Jump to ContentJump to Main Navigation
Containing Nationalism$

Michael Hechter

Print publication date: 2001

Print ISBN-13: 9780199247516

Published to Oxford Scholarship Online: November 2003

DOI: 10.1093/019924751X.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2015. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 30 May 2016

Other Types of Nationalism

Other Types of Nationalism

(p.70) Chapter 5 Other Types of Nationalism
Containing Nationalism

Michael Hechter (Contributor Webpage)

Oxford University Press

Abstract and Keywords

A discussion of peripheral and unification nationalisms. Peripheral nationalism is directly spurred by the onset of direct rule: increasing political centralization threatens local leaders and provides an incentive for them to mobilize nationalist opposition to central authorities. Unification nationalism is a more indirect by‐product of direct rule. Since states adopting direct rule attain geopolitical advantages, the rulers of culturally homogeneous but politically divided territories have an incentive to merge their separate territories into a single political unit for defensive purposes.

Keywords:   Germany, Ireland, Italy, Napoleonic invasions, Norway, periphery, unification

Since state‐building nationalism arises to fill the gap left by the erosion of indirect rule, it is the first type of nationalism to emerge. The other types of nationalism are reactions to, or by‐products of, direct rule. If so, then the onset of direct rule in multinational polities should account for temporal variations in these types of nationalism. To that end, this chapter first discusses the timing and the (rather limited) success of the most prevalent type of nationalism, the peripheral variety. Then it discusses the conditions under which irredentist and unification nationalism occur. The chapter concludes by suggesting that the present analysis also sheds some light on reasons for the varying inclusiveness of nationalist strategies.

Peripheral Nationalism

Peripheral nationalism seeks to bring about national selfdetermination by separating the nation from its host state. As Chapter 3 argues, previous to the late eighteenth century culturally distinct territories located in imperial states were ruled indirectly. This provided a local rather than a central target for political demands. The growth of direct rule fundamentally changes this situation, however. By wresting power and responsibility from local authorities, direct rule increases the population's dependence on the centre. Subsequently, decisions about the administration of the (p.71) periphery are decided in a larger political arena, one in which the interests of peripheral residents are destined to play a minor role. As a result, the centre increasingly becomes the focus of collective action by the periphery. Local authorities have no intention of standing idly by while their prerogatives are stripped from them. Often they instigate peripheral nationalism, or at least lend their support to it. If the growth of direct rule motivates peripheral élites to embrace nationalism, the emergence of a cultural division of labour—often a concomitant of direct rule—does likewise for peripheral masses (Chapter 6).

As there are far more peripheral regions within states than states themselves, peripheral nationalism is by far the most prevalent form of nationalism. Although the bulk of scholarly attention devoted to peripheral nationalism has focused on highly developed societies, such as Britain, Canada, Belgium, and Spain (Tiryakian and Rogowski 1985), it is found in societies at all levels of development, and in the East as well as the West.1

In a previous book, I analysed the sources of peripheral nationalism in the British Isles from 1536 to 1966 (Hechter 1999; see also Canny 1998; Ohlmeyer 1998). Here, the analysis is extended in two directions. First, I explore whether, as the present theory contends, peripheral nationalism follows from the imposition of direct rule in multinational states. Then I discuss some of the factors responsible for the infrequent success of these movements.

The Timing of Peripheral Nationalism in the Ottoman Empire

The Ottoman Empire ranks among the greatest multinational states of all time, reigning over much of Eastern Europe, the Middle East, and parts of North Africa from the fourteenth to the early twentieth century. Like its counterparts, the Empire was a virtual tower of Babel, composed of a bewildering array of linguistic, religious, and ethnic groups. Despite its extreme multinationality, however, the Empire experienced little if any peripheral nationalism prior to the nineteenth century. This absence of peripheral nationalism was the accomplishment of a governance structure based on the timar and millet systems that the Ottomans had devised beginning in the fifteenth century.

(p.72) The timar system was designed to acquire and control territory. Under it, conquered lands were divided into estates (timars) and distributed as fiefs to military officers who had served the state well. The timar system provided both the impetus and the reward for territorial expansion. Once in place, the timar‐holders were obligated to raise a military retinue for service to the centre. The system also provided the economic foundation of the Empire: timar‐holders supervised agrarian cultivation, levied taxes, and passed revenue to the centre. Although the timar system relied on indirect rule, the centre attempted to control the timar‐holders by regularly rotating them from one territory to another, thereby undermining their capacity to develop an independent power base among the local peasantry (Barkey 1994; Barkey and Parikh 1991).

The millet system was a non‐territorial form of indirect rule that was designed to cope with the Empire's cultural diversity.2 It allowed subject peoples to retain governance over their cultural practices and local administration. Millets kept the records of birth, death, marriage, and wills, and were also responsible for education. They had the right to levy taxes on their members and adjudicated disputes between them. These members, in turn, had no direct relations with the Ottoman administration. The head of the millet, the millet‐bashy, was chosen by his community and confirmed by the sultan; his role resembled that of an ambassador (Laponce 1960). In return, the millets had to declare their allegiance to the state and pay annual tribute to it (Karpat 1973).

This two‐tiered form of governance functioned well only as long as the centre could maintain control over the timar‐holders (central control over the leaders of the millets had always been tenuous). For a variety of reasons, its ability to do so began to diminish in the sixteenth century. Increasing provincial autonomy sapped the centre's resources. To attain greater control over its wayward provinces, the centre instituted direct rule. At this point peripheral nationalism made its appearance in many parts of the Ottoman Empire. It is the timing of these nationalist movements that is of concern here. Peripheral nationalism arose first in Serbia, then Greece, Rumelia, Bulgaria, Albania, and, last, in the Arab provinces. Does the imposition of direct rule account for this temporal sequence?

(p.73) Beginning in the late sixteenth century, many areas of the Empire experienced rapid demographic growth with only a small corresponding increase in cultivated land (Goldstone 1991; Inalcik 1970; Shaw 1976). The resulting population pressures pushed many peasants into cities in search of work. An influx of cheap American silver destabilized the Ottoman economy by causing high inflation (Tachau 1984). After the development of modern artillery, the centre replaced the timar cavalrymen and invested in a more modernized army employing janissaries, an élite corps of soldiers known for their military prowess. This shift in military techno‐logy imposed greater costs on the centre.

Soon new social groups eclipsed the timar‐holders. In regions of the Empire closest to Western Europe, local commercial and intellectual notables came to the fore. Elsewhere ayans, a new semi‐feudal aristocracy of landed élites, arose. The ayans became the primary holders of tax farms and arrogated to themselves much local political authority (Davison 1963: 17; Karpat 1968: 71–8; Inalcik 1995: 126; Heper 1985: 30–2; Keyder 1997: 31). The upshot is that both groups were far less dependent on the centre than the timar‐holders had been.

The centre attempted to redress provincial autonomy by instituting direct rule in the late eighteenth century. The road was a rocky one, however. A coalition of janissaries and local authorities revolted in 1807, successfully deposing Sultan Selim III (Inalcik 1995: 130). Further, the leaders of the millets, which had remained intact during the centuries of state decline, became more autonomous as they participated in commerce. They began to organize to resist the increasing incursions by the centre.

The centre first attempted to impose direct rule in its western provinces. This, I argue, is why the Balkan regions were first to develop peripheral nationalism.3 Serbia and Greece were the territories most closely tied to the developing world economy. The opportunity to exchange with the rapidly developing societies in the West enabled local Serbian and Greek authorities to be more autonomous of Constantinople—thus more capable of resisting direct rule—than any others in the Empire. Their autonomy was also fostered by the Western powers' sympathy for the rights of Christian subjects living in what, after all, was a Muslim empire.

(p.74) Serbia was the first Balkan province to develop a strong national identity.4 After enjoying considerable autonomy under a benevolent governor of Belgrade, it was besieged by the centre's crack troops, the janissaries, who killed the governor and many village chiefs. Local notables fought back in the first Serbian insurrection in 1804 (MacKenzie 1996: 210). What began as a series of revolts by local leaders against the janissaries soon acquired the larger goal of national independence as local notables sought to secede from the Empire. In 1813 Turkish armies invaded Serbia in an attempt bring it to heel once more. Nationalist insurrections followed and in 1833, at the behest of Russia, Serbia was granted its autonomy, but it remained within the Empire. Independence was only acquired following the defeat of the Ottomans in the Russo‐Turkish War (1878).

Although it was the most economically developed—hence, economically autonomous—province in the Empire, Greece was the second Balkan province to become nationalist. The Greeks possessed their own administrative, political, and even military institutions (Braude and Lewis 1982: 19). Preserved by the millet system, Greek culture had remained vibrant through the years of Ottoman rule. Had Serbia not faced the wrath of the janissaries, Greece probably would have been the first province to develop peripheral nationalism (Gewehr 1967: 16). The immediate cause of the Greek war of independence (1821–7) was Mahmud II's attempt to impose direct rule.

The central government undertook the Tanzimat reforms (1839–76), an unprecedented attempt to institute direct rule, to respond to events in Serbia and Greece, and to halt the Empire's decline (Davison 1963; Hale 1994; Karpat 1972; Shaw 1968, 1976). These were designed to achieve political centralization, the improvement of social and economic conditions, and the promotion of a sense of ‘Ottomanism’ among all the peoples of the Empire. Naturally, these goals could only be attained by doing away with many of the privileges of local notables and leaders of the millets.

Thus began the next round of peripheral nationalism in the remaining Balkan provinces under central control. Uprisings in Nish and Vidin, two Rumelian territories near Serbia, occurred in (p.75) 1841 and 1850 respectively, the result of upheavals brought about by Tanzimat tax reforms (Inalcik 1976: 18–33). Hostilities in Rumelia took on a religio‐national form as Christians attacked Muslims, attempting to drive them out of the province. Bulgaria also witnessed the birth of peripheral nationalism in this period: anti‐centre revolts occurred in 1835, 1841–2, and 1850.

The tardiness of Bulgarian nationalism, relative to that of Serbia and Greece, was due to a number of factors. Since Bulgaria was the nearest of all the Balkan provinces to Constantinople, it was more remote from Western third parties. Further, the province was more religiously heterogeneous than Serbia and Greece: Muslims constituted a large proportion of the population. Religious heterogeneity impeded territorial solidarity. Like many other provinces, Bulgaria did not manage to secede from the Empire on its own steam. Independence for Bulgaria, Serbia, and Romania only came with the defeat of the Ottomans in the Russo‐Turkish War (1877–8) and the ensuing Treaty of Berlin (1878).

Albania was the last of the Balkan provinces to develop nationalism and the last to secede. It lacked many of the conditions that fostered peripheral nationalism elsewhere (White 1937: 106). Most of its tribes enjoyed almost complete autonomy from central influence and its population was very culturally heterogeneous. Although the Albanian League was established in 1878 to spread Albanian language and literature and raise a national militia, it was banned in 1886. Only after the rise to power of the Young Turks in 1908 was direct rule imposed on Albania in earnest. Attempts to disarm the population in the north in 1910 led to nationalist insurrections. Albania declared independence in 1912.5

Although the Arab provinces were least affected by the centralizing policies of Selim III and Mahmud II,6 they did not escape the Tanzimat reforms altogether. Syria and Lebanon were specifically targeted for conscription, disarmament, and direct taxation. Yet far from producing peripheral nationalism, these reforms instead led to conflict between Christian Maronites and Muslim Druzes (Maoz 1982: 91–2). During the nearly two decades of Ottoman direct rule, civil and cultural unrest escalated, resulting in the widespread massacres of 1861. Sultan Abdülhamid (1876–1908) took care to advance the political and economic interests of local Arab (p.76) notables, who were given important positions in the imperial civil service and military.

The rise of the Young Turks—the Committee of Union and Progress (1908–18)—turned the Arabs' attention toward Istanbul by replacing local notables in key provincial posts, and imposing the Turkish language in government schools, the judicial system, and local administration. Not surprisingly, this met with strong resistance among local Arab notables. Although many of the reforms instituted by the Young Turks were strongly centralist, some—such as the introduction of mass politics, the expansion of educational opportunity, and the development of a more liberal press—facilitated the expression of nationalism (Kayali 1997: ch. 2). Consequently, the seeds of Arab nationalism developed from 1909 through 1918 (Karpat 1972; Keyder 1997; Safi 1992: 347). Ottoman territorial losses during the Balkan Wars and the First World War convinced Arab notables (most notably, Sharif Husayn of the Hijaz) to ‘pursue opportunities other than those emanating from a close indentitifaction with Istanbul that would enhance [their] personal power and prestige’ (Kayali 1997: 172).

All told, the sequence of peripheral nationalism in the Ottoman Empire follows the timing of the imposition of direct rule. Direct rule was first imposed in the western provinces, and last in the eastern ones. Wherever there was sufficient cultural homogeneity to foster territorial solidarity, peripheral nationalism often followed suit.

Yet the relationship between direct rule and peripheral nationalism is not instantaneous. This is well illustrated by the development of Kosovar nationalism in the 1990s. A resource‐rich mountainous region settled by the Slavs, Kosovo became part of Serbia in the twelfth century. Following Serbia's defeat at the hands of the Ottomans in 1389, Kosovo remained under Ottoman rule until 1913. After the First World War, it was incorporated in Yugoslavia. The predominantly Muslim and Albanian‐speaking region became an autonomous region within Serbia in 1945.

During all this time Kosovo was ruled indirectly. Although it remained formally part of Serbia, for all practical purposes Kosovo became self‐governing after 1974. It even had its own representatives in the federal parliament (Hayden 1998). In 1989, however, (p.77) Serbia imposed direct rule on Kosovo. This immediately led to the establishment of a non‐violent opposition (The Albanian Democratic League of Kosovo, the LDK). In 1992, the Kosovo Liberation Army (UCK) was formed to fight for independence and closer affiliation with Albania. It carried out its first armed attack in 1993, but regular and sustained attacks against the centre were not mounted until mid‐1997. By March 1998, the Kosovar nationalists had overrun more than a dozen police stations, looting them of scores of automatic weapons. Serbian police patrols and checkpoints were targeted, and more than fifty Serbian policemen and officials were assassinated. In the most brazen attack to date, the nationalists brazenly stepped into the road near the village of Srbica at midday and killed a local Serbian official on 23 January 1998. The central government response was to conduct brutal military sweeps in territories of nationalist strength. In March 1998, for example, Serbian police, armed with helicopter gunships and armoured personnel carriers, tortured and killed twenty‐four people in the village of Likosane. A mass funeral held near the village was attended by 40,000 mourners. As the spiral of violence escalated, the Serbian government responded by instituting a new round of ethnic cleansing against the Kosovar Albanians. At the present time (May 1999), the North Atlantic Treaty Organization is waging an air war on Serbia to reinstitute autonomy (read: indirect rule)—but not sovereignty—in Kosovo.

The story is familiar.7 As the present analysis suggests, Kosovar nationalism arose just after the imposition of direct rule. Why then did it take eight years to become noticeable to the world at large? The collective action exemplified by the establishment of Kosovar nationalist groups is never instantaneous. A certain amount of time must pass to persuade recruits that the new regime is here to stay. Beyond this, both resources and time must be invested to build effective political organizations. The nationalists have erected their own parallel government, which levies taxes, as well as having shadow institutions—like a school system—that operate from private homes (Hedges 1998b). Beyond this, establishing the underground paramilitary organization required to carry out violent attacks is particularly costly and time‐consuming. Under conditions of secrecy, recruits have to be attracted, trained, and (p.78) provisioned. Weapons have to be acquired—usually from abroad —before paramilitary campaigns can be undertaken to steal more weapons from central authorities.8 The temporal lag between the imposition of direct rule and the emergence of visible nationalism therefore is due, at least in part, to the time intensity that is inherent in organizing militant collective action. Of course, if free riding cannot be prevented, then collective action may never arise at all.

Pure Secession in Norway and Ireland

The distinguishing characteristic of peripheral nationalism is the demand for secession. Secession is the formal withdrawal from a central political authority by a member unit on the basis of a claim to independent sovereign status.9 Many such demands come to nothing: during the summer of 1997, for example, quixotic secessionist actions were mounted in two self‐conscious regions—Texas in the United States, and Veneto in Italy.

If there is one constant in history apart from the imposition of taxes, it is the reluctance of states to part with contiguous territory. Land is the pillar of the state; it provides tax revenue, a labour force, mineral and other geographically based resources, and it is vital for defence. Beyond this, the permanence of state borders is among the most tenaciously held givens in political culture (Lustick 1993: 57–80). For these reasons, it is axiomatic that when rulers contemplate changing state borders, their fantasies are expansionist rather than contractionist. This is why most secessions are the result of fragmentation: they are either imposed on unwilling rulers who are defeated in war, or are the by‐product of endemic imperial weakness or collapse (Posen 1993b).10 Ottoman peripheral nationalisms are all of this type.

Fragmentation is qualitatively different from pure secession, which only occurs when a highly effective state permits a secessionist territory to withdraw from its embrace. The members of nationalist groups located in highly effective states are not likely to regard secession is a realizable political goal. Their perception is, however, extremely sensitive to signs of disarray in the centre. If central authorities signal their willingness to renegotiate the (p.79) terms of exchange with its peripheries—as the Soviet Union did in its waning days by adopting glasnost′—then secession suddenly enters the realm of the feasible.11 Pure secession is among the rarest of political outcomes. It has only occurred twice in the twentieth century: when Norway left Sweden in 1905, and when Ireland left the United Kingdom in 1922.12

Why, then, does pure secession ever occur? To say that rulers will cede territory only when the net benefit of doing so is positive is a truism.13 An answer to this question requires the specification of factors that are most likely to enter into the rulers' political calculations. Even if they are faced with a highly mobilized secessionist movement, central rulers still have a variety of possible reactions that stop short of secession.

They can offer secessionist leaders incentives by rewarding them with good jobs contingent on abandoning their support of secession. The regime also may promise, and perhaps even provide, other kinds of resources (such as development projects) to the region so as to persuade rank‐and‐file members of the secessionist movement that their core interests lie with maintaining a tie to the host state. In similar fashion, they can threaten economic non‐cooperation in the event that secession occurs. Although such a threat is not likely to be regarded as highly credible, it nevertheless increases the costs of secession to potential supporters (Young 1994).

Constitutional reforms are a more costly response to secessionism. They can help forestall secession either by providing federation, devolution, and other institutional changes that effectively vest greater decision‐making authority within the nation (as is ongoing with respect to Quebec), or by instituting administrative redistricting that inhibits the secessionists' potential for collective action, as occurred in Nigeria and India (Wood 1981).

If all else fails, repression is likely to be the state's last resort. Its efficacy, however, depends on at least three kinds of factors. First, geographic: the larger and more mountainous the secessionists' territory, the more difficult it is for the centre to control it militarily. Second, military: the military capacity of the host state relative to that of the secessionist territory depends not only on the secessionists' matériel and training (Do they have an army? If so, how (p.80) effective is it?), but also on each population's morale, which directly affects conscription rates and soldiers' combat performance. For example, it is far more difficult to conscript an army in a population that has recently experienced high rates of military casualties than one that has enjoyed an extended period of peace. Last, geopolitical: how are third parties in the international system likely to react to the host state's repressive moves?14

For the most part, geopolitical factors tend to militate against secession. In the first place, as the origins of most current states are multinational, almost all central rulers can imagine that they too may some day face potential secessionist movements. Supporting secessionist movements elsewhere therefore might help stir up unpleasant problems at home. This provides most central rulers with a reason to collude by universally discouraging secession. Further, support for a secessionist movement necessarily comes at the expense of relations with its host state.15 Given geopolitical realities, the net benefit to be gained by courting secessionists tends to be negative.16 Sometimes international support helps sustain a state built on shaky foundations (as has been argued with respect to Tito's Yugoslavia (Omrĉanin 1976).

Several of these factors come into play in accounting for the unusual success of Norwegian and Irish secession. Norway initially became subject to Swedish rule as the result of an 1813 agreement among the United Kingdom, Russia, and Sweden. Sweden was promised Norway as compensation for its loss of Finland to Russia if it helped the allies defeat Napoleon. Pursuant to the agreement, Denmark handed over the governance to Sweden one year later. Sweden's interest in Norway initially was based on its fear of Russian expansion and was stimulated by Russia's aggressiveness in Finland (Derry 1979: 272). Separatist sentiment in Norway grew during Swedish rule, however, and by 1905 it had become almost universal: a referendum on Norwegian secession passed by 368,208 to 184, with a turnout of 84 per cent of the electorate (Omrĉanin 1976: 10). The referendum results left Swedish rulers with a quandary. Owing to Norway's physical size and topography and to Sweden's small population, military repression was an unattractive option. Swedish defences would be more vulnerable if most of its military resources were expended occupying (p.81) a hostile Norway than if Sweden had a friendly Norway that was responsible for its own security. Occupying Norway would be difficult, not least because, during the years immediately prior to independence, Norway had strengthened its military by buying new field artillery and building forts on the Swedish frontier (Derry 1979: 270).

Further, Sweden's domination of a hostile Norway was an obstacle to Scandinavian regional cooperation. By consenting to Norway's independence, this obstacle was removed and Scandinavian leaders then were able to cooperate as equal partners in establishing the Nordic Interparliamentary Union (1907) and subsequent international institutions that benefited the entire Nordic region (Hancock 1972: 255–6). Last, Sweden's decision may have been influenced by international pressure. Norway had made successful efforts to attain foreign support, and its decision to establish a monarchy may have allayed fears about its future instability (Larsen 1948: 273).

Ireland's incorporation in Great Britain lasted from 1801 to 1922. As in Norway, Irish support for nationalism grew during this period—especially after the Easter Rebellion in 1916—and the nationalists also developed a military capacity. The British state had many military advantages in Ireland relative to that of Sweden in Norway, however. The size and geography of Ireland made occupation feasible, and the British already had been engaged in the occupation of Ireland during much of the late nineteenth century. In addition, British military resources far exceeded Sweden's. Why then did Britain accede to the secessionists' demands?

After the end of the First World War, British public opinion did not support a war in Ireland (Hachey 1973: 23–5). Continuing violence there was undermining the war‐weary public's support for the Liberal government (MacDonagh 1968: 88). The government recognized that an utter lack of public support precluded an all‐out offensive or a prolonged occupation (Beckett 1966).17 In addition, many prominent people—including the king of England and the archbishop of Canterbury—objected to the government's Irish policy and this helped set the stage for negotiations in June 1921. International opinion, particularly in the United States, also inclined the government to concession.

(p.82) Finally, Britain retained some control over the new Irish state and kept total control over Northern Ireland. Although the Irish nationalists sought an independent unpartitioned republic, they were forced to accept a compromise. Ireland was given status as a dominion in the British Empire, rather than outright independence. And Ireland had to give up Ulster, the six counties of Northern Ireland. Ireland remained subject to Britain in many ways: Britain kept control of certain Irish harbours that had been essential to its survival during the war (Beckett 1966: 453; MacDonagh 1968: 91). Britain retained ultimate appellate jurisdiction of court cases (MacDonagh 1968: 91), and also reserved the right to approve the new Irish constitution (Macardle 1965: 818).

Thus the ultimate settlement between Britain and Ireland was a compromise. Britain protected itself from the costs of Irish secession by maintaining close ties. The significance of those ties is revealed by the events that followed the signing of the treaty. Violence broke out in Ireland between those who still wanted a fully independent republic and those who accepted the treaty's terms. To some Irish, the treaty perpetuated Ireland's colonial subjection. Of course, the ties between Ireland and Britain—like those involving other Commonwealth countries—have changed dramatically over time.

Recent events in the Balkans and Eastern Europe have underscored the historical contingency of the role of geopolitical factors in state formation (Heraclides 1990). Geopolitical forces often conspire to promote the fragmentation of states (Lapidus and Zaslavsky 1992). During the post‐Yalta détente between the United States and the Soviet Union, non‐intervention in internal politics was a guiding principle of international relations (for reasons spelt out by Wallerstein 1991). This principle tended to deny secessionists any hope of third‐party intervention on their behalf. Now that the American–Soviet détente is no longer operational, third‐party intervention has been on the rise. The most important contemporary instance was Germany's decision to recognize the sovereignty of Slovenia and Croatia. This move, which was ultimately responsible for recognition of these new states by the European Union and a reluctant United States, would have been unthinkable during the cold war.

(p.83) Most of the new states formed since the break‐up of the Soviet Union and the Warsaw Pact have been the product of fragmentation rather than secession. It has long been appreciated that secession is a highly improbable outcome (Young 1976: 460–504). Why is it so improbable and why it will continue to be so in the future? Since many states are multinational, and many of these contain territorially concentrated national groups, on statistical grounds alone secession should be relatively common. Yet few regions have the structural requirements for the development of a secessionist movement. Not only do these include a national population disproportionately clustered in its ‘own’ territory which has ample intragroup communications capacity—for this requirement is met by many multinational states in the world. Beyond this, the nation must be located in a state that allows considerable freedom of association. Given a demand for sovereignty, the nation must develop political organizations capable of mobilizing the nationals.18 This problem is far more severe in the case of secession than in the case of many other political agendas, because even in the best of circumstances, most people appreciate that host states will tenaciously hold on to their territory. Further, even when mobilization solutions are at hand, it is not even clear that secession is an outcome that most self‐identified secessionists themselves sincerely desire.

The final reason for the improbability of secession is that the host state is far from powerless to protect its territorial integrity. If the secessionist territory is large and/or mountainous, if the regime has too few resources to provide the necessary incentives to coopt secessionist leaders, and if potential constitutional reforms are either too difficult to enact or likely to be regarded as insufficient by the territorial population, then and only then is the host state forced to contemplate acceding to secessionist demands.

Irredentist Nationalism

Irredentism—derived from the Italian irredenta, unredeemed—was first used to describe nineteenth‐century Italian movements to (p.84) annex Italian‐speaking areas under Austrian and Swiss rule. It now refers to any effort to unite national segments of a population in adjacent countries within a common polity (Chazan 1991a). Irredentist nationalism is the least prevalent form of nationalism; for that reason it can dealt with briefly. If secession involves subtracting a national territory from a state, irredentism involves subtracting the territory from one state and adding it to another. Given the obvious difficulties of this, irredentist movements most often are initiated by states interested in annexing territories having a large proportion of their co‐nationals. Nineteenth‐century Romania and Greece relied on irredentist nationalism advocated by the respective governments (Sugar 1969: 50–1). Hitler's annexation of the Sudetenland is perhaps the most transparent recent example.

Unlikely as pure secession is, successful irredentist nationalism is far less likely, even though many nations—especially in post‐colonial Asia, Africa, and the Middle East—are divided by state boundaries (Chazan 1991b; Horowitz 1992). In the last analysis, successful secessionist movements merely require the host state's consent to forego the given territory. Successful irredentist movements require not only that, but the consent of the putative beneficiary state as well. If host states are extremely reluctant to part with their territory, potential beneficiary states may also be leery to accept new territories.

The emergence of an irredentist cause in a given nation is directly related to the receptiveness of the beneficiary state to the prospect (Horowitz 1992: 122–3). Potential beneficiary states tend to waver in their support of irredentist movements. Thus, in 1975 Iran abruptly halted its military assistance to the Iraqi Kurds and eventually closed their borders to them, effectively dooming the movement. In 1987, India ceased its support for the Tamil secessionists in Sri Lanka, reaching an agreement with the Sri Lankan government for Tamil regional autonomy instead, and proceeded to join the Sri Lankan government in suppressing its armed rebels. Moreover, it is by no means certain that the national minority in the donor state will be recognized as kindred by members of the beneficiary state.19 Perhaps the key reason for the improbability of irredentism is that it is seldom in the interest of the nationalist movement's leadership. Whereas successful secession affords them (p.85) the prospect of gaining the reins of power in the government of the new state, irredentism merely provides them with the added competition of the established leadership pool in the beneficiary state. For all these reasons, nationalist groups are far more likely to espouse secession than irredentism.

Unification Nationalism

Unification nationalism is the obverse of state‐building nationalism. Whereas the latter seeks to create cultural homogeneity within the borders of an existing multinational state, the former aims to create an overarching state that supplants a number of smaller sovereign units in a (relatively) culturally homogeneous territory.20 What induces petty sovereigns to undertake collective action that results in a loss of their sovereignty? In general, they are only likely to do so when they perceive a threat to their rule.21 The principal historical examples are the unification of Germany and Italy, whereas Poland offers an instructive counterexample.22

Before 1860, Great Britain and France were the only two prominent countries in Europe to undertake state‐building nationalism.23 Although Spain may have looked like a nation state on the map, in reality it was a multinational empire. Portugal, the Netherlands, and the Scandinavian countries were the closest to being nation states, but they were small and had less geopolitical influence. The characteristic European polities of this era were of two kinds. They were either small states comprising fragments of a nation, like Hanover, Baden, Sardinia, Tuscany, the Two Sicilies, or they were vast multinational states like the Romanov, Habsburg, and Ottoman Empires. For example, the Holy Roman Empire (controlled by the Habsburg dynasty) was split up into 314 territories and towns and into 1,475 free lordships, all of which guarded the sovereign rights guaranteed to them by the European powers after the Peace of Westphalia. The Germans, according to the Imperial Privy Councillor Friedrich Carl von Moser (Von dem Deutschen Nationalgeist) writing in the year 1766: (p.86)

are a puzzle of a political constitution, a prey of our neighbours, an object of their scorns outstanding in the history of the world, disunited among ourselves, weak from our divisions, strong enough to harm ourselves, powerless to save ourselves, insensitive to the honour of our name, indifferent to the glory of our laws, envious of our rulers, distrusting one another, inconsistent about our principles, coercive about enforcing them, a great but also a despised people; a potentially happy but actually a very lamentable people (Schulze 1991: 43).

Whereas feudal restrictions on commercial activity and political participation had begun to crumble in Britain, the Netherlands, and France, these developments had hardly affected the territories of the Reich, save in a few northern trading towns. Much the same could be said for the situation in Italy (Riall 1994: 5). According to Giuseppe Mazzini, the ideologist of Italian unification, ‘The States into which Italy is divided today are not the creation of our own people; they are the result of the ambitions and calculations of princes or of local conquerors, and serve no purpose but to flatter the vanity of local aristocracies for which a narrower sphere than a great Country is necessary’ (Mazzini [1858] 1995: 95).

The impetus toward unification nationalism in Germany and Italy was an exogenous shock—the Napoleonic conquest. This conquest, which seemed to demonstrate the economic and military superiority of direct rule over alternative forms of governance, had important effects on the conquered territories. It consolidated the German and Italian polities, sharply reducing their numbers. And it increased the efficiency of their governance by establishing modern bureaucratic administration and law. In both the (extended) French Empire and all the dependent territories (including all of Germany and Italy in 1810), feudal authority was extinguished. Central governments were given authority over their individual subjects. The nobility lost its privileges in taxation, office‐holding, and military command. The Church was stripped of much of its power as well (Palmer and Colton 1965: ch. 10).

The Napoleonic reforms laid the groundwork for unification nationalism in Germany and Italy in two fundamental ways. In the first place, by threatening the social and political position of local (p.87) authorities the reforms gave prominent political actors a strong incentive to oppose Napoleon, as the extension of direct rule always does. The Napoleonic state spurred opposition among its non‐French subjects by raising taxes in dependent territories to subsidize taxpayers in France.24 Resistance to Napoleon united conservatives, who wished to return to the status quo ante, with liberals who wanted to preserve the liberties introduced by Napoleonic reforms, but insisted on greater self‐government. Such ideological heterogeneity is, of course, typical of nationalist movements. The basis of all nationalist movements is the desire for self‐government, but since they are nearly always coalitions of people whose other interests differ widely, the type of government that particular nationalists envision (whether liberal, conservative, or radical) is likely to be a matter of some dispute (see Chapter 1).

In the second place, the Napoleonic reforms fostered unification nationalism in German and Italy by sharply reducing the number of separate states in each territory, thereby substantially reducing the costs of collective action among them. For most of the eighteenth century, Germans had been the least nationally minded of all the larger European peoples:

They prided themselves on their world‐citizenship or cosmopolitan outlook. Looking out from the tiny states in which they lived, they were conscious of Europe, conscious of other countries, but hardly conscious of Germany. The Holy Roman Empire was a shadow. The German world had no tan‐gible frontiers; the area of German speech simply faded out into Alsace or the Austrian Netherlands, or into Poland, Bohemia, or the upper Balkans. That ‘Germany’ ever did, thought, or hoped anything never crossed the German mind. There was scarcely even a developed language, for the speech habits of Berlin and Vienna were more different than those of London and New York today. The upper classes, becoming contemptuous of much that was German, took over French fashions, dress, etiquette, manners, ideas, and language, regarding them as an international norm of civilized living. Frederick the Great (of Prussia) hired French tax‐collectors and wrote his own books in French (Palmer and Colton 1965: 403).

Napoleon turned all the German states but one, Prussia, into political dependencies in the Confederation of the Rhine (Austria allied itself with Napoleon by marriage in 1810). Even though it (p.88) suffered military defeat resulting in the loss of its territory west of the Elbe, Prussia was the sole German state to hold out against Napoleon.25 Fifty years later, with a highly effective military, Prussia took the lead in German unification. While the Napoleonic conquest greatly simplified the German state system, German romanticism—itself a reaction against the influence of the French Enlightenment, the French Revolution, and Napoleonic rule—spawned nationalist treatises, like those of Herder and Fichte (Greenfeld 1992).26 The defeat of Napoleon and collapse of the continental system, however, reintroduced great political complexity in the German lands. From 1815 to 1866 there were thirty‐nine separate states in Germany.

There were significant parallels in Italy.27 Italy had long been divided into about six large states and a few very small ones. From 1796 to 1814, the French regime in Italy broke the back of the various duchies, oligarchic republics, papal states, and foreign dynasties by which Italy had long been ruled. In the end, a veritable patchwork of sovereignties had been reduced to only three parts. Thus French influence made the notion of a politically unified Italy a reasonable aspiration. And the nationalist ideology of Mazzini articulated this aspiration.28

Napoleon's simplification of the Italian political landscape was undone by the Congress of Vienna. And by 1859—the beginning of the Risorgimento—Italy was once more divided into a number of separate states. However, the actions of the Congress of Vienna had provided the German territory with a more effective overarching political institution—the German Confederation—than had previously existed in Italian territory.29 In the aftermath of the revolutions of 1848, therefore, Germany and Italy were politically divided territories. Both of these territories boasted home‐grown nationalist writings dating from the Napoleonic period. Yet the political impact of these ideals was marginal, for the advocates of nationalism had little effective political organization (Breuilly 1993: 98). Nationalist support was generally confined to gymnastic societies, student fraternities, and to a variety of groups devoted to singing and other cultural and literary pursuits (Düding 1987). Unification did not result from the activities of these groups.

(p.89) The increased effectiveness of the German Confederation was a more important stimulus to the development of a national opposition than the intellectual concerns with romantic nationalism (Breuilly 1993: 100). In taking a firm stand against Napoleonic constitutional reforms, the Confederation alienated liberal businessmen and officials advocating economic and political development. Further, Prussian government support for a customs union among German states helped provide an institutional base for other kinds of cooperation. The Prussian‐backed customs union created an economic ‘Germany’ in the eyes of the world and in itself encouraged a sense of economic nationalism.30

When Bismarck later rejected the idea of joint Prussian and Austrian sovereignty over Germany, he became committed to the idea of a territorially concentrated state to better compete with continental rivals. In a letter written in 1857 Bismarck wrote, ‘Situated in the centre of Europe we cannot afford that kind of passive incompetence which is happy to be left in peace. It will endanger us tomorrow as it did in 1805, and we shall be the anvil if we do nothing to be the hammer’ (cited in Stürmer 1987: 141). Further, Prussian‐led national unification was not the result of obvious popular demand in other German states: most German states were on the Austrian side in the Austro‐Prussian War of 1866. German national sentiment was more a by‐product of German unification than a cause of it.

Much the same could be said of Italian unification. Piedmont played an analogous role to that of Prussia, Cavour was Bismarck's counterpart, and Austria was a common enemy. Although Cavour lacked an institution like the German Confederation, which could have provided an institutional framework to press for political unification, Piedmont benefited from French support against Austria in its bid to extend its control over Northern Italy. As in Germany, there was little popular support for Italian unification.31 Garibaldi's successful invasion of Sicily was both unanticipated and unsupported by Cavour. Garibaldi's military success:

was due to the decrepit state of the Bourbon regime on the island. The speed of his invasion made it possible for both popular opposition and the propertied elements hostile to the regime to see in him an ally and a means of overthrowing the government. Only this general support could have enabled (p.90) the pathetically weak military force under Garibaldi's command to take over the island. The support had nothing to do with the nationalist cause, which was weaker in the south, particularly in Sicily, than in any other part of Italy. It was purely to do with the domestic problems of the Bourbon regime. Indeed, it is reported that when the cry of Viva Italia! was raised during Victor Emmanuel's entry into Naples some natives thought it must refer to his wife (Breuilly 1993: 113).

Even so, Italian unification failed to lead to an effective central state due to a poor communications infrastructure (Riall 1994: 59–60). Rome lacked economic resources relative to those controlled by local élites. As a result, antagonism to the centre persists in peripheral areas (it is worst in Sicily, but seemingly a problem in other regions, as well).32 The Italian example suggests that national unification is far from sufficient to produce effective direct rule.

In both Germany and Italy, various liberal élites pressed for national unification in order to create a modern state that would be recognized by the two European superpowers, Britain and France. As the story of nineteenth‐century Poland attests, however, not all unification nationalist movements manage to win the day. Poland too was split into a variety of separate states, many of which were controlled by foreign powers. As in Germany and Italy, a unification nationalist movement was established among the nobility in order to reassert their political position (Breuilly 1993: 116–18). However, the Polish nobility did not manage to organize militarily to advance their common interests. As a result, Poland never established a unified state that was capable of protecting its territory from its avaricious neighbours.33

Unification nationalism is a by‐product of direct rule, but it occurs via quite different mechanisms. It arises in territories subjected to the geopolitical threat represented by the emergence of direct‐rule states. It has little popular support and is a movement of élites. Because the élites seeking national unity in Germany and Italy were fragmented and without political authority, they looked to existing states for support. Like peripheral nationalism, part of its motivation is defensive: it is difficult to believe that Bismarck or Cavour would have been successful in a Europe without the challenge represented by the existence of other national states.34 (p.91) French state‐building nationalism therefore begat German and Italian unification nationalism as unintended by‐products. Instead of seeking autonomy by seceding from the modernizing state, unification nationalism aims to create a modern state by eradicating existing political boundaries and enlarging them to be congruent with the nation.35 In all of these cases, the impetus to political unity is the perception of an external threat. Once unification occurs, however, the mechanisms of state‐building nationalism encountered in Chapter 4 come into play. Foremost among these is indoctrination in the schools, particularly in the teaching of history.36

Inclusive versus Exclusive Nationalisms

This discussion ignores many other differences among nationalist movements. Chief amongst these are variations in the content of nationalist ideals. Why are some forms of nationalism (like the French) culturally inclusive, and others (like German) culturally exclusive? Max Weber tantalizingly broached the subject of accounting for strategic differences between tactics of monopolization and closure in social groups, but since then little research has focused squarely on the problem (see, however, the suggestive analysis of the dynamics of closure in the context of unionization in Friedman 1990). Recent research explaining differences in the citizenship policies of states offers further insight into the question. State‐building and unification nationalism differ in one important respect: the former is inherently territorially inclusive, whereas the latter is culturally inclusive.37 Thus post Revolutionary France—the pre‐eminent example of state‐building nationalism—has continually defined citizenship expansively and is assimilationist, whereas Germany—the locus classicus of unification nationalism—has defined citizenship as a community of descent based on cultural factors (Brubaker 1992).38

Part of this difference is accounted for by the timing of these two forms of nationalism. Historically, state‐building nationalism preceded unification nationalism. Unlike unification nationalism, (p.92) the initial impetus for state‐building nationalism was largely endogenous. Further, the process occurred in the absence of well‐articulated nationalist ideals: it is no accident that nationalist ideologues like Mazzini, Herder, and Fichte were neither British nor French. When peripheries like Cornwall and Burgundy were folded into the expanding Tudor and Capetian states, no concept of national self‐determination was readily available to help their élites mobilize peasant resistance.

The impetus for unification nationalism, however, was largely exogenous. The rulers of the many different states and principalities in Germany and Italy were induced to form a national state in order to defend their interests against the threat posed by the highly effective direct‐rule states of France and Great Britain. Nationalist ideals came into their own during this process. Ironically, these ideals came back to haunt both Britain and France: they were adopted in Wales, Brittany, and Corsica to resist further extensions of direct rule in the twentieth century. In this way Herder and Mazzini provided the shoulders for an Irish nationalist like James Connolly to stand on.


The rise of direct rule had major consequences for the development of peripheral nationalism, the most common type of nationalism, as well as for unification nationalism. Since direct rule invariably decreases the power of local authorities, its advent provides them with a strong incentive to resist (Tambiah 1996). The local authorities in culturally distinctive peripheries, therefore, are often tempted to play the peripheral nationalist card in their efforts to keep the centre at bay (Brass 1991).39 Direct rule can spur peripheral nationalism even in culturally distinct regions (like nineteenth‐century Wales and the Spanish Basque Country) whose traditional authorities have assimilated to the culture of the centre. In such regions it can afford political space to new kinds of leaders who attempt to mobilize the culturally distinct peasantry on nationalist grounds.

(p.93) This logic has a key implication. Variations in peripheral nationalism among the territories of an empire ought to be explained by the timing of direct rule. The development of peripheral nationalism in the various nations of the Ottoman Empire is consistent with this expectation: the sequence of nationalist mobilization is largely accounted for by the onset of direct rule in each territory.

Direct rule plays a somewhat more indirect role in the genesis of unification nationalism. When combined with state‐building nationalism, it fosters both economic and military efficiency.40 Unification nationalism then emerges as a reaction to the increased efficiency of direct‐rule states. ‘Greatness awaited peoples who could act together as brothers in harmony with one another and with their government, while weakness and ignominy threatened rulers so alienated from their people as to remain incapable of using unstinted popular support’ (McNeill 1986: 51). Thus the seeds of German and Italian unification are to be found in the Napoleonic conquests. Likewise, the origins of direct rule in Japan are defensive: its impetus can be traced to the Admiral Perry's sudden appearence in Edo Bay in 1853.41

Although the extension of direct rule in a multinational polity is a necessary condition of the two most important types of nationalism, it is far from a sufficient one. The first consequence of direct rule is likely to be state‐building nationalism, which aims to destroy the social basis of peripheral nationality. By rendering the boundaries of nation and governance unit congruent, its success erodes any potential for peripheral nationalism. It follows that peripheral nationalism can emerge only when the institutions of direct rule fail to assimilate the distinct nations within a state.

Assimilation is likely to occur only under certain kinds of conditions, however. Also needed, perhaps, is a political opening: the most significant determinants of serious nationalist conflict in the contemporary world are transitional political regimes undergoing democratization, and states that have weathered significant power shifts. Presumably these institutional changes offer new opportunities for political entrepreneurs to build constituencies, and therefore make all kinds of collective action more likely (Gurr 1994: 364).


(1.) It has long been customary to refer to ‘tribalism’ in Africa, and to talk of linguistic and religious conflict in the Middle East and the Indian subcontinent, but the outbreak of peripheral nationalism in western China came as a surprise to many commentators on contemporary politics. Terrorist bombs set in Urumuqi, the capital of Xinjiang, and rioting in the frontier city of Yining in early 1997 signalled the rise of peripheral nationalism among the Muslim Uighurs. Uighur nationalism was stimulated by the immigration of Han Chinese and the subsequent establishment of a hierarchical cultural division of labour there (Tyler 1997).

(2.) The precise nature of the millets, their age, and degree of institutionalization are all highly contested subjects (Grillo 1998: 86–96). For a discussion of analogous institutions designed to provide autonomy to non‐territorial cultural minorities, see Coakley (1994).

(3.) Actually, Crimea was the first province of the Empire to develop nationalism, successfully acquiring its independence in 1774. It was a very special case: Crimea had traditionally been an autonomous vassal state of the Empire and was valued for its strategic geographic position near Russia. Of all the Ottoman provinces, the Khanate (Crimea) maintained more of its independence and sovereignty than any other. ‘It was in fact the first self‐contained province to break away from the Empire at a time when “nationalism” and “national consciousness” had scarcely emerged (p.182) anywhere else’ (Fisher 1977: 59). In the century leading up to Crimea's separation from the Empire, the Ottomans increasingly interfered directly in Crimean politics, particularly influencing the deposition and installation of Khans. In their secessionist struggle, Crimean local authorities stressed their glorious past and their link to the Golden Horde.

(4.) The centre's interference in Serbia's cultural life began in 1766, when the Sultan abolished the Pec patriarchate and subordinated the Serbian church to the Greek patriarch of Constantinople. This intensified the Serbs' sense of national identity. ‘The Serbian church became even more of a people's church, spreading the national mythology that idealized memories of medieval Serbia through art, literature, and numerous saints’ (MacKenzie 1996: 207). The subjection of the Serbian church ensured that the local clergy would support an eventual Serbian nationalist movement (Ramet 1984).

(5.) Likewise, Armenian nationalism began during the latter decades of the Tanzimat era (Suny 1993). It was fuelled from 1894 to 1915 by a series of brutal campaigns that the Turkish government carried out against Armenians. Following Armenian support of Russia during the war, the Turkish authorities ordered a mass deportation of Armenians to Syria and Mesopotamia which led to more than a half‐million deaths. Turkish policies towards Armenians can only be understood if it is appreciated that Ottoman Muslims had themselves been subject to the same kind of brutal ethnic cleansing by Christians in the Balkans, Anatolia, and the Caucausus. For a reading of Armenian–Turkish conflict from the Muslim point of view, see McCarthy (1995).

(6.) One exception was the province of Iraq. In 1830 Sultan Mahmud II sent an envoy to Baghdad with the aim of replacing the local Mamluk notables with a Turkish governor responsible directly to the central government. The city of Karbala, in particular, attracted the attention of Mahmud, as its commercial development permitted it to become a virtually autonomous city‐state in the 1820s and 1830s. In reaction to Mahmud's incursions in Iraq, local notables rallied the population in revolt against the centre (Cole and Momen 1986).

(7.) It is instructive to compare the fate of Kosovo to that of the separatist province of Aceh in Indonesia (Robinson 1998). Like Kosovo, Aceh had attained a special, relatively autonomous, status from the central authorities in Jakarta. The discovery of substantial reserves of liquid natural gas focused Jakarta's attention on Aceh, however. Moves toward direct rule occurred in order to protect what had become one of Indonesia's greatest sources of foreign exchange. A small separatist movement (Aceh Merdaka) was founded to protest increasing direct rule, but until Jakarta launched a massive military effort to repress it the movement fared badly. At the time of writing, Aceh is the site of some of the greatest nationalist (p.183) violence in Indonesia. Christie (1996) provides an analysis of peripheral nationalism in a number of South‐East Asian states.

(8.) For example, the Tamil Tigers of Sri Lanka have obtained surface‐to‐air missiles from Cambodia, assault rifles from Afghanistan, mortar shells from the former Yugoslavia, and explosives from the Ukraine (Bonner 1998).

(9.) Like the previous definitions in this book, this one is analytical rather than normative. To some writers, however, the term secession has definite normative connotations. It is used to describe the severance of a presumptively permanent link between a peripheral territory and a central state, whereas ‘decolonization’ is often used if the link was presumed to have been temporary (Lustick 1993: 22–4). My use of secession conveys nothing whatever about the alleged permanence of the relationship between the territories in question.

(10.) That the sources of fragmentation are often exogenous to a given polity deserves to be emphasized. For a discussion of the fragmentation that beset Yugoslavia after the end of the cold war, see Ch. 8. Laitin (1998: 327–31) describes the fragmentation of the Soviet Union in light of current theories of international relations, and Bunce (1999) contends that institutional differences among fragmenting states help determine how violent the outcome will be. Finally, the government of Indonesia only began to countenance independence for East Timor (January 1999) after an economic crisis that previously caused the fall of the Suharto regime.

(11.) Thus the proliferation of secessionist movements in the Soviet Union followed glasnost′ in time and clearly was precipitated by it throughout Soviet territory (Beissinger 1996).

(12.) Not all peaceful secessions qualify as pure, however (Young 1995: chs. 10, 11). A handful of other 20th‐century political divorces fail to meet this standard. Iceland terminated its union from Denmark in 1944 when the metropole was under German occupation. Likewise, Bangladesh's separation from Pakistan in 1971 does not qualify, because the Pakistani state cannot be considered to have been highly effective at that time. Namibia's independence from South Africa in 1990 was won due to external pressure exerted by the United Nations and by Cuban military forces on the Angolan border. Shortly thereafter the South African regime collapsed. Slovakia's separation from the Czech Republic in 1993 is in key respects different from these other contemporary examples of secession. After the fall of the Soviet Union, the economies of the Czech and Slovak republics were largely incompatible: the latter's economy was highly dependent on non‐competitive heavy industrial production. Here too fragmentation was at work, for Slovakia's independence would have been inconceivable at the height of Soviet power. At the end of the day, the new Czech government willingly (p.184) acceded to Slovak independence rather than compromising its turn toward a market economy.

(13.) For more formal analyses of the bargaining situation between centre and periphery see Putnam (1988) and Tsebelis (1990).

(14.) As the Ottoman case reveals, these third‐party reactions can affect the prospective benefits for secessionists, as well. Thus when the USA made it known that an independent Quebec would neither receive special trade benefits nor be welcomed into North American Free Trade Agreement, this weakened popular support for secession.

(15.) This is one sense in which the analogy between secession and divorce holds (Buchanan 1991). When a third person unilaterally supports one of the parties to a divorce, this is tantamount to rejecting the other party.

(16.) Of course, the leaders of X are likely to support a secessionist movement in Y when they are at war with Y or otherwise committed to Y's destabilization. Thus, Iraq's invasion of Kuwait in 1991 probably increased the likelihood of American support for Iraq's Kurdish insurgents.

(17.) In June 1921, the British chief of the Imperial General Staff wrote, ‘Unless England was on our side we would fail, and if we failed we would break the army . . . Unless England was on our side . . . it would be madness to try and flatten out the rebels’ (Curtis 1936: 451).

(18.) Thus the variable institutional completeness of Soviet‐era nations helps explain variations in their responses to the fragmentation of the USSR (Treisman 1997).

(19.) Ossie/Wessie conflict in Germany is a poignant example of the problem. At the very time that western Germany is making highly visible moves to come to terms with its Nazi past, eastern Germany—facing high unemployment and poor economic prospects in the wake of unification—has become the breeding‐ground for violent neo‐Nazi youth groups. The targets of the neo‐Nazis are ‘leftists, foreigners and others who view themselves as apart from German norms’ (Cowell 1998).

(20.) I borrow the term ‘unification nationalism’ from Breuilly (1993: ch. 4). As discussed previously, the concept of cultural homogeneity is itself vague. It is clear, however, that the adjective ‘cultural’ entails more than language.

(21.) See Ch. 3 n. 12 below.

(22.) The emergence of the central state in a culturally homogeneous Japan may also constitute an instance of unification nationalism.

(23.) Neither had become fully national states, however. Britain contained not only England but an Ireland that would secede in 1921, as well as Scotland and Wales, both of which developed significant peripheral nationalist movements in the 20th century. In addition to its struggles with Germany over Alsace‐Lorraine, France encompassed Brittany, Occitania, and Corsica, all of which later developed peripheral nationalism.


(24.) ‘Napoleon considered Germany chiefly as a base for Imperial recruitment to his Grand Armée and as the object of financial and economic exploitation. The burdensome billetings and devastating marches through the country by foreign armies, the financial burdens, which Prussia in particular had to bear after the Peace of Tilsit, and which [caused] the impoverishment of various sections of the population, were followed by a tariff system which protected the French economy at the expense of the remaining European states, entailing price rises and economic collapse. Thus, the population's original indifference was transformed within a few years into hatred of the occupying power’ (Schulze 1991: 50–1). Conspiratorial anti‐French associations arose around the so‐called Club movement and among student fraternal groups. Beethoven originally named his Third (‘Eroica’) Symphony Sinfonia grande Napoleon Bonaparte. On learning that Napoleon had proclaimed himself emperor of Germany, Beethoven tore up its title‐page in wrath and changed the name of the symphony to Sinfonia eroica composta per festeggiare il souvenire d'un gran uomo (Heroic symphony, composed to celebrate the memory of a great man) (Taylor and Kerr 1954: 52).

(25.) The effects of Napoleon's continental blockade against Britain fostered economic prosperity in some German territories (protection against British imports stimulated textile production in Saxony), whereas it led to economic collapse in regions depending on exports to British markets (Kiesewetter 1987). That East Prussia, a monocultural grain export economy, suffered particularly badly as a result of the blockade no doubt contributed to its resistance to Napoleon.

(26.) Despite the patent insubstantiality of some of the claims in these nationalist treatises, their popular appeal cannot be denied. Thus in his ‘Speeches to the German Nation’ delivered in 1807–8 to large audiences in French‐occupied Berlin, Fichte argued that the German language was superior to its French counterpart. German was a natural language (Ursprache) which had grown from living roots, while French was a neo‐Latinate language that was superficially living but dead at its roots (Düding 1987: 23). Since language was the only possible basis of nationality in the German territories, Fichte's argument about the natural superiority of German legitimated the pursuit of German nationhood.

(27.) There were also significant differences. On the one hand, Italy's Restoration states were more vulnerable to international pressure than were Germany's—to a considerable extent, their destiny was decided by statesmen in Vienna, Paris, and London. On the other, the physical presence of the Pope in Italy and the Church's strong hostility to national unification had no German parallel (Riall 1994: 78–9).

(28.) ‘Napoleon reorganised the Italian and German lands along the lines of a reduced number of more rationally administered and territorially bounded states. The modern‐minded men who oversaw this (p.186) transformation sought to defend these new creations as best they could, sometimes collaborating with Napoleon and sometimes opposing him. Generally speaking these men and their political creations survived Napoleon's defeat. Within this new political framework ideas of German and Italian nationalism were formed . . . in ways which would have been inconceivable under the patchwork ancien regime order’ (Breuilly 1993: 99).

(29.) This difference in institutional arrangements in the two territories may have resulted from different combinations of geographic and political conditions. The German territories lay between two great powers, France and Russia, and were dominated by two strong states, Prussia and Austria. To minimize external attempts to stimulate political fragmentation within the territory, Prussia and Austria developed a formal institutional structure. Italy likewise faced a powerful France, but there the Habsburgs were virtually unchallenged and had no need to develop a formal institutional arrangement to exercise their control (Breuilly 1993: 100).

(30.) The customs union (Zollervein) was only one of three important pan‐German economic reforms enacted in the early 19th cent. The first reform (Gewerbefreiheit) extended the freedom to engage in trade and industry. The last development—and perhaps the one with the greatest effect on German nation‐building—was the construction of the German railway network. Although each of these factors contributed to the growth of a pan‐German economy, they had no necessary implications for German political unification. Further, as might be expected, political fragmentation severely inhibited the establishment of a single German customs union (Kiesewetter 1987). The union was acceded to in much of the territory ‘only when economic depression and empty exchequers made further resistance to Prussia impossible’ (Henderson 1984: 95).

(31.) Only about 2.5% of the inhabitants of unified Italy were fluent in Italian in 1870; Cavour declared his loyalty to Italy by saying ‘Je suis Italien avant tout’. When Garibaldi spoke to southern Italian peasants about ‘Italia’, they thought he was talking about his mistress (Gibson 1994: 179).

(32.) The Northern League has significant political support at the time of writing (spring 1999), and recently a small secessionist movement in Veneto was put down militarily.

(33.) Likewise, it is instructive to compare the modern German and Italian experiences with that of ancient Greece. Although the members of each polis recognized the members of other polities as Greek, nothing like unification nationalism ever developed there (Yack 1996: 203–4). The present analysis suggests that the absence of a Persian state based on direct rule was ultimately responsible for the failure of a Greek polity to (p.187) emerge in the ancient world. France, in particular, played such a role for Germany and Italy.

(34.) ‘It was undoubtedly the example of the west, of Great Britain and France, successful and flourishing because they were unified nations, that stimulated the amibitions of other peoples to become unified nations too. The period after 1815 was in Germany a time of rising agitation over the national question, in Italy of the Risorgimento, in eastern Europe of the Slavic Revival. The movement was led by intellectuals, who often found it necessary to instil in their compatriots the very idea of nationality itself . . . Since such ideas could not be fully realized without the overthrow of every government in Europe east of France, thoroughgoing nationalism was inherently revolutionary’ (Palmer and Colton 1965: 438).

(35.) The attempt to build a European Union is a contemporary example of unification nationalism; its fitful development reveals just how torturous a process this is. As was argued in Ch. 3, for further unification to proceed, the threat posed by commercial and political alliances between American states, and between American and Pacific Rim countries, must become more tangible to the member states of the European Union. In the absence of such external threats, there is an insufficient incentive for member states' leaders to forsake their sovereignty; cf. Streeck and Schmitter (1991: 148).

(36.) A telling example comes from a Bavarian teacher's manual at the turn of the century concerning the teaching of a poem celebrating a costly military victory in the war of 1870–1: ‘If during the treatment of such a piece the eyes of our boys do not flash, the cheeks do not glow, the heart does not pound, if the fist is not clenched with the thought: If only I could have ridden there!—then either the discussion or the delivery lacked the proper inspiring fires, or the boy is no good’ (Kennedy 1982: 262).

(37.) Friedrich Meinecke distinguished between state nations (Staatsnation) and cultural nations (Kulturnation). In state nations of the West, state‐ and nation‐building coincided, so the nation became defined through the political principles of constitution and citizenship. In Germany, where state‐building occurred belatedly, the nation became defined as a prepolitical community of culture, language, and ethnicity. This ethno‐cultural notion of nationhood became perpetuated over time and planted the seeds for the exclusivist conception of Germanness ever since (Forsythe 1989).

(38.) This difference has been characterized in terms of various distinctions: e.g. between ethnic and civic, Eastern and Western, ethnos and demos, and German versus French understandings of nationhood. Like many generalizations, this one is overdrawn. On the one hand, French history reveals the importance of both kinds of self‐understandings (Grillo 1998: 121–40). On the other, these baseline tendencies are conditioned by historical particulars that can change as circumstances do. Thus the present (p.188) Social Democratic regime in Germany has proposed legislation toward immigrants that is far more culturally inclusive than that entailed in current French policy.

(39.) Opposition to direct rule was not always based on nationality; often it was based on religion. Breuilly (1993: 76–81) contains a useful discussion of the similarities and differences between religious and nationalist opposition to the imposition of direct rule. Breuilly concludes that Namier is to a large degree justified saying that religion is a 16th‐cent. word for nationalism. There is a fundamental difference between these two bases of opposition, however: unlike the national variety, religious opposition is guided by an explicit ideology (Hobsbawm 1992: 175–7).

(40.) Although England adopted direct rule early, the origins of British nationalism largely arose as a defensive response to direct rule in France: ‘Time and time again, war with France brought Britons, whether they hailed from Wales or Scotland or England, into confrontation with an obviously hostile Other and encouraged them to define themselves collectively against it. They defined themselves as Protestants struggling for survival against the world's foremost Catholic power. They defined themselves against the French as they imagined them to be, superstitious, militarist, decadent and unfree’ (Colley 1992: 5). See also Newman (1987).

(41.) Interestingly, the Japanese employed direct rule in their Asian colonies (especially Korea) to a far greater extent than other colonial powers. By concentrating political power in central agencies, rather than relying on local notables (as in indirect rule), this may have inadvertently set the stage for strong states in Korea and Taiwan in the era following the Second World War (Migdal 1988: 272).