Having chosen English as the preferred language in the European Union (EU), the European Parliament has commissioned a feasibility study in ways to improve efficiency in communications between government departments.

European officials have often pointed out that English spelling is unnecessarily difficult. What is clearly needed is a phased program of changes to iron out the anomalies. The program would be administered by a committee appointed by the participating nations.

In the first year, for example, the committee might suggest using "s" instead of the soft "c." Sertainly, sivil servants in all sities would reseive this news with joy. The hard "c" could then be replased by "k" sinse both letters are pronounsed alike. This would klear up a lot of konfusion in the minds of klerikal workers.

In the sekond year, bekause of growing enthusiasm, "ch" would be replaced by "c." This would make words like "switc" 20% shorter. Similarly, it will be announsed that the troublesome "ph" would henseforth be replaced by "f."

In the third year, publik akseptanse of the new spelling kan be expekted to reac the stage where more komplikated canges are possible. Governments would enkourage the removal of double letters whic have always ben a deterent to akurate speling.

We would al agre that the horible mes of silent "e" 's in the languag is disgrasful. Therefor, we kould drop thes and kontinu to read and writ as though nothing had hapend.

By this tim it would be four years sins the skem began, and peopl would be reseptiv to steps suc as replasing "th" by "z." Perhaps zen ze funktion of "w" kould be taken on by "v," vitc is, after al, half a "w."

Finaly, ze unesesary "o" kuld be dropd from words kontaining "ou." Similar arguments vud of kors be aplid to ozer kombinations of leters.

Kontinuing zis proses yer aftr yer, ve vud eventuli hav a reli sensibl riten stil. Aftr tventi yers zer vud be no mor trubls or difikultis and evrion vud find it ezi tu understan esh ozer. Ze drems of ze EU vud finali kom tru. (Anonymous joke posted on the Internet)

Through the lens of language, one can look at the process of European state formation from opposite ends. From one perspective, the tenacity of state languages embedded in bureaucracies, educational systems, and the patterns of everyday life suggests that a new European state will never become a reality. From a second perspective, however, the dual emergence of regional languages within current states as official organs of business and of English as a de facto lingua franca suggests that the new European state is following a general pattern for the consolidation of states characteristic of the twentieth century. Looking backward through the lens of history, it seems apparent that Europe of the twenty-first century will not replicate Spain of the eighteenth century. But looking forward to the dynamics of twentieth-century state building, European state consolidation has characteristics similar to what is emerging in India.

The principal argument of this article is that once we recognize that European state consolidation is a late twentieth-century phenomenon, we will understand why European cultural identities will remain multiple like they are in India rather than homogenized as became the case in France. This article also identifies two mechanisms by which a multicultural state can form. First, it looks at the rhetorical use of "Europe" by regional actors as a tool to fortify their claims for autonomy from central states. A consequence of this strategy could well be to normalize multicultural identities for which "Europeanness" is one element. Second, the article analyzes an emergent political coalition-Eurocrat centralizers aligned with regionalists-that has the potential of furthering a multicultural model of state consolidation. The result of these two mechanisms in the context of late twentiethcentury state building will be an institutionalized "constellation" of languages that will be a unique European constellation, reflecting a unique European identity.1 The languages in the constellation will not all be unique to Europe, but the specific constellation of them will be the core of the European linguistic identity.

Language, to be sure, is but one cultural component of national identities. Religion, family structure, dress, the arts, the media, and sport form other components, helping to constitute an "imagined community" of people who have mostly never met.2 Yet, as Tocqueville observed, "The tie of language is perhaps the strongest and most durable that can unite mankind." He felt that religious diversity would strengthen America, yet linguistic sharing was the key to a common identity.3 Many theorists of the nation-state, from the nineteenth-century romantics, to Soviet Communists exemplified by Stalin, to twentieth-century modernization theorists such as Deutsch and Gellner, focus primarily on language as the core cultural component of national identities.4 The fact of great linguistic diversity in Europe often has been cited as the highest barrier preventing the emergence of a genuine European nation. My argument here is that in twentiethcentury state building, a shared language constellation, rather than a particular language, is the marker of a successful national project. Europe, I shall show, is on that road toward a coherent language constellation.

CONTRADICTORY TRENDS IN CONTEMPORARY EUROPE

Among the observers of European integration, there has long been a divide between the tortoises and the hares. While some saw a European state emerging out of elite bureaucratic processes, others focused their attention on the slow processes in which ordinary people change their national identities. Sophisticated syntheses modeled the contradictory processes in a more dialectical form. Once a dialectical process is accepted, the force of nationalism becomes ambiguous. On the one hand, nationalism brakes the wave toward a new Europe; on the other hand, the idea of Europe pushes the wave (often those same people) toward new shores.5

Since the "fall" (in both senses) of 1989, the enigmatic power of national and cultural identity has impressed itself even further on the sociological imagination. On the one hand, the national idea crushed the capitalist prudentialism of the Bundesbank, which tried to slow down the process of German reunification; it set Czech and Slovak intellectuals, formally united in anti-communism, against each other; and it unleashed independence movements in Slovenia, Georgia, Armenia, the Baltics, Uzbekistan, and elsewhere in a collapsing Soviet "Union." Eric Hobsbawm's stunning exegesis, written just before the "fall," confidently asserts that the national quest is, by the late twentieth century, a spent historical force. Brilliant he was-but wrong. The national idea seems to be as powerful and as inexplicable today as it was to Lenin at the outbreak of the Great War in 1914.6

On the other hand, the movement for a united Europe proceeds apace. Crucial elements of cultural distinction must give way to continental standards. Germans might well have to live with speed limits on their autobahns; Spaniards feel pressured to work with only a single hour for lunch. And the role of national languages in Western European education is becoming somewhat more restricted. In a study of eight leading business schools in Europe by L'Expansion, as reported by The New York Times, all relied on English as a medium of instruction, and in four of them English was the sole medium of instruction.7 The brouhaha in France over the issue of English as the medium of scientific exchange in the journal of the Louis Pasteur Institute seems arcane today. Even in the sensitive area of movies, French "Cesars" (the U.S. Oscar equivalents) in all categories (save "best film") can go to films produced in any language so long as there is significant French involvement.8 National culture, nearly all elites in France now agree, must give way to the realities of globalization.

Nationalism versus globalization-these counterpressures are seen most starkly over issues concerning language. The statuses of English in Western Europe, of Russian in the former Soviet Union, and of English and French in the states that received independence since 1945 all raise sensitive political issues. On the one hand, globalization makes these languages tools for international communication. People do not want to be left behind on the train of history, and they will equip themselves with language repertoires that meet current needs. On the other hand, the pressures for national identity are most keenly felt in the domain of language. People want to keep their mother tongues alive, even if those languages are left behind in the world of technology and interdependence. Will globalization in language outpace the national idea? If so, how is it that the national idea can mobilize people to risk their lives and property for independence yet allow people to give up perhaps the most visible aspect of their national identities to the demands of globalization? Observers of nationalism are indeed both awed by its power and dumbfounded by its weakness. These contradictory trends must be kept in mind when one assesses the future of the European project. Highlighting the overlapping language repertoires constituting a continental constellation, as I do in this article, gives due weight to both the power and the weakness of nationalism.

Looking Backward: Language and Early State Construction

Rationalization in general, and in language as well, has been a characteristic process in the construction of the modern state. Max Weber used the term rationalization to refer to the process of efficient and orderly rule.9 The development of a professional civil service with a well-specified division of labor was for Weber the essence of rationalization in the modern state. The establishment of sharp territorial boundaries, the standardization of the calendar and of weights and measures, and the issuance of a common currency are other important examples of state rationalization.

Weber did not systematically explore language rationalization.10 Yet, the use of state power, through administrative regulation and public education, to standardize language within the boundaries of the state is precisely what he had in mind with his concept of rationalization. Legal uniformity is easier to ensure when court decisions are delivered and recorded in a common language. Taxes can be collected more efficiently and monitored more effectively if merchants all keep their books in the same language. State regulations can be disseminated more efficiently if translations are not necessary for compliance to take place. And territorial boundaries are easier to patrol if the population at the boundary speaks the language of its political center, one that is distinct from the language of the population on the other side of the boundary. Given these considerations, it is not surprising that rulers of states have sought to transform their multilingual societies into nation-states through policies that can be called "language rationalization."

Language rationalization policies usually entail the specification of a domain of language use (e.g., appeals court cases, church sermons) and a requirement that the language chosen by the ruler be employed within that domain. When rulers have established power over a territorially distinct speech community, they are easily able to induce some of its members to become bilingual so as to translate documents from the language of the speech community to the language of the ruler. To the extent that political rule is stable, more and more members of the newly incorporated speech community will find it useful to learn the language of the ruling elites. Language rationalization is successful when there is a sufficient number of bilinguals among linguistically distinct communities so that the business of rule can be transacted in a single language.

Language rationalization strategies were employed so successfully in France, Japan, and Spain that these countries began to be considered in the nineteenth century (not by all people, and certainly not by people who considered themselves part of minority language groups) as "natural" nations. The image of a rationalized nation-state was a powerful one, enough to impel political movements in Italy, Germany, Serbia, and Tamil Nadu seeking to construct new states based on linguistic boundaries.

To give a sense of the process of language rationalization, consider the case of Spain. Spain was a multilingual empire when the Catholic monarchs, Ferdinand and Isabella, presided over the final reconquest of the peninsula from Muslim rule. Castilian, Catalan, Basque, and Galician were the major languages of Spain. The Habsburg kings, following the treaties of Ferdinand and Isabella, respected regional differences in language and in law. Spain's wealth from overseas conquest, however, attracted artists and writers from all over Europe, and Castilian became a language of prestige throughout the peninsula. The literary fluorescence of the Golden Century (mid-sixteenth through mid-seventeenth century) induced well-to-do families throughout the kingdom to educate their children in Castilian.

It was not until 1716 and the Decree of the Nueva Planta, under Spain's first Bourbon king, Philip V, that Castilian became Spain's language for official business. A series of decrees in 1768-1771 required all primary and secondary education to be in Castilian, and in 1772 all commercial establishments were required to keep their accounts in Castilian. Despite these laws, multilingualism persisted in Spain, although virtually all citizens became fluent Castilian speakers and all educated Spaniards became literate in Castilian. Spain's status as a nation-state was often questioned by Basque and Catalan nationalists, but the Spanish state through the regime of Francisco Franco (1936-1975) relied exclusively on Spanish to conduct its business.

The construction of the modern state, then, has long been considered synonymous with the construction of a nation-state, a political unit in which there would be greater cultural homogeneity within its boundaries than across them.11 To be sure, the historical record provides anomalies such as Switzerland, where societal multilingualism coincides with strong feelings of national solidarity. What is interesting about the Swiss case, however, is the rigidity of cantonal language rationalization that has brought greater levels of language homogenization within cantons than within most nation-states. Despite historical anomalies and the farsighted analysis of the Austro-Marxists who indeed saw the multilingual state as normal, the idea that real states have unique national languages became the conventional wisdom by the turn of the twentieth century for people as diverse as Lenin and Woodrow Wilson.

Looking Forward: Language and the Postcolonial State

State construction for many countries that achieved independence from colonial rule in the twentieth century has not followed the pattern of early developers. To explain this difference, analysts of postcolonial state construction at first misread the historical evidence and equated the naturalization of language homogeneity in the modern European state with the supposed naturalness of European boundaries. In this regard, colonial boundaries were considered deficient or arbitrary. Tanzania's founding President Julius Nyerere once noted that Africa's boundaries are so absurd that political prudence required sanctifying them. His ironic observation implicitly assumed that boundaries in Europe were natural.

Once the historical record is clarified, one might surmise that the postcolonial states may well remain linguistically heterogeneous for centuries (as did their European counterparts in similar stages of state construction) but that eventually unique national cultures (with a single official language) would be fashioned. This surmise is based on the assumption that state construction follows standard patterns no matter what the world historical time in which the process began.12

State development, however, needs to be historicized. Consider India. India is a postcolonial multilingual state that is not moving toward rationalization or linguistic uniformity. During its independence struggle, the Congress Party leaders assumed that on achieving independence, some form of Hindi should be India's official language. But the Indian Administrative Service and other bureaucratic agencies operated entirely in English, the language of colonial rule. Officials in these agencies had a strong interest in preserving English as the language of administration. Although the Indian constitution called for the imposition of Hindi as the All-Union language fifteen years after independence, opposition from bureaucrats and from citizens in non-Hindi-speaking areas indefinitely delayed the change. Consequently, there are now two All-Union languages, each of which can be used for official dealings within the Indian state.

Meanwhile, during the 1950s, citizen pressure at the regional levels compelled the government to redraw federal boundaries consistent with language zones. Each state has an official language today, and the governments of most of them are zealously pursuing language rationalization policies within their states. By one official measure, as of 1980, only 2.7 percent of the Indian population has as its primary language a language different from the official language of its state.13 The Indian constitution, however, assures all minorities the right to an education in their own language.

India thus has a multilingual state in which citizens who wish to have a broad range of mobility opportunities must learn 3 I languages. English and Hindi are necessary languages for communicating with the central state. It is necessary as well to speak the language of the state in which you live. This makes for a 3-language outcome. Acitizen of Tamil Nadu must learn English, Hindi, and Tamil to be able to operate in a wide range of activities within India. Those citizens who live in states where Hindi or English is the official language need learn only 3 1 or 2 languages. Citizens who are language minorities in some states (e.g., Marathi speakers in Karnataka) must learn 3 + 1 or 4 languages: Hindi, English, Marathi, and Kannada (the state language of Karnataka). Thus, there is a range of from 2 to 4 (i.e., 3 +/- 1) languages that citizens must know. The resulting language constellation, unique to India, has become accepted by most groups in India, and therefore it is probable that India will remain a multilingual state.

What explains the inexorable move toward rationalization in early statebuilding experiences and the stable 3 +/- 1 outcome in India? A first difference has to do with changes in the nature of the business of rule. By the twentieth century, virtually all states engaged in activities in which the language used has noticeable effects on the general population. As states were in competition with one another for centuries, successful innovations in one state became a point of reference for others. Those states that provided compulsory education to the young, drafted "citizens" into a national army, and employed large numbers of literates in a rationalized bureaucracy became very powerful and were consequently attractive models for less prominent states. For the initial cases of state consolidation, the expansion in functions occurred after state rationalization of language had been successfully completed. In France, for example, there was sufficient knowledge of French (but it was not widespread) in the mid-nineteenth century so that a teaching corps and an officer corps could run a school system and army in French.14

With twentieth-century state building, rulers have felt it necessary for their states to perform all evolutionarily acquired state functions. This phenomenon can be partly explained by the competitive model of state functions and by the "modular" (i.e., easy to copy in outline) aspect of nationalist ideology.15 The ideology of necessary state function comes out clearly in the words of the Kher Commission examining the question of a national language for India. "Modern governments," it reasoned, "concern themselves so intimately and so extensively with all aspects of social and even individual existence that inevitably in a modern community the question of the linguistic medium becomes an important matter of concern to the country's governmental organization. In the conduct of legislative bodies, in the day-to-day dealings with citizens by administrative agencies, in the dispensation of justice, in the system of education, in industry, trade, and commerce; practically in all fields in which it has to interest itself in modern times, the state encounters and has to tackle the problem of the linguistic medium.16 During the 1950s, no serious politician in India suggested that India, as a new country, should perform only those functions performed by European states in their early periods of rationalization.

This historical change implies that rulers of the new states have needed to be far more sensitive to the linguistic repertoires of their citizens than were rulers that consolidated states in earlier centuries. Imposing a specific language as the sole language for rule on a population that does not speak it will more likely mobilize the population when the state is already providing public education and local health services than when it was not involved in such activities. The data from India show that people from non-Hindi-speaking zones are often quite willing to learn Hindi (e.g., to watch movies and TV docudramas of Hindu sagas) but become rebellious if the use of Hindi is enforced in schools or is used (at the expense of local languages) in state hospitals. Standard policies of language rationalization similar to the Nueva Planta in Spain cannot be legislated in India at low political cost.

A second historical factor that distinguishes the Indian case from the European ones concerns the effect of modern colonialism on political-bureaucratic relations. In the postcolonial state, there is a conflict of linguistic interest between national politicians and senior bureaucrats, one in which the latter group has a strategic advantage. State builders of early modern Europe had an administrative service loyal to them. Weber, in his classic study of bureaucracy, notes that modern states are distinctive in that they employ officials who earn a salary that is paid irrespective of their loyalty to the ruler. The burden of contemporary state builders is that they were handed modern bureaucracies in order to accomplish tasks best performed by loyal knights and retainers.

Civil servants in postcolonial states regularly deviated from Weberian bureaucratic norms.17 But while these bureaucrats quickly gave in to practices that are corrupt and therefore subverted the norms of a neutral civil service, they never abjured the perquisites of office (e.g., regular salary payments, health benefits) enjoyed by their European predecessors. An expensive and entrenched bureaucracy with high status presented a challenge to political leadership in the new states that earlier monarchs never had to face. This problem also applies to the issue of language rationalization. The party elites who fought for independence had different interests from the administrative elite that remained on salary during the period of transition from colonialism to independence.18 The bureaucrats had a vested interest in the perpetuation of the colonial language as the official language of state while the politicians had a mixed interest. They wanted to give official status to indigenous languages, for this would draw them closer to the people. But they were also interested in getting compliance and support from the bureaucracy, which would be charged with carrying out their social and economic programs. Because the politicians were willing to relinquish their goal of a unique national language in order to get compliance from the bureaucracy for the administration of an economic development program, the language rationalization project was abandoned.

The 3 +/- 1 language outcome in India was therefore the unintended outcome of an intense political process that involved Congress politicians, bureaucrats, state-level politicians, and the mobilized public. Central politicians were willing to live with English as the de facto language of central rule just so long as Hindi were sufficiently promoted (as a legal language of interstate communication and as a required language for secondary school graduation) to ensure its hegemony among indigenous languages. Central bureaucrats were occasionally bothered by dealing with Hindi language requirements, but they were clear winners in ensuring the dominance of English in their domain. State-level politicians were able to ensure their autonomy by running their state business in state languages. Appointments to government positions within the states were thereby practically reserved for sons of the soil. Finally, citizens could hardly mobilize against the 3 +/- 1 outcome. Practically no citizen was denied the right to education in mother tongue or mobility opportunities in having access to the state language, Hindi or English.

To support the claim that the 3 +/- 1 outcome is stable, one might examine the 1977 election of the Janata government, which was brought to power in the wake of Prime Minister Indira Gandhi's ill-fated state of emergency in which civil liberties were suspended. In the postemergency government, 221 out of 299 elected representatives of Janata came from Hindi-speaking areas. This represented a government clearly committed to a Hindi vision of India. Once in government, however, the Janata immediately stood behind Nehru's language concessions.19 English would remain the language of elite domains; the non-Hindi states would not face Hindi imposition. In light of this experience, the 3 +/- 1 language policy appears now to be in equilibrium in the sense that the costs for change, for any party, outweigh the benefits of the status quo.20

Part of what is meant by a multiple language equilibrium is an institutionalized set of norms governing which language is to be used for which social domains. Thus, the Indian multilingual equilibrium does not mean that all languages can be used for all purposes. Rather, the domains for languages are functionally specific, and clear social norms give guidance as to which language is proper for which type of interaction. Nor does the multilingual equilibrium imply that the languages are on a strict hierarchy, with English for "high" functions and state languages for "low" ones, a situation closely related to the phenomenon known as "diglossia."21 Jobs and other opportunities are associated with fluency in Hindi and the state languages, so they will not be marginalized to live only in the world of nursing mothers. The languages are not of equal prestige, to be sure, but Hindi and the state languages continue to have considerable elan in a range of important social, political, and economic domains. The number of languages, then, and their functionally specific domains of use help constitute India's 3 +/- 1 equilibrium.

The dynamic leading to a 3 +/- 1 outcome can be portrayed in a simple model, one that applies to other postcolonial states.22 Consider a game in which the ruler of the central state can choose as official either the colonial language or the dominant indigenous lingua franca. Should the ruler choose the latter, the bureaucrats (who got their jobs based on facility in the colonial language) get to choose whether to accept the new language regime or to subvert it (by continuing to operate in the language of colonial rule). Once the bureaucrats choose, regional leaders (representing distinct language groups) can decide whether to promote the local vernaculars in a serious way or to promote them as merely a symbol of local identity.

If nationalist fervor pushes the ruling political elites to choose a national language as the official one for state business, the bureaucrats will reject this solution and defy the center. With a de facto situation of a divided center, regional elites are not under great pressure to conform and have some leeway in promoting local vernaculars for education and for jobs in the local administration, which would give their groups (as opposed to migrant communities) virtual monopoly control over a set of state resources. In reaction to this cacaphony, the center can neither punish the bureaucrats (whom the center needs for basic administration) nor punish the local leaders (who could defy the center through an alliance with the central bureaucrats). All the center can do is make an alliance with minorities in the states, assuring them that their languages will not be ignored in the school systems where they are minorities. This model demonstrates the logic of the 3 +/- 1 outcome in India, and it has implications elsewhere in the postcolonial world.

Indeed, many other postcolonial countries, including Nigeria, Zaire, Kenya, and the Philippines, may follow India's track, with a colonial and an indigenous language sharing central stage in the business of rule. In others, such as Algeria and Zimbabwe, the colonial language will remain as a lingua franca, but not at the expense of the continued official reliance on indigenous languages, at least on the regional level. To be sure, some postcolonial states, such as Indonesia, Cambodia, Malaysia, Tanzania, and Somalia (despite the civil war), are on the road toward rationalization. But given the changed international environment of state building in the twentieth century, the "normal" path seems to be toward some form of institutionalized multilingualism.

MULTIPLE IDENTITIES AND THE NEW EUROPEAN STATE OR THE EMERGENCE OF A 2 +/- 1 STANDARD LANGUAGE REPERTOIRE?

Europe, to the extent that it is consolidating as a state form, will more likely follow twentieth-century patterns than patterns consistent with sixteenth- to eighteenth-century state building. To expect otherwise would be to dehistoricize state-building processes. To be sure, the analogy with India is not perfect. Europe does not have a popularly based national movement supporting a nation-building strategy. Nor can Europe's states be usefully compared with India's, as the latter states did not experience centuries of sovereign rule. Yet, there are some striking trends on the road toward European integration, quite similar to those in Indian state building, with similar implications for language repertoires.

The initial trend worth noting is the true Europeanization of English as the lingua franca of the new European state. The second trend is the tenacity of the state languages in the educational systems of the present member states. The final trend is the successful incursion of regional languages into regional educational systems and bureaucracies. These trends have parallels to what has taken place in India, and they provide a basis for predicting the language repertoires of future European citizens and a unique language constellation that will thereby emerge.

English as Europe's Lingua Franca

Almost any accounting procedure will show that English has come to dominate French and German, and indeed all languages of Europe, as the language of internationality communication. "Eurobarometer" surveys show that 51 percent of European Community (EC) citizens report knowing English, dominating French (42 percent) and German (33 percent).23 In a quantitative study of the European language constellation, De Swaan distinguishes between the payoffs for learning a language that would enable either direct communication (i.e., with native speakers of that language) or indirect communication (i.e., with people who speak it as an auxiliary language). The power of English, he found, is in the latter category. His conclusion:

[In the constellation after the United Kingdom's entry into the EC,] English now was the most attractive single language for outsiders, and what is more, the speakers of all other languages in the constellation preferred English as an additional language over any other. Since any speaker could therefore expect all other language learners in the constellation to prefer English, they had every incentive also to choose English. Only the English had an incentive to learn German or French as their first foreign language, but they might refrain from learning any languages at all, confidently waiting until the rest of the constellation had joined them.24

With the collapse of the Berlin Wall and the possible inclusion of many East Europeans into the EC language constellation, the number of native German speakers is greater than that of native English speakers. However, since so many of these new entrants into the constellation speak English as a second language, English remains in first place. In a study supported by the Commission of the European Communities (European Commission), it was found that 83 percent of secondary school pupils in the European Union (EU) were learning English as a foreign language, compared with 32 percent learning French and just 16 percent learning German.25 Even in a country such as Estonia, which had a German ruling class for nearly a millennium and continued business interchanges, independence from Soviet control has induced a cascade toward English as the language of international communication. In a 1993 survey of 1,454 Estonian respondents to the question of which language is most important for foreign business contacts, 90 percent said English and only 7.9 percent said German.26 Even on the German side of the European periphery, then, English has become the dominant second language. Perhaps more important, the new relative prominence of German in Eastern Europe seriously weakens the role of French in the wider European constellation.

Perhaps the best indicator of the tide toward English is the fact that people throughout Europe (and the world) are willing to pay privately for its acquisition. "English as a Foreign Language" is a global industry, and the United Kingdom earns about 6 billion pounds annually on language exports. Meanwhile, according to Coulmas, Germany spends about 450 million deutsche marks to promote German culture overseas, with some significant part for language, and France expends as much as 1.5 percent of its gross national product on the defense of French.27 People are willing to pay high personal costs to learn English; they have to be bribed to learn French or German. The microeconomic handwriting is on the wall.

French, to be sure, still plays an important role in EC affairs. It was the sole official language of the Coal and Steel Community, the precursor of the EC. In Brussels, Strassbourg, and Luxembourg, three core Eurocratic centers, French is a dominant language. And French remains the principal language of a significant percentage of Eurocrats to this day. There is no doubt that French political and bureaucratic pressure can stem the tide of English for a considerable time, but the failure of French to prosper on its own terms in the current constellation of languages cannot be denied.

Germany is a more interesting case in that its currency is unbeatable while its language is marginal. Despite the high relative number of native German speakers in the EC, German has an official role in only 8 countries in the world (as opposed to 63 for English), and because of the issue of guilt in World War II, German plays no role in the United Nations. Even in Germany, scientists publish more original research results in English than in German. (Springer-Verlag, a leading scientific publisher, has 80 percent of its list in English.)28 Germany would, more so than France, be able to challenge English's hegemonic language role, but it has much less of an incentive to do so inasmuch as most elite Germans already operate in English in international forums.

Because the issue of language is so sensitive, the European Commission has tried to avoid even recognizing the trends. In 1988, it urged in a policy statement that "member states should be encouraged to ensure that all official [European] Community languages are on offer within their educational system, even if there is an increasing trend towards certain languages."29 It has even been suggested that perhaps some "antitrust" legislation should be enacted against English.30 In 1996, the unveiling of the future Eurocurrency had no recognizable words in any known language. But the logic of the market is clear: English has become the virtual lingua franca of the EC. For political reasons, however, this social fact has not been officially acknowledged.

State Languages Maintain Their Functions

Whatever the trends toward a lingua franca, there is no evidence of any trend away from the state languages within state boundaries.31 In all EC countries. children learn their national languages first and go to school where these languages remain the principal media of instruction. Newspapers, TV, leisure-time reading, advertising, and official local and state services all continue to operate in the state languages. The rise of English in Europe has not displaced the state languages on any of the avenues of daily life.

In fact, the EC, composed of member "states," has been assiduous in protecting state languages. A well-funded program called LINGUA promotes state languages in other countries so that a French parent working in, say, Italy will be able to get a French-medium education for her children.32 More politically, the European Commission has not questioned the member states' prerogative in subsidizing publications in the state languages. Such activities would normally be seen as an unfair trade practice, but if a language (as opposed to a wine) is to be protected, the EC has been reluctant to intervene.33 To be sure, the European Court of Justice declared against a Netherlandic law that prohibited Dutch-language advertisements on foreign channels. This was a strange law anyway, and it was meant to protect Dutch products, not the Dutch language. But the EC has permitted governments to require a certain percentage of broadcasts within a country to be in a specified language, and this restraint of trade is overlooked due to the sensitive matter that is at stake.34 Whatever the rise of English, then, the state languages have a firm social footing in the realities of everyday life in Europe and in the corridors of state power. There may well be a "hollowing out" of the nation-state, as some post-Fordist commentators have argued, but it is not happening in the realm of language.35

New Institutional Support for Regional Languages

The rise of one language (English) has not even meant the fall of regional languages. In fact, the European language constellation in the past generation has been more favorable to regional languages (minority languages within established states) than have the states themselves. In 1981, the European Parliament adopted the Arfe Resolution, which called for a charter for regional languages. As a result, a European Bureau for Lesser Used Languages was established in Dublin. By 1983, policies protecting language minorities became a recurrent budget item in the European budget.36

Regional language groups, politically dead for centuries, have become mobilized and mobilizable. In Italy, significant Albanian, Catalan, Provencal, Friulan, German, Slovene, Occitan, and Sardinian minority language groups have been identified, with the German and Slovene communities already politically mobilized.37 In France, German, Occitan, Breton, Catalan, Flemish, and Corsican all have more than 100,000 speakers. In Germany, there is a large and growing Polish-speaking community. In the United Kingdom, Irish, Scottish Gaelic, and Welsh are the principal regional languages. Their official recognition is dubious. But a commission led by Sir John Banham, and supported by both parties, is to report in 1997 on a regional government plan for the United Kingdom that will deal with Wales, Scotland, and Northern Ireland, with an expectation that the treaties incorporating these territories into a United Kingdom will be substantially revised.38 In Belgium, with two official languages, there is also a moderately sized German language community, which, since the accords of Saint-Michel of 1992, has an elected German Assembly.39 The Belgian regions have their own delegated official within the official Belgian permanent representative to the EU. Denmark, too, has a significant German-speaking community. And Holland's Frisian-speaking community has already gotten political support from the EC, to the chagrin of Dutch authorities.40 The recognition of these language communities by the EC, like the korenizatsa policies of the early Soviet state, gives a legitimacy and a political agenda for mobilized elites from these groups to further press new language demands on the European political stage. The semi-official recognition of Catalan as a community language is the first elevation of a regional language into a community-wide function.41 While it is true that state languages get stronger and more regular subsidies than the regional languages,42 and while it is also true that member states have the right to define which languages spoken within their borders are official minority languages,43 the footing and resources of these regional languages have been strengthened by EC intervention. Indeed, the constraints against the development of the regional languages in most of Europe seem high. Simons reports on the oc revival movement in southern France. Despite the missionary zeal in which these languages are promoted, she also notes the habits and practices of using French that go back for generations and also the minimal resources available to the revivalists. The effort appears quixotic.44 One therefore gets the impression that regional languages may be more of a luxury consumption item for the few than a serious revival movement that will coercively demand language competence in the regional language for all who live within the regional boundaries. Local educational authorities might require a year or two of study of the regional language in primary schools. Urban professionals living outside the region may enroll their children in summer courses, give a contribution to the local language activists, or even take a course themselves to reconnect with their roots. However, these symbolic acts can have longer term consequences. Once funded, language activists, allied with politicians and businessmen seeking higher levels of autonomy from the central state, can set a coercive language regime far beyond the expectations of those who began to support it as a leisure activity.

The Role of "Europe" in Regional Nationalism

The concept of "Europe" will play an important role as a mechanism institutionalizing these trends inducing individuals to adopt stable multicultural identities. This point can be illustrated by examining 1980s rhetoric in Catalonia. Catalonia has for millennia been a cosmopolitan center for trade and culture, yet for the past century it has also been the center of a language and cultural revival movement that is apparently provincial. Catalonia's split image was ambiguously represented during the 1992 Olympic Games, where Barcelona's internationalist art, cuisine, and intellectual currents were juxtaposed to its political program of proclaiming Catalonia as a country separate from Spain.

In a content analysis of the Catalan language press in 1984-1985, Laitin and Rodriguez found that in editorial discourse the two leading Barcelona newspapers used the word "Europe" the most, and this followed by "Catalonia."45 These two powerful images reinforced each other in the overarching ideological theme that was evident in both the newspaper of the left and of the center right.

This discourse strategy serves an array of ideological purposes. First, Catalan nationalists do not wish to be accused of provincialism in their self-distancing from Madrid. They are often accused of this when they appeal to local language and regional customs. To be part of Europe is bigger, and more international, than being part of Spain. During the civil war (1936-1939), many Catalans found freedom and had encounters with "modernity" or with a "progressive world" by traveling back and forth between Catalonia and France; it was Franco, Catalanists claim, who was the provincial, who never traveled outside Spain in his four decades of rule. To counter the French epithet that Africa begins at the Pyrenees that was reinforced by Franco in his claim that "Spain is different," Catalans have sought to promote themselves as more international, more cosmopolitan than "Spaniards." Catalans, then, identify with Europe because they see themselves as more European than Spanish, and they have often talked about post-Franco democratization as part of a vocacio europea ("European commitment") that will allow the Generalitat (the Catalan administration) to "make explicit what is the reality [of the Spanish state] which the Constitution imposes." The reality of Spain as a compound of autonomies will be fulfilled, Avui (the center right Catalanlanguage newspaper) argues, due to the "European and progressive commitment which characterizes the Catalan people."46 Therefore, those who have framed Catalan discourse connect themselves to a wider political entity that is democratic and modern, in large part to prove that they themselves, in their promotion of a regional language, are not narrow, autocratic, or provincial.

The second ideological purpose that Europa serves in Catalanist ideology is analogic. Europe is a political category that is growing in authority yet is not a "state." Catalans want their government, which is also not a "state," to grow in authority as well. These Catalanist claims for political power by stating an explicit relationship between Catalonia and the supranational entity, l'Europa ampliada ("expanded Europe"), are expressed by Avui when its editors write, President Pujol [of the Catalan autonomous government] had clearly enunciated the philosophy which underlies his government's project. It is a philosophy that attempts to make the Europe of the States compatible with the Europe of the regions-or of the nations. This is because it is only realistic to recognize that at this time it is easier to build the Europe of the 12 than the Europe of the 110. At this point, the Generalitat should deepen its understanding of the president's political philosophy in order to make it more consistent, and even more important, to enrich it with specific propositions that could help to upgrade the role of the regions.

Europe, for Avui, is therefore a model of non-state authority in the modern world that has relevance for Catalonia's claims.47

Third, Europe is relevant within a Catalan nationalist discourse because to conceive of Europe as a compound of regions i nacions sense Estat ("regions and nations without a state") or to write about Europe as constituted by autonomias and pobles, as the Catalan-language newspapers do, implies that all states are multinational. This is why the editorial refers to Spain as the l'Estat espanyol. Avui here positions itself as an outsider to that state and uses this language to emphasize that states are far more internally diverse than their self-image of "nation-states" implies and that they must all confront the realitat plural ("pluralist reality") within their boundaries. We can see this argument in the following discussion:

[The construction of the EC's] Council of Regions . . . was . . . a first important step in the pursuit of new bonds among the diverse peoples of the European Community. . . a Europe which is much more diverse than the Europe of the 12 and which represents in a much more correct manner the various peculiarities of the European peoples. The recreation and the institutionalization of the Europe of the regions is not at all easy. The internal organizations of each one of the member states are sufficiently different and they don't have tight relations. On the other hand, that which we can call the historical inertia of the States weighs like a gravestone on the construction of this new way of understanding what Europe is all about.48

In a democratic EC, the Catalanist argument goes, the peoples of Europe (the Catalans, the Flemish, the Alsatians, the Corsicans, the Bavarians, the Scots) will not be deterred by the vagaries of ancient history. They will have an equal claim to representation in Europe as do the states.

Diari (the left-leaning Catalan-language journal under study) refers to Europe in a similar tone as does Avui. For the editors of Diari, Catalanist claims for regional autonomy are reinforced by the idea of a Europe that overcomes sociopolitical configurations along the lines of statehood. Diari relies on expressions such as Europa sense fronteres ("Europe without borders"), Comunitat Europea ("European Community"), Europa unida ("united Europe"), and Comunitat Economica Europea ("European economic community") to reify a notion of an already-existing European entity. In an editorial called "L' Europa del 92: Un punt de partida" ("Europe of 1992: A Point of Departure"), Diari only once refers to Europe in terms of states, but in a vague manner: l'Europa dels Sis ("the Europe of the six"). For the remainder of the editorial, we read about nationalities, but not about states. Consider this fragment:

In dislodging the economic issue, the Greeks on the one hand, and the Spaniards and French on the other, will have the task in the next 18 months of carrying out the Germans' policy [when their representative was the EC president] and to close the social and political dossiers which are necessary for the Act of Union.49

Lexically, Diari writes as if it is the pobles of Europe rather than their states who are engaged in the key negotiations. This is because Diari uses pluralized nouns to refer to the social agents (e.g., els grecs, els espanyols, els francesos) rather than their "nation-state" terms (e.g., Greece, Spain, France). Thus, Diari, like Avui, legitimates European nations at the expense of its states. In Catalanist political discourse, then, it is the mythical homogeneous nation-state that is defunct; the rise of a multinational Europe and a non-state Catalonia will simultaneously erode the legitimacy of contemporary state boundaries.

The model of European state building from a Catalanist perspective is not one that would erase cultural difference; rather, it is one in which putative state builders would be politically compelled to acknowledge and promote nationalcultural difference. While the rhetorical use of "Europe" has helped serve Catalan political interests, an important consequence of this rhetorical practice is to normalize the idea of "one region, many identities" in a way that the ideal of "one language, one state" pervaded European political discourse throughout the nineteenth century.

"One region, many identities" as an ideology is consistent with strategies to promote Catalan (as the regional language), Spanish (as the state language), and English (as the European language) as the standard repertoire for well-educated Catalans in the next century. Linz's data on the multiple identities of Catalans, who feel both Spanish and Catalan at the same time, lend support to this trend toward stable multilingualism.50 To my great surpise, however, while participating in a conference of a Catalan scientific organization in Girona in April 1994, I met a few graduate students who were in effect monolingual in Catalan. If these incidents are a signal of a broader trend toward a kind of regional tribalism, the thesis in this article will be undermined. The incentives nonetheless remain for Catalans to develop full facility in three languages.51

Is There a Stable 2 +/- 1 Language Outcome in Europe?

There is, then, a rhetorical/ideological basis for a multilingual European state. Is there a comparable political basis? Many Euro-ideologues have speculated on a multicultural theme in order to gain support for the European idea. Lepsius writes of a "Nationalities State" to elide the problem of what the European "nation" would look like.52 Du Granrut seeks to paint a "utopian" picture of a twenty-first-century state (a Europe of the Regions) in which people will remain closely aligned with their historical regions and be protected to do so by membership in Europe.53 But such Eurovisions, as more sober scholars like those represented in the Wildenmann working group point out, understate the continuing power of the member states.54

The prediction offered here requires no utopian vision. Many of the features that I have presented-like the consolidation of English as the lingua franca-are robust equilibria already forming. And for the prediction to be borne out, there is no requirement for a weakened nation-state. Yet, for the trends outlined here to occur requires a political jump-start for the regions.55 Is there any evidence for this? The answer here is only a weak affirmative. The major impetus for an EC language policy is budgetary. With each new official language, the cost of translation goes up geometrically, as translators are needed for every dyadic language communication. Already about 14 percent of the European Commission's ll,OOO staff work in translation and interpretation services, and this constitutes 40 percent of the administrative budget. The annual cost of language issues is approaching a billion ECUs per year.56

The principal administrative justification for these costs is that since each EC law automatically becomes a law for the member states, all regulations must be in all state languages. But this is hardly a convincing explanation inasmuch as most of the reports and memoranda produced by the EC are not laws yet are translated. Furthermore, sophisticated work on "language risk insurance" would permit the use of lingua franca contracts and regulations in which signatories would be protected if they broke the law due to language failure.57 Insurance for misunderstanding would be far cheaper than the titanic load of translation. The opposition to a single official language is political rather than legal or technical SR

Could a slightly different political landscape change these feelings? Suppose there were a bargain between the Eurocentralizers in the European Commission (who would want to rationalize European administration) and the regional decentralizers in the member states (who would want to decrease the ideological power of the state languages). Like the bureaucrats in the Indian Administrative Service, members of the European Commission do not carry briefs for the language interests of the states they represent. Both the Eurocrat and regionalist positions could be advantaged by promoting a cosmopolitan language as the European lingua franca (although many present Eurocrats would hold on to French as long as possible), thereby reducing the official role of the state languages. To be sure, the state languages would have to be protected (and the states be allowed to break subsidy norms by promoting TV and literature in the state languages) within the boundaries of the states. But the regional languages would be protected as well in the realms of TV, education, and local services.

This bargain faces high hurdles to be effective in the European Parliament. For one, many members of the European Parliament (especially from Greece, Spain, and Portugal) do not have facility in English. Second, European parliamentarians are not likely to carry the brief for regionalists. To be sure, a number of members of the European Parliament feel marginalized by members of Parliament in their own states and see an alliance with regionalists within their own states as a way to increase their relative power compared to their state legislature counterparts.59 But this is not a powerful reason for non-regionalists in the European Parliament to act as regionalist agents. Furthermore, there are few real regionalists in the European Parliament. The rules of voting for the European Parliament are such that in most countries, voters see only "national" lists (rather than regional ones). The kinds of parties that thrive on regional interests cannot survive in this electoral atmosphere. In the 1989 elections, France, Greece, Portugal, Germany, Denmark, and the Netherlands had complete or virtual one-constituency rolls for each whole country. Britain had single-member districts (vitiating proportional representation), and this too limits the vote-getting capacities of regional parties. As a result, the number of members of European Parliament (MEPs) associated with regional interests were very few. In Spain there were 6 such deputies, in Belgium 2, in Italy 4 (S. Tyrol People, Sardinian Action, and Lombardy Region), and in the United Kingdom 1 (plus 3 from Northern Ireland). With 518 MEPs in total, these regionalists could hardly make a dent.60 Their votes were insufficiently worthy for them to procure an alliance with any major parliamentary force; certainly, their voting strength was too paltry to press for a centralizer/regional alliance. The regional minorities would have to become far more mobilized politically than they are now, press for changes in the voting rules for the European Parliament, and then put pressure on the EC for the promotion of regional languages along with acceptance of a European lingua franca. Early evidence suggests that regional movements are seeking to change state-level voting rules for elections to the European Parliament, but so far little headway has been made.

The European Commission has been more successful in its alliance with the regionalists outside of the parliamentary domain. The preamble of the Maastricht Treaty ensures that "decisions are taken at greatest proximity to the people." This statement helped the commission, along with a coalition of regionalists, to foster the "Committee of the Regions." It is really made up of a patchwork of representatives from German Lander; regional autonomies of Belgium, Italy, and Spain; the French regions; and English Counties, as well as mayors of small communities, mayors of big cities, and presidents of French departments. Yet in 1996, it attained a status equal to the EU's Economic and Social Committee, which is institutionally weak but much greater than the previous situation of the "Consultative Council." Very propitious for this alliance is that it has been propelled by the actions of the German Lander, seeking to halt the EC from intervening in their competencies. In conjunction with this, the structural funds distributed by the EU in its regional policy have grown from 64 billion ECUs in 1989-1993 to 141 billion ECUs for 1994-1999 and now is a third of the EU budget. Since 1988, these funds have been disbursed according to commission-set criteria rather than member state agreements, and the technical criteria involves funding projects based on "community initiatives," which worked to the advantage of regional applicants, especially the Lander. The national governments have not pushed this trend; rather, the force is with the European Commission with its regionalist allies.61 To the extent that the Lander push for greater power for the Committee of the Regions (for purposes of their autonomy and to procure development funds), they may be perfectly happy to support the linguistic demands of their allies on the committee as part of a wider bargain. This would be far more powerful than if only the regions with linguistic programs fought for themselves.

This political dynamic leading to a stable European language constellation is a somewhat distinct dynamic from the model portrayed for postcolonial states. In Europe, English has become the de facto lingua franca (as Hindi had become in at least north India). As Eurocrats seeking efficiency seek to ride this trend-promoting English at all opportunities-they face subversion, especially from the bureaucrats of the national states that constitute Europe. Eurocrats must agree to protect state bureaucracies and languages in order to survive. In this situation of the division of linguistic authority, the regional leaders have a historic opportunity. By promoting local vernaculars, they get the unexpected support of the Eurocrats, who see in a 2 +/- 1 language outcome one that allows for a single European language for interstate and all-union business. This outcome has the same equilibrium qualities as does the 3 +/- 1 outcome in India.

The Rule of "Europe" in Post-Soviet Nationalism

With the collapse of the Soviet bloc, Europe of course is in the process of expansion. Yet, some of the forces unleashed by the fall of communism will sustain the 2 +/- 1 European equilibrium. After all, Catalonia is by no means the only case where the ideology of Europe justifies multilingual repertoires. In Estonia (and in other multinational post-Soviet states), very similar phenomena can be observed. Since the Russian conquest of today's Estonia in 1710, a slow but significant immigration (first officials, then merchants, and finally nobles who received manors as grants from the czar) of Russians began to take place. After the completion of the St. Petersburg-to-Tallinn railway in 1870, a more significant wave of workers swept into the newly built factories and settled in Estonian towns.62 After World War II, a massive wave of immigration (and the forced displacement of the Estonian population) "Russianized" the northeastern region of Estonia. In the border city of Narva, for example, by the 1989 census nearly 95 percent of the population identified itself as Russian speaking. Although there were many Ukrainians, Belarusans, and Jews in this wave, today's rhetoric classifies them all as part of a "Russian-speaking population."63

The Estonian independence movement in the wake of the Soviet collapse could not ignore the monolingual Russian-speaking population. Although many Estonian leaders worked to induce the Russian speakers to migrate back to their "homeland," most of these people considered Estonia to be that homeland and had no intention (nor any place) to "return." The diverse Estonian political factions eventually accepted the reality of the Russian presence but agreed among themselves that fluency in Estonian would become a sine qua non for Estonian citizenship. Russian speakers would need to pass an examination in the Estonian language (a language that is not of the Indo-European family and quite difficult for Slavic speakers to learn) to become citizens or to work in a whole range of white-collar jobs.

While this reads like a standard battle over cultural hegemony within a single state, "Europe" plays a key role in its development. The Estonian Parliament was considering a "Law on Foreigners" (read "Russians") in the summer of 1993 that stipulated the documents that non-citizens would need to procure in order to maintain their rights as working residents of Estonia. Leaders of the Russianspeaking population were outraged (more by the language of the law's title than by its contents) and appealed to international standards of human rights to block its passage. Sergei Sovetnikov, a former parliamentarian, wrote an open letter to President Lennart Meri in which he pointed out that Estonia's recent entry into the Council of Europe was premised on the acceptance of international standards of rights. The proposed law, he wrote, "means that no sooner have we entered into the European home than we are taking our first step to leave it."64 The law was passed by Parliament, but the constitution required the president to sign it. In an unprecedented step, he asked for and received advice from a committee of jurists from the Council of Europe that recommended certain changes to better ensure the rights of non-citizens. Meri outraged his prime minister by refusing to sign the law and went on state television to justify his unusual step. "Estonia is constructing a European precedent," he said, "and this will give impulse to European integration. Estonia is answering to Europe, and Europe will answer to Estonia."65

In terms of language, anti-Soviet idelolgy provides strong incentives for Estonians to learn English, replacing Russian as their language of international communication. Most Estonians see only Estonian and English as absolutely necessary to establish themselves as Estonians and Europeans. Although many speak Finnish, Swedish, German, and/or Russian (which will decrease significantly within a generation), the two-language repertoire (Estonian and English) will be the standard minimum for Estonians. "Europe" for the Russian-speaking population also means learning English, and many of these people feel desperately behind the Estonians in this quest. But being part of Europe requires for them (as they see Russia falling into a chaos that they identify as non-European) to learn the state language of Estonia. Middle class professionals thereby spend their evenings and free time at state language centers learning Estonian in order to qualify for their jobs and citizenship. They complain, in no uncertain terms, to the state authorities not of the near impossibility of their burden but of the insufficiency of Estonian instruction for their children. To be sure, learning English and Estonian is not at all seen as a substitute for Russian. The Russian-speaking leaders have demanded, and the Estonian state is committed to providing, schools where the medium of instruction will remain Russian. Maybe not in this generation, but by the next, the Russian-speaking population of Estonia will have a standard threelanguage repertoire (Russian, Estonian, and English), which will signal their nationality, their citizenship in Estonia, and their simultaneous membership in "Europe."

"Europe" is important to the Estonians and Russian Estonians just as it is to the Catalans. It cleans nationalism from its ugly connotations and sets up a model of human rights in which minorities (and minority languages) receive protection from higher authorities. It is a potential coalition of minorities and small nationalities, all locked into standard multicultural (and multilingual) identities, that seeks to build a European reality. Even supposing that post-Soviet states successfully nationalize-with a single official language, that of the titular nationalitythere is reason to believe that their entry into Europe (something for which there is already a strong societal consensus) will give added support to the 2 +/- 1 European language constellation.

CONCLUSION

It would be historically inaccurate to argue that because there is little cultural rationalization going on in Europe, there is not any evidence of a new European state. This view neglects that states constructed in the twentieth century have different cultural configurations than those that were constructed in the sixteenth to eighteenth centuries. The argument of this article is that there is a process of state building in Europe going on today, except that it looks more like India's experience since 1947 than France's since 1516.

The observation that European state consolidation is underway does not imply that the road to this future is well paved. All state-building projects suffer innumerable setbacks, including violence. In fact, there is no assurance of successful consolidation at the end of the road. Strong regional identities, when backed with economic dynamism and temporary weakness at the center, can unleash powerful incentives for separation. Leaders in poorer regions may see advantage in control over the civil service jobs that are associated with sovereignty and thereby press for separation as well.66 Separatist movements, as in Catalonia, can be peaceful; they can also, as in Basque country, involve terrorism. These regional movements were both moderately effective in part because of postFranco Spain's commitment to become part of a European project. But the Indian experience, with the holocaust of partition and subsequent communal violence, is not a happy alternative. The point being made here is not that the future is bright but rather that the citizen of a future European state will have a cultural or national identity that involves a far more complex language repertoire than was required of citizens of early-developing European states.

The future European citizen will have multiple languages and multiple cultural identities. But a certain "European" constellation will become institutionalized, and this will represent a new state identity. She or he will be European (probably expressed through gutless English), a citizen of a core European state (the official members of the EC), and possibly also a member of a protected minority (or regional) language community. Some citizens will need to learn only one language (English speakers in England); most will require two (German, French, and Dutch speakers in Germany, France, and the Netherlands, respectively). Many others will be fluent in three, such as Catalans living in Spain (English, Spanish, and Catalan), Frisians living in the Netherlands (Frisian, Dutch, and English), and Slovenians living in Italy (Slovene, Italian, and English). Individual multilingualism and multiple identities will become normal. States consolidating in the twenty-first century will promote an international lingua franca, central state languages, and regional languages as well. Whether this outcome gets recognized or not in a political bargain, it is likely to emerge as the de facto European language constellation of the next century.

Earlier versions of this article were presented at the "European Identity and Its Intellectual Roots" conference at Harvard University (organized by Michael Herzfeld and supported by the Joint Committee on Western Europe of the Social Science Research Council), May 1993, and at the "Quo Vadis Europa, 2000?" conference at the University of California, Los Angeles (organized by Michael Mann and supported by the Center for European and Russian Studies at UCLA), March 1994. Sections from previous publications-David Laitin, "The Game Theory of Language Regimes," International Political Science Review 14, no. 3 (1993): 227-39; David Laitin, Language Repertoires and State Construction in Africa (Cambridge, UK: Cambridge University Press, 1993); and David Laitin and Guadalupe Rodriguez, "Language, Ideology and the Press in Catalonia," American Anthropologist 94, no. 1 (1992): 9-3-are incorporated into this article without quotation marks or direct citation. The author would like to thank Peter Katzenstein, Rogers Brubaker, and James Fernandez for comments on an earlier draft of the article.

NOTES

1. Abram De Swaan, "The Evolving European Language System: A Theory of Communication Potential and Language Competition," International Political Science Review 14, no. 3 (1993): 241-55.

2. Benedict Anderson, Imagined Communities (London: Verso, 1983). 3. Alexis de Tocqueville, Democracy in America (New York: Schocken Books,1961), 14.

4. On the German romantics, see, for example, Wilhelm Von Humboldt, Linguistic Variability and Intellectual Development (Philadelphia: University of Pennsylvania Press, 1971), chap.1; Joseph Stalin, Marxism and the National Question (New York: International Publishers, 1942); Ernest Gellner, Nations and Nationalism (Ithaca, NY: Cornell University Press,1983), esp. chap. 6; and Karl Deutsch, Nationalism and Social Communication (Cambridge, MA: MIT Press, 1953).

5. Peter Katzenstein, "Hare and Tortoise: The Race toward Integration," International Organization 25, no. 2 (1971): 290-95; Ernst Haas, The Uniting of Europe (Stanford, CA: Stanford University Press, 1958); Karl W. Deutsch, Sidney A. Burrell, Robert A. Kann, Maurice Lee, Jr., Martin Lichterman, Raymond E. Lindgren, Francis L. Loewenheim, and Richard W. Van Wagenen, Political Community and the North Atlantic Area (Princeton, NJ: Princeton University Press, 1957); John Ruggie, "Collective Goods and Future International Collaboration," American Political Science Review, 66, no. 3 (1972): 874-93.

6. Eric Hobsbawm, Nations and Nationalism Since 1780 (Cambridge, UK: Canto, 1990).

7. Steven Greenhouse, "In Europe, a Boom in M.B.A.'s," The New York Times (29 May 1991).

8. John Rockwell, "French Strike a Blow for la Gloire in Their Film Industry's Oscars," The New York Times (18 January 1993).

9. Max Weber, Economy and Society, 2 vols. (Berkeley: University of California Press, 1968), 71 (economic rationalization), 655, 809-38 (legal rationalization), 1108 (educational rationalization).

10. See the discussion of language, the nation, and the state in Max Weber, "The Nation," in H. H. Gerth and C. W. Mills, eds., From Max Weber (New York: Oxford University Press, 1958), 177-79.

II. The classic statement of this position is that of Deutsch, Nationalism. For a powerful empirical demonstration of this point, see Susan Watkins, From Provinces into Nations (Princeton, NJ: Princeton University Press, 1991).

12. This position is consistent with Deutsch's theories in Nationalism. See Cyril Black, The Dynamics of Modernization (New York: Harper & Row, 1966). 13. J. E. Schwartzberg, "Factors in the Linguistic Reorganization of Indian States," in P. Wallace, ed., Region and Nation in India (New Delhi: Oxford University Press, 1985). 14. Eugen Weber, Peasants into Frenchmen (Stanford, CA: Stanford University Press, 1976).

15. Anderson, Imagined Communities, chap. 5.

16. B. G. Kher, Report of the Official Language Commission (New Delhi: Government of India Press, 1956), II.

17. Robert Price, Society and Bureaucracy in Contemporary Ghana (Berkeley: University of California Press, 1975).

18. Politicians and bureaucrats can well be the same people or from the same family. Nehru, for example, had three close relatives in the Indian Civil Service. See David Potter, India's Political Administration (Oxford, UK: Clarendon, 1986), 129. When this occurs, we can say the individual or family is cross-pressured, as the theory here is positing "roles" rather than individuals.

19. K. L. Gandhi, The Problem of Official Language in India (New Delhi: Arya Book Depot, 1984), chap. 3.

20. Game theorists would be reluctant to call the outcome described as an equilibrium. Through backward induction, they could point out, the Congress leaders should have known they would have lost to the bureaucrats and sought to rationalize English rather than Hindi. The reason that all parties are satisfied lies in the phenomenon of "sour grapes." See Jon Elster, Sour Grapes (Cambridge, UK: Cambridge University Press, 1983). Indian politicians now glorify the 3 1 outcome as if it were their original goal and consistent with India's historical diversity.

21. Charles Ferguson, "Diglossia," Word 15 (1959): 325-40. 22. This model is specified more fully in David D. Laitin, "Language Policy and Political Strategy in India," Policy Sciences 23-24 (1989): 428-30. 23. Commission of the European Communities, Eurobarometer, no. 28 (December): 78.

24. De Swann, "The Evolving Europen Language System," 248-49. Although De Swann does not address this issue, clearly the economic, military, political, and cultural dominance of the United States-and not the mere entry of the United Kingdom into the EC-explains the power of English as the continental lingua franca.

25. "Dumbstruck," The Economist (14 January 1995): 45.

26. These data are from a survey run by Jerry Hough and David Laitin under a National Science Foundation grant (SES9212568), "Nationality and Politics: The Dismemberment of the Soviet Union." Data from that project are deposited in the archive of the Inter-University Consortium for Political and Social Research, University of Michigan.

27. Florian Coulmas, ed., A Language Policy for the European Community: Prospects and Quandaries (Berlin: Mouton de Gruyter,1991), 24-25. This figure must surely reflect a very broad understanding of what it means to defend French. Yet, the fact that France does spend considerable sums of money on this activity is without dispute. 28. Ulrich Ammon, "The Status of German and Other Languages in the European Community," in Coulmas, A Language Policy, 241-54. 29. Coulmas, A Language Policy, 13. 30. Ibid., 26-27.

31. See Marlise Simons, "A Reborn Provenqal Heralds Revival of Regional Tongues," New York Nmes (3 May 1993): 1, 6 (national edition). In her entertaining report, she correctly points to the energetically administered programs within the regions of France compelling students to learn "their" language. But when she writes that "the [central] government has concluded that France's regional languages enrich the national heritage rather than pose a threat to the country's identity," she underemphasizes the continuing state interest in the promotion of French.

32. Bruno De Witte, 'The Impact of European Community Rules on Linguistic Policies of the Member States," in Coulmas, A Language Policy, 174-75. 33. Ibid., 173. 34. Ibid., 167.

35. Bob Jessop, "Post-Fordism and the State," in Ash Amin, ed., Post-Fordism: A Reader (London: Blackwell, 1994). 36. Coulmas, A Language Policy, 14-16.

37. Elisabetta Zuanelli, "Italian in the European Community: An Educational Perspective on the National Language and New Language Minorities," in Coulmas, A Language Policy, 293.

38. Claude du Granrut, Europe: Le Temps des Regions (Paris: Librairie Generale du Droit et de la Jurisprudence, 1994). 39. Ibid., 59.

40. Some data for this paragraph are from the Commission of the European Communities, Linguistic Minorities in Countries Belonging to the European Community (Luxembourg: Office for Official Publications of the European Communities, 1986). The data do not cover Spain, Portugal, or Greece.

41. See the resolution of the European Parliament, "On Languages in the Community and the Situation of Catalan," A3-0169190. The Catalans hardly got what they asked for in this resolution, but the call for inter alia "the publication in Catalan of the [European] Community's treaties and basic texts" was considered by Catalans as a foot in the door. 42. Harold Koch, "Legal Aspects of a Language Policy for the European Communities: Language Risks, Equal Opportunities, and Legislating a Language," in Coulmas, A Language Policy, 174-75.

43. Andree Tabouret-Keller, "Factors of Constraints and Freedom in Setting a Language Policy for the European Community: A Sociolinguistic Approach," in Coulmas, A Language Policy, 47. 44. Simons, "A Reborn Provencal."

45. David Laitin and Guadalupe Rodriguez, "Language, Ideology and the Press in Catalonia," American Anthropologist 94, no.1 (1992): 9-30. 46. "La realitat autonomica," Avui (14 June 1985). 47. "Catalunya i el futur europeu," Avui (6 June 1985). 48. "Un primer pas," Avui (19 June 1985). 49. "L'Europa del 92: Un punt de partida," Diari (30 June 1988).

50. Juan J. Linz and Alfred Stepan, "Political Identities and Electoral Sequences: Spain, the Soviet Union and Yugoslavia," Daedalus 121 (1992): 128.

51. Keating suggests that among the Scottish and Basques, revivalists are connecting regional autonomy to the European project in ways similar to how I have portrayed the Catalans, although he suggests the Catalans are less inclined in this direction than the Scots and Basques. See Michael Keating, "The Continental Meso: Regions in the European Community," in L. J. Sharpe, ed., The Rise of Meso Government in Europe (London: Sage, 1993), 309.

52. Cited in Rudolf Wildenmann, ed., Staatswerdung Europas? Optionen fir eine Europaische Union (Baden-Baden, Germany: Nomos, 1991), 19-41. 53. Du Granrut, Europe, 198-200.

54. See especially Fritz W. Scharpf, "Kann es in Europa eine Stabile foderale Balance geben?" in Wildenmann, ed., Staatswerdung, 415-29. 55. Grin presents an interesting but unconvincing economic logic to the success for some of Europe's regional languages, having to do with trends in intra-European trade. See Francois Grin, "European Economic Integration and the Fate of Lesser-Used Languages," Language Problems and Language Planning 17, no. 2 (1993): 107. 56. Coulmas, A Language Policy, 22-26. 57. Koch, "Legal Aspects," 147 ff.

58. See the Parliamentary admonitions demanding equality of all languages in parliamentary affairs in Coulmas, A Language Policy, 7. 59. Keating, "The Continental Meso," 303.

60. Data for this paragraph are from Enid Lakeman, "Elections to the European Parliament, 1989," Parliamentary Affairs 43, no. 1 (1990): 77-89; and The Times (London), Guide to the European Parliament (London: The Times, 1989). 61. Christopher Ansell, Keith Darden, and Craig Parsons, "Dual Networks in the European Union" (unpublished manuscript, University of California, Berkeley, n.d.); Du Granrut, Europe, 185-86.

62. Alexander Loit, "Nation-Building in the Baltic Countries, 1850-1918," in Justo G. Beramendi, Ramon Maiz, and Xose Nunez, eds., Nationalism in Europe: Past and Present (Santiago de Compostela, Spain: Universidade de Santiago de Compostela, 1994), 479504.

63. David D. Laitin, Identity in Formation: The Russian-Speaking Populations in the Near Abroad (Ithaca, NY: Cornell University Press, forthcoming). 64. "Evropu my uvazhili, teper' by uvazhit' evoi narod," Narvskaia Gazeta (29 May 1993).

65. "Telivizionnoe obrashchenie prezidenta respubliki," Narvskaia Gazeta (13 July 1993).

66. These two sources of separatist movements in regard to the former Soviet Union are discussed in David D. Laitin, "The National Uprisings in the Soviet Union," World Politics 44, no. 1 (1991): 139-77.
Reproduced with permission of the copyright owner. Further reproduction or distribution is prohibited without permission.