superior' activities. Superior activities are the opposite of the inferior activities described above: if real income raises by n% and if a given activity then increases by more than n%, it may be called superior. A superior activity is a prestige activity, since individuals engage in it more when they become richer. This provides another rule of thumb for evaluating the probable efficiency of an income-increasing language policy: instead of asking whether some goods-intensive activities are available in the minority language, it may be as relevant to ask whether superior activities are available. Ultimately, what determines the 'inferior', 'normal' or 'superior' status of a particular activity is an individual's set of preferences, or, to put it another way, the shape of the individual's utility function. The implications for language policy are quite clear: there is no point in increasing incomes in minority-language areas unless-the people themselves view at least some minority-language activities as prestigious. Raising local incomes may therefore be a necessary condition for minority language survival, but it certainly is not a sufficient one. Exposure and Availability The comparative statics stemming from changes in exposure and/or availability provide a series of results of which only a few will be presented here. Under the set of assumptions used, it can be shown that increased exposure and/or availability will not necessarily lead to an increase in minority language use. Let us return to the specific case of Welsh. The elasticity of the amount of time spent doing things in Welsh with respect to its marginal productivity will be called W. Its value in our model is: W = hwswfw - fw + ew (swfw - fw).

< previous page

[11]

page_167

next page >

< previous page

page_168

next page > Page 168

This rather simple expression is a combination of slightly more involved variables: hw is the full income elasticity of activities done in Welsh. It is negative for an inferior activity, greater than 1 for a superior activity, and otherwise between 0 and 1. We will assume that activities in Welsh are non-inferior, so that hw will be systematically positive. 14 ew is the relative price elasticity of activities done in Welsh. It is unambiguously negative, and this simply means that if the cost of doing things in Welsh increases relative to other activities, fewer activities will take place in Welsh. sw is the weight of the expenditure on activities in Welsh in the individual's total expenditure. Its value is positive, and will be higher in the heart of Welsh Wales than in 'zones of collapse' where many activities are already conducted in English.15 fw is the time-intensiveness of activities done in Welsh. It is higher for reading books and lower for going out to a smart night-club. Since all activities involve the use of at least a little time, fw is always positive. Obviously, W may be positive or negative: increased exposure and/or availability can increase or reduce the amount of time spent doing things in Welsh. We can derive W with respect to the four variables hw, ew, sw and fw to evaluate their respective influence on the efficiency of an exposure-increasing language policy. W/ hw is positive. In other words, the better the image of Welsh, the more efficient increased exposure will be. W/ ew is negative. This means that time spent doing things in Welsh is more likely to increase if it is used in activities that are very sensitive to costs. W/ sw has an ambiguous sign. This means that there is no telling if increased exposure will prove more efficient in 'zones of collapse' or in areas where the language is still widely spoken. Increased exposure will tend to be more efficient in the latter if full income elasticity is high, and/or if the absolute value of relative price elasticity is high. W/ fw also has an ambiguous sign. Higher average time-intensiveness of things done in Welsh will boost the use of Welsh if the absolute value of relative price elasticity is high. Since fw is positive, the necessary and sufficient condition for increased exposure to have a positive influence on language use simplifies to: hwsw + ew (sw-1) > 1.

[12]

This condition may be interpreted as follows: the income effect plus the relative price effect ('the individual does more things in Welsh') must

< previous page

page_168

next page >

< previous page

page_169

next page > Page 169

outweigh the production effect ('less time than before is needed to do the same amount of things in Welsh'). This may not be the case, and increased exposure is therefore not certain to lead to an increase in language use, even if it succeeds in raising the proficiency of Welsh-speakers. 16 I believe a strong case can still be made in favour of exposure-increasing language policies; however, it is based on causations not taken into account in the present version of this model. Indeed our model captures proficiency only in terms of time productivity, but higher proficiency resulting from increased exposure is liable to increase the productivity of 'Welsh goods' as well. The relative cost of doing things in Welsh will therefore decline even further, inducing a net increase in time spent doing things in Welsh. Evaluation As stressed earlier, our model may be used to shed light on some of the significant causations involved, but cannot cover all of them. The six issues mentioned below seem particularly important, and some can be accommodated by alterations to the type of approach used here. The precise nature of such modifications will not be discussed, but may be put on the agenda for further research. Stable Preferences and the Rationality Hypothesis As with 99% of economic theory, the model presented here assumes that the individual's preferences do not change in the course of time. Formally, this implies that the parameters of the utility function are stable. They are, however, liable to change in response to higher exposure. This provides the strongest argument in favour of increasing the number of Welsh programmes on television and radio, and making access to Welsh goods and services easier, because this type of language policy may induce a shift in preferences away from English and towards Welsh. Such changes in tastes should not be interpreted in terms of superior or inferior activities, since the latter classification only makes sense in an environment where tastes are stable. Modelling endogenous tastes usually proves quite impracticable, and I do not believe it is the best line to follow. It may prove more fruitful to relax the rationality hypothesis, and to replace the utility maximisation paradigm by Simon's limited rationality approach. Limited rationality theory has not yet produced a set of widely recognised axioms about human behaviour. However, it pays more attention to psychological mechanisms that govern individual decisionmaking processes. In order to work in this general direction, some of the eight suggestions made by Lavoie (1983: 4041) in the specific case of languages may prove easier to integrate into future approaches.

< previous page

page_169

next page >

< previous page

page_170

next page > Page 170

Working Hours In line with Becker's (1965) theory of the allocation of time, working hours have been assumed to be flexible. It has proved easy enough to relax this hypothesis informally, but a formal model of minority language use with fixed hours of work still has to be developed. It may provide results not obtained here. Substitution Between Time and Goods Formal results allowing for substitutability between time and goods are already available. However, they are less manageable and the conclusions offered are less definite since a wider range of behavioural options is included. Before adapting such a model to the case of minority languages, it is worth making sure that substitutability between time and goods is relevant in the case of languages. It probably is not, but this point remains to be investigated. Adapting the Model to Specific Cases It may be worth applying the model to real situations in which field research has provided estimates for sw. It is also possible to restrict our attention to particular activities, so that hypotheses can be made about the value of fw. The range of possible results is likely to narrow down and to provide more precise guidelines for language policy. Aggregation In this paper I have quite casually moved back and forth between the individual and aggregate levels. This particular point needs to be looked at more carefully, if only to take a very important feature of languages into account: the higher the number of speakers, the more valuable it is to be able to speak the language concerned. Language Quality Language quality is not featured as such in the model. It only helps explain language quantity evaluated in time units. However, the phenomenon of language quality may deserve more attention. There is indeed a clear difference between communication that dispenses with borrowings from the majority language, and communication so heavily influenced by the latter as to include complete sentences in it. This particular problem can be linked to archaic usage and word creation. More generally, the causations selected here may need further analysis in order to reflect more accurately the actual weight of certain influences on

< previous page

page_170

next page >

< previous page

page_171

next page > Page 171

language use. I believe this task may be an area of fruitful collaboration between economics and other social sciences. Summary and Conclusion In this paper I have provided a brief survey of existing contributions to the economics of language, noting that most of these focused on specific issues relating language use to some traditional economic variable. One striking point is that the case of minority languages as we know them in Western Europe today does not seem to have been approached from an economic perspective. By focusing less on economic variables and more on fundamental economic reasoning, it is possible to suggest a model of language use that sheds light on some of the causations involved. The results obtained can help us evaluate the effectiveness of language policies. As such, subsidising goods used in minority language activities will prove efficient if they are rather expensive anyway; for example, subsidising books in the minority language will not make much difference. Increasing incomes in minority language areas is not necessarily the answer: along with improved financial opportunities, an individual needs ways to enjoy higher spending power in the minority language. Increasing exposure to the minority language is on the whole likely to have a positive influence on language use; the effect may be disappointing, however, if there is a lack of attractive things to do in the minority language, namely activities that respond more than proportionately to full income increases. For any of the above policies to work, the minority language needs to have a sufficiently good image. Any language policy that provides money, but avoids sincere commitment to boosting the image of the language, is therefore likely to fail. There seems to be no way around this: for a minority language to survive, its image must be positive. Acknowledgements The author wishes to thank Kevin Cook for revising the manuscript. Notes 1. It might be sounder economic theory to introduce a further distinction, and to ask why people are willing to use a minority language, since our model actually defines a demand function for language use; actual language use would result from the combination of supply and demand, where supply denotes the availability of real situations when the language can be used. This point will not be discussed here, and we shall assume that 'supply' will meet demand no matter where demand is located. In technical terms, we could say that supply is perfectly elastic.

< previous page

page_171

next page >

< previous page

page_172

next page > Page 172

2. French is a minority language in Canada, but a majority language in Quebec. Whether the Province of Quebec or all of Canada should be used as a reference area is a matter open to discussion. In the same way, it is not satisfactory to describe Marathi (which is spoken by some 77% of the population of the State of Maharashtra) as a minority language, even though it is spoken by a minority of Indian citizens. 3. A clear criterion would be to restrict attention to languages that still have monoglot speakers. The drawback then is that there would scarcely be any European minority languages left to study. 4. Simpson (1981: 235237) suggests a list of criteria defining minority languages. However, as some of these criteria are very general while others are quite specific, they are not suited to our purposes; they may nevertheless be used to adapt the model in the following sections to more specific cases. 5. The case of Romany is more difficult since it cannot be assigned to any reference area. 6. This is less obvious for 'reading' than it is for other activities, such as socialising. Suppose for instance that the individual goes out to a pub to meet people. If there is severe group pressure not to speak Welsh because of the negative image of the language, other customers will abstain from using Welsh. Should the individual insist on socialising in Welsh, on average more time will be needed to meet someone that agrees to speak the language. Here again the productivity of time will be adversely affected. 7. To keep the model simple, we will assume that time and goods are used in fixed proportions. In other words, it would not make sense for the individual to attempt to produce more 'reading' by substituting goods for time. 8. To keep things simple, the distinction between average and marginal productivity will not be made here. Assuming a constant returns production function, both measures will be equivalent at the equilibrium point. 9. The reader is referred to Becker (1965) or Michael (1972) for the mathematical treatment of this particular kind of model. 10. The two constraints could also have been kept separate by assuming fixed hours of work. 11. These demand functions are homogeneous of degree 0 in prices and wage rate. 12. The mathematical steps involved can be found in Grin (1989: 297314). 13. We are referring to an increase in real wage rate: the pay rise must not be cancelled out by inflation. 14. If activities in Welsh turn out to be inferior, no increase in the use of Welsh can be expected from increased exposure, from subsiding Welsh goods or from better material prospects in minority-language areas. The first goal of language policy should then be to boost the language's image, so that the individual's utility function changes, causing the prestige of things done in Welsh to increase. 15. See Ambrose & Williams (1981) for a definition of the 'zone of collapse'. 16. If the utility function is first-degree homogeneous, W = 0. On the other hand, should the hypothesis of fixed working hours be dropped, increased exposure would produce results similar to a decrease in the price of Welsh goods. References Ambrose, J.E. and Williams, C.H. (1981) On the spatial definition of 'Minority': Scale as an influence on the geolinguistic analysis of Welsh. In E. Haugen et al. (eds) Minority Languages Today. Edinburgh: Edinburgh University Press. Becker, G. S. (1965) A theory of the allocation of time. The Economic Journal 75, 493517. (1976) The Economic Approach to Human Behavior. Chicago: University of Chicago Press. Breton, A. (1964) The economics of nationalism. Journal of Political Economy 62, 37686. (1978) Nationalism and language policies. Canadian Journal of Economics 11, 65668. Breton, A. and Mieszkowski, P. (1979) The economics of bilingualism. In W.E. Oates, (ed.) The Political Economy of Fiscal Federalism. Lexington: Lexington Books. Carr, J. (1976) Le bilinguisme au Canada: l'usage consacre-t-il l'anglais monopole naturel?. In F. Vaillancourt, (ed.) (1985) Economie et langue. Montréal: Editeur Officiel du Québec.

< previous page

page_172

next page >

< previous page

page_173

next page > Page 173

Grenier, G. (1982) Language as human capital. Theoretical framework and application to Spanish speaking Americans. Princeton: PhD Dissertation, Princeton University. Grin, F. (1989) L'impact de l'efficacité personnelle sur l'allocation du temps. Geneva: PhD Dissertation, Université de Genève. Hocevar, T. (1975) Equilibria in linguistic minority markets. Kyklos 28, 33757. Lavoie, M. (1983) Bilinguisme, langue dominante et réseaux d'information. L'actualité économique 59, 3862. Marshak, J. (1965) Economics of language. Behavioral Science 10, 13540. Michael, R.T. (1972) The Effect of Education on Efficiency in Consumption. N.B.E.R. Paper No. 116. New York: Columbia University Press. Migué, J.-L. (1970) Le nationalisme, l'unité nationale et la théorie économique de l'information. Revue canadienne d'économique 3, 18398. Raynauld, A. and Marion, G. (1972) Une analyse économique de la disparité inter-ethnique des revenus. Revue économique 23, 119. Sabourin, C. (1985) La théorie des environnements linguistiques. In F. Vaillancourt, (ed.) Economie et Langue. Montréal: Editeur Officiel du Québec. Simpson, J.M. (1981) The challenge of minority languages. In E. Haugen et al. (eds.) Minority Languages Today. Edinburgh: Edinburgh University Press. Vaillancourt, F. (1985) Les écrits en économie de la langue. In F. Vaillancourt (ed.) Economie et Langue. Montréal: Editeur Officiel du Québec.

< previous page

page_173

next page >

< previous page

page_i

next page > Page i

Fourth International Conference on Minority Languages Vol. I: General Papers

< previous page

page_i

next page >

< previous page

page_175

next page > Page 175

Summing Up Anders Ahlqvist Helsingfors Universitet and Coláiste na hOllscoile, Gaillimh As a way of beginning this summing up, 1 I should like to quote to you something that Marshal Mannerheim of Finland said, greeting soldiers returning from one of the far too frequent wars that Finland has been forced to be involved in: 'Olen taistellut monilla tantereilla, mutta en ole vielä nähnyt vertaisianne sotureita,'2 'I have fought on many battlefields, but never yet seen soldiers like you'. This applies, it seems to me, to this conference, which really has been a very extraordinary occasion. This time, we have had a proper theme for the conference. Earlier this morning, we saw Howard Giles and John Edwards addressing the theme very expertly indeed. But we started our work by hearing Joshua Fishman reminding us of some very fundamental matters, especially the importance of the home environment, which may sometimes feel so self-evident that we occasionally lose track of it. More generally, I have been rather struck, over these last few days, by the very civilised fashion in which we have been able to conduct our business. This is not unimportant, since many issues relating to minority languages can be rather emotive. It is salutary to remember that we are ourselves a rather peculiar, not to say queer mixture. On the one hand, we are scholars (most of us, at any rate), who are trying quite hard to assess and analyse difficult problems in a detached fashion. At the same time, most of us happen to be, ourselves, speakers of at least one of the languages that we are trying to study, so that we may feel certain emotions about them. Certainly, it would be dishonest of me not to confess that I do not myself, as a native speaker of the Swedish of Finland, or as a learner of Irish,3 always find it terribly easy to look at things objectively. Stiff, we have been able to present our different viewswe certainly have many differing opinons about various mattersin a very courteous fashion. After Barbara Boseker's most interesting and very lively paper on the 'US English' movement, for instance, Yvo Peeters was able to tell us about his positive reactions to this movement, in spite of the fact that probably very few of us actually share them. Nevertheless, we listened to and pondered his views. By its nature, a conference like this has to be interdisciplinary, as John Edwards has emphasised. Thus, we have had quite a number of papers stressing the importance of history in what we are trying to achieve. This

< previous page

page_175

next page >

< previous page

page_176

next page > Page 176

is of course rather common-place: everybody knows this, but simple historical facts do need to be re-affirmed from time to time. This came out very well in Pádraig Ó Riagáin's' lucid account of the situation of the Irish language. I was of course also highly interested to hear about the history of the Estonian-speaking communities in Siberia, which Jüri Viikberg presented such a fine account of. In this connection, I think that it is appropriate to put on record how delighted we all were that our two Estonian colleagues undertook what for them actually was a very arduous journey to come here. We have seen all kinds of different approaches to our studies, sociological, historical, linguistic and even one relating to the science of economics. As far as linguistic matters are concernedby this I mean matters pertaining to the discipline known as linguisticswe have had very good and very interesting papers. Obviously, as an Irish scholar myself, I liked to hear Tina Hickey's detailed and careful and (on the whole) very correct analysis of some patterns in the way certain lexical items are acquired in Irish. We have also had a very interesting plenary paper by Germen de Haan about the languages of this part of the world. This was for me, at any rate, specially valuable in that it actually told me quite a lot about them, qua languages. I know that his conclusions are controversial among people who are specialists in the languages concerned, so that is probably safest for me to leave the specifics at that. On the other hand, in a more general way, I may perhaps point out that it seems to me that features like syntactic interference and the like are actually quite common in the languages of the world. Of course, there is a problem of definition involved here: if one says that something that goes on in one language triggers something that is 'latent' in another language, is this then interference or not? To my mind, it probably is, as a few features from the languages of Europe will demonstrate. Thus, most languages in Western Europe have a verb meaning 'to have', which many of these languages have managed to grammaticalise into an auxiliary. Outside the area in question and even on its fringes (as for instance Finnish and Insular Celtic will tell us), such a verb is rarely to be found. It also seems quite clear that this feature has spread from one European language to another: it is not shared by them for genetic reasons. The same applies to the definite article, which again is a feature shared by many European languages (and a few others around the world), but which Proto-Indo-European did not have, as proven by its absence from most of the older Indo-European languages and even some modern ones. It is a feature that can spread from one family of languages to another, as for instance in present-day urban Finnish, which is in the process of acquiring a definite article (in spite of what language planners may wish). Then we have the situation (cp. Henry, 1958 for a good introduction) of Hiberno-English (with many syntactic features derived from Irish that are not easily explained if compared to British English only). Lastly, it is useful to refer to the much freer word

< previous page

page_176

next page >

< previous page

page_177

next page > Page 177

order in the Swedish of Finland (if compared to that of Sweden), when discussing 4 the phenomenon of linguistic interference. As mentioned earlier, most of us are not only scholars looking at linguistic phenomena in the same detached way that scientists seek to apply to natural phenomena. We have emotions about the languages we speak and we cannot but show them, in spite of our scholarly aspirations. This came out rather clearly in the most interesting discussions we have had about the notion of 'linguicism'. John Edwards has, to be sure, provided us with some valuable remarks about the perils of allowing too many unnecessary neologisms to befuddle our work. This one probably does not come into that category, because it seems to me that Tove Skutnabb-Kangas and Robert Phillipson have been very helpful in telling us what linguicism is, how it may work and how it probably affects many of us in various ways. This is of course a touchy area (politically speaking): it is one thing to identify something that is happening to oppress minority languages and perhaps even majority ones, but once you have identified the process, is it very easy to find the root causes of it? Are we watching an evil conspiracy at work or are we just dealing with mere human unthinkingness and stupidity? The latter would to me seem to be behind some minor manifestations of linguicism, such as one in itself rather small thing that happens in Finland, in spite of otherwise fairly successful efforts towards making bilingualism visible. I am referring to the way some airports have managed to suppress the Swedish language almost completely, at least from being visible to the public, by the simple expedient of pretending that all the placenames on the various notice-boards etc. are in English (which is of course deemed to be the language of international air travel). In spite of covering much ground, Tove Skutnabb-Kangas did not allow herself the time necessary to tell us about the details of what is wrong with the Swedish education system: perhaps we shall have the pleasure of finding out about it at a future conference. At this point, we may turn, very briefly, to some fundamental issues. In an informal sort of way, we all of course know what a minority language is; in an equally informal way, we probably even know what a language is (as opposed to a 'dialect'), but exact definitions unfortunately still seem to elude us. This also applies to the accepted dogma,5 among participants at this conference, at any rate, according to which all languages are equal (in the sense of being linguistically speaking equally functional). Yet we all know that very few people out there in the real world actually would agree with us. We also know that language planning is a thriving industry. However, there is a contradiction here, for if we believe that all languages are equally functional, communicatively speaking, then, what is the point of trying to improve on them? That brings me to another normative aspect of our work. Law and minority languages belong together' in a way that has crucial implications for policy-making. Human society being what it is, we have to regulate our

< previous page

page_177

next page >

< previous page

page_178

next page > Page 178

activities. We have had some very valuable contributions to these questions, notably Boseker's paper already mentioned. Joseph Turi's contribution to our understanding of what is being and has been done in Quebec was also very useful, because some rather radical andto an outside observerperhaps even rather extraordinary things (for a democratic society, that is) have actually happened there. It will be very interesting to see how they work out in practice: in Canada a French-only policy applies in part of the country and in the United States, an English-only policy has been proposed (the political backgrounds are different, however). This then brings me to a different but related matter, which is that of the conference language(s) to adopt. Some of us find it a bit contradictory that we are dealing with minority languages, yet the vehicle for so doing happens to be what now undoubtedly is the majority language of the world, or so it would appear at this gathering, at any rate. Our very kind and good publisher and editor prefer papers in the Journal of Multilingual and Multicultural Development to be written in English only. I am wondering whether the time might not have come for the two other international congress languages to be allowed: I am of course referring to French and German. I hope that this modest proposal will be taken seriously, since its adoption would, in the case of the previous conference at any rate, have made one editorial mistake in the published proceedings unnecessary. In this connection, I must mention that, at the time of giving this summing up, a certain review (Williams, 1988) of those books (Mac Eoin, Ahlqvist and Ó hAodha (eds) 1987ab) is known to me only from a reference made to it in Giles's paper, so that I cannot comment 6 on it here at the conference. With that, it is time to return to Friesland. We have been looked after most excellently by our hosts and I hope that some of us have been able to give our hosts something in return. If, for instance, this conference were to prompt officials from this country to go and look at roadsigns in Wales, for instance, that might be quite helpful to those who would like to improve the visibility of the Frisian language. Something must also be said about the way in which the conference has been organised. I can only say that it seems to me that it has been absolutely superb. I know that Durk Gorter has been very strict with us about time; very few of us have (in spite of a few quite blatant attempts) got away with speaking for too long. I should like to stress that this has been entirely for our own good. Having a conference in a location like this one seemed like an unexpected choice, but I am sure all will agree with me and with Mikael Reuterwho spoke eloquently about this at the stupendous banquet last nightthat it was an absolutely wonderful idea, in all ways. The Hotel Management School staff and students have performed magnificently in looking after us all so well. Also, the visit to the Frisian Academy was very valuable, in that it gave us a very welcome opportunity of seeing how a small academic institution may be properly run. Together, the Academy and the School have performed superlatively welland I am trying not to be excessively

< previous page

page_178

next page >

< previous page

page_179

next page > Page 179

recursive in my use of superlatives! Thus, perhaps an ignorant foreigner may be permitted a quiet meditation about a day dawning, when these two fine bodies might perhaps come together to form a new Universiteit fan Fryslân. There certainly is a very fine nucleus for this purpose, here in Ljouwert. With these few words, I have attempted to sum up our activities and, on the behalf of all participants, I now wish to thank our Frisian friends formally but, above all, as cordially as I can, for all their very disciplined and extremely hard work to make this conference such a wonderful success. Notes 1. I have edited it from a tape of the talk itself, adding a few references and footnotes, where I felt them to be appropriate. Also, I have to put on record my gratitude to the Academy of Finland, which enabled me to attend the Conference and deliver this paper. 2. This comes from an order of the day issued at the end, in March 1940, of the Winter War, and published two years later, in Finnish (Mannerheim, 1942a: 38), and Swedish (1942b: 38). 3. Especially one that has come to use this particular minority language a great deal in everyday life. 4. In this connection, I wish to draw attention to the following wise words (Lindström, 1888:171 [quoted by Laurén, 1985: 107]): 'Man kan ej gärna tänka sig, att inom ett område, där tvänne språk, lika berättigade och lika aktade, ständigt och om hvarandra ljuda, ej dessa språk skulle starkt influera på hvarandra såväl med hänseende till ordförråd som än mer till satsbyggnaden. Och emot detta inflytande skall man förgäfves i längden kämpa'. See further Hansén, 1987: 80 and Strömman, 1987 for valuable accounts of the present-day situation. 5. Cp. Markey, 1987 for suggestions against this. Judging by one comment (Williams, 1988: 1734), his views seem to be falling on rather deaf ears. 6. Since then, I have naturally made it my business to read it, but I can think of no adequate comments on the sociological theory (if that is what it is) involved. On the other hand, only very sloppy scholarship can have allowed its author to turn such a frequent blind eye to the correct spellings of many proper names, offensively misprinted in the review. References Hansén, S.E. (1987) Mother-tongue teaching and identity: The case of the Finland-Swedes. In G. Mac Eoin, A. Ahlqvist and D. Ó hAodha (eds) (1987a) pp. 7582. Henry, P.L. (1958) A linguistic survey of Ireland: Preliminary report. Lochlann 1, 49208. Laurén, C. (1985) Normer för finlandssvenskan. Helsingfors: Holger Schildts. Lindström, K. (1888) Huru bör undervisningen i modersmålet förhålla sig gent emot de afvikelser från riksspråket, hvilka kunna påvisas i den finländska svenskan? Tidskrift utgifven af Pedagogiska föreningen i Finland 25, 16776. Mac Eoin, G., Ahlqvist, A. and Ó hAodha, D. (eds) (1987a) Third International Conference on Minority Languages: General Papers. Clevedon & Philadelphia: Multilingual Matters. (eds) (1987b) Third International Conference on Minority Languages: Celtic Papers. Clevedon & Philadelphia: Multilingual Matters. Mannerheim, C.G. (1942a) Puhtain asein. Sotamarsalkka Mannerheimin päiväkäskyjä vv. 19181942. Helsinki: Otava.

< previous page

page_179

next page >

< previous page

page_180

next page > Page 180

(1942b) Med rena vapen. Fältmarskalk Mannerheims dagorder 19181942. Helsingfors: Holger Schildts. Markey, T.L. (1987) When minor is minor and major is major: Language expansion, contraction and death. In G. Mac Eoin, A. Ahlqvist and D. Ó hAodha (eds) (1987a) pp. 322. Strömman, Solveig (1987) Trade slang as a manifestation of language-group relations at work places. In G. Mac Eoin, A. Ahlqvist and D. Ó hAodha (eds) (1987a) pp. 95101. Williams, G. (1988) Review of Mac Eoin, Ahlqvist and Ó hAodha (eds) (1987ab) Language Culture and Curriculum 1/2, 16978.

< previous page

page_180

next page >

< previous page

page_181

next page > Page 181

List of Participants of the Conference Ahlqvist, Anders, Room 512, Tower 1, University College, Galway, Ireland. Aizpurua, Xabier, Hiskuntz: Politikarako Idazkaritza Nagusia, Duque de Wellington, 2, 01011 Vitoria-Gasteiz (Basque Country), Spain. Alkain, Myriam, Hiskuntz: Politikarako Idazkaritza Nagusia, Eusko Jaurlaritza, Duque de Wellington, 2, 01011 VitoriaGasteiz (Basque Country), Spain. Alkan, Tove, Villa Aavold, Madserud allé 1., 0268 Oslo 2, Norway. Alkema, Hinka, Woerdes 16, 9407 CG Assen, Netherlands. Anderson, Alan B., Dept. of Sociology, University of Saskatchewan, Saskatoon, Sask., Canada. Araújo Carreira, Maria Helena, 9 Rue de Sèvres, app 242, 92100 Boulogne, France. Bahat, Esther, 13, Hakalanit st, Ramat Hasharon, 47243, Israel. Bakker, Marjan, Lieuwenburg 100, 8925 CL Ljouwert, (Fryslân) Netherlands. Bilash, Borislaw, 209 Scotia Street, Winnipeg, Manitoba, R2W 3X2, Canada. Bodi, Marianne, Dept. of Teacher Education, Footscray Institute of Technology, P.O. Box 64, Footscray, Vic. 3011, Australia. Boersma, Tjerk, Hoeveweg 18, 8394 VR De Hoeve, (Fryslân) Netherlands. Boseker, Barbara J., Education Dept., Moorhead State University, Moorhead, Minnesota, 56560, USA. Bosma, Meine, Fryske Akademy, Doelestrjitte 8, 8911 DX Ljouwert (Fryslân), Netherlands. Bostock, William, University of Tasmania, G.P.O. Box 252 C, Hobart, Tasmania 7001, Australia. Bot, Kees de, K.U.Nijmegen, Vakgroep Toegepaste Taalwetenschap en Methodologie, Postbus 9103, 6500 HD Nijmegen, Netherlands. Brezigar, Bojan, 1-34011 Aurisina 150, Trieste, Italy. Brijnen, Hélène, Hugo de Grootstraat 9, 2311 XJ Leiden, Netherlands. Childs, Trudy, Fryske Akademy, Doelestrjitte 8, 8911 DX Ljouwert (Fryslân), Netherlands. Coerts, Jane, Inst. voor Algemene Taalwetenschap, Universiteit van Amsterdam, Postbus 19188, 1000 GD Amsterdam, Netherlands. Comet i Codina, Robert, 41 Route de Cartigny, 1236 Cartigny, Geneva, Switzerland. Craen, Pete van de, Vrije Universiteit Brussel, Germaanse Filologie, Pleinlaan 2, 4050 Brussel, Belgium. Daan, Jo, Bremstraat 8, 7244 BL Barchem, Netherlands. Deprez, Kas, J.Wauterstraat 28, B-3200 Kessel-lo, Belgium. Edwards, John, Dept of Psychology, St Francis Xavier University, Antigonish, N.S. B2G 1CO, Canada. Eisenga, Jan, Provinsje Fryslân, Postbus 20120, 8900 HM Ljouwert (Fryslân), Netherlands. Erasmus, Hans, Laan van Oostenburg 40, 2271 AP Voorburg, Netherlands. Escure, Genevieve, Dept of English, University of Minnesota, 207 Lind Hall, 207 Church Street S.E., Minneapolis, Minnesota 55455, USA. Faddegon, Gerard Herman, De Savorin Lohmanstraat 21, 1064 LZ Amsterdam (Slotermeer), Netherlands. Falkena, Roel, Haadstrjitte 4, 9244 CN Beetstersweach, (Fryslân) Netherlands. Fase, Willem, Erasmus Universiteit Rotterdam, Postbus 1738, 3000 DR Rotterdam, Netherlands. Feitsma, Anthonia, De Boeijer 23, 9001 JJ Grou, (Fryslân) Netherlands. Fiilin, Ullamaija, Huvilarinne 32, 02730 Espoe, Finland. Fishman, Joshua A., 3340 Bainbridge Ave, Bronx, New York 10467 NY, USA. Fokma-Brouwer, A.J., Lytse Heechstrjitte 8, 8951 HE Ljouwert, (Fryslân) Netherlands. Garmendia, Maria Karmen, Hiskuntz Politikarako Idazkaritza Nagusia Eusko Jaurlaritza, Duque de Wellington, 2, 01011 Vitoria-Gasteiz (Basque Country), Spain. Gildemacher, Karel, Vegelinswei 17, 8501 ZC De Jouwer, (Fryslân) Netherlands.

Giles, Howard, Communication Studies, UCSB, Santa Barbara, CA 93106, USA. Goot, Auke van der, Provinsje Fryslân, Postbus 20120, 8900 HM Ljouwert, (Fryslân) Netherlands. Gorter, Durk, Fryske Akademy, Doelestrjitte 8, 8911 DX Ljouwert, (Fryslân) Netherlands. Graaf, Tseard de, Vakgroep Taalwetenschap, Rijksuniversiteit Groningen, Grote Rozenstraat 31, 9712 TG Groningen, Netherlands.

< previous page

page_181

next page >

< previous page

page_182

next page > Page 182

Grin, François, C.R.D.E. Université de Montréal, C.P. 6128, Succ. A, Montréal, Québec H3C 3J7, Canada. Grover, Mike, Multilingual Matters Ltd, Bank House, 8A Hill Road, Clevedon, Avon BS21 7HH, United Kingdom. Gunther, Wilf, ''Hillcrest", Puddington, EX16 8LW, United Kindom. Haan, Bouke de, Ministerie van Onderwijs en Wetenschappen, Postbus 25000, 2700 LZ Zoetermeer, Netherlands. Haan, Germen de, Vakgroep Linguïstiek, RUU, Trans 14, 3512 JK Utrecht, Netherlands. Hajer, Maaike, Joh. v.d. Waalsstraat 1, 1098 PH Amsterdam, Netherlands. Hansén, Sven-Erik, The Faculty of Education, Abo Akademi, Kyrkokoesplanaden 1214, SF-65100 Vasa, Finland. Haruyama, Shoichi, Välkkylä 1B 135, 90100 Oulu, Finland. Herberts, Kjell, SF-66580 Kuni, Finland. Hickey, Tina, Institiúid Tangeolaífochta Eireann, 31 Plás Mhic Liam, Baile Atha Cliath 2, Ireland. Hoekstra, Ruurd, Gauke Boelensstrjitte 116, 9203 RS Drachten, (Fryslân) Netherlands. Hoekstra, Jarich, Fryske Akademy, Doelestrjitte 8, 8911 DX Ljouwert, (Fryslân) Netherlands. Hoiting, Nini, Royal Insitute for the Deaf, Rijksstraatweg 63, 9752 AC Haren, Netherlands. Hughes, Medwin, The Welsh Department, Trinity College, Carmarthen, Dyfed, SA31 3EP Wales, United Kingdom. James, Allan, Debussystraat 92, 1817 GM Alkmaar, Netherlands. Jansma, Lammert G, Fryske Akademy, Doelestrjitte 8, 8911 DX Ljouwert (Fryslân), Netherlands. Jelsma, Gjalt H., p/a Fryske Akademy, Doelestrjitte 8, 8911 DX Ljouwert (Fryslân), Netherlands. Johnston, Tom, Govert Flinckstraat 246 hs, 1073 CD Amsterdam, Netherlands. Jones, Thomas Prys, Hafren, 8, Bedd Gwenan, Llandwrog, Caernarfon, Gwynedd, (Wales), United Kingdom. Jones, Glyn Evans, Dept of Welsh, Coleg y Prifysgol, PO Box 78, Caerdydd CFI 1XL (Wales), United Kingdom. Jong, Sikko de, Ynspeksje basisûnderwiis, Willemskade 27, Ljouwert (Fryslân), Netherlands. Jonker Roelants, J.J., Ministerie van Buitenlandse Zaken, Postbus 20061, 2500 DB 's-Gravenhage, Netherlands. Jonkman, Reitze, J., Fryske Akademy, Doelestrjitte 8, 8911 DX Ljouwert (Fryslân), Netherlands. Jorgensen, J. Normann, Dept of Danish, Royal Danish School of Educational Studies, Emdrupborg, DK-2400 Copenhagen NV, Denmark. Káretu, Timoti, Te Taura Whiri i te Reo Máori, P.O. Box 411, Wellington NZ, New Zealand. Kikstra, Atty, Prof. Vitringstrjitte 24, 8801 DL Frjentsjer, (Fryslân), Netherlands. Kodama, Hitoshi, 1758-62 Kinosaki, Noda-shi, Chiba, Japan. Koplewitz, Immanuel, 20 Bet Karerem St. 96343 Jerusalem, Israel. Lainio, Jarmo, Finsk-Ugriska Inst. Uppsala Universitet, Box 513, S-75120 Uppsala, Sweden. Langevelde, Ab van, Provinsje Fryslân, ôfdieling ûndersyk, Postbus 20120, 8900 HM Ljouwert (Fryslân), Netherlands. Liemburg, Jehanneke, Provinsje Fryslân deputearre, Postbus 20120, 8900 HM Ljouwert (Fryslân), Netherlands. MacKinnon, Ken, The Hatfield Polytechnic, P.O. Box 109, Div. of Social Sciences, College Lane, OANE, Hatfield, Herts, United Kingdom. Mägiste, Edith, Department of Psychology, Uppsala University, Box 227, 751 04 Uppsala, Sweden. Majewicz, Alfred F., Japanologisches Seminar der Universität Bonn, Regina Pacis Weg 7, D-5300 Bonn 1, WestGermany. McNamara, Tim, Dept. of Russian and Language Studies, University of Melbourne, Parkville, Victoria 3162, Australia. Meestringa, Theun, S. Veldstraweg 50, 9833 PB Den Ham (Gn), Netherlands. Misiego, Rose, A. Spazierhof 29, 1065 SJ Amsterdam, Netherlands. Muller, Migiel, Wynzerdyk 25, 9062 GP Oentsjerk, (Fryslân) Netherlands. Neldee, Peter H., O.V.M., Vrijheidslaan 17, B1080 Brussel, Belgium. Nicholas, Howard, La Trobe University, Bundoora Victoria, 3083, Australia. Nouwen-San Giorgi, Marina, Groene Wetering 7, 3062 PB Rotterdam, Netherlands. O'Coileáin, Antoine, Bord na Gaeilge, 7 Cearnóg Mhuirfean, Baile Atha Cliath 2, Ireland.

O'hIffearnáin, Tadhg, J504, An Roinn Staidéir Eor-paigh, Ollscoil Uladh, Cúil Raithin, BT 521 SA, Co. Dhoire (N. Ireland), United Kingdom. O'Riagáin, Pádraig, ITE, 31 Fitzwilliam Place, Dublin 2, Ireland. Oda, Masaki, Dept. of Chinese and Japanese, 432A ICC, Georgetown University, Washington, DC 20057, USA. Overduin, Hannie, Mauritsstraat 57, 3583 HJ Utrecht, Netherlands. Ozolins, Uldis, Dept. of Language and Culture Studies, Victoria College, Toorak Campus, P.O. Box 224, Malvern, 3144, Victoria, Australia. Paulsen, Frederik, D-2271 Alkersum, Föhr, Nordfriesland, West-Germany. Peeters, Yvo, Lenoirstraat 13, 1090 Brussel, Belgium.

< previous page

page_182

next page >

< previous page

page_183

next page > Page 183

Phillipson, Robert, Tronninge Mose 3, 4420 Regstrup, Denmark. Pieters, Wilfried, Kikvorsstraat 955, B-9000 Gent, Belgium. Priestly, Tom, Dept. Slavic & East European Studies, University of Alberta, Edmonton, Alberta T6G 2H1, Canada. Prieto, Domingos, Inst. ITT/ATW, Faculteit der Letteren RUG, Grote Kruisstr 2-1, 9712 TS Groningen, Netherlands. Raatgever-Fraser, Mavis, Granaathorst 29, 2592 SN 's-Gravenhage, Netherlands. Rannut, Mart, Insitute of Language and Literature, Lauristini 6, Tallinn 200106 Estonia, USSR. Reuderink, Ronald, Binnenhôf 19, 8802 MD Frentsjer, (Fryslân) Netherlands. Reuter, Mikael, Svenska sprakbyran, Fabiangatan 7B, SF-00130 Helsingfors, Finland. Riemersma, Aleks, P.K.F., Provinsjehûs, Postbus 20120, 8900 HM Ljouwert (Fryslân), Netherlands. Sala, Rafael, Modern Languages Centre, University of Bradford, Bradford, West Yorkshire BD7 1DP, United Kingdom. Schaarschmidt, Gunter, Department of Slavonic Studies, University of Victoria, Victoria, B.C. V8W 2Y2, Canada. Sharp, Derrick, Multilingual Matters Ltd, Bank House, 8A Hill Road, Clevedon, Avon BS21 7HH, United Kingdom. Sikma, Jantsje, Fryske Akademy, Doelestrjitte 8, 8911 DX Ljouwert (Fryslân), Netherlands. Skutnabb-Kangas, Tove, Tronninge Mose 3, 4420 Regstrup, Denmark. Slomanson, Peter, 137 West 78 Street, New York, N.Y. 10024 USA. Spolsky, Bernard, Dept. of English, Bar-Ilan University, 52100 Ramat-Gan, Israel. Starkenburg, Jeltje, Heemskerklaan 23, 6881 EK Velp (Gld), Netherlands. Steenwijk, Han, Slavisch Seminarium, UvA, Spuistraat 210, 1012 VT Amsterdam, Netherlands. Stern, Asher, Ulpan Akiva, P.O.B. 6086, Netanya 42160, Israel. Stolp, P. J., Kerk van Jezus Christus van de Heiligen van de Laatste Dagen, sectie vertaalbegeleiding, Postbus 224, 3430 AE Nieuwegein, Netherlands. Strömman, Solveig, Vasaesplanaden 20A, SF 65100 Vasa, Finland. Strukelj, Inka, University of Ljubljana, Institute of Sociology, Ljubljana, Yugoslavia. Tandefelt, Marika, Tavastävgen 10 A 15, SF-00530, Helsinki 53, Finland. Tomaszczyk, Jerzy, Institute of English Studies, University of Lódz, Kosciuszki 65, 90-514 Lódz, Poland. Tsow, Ming, Inner London Education Authority, The County Hall, London SEI 7PB, United Kingdom. Tuk, C., Vereniging van Leraren in Levende Talen, Schubertlaan 16, 1411 HZ Naarden, Netherlands. Turi, Joseph, Commission de protection de la langue française, 800, place Victoria, 40e étage, Montréal (Québec), Canada. Ugalde, Mikel, Hiskuntz Politikarako Idazkaritza Nagusia Eusko Jaurlaritza, Duque de Wellington, 2, 01011 VitoriaGasteiz (Basque Country), Spain. Ureland, P. Sture, In der Täsch 11, D-6906 Leimen, West-Germany. Urreiztieta, Irel, California State University, Stanislaus, College of Arts, Letters and Sciences, 801 West Monte Vista Avenue, Turlock CA 95380, USA. Viikberg, Jüri, Institute of Language and Literature, Lauristini 6, Tallinn 200106 Estonia, USSR. Vollmer, Helmut, J., Fb Sprach- und Literaturwissenschaft, Universität Osnabrück, Postfach 4469, 4500 Osnabrück, West-Germany. Vries, John de, Dept. of Sociology, Carleton University, Ottawa, K1S 5B6, Canada. Waas, Margit, 49 Lansdowne St., Surry Hills, N.S.W., 2010, Australia. Wal, Ytsje van der, Provinsie Fryslân, Ofdieling ûndersyk, Postbus 20120, 8900 HM Ljouwert, (Fryslân) Netherlands. Wellinga, Willeke, Kamgras 135, 8935 EG Ljouwert, (Fryslân) Netherlands. Wicherkiewicz, Tomasz, ul. Bulgarska 124/3, 60-381 Poznan, Poland. Williamson, May, 15 Falcon Road West, Edinburgh, EHIO 4AD, (Scotland), United Kingdom. Willkommen, Dirk, Postfach 2804, D2350 Neumünster, West-Germany. Wingstedt, Maria, Stockholms Universitet, Institutionen för Lingvistik, 106 91 Stockhom, Sweden.

Wirrer, Jan, Universität Bielefeld, Fakultät für Linguistik und Literaturwissenschaft, D-4800 Bielefeld 1, BRD, WestGermany. Wynants, Armel, Stationsstraat 108, B-3790 Voeren, Belgium. Wijnhoud, J., Ministerie van Algemene Zaken, Postbus 20001, 2500 AE 's-Gravenhage, Netherlands. Ytsma, Jehannes, Fryske Akademy, Doelestrjitte 8, 8911 DX Ljouwert (Fryslân), Netherlands. Zeldenrust, Auke, p/a Fryske Akademy, Doelestrjitte 8, 8911 DX Ljouwert, (Fryslân) Netherlands. Zondag, Koen, Mienskiplik Sintrum foar Underwiisbegelieding, Sixmastrjitte 2, 8932 PA Ljouwert, (Fryslân) Netherlands. Zondergeld, Gjalt, Herensingel 162, 1382 VV Weesp, Netherlands.

< previous page

page_183

next page >

< previous page

next page >

page_184

Page 184 Contents of Volume II Preface Durk Gorter, Jarich Hoekstra, Lammert G. Jansma and Jehannes Ytsma 1 The Reformation and the Vernacular Anthonia Feitsma

vii

1

2 Characterising a Minority Language: A Social Psychological Comparison Between Dutch, Frisian and the Ljouwert Vernacular Reitze J. Jonkman

11

3 Sweden Finnish - Development or Deterioration? Jarmo Lainio

21

4 Flemish Primary Schools in Brussels: Which Prospects? Kas Deprez and Armel Wynants

43

5 Language Conservancy, or: Can the Anciently Established British Minority Languages Survive? Wilf Gunther

53

6 Language Maintenance and Viability in the Contemporary Scottish Gaelic Speech-Community: Some Social and Demographic Factors 69 Kenneth MacKinnon 7 Migrant Pupils: Welsh Linguistic Implications Thomas Prys Jones

91

8 Minority Languages in Spain Robert Comet i Codina

103

9 A Demolinguistic Analysis of the Basque Autonomous Community Derived from the Census of 1986 115 M. Karmen Garmendia and Xabier Aizpurua

< previous page

page_184

next page >

< previous page

next page >

page_185

Page 185 10 'Our Dialect Sounds Stupid': The Importance of Attitudes to SoCalled Sub-Standard Language Codes as a Factor in the (Non)Retention of Slovene in Carinthia, Austria Tom Priestly

135

11 National Minority Languages in Media and Education in Poland Alfred F. Majewicz and Tomasz Wicherkiewicz

149

12 The Siberian Estonians and Language Policy Jüri Viikberg

175

13 The Use and Integration of Hebrew Lexemes in Israeli Spoken Arabic Immanuel Koplewitz

181

Contents of Volume I

196

Index

198

< previous page

page_185

next page >

< previous page

page_186 Page 186

Index B Bilingualism 101-118 C Canada 57-76 Cognitive 37-55 Comparative Analysis 119-136 Culture Planning 5-36 D Demolinguistics 57-76 E Ecology of Language 137-151 Economics of Language 153-173 Ethnicity 77-100 Ethnolinguistic - Minorities 119-136 - Vitality 137-151 F Finland 57-76 Frisian 101-118 G Grammatical Borrowing 101-118 I Inductive Method 137-151 Integration 77-100 Interactional 37-55 Intercultural 77-100 Interdisciplinary 37-55 Intergenerational Mother Tongue Continuity 5-36 L Language - Decline 37-55 - Maintenance 5-36, 37-55 - Planning 5-36 - Policy, efficiency of 153-173 - Revival 5-36 - Shift 57-76 - Survival 37-55 Linguicism 77-100 Linguistic Human Rights 77-100 M

Minority Language 137-151 - Use 153-173 Racism 77-100 Reversing Language Shift 5-36 S Sociolinguistics 101-118 Sociological 37-55 Sociopolitical Framework 119-136 Sociopsychological 37-55 T Theory 37-55 Typologies 137-151

< previous page

page_186