LANGUAGE AND FEDERALISM: THE MULTI-ETHNIC CHALLENGE

International Conference on Federalism Mont-Tremblant, October 1999 BACKGROUND PAPER LANGUAGE AND FEDERALISM: THE MULTI-ETHNIC CHALLENGE Subrata K. Mitra Department of Political Science South Asia Institute Heidelberg, Germany Introduction Largely for historical reasons, the role of language has remained outside mainstream research on federalism. Classic studies of federalism, particularly those based on the United States as the main empirical exemplar have concentrated on powersharing between two sets of governments, one representing the constituent units and the other a central government as the main issue of federalism. The empirical research on federalism thus concentrates on the key institutions of state like a written constitution stipulating the division of powers between the two sets of governments, separate sources of revenue, and, a Supreme Court to act as an impartial arbitrator in cases where the division of powers is not clear. These principles have served well in countries where a core national identity based on language existed before the establishment of a federal state. Language, in that sense, has been a non-issue in the United States and Australia, or for that matter, in language-proud and linguistically homogeneous Germany. However, in post-colonial states like India (where the status of Hindi as the national language was contested by non-Hindi speaking people after the transfer of power by the British to the Congress party, dominated by Hindi speaking leaders), Pakistan (which subsequently split on this issue, giving rise to Bengali-speaking Bangladesh) and Ceylon (subsequently SriLanka, currently locked in a conflict on the issue of Tamil) where the task of creating a nation out of many distinct social groups based on their own language followed the founding of the state, language can emerge as the most severe challenge to national unity. Language, especially when it is believed to be a basis of economic discrimination, can become a main issue in rich, industrial, liberal democratic states like Canada or Switzerland as well when implicit cleavages on the lines of languages, long dormant, are revived. In these cases, the institutional structure of federalism, rather than being a sine qua non of political community, becomes a political battle ground. With competing languages vying with one another for supremacy as its main focus, this paper suggests that the success of a state in coping with the challenge of linguistic diversity depends on its capacity to respond through constitutional accommodation of diversity and adequate policy responses through firm and clear social, political and economic initiatives. The Puzzle Language plays a double role with regard to the problems of legality and legitimacy in the modern state. As a thin bond, language is the basis of communication between the institutions of the state, and the basis of the negotiation for power and resources between competing social groups. As such, the availability of a common language that minimises uncertainty, loss of information in course of transmission and equalises the chances for competing groups considerably enhances both the legality and the legitimacy of the state. As a thick system of meanings, carrying the burden of history, religion, culture, ritual and memory, language is a moral bond between the state and the individual. The issue here is not so much what the individual can get out the state as whether the individual can identify with the state which, short of this identification, remains only a formal, distant and oppressive presence. On the basis of these theoretical assertions, one can construct cases where, as in the United States or Australia, the historically dominant Anglo-Saxon ethos has acted as the gate-keeper for generations of immigrants and made the acceptance of linguistic homogeneity a condition of their integration. At the polar opposite are cases like India or Pakistan after the end of colonial rule which, having become free-standing states in their own right, set about constructing nations. Poised between the two types are continental states like Switzerland where a specific language could not, for historical and political reasons, be given the uncontested status of an exclusive, national symbol. Drawing on Switzerland and India as empirical exemplars of the liberal-democratic multi-lingual state and the post-colonial multi-lingual state, the paper raises a number of specific questions. In terms of which institutions and processes do these states formulate and solve the problems of legality and legitimacy with regard to the issue of multi-lingualism? Is bi(multi) lingualism practised at a territorial or personal level? How do “minority” languages survive? Taking into account the uneven strength and development of the different languages, how does the imperative of achieving federal balance cope with the issue of language asymmetry? Finally, how does the state avoid the perils of irredentism, linked to the status of minority languages? Switzerland: The Institutionalisation of Linguistic Diversity The statistics about the percentage of people who identify with the four main languages of Switzerland gives an insight into the potential problems that the state could face. (Table 1) Table 1: Residents of Switzerland by mother-tongue (in %) Source: Statistisches Jahrbuch der Schweiz 1995 (Bern, 1995) cited in Steinberg, p 131 Those unfamiliar with Switzerland will wonder whether the figures reported above translate themselves to an automatic supremacy for German. The reality however is very different and quite complex. As a matter of fact, Switzerland has not one but four national languages and the role of language is institutionalised in a manner where it is not usually perceived as a threat to the identity or the livelihood of any particular group In fact so successful has Switzerland been in this manner that in the words of a renowned specialist, « language defines and at the same time denies Swiss identity; it reinforces the peculiarities of political practice and reflects them. » (Steinberg: 1991; 130) This remarkable phenomenon can be explained in terms of the following arguments. In the first place, though the fact that German is mentioned as the mother tongue of the majority of the Swiss population, in philological terms, what goes by the name of Swiss German is Schwyzerdütsch, a broad category that describes an astonishing number of dialects spoken in a relatively small geographical area. The broad category can be further subdivided into three linguistic groups, Low Alemannic, High Alemannic and Highest Alemannic. While the broad categories create an overall fuzziness when it comes to Schwyzerdütsch as a linguistic category, the internal differences are further sharpened thanks to the pride that the Swiss Germans take in their Dialekt, a far cry from the condescending approach with which the French refer to patois. Dialect is a constant subject of discussion among Swiss German-speakers and a distinct form of identity. In the second place, dialect is identity but it is also a form of social communication. Third, an implicit two language formula facilitates dialect at home, and high German at school, Church and work. The juggling involves what Steinberg, citing the Czech philologist Olga Neversilova (p 138) calls “to learn to play the piano and drive a car at the same time” but it surely takes the edge off the complexes of majority-ism. While these points explain why the Swiss-German identity does not crush the others taking advantage of its huge share in the population, the following describes why the French, for all purposes a proud minority, does not take advantage of its geographic proximity to France to endanger the federation by starting an irredentism movement. Historically, the cultural identity of French Switzerland is defined in opposition to France as much as with reference to it. In contrast to France, culture in Swiss cantons has never been tied to the state, nor been an instrument of state power. Culture has had its existence in little compartments which have never been unified by a central power or made uniform as was the case of the French provinces under successive regimes. The French cantons are old republics founded on the autonomy of the communes. Protestantism is dominant in French Switzerland; it has determined the greater part of customs, profound moral concerns and distrust of ceremony, in contrast to French Catholicism. Unlike in some states like SriLanka where the cumulation of linguistic and religious cleavages endangers national unity, in Switzerland they are fortuitously cross-cutting, with 61.3% of the German-speaking declaring themselves to be Protestant and 37.2% Catholic, and 53.7% of the French speaking being Protestant and 44.4% being Catholic. Finally, the French cantons are not only neighbours to a German world; they are in constant interaction with it, more than is commonly realised. Unlike the Swiss Germans, the “Swissness” of the Suisses romands is not primarily based on language. History, not an exclusive language, has made them Swiss, and religion, politics and economics forces have kept them that way. The most striking characteristic of French speaking Switzerland is its diversity. To begin with, the six territories are geographically divided, and their river systems, unlike those of the German areas, do not lead to one great central valley. They are divided by religion, which, in turn, means that they are divided by culture, education, and social custom. They are divided politically, including the most radical and most conservative communities in Switzerland. Each canton has its own school system, its own university (with the exception of the comparatively recent French mono-lingual canton of Jura) and higher secondary systems and its own tax laws. In terms of cultural one-upmanship, the membership of a great cultural, Paris-centred world, gives French-speaking Swiss an aura of superiority over their more provincial Swiss German cousins, who have nothing comparable to offer in reply. On the other hand, for the Parisian French, the Swiss are hopelessly provincial, “les petits Suisses”. Swiss French live in an ambivalent position toward both their fellow citizens of different language and their fellow French-speakers of different citizenship. Multi-lingualism is a national not an individual ideal. Muliti-lingualism is a fact about the Swiss as a people; it is much harder to say if and to what extent it is an attitude of individual Swiss men and women. The government working party on language stated in 1989 that “multi-lingualism in our country is a social fact, i.e., that four linguistic regions make the whole, but the inhabitants of these regions are predominantly monolingual.” The territorial principle, recognised by article 116 of the Swiss Federal Constitution, is a recognition of this fact. There is, however, enough personal bi-lingualism available to make interaction between cantons possible. The census of 1990 confirms this fact. The functional necessity of personal bi-lingualism is indicated by the fact that it is inversely proportional to the size of the linguistic group. Whereas 65.4% of Swiss Germans declared themselves to be mono-lingual, only 43.4% of Swiss French, 40% of Swiss Italians and 20.3% of the Swiss Romansch did so. Evidently, the smaller the linguistic community the more likely it is that its speakers will use other languages regularly. The monolingual percentage of Swiss Germans matches the almost identical percentage from the same census, 66.4% who state that they never speak High German. Whereas fewer than 1% of French Swiss and 0.4% of Italian Swiss state that they only use dialect, two-thirds of Swiss German live in the world of dialect exclusively. An intricate system of subsidies, rotation of elective office, and the tradition of direct democracy which gives an opportunity to exercise power at the local level to the nationally fragmented but locally homogeneous linguistic groups has facilitated the interweaving of both the thin and thick bonds in the creation of a state and a nation without a specific national language. India: Linguistic diversity and the problem of national integration Seen against its western counterparts, the federal system of India at first appears like an unlikely member of the club. Compared to the relatively longer existence of the four key federal states, namely the United States (1789), Switzerland (1848), Canada (1867), and Australia (1901), their racial homogeneity, and high literacy and standards of living considered necessary for the sophisticated power-sharing that a federal system requires, India presents a set of apparently insurmountable obstacles against a likely federal solution. However, with a clear, constitutionally guaranteed division of power between the central government and the constituent States, effectively policed by an independent Supreme Court, separate, direct elections to the central and regional governments monitored by an independent Election Commission, and, the capacity of the political process at sustaining a dynamic balance between the jurisdictions of the two sets of governments, India exhibits many of the features of federalism. But scholarly scepticism persists nevertheless and surfaces as part of a larger question: with her multi-ethnic society, structural asymmetry of constituent units, mass illiteracy and poverty and the uphill task of simultaneous state formation and nation-building why does federalism even in a broader sense survive in India at all? India’s linguistic diversity was a source of scepticism about the durability of the state in the early years after independence. “As a civilisation and as an integrated cultural whole, India has shown a power of survival rivalled only by China. But multilingual India’s separate territories have failed as consistently as Europe’s to hold together as a political unity.” (Harrison: 1960; 3) Harrison went on to predict the failure of “so many linguistically differentiated peoples, all of them so self-aware, all numbered in millions and tens of millions” to hold together “within a single national body politic”, thus creating the prospects of “’anarchy’, ‘Fascism’, and ‘totalitarian small nationalities’.” That Indian experience belies such dire predictions is a homage to good constitutional and political management, and a particularly innovative use of the potential of federalism to accommodate linguistic diversity. While originally it was hoped by the Fathers of the Constitution that India would become monolingual in the course of a few years, it has actually proven necessary not only to accept multi-lingualism, but also to restructure the country’s effective divisions in accordance with diversified cultural and linguistic ambitions. Quite understandably, following independence from British rule and the partition of India on religious lines, unitary sentiment ran strong and in testimony to these feelings, the constitution (art 343) categorically affirmed that “the official language of the Union shall be Hindi in Devnagari script” and (art 351) that “it shall be the duty of the Union to promote the spread of the Hindi language… so that it may serve as a medium of expression for all the elements of the composite culture of India…”. Yet, even though the federal scheme of the constitution of 1935 had proved unworkable, there was virtually no sentiment for the establishment of a constitutionally unitary state, but there was powerful sentiment for giving the central authorities sufficient power to combat the centrifugal tendencies in so vast a country with so rich a diversity of cultures, languages, and religions. With India’s complex linguistic profile (table 2) it is difficult to envisage how indeed it could have been otherwise. Among the most interesting challenges to centralising tendencies and the former unitary propensity was the political opposition to “Hindi imperialism” from non-Hindi speakers to policies designed to carry out the constitutional injunction for establishing Hindi as the common language of India. As a result, English continues as the available lingua franca of a good part of the governing and business elites of India, though, regional languages, thanks to the exigencies of competitive, democratic politics and the need for communication with the masses in their mother tongue by regional and local politicians. Under the compulsions of mass democracy and the imperative of coalition-building, the state has followed the path of least linguistic resistance. The combination of constitutional design and democratic politics has created a complex language formula that accommodates Hindi and English as national and link languages, uses major languages as official language in respective regional States but protects minority languages, ensconced within largely mono-lingual regional States through a constitutional guarantee of group rights to culture and identity. Table 2: Major Language Groups in India by Population Source: Hardgrave and Kochanek, p 11 Faced with the challenge of linguistic diversity, India has come up with solutions that combine constitutional innovation and political consensus-building. The most important of the former is the gradual dilution of the strong emphasis on the role of Hindi as the national language. No consensus emerged with regard to the apprehensions of non-Hindi speakers who form a majority of the population of India during the fifteen years following independence. As such, India settled down for a three language formula whereby Hindi was given the status of the national language with the proviso that English will be indefinitely retained as the link language and regions will have one or more official languages. The territorial multi-lingualism dovetails into the federal division of powers where most of the policy aspects of language are covered by the subjects that fall within the State list, under the area of competence of the regional government. In the second place, since this three-plus language formula could still create linguistic minorities out of groups who are majorities elsewhere, (see table 3 for the description of regions by language) a national watchdog body has been created for this purpose. The Commission for Linguistic Minorities in India is appointed by the President of India and for all purposes, is independent of the cross-currents of India’s national and regional politics. This Table 3: States and Territories of the Indian Union and their Principal Languages Source: Hardgrave and Kochanek, p 143 institutional defence of the rights of minorities is of paramount importance. The following citation from the recommendation of the Commission, in this particular case directed against the government of Punjab, will give a flavour of the functioning of this remarkable institution: The governments of Punjab have not been furnishing replies to the annual questionnaires since the re-organisation of Punjab on linguistic basis in 1966. They had stated that the question of furnishing replies does not arise as Punjab has become a unilingual State and, therefore, there are no linguistic minorities in the State. The stand taken by the State Government is not in consonance with the all-India decision as well as the recommendations of the States Reorganisation Commission. According the Census figures of 1971, the speakers of Hindi and Urdu constitute 19.9 and 0.2 per cent respectively of the total populations of the State. Punjab is thus not completely unilingual and the linguistic minorities in the State are entitled to the various safeguards. Keeping these facts in view, the Government of Punjab may appropriately reconsider their stand and take necessary action to implement various safeguards provided for linguistic minorities. (Government of India, Report of the Commission for Linguistic Minorities, 1985; 97) For the protection of the other languages in use, the following directives are provided. For the submission of representation for the redress of any grievances to any officer or authority of the Union or a State the petitioner is authorised to use any of the languages used in the Union or in the State, as the case may be (art 350). Every State and other local authority within a State is directed to provide adequate facilities for instruction in the mother tongue at the preliminary stage of education to children belonging to linguistic minority groups and the President is authorised to issue such directions to any State as he may consider necessary for the securing of such facilities. (art 350a) A Special Officer for linguistic minorities is appointed by the President to investigate all matters relating to the safeguards provided by the Constitution for linguistic minorities and to report to the President upon those matters. It shall be the duty of the President to cause all such reports to be laid before each House of Parliament and also to be sent to the Government of the State concerned. (art 350B) Further, attempts are made to encourage personal bi-lingualism at the individual level for civil servants and academics. Upward mobile politicians, businessmen and professionals with the all-India market in perspective have seen the importance of bilingualism. The functional interface of competing languages is provided by bi-lingualism, which like in Switzerland, is present in greater numbers in smaller language groups where the facility of speaking another language considerably enhances career options. The figures for persons speaking a language subsidiary to their mother tongue are quite revealing: Hindi (5.10%), Oriya (5.75%), Malayalam (7.11%), Gujarati (7.31%), Tamil (8.11%), Bengali (8.65%), Assamese (8.96%), Marathi (10.47%), Kashmiri (10.69%), Telugu (14.03%), Punjabi (14.16%), Kannada (14.43%), Urdu (22.09%). Besides, differences arising out of language and religion are non-overlapping. Both Hindus and Sikhs speak Punjabi, Urdu is understood by Muslims, Hindus, Sikhs, and the Bengali literary pantheon counts several Muslim writers along with Hindus among its stars. The Indian Constitution provides for a clear division of powers between the Union government and the State governments in the seventh schedule. The Union controls the “central list” consisting of areas which involve inter-State relations, national security and foreign affairs. Subjects of primary interest to the regions, called the State list, are under the jurisdiction of the States. The concurrent list holds subjects of overlapping interest like land reform laws, or issues relating to the cultural or religious minorities where both centre and State can make laws with the understanding that in case of conflict the central laws will take precedence. This ensures the institutionalisation of unity as well as diversity. The fourth General Election of 1967 which drastically reduced the overwhelming strength of the Congress party in the national parliament to a simple majority and saw half of the States moving out of Congress control into the hands of opposition parties or coalitions, caused a radical change in the nature of centre-State relations. No longer could an imperious Congress Prime Minister afford to “dictate” to loyal Congress Chief Ministers in the regions. But, even as the tone became more contentious, the essential principles of accommodation and consultation continued during the crucial 1967-1969 period of transition. The Congress dominated centre started cohabiting with opposition parties at the regional level. The balance was lost once the Congress party split (1969) and Indira Gandhi, her party reduced to a minority in the Parliament, took to the strategy of radical rhetoric and authoritarian leadership. However, after the authoritarian interlude of 1975-1977, which, both in law and fact reduced India’s federal system to a unitary state, the system reverted to the earlier stage of co-operation between the centre and the States. That balance has stabilised during the alternating rules of the Congress party, the Janata coalitions and endures up to the present day The Hindu nationalist Bharatiya Janata Party which leads the current ruling coalition has been solicitous in its adherence to the norms in centre-State relations established by its predecessors, including such hallowed principles of the Indian Union as the three language formula in spite of its advocacy of Hindi as India’s national language during the long years in the opposition. Having established themselves in their regions, the erstwhile champions of regional languages have, in this new phase of Indian politics that we have now entered, set their sights at constructing the kind of nation that would be appropriate to the new scheme of things. Increasingly, rather than remaining content with their own region, they are stretching out their hand, using their alliances with similar forces from outside, to define the nature of the national community in their own ways. Recent experience in different parts of the country has demonstrated that the pursuit of these goals can not only coexist with similar aspirations elsewhere; regional movements can, in fact, reinforce one another by pooling their political resources. Hence the unprecedented scenes of regional leaders from one part of India campaigning (in languages other than their own!) for regional parties in other parts of the country. The Congress System encapsulated the expressions of local and regional interests and symbols at lower levels of the system; the new element in Indian politics makes these processes of consultation an explicit and a systematic way of bringing out India’s outlying areas and peoples, and weaving them into different ways of defining what the nation is about and who has the legitimate right to speak in its name. Conclusion: Why do some states succeed in coping with linguistic heterogeneity and others fail? On the basis of the comparative experiences of federal states in dealing with linguistic diversity, we can conclude that there are in the main three structural grounds that produce cause of disaffection: namely, (1) the linguistic “distance”, both philological and in terms of numbers, between the dominant and minority linguistic groups; (2) the real and perceived links between language and livelihood, and as such, the potential for discrimination on the basis of linguistic competence; and (3) the exclusion of the leaders of minority languages from the national elite. These structural causes of disaffection are exacerbated in practice due to the following processes: (1) the perceived control of the central government on the areas inhabited by the minority language groups; (2) geopolitical conditions helpful to separatists, such as the expression of symbolic or material support by neighbouring countries, and an international climate favourable to their cause; and, (3) a social network that can be transformed into a political organisation by language leaders, generating symbolic and material support and acting as a vigilante organisation to punish recalcitrant members of the language group. The capacity of the state to manage the potential discontent can be enhanced through constitutional amendments, particularly leading to the creation of mono-lingual regional units. But these need to be reinforced with some form of protection for minorities, such as institutional innovation like setting up commissions to enquire into grievances and act upon recommendations; and, a fair and effective electoral process which can accommodate local and regional elites within the national agenda; cross-cutting political parties, coalitions which can provide informal avenues of the accommodation of minority languages; the co-option of opposition leaders by the regional and national governments; and, finally effective and fair management of law and order which make revolts costly, and as such, enhances the value of negotiation. The Indian and Swiss examples hold general lessons for all states facing the challenge of linguistic diversity. Though the Swiss federal constitution, unlike the Indian constitution, does not explicitly guarantee group rights to language, a solution of territorial protection for languages is available through the flexibility in the creation of mono-lingual cantons as we have seen in the case of the creation of the French mono-lingual canton of Jura. With remarkable prescience, the framers of the Indian constitution have equipped the Indian state to respond to the demands for linguistic autonomy through the double mechanism of individual and group rights. During the first phase of India’s constitutional development, some of these instruments were useful by empowering political majorities below the level of the national state through the effective enactment of provincial administrations. The second phase of constitutional development through the States Reorganisation of 1957 which created linguistically homogeneous States and counter-balanced the potential chauvinism through the promotion of the three language formula, requiring the use of Hindi, English and the regional language, made it possible to institutionalise the multi-cultural nature of the Indian state. One can see the wisdom behind the Indian policy when one contrasts this with parallel developments in Pakistan and Sri Lanka during the same period, with the former unwilling to accommodate the Bengali language movement in East Pakistan leading to the violent end of the federation, and the latter, thanks to the “Sinhala only policy” that virtually reduced the Tamil minority to second class citizenship, eventually sliding into the civil war in which the country is engulfed today. These normative developments of the federal principle and their adaptation to India’s cultural and historical context has been ably complemented by the democratic political process. During the critical years of transition from British rule and the consolidation of popular democracy in India, the Congress party provided the link between the modern state and the traditional society. Congress rule both at the centre and in the States provided informal channels of communication and the balancing of national, regional and sectional interests. The politics of coalitions that has replaced Congress hegemony has given a public articulation to the process of the integration of the Indian languages for the purpose of launching a new debate on the nature of the nation and for identifying the variable boundaries of the nation and region. In consequence, looking for regional allies has now become an imperative for all national parties which has greatly reinforced the munti-lingual character of Indian politics. The constitutional, legislative and policy instruments that India has drawn upon to reach the positive outcomes with regard to the development of federalism have an important implication for the comparative analysis of federalism. The empirical analysis undertaken here suggests some general inferences about the potential for the federal solution to conflicts arising out of linguistic diversity as well as its limits. In the first place, while the federal principle works on the basis of the recognition of separate identities born out of language and their empowerment through legislative, financial and moral empowerment, there are practical limits to the process. Linguistic groups that are territorially dispersed over the national state, or, that are very small, might find it difficult to form or sustain a separate political unit. In this case, accommodation of the interests of the specific linguistic group at the national level through a consociational arrangement rather than a new federal unit may be preferable. In the second place, a national linguistic minority group might have in its midst smaller groups, who in the event the minority group is recognised as a regional unit would constitute a new, persecuted minority group, whose sense of deprivation would thus be magnified by the very act of conceding the demands of the original claimant to the status of a regional group. This was the case with the Bengalis in Assam after the States’ Reorganisation of 1957, or, of Gorkhas in West Bengal in recent times. The Indian solution of creating “sub-States” where such minorities are granted some cultural rights while the national state carefully monitors the implications for “new” minorities provides a workable solution. (Mitra and Lewis, 1996). Finally, by opening up the Pandora’s Box of pent-up demands for the expression of linguistic identity, the process of constitutional engineering might endanger the unity and security of the state and create a backlash that would endanger the long-established and legitimate demands for linguistic diversity. Multi-lingual federations need, therefore, to safe-guard national interests, inter-group communication and national integration through an effective national language. This calls for careful institutional safeguards to ensure that no group feels privileged or discriminated in the process. The Indian “three-language formula” which permits the simultaneous use of Hindi (national language), English (link language) and the regional language is an example of a relatively successful compromise. Bibliography Harrison, S. (1960) India: the Most Dangerous Decades, Princeton: Princeton University Press. Lijphart, A. (1996) “The Puzzle of Indian Democracy: A Consociational Interpretation”, in American Political Science Review 90(2): 258-268. Mitra, S. K. (1996) “India”, in G. Almond and G. Bingham Powell Jr. (eds.) Comparative Politics Today, 6th Edition, New York: HarperCollins. Mitra, S. K. and Alison Lewis (eds.), (1996), Subnational Movements in South Asia, Boulder/Colorado: Westview. Steinberg, Jonathan (1996) Why Switzerland?, Cambridge: Cambridge University Press. Watts, R. L. (1981) “Federalism, regionalism and political integration”, in D. Cameron (ed.) Regionalism and Supranationalism, pp. 3-19, Montreal: Institute for Research on Public Policy. Watts, R. L. (1998) “Federalism, Federal Political Systems and Federations » in Annual Review of Political Science, (1): 117-137. Wheare, K. C. (1964) Federal Government, 4th ed., New York: Oxford University Press. Forum of Federations / Forum des fédérations forum@forumfed.org