Respecting Linguistic Diversity in the European Union (Studies in World Language Problems, Volume 2)

Respecting Linguistic Diversity in the European Union

Studies in World Language Problems Studies in World Language Problems (WLP) focuses on political, sociological, and economic aspects of language and language use. It is especially concerned with relationships between and among language communities, particularly in international contexts, and in the adaptation, manipulation, and standardization of language for international use. It aims to publish monographs and edited volumes that deal with language policy, language management, and language use in international organizations, multinational enterprises, etc., and theoretical studies on global communication, language interaction, and language conflict. Published in cooperation with the Centre for Research and Documentation on World Language Problems.

General Editor Humphrey Tonkin University of Hartford

Editorial Board E. Annamalai

François Grin

Richard B. Baldauf, Jr.

Kimura Goro

Ina Druviete

Timothy Reagan

Central Institute of Indian Languages & Yale University University of Queensland University of Latvia

University of Geneva Sophia University, Tokyo Central Connecticut State University

Mark Fettes

Simon Fraser University, Vancouver

Volume 2 Respecting Linguistic Diversity in the European Union Edited by Xabier Arzoz

Respecting Linguistic Diversity in the European Union Edited by

Xabier Arzoz University of the Basque Country

John Benjamins Publishing Company Amsterdam / Philadelphia

8

TM

The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences – Permanence of Paper for Printed Library Materials, ansi z39.48-1984.

Library of Congress Cataloging-in-Publication Data Respecting linguistic diversity in the European Union / edited by Xabier Arzoz. p. cm. (Studies in World Language Problems, issn 1572-1183 ; v. 2) Includes bibliographical references and index. 1. Multilingualism--Europe. 2. Language policy--Europe. 3. European Union. I. Arzoz, Xabier. P115.5.E85R48    2008 306.44'6094--dc22 isbn 978 90 272 2833 8 (Hb; alk. paper)

2007036244

© 2008 – John Benjamins B.V. No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publisher. John Benjamins Publishing Co. · P.O. Box 36224 · 1020 me Amsterdam · The Netherlands John Benjamins North America · P.O. Box 27519 · Philadelphia pa 19118-0519 · usa

Table of contents Preface Xabier Arzoz Introduction: Respecting linguistic diversity in the European Union Xabier Arzoz

vii 1

Part 1. Factual and theoretical approaches chapter 1 Linguistic diversity as curse and as by-product Philippe Van Parijs

17

chapter 2 Linguistic diversity in the European Union: An overview Patxi Juaristi, Timothy Reagan and Humphrey Tonkin

47

chapter 3 Principles of policy evaluation and their application to multilingualism 73 in the European Union François Grin chapter 4 A one-dimensional diversity? European integration and the challenge 85 of language policy Peter A. Kraus Part 2. The protection of linguistic diversity in EU law chapter 5 Union citizenship and language rights Peter Hilpold

107

chapter 6 EC law and minority language policy: Some recent developments Niamh Nic Shuibhne

123

vi

Table of contents

chapter 7 The protection of linguistic diversity through Article 22 145 of the Charter of Fundamental Rights Xabier Arzoz chapter 8 The protection of linguistic diversity through provisions of the EU Charter other than Article 22 175 Bruno de Witte chapter 9 Languages that are official in part of the territory of the Member States:  Second-class languages or institutional recognition in EU law? Antoni Milian-Massana

191

appendix Selected provisions on language issues from EU law

231

Contributors

261

Index

265

Preface Xabier Arzoz

The origins of this book lie in a workshop organised on 24–25 May 2005 in St Antony’s College, University of Oxford, sponsored jointly by the Basque Studies Society (Eusko Ikaskuntza) and the European Studies Centre of St Antony’s College, Oxford. The organisation of that workshop was the most visible obligation during my Basque Visiting Fellowship at St Antony’s in 2004/05, which was supported by the Basque Studies Society. I am indebted to both institutions, as well as to the participants in the Oxford workshop. My special thanks go to the contributors who kindly accepted an invitation from a young scholar to produce papers for this volume. I would particularly like to thank Jeremy McClancy of Brookes University (Oxford) for his advice and support throughout the workshop. I would also like to express my gratitude to Grégoire Webber (Oxford), Patrick Carlin (Cardiff) and Niamh Nic Shuibhne (Edinburgh), who commented on and checked parts of the book. Finally, I must thank Professor Humphrey Tonkin, the general editor of the series, for his engaged and supportive editorial management in the process of preparing the manuscript for publication. At the time of the workshop, after the accession of ten new Member States in 2004, the number of official EU languages had increased from eleven to twenty. Immediately after the workshop took place, on 13 June 2005, the Council of the European Union authorised an improvement of the existing legal framework for Irish (the twenty-first official language from 2007 on) and for other languages, such as Basque, Catalan and Galician, which, according to the constitutional order of their respective Member States, are official in all or part of the territory of those Member States. Later, in September 2006, Bulgaria and Romania were definitively accepted as new Member States as of 1 January 2007, with the consequence that the number of EU official languages has again increased, to 23. These events highlight the urgency of the EU search for guidance in addressing its internal linguistic diversity. Although the European scene has continued to change rapidly, especially over the past few years (for example, the innovations concerning Irish and the use

viii Preface

of other languages, the accession of two new Member States, and the collapse of plans for the European Constitution), all the original papers have been updated with the latest developments. Subject areas not covered in the original workshop have also been added, making the publication more complete. Each chapter has been written by a recognised expert in the field. The Charter of Fundamental Rights of the European Union is cited throughout the volume in accordance with its original numeration. Its text was solemnly proclaimed by the three EU institutions in Nice in December 2000 and it preserves its value as a political declaration. The Charter provisions received a new numbering system in the Intergovernmental Conference after their incorporation as Part II of the Treaty establishing a Constitution for Europe (beginning with Article II–61). After two years of uncertainty over the Union’s treaty reform process, the European Council decided at its meeting of 21–23 June 2007 to abandon the ratification process for the Treaty establishing a Constitution for Europe. Following this mandate, the Presidency of the Council convened an intergovernmental conference for the purpose of determining by common accord the amendments to be made to the Treaties on which the European Union is founded. The “draft Treaty amending the Treaty on European Union and the Treaty establishing the European Community” (CIG 1/07) submitted to the intergovernmental conference includes the provisions and the declaration on languages that were contained in the Treaty establishing a Constitution for Europe. Consequently, these provisions and the declaration will very probably be incorporated in the future Treaty. The new “draft Treaty” also recognizes the rights, freedoms and principles set out in the Charter of Fundamental Rights of 7 December 2000, and it foresees that the Charter will have the same legal value as the Treaties. This book aspires to contribute to our understanding of the challenge of respecting linguistic diversity within the EU. It is intended as a useful introduction to the issue for those not already familiar with EU law and, at the same time, it aims to provide an analysis of the potential of the Charter of Fundamental Rights of the European Union to enhance respect for linguistic diversity. The appendices are an attempt to give access to the most basic legal norms concerning linguistic diversity within the EU institutions. October 2007

Introduction Respecting linguistic diversity in the European Union



Xabier Arzoz The rise of concern for cultural and linguistic diversity Concern for cultural diversity, of which linguistic diversity is a fundamental element, is increasingly being raised in the international arena and in the research sphere, most likely because of awareness of the far-reaching effects of globalisation. For some time, cultural diversity was an issue that concerned only the sociopolitical movements of non-dominant groups confronting the phenomenon of acculturation within the societies where they lived. The nation-state as it evolved following the French Revolution was considered the natural enemy of minorities. Now, it would seem that globalisation and regional supranational integration are emerging as processes that cause similar concern among the populations of nation-states. Paradoxically, fear of the effects of globalisation is a cause that reunites, in an odd coalition, actors that were separate or even opposed before, namely “state-led nations” on the one side, and “state-seeking nations” and ethnic groups on the other. The globalisation of the economy tends to homogenise or to sweep aside cultural obstacles to the standardisation of goods and practices required by the market, whatever the origin of those cultural obstacles, whether cultural practices and forms or legal requirements imposed by national legislators. There appear to be two complementary lines of thought behind the rising concern for cultural diversity, which have also been picked up by international organisations such as UNESCO and, to some extent, the Council of Europe. One line of thought concerns more generally the need for protection of cultural diversity as an “ethical imperative”.1 On 2 November 2001, the Universal Declaration on Cultural Diversity was adopted unanimously by the 185 member states represented at the thirty-first session of the UNESCO General Conference in Paris.2 This instrument recognises, for the first time, cultural diversity as a “common heritage of humanity” and considers its safeguarding to be an “ethical imperative, inseparable from respect for human dignity”. Article 1 proclaims:



Xabier Arzoz

Culture takes diverse forms across time and space. This diversity is embodied in the uniqueness and plurality of the identities of the groups and societies making up humankind. As a source of exchange, innovation and creativity, cultural diversity is as necessary for humankind as biodiversity is for nature. In this sense, it is the common heritage of humanity and should be recognized and affirmed for the benefit of present and future generations.

The fact that the Declaration came in the wake of 11 September 2001 should not diminish its importance: independently of the aim of promoting intercultural dialogue for the sake of peace and mutual understanding, it recognises the link between culture and human dignity and the right to enjoy and develop multiple identities, and it stresses the key relevance of human rights, including the rights of persons belonging to minorities, for preserving cultural diversity. The Universal Declaration is seen by UNESCO as the founding act of a new ethic in relation to cultural diversity. In fact, only four years after the Declaration’s adoption, on 20 October 2005, a Convention on the Protection and Promotion of the Diversity of Cultural Expressions was adopted at the 33rd session of the UNESCO General Conference.3 This instrument recognises the need to take measures to protect the diversity of cultural expressions, including their contents, especially in situations where cultural expressions may be threatened by the possibility of extinction or serious impairment. The Convention reaffirms the Parties’ sovereign right to formulate and implement their cultural policies, to adopt measures to protect and to promote the diversity of cultural expressions, and to strengthen international cooperation to achieve the purposes of the Convention: “When a Party implements policies and takes measures to protect and promote the diversity of cultural expressions within its territory, its policies and measures shall be consistent with the provisions of this Convention” (Art. 5). The Convention recognises the right of the Parties to adopt “measures that, in appropriate manner, provide opportunities for domestic cultural activities, goods and services among all those available within national territory for the creation, production, dissemination, distribution and enjoyment of such domestic cultural activities, goods and services, including provisions relating to the language used for such activities, goods and services” (Art. 6.2 (b)). From a more general perspective, it is interesting to stress here three of the eight “guiding principles” set out in Article 2 of the Convention: 1. Principle of respect for human rights and fundamental freedoms Cultural diversity can be protected and promoted only if human rights and fundamental freedoms, such as freedom of expression, information and communication, as well as the ability of individuals to choose cultural expressions, are



Introduction

guaranteed. No one may invoke the provisions of this Convention in order to infringe human rights and fundamental freedoms as enshrined in the Universal Declaration of Human Rights or guaranteed by international law, or to limit the scope thereof. 3. Principle of equal dignity of and respect for all cultures The protection and promotion of the diversity of cultural expressions presuppose the recognition of equal dignity of and respect for all cultures, including the cultures of persons belonging to minorities and indigenous peoples. 6. Principle of sustainable development Cultural diversity is a rich asset for individuals and societies. The protection, promotion and maintenance of cultural diversity are an essential requirement for sustainable development for the benefit of present and future generations.

The Convention, which will enter into force three months after the date of deposit of the thirtieth instrument of ratification, is open for accession also by any regional economic integration organization such as the EC/EU. Therefore, in the event of an accession by both the EC/EU and one or more of its Member States, there could be a distribution of responsibility for the performance of their obligations under the Convention.4 The other line of thought behind the rising concern for cultural diversity refers to the more specific field of linguistics. Since the early 1990s, there has been a universal upsurge of professional linguistic concern about the unprecedented scale of language death facing the world.5 According to estimates, half of the world’s 6,000 languages will die in the next 100 years.6 It is itself a statistic at least as impressive and worrying as the estimated damage caused by the greenhouse effect in an equivalent period of time. Part of the contemporary emphasis on the issue of endangered languages is ecological in character, focusing on the relationships between people, their environment, and their thoughts and feelings (see Crystal 2000: 25, 32–36 and Skutnabb-Kangas 2003). In this sense, it is claimed not merely that diversity is as necessary for humankind as biodiversity is for nature (as the UNESCO Universal Declaration, mentioned above, seems to argue), but that linguistic diversity is an integral part of biodiversity.7 This does not mean that linguistic diversity operates in the same way as biodiversity, or that similar forms of legal protection should apply to them. There are fundamental differences between languages and living species, as the former are socio-cultural goods. The major conceptual difference consists of transmissibility and accumulation: we can acquire a language but we cannot get wings; we can acquire many languages but, if we get wings, we lose our arms (Junyent 2002: 14).





Xabier Arzoz

However, concern about the loss of language diversity does not reach the same level of concern as the loss of biodiversity.8 Even the many socio-political movements claiming territory or rights for a language seek to replace the dominance of one language with another, and do not seek diversity for its intrinsic value (Annamalai 2004: 2). Annamalai has reflected on the political and philosophical reasons for public indifference to language diversity. First, “the technologically driven, socially stratified and centrally administered social system [is] conducive to language homogeneity.” In other words, the economic and political supremacy of certain communities is transferred to their language. But Annamalai’s second reason deserves a more careful reflection, since it associates the prevailing “western” or “modern” attitude with both nature and linguistic diversity and contrasts it with the alternative philosophical view of traditional societies. For Annamalai, “the western languages, which embody the philosophical position of subordinating nature to the quest of human progress, facilitate a way of speaking and thinking that values reduction of diversity as the means to progress through increased efficiency.” As the sociolinguist Mackey (1991: 57) rightly observes, “as for all systems, language as it spreads, tolerates fewer conflicting standards.” The process of European integration shows expansionist standardising tendencies in the field of communications, economics, social protection and justice. Is language to be handled according to the same criteria? The efficiency discourse is one we frequently hear in the context of EU multilingualism.

Concern for linguistic diversity within the EU The proportion of global linguistic diversity which attaches to Europe is comparatively small (ca. 3–4%); it is even smaller if we concentrate on the European Union and exclude the overseas countries and territories of the Member States (ca. 1%). Europeans have always been conscious that linguistic differences exist among them and between them and other peoples (Martinell Gifre & Cruz Piñol 1996, Burke 2004 and Mackey 1991: 5–61). In fact, nation-building in Europe was based primarily on linguistic terms: modern nations are basically “linguistic nations” in the sense that having a common language was the instrument and the symbol of each nation-state.9 The emergence of the European Union as a supranational entity has not modified this situation. On the contrary, concern for cultural and linguistic diversity appears to be rising precisely within the EU. It is difficult to say to what extent concern for linguistic diversity has been influenced by the aforementioned general arguments (diversity as an “ethical imperative” and as an “ecological necessity”) or by internal developments such as the processes of regionalisation in some member states, the ethnic tensions in



Introduction

central and eastern Europe and the successive enlargements of the EU. It is certain that, as in any region of the world, languages are dying and disappearing in Europe too.10 Traditionally the protection of linguistic diversity was considered to be an issue outside the European Union’s competences and tasks. In fact, the most interesting legal developments in this field, the Framework Convention for the Protection of National Minorities and the European Charter for Regional or Minority Languages, are regional instruments of the Council of Europe. These legal instruments represent the most advanced notion of minority protection existing in the world. Some recent developments illustrate that concern for linguistic diversity and for the challenges of multilingualism within the EU is increasing. As those developments are related to one another, the following exposition does not necessarily imply a causal sequence of events. First, there is a movement towards giving a more central position to the value of linguistic diversity within the EU. The Charter of Fundamental Rights of the EU, which was proclaimed in December 2000, committed the Union to respecting cultural, religious and linguistic diversity (Art. 22). Furthermore, the Treaty establishing a Constitution for Europe establishes “United in Diversity” as the motto of the Union and includes explicitly among the Union’s objectives respect for “its rich cultural and linguistic diversity” and ensuring that “Europe’s cultural heritage is safeguarded and enhanced” (Art. I–3(3)). It is not clear whether, and to what extent, a concern for linguistic diversity could be mainstreamed, on that basis, in the framework of the Union’s policies; nor is it clear whether the institutions of the Union could prompt the Member States to take that concern into account in the implementation of Union law and policies. The phrase “cultural and linguistic diversity” also appears in another provision of the Constitutional Treaty, in the context of the common commercial policy, in relation to the conclusion of international agreements in the field of trade in cultural and audiovisual services that “risk prejudicing the Union’s cultural and linguistic diversity” (Art. III–315.4 (a)). Second, the successive enlargements of the EU have caused the number of EU official languages to rise from four to twenty-three. As long as new candidates are available and willing to join the EU, the expansion of the number of EU official languages seems likely to increase accordingly. However big this number may appear, it constitutes only a tiny part of the European linguistic mosaic: the Union embodies twenty-seven Member States and 490 million people who speak over 60 European autochthonous languages. The privileged status of the appointed twenty-three “official languages” contrasts sharply with the unprivileged status of the majority of languages spoken within the borders of the EU. The contrast became sharper when languages with even smaller numbers of speakers acceded





Xabier Arzoz

to the EU. These arrangements may have the effect of touching on the self-esteem of linguistic communities whose native languages are not given equal status. At issue is the symbolic affirmation of national languages on equal terms. Malta, for example, insisted on its equal right to include Maltese among the official and working languages of the Union, although its population could also be served through English (Malta’s other national language according to its Constitution11). If Maltese is included, with Malta having 400,000 inhabitants, what should legally and politically prevent Ireland and Luxemburg from asking for the same treatment for Irish12 and Lëtzebuergesch, national languages of those Member States? However, the EU official language status is not limited to the issue of symbolic affirmation. Apart from the number of positions for translators and interpreters being created, whose main beneficiaries are nationals from the respective Member States, the very decision of translating and publishing every page of the Official Journal provides crucial chances for further development of relatively lesser-used languages, such as Greek, Swedish, Latvian, Lithuanian, Slovenian or Maltese. Languages are not all equally standardised: some languages are more standardised than others (Mackey 1991: 55). The governments of many European states, including those speaking an international language, like France or Spain, spend money on the standardization of the national language (for instance, subsidising reference works or standard encyclopaedias and creating terminology centres to plan and coordinate terminological research). Since EU legal and political acts cover relevant areas of modern life (from the environment to technical standards and new technologies), less standardized languages with EU official status obtain a free development of their language corpus, while the costs of similar efforts for standardization of languages such as Catalan, Basque or Welsh have to be assumed integrally by sub-state authorities. On the other hand, it is odd that the language arrangements of a supranational polity are based, in practice, exclusively on national interests and sovereign considerations of the Member States, without taking into account the relative size of each language and the communicative needs of EU citizens (see Milian-Massana 2004; Creech 2005: 153). It is up to the Member States to decide freely what language(s) will be accorded the status of official language(s) within their jurisdiction. Since a unanimous decision of the Council is required to determine the rules governing the languages of the EU institutions,13 the choice of each Member State automatically becomes an official language of the Union, unless, apparently, the Member State in question renounces, for whatever reasons, the inclusion of one or more of its national languages (as was the case for Ireland with regard to Irish, from its accession in 1973 until 2005, Luxemburg with regard to Lëtzebuergesch from 1984 to date, and Cyprus with regard to Turkish from its accession in 2004).



Introduction

That is why the Catalan language does not enjoy official status within the EU, although it is spoken by seven million EU citizens and therefore has more speakers than each of eight of the current EU official languages (Danish, Finnish, Slovak, Slovenian, Lithuanian, Latvian, Estonian and Maltese). Events following the recent vast enlargement of the EU demonstrate that the contrast between privileged and non-privileged status has become sharper. At the request of the Irish Government, the Council decided on 13 June 2005 to amend Regulation No 1/1958 determining the languages to be used by the European Economic Community, which, for more than forty years, had been changed only on the occasion of the successive enlargements (that is, by virtue of primary law): the aim was to accord the Irish language, the first national language of the Republic of Ireland, which had the status of a mere “language of the Treaty”, the status of “official and working language” of the Union, i.e. to accord to it the same status as that accorded to the national official languages of the other Member States.14 Nevertheless, for practical reasons and on a transitional basis, the EU institutions will not be bound to draft and translate all acts into Irish (or, for different reasons, into Maltese), but only the Regulations adopted jointly by the European Parliament and the Council.15 At the same meeting on 13 June 2005, the Council approved some conclusions to govern the use of languages that, although not included in Regulation No 1/1958, enjoy, according to the Constitution of a Member State, official status in all or some part of the respective State territory: the use of additional languages has to be approved by the Council on the basis of an administrative agreement between the Council and the particular Member State, and it will involve the mediation of a mechanism of translation for which cost and responsibility are to be assumed by that Member State. Spain has undertaken to make such agreements with the EU institutions and bodies to allow the use of Catalan, Basque and Galician before them. This solution might also be used for those languages that, although enjoying official status throughout the territory of a Member State, do not possess a number of speakers that justifies a full official and working status within the EU. In many cases, unfortunately, advocacy of linguistic diversity at the supranational level does not correspond with equal advocacy at the state and sub-state levels. Many voices support multilingualism within the EU institutions and the granting of equal status to every Member State language, while they insist on, or are in no way concerned about, cultural and linguistic homogeneity within their nation-states and the preservation of the privileged status of a national language. They are concerned about the impending demise of their national language at the supranational level, but they do not care about the impending demise of minority and regional languages at the sub-state level. There is a double standard here. As





Xabier Arzoz

Fishman rightly observes, “the states are even more reluctant to apply the principle of ethnolinguistic democracy below (or within) the level of the state than they are to set aside the principle of ethnolinguistic equality above the level of the state” (1994: 55). Moreover, the concept of limits (costs, efficiency, workability, manageability and so on) is vociferously put off in one context (the “manageable” supranational case) and prematurely applied in the other (the “obviously unmanageable” intra-state case): “The Netherlands is very certain that Nederlands [Dutch] is as good a language as English for the operation of EC, but it is not sure that Frisian is as good a language as Nederlands for the operation of local public services in Friesland” (Fishman 1994: 55). Some basic principles of fairness for conducting ethnolinguistic affairs are urgently needed to redress these examples of double standards.

The themes of this volume The essays collected in this volume take the background set out above as their departure point. The authors examine the present situation of linguistic diversity and the respect for diversity within the Union. The volume is structured in two parts. The four essays constituting the first part of the book address the issue of linguistic diversity from a multidisciplinary perspective (sociology, economics, and political science). In Chapter 1 of the book, Philippe Van Parijs examines the notion of linguistic diversity. He first distinguishes three dimensions of diversity (richness, evenness and distance) and two levels of linguistic diversity (local and territorial). Then, he analyzes the structural tensions that tend to develop between these levels. With this as background, he argues that linguistic diversity, by itself, cannot plausibly be regarded as a good, all things considered, especially because of its negative impact on the prospects for economic solidarity; but that territorial linguistic diversity will need preserving as the by-product of a concern for the equal dignity of the identities closely associated with native languages. In Chapter 2, Patxi Juaristi, Timothy Reagan and Humphrey Tonkin offer a wide panorama of linguistic diversity within the European Union. European languages are divided into different categories from “big” and “medium” languages to endangered languages, and the European Union’s linguistic diversity (1%) is compared with the world-wide picture. The chapter also introduces five basic linguistic models through which European states attempt to accommodate or reduce the linguistic diversity existing within their borders. François Grin’s contribution, in Chapter 3, proposes an innovative treatment of language regimes, by looking at them as policy options that can be



Introduction

methodically evaluated. The author applies the tools of policy analysis to EU multilingualism, in terms of resource allocation (or efficiency) and of resource distribution (or fairness). The role of policy analysis in general and his discussion in particular aim at providing “tightly argued and solidly documented inputs to the political debate.” Against what is often presented and disseminated unquestioningly, he argues, there is no obviously superior solution in terms of resource allocation. Moreover, from the perspective of resource distribution (linguistic justice), he defends as sensible options some neglected alternatives such as Esperanto. Diversity is not only an appropriate description of the sociolinguistic situation in Europe, but, as Peter A. Kraus puts it in Chapter 4 of the book, it also “has become a pivotal concept in the normative meta-language that we use as Europeans in addressing one another. [...] Thus, in the archipelago of European identities, diversity may be considered to be the waterway uniting what it separates.” Kraus examines this second dimension and addresses the many ambiguities of the EU’s official discourse on linguistic and cultural diversity. In his view, part of the motivation behind the commitment to protect diversity is not high-minded, but “obeys tactical criteria.” It could be “understood primarily as a concession to the importance of the tradition of the national language in Europe.” Thus, diversity is “basically framed as an additive diversity of state languages.” Besides, he reminds us that market integration and diversity follow competing rationales: the meta-language that the process of market integration speaks “ousts the language of diversity in silent but effective ways.” Against the homogenizing instrumentalism of market integration, the author advocates “an innovative political approach allowing the legitimate articulation of difference in a multinational community of Europeans.” The five essays grouped in the second part of the volume focus on the basis and shortcomings of legal protection of linguistic diversity within the EU. In Chapter 5, Peter Hilpold examines the relationship between two concepts that, independently of each other, have undergone a profound evolution in the last decade: EU citizenship and language rights that are awarded to national citizens in some Member States. The author tends to believe that, from the perspective of those who consider linguistic diversity an enrichment of the project of European integration, the legal environment seems now to be far more propitious than before. In his view, Union citizenship harbours a great potential for the future development of language rights in Europe. Paradoxically, while the Community Courts appear to be determined to elaborate progressively the intra-state linguistic dimension of EU citizenship (that is, vis-à-vis the authorities of Member States), they are reluctant to build the supra-state linguistic dimension of EU citizenship (that is, vis-à-vis the non-institutional EU bodies). This is the topic dealt with in Chapter 6. In this chapter, Niamh



10

Xabier Arzoz

Nic Shuibhne examines recent developments in the field of EC law and minority language policy. She argues that the EU legal framework on language is “a formalistic straitjacket hopelessly unsuited to the reality of EU language needs.” By formally preserving an equality of status for twenty-three languages, the Community system ignores the complexity of language needs and language functions. Meanwhile, however, linguistic hierarchies are intensifying, to the extent that minority languages can no longer be conceptualised as simply those outside the official EU language framework. The author sharply criticises the lack of coherence in EU language planning. She finds worrying that “[w]hile, politically, the Member States have achieved partial expansions of the official language framework […], the Community Courts have meanwhile been moving instead to confine both the scope of that framework and the ideological basis underpinning it.” In particular, Nic Shuibhne comments on the Kik case where the European Court of Justice seems implicitly to reject the proposition that linguistic equality is a general principle of Community law or an inherent aspect of nationality discrimination. Chapter 7, by Xabier Arzoz, offers a preliminary legal assessment of the conditions and limitations under which linguistic diversity is protected by Article 22 of the Charter of Fundamental Rights (“The Union shall respect cultural, religious and linguistic diversity”). It is argued that Article 22 is a minority protection clause, since it addresses the most basic protection needs of minorities. It is further argued that it provides cultural, religious and linguistic minorities with an enforceable right to non-interference on the part of the Union in order to preserve their minority characteristics. This minimum form of protection is consistent with the weak formulation of Article 22. Bruno de Witte analyses in Chapter 8 the potential for promoting and – what may seem less evident – for restricting linguistic diversity through provisions of the EU Charter of Fundamental Rights other than Article 22. Special attention is devoted to the right to “good linguistic administration” in Article 41(4) and to the implications, in Article 21, of the prohibition of discrimination on grounds of language for access to employment and in the workplace, and for language funding by the European Union. Antoni Milian-Massana deals in Chapter 9 with the special situation of languages that are official in all or part of a Member State territory, but lack an EU official status. He analyses these ambivalences of European integration. Although the basic situation remains unchanged, a recent Council Conclusion allowing the exercise of additional languages that, according to the Constitution of a Member State, are official in all or part of its territory, will give some visibility to these languages, such as Catalan, Basque and Galician. This edited collection contains a diverse range of perspectives as well as strong views and arguments on such issues as the convenience of maintaining or reduc-



Introduction

ing the number of official languages within the EU institutions and the possible character of a pan-European language policy. While differences of opinion exist, the contributors appear united in the belief that a culture of linguistic diversity is not only possible within the European Union, but that respect for linguistic diversity would also be a “good thing”, either intrinsically or as a by-product of the pursuit of linguistic justice and equal dignity. This vision of a culture of respect for linguistic diversity that should go beyond the reciprocal recognition of national languages among the Member States could perhaps seem too daring or even too threatening to some people. The objective of a culture of respect for linguistic diversity within the EU has not been achieved so far. Neither the Charter of Fundamental Rights nor the Treaty establishing a Constitution for Europe contains a comprehensive constitutional concept of linguistic diversity going beyond the recognition of the right to use any of the twenty-three “languages of the Constitution” vis-à-vis the EU institutions. The contributors to this book advocate an open debate on a meaningful culture of respect for linguistic diversity both within EU institutions and in the transnational public sphere. The challenge that they present is to convince people that respect for linguistic diversity should have a place among the common values that hold all Europeans together and that, consequently, it should require institutional arrangements for its protection and promotion.

Notes 1. For a scholarly discussion see Van Parijs (2003, 2004) and Grin (2003). 2. Link to official text: http://unesdoc.unesco.org/images/0012/001271/127160m.pdf. 3. On the background and the content of this Convention see Ruiz Fabri (2007). 4. On the Council of Europe’s activities in this field, see the recent Framework Convention on the Value of Cultural Heritage for Society, of 27 October 2005. 5. For an overview, see Crystal (2000) and Hagège (2000). Similarly, Annamalai (2004) locates the beginning of the professional interest in preserving language diversity in recent times, as evidenced by the formation of a special Committee on Endangered Languages and their Preservation (CELP) by the Linguistic Society of America in 1992; but he underlines that “the active participation of linguists in preservation remains, nevertheless, peripheral to the discipline.” Yet the “ecological” line of thought has penetrated the scholarly writings of other disciplines such as law or political philosophy even less intensively. 6. For the many problems concerning the establishment of the language pool and the size of the problem, see Crystal (2000: chapter 1); for the reasons why we should care, see Crystal (2000: Chapter 2) and Skutnabb-Kangas (2003).

11

12

Xabier Arzoz

7. For an earlier discussion of the conceptual assimilation of linguistic diversity protection with environmental protection see Kloss (1969: 287–304). 8. See Annamalai (2004), commenting on the reasons for such public indifference to linguistic diversity. 9. For an introduction to the history of European languages see Harmann (1993) and Walter (1994). On endangered European languages see Extra & Gorter (2001) and Salminen (1999). 10. On the process of extinction of some isolated or marginalised European communities (mostly linguistically defined) such as the Sephardic Jews of Sarajevo, the Germans of Gottschee (Slovenia), the Sorbs of Germany, the Albanians of Calabria and the Aromanians (or Vlachs) of Macedonia, see Gauß (2001); on the opposite process of revival of some linguistic communities, see Abley (2005). 11. See the criticism on this point from an EU lawyer: Wägenbaur 2003. 12. As in fact happened in June 2005. See below. 13. Art. 290 of the Treaty establishing the European Community. 14. Council Regulation (EC) No 920/2005. 15. Art. 1 of the Council Regulation (EC) No 930/2004 and Art 2 of the Council Regulation (EC) No 920/200.

References Abley, Mark. 2005. Spoken Here – Travels Among Threatened Languages. London: Arrow Books. Annamalai, E. 2004. “Public Perception of Language Diversity”. Paper presented at the Linguapax Congress “Dialogue on Language Diversity, Sustainability and Peace”, Barcelona, May 2004. Burke, Peter. 2004. Languages and Communities in Early Modern Europe. Cambridge: Cambridge University Press. Creech, Richard L. 2005. Law and Language in the European Union. The Paradox of a Babel “United in Diversity”. Groningen: Europa Law. Crystal, David. 2000. Language Death. Cambridge: Cambridge University Press. Extra, Guus & Durk Gorter, eds. 2001. The Other Languages of Europe. Clevedon: Multilingual Matters. Fishman, Joshua A. 1994. “On the Limits of Ethnolinguistic Democracy”. Linguistic Human Rights. Overcoming Linguistic Discrimination ed. by Tove Skutnabb-Kangas & Robert Phillipson, 49–61. Berlin: Mouton de Gruyter. Gauß, Karl-Markus. 2001. Die sterbenden Europäer. Vienna: Zsolnay. Grin, François. 2003. “Diversity as Paradigm, Analytical Device, and Policy Goal”. Language Rights and Political Theory ed. by Will Kymlicka & Alan Patten, 169–188. Oxford: Oxford University Press. Hagège, Claude. 2000. Halte à la mort des langues. Paris: Éditions Odile Jacob.



Introduction

Harmann, Harald. 1993. Die Sprachenwelt Europas. Frankfurt am Main: Campus Verlag. Junyent, Carme. 2002. “El concepte de diversitat”. El dret a la diversitat lingüística. Reflexions al voltant de l’article 22 de la Carta dels Drets Fonamentals de la Unió Europea ed. by CIEMEN, 13–19. Barcelona: Mediterrània. Kloss, Heinz. 1969. Grundfragen der Ethnopolitik im 20. Jahrhundert. Vienna: Braumülller. Mackey, William F. 1991. “Language Diversity, Language Policy and the Sovereign State”. History of European Ideas 13: 51–61. Martinell Gifre, Emma & Mar Cruz Piñol, eds. 1996. La conciencia lingüística en Europa. Barcelona: PPU. Milian-Massana, Antoni. 2004. “Le régime juridique du multilinguisme dans l’Union européenne. Le mythe ou la réalité du principe d’égalité des langues”. Revue juridique Thémis 38: 211–260. Ruiz Fabri, Hélène. 2007. “Jeux dans la fragmentation: La Convention sur la promotion et la protection de la diversité des expressions culturelles”. Revue internationale de droit international public 1: 43–87. Salminen, Tapani. 1999. UNESCO Red Book on Endangered Languages: Europe. http://www. helsinki.fi/~tasalmin/europe_report.html. Skutnabb-Kangas, Tove. 2003. “(Why) Should Diversities Be Maintained? Language Diversity, Biological Diversity and Linguistic Human Rights”. Glendon Distinguised Lecture, York University, Toronto, 2003. Van Parijs, Philippe. 2003. “Linguistic Justice”. Language Rights and Political Theory ed. by Will Kymlicka & Alan Patten, 153–168. Oxford: Oxford University Press. Van Parijs, Philippe, ed. 2004. Cultural Diversity versus Economic Solidarity. Brussels: De Boeck. Wägenbaur, Bernd. 2003. “Die Erweiterung der Union zwischen Sprachenvielfalt und Sprachlosigkeit”. Europäische Zeitschrift für Wirtschaftsrecht 23: 705. Walter, Henriette. 1994. L’ aventure des langues en Occident: Leur origine, leur histoire, leur géographie. Paris: Robert Laffont.

13

part 1

Factual and theoretical approaches

chapter 1

Linguistic diversity as curse and as by-product1 Philippe Van Parijs

What is linguistic diversity? How can it be measured? Can the spreading of a lingua franca be expected to reduce it? And if so, does it matter? In an attempt to answer these questions, this paper first distinguishes the three dimensions of diversity – richness, evenness and distance – and describes the tension that is bound to arise between linguistic diversity in various senses and the promotion of multilingualism. It next distinguishes between the two levels of linguistic diversity – local and territorial – and describes the structural tension that tends to develop between local and territorial linguistic diversity. Against this background, it argues (a) that linguistic diversity, by itself, cannot plausibly be regarded as a good, all things considered, especially because of its negative impact on the prospects for economic solidarity; (b) that the erosion of local linguistic diversity should therefore be witnessed with equanimity; but (c) that territorial linguistic diversity will nonetheless need preserving as the by-product of a concern for the equal dignity of the identities closely associated with native languages.

The three dimensions of diversity Our intuitive notion of diversity contains three dimensions. These I shall label, borrowing from the discussion on biodiversity, richness, evenness and distance. Take a population A consisting of three communities, each of them speaking only one language, and another population B consisting of five communities, each of them also speaking only one language. The richness of B is then said to be greater than that of A, as the number of distinct types – whether species, races or, in this case, native languages – is larger in B than in A. It may therefore be tempting to

18

Philippe Van Parijs

infer that population B is linguistically more diverse than population A. But this would be premature. Why? Suppose that the three communities that make up population A are of about equal sizes, whereas in population B one of the five languages is the native language of 99% of the population. In the light of this additional information, we shall have no difficulty agreeing that population A is, after all, linguistically more diverse than the nearly homogeneous population B. Diversity, we conclude, cannot be only a matter of richness, i.e. of number of types, but also of evenness, i.e. of how equally the population is spread between those types, or of how little the members of the population are concentrated in one or few types. Just as richness will not do without evenness, evenness will not do without richness. Just imagine that population A is joined by one individual with a distinct native language. Surely diversity increases. Yet evenness unambiguously declines. Hence evenness cannot be all there is to diversity. Richness matters independently.2 How should these two dimensions be combined? Several indices of fragmentation (or fractionalization, or segmentation) have been proposed for this purpose. The most widely used among them is the Simpson index of fractionalisation: 2 F = 1– Σ ․ si ,

where si is the share of type i in the population (ni/N). The negative term in this expression (the sum of the squares of the shares) is also known as the Herfindal index of industrial concentration when si is interpreted as the share of a firm in total sales on a specific market.3 The Simpson index increases monotonically with both richness and evenness and it can be intuitively interpreted as the probability that any particular member of the population concerned meets someone belonging to a type different from hers in random encounters within the population.4 Richness may be enough to capture our intuitive notion of diversity qua variety or plurality. Richness and evenness together may be enough to capture our intuitive notion of diversity qua fragmentation. But they do not exhaust the whole of our intuitive notion of diversity. If island A houses three species of mosquitoes and island B, in the same proportion, one species of mosquitoes, one species of parrots and one of crocodiles, we shall have no difficulty agreeing that there is more diversity in B than in A. Making such a judgement presupposes some notion of distance. In the case of biodiversity, the extent to which the genetic equipments characteristic of two species differ from one another (genomic distance), or the number of nodes that separate them in the most plausible conjectural genealogical tree (taxonomic distance), has been used for this purpose (see, for example, Weitzman 1992). This third dimension of diversity can be added to richness and



Chapter 1.  Linguistic diversity as curse and as by-product

evenness, not just in the area of biodiversity, but whenever some sensible measure of distance can be devised. Obviously, diversity may increase in terms of Simpson-style fragmentation while decreasing in terms of distance, or the other way around. In many relevant cases, however, the partial ordering produced by the intersection of the three dimensions just discussed (number of types, spread among types, distance between types) should be sufficient to enable us to say, without much hesitation, that diversity is increasing or decreasing in a particular population, and in many relevant cases that diversity is greater in one population than in another. On the other hand, if some sufficiently robust and relevant notion of distance is available, one might be tempted to side-step types altogether. Rather than trying to construct a compound out of the number of types, the distances between types and the spread of individuals among types, one might wish to go straight for the average distance between individuals. This is exactly what is proposed by Bossert, D’Ambrosio & La Ferrara (2006) with their generalized fractionalization index: 2 G = 1– ( iiΣ ijΣ ․ pij )/ N

where pij is the degree of proximity (normalized so as to fall between 0 and 1) between individuals i and j, and N is the number of individuals in the population. In the special case in which pij is posited to be 1 for two individuals belonging to different types, and 0 for two individuals belonging to the same type, this generalized index of dissimilarity coincides with the Simpson index of fractionalization presented above.5 The latter can therefore be interpreted as relying on a very rough assessment of distance. But this need not make it less relevant than the generalized index in most contexts. Take race, for example – as defined by the colour of the skin. Degrees of darkness may matter for some purposes, but the most useful index of racial diversity is most likely to remain one defined in terms of that small set of discrete types in terms of which people perceive themselves and are perceived by others.6 As regards linguistic diversity, Simpson-style indices can be devised and applied easily enough as soon and as long as one can draw a sensible list of distinct languages (no continuum of dialects) and uniquely ascribe each individual to one and only one of them (no multilinguals). To capture distance, on the other hand, two types of indices have been proposed, parallel to taxonomic and genomic distance in the biological case. Thus, Laitin (2000) and Fearon (2003) use an index based on the number of branches two languages share in a hypothetical family tree of languages, while Pinelli (2005) uses an estimate of the time elapsed since the linguistic communities involved were separated. One problem with such indices is that they overlook proximity generated through lexical borrowing and

19

20 Philippe Van Parijs

other linguistic influences after separation. This could in principle be remedied by using measures of linguistic proximity. Thus, Desmet and others (2005) use Dyen and others’ (1992) index of lexical distance among a large number of IndoEuropean languages based on the proportion of words with a common origin in a small sample of basic words. The problem is that basic words are less likely to be borrowed than less frequent ones, and that use of the Dyen index therefore oddly implies that French is lexically closer to German than to English, despite the massive import of French words into English. However, one may wonder whether making linguistic difference a matter of finely measured degree serves any purpose. Linguistic difference as a causal factor would seem to be essentially a matter of hindered oral communication through lack of mutual intelligibility, a threshold quickly reached even by comparatively closely related idioms. Nonetheless, Desmet and others (2005) show that an index of linguistic diversity that takes linguistic distance into account is a better predictor of the degree of redistribution (as the share of social spending in GDP) than are indices of fractionalization. The underlying mechanism they suggest, however, is not a direct causal impact of linguistic distance. Linguistic similarity is rather taken as a proxy for how close and recent contact has been between the populations concerned.7 The discussion so far takes for granted that each human being can be assigned to one language, just as every organism can be assigned to one species. Human beings, however, are endowed with the capacity to become competent in several languages. This makes room for a distinct and often more relevant notion of linguistic distance. Take the case of a population consisting initially of two unilingual communities A and B, and suppose that half of B learns the language of A. This learning generates a new mixed type AB, and diversity, as measured by the Simpson index, would unambiguously rise. Surely, it makes far more sense to assert that diversity has thereby been reduced: by being turned into ABs, some of the As have come linguistically much closer to the Bs, and have thereby reduced the average distance between the linguistic repertoires of the population. As soon as some degree of multilingualism is present, in other words, it is natural to define linguistic distance as the lack of overlap between linguistic repertoires, and linguistic diversity as average linguistic distance. The more languages two people have in common, and the better they know these languages, the smaller the linguistic distance between them. And the smaller this distance, on average, between members of a population taken two by two, the less diverse the population.



Chapter 1.  Linguistic diversity as curse and as by-product

Multilingualism against linguistic diversity? In the light of this conceptual clarification, let us examine the relationship between linguistic diversity and convergence towards one lingua franca. Consider, for example, the two chief official goals of the European Union’s language policy: the protection of linguistic diversity and the promotion of multilingualism.8 At first sight, there is a natural complementarity between these two objectives: multilingualism is inconceivable without linguistic diversity, and linguistic diversity is pointless in the absence of multilinguals capable of enjoying it. There would indeed be no tension whatever between the two objectives if the multilingualism of the European Union took the form of the Germans learning Cantonese and Quechua, the French Afrikaans and Telugu, the Brits Javanese and Lingala, and so on. But of course it does not, for two main reasons.9 First, there is an officially declared bias favouring the learning of other EU languages rather than, say, Arabic or Chinese. Secondly, there is an officially unintended bias towards the learning of English, driven by the interaction of probability-sensitive learning and minimex language use.10 Consequently, as the European population is growing more multilingual (or at least bilingual) by increasing its competence in at least one foreign language, it is becoming more diverse in terms of richness and evenness, through the appearance and gradual expansion of mixed types. But it is by the same token becoming less diverse in the more plausible sense spelled out at the end of the previous section that incorporates distance between linguistic repertoires. Reduced linguistic diversity so understood is a direct and unavoidable consequence of the spreading of a lingua franca. There is a second reason, this time causal, not logical, why widening competence in a shared language can be expected to reduce linguistic diversity. Whenever natives of some language learn another language, this expands the possibility of borrowing and other forms of influence. However, as most native speakers of a given language become competent in the same non-native language, this possibility becomes a strong probability, and the language they all learned will tend to exert a lasting influence on their native tongue – most obviously through the import of vocabulary, sometimes also through morphological and syntactic changes (see for example McWorther 2001: Chapter 3). Moreover, as the natives of several languages all become competent in the same lingua franca, such influences bring their languages closer not only to this lingua franca but also to one another. There must be few languages today in which it has not become “cool” to “google” “blogs” on the “web”. Owing to this process, it is not just the distance between linguistic repertoires that shrinks  – trivially – as a result of the emergence of a lingua franca. It is also the distance between the native languages themselves.

21

22

Philippe Van Parijs

Thirdly, the spreading of the lingua franca can be said to reduce linguistic diversity even in a sense that abstracts from distance between languages and repertoires and takes only richness and evenness into account. To see this, one needs to make a distinction between the distribution of competences and the distribution of performances – a distinction which has no analogue in the case of biodiversity. To illustrate, consider jointly the following two trends, both very tangible in Europe today. One is the spreading of competence in English through the explosive interaction of probability-sensitive learning and minimex communication.11 The other is the growth of the proportion of linguistic interaction occurring between people with different mother tongues. Under some mild assumptions, the necessary outcome of the combination of these two mutually reinforcing trends is an increase in the proportion of conversations held in English. As we move away from a world of essentially unilingual communities whose members talk only to each other, the proportion of conversations held in a language stops being roughly equal to the proportion of natives of that language.12 And once the gap between the two proportions is no longer insignificant, it makes sense to redefine richness and evenness in terms of whether and how much the various languages are being used in conversations, rather than in terms of their presence and distribution in people’s competences. As people add competence in English to competence in their native language and interact with people who lack the latter competence, they substitute English for their mother tongue in a growing share of their conversations. Hence, while evenness need not decline and may even increase as regards competences (as mixed types catch up with pure types), it will exhibit a strong tendency to decrease in terms of performances. We are of course very far from a situation of random mixing, but transnational mobility and communication are sharply on the increase, and further facilitated by the very spread of the lingua franca. Moreover, in some non-Anglo countries as different as India and the Netherlands, many people have achieved such a high level of competence in the lingua franca that they find it easier to express themselves on some subjects in that language even with people sharing their native tongue. When this last stage is reached, however, one must be prepared for a decline in diversity in yet another, fourth sense, which concerns competence, as the first two did, not just performance, and which obtains even if we ignore the dimension of distance, as the third sense of diversity did. Universal diglossia, i.e. the generalization of bilingualism in a linguistic community through the learning by all its members of the same more widely spread language, is commonly regarded by sociolinguists as the last stage before the local language starts withering away. As Antoine Meillet (1928: 117), among many others,13 puts it, “The local idiom is useless the day the whole population, knowing the common language, is bilin-



Chapter 1.  Linguistic diversity as curse and as by-product

gual. The young then no longer feel the need to know the local idiom: even if they practiced it in their childhood, they forget it as they grow older.” Whereas few people seem to be greatly concerned with loss of linguistic diversity in the first three senses considered above, many are indeed concerned with loss of linguistic diversity in this fourth sense. This time, competence diversity shrinks – not just performance diversity as in the third sense; and it does so through impoverishment, no longer through enrichment as it did in the first two senses. Yet, throughout the linguistic history of the world, diversity loss in the first sense has typically led to diversity loss in the fourth one. In many cases, oppression and shame accelerated the process. But in a high mobility context, the explosive interaction of probability-sensitive learning and minimex communication is powerful enough to complete the job unassisted. As diversity loss in the first sense is inseparable from the very adoption of a lingua franca, the erosion of linguistic diversity in this fourth sense is a sensitive issue to which I shall return.

Local diversity versus territorial diversity Just as important as the distinction between three dimensions of diversity – richness, evenness and distance – is the distinction between its two levels. Here again, it is helpful to look at the literature on biodiversity, where a distinction is commonly made between α-diversity, or the number of species within a particular habitat, and β-diversity or the number of species within a particular landscape consisting of a set of habitats. Both α-diversity and β-diversity express richness – or “variety” or “inventory diversity” – at the local and at the global level, respectively. By contrast, β-diversity is meant to express differentiation – or “distinctiveness” or “specialization” or “segregation” – i.e. the extent to which habitats differ from one another within a given landscape. β-diversity can be defined, as it was initially in the biodiversity literature, as the ratio of γ-diversity to average β-diversity, i.e. the ratio of the total number of species in the landscape to the average number of species in its habitats. Alternatively and more conveniently, it can be defined as the difference between γ-diversity and average α-diversity.14 Under either definition, it reaches its minimum (1 and 0, respectively) when all species present in the landscape are present in each habitat, or in as many habitats as their (possibly small) sizes allow, and its maximum is reached when each species is gathered in a single habitat, or in as few habitats as its (possibly large) size allows. Beyond the case of biodiversity, we can analogously make a more general distinction between (α) diversity (or variety) within some component (habitat, neighbourhood, region, country, etc.) of a broader population (landscape, city, country, world, etc.) and (β) diversity (or differentiation) across such components.

23

24

Philippe Van Parijs

These components are usually defined territorially, but need not be. They could correspond, for example, to the various sections of a school or to the various regiments that make up an army. To the extent that the components are local units, defined in territorial terms, I shall speak of local diversity to refer to α-diversity within some local unit, and of territorial diversity to refer to β-diversity across local units. In the case of ethnic, or cultural or linguistic diversity, this distinction is useful to contrast, on the one hand, the local diversity that exists, say, in a town (typically as a result of recent immigration, but also sometimes as a result of more ancient immigration, with distinctness perpetuated by religious differences, as in the Jewish ghettos of medieval cities); with, on the other hand, the territorial diversity that exists between different geographical areas of a particular country, such as Switzerland or Nigeria, typically as a result of its incorporating territories in different languages that have been co-habiting “forever”. To capture the notion of territorial diversity properly, however, it is essential not to confine it, as I have done so far, to the richness dimension, i.e. to the number of types (species, languages, races, etc.). Take for example an island consisting of two regions. At an early stage, each of the two regions has native Greek speakers and native Turkish speakers in equal proportions, say 50/50. At a later stage, one region has a 90/10 majority of Greeks and the other one a 90/10 majority of Turks. Despite the dramatic shift, the island’s territorial diversity, using the measure specified above, has remained unchanged at its minimum level, since the average number of native languages per region has remained equal to the total number of native languages on the island (β = 2 – (2+2)/2 = 0). As this example shows, a useful notion of territorial diversity should at least take the evenness dimension of diversity into account. Useful indices which do precisely this have been developed for very different purposes by sociologists and economists. For example, the isolation index developed in the sociological literature on segregation is the probability that a person will meet another member of her own ethnic group if she were to meet at random other dwellers of her neighbourhood. This index can obviously be generalized to any interpretation of both type and local unit. Its average value across all types reaches its minimum when the distribution of types is the same in all units. Moves away from such a homogeneous distribution of types across units are reflected in a rise of the isolation index, which therefore provides a more satisfactory index of territorial diversity than β-indices of biodiversity. In our island example, for instance, the probability that a Greek will meet a Greek when randomly meeting inhabitants of her or his region is significantly higher under a 90/10 distribution than under a 50/50 distribution.15 In a very different context, economists developed the country Gini coefficient in order to capture how specialized a country is, i.e. to what extent the distribution of its output between industrial sectors diverges from the distribution of the



Chapter 1.  Linguistic diversity as curse and as by-product

output of a larger entity (say, the European Union) of which it is a part. When averaged over all local units (here, countries), this index can be used to measure the degree of territorial diversity, i.e. in this case the degree of sectoral differentiation of countries. It again reaches its minimum when the distribution is the same in all units, and the larger it becomes, owing to the specialization of each unit in one or more types, the greater the larger entity’s territorial diversity. When applied to our island example, this index also yields a more satisfactory verdict than simple β-indices that only take the number of types into account. Both regions’ Gini coefficients (which capture how “ethnically specialized” they are) and hence also their average value obviously increase as one moves from a 50/50 to a 90/10 distribution of Greeks and Turks.16 Suppose then that we have some sensible notion of both local and territorial diversity. As suggested long ago by Claude Lévi-Strauss (1961) in connection with cultural diversity, there is a tension between the two. Maximum local cultural diversity means that every culture can be found in every place, whereas maximum territorial diversity requires that each place has a culture that cannot be found anywhere else. And the same can be said for linguistic diversity.17 It follows that those wishing to promote local cultural or linguistic diversity must realize that this will come at the cost of reduced territorial diversity. By contrast, if preserving territorial diversity is a meaningful objective, the development of local diversity will need to be counter-acted. Thus, typical “multiculturalism policies” entail fostering local diversity at the expense of territorial diversity, whereas the imposition of an official language on a whole country or part of it amounts to favouring territorial diversity at the expense of local diversity. What about the adoption of a lingua franca? Like multiculturalism policies, it tends to systematically reduce territorial linguistic diversity as a result of the four processes discussed above. Unlike multiculturalism policies, however, it does not do so by systematically increasing local diversity. On the contrary, if we start from a local situation that is not totally homogeneous, it will tend to reduce local diversity too. And even if we start from a unilingual situation, the gradual spread of the lingua franca may first increase local diversity, but once it has reached everyone it will have reduced it to its initial level, and the erosion of the “superfluous” language will start, thus further reducing territorial diversity.

Is linguistic diversity valuable? This stylized picture of the processes at work brings out the ineluctability of a deep long-term tension between the spreading of a lingua franca and linguistic diversity, whether the latter is defined in a way that includes only richness or also

25

26 Philippe Van Parijs

evenness and distance, whether it is interpreted as local or as territorial diversity. Is this a problem? The responses I offered to the cooperative and distributive injustices generated by the adoption of a lingua franca (LJ chapters 2–3) did nothing to alleviate this threat. On the contrary, by endorsing linguistic territoriality, my response to linguistic injustice as unequal dignity does involve a firm protection of some degree of territorial linguistic diversity; but this protection (not maximization) of territorial linguistic diversity is just a by-product of the proposed solution to this dimension of linguistic injustice (LJ chapter 4).18 It is not appealed to as an argument in its favour, whether as an aim in itself or as instrumentally useful for some important aspects of our common good. Is this right? Should diversity not rather be regarded as having great intrinsic or instrumental value, as seems implied, for example, by both the Indian Union and the European Union choosing “Unity in Diversity” as their mottos, in contrast with the United States’ “E pluribus Unum”, or by the European Union’s Charter of Fundamental Rights (2000) making sure to state, in its article 22, that “the Union shall respect cultural, religious and linguistic diversity”?19 And should this not play a major role in the normative discussion of linguistic matters? Let us acknowledge straight away that there are many people who find linguistic diversity a plain nuisance. Among them are those who care for nothing but business.20 But they are not alone. Thus, no doubt to the surprise of many of his contemporary colleagues, the distinguished linguist Meillet (1928: 244) wrote: The small national languages are a stage through which poorly cultured peoples pass on their way to universal civilization. But the multiplicity of the languages currently used in Europe, already inconvenient today, prepares crises which will be hard to resolve, as it goes against the general trends of civilization. The unity of the common language is an immense strength for those who possess it.21

If a powerful language were to drive all others into gradual extinction, not only would we all enjoy the convenience of being able to use our mother tongue in all the conference rooms and hotel lobbies of the world, but incomparably more would be within our reach: even in the most remote bazaars, farmyards and playgrounds, we would be able to understand directly what the locals were saying to each other, while the reach of diverse yet world-wide media and the massively enhanced transnational mobility would prevent the stable development of mutually unintelligible dialects. Once again, all human beings would “speak the same language and form a single people,” and hence possibly “no goal will be unachievable for them” (Genesis 11.6). Is there anything to prevent us from looking forward to this new stage in the progress of mankind, apart from the irrational fear that a jealous Yahweh may strike once more and cruelly thwart our neo-Babelian hubris?22



Chapter 1.  Linguistic diversity as curse and as by-product

Unlike Meillet, most professional linguists are understandably not keen to see most languages in the world quickly vanish, many of them only poorly documented, thereby irreversibly amputating the subject matter of their discipline. But most of them are also reasonable enough to admit, when pressed, that it would be unfair to attempt to induce some people to keep learning, talking and teaching a language they would otherwise abandon, for the sole purpose of enabling a small bunch of inquisitive scholars to indulge their intellectual curiosity and write about them in academic journals. To broaden their coalition, linguists will have no difficulty enlisting translators and interpreters, who would obviously lose their source of income were linguistic diversity to disappear. But this would hardly make the argument less suspiciously corporatist. Nor would their case be much strengthened by the support of the aesthetes who love steeping themselves in delightfully varied linguistic environments. One can sympathize with those who believe that the attraction of Florence would be diminished if Italian had gone into disuse, and that the charm of Rio would suffer if Brazilian were replaced by American. But this seems of precious little weight in regard to the great collective benefits of universal communicability. Those who want to make a persuasive case for the value of linguistic diversity therefore realize they need arguments that appeal to less factional interests.23 One possible line of argument emphasizes that each language is a unique repository of human knowledge. It comes in two main variants. The more subtle one rests on the interesting observation that the syntax of a language, its phonological system, its morphology and its lexicon contain information about the history of the peoples who have been speaking it through the centuries, most obviously about where they came from, about which other peoples they are related to, about which peoples they interacted with. With any language that goes extinct without having been fully recorded, knowledge of this sort is lost forever, a great pity for anyone interested in the relevant segments of human history. But it should be no insult to the honour of their profession that, as an argument for preventing people from giving up their ancestral language, the irreversible loss of such potential knowledge is no more persuasive than the shrinking of the subject matter of professional linguists. The second variant is less subtle but incomparably broader in scope, and hence more promising as a non-corporatist argument. It rests on the plausible assumption that some things have been known only to people of a particular language, uniquely equipped with the terminology needed to formulate them. To illustrate, take the attempt made by Skutnabb-Kangas (2003: Section 3) and others to seal a strong alliance between advocates of biodiversity and linguistic diversity. There is an undeniable positive cross-regional correlation between linguistic diversity and biodiversity, with both languages and species particularly numerous

27

28

Philippe Van Parijs

(relative to the sizes of human populations) in equatorial areas. What explains the correlation, it is suggested, is a causal link from linguistic diversity to biodiversity: in the absence of the nature-respecting knowledge incorporated in the many local languages, species diversity would soon be reduced. The correlation, however, is bound to have far more to do with the relative attractiveness of certain natural conditions (climate, topography, etc.) for nature- and culture-destroying colonization and industrialization. The relationship between biodiversity and linguistic diversity, as we saw above, can be instructive, but to appeal to a spurious correlation is an unpromising way of harnessing interest in the former in order to generate support for the latter. The impact language conservation may have on species conservation must be, if at all real, very modest, and likely to be offset by the potential of knowledge dissemination which a switch to a more widespread language would create. This last remark applies more generally to any instance of this second variant of the argument: people who possess some knowledge can express it in the language they have learned (and which threatens to displace the old one), even if by importing the terminology from their native language. And by doing so, they would not only preserve the knowledge in question, but also make it more widely available. A distinct family of arguments rests on the connection between linguistic diversity and cultural diversity. Culture can be roughly defined as a set of ways of thinking and behaving that is durably shared by, and distinctive of, a community. Linguistic diversity is linked to cultural diversity in two ways. One is that, like religion for example, a community’s language directly constitutes and shapes its distinctive thoughts and practices. The other is that, more than religion or any other aspect of a culture, it affects the patterns of interaction and hence the flows of information, education, persuasion or imitation that constantly shape and reshape all aspects of culture. Given the nature and reach of present and future media, so this line of argument goes, linguistic diversity is the firmest, and increasingly the only serious protection of cultural diversity. Worldwide, this preserves more options for people to choose from, and hence leaves more room for collective experimentation in private and social life, from which mankind as a whole may benefit in the long run.24 Though unavoidably speculative, this line of argument must be taken seriously. But it is weakened by two serious difficulties. First, if the point of cultural diversity is the cultural freedom it gives, and if the exercise of the latter must therefore be fostered, then cultural diversity, i.e. diversity of thoughts and practices linked to distinctive multi-generational communities, is in the process of being eroded by the very dynamics it sets in motion. Cultural diversity makes cultural freedom possible, but the exercise of cultural freedom replaces cultural with non-cultural diversity, which linguistic diversity has no particular capacity



Chapter 1.  Linguistic diversity as curse and as by-product

to protect. Secondly, the price to be paid for linguistic diversity (also, though to a lesser extent, in the presence of a lingua franca) is the less general and less rapid availability of whatever exists or is invented in any particular culture (except the one associated with the lingua franca). Assuming there is a positive impact of linguistic diversity on cultural diversity, therefore, the positive impact on the general interest will be offset, perhaps only partly but perhaps also to the point of becoming negative, by the lesser availability of whatever cultural diversity there is.

Local linguistic diversity against economic solidarity? In this light, the case for linguistic diversity as an efficient way of pursuing some general benefit turns out to be rather weak. As if this were not bad enough, there appears to be a strong case against linguistic diversity as an obstacle to the achievement of distributive justice (as understood in LJ Chapter 2). Thus, a growing number of econometric studies have tended to establish a robust negative correlation between linguistic diversity and a number of variables that can be regarded as more or less plausible proxies for the extent of economic solidarity or of redistribution of the better off to the worse off. For example, using a sample of over two hundred countries, Alesina and others (2003: Table 13e) show that, after controlling for many variables, there remains a significant negative correlation between, on the one hand, ethnic diversity (using the Simpson index), especially when defined exclusively in linguistic terms, and, on the other, the share of transfers and subsidies in the country’s GDP. This result is even strengthened when the linguistic diversity index is redefined to incorporate linguistic distance (Desmet et al. 2005: Section 5.2). This does not rule out that, owing to some other relevant difference, a linguistically more diverse country may nonetheless achieve greater solidarity than one that is far more homogeneous in this respect.25 All the correlations establish, and need to establish, is the existence of a ceteris paribus relationship. To make sense of this negative relationship between linguistic diversity and economic solidarity, several mechanisms have been suggested. How generous an economic solidarity system manages to be can schematically be said to depend on the willingness of the better-off to share with the worse-off and on the ability of the worse-off to organize so as to force the better-off to share.26 Both factors can be affected by the degree of linguistic diversity through two mechanisms. On the one hand, linguistic diversity makes identification more difficult: a different language makes one part of the population perceive another as alien, as not belonging to the same kind, and hence as less trustworthy, less likely to reciprocate, and less likely to have reciprocated had roles been reversed.27 This may be either

29

30

Philippe Van Parijs

because the alien language is, as such, a feature of the alien culture that makes the other difficult to identify with, and/or because it reduces interaction and thereby makes the other more impenetrable. Lesser identification makes both the betteroff more reluctant to accept economic solidarity and the worse-off less capable of organizing collectively to demand it effectively. On the other hand, linguistic diversity also affects solidarity simply by making communication more laborious: in the absence of an effective medium of communication, it is more difficult for the better-off to be persuasively exposed to arguments of fairness in favour of the worse-off, for the worse-off to coordinate effectively their struggle against the better-off, and for all to settle on the fine grain of the organization of solidarity.28 The tension thus highlighted can be expected to hold for both local and territorial diversity. Let us consider local diversity first. By ensuring that everyone learns the local official language, the implementation of linguistic territoriality (as advocated in LJ Chapter 4) reduces local linguistic diversity in all four senses distinguished above, and there is no doubt that it also tends to reduce local cultural diversity by increasing every local resident’s potential exposure to the information and ideas available to every other local resident. The intensification of contact may threaten the local survival of languages other than the official one, but the aim should not be to eradicate them. Getting everyone into a common demos made possible by a shared language is essential to the pursuit of distributive justice. People from all layers of society must be willing and able to explain their standpoint to each other, to listen to each other, to take decisions that can seriously claim to have been made for the common good, that can hope to reflect an equal respect for the diversity of conceptions of the good life and an equal concern for the interests of everyone affected. But this does not amount to merging everyone into the same ethnos, with a common language as a core component of the common culture. In other words, the strong identification favoured by linguistic homogeneity may be dispensable if the communication enabled by proficiency in a common medium fully performs its job both on the side of acquiescence by the better-off and on the side of mobilization of the worse-off. In particular, a plurality of languages and the associated cultures can be transmitted from generation to generation in addition to the language known in common. Competence in a shared language is essential to facilitate dialogue, discussion, argumentation, understanding among all the community’s members, but there is no need to turn the community into a cultural monolith. From this perspective, there cannot be either a general endorsement or a general condemnation of linguistically relevant multiculturalism policies, i.e. policies aiming to preserve or respect cultural diversity. There is no conclusive evidence as to whether the adoption of multiculturalism policies tend to correlate with increases or decreases in institutionalized solidarity.29 But even if there



Chapter 1.  Linguistic diversity as curse and as by-product

were such evidence, no general conclusion could be drawn. Suppose for example that the data showed a positive correlation. For critics of multiculturalism policies this need not come as any surprise. Both multicultural policies and welfare policies, some of them claim, stem from the same “providential” dynamics of the contemporary democratic state, which drives the latter to accommodate an ever expanding set of sectoral demands.30 No wonder, therefore, that they might be positively correlated, at least in the short run. But this is consistent with multiculturalism policies slowly undermining the welfare state, if only by hindering a rapid reduction of the linguistic cultural diversity that keeps being amplified by immigration. Because of the time required for these sociological processes to work themselves out and the randomness involved in their being politically exploited, this effect is most unlikely to show up immediately, even though the weakening of the fellow-feeling between all citizens and the decreasing ability of the worse-off to join forces in a common struggle make the “providential” set-up increasingly vulnerable. Moreover, the fine grain of the multiculturalism policies is of crucial importance. By way of illustration, consider the teaching of, or in, the immigrants’ mother tongues – a multiculturalism policy that can be sensibly justified both as a way of symbolically asserting the dignity of the languages concerned and the associated identities and, in some cases, as a way of formalising and strengthening the children’s valuable competence in a major world language such as Spanish, Arabic, Turkish or Bengali. The most straightforward way of doing this consists in offering this option in those schools in which there is sufficient demand for a particular language, owing to a high proportion of pupils with a particular origin. In countries where school choice is free, the provision of such courses will create an incentive for parents of the relevant origin to send their children to those schools. As a result, whatever degree of ethnic mixing has been achieved in the school system will be reduced, and given that children’s acquisition of the local language depends more on interaction with their peers than on formal teaching, the long-term threat posed to social cohesion (through the causal chain of poor linguistic competence, low productive skills generally, low probability of landing a good job, low chance of social and geographical mobility) is quite considerable. In countries in which school choice is strongly constrained by districting, the threat will be slower to show but deeper, as the provision of immigrant language courses will not only create an incentive to change schools, but also to move, thus fostering segregation not just in schooling but also in housing. Much of this effect can be switched off, however, if instead of being organised as part of the curriculum of a particular school, the courses were open to pupils from different schools, which, in an urban context, is often a realistic possibility.31 A small organisational difference that could not reasonably be expected to affect an index of the extent

31

32

Philippe Van Parijs

of multiculturalism policies is therefore arguably of crucial importance for the strains these may help perpetuate or amplify with economic solidarity.32 Presupposed in this discussion is the notion that the persistent segregation of linguistic communities is to be counteracted. This should not be regarded as self-evident. Thus, to explain that support for the welfare state in Canada does not seem to suffer from growing overall levels of ethnic diversity, Soroka, Johnston and Banting (2004) suggest, that “the high level of geographic concentration of immigrant minorities in certain regions and especially certain urban areas” may be better for interpersonal trust, which itself clearly has a positive effect on support for the welfare state.33 One should not overlook the direct impact of social capital on distributive justice (what is distributed through local informal solidarity may matter as much as what is distributed by formal institutions) and the welfare state may well be less vulnerable politically with tight ethnic communities than with general anomy.34 But especially if the ethnic divide is a linguistic divide, the vicious circle of a persisting linguistic handicap sketched earlier makes segregation most unpromising for distributive justice, both because of its direct impact on opportunities and because of its impact on competent participation in a common public forum and in the mobilization of the worse-off.35 Hence, multiculturalism policies that breed segregation by compelling, encouraging or even simply allowing linguistically distinct immigrant communities to have their own schools, sports clubs or neighbourhoods can be expected to affect negatively the prospects of economic solidarity. It is of course not only multiculturalism policies that affect the extent of separation between the various linguistic communities. The fine grain of the institutions of economic solidarity is no less relevant.36 This holds, for example, for the organization of the health care system, which can be more or less segregated. It also holds for the aspects of the welfare state that most affect the labour market. Suppose that basic economic security is implemented through employment­unfriendly means-tested benefits, which are withdrawn as soon as a member of the household performs a declared paid job and which may prove difficult to recover once that job is lost, owing to opaque and possibly discriminatory rules. For a given level of generosity of the welfare state, the opportunity and motivation to acquire and retain linguistic skills will be far less than under a more employmentfriendly regime, with a cumulative negative impact on both the opportunities and participation of linguistic minorities. It does not follow that one should go for a punitive workfare state, which would amount to reducing the extent of solidarity (distributive justice, as interpreted in LJ Chapter 3, is not only a matter of income). There are other versions of the “active social state” that consist in spreading solidarity to low-income working households, whether through wage subsidies, through varieties of earned income tax credits or (my preferred variant) through



Chapter 1.  Linguistic diversity as curse and as by-product

the provision of a universal income floor in the form of non-means-tested child benefits and basic pensions and a modest citizen’s income for all adults.37 The general conviction that underlies this discussion is that a key factor in sustaining generous economic solidarity despite growing local linguistic and cultural diversity is the presence of sufficiently intensive interpersonal contact, through school, work and other channels, across the boundaries that tend to form around linguistically distinct communities.38 It must, however, be conceded that the more contacts of these sorts there are, the stronger the tendency for local linguistic and cultural diversity to wither away or to grow less (as a result of continuous immigration) than it would otherwise have done, as a result of interaction, including intermarriage, and of exposure to common circumstances, information and other influences.39 Unlike the erosion of territorial diversity, to which I will return shortly, this erosion of local diversity is not a problem. Fair economic opportunities and appropriate weight in the political process must have precedence over the preservation of local linguistic and cultural diversity wherever it happens to emerge as a result of immigration. Even if this results in some languages and cultures being squeezed out locally altogether, no unfairness, or lack of respect, is thereby being inflicted on anyone, providing the background assumption is one of reciprocity, however counterfactual. Had the roles been reversed, had the (current) autochthonous population been migrating into the homeland of the (current) immigrants, they could not have claimed or expected more by way of preservation of the local cultural diversity which they would have been causing by moving there. By deliberately reducing or containing local linguistic diversity, therefore, the learning of a common language by all those who share a territory does not hurt justice as equal dignity (see LJ Chapter 4), while promoting the conditions for furthering distributive justice.40

Territorial linguistic diversity against economic solidarity? Thus, the linguistic territoriality regime offered to handle linguistic injustice as unequal dignity sides conveniently with distributive justice at the expense of some diversity at the local level. At the territorial level, by contrast, it has the effect of protecting diversity, and hence would seem to undermine rather than improve the conditions for distributive justice on a broader scale. If the segregation of linguistic communities at the local level is a bad thing for the sake of economic solidarity, fortunately counteracted by the linguistic territoriality principle, how can it fail to be a bad thing on a larger scale too? By hindering both identification and communication, as explained above, linguistic diversity can be expected to

33

34

Philippe Van Parijs

weaken the prospects for economic solidarity at the territorial no less than at the local level. Before addressing this challenge, we need to consider a puzzling argument to the effect that this challenge is a fiction, that there are on the contrary good reasons to expect territorial linguistic diversity to go hand in hand with a developed welfare state. The point of departure of this argument – and part of what it is meant to explain – is the contrast between the United States and the European Union taken as a whole. At present, the level of economic solidarity can safely be said to be higher in the latter than in the former. Yet it is obviously also the European Union that exhibits the higher level of territorial linguistic diversity, firmly preserved by national boundaries that have gradually come to coincide by and large with linguistic boundaries (through the two mechanisms sketched in LJ, Chapter 4). Economic reasoning suggests that this positive correlation between solidarity and diversity is no coincidence. Here is the argument. Industrial development relies crucially on specialised skills. But heavy investment in these skills will happen only if enough insurance is provided in case local demand for them never materializes or disappears. One way of providing such insurance is by unifying linguistically a large area within which one can then move at comparatively low cost in search of another employment for the same skills. Another is through a developed welfare state. With a territory cut up into smaller linguistic areas, and hence with a higher average cost of moving in search of another use for one’s skills, the optimal welfare state is bound to be, on average, considerably larger in the European Union than in the United States.41 Rather than happily concluding that, far from being antagonistic, territorial linguistic diversity and generous economic solidarity are complementary, it is important to note, first, that what is shown to be optimal, under conditions of greater linguistic diversity, is greater social insurance, not greater genuine (ex ante) solidarity. That Europe should have a larger truly redistributive welfare state can therefore be explained by this argument only to the extent that it forms an unavoidable by-product of a strong social insurance system, as administrative simplification and political dynamics push the transfer systems of each nation-state beyond what fits under the umbrella of the insurance principle. This is an important qualification. But there is an even more serious objection to a complacent reading of this argument. Suppose that the national solidarity systems become immersed in a common market, in which capital and commodities move freely, while people remain essentially stuck within national borders as a result of language differences. Considerations of competitiveness will put the truly redistributive, so-called “compassionate” aspects of the welfare state under growing pressure, as mobile capital and consumer demand will tend, other things



Chapter 1.  Linguistic diversity as curse and as by-product

being equal, to move to those places where redistributive taxation weighs less heavily on the efficient remuneration of factors of production. The pressure is further increased as the upper layers of the skilled labour force become more mobile transnationally ­ (precisely by virtue of having become competent enough in an international lingua franca), thereby increasing considerably the (redistributive) tax elasticity of the domestic supply of human capital. Governments will need to respond by lowering the higher rates of income tax, by substituting regressive consumption taxes for progressive income taxes, by expanding lax and generous “expatriate” or “non-resident” statuses, by deliberately tolerating tax loopholes that primarily benefit affluent taxpayers, and by shifting government expenditures to the advantage of high earners. Put differently, they will need to shift resources towards subsidised opera performances, public golf courses and convenient airports, and away from subsidies to low-paid jobs, cheap public housing and benefits for the unemployed; towards the cleaning or policing of the better neighbourhoods at the expense of education or public transport in the poorer ones. Immersion in a competitive transnational market turns states into firm-like entities, under constant pressure to downsize their redistributive ambitions, to shrink those aspects of their welfare systems that go beyond insurance, to reduce public expenditures that effect genuine transfers from the high earners to the low earners, from the more talented, the more skilled, the more mobile, towards the less qualified, the less able, the less mobile.42 Of course, as solidarity becomes more difficult to organize, for the reasons just sketched, at the level of individual nation-states immersed in a common market, one might hope that a larger political entity, pitched at a level closer to the one at which the market is operating, say the European Union, could take over the task. Indeed, in the US, the bulk of the net redistribution accomplished by the taxand-transfer system is the work of federal, not of state programmes.43 But as the lingua franca spreads among the highly skilled, the economic case just sketched in favour of developed social insurance systems – and hence the expectation of some genuine solidarity as a by-product – would keep weakening. More importantly, by switching to this higher level, the EU’s advantage over the US in terms of prospects for sustainable economic solidarity would be turned into a handicap, since its far greater territorial linguistic diversity, entrenched by the linguistic territori­ ality principle, tends to make identification and communication at the relevant level far more difficult than in the US. Moreover, precisely because of the grip of the linguistic territoriality principle, there is a strong case for keeping many policies decentralised at a level at which they can be discussed and explained in the language people are most familiar with. This is in principle compatible with organising the bulk of economic solidarity at a more centralised level. But a tension unavoidably arises between

35

36

Philippe Van Parijs

centralised solidarity and decentralised competences in other fields (say, education, town planning or environmental policy), as some of the benefits and costs of sound or sloppy policy in decentralised matters are exported upwards to the central level, and hence shared by decentralised units which bear no responsibility for them. Thus, the partition of the near totality of Belgium’s surface into two unilingual areas (in the 1930s) led to the growing separation of two distinct public discussion spaces, subsequently to demands on both sides for the devolution of powers to entities matching these spaces, and eventually, in the 1990s, to the transformation of the country into a full-fledged federal state along linguistic lines. In short: no logical entailment, but a natural dynamics. However, the need to allocate to the decentralised authorities responsibility for the consequences of their policies in ways other than the implementation of solidarity further contributes to preventing more centralised, and therefore economically less vulnerable, solidarity from being sustained at as high a level as would be the case with less territorial linguistic diversity. Added to the obstacle it creates for strong identification and fluid communication, this is a third reason why territorial linguistic diversity hinders the pursuit of economic solidarity. There is no point denying this tension between territorial diversity and economic solidarity, and hence between the pursuit of justice as equal dignity through a linguistic territoriality regime and the pursuit of justice as equal opportunity through redistribution. Let us recall the rationale for the territoriality principle (LJ Chapter 4). If we want to respect the equal dignity of the various linguistic communities in a context in which the language of one of them is granted the status of a common lingua franca, demystification and symbolic affirmation are not enough. We must also accept that each of the communities can adopt coercive measures which make it realistic and legitimate to give their language top status in the territory in which it happens to prevail and to effectively protect it against gradual displacement. Although such coercive measures provide some material advantages to the natives of the language, the measures can be expected, under contemporary conditions, to involve a significant long-term net cost for the communities concerned. This cost should be borne by each of the communities which make the choice of imposing protective measures. Some of them will consider it prohibitive, and they will waive their right to implement the territoriality regime which these measures would constitute. The preservation, owing to this set-up, of a significant degree of territorial linguistic diversity constitutes a prima facie hindrance to the pursuit of economic solidarity and hence of distributive justice as equal opportunity, both across linguistic communities and (because of tax competition in the case of redistributive



Chapter 1.  Linguistic diversity as curse and as by-product

schemes at the level of political units immersed in a common market) within each of them. But it is precisely this hindrance which the spreading and democratization of competence in the lingua franca is meant to reduce, by making communication, though not necessarily identification, about as easy and reliable among natives of different languages as it is among natives of the same one. As argued elsewhere (LJ Chapter 4), the very spreading of the lingua franca hinders the effective implementation of the linguistic territoriality principle, both because of universal diglossia among natives and because non-native immigrants can get away with using just the lingua franca. Consequently, the level of coerciveness required to prevent erosion of the local language will need to increase. At the same time, the extent to which the identity of the dwellers of the territory is linked to its language may be sharply reduced.44 We may then be approaching a situation analogous to the terminal stage of many “dialects”, whose native speakers were persuaded to identify with a more or less cognate grander language. No one’s honour or dignity is being threatened if a language with whom no one identifies is not given pride of place anywhere, if it is left to agonize and die. As identification with their language declines, more communities may judge that the preservation of their linguistic distinctiveness is not worth the cost and coercion it imposes and decide to waive in turn their right to protect it. The hindrance to the pursuit of transnational distributive justice will thereby be further reduced. Presumably, if people’s identities were re-shapeable at will, their development of a broader identity strong enough to obliterate their identities as speakers of the weaker language (and heirs to the traditions associated with it) would eliminate any injustice-as-unequal-dignity that might otherwise have been present: injustice as unequal dignity is not between languages, but between people who identify with those languages. Identities, however, are not so malleable. For the time being, therefore, and for the foreseeable future, justice as equal dignity will need to ascribe a major role to the linguistic territoriality principle and hence significant salience to the many quibbles to which its implementation is bound to give rise. As a consequence, a considerable degree of territorial linguistic diversity will need to remain entrenched, to the chagrin of those who believe justice is entirely a matter of material distribution. The celebrators of linguistic diversity will be relieved to read this. But they must realize that they are just being lucky. The reason why linguistic diversity must and will be preserved is not that it is intrinsically valuable, nor that it can be expected, all things considered, to have beneficial consequences. It is simply that it constitutes, for the foreseeable future, a by-product of the pursuit of linguistic justice as equal dignity.

37

38

Philippe Van Parijs

Notes 1. This paper constitutes a preliminary version of Chapter 5 of a book in progress (Philippe Van Parijs, Linguistic Justice for Europe and for the World, Oxford University Press), referred to below as LJ. The arguments it offers rely on a broader framework developed in this book, in particular the “minimex” analysis of language dynamics, and the distinction between linguistic justice as fair cooperation, as equality of opportunity and as equal dignity. This framework is sketched in Van Parijs (2004). Earlier versions were presented at the seminar “Social Justice and Cultural Diversity” (Amartya Sen and Philippe Van Parijs, Harvard University, Spring Semester 2005), at the Conferences “Understanding Diversity: Mapping and Measuring” (European network “Sustainable Development in a Diverse World”, Milan, 26 January 2006) and “Challenges of Multilingual Societies” (CORE & ECARE, Brussels, 9–10 June 2006) and at the workshop on “Language and Politics” of the World Congress of Political Science (Fukuoka, 10 July 2006). 2. It is here taken for granted that linguistic diversity relates to people’s native linguistic competence. But richness and evenness could also be applied at the level of linguistic performance. This distinction does not make sense in the case of biodiversity: if you are born a chimpanzee, you will not behave and look like a mosquito, nor conversely. But natives of one language can end up speaking to one another most of the time. Diversity at the level of performance may therefore diverge significantly, typically downward, from diversity at the level of competence. The relevance of this distinction will be illustrated below. 3. The Simpson index can be shown to correspond to a particular value (1) of a parameter in a more general index of diversity combining richness and evenness (Patil & Taillie 1982), with richness corresponding to another extreme value of that parameter (-1) and the ShannonWeaver index of entropy corresponding to an intermediate value. For an instructive formal discussion of these indices, see Ottaviano & Pinelli (2005). 4. The probability for some member of type i to meet a non-member of i, assuming the probability of meeting any member of the population (including herself) is the same, is given by (N-ni)/N = 1-si. The weighted average of this probability over all types i is given by Σsi.(1-si) = 1- Σsi2, which is precisely the Simpson index. 5. See Bossert, D’Ambrosio & La Ferrara’s (2006) illuminating axiomatic derivation of their generalized index. 6. See Lee & Bean (2004) and Hochschild, Burch & Weaver (2005) for insightful empirical studies of the distinct effects of colour and race. 7. Indices of diversity that incorporate distance also have the advantage of making results less dependent on the choice of considering two dialects (say, Neapolitan and standard Italian) as two variants of the same language or as two distinct languages. Simpson-type indices can jump as a result of choosing, more or less arbitrarily, the second option, whereas distance-sensitive indices behave more smoothly. 8. See Strubell (2006) for an overview of the declarations by the Commission, the Council and the Parliament.



Chapter 1.  Linguistic diversity as curse and as by-product

9. Robert Phillipson (2003: 193–8) emphasizes the intrinsic tension between the EU’s ritual assertion that, on the one hand, it wants to encourage contact and mobility in all sorts of ways, and hence multilingualism, and, on the other hand it wants to encourage the preservation (if not the promotion) of linguistic diversity. To be fair, the European Commission is not unaware of the tension: “While recognising the emergence of English as the most widely-spoken language in Europe, the Union also wants to make sure that this does not become, over time, a factor limiting linguistic diversity within its frontiers. This is why the Commission’s Action Plan has set the target of ‘mother tongue-plus-two’” (European Commission 2004: 22). This remedy thus offered by the Commission, however, is hardly promising for reasons linked to the core of the language spread mechanism (see LJ Chapter 1) and, even if it did achieve universal trilingualism, it would only mitigate, not remove, the fourfold tension delineated below. 10. As argued in LJ Chapter 1, the core of the dynamics of language spread under present conditions can be understood as the mutually reinforcing interaction between on the one hand the impact of the probability of using a language on the speed with which it is learned, and on the other the systematic adoption, in communication between plurilinguals, of the language of minimum exclusion (or minimex), i.e. the language best known by the participant who knows it least. See also Van Parijs (2004) for a sketchier formulation. 11. See previous note. 12. At the limit (with universal asymmetric bilingualism and random pairing of speech partners), the proportion of conversations held in any language i other than the lingua franca f shrinks from ni/N to (ni/N)2 (with ni the number of natives of language i and N the total population), while the proportion of conversations held in the lingua franca rises from nf/N to 1Σ(ni/N)2. Take, for example, the case of five languages, each with 20% of the native speakers. If the members of each language group only speak among themselves, 20% of the conversations are conducted in each of the five languages. If one of them becomes a lingua franca and conversation partners are picked randomly, 84% of all conversations happen in the lingua franca and 4% in each of the other four languages. 13. See, for example, Crystal (2000: Chapter 3) on this unstable “stage 2” of language shift. In a highly abstract model of trade interaction between two countries of equal sizes but unequal wealth, Choi (2002) shows how the greater relative profitability of specializing in jobs requiring bilingualism for members of the poorer country eventually leads, after several generations, to the exclusive use of the language of the richer country. 14. The former definition was proposed by Whittaker (1972) and the latter by Lande (1996). When all species are present in each habitat, Whittaker’s index is equal to 1 and Lande’s to 0. When the habitats are as specialized as the sizes of the species allow, Whittaker’s index is equal to the number of habitats K or the number of species in the landscape γ, whichever is smaller, while Lande’s is then given by (γ – γ/K). Lande’s measure is more convenient because it enables the richness of the landscape (α) to be decomposed into the sum of the average richness of the habitats (Σαi/K) and their differentiation (β). 15. This probability is (1/10)2 + (9/10)2 = 82/100 under the uneven 90/10 distribution, whereas it is (1/2)2 + (1/2)2 = 50/100 under the homogeneous distribution.

39

40 Philippe Van Parijs

16. See LJ Appendix 1 for a more precise presentation of this country Gini coefficient (which captures the degree of industrial specialization of countries), as well as of the symmetric industry Gini coefficient (which captures the degree of geographical concentration of industries). Inequality indices can thus serve to measure diversity, albeit in a rather convoluted way. 17. This trade-off is self-evident under the simple interpretation of territorial diversity that only takes richness (the number of languages) into account. But it also holds under interpretations that incorporate evenness (the relative spread of languages), as captured in the indices of isolation or specialization. Starting from a situation of minimum territorial diversity (same distribution of languages in each local unit), it is possible to increase territorial diversity while also increasing local diversity in some of the local units (through swaps that make the distributions more even in each of these), but only at the cost of sharpening concentration (and hence reducing local diversity) in others. 18. The proposed solution falls far short of maximizing linguistic diversity, or even of maintaining as much as possible the existing linguistic diversity, for two reasons. One is that only those languages will survive which manage to grab a territory with sufficient firmness, and hence only those linguistic communities which have the critical mass and the economic confidence that make the cost of this grabbing reasonable. Secondly, whereas the linguistic territoriality principle fosters territorial linguistic diversity by blocking the full replacement of local languages by the worldwide lingua franca, it also counteracts local linguistic diversity by operating as an unfettered local lingua franca. 19. Diversity is sometimes distinguished from sheer variety precisely on the ground that the former, unlike the latter, is regarded as intrinsically valuable. See, for example, Heyd (2005). 20. See, for example, Feld (1998: 199), quoted by Phillipson (2001: 113–114): “It is worthwhile to consider whether the EU should answer the call for uniformity on the issue of language business transactions and further protect itself against the potential onslaught of language regulation by each individual Member State. One potential action the EU might take would be to declare a common language in the EU market.” 21. Less explicit on the linguistic dimension but fundamentally on the same line is Mill’s (1861: 294–5) famous passage: “Nobody can suppose that it is not more beneficial to a Breton, or a Basque of French Navarre, to be brought into the current of the ideas and feelings of a highly civilised and cultivated people – to be a member of the French nationality, admitted on equal terms to all the privileges of French citizenship, sharing the advantages of French protection and the dignity and prestige of French power – than to sulk on his own rocks, the half-savage relic of past times, revolving in his own little mental orbit, without participation or interest in the general movement of the world. The same remark applies to the Welshman or the Scottish Highlander, as members of the British nation.” 22. Troja provides a somewhat less mythical parable for the curse of linguistic diversity. Why did the Acheans win the war? Perhaps because of the cunning of the wooden horse, but incomparably more because their koine, their shared language, enabled them to coordinate effectively, whereas the cacophony that prevailed between Trojans and their allies turned out to be a decisive disadvantage (see Ross 2005). What the Iliad documents is arguably just a special case of the tension between ethnic diversity and the efficient production of public goods, as studied for example by Alesina & La Ferrara (2000).



Chapter 1.  Linguistic diversity as curse and as by-product

23. For a critical survey of the most common arguments, see Crystal (2000: Chapter 2). 24. See, for example, UNDP (2005). 25. One interesting example is offered by the comparison of South Africa and Brazil, two recent federal democracies with a comparable level of development and a similarly high level of gross income inequality (Seekings 2004). Given the massive prevalence of Portuguese in Brazil, any reasonable measure of (evenness-sensitive) linguistic diversity should uncontroversially rank South Africa above Brazil. Yet South Africa displays a far higher level of economic solidarity (transfers reduce the Gini coefficient by 25%, compared to 0–7% in other developing countries), at least before the expansion of Brazil’s centrally funded social assistance programmes from the late 1990s onward. The bulk of this is achieved through a non-contributory old-age pension scheme, first restricted to whites, then extended to all, then differentiated in discriminatory fashion, and finally reunified – all under the apartheid regime. Paradoxically, in the South African case, the racial divide may have helped. The crucial step – the reunification of its old-age assistance scheme in the final years of the apartheid regime – became possible as a result of the establishing of strong intra-racial solidarity being followed by an attempt by the ruling racial group to deflect both domestic revolt and international opprobrium through universalising this solidarity. 26. Generous solidarity is of course more a matter of conquest than of generous sentiments. See, for example, Stephens (1979) on the negative correlation between ethnic diversity and the strength of the labour movement. La Ferrara’s (2004) findings about the negative impact of ethnic heterogeneity on participation in the production of a public good is also, albeit less directly, relevant. 27. On the basis of a large US survey, Putnam (2005) shows that there is a robust negative correlation between the degree of ethnic diversity (using a Simpson index and the five US Census categories: White/Black/Asian/Hispanic) and the level of trust (in members of one’s own as well as of the other groups). 28. Miller (2004) argues, on the basis of international comparisons of people’s revealed sense of justice, that there is little ethnic variation as regards conceptions of distributive justice. However, the finer structure of legitimate solidarity (“What counts as an illness?”, “To what extent and how should it be cared for collectively?” etc.) may vary significantly in culturally diverse communities, and, conceivably, as suggested by Anne Phillips (2004), for reasons that owe more to differences in their recent histories and present conditions than to differences in their remote cultural roots. 29. Against the claim made, for example, by Todd Gitlin (1995) or Brian Barry (2001), that multiculturalism policies undermine economic solidarity, Banting and Kymlicka (2004) showed, using various indices of the degree of economic solidarity achieved by tax and transfer systems, that this cannot be said to have been the case so far. Within the small sample consisting of the four Anglo countries in the context of which the above-mentioned claim is being made, the opposite seems to have happened: the more MCP-intensive countries (Canada and Australia), as far as immigrant minorities are concerned, fared better than the other two in terms of economic solidarity trends. However, if the sample is expanded to include a larger set of OECD countries, no pattern can be detected either way.

41

42

Philippe Van Parijs

30. See, for example, Schnapper (2002, 2004). 31. Forcing the autochthonous pupils of a school to attend the immigrant language classes when offered would no doubt be even more counterproductive than introducing it as an optional subject. On the contrary, managing to convince some of them (and their parents) of the interest they may have in learning languages such as Arabic and Turkish would be a welcome achievement, providing it is not so successful that it reduces significantly the opportunity and incentive for immigrant children to learn the local language. 32. There is a hint at the importance of considering the fine grain of MCPs in Banting and Kymlicka’s (2004) remark that some critics of MCPs seem to understand the latter in a narrow sense that entails separateness. But separateness need not be part of the explicit content of the policy for it to emerge from its implementation. In my example, it could be said that separateness is greater in the variant of the policy that will, I argue, end up producing less separateness: teaching Arabic as an optional part of the curriculum of nominally multi-ethnic schools looks far less “separatist” than gathering in one place children from various schools who (or whose parents) share the wish (for them) to study Arabic. 33. In the same vein, Putnam’s (2005) data indicate that local ethnic heterogeneity has a negative impact on trust and other indicators of social capital. They therefore suggest that increasing local homogeneity through segregation would increase the level of social capital. 34. See also David Miller’s (2004) conjecture that “segregation”, though worse than “integration”, should be better than “alienation” as far as trust and hence solidarity are concerned. 35. The twofold argument is parallel to the argument in favour of dissemination (versus compensation) as the most appropriate strategy for pursuing distributive justice in the context of the spread of the lingua franca (LJ, Chapter 3). 36. Soroka, Johnston and Banting (2004) persuasively observe, as one possible explanation for the resilience of the Canadian welfare state in the face of increasing ethnic diversity, that a welfare state can accommodate cultural diversity more easily if it relies on contributory benefits (as Canada does to a greater extent than Australia or the US): strengthening the relative importance of the insurance component of a welfare state is no doubt less demanding in terms of identification, but it also amounts to reducing the extent of the genuine (ex ante) solidarity it realizes. The point I am making here is independent of this observation. Even for a given level of generosity, the way the welfare state is structured matters to the tension there may be between diversity and solidarity. 37. See Van Parijs, Jacquet and Salinas (2000) for a detailed comparison of the various versions of this non-punitive “active welfare state”) and Vanderborght and Van Parijs (2005: Chapter 2) for a less technical treatment. 38. Another example is the extension of voting rights at local elections to all non-citizens. This arguably strengthens the weak channels of communication across communities by giving more reasons and pretexts to talk, more opportunities for friendships, connivances and solidarities. It would also increase the electoral incentive to look after neglected urban neighbourhoods, thereby counteracting their ethnic homogenisation. As mutual identification, joint responsibility for the common good, pride (and shame) of place keep being constructed through this and other means in mixed communities, starting from the most local level up, the challenge posed



Chapter 1.  Linguistic diversity as curse and as by-product

to economic solidarity by persistent cultural diversity – in particular in the form of maintenance of immigrant languages from generation to generation – will arguably lose some of its sharpness. 39. As the point is sometimes put (e.g. Salas Astrain 2004: Chapter 1), the “inter-cultural” saps the “multi-cultural”. 40. It is the principled symmetry (or reciprocity) of territoriality regimes that makes a blatant lack of symmetry (between the official language and the others) acceptable without embarrassment as regards local diversity. 41. See D’Antoni and Pagano (2002) and Pagano (2004). What they have in mind is essentially the cash transfer system. But the argument can plausibly be stretched to explain why the optimal level of public funding of higher education should be higher in Europe than in the US, or in Francophone than in Anglophone Canada – and indeed why this is actually the case. This is just another way of collectivising part of the risk involved in the expensive acquisition of potentially remunerative skills. And so is the adoption of rigid pay scales linked to the educational level, also prevailing far more in Europe than in the US. 42. Instead of lamenting all this, should one not rejoice at the constraints to which the national Leviathans are thereby subjected? The disciplining of rulers by a mobile tax base may sometimes provide a powerful and salutary lever for instilling respect for the rule of law, or for fostering the efficiency of the public sector, or for promoting a better match between the public goods supplied by a government and those the populations really want. But we live in a world in which globalisation, privatisation and trade-union decline make factor incomes ever more unequal. We also live in a world in which secularization, marital instability and geographical mobility keep eroding the once powerful income-sharing function of the family. For these two sets of reasons, the redistributive role of the tax system is more crucial than ever to the achievement of anything remotely resembling distributive justice. If it turns out that, in the wake of financial and industrial capital, human capital has to be immunised from redistributive taxation, it will be impossible for anyone who cares about distributive justice not to be deeply concerned. 43. This pressure of market competition on decentralised polities helps account for the negative correlation between federalism (vs unitarism) and economic solidarity. This is pointed out, for example, by Banting and Kymlicka (2004). 44. David Crystal (2000: 81) notes that, in a situation of universal diglossia, the dominated language is there “to express the identity of the speakers as members of their community… The dominant language cannot do this.” But also, “Only at the point where people have completely lost their sense of identification with their ethnic origins will the new language offer an alternative and comfortable linguistic home (at which point, the cultural assimilation would be complete)”.

References Alesina, Alberto & Eliana La Ferrara. 2000. “Participation in Heterogeneous Communities”. Quarterly Journal of Economics 115: 3.847–904.

43

44 Philippe Van Parijs

Alesina, Alberto, Arnaud Devleeschauwer, William Easterly, Sergio Kurlat & Romain Wacziarg. 2003. “Fractionalization”. Journal of Economic Growth 8: 155–94. Banting, Keith & Will Kymlicka. 2004. “Do Multiculturalism Policies Erode the Welfare State?” Van Parijs 2004b. 227–284. Bossert, Walter, Conchita D’Ambrosio & Eliana La Ferrara. 2006. “A Generalized Index of Fractionalization”. Paper presented at the conference of the Eni Enrico Mattei Foundation “Understanding Diversity: Mapping and Measuring”, Milan, January 2006. Choi, E. Kwam. 2002. “Trade and the Adoption of a Universal Language”. International Review of Economics and Finance 11: 265–75. Crystal, David. 2000. Language Death. Cambridge: Cambridge University Press. D’Antoni, M. & Ugo Pagano. 2002. “National Cultures and Social Protection as Alternative Insurance Devices”. Structural Change and Economic Dynamics 13: 367–386. Desmet, Klaus, Ignacio Ortuño-Ortín & Shlomo Weber. 2005. “Peripheral Linguistic Diversity and Redistribution”. Paper presented at the CORE conference, Université Catholique de Louvain, November 2005. European Commission. 2004. Many Tongues, One Family. Languages in the European Union. Luxemburg: European Commission, July 2004. Fearon, James D. 2003. “Ethnic and Cultural Diversity by Country”. Journal of Economic Growth 8: 2.195–222. Feld, Stacy Amity. 1998. “Language and the Globalization of the Economic Market: the Regulation of Language as a Barrier to Free Trade”. Vanderbilt Journal of Transnational Law 31: 153–202. Gitlin, Todd. 1995. The Twilight of Common Dreams: Why America is Wracked by Culture Wars. New York: Metropolitan Books. Heyd, David. 2005. “Cultural Diversity and Biodiversity: A Tempting Analogy”. Paper presented at the conference of the British Academy “Democracy, Equality and Justice”, London, December 2005. Hochschild, Jennifer, Traci Burch & Vesla Weaver. 2005. “Effects of Skin Color Bias in SES on Political Activities and Attitudes”. Paper presented at the Wiener Center Inequality and Social Policy Seminar, Harvard University, February 2005. Kruskal, Joseph B., Paul Black & Isidore Dyen. 1992. An Indo-European Classification: A Lexicostatistical Experiment. Transactions of the American Philosophical Society 82: part 5. Kymlicka, Will. 2004. “Concluding Reflections”. Van Parijs 2004b. 353–356. La Ferrara, Eliana. 2004. “Solidarity in Heterogeneous Communities.” Cultural Diversity versus Economic Solidarity ed. by Philippe Van Parijs, 69–80. Brussels: De Boeck Université. Laitin, David D. 2000. “What is a Language Community”. American Journal of Political Science 44: 1.142–155. Lande, R. 1996. “Statistics and Partitioning of Species Diversity, and Similarity among Multiple Communities”. Oikos 76: 5–13. Lee, Jennifer & Frank D. Bean. 2004. “America’s Changing Color Lines: Immigration, Race/Ethnicity and Multiracial Identification”. Annual Review of Sociology 30: 221–42. Lévi-Strauss, Claude. 1961. Race et histoire. Paris: Gonthier. Maignan, Carole, Gianmarco Ottaviano, Dino Pinelli & Francesco Rullani. 2003. “Bio-Ecological Diversity versus Socio-Economic Diversity: A Comparison of Existing Measures”. Paper presented at the conference of the Eni Enrico Mattei Foundation, Milan, January 2003. Massey, K.G. & Nancy A. Denton. 1988. “The Dimensions of Residential Segregation”. Social Forces 67: 281–315.



Chapter 1.  Linguistic diversity as curse and as by-product

Meillet, Albert. 1928 (1918). Les Langues dans l’Europe nouvelle. 2nd edition. Paris: Payot. Mill, John Stuart. 1861. “Considerations on Representative Government”. On Liberty and Other Essays ed. by J. Gray 3–467. Oxford: Oxford University Press, 1991. Miller, David. 2004. “Social Justice in Multicultural Societies”. Van Parijs 2004b. 13–32. Ottaviano, Gianmarco & Dino Pinelli. 2005. “Bio-Ecological Diversity versus Socio-Economic Diversity: A Comparison of Existing Measures”. Paper presented at the conference of the Eni Enrico Mattei Foundation & University of Bologna. Pagano, Ugo. 2004. “Cultural Diversity, European Integration and the Welfare State”. Van Parijs 2004b. 315–330. Patil, G.P. & C. Taillie. 1982. “Diversity as a Concept and its Measurement”. Journal of the American Statistical Association 77: 548–61. Phillips, Anne. 2004. “Comments on Miller and Soroka, Johnston & Banting.” Van Parijs 2004b. 59–64. Phillipson, Robert. 2001. “Principles for a Supra-National EU Language Policy”. L’Europe parlera-t-elle anglais demain? ed. by R. Chaudenson, 103–18. Paris: Institut de la francophonie & L’Harmattan. Phillipson, Robert. 2003. English-only Europe? Challenging Language Policy. London: Routledge. Putnam, Robert. 2005. “Can We Reconcile Community and Diversity?” Harvard University: Justice, Welfare and Economics Lecture, 6 April 2005. Ross, Shawn A. 2005. “Barbarophonos: Language and Panhellenism in the Iliad”. Classical Philology 100: 4.299–316. Salas Astrain, Ricardo. 2004. Etica Intercultural. (Re)lecturas del pensamiento latinoamericano. Santiago: UCSH. Schnapper, Dominique. 2002. La Démocratie providentielle. Essai sur l’égalité contemporaine. Paris: Gallimard. Schnapper, Dominique. 2004. “Linguistic Pluralism as a Serious Challenge to Democratic Life”. Van Parijs 2004b. 213–226. Skutnabb-Kangas, Tove. 2003. “(Why) Should Diversities Be Maintained? Language Diversity, Biological Diversity and Linguistic Human Rights”. Glendon Distinguished Lecture, York University, Toronto, 2003. Soroka, Stuart N., Richard Johnston & Keith Banting. 2004. “Ethnicity, Trust, and the Welfare State”. Van Parijs 2004b. 33–58. Stephens, John. 1979. The Transition from Capitalism to Socialism. Urbana: University of Illinois Press. Strubell, Miquel. 2006. “Building a Multilingual Europe: Discourse or Rhetoric?” The Challenge of Multilingualism in Law and Politics ed. by D. Castiglione & C. Longman. Oxford: Hart. UNDP. 2005. Human Development Report 2004: Cultural Liberty in Today’s Diverse World. New York: UNDP. Vanderborght, Yannick & Philippe Van Parijs. 2005. L’ allocation universelle. Paris: La Découverte. Van Parijs, Philippe, Laurence Jacquet & Claudio Salinas. 2000. “Basic Income and its Cognates”. Basic Income on the Agenda ed. by R.J. van der Veen & L. Groot, 53–84. Amsterdam: Amsterdam University Press. Van Parijs, Philippe. 2004a. “Europe’s Linguistic Challenge”. Archives européennes de sociologie 45: 1.111–152.

45

46 Philippe Van Parijs

Van Parijs, Philippe, ed. 2004b. Cultural Diversity versus Economic Solidarity. Brussels: De Boeck Université. (Downloadable from http://www.etes.ucl.ac.be/Francqui/Livre/Livre.htm.) Van Parijs, Philippe. Linguistic Justice for Europe and for the World. Oxford: Oxford University Press, in progress (referred to as LJ). Whittaker, R.H. 1972. “Evolution and Measurement of Species Diversity”. Taxon 21: 213–251.

chapter 2

Language diversity in the European Union An overview Patxi Juaristi, Timothy Reagan and Humphrey Tonkin

An examination of the extent and diversity of languages in Europe and specifically in the European Union reveals that there are over sixty indigenous languages currently spoken in, and historically established in, the territory of the EU, along with almost thirty sign languages. The authors identify and briefly examine the spoken languages, with due attention to the difficulty of working with language statistics. European languages within the EU can be divided into three categories, based on the number of native speakers of a given language in the EU itself. The authors also compare European linguistic diversity with linguistic diversity worldwide, and present a brief overview of language policies in the EU with respect to the designation of official languages in various member states.

Introduction There are, according to most estimates, somewhere in the general neighbourhood of 6,000 languages spoken around the world at the present time. The relative diversity of languages varies considerably, though, from one part of the globe to another, and Europe is an area in which linguistic diversity is comparatively fairly limited. Of the roughly 6,000 languages in the world (if we accept that total), only 200 are European (i.e. 3%), while 1,900 (32%) are African, 1,900 (32%) are Asian, 1,100 (18%) are Pacific and Oceanic, and 900 (15%) are American. To be sure, there are on-going debates about the precise criteria for differentiating languages from dialects, and how these criteria might apply in particular cases, but the overall picture is clear (see Comrie, Matthews & Polinsky 1996; Crystal 2000; Grenoble & Whaley 1998; Hagège 2000; Nettle 1999; Wurm 2001).

48 Patxi Juaristi, Timothy Reagan and Humphrey Tonkin

The relative lack of linguistic diversity in Europe, though, is a lack only in relative terms: the presence of linguistic diversity in Europe in general, and in the European Union (EU) in particular, is an important social, cultural, economic and political fact of life, and one which has significant implications both for Europe itself and for many other parts of the world. In this chapter, we will provide a general overview of the languages spoken in the EU, as well as the numbers of speakers and the status of these different languages. The information used in this chapter comes, in part, from the World Languages Report, a research project that began in 1998 and was finished in 2003 (Martí et al. 2005; see also www.amarauna-languages.com). This report had two very important aims to fulfil: on the one hand, to provide a description of the languages in various regions of the world and the problems affecting them; on the other hand, to try to recover languages in danger of extinction by facilitating decision-making on policies.1

Languages in the EU Because of the problem of deciding whether a variety of a language is a freestanding language or a dialect, it is quite difficult to specify the exact number of languages spoken in the EU: one’s estimate will vary depending on the criteria used to define what constitutes a language. For example, while Grimes (2000) mentions seven different Germanic languages spoken in Germany, many other scholars consider these “languages” to be varieties of German. Should Luxembourgian, Alsatian, Mocheno, Francique, Walser, or Zimbrian be considered dialects of German, or are they different languages? The answer, of course, is ultimately extra-linguistic in nature; as the American linguist Max Weinreich is credited with observing early in the 20th century, “A language is a dialect with an army and navy.” Thus, in the modern European context, Luxembourgian has recently become one of the official languages in Luxembourg along with German and French, although not long ago it was considered a dialect or variety of German. A similar question could be asked: Is Tsakonian a dialect of Modern Greek or is it a separate language? As Price (2000) has observed, Tsakonian is a dialect of Greek spoken in a number of villages in the SE Peloponnese between Mount Parnor and the Argolic Gulf. Not being mutually intelligible with other dialects, it has sometimes been reckoned to be a distinct language. According to different sources there are 300 speakers of Tsakonian nowadays.

The answer, again, is extra-linguistic rather than linguistic in nature. In this chapter, all language varieties which have any kind of official status, such as Luxembourgian, Meänkieli, or Rusyn,2 will be considered, but those dialects without any



Chapter 2.  Language diversity in the European Union

official status, such as Alsatian, Mocheno, Francique, Walser, Zimbrian, Tsakonian or Võro, will be ignored.3 An additional challenge in addressing linguistic diversity in the EU is presented by immigration into Europe, with the concomitant establishment of language communities speaking non-European languages. The languages spoken in these immigrant communities for the most part lack official recognition, regardless of their numbers of speakers. For example, because of the migrations of recent decades, more than 125 languages are spoken nowadays in the EU (Siguán 1995). In France, for instance, Algerian Arabic is spoken by more than 600,000 speakers; Kabyle, a language of the Algerian Berber family, by more than 500,000 speakers, and Tunisian Arabic by more than 200,000 speakers (Martí et al. 2005). In London, Punjabi has approximately 143,000 speakers, Gujarati around 138,000 and Hindi/Urdu 125,900 (Baker & Eversley 2000).4 Clearly a public debate about the role and rights of languages spoken by immigrants is needed in the EU, and to some extent this debate has already begun. In September 2005 the European Commission adopted A Common Agenda for Integration – Framework for the Integration of Third-Country Nationals in the European Union. This document provides some suggestions for action at both the EU and the national level, and member states are encouraged to strengthen their efforts with a view to developing comprehensive national integration strategies, while new ways of ensuring consistency between actions taken at the EU and national levels are proposed. Nevertheless, action with respect to the linguistic rights of immigrant communities remains largely unaddressed. Given the limitations of this chapter, such languages will not be discussed here in spite of their importance. We do want to note, though, the significant importance of such languages for the EU; in many instances, the number of speakers of these languages far exceeds those of many of the indigenous languages that we will be discussing here. Of the languages indigenous to Europe which are recognized in some way officially (though this recognition, and its implications, can vary dramatically), we find some 63 languages spoken as first languages in the EU, plus a few “special cases.”5 Our use of the term “first languages” leads us to issue a warning: in examining the following list and its accompanying comments, the reader should bear in mind the extreme unreliability of even the best of language statistics. What constitutes a speaker of a language? How are the statistics gathered? How reliable is the source of the statistics? Everything from methods of census gathering (including self-reporting) to hidden political agendas (and even the names used for individual languages) can skew statistics significantly, and, at the end of the day, one is still not entirely sure whether the numbers are limited to native or near-native

49

50

Patxi Juaristi, Timothy Reagan and Humphrey Tonkin

speakers (or speakers of the languages as first languages, which may or may not be the same thing), or include individuals who are able, or judge themselves able, to read, write, speak and/or merely understand the language in question6. We therefore suggest that the following statistics be taken only as rough guides and that the reader be wary of using them to make further statistical extrapolations. The languages are as follows: Albanian. In addition to speakers of Albanian in Albania itself, there are an additional 80,000 speakers in Italy (the Albanian spoken in Italy is called Arbëreshë; Price 2000 gives a figure of 100,000 speakers) and 150,000 speakers in Greece (the Albanian spoken in Greece is called Arvanitika). Aragones, spoken in the province of Huesca in Spain, specifically in the neighbourhood of four villages: Ansó, Ayerbe, Fonz and Benasque. Armenian, spoken, apart from Armenia itself (and neighbouring countries), in small communities in Cyprus (around 2,600 speakers), Poland (about 872 speakers), and Hungary (294 speakers). Asturian. The Asturias region of Spain has 1,083,576 inhabitants. Some 7.6% of the population of Asturias can understand, speak, read and write the language; 14.6% can understand, speak and read it; 26.8% can understand and speak it, and 33.4% can only understand it (Llera & San Martín 2003: 132). Basque, spoken in Spain, in the Autonomous Community of Navarre and in the Basque Autonomous Community, and also in France in some areas of the Département Pyrénées Atlantiques. The level of familiarity with and use of the language varies considerably by geographical area. The Basque Autonomous Community (Spain) is where the best organised plans for positive promotion have been set up and where the language receives most recognition, prestige, understanding and protection. See Aztiker (2002), Basque Government (2005), Hualde & Zuazo (2007). Belarusian, spoken by roughly eight million speakers in Belarus. There are an additonal 220,000 speakers of Belarusian in Poland. Bosnian, spoken in Slovenia. It is one of the three official languages of Bosnia and Herzegovina (the other two being Croatian and Serbian). Breton. According to Broudic (1999), the number of speakers of Breton, spoken in NW France, has decreased markedly: at the beginning of the 20th century there were around 1.5 million speakers, and in 1950 1.2 million speakers. Today, there are 369,000 speakers left.



Chapter 2.  Language diversity in the European Union

Bulgarian. Outside Bulgaria, the language is spoken in Greece (where it is called Pomak) by 30,000 inhabitants, and in Hungary. Catalan. There are 49,000 speakers in Andorra, 6.3 million in Spain, 203,000 in France, and 20,000 in Italy (Sardinia). Corsican, spoken on the island of Corsica (France), alongside French, which is the official language. According to Salminen (2005), the Corsican language is currently in danger of extinction. Greater protection for the language has been discussed as part of proposals for the increased autonomy of the region of Corsica from France. Croatian. Croatia itself is a candidate for EU membership, but there are also 103,000 speakers of Croatian in Austria, 32,132 in Hungary, and 3,500 in Italy. Croatian is also spoken in Slovenia. Cypriot Arabic, spoken by fewer than 1,000 people belonging to the Maronite community of Cyprus. Czech, spoken in the Czech Republic. Although with the establishment of the Czech Republic and Slovakia as independent nations, Czech and Slovak are now considered separate languages, they are in fact usually mutually intelligible. Danish. In addition to Denmark, Danish is spoken in Germany by 21,000 speakers, and in Greenland by 8,000. Although Greenland belongs to Denmark, it does not form part of the EU. Dutch, spoken in the Netherlands, Belgium, as well as in France on the Belgian border. English, the language of the UK and the bulk of the population of Ireland. It is also spoken in Gibraltar and on the island of Malta. Estonian. In addition to Estonia itself, there are 6,000 speakers of Estonian in Finland. There are also speakers of Estonian in Latvia (and in Russia: Nanovfszky 2004: 183). Faroese. Danish and Faroese are co-official in the Faeroe Islands, which belong to Denmark but are self-governing in most matters. Finnish. In addition to speakers in Finland, there are approximately 200,000 speakers of Finnish in Sweden. Franco-Provençal. Franco-Provençal is spoken in Switzerland, France and Italy.

51

52

Patxi Juaristi, Timothy Reagan and Humphrey Tonkin

French. There are 51 million speakers of French in France (and its overseas territories), 4 million in Belgium, 100,000 in Italy and 13,100 in Luxembourg. Frisian, spoken in the Netherlands (350,000) and Germany (12,000). Friulian, spoken in Italy and some villages of Austria and Slovenia. Gagauzian, a Turkic language spoken (according to Jacques Leclerc, of the University of Laval) by a few of the 12,000 Gagauzian people living in Bulgaria. An additional 154,000 people in the Republic of Moldova have Gagauzian as their first language. Galician, spoken in Spain in the Autonomous Community of Galicia and some areas of Asturias, Leon and Zamora; and in northern Portugal. German. German is the most extensively spoken language of the EU. Ethnologue states that there are 75.3 million speakers in Germany, 7.5 million in Austria, 150,000 in Belgium, 50,000 in the Czech Republic, 23,000 in Denmark, 250,000 in Hungary, 225,000 in Italy, 10,900 in Luxembourg and 500,000 in Poland. German is also spoken in Switzerland. Greek. There are 9.9 million speakers in Greece, 578,000 in Cyprus and 20,000 in Italy. Hungarian, spoken by 10.45 million speakers in Hungary, 1.4 million in Romania, 15,000 in Austria, 765,000 in Slovakia, and 10,000 in Slovenia. Irish, spoken in Ireland and in areas of Ulster (UK). The Irish Constitution recognizes two official languages: Irish and English. However, Irish is spoken only by a small number of people. See Martí et al. (2005: 98). Italian, spoken in Italy, the southern edge of Switzerland, the French side of the Franco-Italian border, and by 4,000 speakers in Slovenia. Karelian, spoken in Finland. There are also 118,000 speakers of Karelian in the Russian Federation (Nanovfszky 2004: 193 gives a figure of 95,000). Kashubian, spoken in Poland. Estimates of the number of native speakers vary dramatically: Price (2000) claims 150,000, Moscal (2004) claims 52,600. The Ethnologue figures are still lower. Ladin, spoken in Italy, in the provinces of Bolzano, Trento and Belluno. Latgalian, spoken in Latvia. 150,000 people speak it as a mother tongue. In 1919, it was declared the official language of Latgale, a region of Latvia.



Chapter 2.  Language diversity in the European Union

However, the present Latvian legislation considers Latgalian a historic variant of the Latvian language, although others regard it as a separate language. Latvian, spoken by 1.6 million in Latvia and an additional 200,000 in Estonia, Russia, Belarus, and Lithuania. Lithuanian, spoken in Lithuania, and an official language since 1922. Luxembourgian. Luxembourg declared its three languages – French, German and Luxembourgian – official over the whole of its territory in the 1980s. Macedonian. In the Republic of Macedonia (a candidate for EU membership) there are 1.4 million speakers, with an additional 10,000 in Greece, 15,000 in Albania and 180,000 in Bulgaria. Macedonian is linguistically close to Bulgarian and regarded by many in Bulgaria as a variety of Bulgarian (Kramer 1999). Maltese. In Malta, “Maltese is the national language, and co-official with English according to the Constitution. It is the language of the Parliament and the Law Courts. It is spoken generally at all administrative levels, but most (. . .) writing is carried out in English” (Martí et al. 2005: 122). Meänkieli, spoken in the most northern parts of Sweden (mainly in Norrbotten), around the valley of the Torne River. From a linguistic point of view Meänkieli is a mutually intelligible dialect of Finnish, but for historical reasons it has been awarded the status of a minority language in Sweden. On April 1, 2002, Meänkieli became one of the five nationally recognized minority languages of Sweden. Mirandese, spoken in Portugal, close to the Spanish border, in the town of Miranda do Douro and environs. Some regard Mirandese as a variety of Asturian. Occitan, spoken in Spain (Val d’Aran) and between Italy and France. There are around 5,000 speakers in Spain. Occitan, under the name of Aranese, is the official language of the Pyrenean valley of Aran alongside Spanish and Catalan. Polish, spoken not only in Poland, but also by 240,000 in Germany and 50,000 in Slovakia. Portuguese. There are approximately 10 million speakers of Portuguese in Portugal itself, but in excess of 177 million worldwide.

53

54

Patxi Juaristi, Timothy Reagan and Humphrey Tonkin

Romani. Liégeois (1992) says that policies have always favoured denial of the existence of the Gypsy people, their culture and their language, and Hagège (2000) points out that many speakers of Romani have disappeared through various attempted genocides. “There are a large number of Gypsies in Hungary, the Former Yugoslav Republic of Macedonia, and Romania, and they can be found in smaller numbers in other countries in Western and Eastern Europe . . . Romani is perhaps one of the European languages whose outlook is most uncertain and inevitably tied to marginalisation and assimilation of the Gypsy People” (Martí et al. 2005: 70). Romani is, however, an officially recognised minority language of Sweden, co-official in 79 rural communes and in one town (Budesti) in Romania, and the official language of Shuto Orizari (Macedonia) alongside Macedonian. Romanian, spoken in Romania and by 100,000 in Hungary. Russian, spoken also as a native language in Estonia (474,834 speakers), Latvia (861,600) and Lithuania (562,000). Through the influence of the Soviet Union, Russian has been until recently one of the most widely spoken foreign languages in countries surrounding Russia, such as Poland, Slovakia and the Czech Republic. Rusyn, also referred to as Ruthenian, Carpato-Ukrainian, or Lemkish, spoken in Slovakia, in the region of Preshow, and in Poland, in the Lemko region. It is also spoken in Ukraine, in NW Serbia (it is one of the official languages of the Serbian Autonomous Province of Vojvodina) and in eastern Croatia. Rusyn is regarded by some as a dialect of Ukrainian and by others as a distinct language. See Magosci (1993). Sami. Sami speakers are to be found in territory belonging to Norway, Sweden, Finland and the Russian Federation. See M. Aikio (1990), Helander (1990, 1994), Magga (1990), Kuoljok & Utsi (1993), Nanovfszky (2004). Sardinian, spoken on the island of Sardinia, Italy, it has recently been recognised as an official regional language in the Sardinian Autonomous Region; it can therefore be used for official purposes on the island. Scottish Gaelic. In 1891, Scottish Gaelic had 254,415 speakers (Aitken & McArthur 1979). The 1991 census shows only 69,510 able to “speak, read or write” Gaelic and 65,978 able to speak it. The numbers had declined to 65,674 and 58,652 by the time of the 2001 census. Therefore, of Scotland’s 5.5 million inhabitants, only a little over 1% speak the language. It receives some government support, but does not have formal official status, although it does have limited government recognition.



Chapter 2.  Language diversity in the European Union

Serbian, outside Serbia, spoken also by a small minority in Hungary. In Slovenia, 31,329 people indicate that Serbian is their first language, while another 36,265 indicate Serbo-Croatian (see Croatian, above). Slovak. Slovak is spoken in Slovakia by 4.9 million speakers, in Hungary by 11,000, and in Serbia by 80,000 speakers. It is also spoken in Ukraine, Poland, the Czech Republic, Croatia and Romania. Slovenian. In addition to Slovenia, Slovenian is also spoken in Austria (30,000 speakers) and in Italy (100,000). Sorbian, spoken in Germany. According to Martí et al. (2005), the language is in danger because “in many Sorbian-German marriages (and also in SorbianSorbian marriages), the parents do not pass on their mother tongue (Sorbian) to their children. Many parents think it is difficult for their children to learn two languages from the start and they reject Sorbian because they think German is the more important and useful language.” Spanish. According to Crystal (1997), Spanish is official or co-official in 20 states in the world. Swedish. Swedish has 7.8 million speakers in Sweden and 300,000 in Finland. Citizens of the Finland-Swedish minority, about 5% of the Finnish population, have the right to communicate with the authorities in their mother tongue. Moreover, after an educational reform took place in the 1970s, both Swedish and Finnish are compulsory school subjects. Tatar. While there are some 5 million speakers of Tatar in Russia, Tatars have been living in Estonia since the 1870s. There were 4,058 Tatars in Estonia in 1989, 3,315 in 1997 and 2,582 in 2000, of whom 1,229 (47.6%) spoke Tatar as a mother tongue, 1,295 Russian and 51 Estonian. Turkish, spoken in the EU in Germany (as an immigrant language), Bulgaria (845,000), Cyprus (177,000), Greece (128,380), and Romania (28,000). Ukrainian, also spoken in Poland (around Chelm), in the border area between Ukraine and Poland around the San River valley (Ukraine) and across the border in the Subcarpathian Voivodship (Poland). Walachian, also known as Aromanian or Vlach, spoken in areas of northern Greece and in Bulgaria (the Aromanians are also recognized as a minority in Macedonia).

55

56

Patxi Juaristi, Timothy Reagan and Humphrey Tonkin

Welsh. Wales has approximately 2.5 million inhabitants, of whom some 20% claim to have knowledge of Welsh. Of these, 69% can speak, read and write the language; 17% can only speak it. The rest can either speak and read, speak and write or simply read and write. Yiddish. In 2005 there were 3 million speakers of Eastern Yiddish. Western Yiddish is reported to have had an “ethnic population” of slightly below 50,000 in 2000. There are Yiddish-speaking communities in some eastern European countries (Estonia, Lithuania, Hungary, Poland, and the Czech Republic). In 2000, Sweden recognised Yiddish with four other minority languages: Finnish, Meänkieli, Sami and Romani.

Four virtually extinct languages, or languages currently returning from extinction as examples of language revival, should be added to this list: Cornish, spoken in Cornwall (UK). About 2,000 people understand the language to some degree and have varying experience in using it. About 200 claim fluency. Cornish is an interesting case in that it has no native speakers, having become extinct. What is now occurring is a case of language revival (see Payton 2004). Karaim. All together, around 2,800 people speak Karaim: 291 people in Lithuania, 680 in Russia, 1,400 in Crimea (Ukraine), fewer than 100 in Poland, fewer than 100 in Turkey, and several hundred in Israel. According to Schur (1995), M. El-Kodsi found 280 Karaites in Lithuania in 1991: 150 in Vilna, 50 in Panevezys and 80 in Troki. In Poland, M. El-Kodsi found some 150 Karaites in 1991: 50 each in Warsaw, Gdansk and Varcelova and four in Pele. Livonian, spoken in Latvia. This Finno-Ugric language is almost extinct, with only a handful of speakers of Livonian remaining (Nanovfszky 2004:199 gives the number as six). The UNESCO Red Book on endangered languages (Salminen 2005) describes Livonian as a nearly extinct language – though there are efforts at revival (Nanovfszky 2004: 201). Manx. The Isle of Man holds neither membership nor associate membership in the European Union. Although the UK takes care of its external and defence affairs, and retains paramount power to legislate for the island (it is a Crown dependency), the Isle of Man is self-governing. According to official census figures, 9.1% of the population claimed to speak Manx in 1901; in 1921 the percentage was only 1.1%. By the mid-20th century, only a few elderly native speakers remained (the last died in 1974), but by then a scholarly revival had begun to spread to the populace and many had learned Manx as a second language. In this respect, Manx resembles Cornish as a revived language. In the 2001 census, 1,689 out of a



Chapter 2.  Language diversity in the European Union

population of about 71,000 claimed to have some knowledge of Manx. See Nettle & Romaine (2000). Finally, there is a significant quantity of sign languages used by deaf people throughout the EU.7 Although most of these sign languages do not have official status, some do (thus Swedish Sign Language has had official status for deaf people in Sweden since 1981). In addition, in practice in the EU the deaf are widely provided with interpreters and other accommodations regardless of explicit constitutional or legal guarantees. These sign languages are most often passed on from student to student in residential schools for the deaf. These sign languages are not related to the spoken language in the surrounding hearing community – that is to say that, for example, French Sign Language is not simply French in a gestural/visual modality, but rather is a completely separate language with its own linguistic system. Among the identified sign languages used in the EU are the following: ustrian Sign Language A Belgian Sign Language British Sign Language Catalonian Sign Language Czech Sign Language Danish Sign Language Dutch Sign Language Estonian Sign Language Finnish Sign Language Finnish-Swedish Sign Language8 French Sign Language German Sign Language Greek Sign Language Hungarian Sign Language Irish Sign Language Italian Sign Language Latvian Sign Language Lithuanian Sign Language Lyons Sign Language9 Maltese Sign Language Norwegian Sign Language Polish Sign Language Portuguese Sign Language Slovakian Sign Language Spanish Sign Language Swedish Sign Language

57

58

Patxi Juaristi, Timothy Reagan and Humphrey Tonkin

S wiss-French Sign Language10 Swiss-German Sign Language Swiss-Italian Sign Language

In spite of having 7% of the world’s total population, the EU thus includes only 1% of the world’s linguistic diversity (with Europe as a whole including only 3% of that diversity) (see European Commission 2006). Why is this so? The reasons are historical, economic, political and cultural. Nevertheless, most of the governments of the 27 member states of the EU have applied and are still applying all too efficient policies of discrimination, marginalisation and assimilation (Martí et al. 2005).

Speakers of languages in the EU There are many ways of classifying the languages spoken in the EU. The following estimates are based on “native speakers” (a slippery category that requires that we take these statistics only as a very rough guide: see our caveat above) within the EU. Thus, the figures provided may seem in some instances quite low (as in the case of English), because neither native speakers in other parts of the world, nor second or additional language users are included. Using the approximate number of native speakers of the language as the criterion for classification in this manner, the languages of the EU can be divided into three groups: those with more than ten million native speakers, those with between one million and ten million, and last, those with fewer than one million. The first group of languages consists of those languages with more than ten million speakers each – in all, 411 million EU citizens – meaning that nearly 85% the EU population of around 490 million speaks fewer than 20% of its languages. These languages, in order of the number of native speakers in the EU, are: erman (c. 84 million) G English (c. 55 million) French (c. 55 million) Italian (c. 55 million) Spanish (c. 40 million) Polish (c. 39 million) Dutch (c. 21 million) Romanian (c. 19 million) Hungarian (c. 12 million) Greek (c. 10.5 million)



Chapter 2.  Language diversity in the European Union

zech (c. 10.3 million) C Portuguese (c. 10 million)

Although the future of these languages is not based solely on the number of native speakers, this group seems to be very prosperous. The existence of independent political states associated with these languages has meant that they enjoy a favourable situation (Martí et al. 2005). Moreover, as well as being official in their own countries and in the EU, a few of these eleven languages, such as English, French, Portuguese and Spanish, are among the most widespread languages of the world. English, Spanish, French and Portuguese are today official languages in most of America and Africa, and in large areas of Asia and the Pacific. According to Crystal (1997), English is the official or co-official language in more than 70 states of the world, French in 30 states, Portuguese in 6 states and Spanish in 20 states. The second group of languages in the EU comprises fifteen languages having between approximately one million and ten million speakers. In order of the number of native speakers in the EU, these languages are: ulgarian (c. 8.5 million) B Swedish (c. 8 million) Catalan (c. 6 million) Slovak (c. 5.5 million) Danish (c. 5 million) Finnish (c. 4.7 million) Galician (c. 3 million) Lithuanian (c. 2.9 million) Slovenian (c. 2 million) Russian (c. 1.9 million) Latvian (c. 1.8 million) Sardinian (c. 1.6 million) Occitan (c. 1.5 million) Romani (c. 1.5 million) Estonian (c. 1 million)

Most of the languages in this group lack the power and distribution of the languages in the former group, but most of them are official both in their own states and in the EU. Nevertheless, some of the languages with more than one million speakers, such as Romani, or others in a minority situation, such as Occitan, seem to have a very uncertain future unless decisive actions are taken in their favour. As mentioned previously, the longevity of the languages does not lie only in the number of speakers. Although Russian has been included in this group, its significant place in some of the member states (and, of course, its status as a major language

59

60 Patxi Juaristi, Timothy Reagan and Humphrey Tonkin

outside of the EU) should be taken into account. For example, in Estonia some communities located near the Russian border have Russian as their dominant language, and in Latvia and Lithuania 29% and 6% of the population, respectively, is Russian-speaking. Furthermore, Russian is one of the most widely spoken foreign languages in countries surrounding Russia, such as Lithuania (where 83% of the total population speaks some Russian), Latvia (59%), Estonia (53%), Poland (28%), Slovakia (30%) and the Czech Republic (21%) (See Eurobarometer, Report Number 50). It is important to note here that these percentages refer to individuals who have studied Russian as a second or foreign language, rather than to native speakers. The third group of languages in the EU includes languages with fewer than one million native speakers, as well as all of the sign languages mentioned earlier. Estimates for numbers of speakers of various sign languages are difficult to obtain, but are, without exception, tiny compared to other languages. As for the other (spoken) languages with less than one million native speakers, listed here, statistics on numbers of speakers vary wildly. Because of this difficulty, the list is presented in alphabetical order (we exclude Cornish, which has no native speakers, and Manx). lbanian A Aragones Armenian Asturian Basque Belarusian Bosnian Breton Corsican Croatian Cypriot Arabic Faroese Franco-Provençal Frisian Friulian Gagauzian Irish Karaim Karelian Kashubian Ladin Latgalian



Chapter 2.  Language diversity in the European Union

ivonian L Luxembourgian Meänkieli Macedonian Maltese Mirandese Rusyn Sami Scottish Gaelic Serbian Sorbian Tatar Turkish Ukrainian Walachian Welsh Yiddish

There are three languages in this list (Irish, Maltese and Luxembourgian) which are official in their respective EU member states, and nine languages which are official in other European states outside the EU (Albanian, Armenian, Belarusian, Bosnian, Croatian, Macedonian, Serbian, Turkish, and Ukrainian), the others being considered minority languages, meaning that they are only spoken by a minority of the population of a state. There are two additional languages spoken in the UK that need to be mentioned here as well: Cornish and Manx. Although these languages were extinct (the last native speaker of Cornish, for instance, died in 1777), language revival efforts are currently underway for both of them, and each has a small but vibrant speaker community. The political status of the Isle of Man is, however, unusual (see above). We can further distinguish three groups among these minority languages. First, there is the singular case of Irish, the national language and main official language of the Republic of Ireland, and an official and working language of the EU. Second, there are 13 languages that have co-official status or some kind of ­official recognition at the substate level. These languages are: asque B Faroese Franco-Provençal Frisian Friulian

61

62

Patxi Juaristi, Timothy Reagan and Humphrey Tonkin

adin L Meänkieli Mirandese Sami Scottish Gaelic Sorbian Welsh Yiddish

The remaining 14 indigenous languages include the most threatened of the languages in the EU, as well as languages with small numbers of speakers in the EU but larger numbers and/or official status outside of the EU. These languages are: ragones A Asturian Breton Corsican Cypriot Arabic Gagauzian Karaim Karelian Kashubian Latgalian Livonian Rusyn Tatar Walachian

Nettle (1999) indicates the figure of 10,000 speakers as the threshold below which a language can have problems surviving. In the EU, four languages have around 10,000 speakers – Aragones, Karelian, Mirandese and Macedonian, and six languages have fewer than 10,000 speakers – Armenian (4,000), Cornish (200), Cypriot Arabic (under 1000), Karaim (400), Livonian (10), Tatar (around 2000) (we do not include the special case of Manx; see above). Armenian, Macedonian and Tatar, though numerically tiny in the EU itself, are official languages elsewhere (Tatar is a formally recognised regional language in Russia), and Karelian and Gagauzian have a considerably larger population outside the EU than within it, but, with these exceptions, there is a strong possibility that the other languages will be extinct in a matter of years. To summarize, the EU can be considered a diverse community even though it has fewer languages than other parts of the world; that is, there are only 68 spoken languages native to the territory covered by the EU (1% of the world’s spoken



Chapter 2.  Language diversity in the European Union

languages, and 34% of all European autochthonous languages, if we agree that there are 200 [spoken] languages in Europe). Very few languages are widely distributed and most are spoken by relatively few people. As we have noted, some 85% of the EU population – 411 million citizens – is registered as speaking under 20% of all EU languages (please note our caveat about immigrant languages, which are not included in this survey: see above). Upwards of half of all EU languages can be classified as minority languages, and they are spoken (as first languages, or native languages, or additional languages – depending on the way the question is posed, the statistics are gathered, and minority status is defined) by only a very small percentage of the population. In most cases, the official status and the sociolinguistic situation of the languages correspond to the number of speakers.11

Language status and language policy in the EU Different methods for dealing with language diversity are used by member states within their borders. We can distinguish five principal models with respect to the issue of official status: •

Member states with a single official language: Bulgaria, Estonia, France, Greece, Latvia, Lithuania, and Poland. • Member states with a single official language, but in which other languages enjoy certain state support in some parts of the country or official recognition limited to particular municipalities: Austria, the Czech Republic, Germany, Hungary, Portugal, Romania, Slovakia, Slovenia, Sweden, and the United Kingdom. • Member states which recognise co-official languages in certain geographical areas: Spain, Denmark, the Netherlands and Italy. • Member states which are language federations: Belgium. • Bilingual and multilingual member states: Cyprus, Finland, Ireland, Luxembourg, and Malta.

1.

Member states with a single official language: Bulgaria, France, Estonia, Latvia, Lithuania, Greece, and Poland

There are seven states within the EU which have a single official language and encourage its exclusive use (the best known example of this language model is France). The language policy followed by these countries is based on the belief that the use of just one language facilitates the cohesion and progress of the state.



63

64 Patxi Juaristi, Timothy Reagan and Humphrey Tonkin

In other words, these states have married the concepts of a common language, nation and state (Lastra 1992), and hold to the view that a modern state must have a single language common to all its citizens. The consequences of this policy have been to marginalise other languages, lower their prestige, endanger them and even cause their disappearance (Crystal 2000). Moreover, the states following this language model generally promote a monolingual school system, which poses enormous contradictions for minority language communities. Consider the evolution of Breton in France: as we have already noted, at the end of the nineteenth century Breton had almost one and a half million speakers. Nowadays, there are around 369,000. But Breton is not alone. A similar situation confronts all minority languages in France, such as Basque, Catalan, Corsican, Franco-Provençal and Occitan. The same is true of Tatar in Estonia, Romani throughout Eastern Europe, and Kashubian in Poland. If language policy is not changed in these countries, all these languages are in danger of disappearing in the near future. In addition, in several of these cases language policy is explicitly utilised for ideological and political ends. In Estonia, Latvia, and Lithuania, for example, such policies are clearly directed at the deliberate assimilation of speakers of Russian.

2.

Member states with a single official language, but in which other lan- guages enjoy certain state support in some parts of the country or official recognition limited to particular municipalities: Austria, the Czech Republic, Germany, Hungary, Portugal, Romania, Slovakia, Slovenia, Sweden, and the United Kingdom

Ten member states of the EU, apart from having one official language, recognise other languages in some geographic regions of the country, usually in specific municipalities or areas. In 2002, Sweden recognised Sami as its fifth minority language (the others, previously recognized, are Meänkieli, Standard Finnish, Romani and Yiddish), giving it formal standing in government agencies, courts, preschools and nursing homes in four municipalities (Arjeplog, Gällivare, Jokkmokk and Kiruna). As we have already noted, Sweden has also recognised Swedish Sign Language. Italian is co-official in some cities of Istria (Slovenia), as is Mirandese in some municipalities of northeastern Portugal, mainly in Miranda do Douro and surrounding areas, where it was granted official recognition by the Portuguese Parliament in 1998. In the United Kingdom, Welsh, Scottish Gaelic, Cornish, Manx and Irish do not enjoy official status in their respective territories (Wales, Scotland, Cornwall, the Isle of Man, and Northern Ireland). Consequently, English is slowly spreading



Chapter 2.  Language diversity in the European Union

in the traditional territories of these languages (Moreno Cabrera 2000). Certainly, the Welsh Language Act of 1993 purports to give “effect, so far as is both appropriate in the circumstances and reasonably practicable, to the principle that in the conduct of public business and the administration of justice in Wales the English and Welsh languages should be treated on a basis of equality,” but it does not accord individual linguistic rights to Welsh speakers. However, it should be noted that while this is true in broad terms, Welsh speakers do in fact have an absolute right to use the language in courts of law, which is certainly, albeit in only one domain, an individual linguistic right. In the case of Welsh, a language board has been established to promote and facilitate the use of the language, and in particular to advise public bodies which provide services to the public in Wales, on the ways in which effect may be given to the principle of treatment “on a basis of equality.” A further example of such a policy is found in Romania, where “there are two hours a week on Romanian television for the Roms. There are five magazines for the Roms with writing in Romany. There are also three radio programmes in Romany” (Martí et al. 2005: 206). Also focused on the domain of language use rather than on a particular geographic area, this case would appear to be similar to the others included here, although no official status is granted to Romani. Because they do not enjoy official status, most of the languages in this group may have difficulty in surviving, in spite of having limited state support.

3.

Member states which recognise co-official languages in certain geographical areas: Spain, Denmark, the Netherlands and Italy

Four EU member states, instead of having several official state languages, have chosen to have a single official language in the country as a whole, while granting official status to regional languages in the territory in which they are spoken, alongside the official language of the state as a whole. For example, Catalan, Galician, Basque, Faroese, Frisian, Sardinian, Friulian, Ladin and Occitan are all official only in some areas of the relevant countries (Spain, Denmark, the Netherlands and Italy). This is also the case, for instance, of Frisian in Germany, where Frisian is officially protected as a minority language, according to the European Charter for Regional or Minority Languages, in the province of Schleswig-Holstein and in the Saterland region of Lower Saxony. Although German politicians are currently passing laws to promote this language, there remain serious problems implementing these laws. As Hagège (2000) points out, an officially recognised language has a degree of legal protection, and this is the first step toward the preservation and promotion of a minority language – but having official status

65

66 Patxi Juaristi, Timothy Reagan and Humphrey Tonkin

does not mean that the future of a language is secure: such recognition mush be accompanied by decisive action in its favour. For instance, on the other side of the German border, in Friesland, a province of the northern Netherlands, Frisian enjoys a status equivalent to Dutch as the official language of the territory. For all these efforts on behalf of Frisian, it remains one of the most endangered minority languages in the EU (Fishman 1991). The Basque language took a very important step forward when it was proclaimed an official language in the Basque Autonomous Community in 1979 and in the Autonomous Community of Navarre in 1982. This recognition was arguably a key step for the survival of the language, although there remain many problems in preserving this linguistic heritage, especially in Navarre, due to the fact that any action, law, movement or initiative taken in favour of preserving Basque is apt to be resisted by the current regional government. The countries included in this group may, of course, grant official status to some languages while denying the same status to other languages spoken within their territory. For instance, Basque, Catalan, Galician and Occitan are official in Spain, while Asturian is not. As we have noted, in some cases the same linguistic community extends to territories belonging to different states: sometimes a language is co-official in one state and not recognized at all in another. Examples of this phenomenon include Basque (Spain and France) and Occitan (France, Spain and Italy). As long as this situation does not change, such linguistic communities will have difficulties promoting the language and surviving.

4.

Member states which are language federations: Belgium

Belgium is the only language federation in the EU at the present time. Three languages – Dutch, French and German– are official. The three are prestigious languages, and they are all promoted. Belgium is divided into three communities, the Dutch-speaking Flemish community, the French-speaking community and the German-speaking community. At the same time, it is also divided into three regions: Brussels, mainly consisting of a mixed Dutch- and French-speaking population of 980,000; the Flemish region, mainly Dutch-speaking, with 5.9 million inhabitants; and Wallonia, mainly French-speaking, with 3.36 million inhabitants. About 60 percent of the country is therefore Dutch-speaking; French is the second-most-spoken language (with about 40 percent) and finally German is spoken only by less than 1% of the population. Although each language has a completely different sociolinguistic situation, all three languages are official and treated as such. It is also important to note that Belgium is a federal country, and



Chapter 2.  Language diversity in the European Union

so Dutch and French, in particular, have relatively little standing outside their own regions.

5 .

Bilingual and multilingual member states: Cyprus, Finland, Ireland,  Luxembourg, and Malta

There are five bilingual and/or multilingual countries in the EU: Cyprus (Greek and Turkish), Finland (Finnish and Swedish), Ireland (Irish and English), Luxembourg (French, German and Luxembourgian), and Malta (Maltese and English). With the exception of Cyprus, where a special political situation exists, in all the other member states the constitution indicates that there is more than one official language in the country. Consequently, in theory, all constitutional languages are protected by the law. According to its constitution, Finland is a bilingual country; that is to say, citizens of the Swedish-speaking minority, constituting about 5% of the Finnish population, have the right to communicate with the authorities in their mother tongue. Moreover, since the educational reform of the 1970s, both Swedish and Finnish are not only compulsory school subjects, but are also mandatory in final examinations. Nonetheless, even in Swedish-speaking towns and municipalities, Finnish is the dominant language. The Sami people living in Finland were also granted the right to use the Sami language for all government services by the Sami Language Act of 1991 (P. Aikio 1994). Finally, the Sami Language Act of 2003 granted the Sami language official status in five municipalities of Finland. Ireland is also a bilingual country. The constitution recognizes two official languages, Irish being the “national language” of the Republic of Ireland and its main official language, and English the second official language. However, English is spoken by the majority of the population and Irish only by a small number. Thus, official bilingualism does not necessarily mean all languages mentioned in the constitution are equally secured.

Conclusion To understand the sociolinguistic map of the EU, we must take into account that various member states use different criteria to deal with language diversity: a wide variety of legislation, policies and constitutional provisions deals with language issues. In this chapter we have provided a brief overview of explicit official language policies in EU member states, but this is only the tip of the iceberg for understanding the complexities and intricacies of language matters in the EU. Perhaps

67

68 Patxi Juaristi, Timothy Reagan and Humphrey Tonkin

as important as the official language policies of member states are policies at the macro-level that address the EU as a multinational and multilingual community, and policies at the micro-level that define the individual’s language rights. These are both areas in which there has been extensive debate and both political and scholarly discussion in recent years. Furthermore, they are both areas in which no consensus has been reached in spite of the urgency to do so. And, of course, as the EU expands, the challenges at both levels will only increase. The lack of decision in these areas may constitute, unfortunately, a decision in its own right. If the EU as a whole continues to avoid making difficult macrolevel decisions, and if individual member states continue to be reluctant to pass legislation on minority languages, this may well result in damage to currently endangered languages in the EU. We believe that public debate is essential in four broad areas: •

The role of languages which currently lack formal recognition within the EU; • The situation of linguistic communities that speak official languages of the EU but whose linguistic rights are not protected in the areas where these communities are settled; • Criteria for the linguistic operation of the institutions of the EU; • Most importantly, the prospects for a global language policy for the EU. What is intriguing to us is that none of these issues is new or novel. They are all issues about which there are extensive bodies of relevant literature, and about which thoughtful individuals have spent a great deal of time and effort (see, for instance, Extra & Gorter 2001, Fettes & Bolduc 1998, Maurais & Morris 2003, Phillipson 2003, Spolsky 2004, and Wright 2004). And yet, decisions continue to be avoided, perhaps for understandable political reasons. In any event, we would hope that this volume, and the chapters it contains, would help to advance this important on-going dialogue, and might perhaps even serve to elevate the discussion.

Notes 1. Unless otherwise indicated, estimates for the numbers of speakers for particular languages discussed in this chapter are drawn from either Ethnologue (see Grimes 2000) or from Euromasaic data. 2. Luxembourgian (Lëtzeburgesch) is an official language in Luxembourg; Meänkieli has official recognition in Sweden; and Rusyn is one of the official languages of the Serbian Autonomous Province of Vojvodina.



Chapter 2.  Language diversity in the European Union

3. The Võro language is a dialect of Estonian, but it has its own literary language and is in search of official recognition as an autochthonous regional language of Estonia. We do not mention Zimbrian, or Cimbrian, in our list of languages, but its situation is not dissimilar. Cimbrian is an ancient Bavarian dialect, but now sufficiently isolated as to constitute a distinct idiom – and it enjoys protection under Italian State Law 482 (see Coluzzi 2005). Does this make it a language, and does it make it official? We are very aware of the relative arbitrariness of the language/dialect distinction. The question of what constitutes an “official” language is again a vexed and ultimately somewhat subjective issue. We return to this issue later in the chapter. 4. For a review both of the smaller languages of Europe and of immigrant languages, see Extra & Gorter 2001. 5. The list is based in part on Martí et al. 2005 and in part on data from Euromosaic, http:// europa.eu.int/comm/education/policies/lang/languages/langmin/euromosaic/index_en.html. For Bulgaria and Romania, see http://www.tlfq.ulaval.ca/axl/europe/europeacc.htm. Additional information comes from Grimes 2000. We have limited our survey to the continent of Europe, omitting languages indigenous to French overseas territories and to the Spanish enclaves in North Africa. We have also omitted Esperanto, spoken as a second language by small numbers of people in most EU countries, and occasionally as a native language. On languages supported by the European Bureau for Lesser-Used Languages (EBLUL), see Ammon 2003. 6. Of course, the relationship between these skills will vary greatly, and native ability to speak may not signify, even among those literate in other languages, the ability to read or write the language. Thus, in the 2001 UK census, 58,652 people claim to speak Scottish Gaelic, but only 45,320 read it and 33,814 write it. The two latter figures are actually higher than they were ten years ago – probably an indication that the language is receiving more serious attention in the schools, possibly among non-native speakers. On the problems of linguistic data-gathering, see Edwards 1994: 35–39. 7. See Reagan (2006). Tove Skutnabb-Kangas has gone so far as to suggest that, “There may be as many Sign languages as oral languages – all fully-fledged languages, logical, systematic, capable of expressing everything in the world, provided enough resources are devoted to their development – just as all oral languages are fully-fledged languages, logical, and so on . . .” (2000: 226). 8. Finnish-Swedish Sign Language is a remnant of a closed Swedish school for the deaf. It is used only by older adults in private spheres, and there are at most 150 users. 9. Lyons Sign Language is used in an area approximately 250 miles from Paris. No figures are available on the number of users. Interestingly, LSL has little mutual intelligibility with French Sign Language. 10. In Switzerland, there are three distinct sign languages, each a variety of the sign language used by the hearing linguistic community in the relevant country; thus, Swiss-French Sign Language is related to French Sign Language, Swiss-German Sign Language is related to German Sign Language, and Swiss-Italian Sign Language is related to Italian Sign Language. In all three cases, however, the Swiss variety is distinctive.

69

70 Patxi Juaristi, Timothy Reagan and Humphrey Tonkin

11. Ammon 2005 provides a useful taxonomy of languages used in Europe, ranging from “European lingua franca(s)” at the top to “languages with debated language or dialect status” and “dialects” at the bottom.

References Aitken, A.J. & Tom McArthur. 1979. Languages of Scotland. Edinburgh: Chambers. Aikio, Marjut. 1990. “The Finnish Perspective: Language and Ethnicity”. Arctic Languages: An Awakening, ed. by Dirmid R. F. Collis, 367–400. Paris: UNESCO. Aikio, Pekka. 1994. “Development of the Political Status of the Sami People in Finland.” Diedut 1.39–43. Ammon, Ulrich. 2003. “Present and Future Language Conflicts as a Consequence of the Integration and Expansion of the European Union (EU)”. Ecologia Linguistica, ed. by Ada Valentini, Piera Molinelli, Pierluigi Cuzzolin & Giuliano Bernini, 393–405. Rome: Bulzoni. Ammon, Ulrich. 2005. “Some Problems of EU Language Policy and Discussion of Possible Solutions.” Atti del 4º congresso di studi dell’Associazione Italiana di Linguistica Applicata, Modena, 2004, ed. by Giorgio Banti, Antonietta Marra & Edoardo Vineis, 193–207. Perugia: Guerra. Aztiker Ikergunea. 2002. Euskal Herria Datuen Talaiatik. Astigarraga: Udalbiltza. Baker, Philip & John Eversley. 2000. Multilingual Capital: The Languages of London’s Schoolchildren and Their Relevance to Economic, Social and Educational Policies. London: Battlebridge Publications. Basque Government. 2005. III. Mapa Soziolinguistikoa. Vitoria-Gasteiz: Basque Government Publishing. Broudic, Fañch. 1999. “L’ évolution des bilingues bretonnants au XXe siècle”. Klask 5.179–192. Cerquiglini, Bernard. 1999. “Les langues de la France”. Rapport au Ministre de l’Education Nationale, de la Recherche et de la Technologie, et à la Ministre de la Culture et de la Communication, Paris, April 1999. http://www.culture.gouv.fr/culture/dglf/lang-reg/rapport_cerquiglini/langues-france.html#ancre276758) Coluzzi, Paolo. 2005. “Language Planning for the Smallest Language Minority in Italy: The Cimbrians of Veneto and Trentino-Alto Adige”. Language Problems and Language Planning 29.3: 247–269. Comrie, Bernard, Stephen Matthews & Maria Polinsky. 1996. The Atlas of Languages: The Origin and Development of Languages Throughout the World. New York: Facts On File. Crystal, David. 1997. English as a Global Language. Cambridge: Cambridge University Press. Crystal, David. 2000. Language Death. Cambridge: Cambridge University Press. Dixon, Robert M.W. 1997. The Rise and Fall of Languages. Cambridge: Cambridge University  Press. Edwards, John. 1994. Multilingualism. London: Routledge. European Commission. 2006. EU Integration Seen Through Statistics: Key Facts of 18 Policy Areas. Luxembourg: Eurostat. Extra, Guus & Durk Gorter, eds. 2001. The Other Languages of Europe. Clevedon: Multilingual Matters.



Chapter 2.  Language diversity in the European Union

Fettes, Mark & Suzanne Bolduc, eds. 1998. Al Lingva Demokratio / Towards Linguistic Democracy / Vers la Démocratie Linguistique. Rotterdam: Universala Esperanto-Asocio. Fishman, Joshua. 1991. Reversing Language Shift: Theoretical and Empirical Assistance to Threatened Languages. Clevedon: Multilingual Matters. Grimes, Barbara, ed. 2000. Ethnologue: Languages of the World (13th ed.). Dallas: Summer Institute of Linguistics. See also www.ethnologue.com. Hagège, Claude. 2000. Halte à la mort des langues. Paris: Odile Jacob. Hale, Ken, and others. 1992. “Endangered Languages.” Language 68.1: 1–42. Helander, Elinia. 1990. “Situation of the Sámi Language in Sweden”. Arctic Languages: An Awakening, ed. by Dirmid R. F. Collis, 401–417. Paris: UNESCO. Helander, Elinia. 1994. “The Sami People: Demographics, Origin, Economy, Culture”. Diedut 1.23–34. Hualde, José Ignacio & Koldo Zuazo. 2007. “The Standardization of the Basque Language”. Language Problems and Language Planning 31.2: 143–168. Junyent, Carme. 1999. La Diversidad Lingüística: Didáctica y Recorrido de las Lenguas del Mundo. Barcelona: Octaedro. Kramer, Christina. 1999. “Official Language, Minority Language, No Language at All: The History of Macedonian in Primary Education in the Balkans”. Language Problems and Language Planning 23.3: 233–250. Kuoljok, Sunna & John Erling Utsi. 1993. The Sami: People of the Sun and Wind. Ájtte & Jokkmokk: Swedish Mountain & Saami Museum. Lastra, Yolanda. 1992. Sociolingüística para Hispanoamericanos. México: El Colegio de México. Liégeois, Jean-Pierre. 1992. “Les Tsiganes: Situation d´une minorité non territoriale.” Les Minorités en Europe: Droits linguistiques et droits de l´homme ed. by Henri Giordan, 419–443. Paris: Kimé. Llera, Francisco & Pablo San Martín. 2003. II Estudio Sociolingüístico de Asturias 2002. Uviéu: Academia de la Llingua Asturiana. Magga, Ole Henrik. 1990. “The Sámi Language in Norway”. Arctic Languages: An Awakening, ed. by Dirmid R. F. Collis, 419–436. Paris: UNESCO. Magosci, Paul Robert, ed. 1993. The Persistence of Regional Cultures: Rusyns and Ukrainians in the Carpathian Homeland and Abroad. New Jersey: East European Monographs. Martí, Felix, Paul Ortega, Andoni Barreña, Itziar Idiazabal, Patxi Juaristi, Carme Junyent, Belen Uranga & Estibaliz Amorrortu. 2005. Words and Worlds: World Languages Review. Clevedon: Multilingual Matters. See also www.amarauna-languages.com Maurais, Jacques & Michael Morris, eds. 2003. Languages in a Globalising World. Cambridge: Cambridge University Press. Moreno Cabrera, Juan Carlos. 2000. La dignidad e igualdad de las lenguas: Crítica de la discriminación lingüística. Madrid: Alianza Editorial. Moskal, Marta. 2004. “Language Policy and Protection of Endangered Languages in Poland”. Endangered Languages and Linguistics Rights: On the Margins of Nations ed. by Joan A. Argenter & R. McKenna Brown, 93–98. Barcelona: FEL. Nanovfszky, György, ed. 2004. The Finno-Ugric World. Budapest: Teleki László Foundation. Nettle, Daniel. 1999. Linguistic Diversity. New York: Oxford University Press. Nettle, Daniel & Suzanne Romaine. 2000. Vanishing Voices. The Extinction of the World´s Languages. Oxford: Oxford University Press. Payton, Philip. 2004. Cornwall: A History. Fowey: Cornwall Editions. Phillipson, Robert. 2003. English-Only Europe? Challenging Language Policy. London: Routledge.

71

72

Patxi Juaristi, Timothy Reagan and Humphrey Tonkin

Price, Glanville. 2000. Encyclopedia of the Languages of Europe. Oxford: Blackwell. Reagan, Timothy. 2006. “Language Policy and Sign Languages”. An Introduction to Language Policy: Theory and Method ed. by Thomas Ricento, 329–345. Oxford: Blackwell. Salminen, Tapani. 2005. UNESCO Red Book on Endangered Languages: Europe. Paris: Unesco. See also http://www.helsinki.fi/~tasalmin/europe_index.html#biblio. Schur, Nathan. 1995. The Karaite Encyclopedia. Frankfurt: Peter Lang. Siguán, Miquel. 1995. La Europa de las lenguas. Barcelona: Edicions 62. Skutnabb-Kangas, Tove. 2000. Linguistic Genocide in Education – or Worldwide Diversity and Human Rights? Mahwah, NJ: Erlbaum. Spolsky, Bernard. 2004. Language Policy. Cambridge: Cambridge University Press. Wright, Sue. 2004. Language Policy and Language Planning: From Nationalism to Globalisation. London: Palgrave. Wurm, Stephen. 2001. Atlas of the World’s Languages in Danger of Disappearing. Paris & Canberra: UNESCO Publishing & Pacific Linguistics.

Web sources www.amarauna-languages.com www.ethnologue.com http://europa.eu.int/comm/education/policies/lang/languages/langmin/euromsaic/et2_ en.html#11 http://www.helsinki.fi/~tasalmin/europe_index.html#biblio http://www.culture.gouv.fr/culture/dglf/lang-reg/rapport_cerquiglini/langues-france. html#ancre276758 http://ec.europa.eu/education/policies/lang/languages/langmin/euromosaic/index_en.html

chapter 3

Principles of policy evaluation and their application to multilingualism in the European Union François Grin

This policy analysis-based treatment of language choices in the European Union departs from usual approaches that stress educational, legal or political aspects, and attempts instead to develop an evaluative framework resting on the standard breakdowns of policy analysis, particularly the distinction between efficiency and fairness. The paper focuses on the narrower question of ensuring communication at reasonable cost in European institutions, and discusses the advantages and drawbacks of seven different models, or “language regimes”. It is shown that depending on the relative importance given to different criteria such as communicative speed, organisational simplicity, or inclusiveness, any of these models can be considered “best”. This general result points to the need to clarify policy goals, to beware of seemingly “obvious” solutions, and to develop language policies that combine different language regimes.

The European Union arguably represents one of the most eloquent examples of the need to develop novel solutions to the problem of communication between social actors with a large number of different mother tongues. Whereas the wellknown Regulation No 1 adopted in 1958 and in force to this day provides for the equal treatment of the official languages of all Member States, actual practice has departed from this principle of “full multilingualism” in a number of ways, even before the enlargement that took place in May 2004. Restricted language regimes (according to which only a subset of national languages are used) have been adopted in various aspects of the formal operations of the Union (Assemblée Nationale 2003; Gazzola 2005), and the day-to-day operations of the European Commission clearly favour English and, to a decreasing degree, French. German turns up as a very distant third.

74

François Grin

Beyond formal reassurances and ritual endorsements of plurilingualism, the choice of a language regime – or regimes – for the European Union is increasingly recognised as a highly sensitive issue. However, most of the contributions on this matter rest on legal considerations (with a reference to rights, sometimes more specific “language rights” or “linguistic human rights”), or on socio-political analyses that highlight notions of identity, nationality and citizenship. This paper attempts to address multilingualism in the institutions of the European Union (EU) in a way that deliberately departs from these standard approaches. Instead, it looks at the choice of language regimes as a public policy issue. This requires identifying and examining the allocative and distributive implications of different language regimes – or, putting it differently, to reason in terms of efficiency and fairness. I will try to establish the following two points: first, that language issues cannot be ignored and left to themselves, unless we are prepared to accept grossly inefficient and grievously unfair outcomes; second, that contrary to what is often claimed, there is no “obviously” superior solution, and that selecting a language policy for the European Union requires a much more systematic assessment of competing scenarios than what is often proposed. The paper is organised as follows. In the next section, I lay out the basic policy-analysis aspects of the problem. Following this, I present seven different “language regimes” and compare them in terms of resource allocation (or efficiency), while a separate section is devoted to the comparison between them in terms of resource distribution (or fairness). The main results are summarised in a brief conclusion.

The public policy perspective If choices are to be made regarding language policy options, the most fundamental instrument available remains policy analysis. Whereas it is commonplace to apply policy analysis tools to matters regarding transport, health, or the environment, its application to language policy is a more recent development (Pool 1991a, 1996). However, policy analysis is also useful in language policy because it can be located at a higher degree of generality than the approaches most frequently brought to bear on language policy issues (Grin 2000, 2003a). The latter tend to rely on three clusters of disciplines. The first cluster comprises applied linguistics and the education sciences, where contributions usually focus on (foreign) language teaching and language use in multilingual settings. Useful as such analyses may be, they emphasise the question “how” (namely, how to teach languages, how languages are learned, in what ways language learning processes may be related to other social processes, or how actors choose to use



Chapter 3.   Principles of policy evaluation and their application

one language or another), but this does not tell us what languages to learn, or for what reasons. The second cluster is anchored in legal perspectives. Again, these are necessary, but they usually take a number of parameters for granted – for example, certain standards embodied in international legal instruments. This approach leads to an examination of the compatibility of specific (e.g. national) legal provisions with those international instruments, a question rather removed from the fundamental one, namely, what language regime a particular society should choose, and for what reasons it should do so. To some extent, we could say that legal questions arise downstream from the policy analysis ones, since legal discourse serves to give a formal shape to the norms adopted as part of the selection and design of public policies. The third cluster sits astride the disciplines of international relations and political science. In so far as it incorporates the field of normative political theory, it is certainly closer to the core concerns of policy analysis. However, many contributions hailing from this cluster, rather than providing a general treatment of the issue, branch off into a discussion of the role of concepts like identity, nations or nationalism. There again, these are undoubtedly relevant ingredients in the examination, but they do not constitute the broad type of framework required to lay out social choices. Policy analysis includes two main questions, namely, resource allocation and resource distribution. When applying the former to language policy issues, the analyst tries to assess whether different language policy choices represent an efficient use of scarce resources. This requires assessing the net value (that is, “benefits” minus “costs”) of each policy option. It is important to note that both benefits and costs can be of the “market” and “non-market” kind (which therefore include symbolic benefits and costs), and need to be evaluated at the individual and social level.1 In general, the policy that generates the highest net value should be preferred to others. The examination of resource distribution – and redistribution following the implementation of a particular policy – aims at a systematic identification of the winners and losers, and of the amounts gained or lost by both groups. One particularly delicate point, of course, is the choice of criteria for establishing categories of people who may be winners or losers; in our context, people’s mother tongue is a key criterion.2 The question, then, is whether these gains and losses (which often take the form of undesired transfers from losers to winners) are seen as fair, or if they should be corrected by compensating transfers.3 Clearly, the type of examination proposed by a policy analysis framework can accommodate issues of identity, national sentiment and the like as investigated by contributions from “cluster 3”. Through the precise estimation of the effectiveness

75

76

François Grin

of different scenarios, it encompasses the language education and language use issues which “cluster 1” research usually emphasises; and its results can be used in the selection and design of policies that can subsequently be codified in legal discourse, using the apparatus provided by “cluster 2” (Grin 2003c). This is not to say, however, that policy analysis is enough to make appropriate public policy choices. Policy analysis is only an instrument for political debate and governance. It does not dictate solutions, but helps to expand citizens’ and decision-makers’ knowledge base, in order to allow for more informed choices.

Choosing official and working languages for the European Union The debate over the choice of official and working languages (hereafter: OWLs) in the European Union tends to side-step the policy analysis considerations outlined in the preceding section, and to rely instead on (in my view), incomplete characterisations of the problem in terms of the three “clusters” of disciplines mentioned above. The resulting shortcomings, which can also be observed in official EU literature (see, for example, European Commission 2003), and the attendant political questions this raises, have been discussed by numerous authors (see, for example, Ives 2004) and will not detain us here. Suffice it to say that issues of power are typically played down in the debate, and that a notable ambiguity appears to surround invocations of the principle of subsidiarity. In order to investigate language regimes, I start out from a model initially developed by Pool (1991b), which focuses on a limited range of “benefits” and “costs”. The variables considered are, for each regime, the number of OWLs, the nature of OWLs, the number of directions of translation and interpretation required, and the corresponding foreign language learning needs by EU officers, civil servants and Members of Parliament. Seven regimes are discussed in this paper, although additional regimes could certainly be identified. For example, “intercomprehension” (that is, banking on people’s receptive competence in languages structurally related to their mother tongue, for instance within the family of Romance or Germanic languages) considerably widens the range of possible language regimes. In this paper, we shall, however, confine ourselves to the six regimes investigated by Pool, plus a seventh regime introduced by Gazzola (2005). These seven regimes are summarised in Table 1, where the following assumptions have been made: 1) English is the only official language or the pivot language if the “monarchic” or “hegemonic” system is chosen;



Chapter 3.   Principles of policy evaluation and their application

2) this applies to Esperanto if the “synarchic” or “technocratic” regime is chosen; 3) if the “oligarchic” regime is chosen, a three-language troika is picked, comprising English, French and German; 4) in a regime of “triple symmetrical relay”, three languages are chosen as pivot languages, but no further assumption needs to be made regarding the languages concerned. Therefore, the seven regimes can be succinctly defined as follows: 1) the “monarchic” regime has only one official and working language, namely, English. This means that no translation or interpretation services are necessary, but that English must be learned as a foreign language by all non-Anglophones; 2) the “synarchic” regime has only one official and working language, in this case Esperanto (although Esperanto could be replaced by any language which is not native to the citizens of any Member State); there again, no translation or interpretation services are needed, but Esperanto must be learned by all; 3) in the “oligarchic” regime (or at least, in the version of this regime presented here, three languages (English, German and French) are treated as official and working languages; this keeps the number of directions of translation and interpretation that must be provided to six (that is, 3x(3–1)), but either English or German or French must be learned by persons whose mother tongue is other than any of these three; 4) the “panarchic” regime resembles current theoretical practice at the European Parliament: there are 23 official and working languages, which implies a staggering 506 (that is, 23x(23–1)) directions of translation and interpretation; however, there are in principle no foreign language learning needs; 5) in the “hegemonic” regime, all 23 languages are official; all written and oral communication gets translated (or interpreted) into the other 22 languages, but for 21 of them, translation and interpretation are indirect and take place via a “pivot” language, in this case English; the number of directions of translation and interpretation that must be guaranteed is therefore equal to 44 (that is, 2x(23–1)); however, no foreign language learning is necessary; 6) the “technocratic” regime is very similar to the hegemonic one, except that one official language such as Esperanto is added to the list of official and working languages, and it is used as the “pivot” language; therefore, 46 (that is, 2x(24–1)) directions of translation and interpretation must be covered, but there are no foreign language learning needs;

77

78

François Grin

7) finally, “triple symmetrical relay” also treats all 23 languages as official and working languages, and translation and interpretation take place through any of three pivot languages (instead of just one); it is simply assumed that for each of these languages, the amounts of translation and interpretation into and out of this language are equal; thus, the number of directions of translation and interpretation that must be guaranteed is equal to 126 (that is, 3(2x23–3–1)), but there again, no foreign language learning is needed. Table 1.  Seven language regimes4 Regime

Number of OWLs

Nature of OWLs

Translation and interpretation needs (directions)

Foreign language learning needs

Monarchic

1

English

0

English by all  non-Anglophones

Synarchic

1

Esperanto

0

Esperanto

6

English or German or French by others

Oligarchic

English, 1