The most influential theorist of nationalism, Ernest Gellner, was wont to claim that one should assimilate or get one’s own state given that the only remaining alternative was to be killed. Each nation must have its own state, and each state its own nation.Footnote ¹ Homogeneity is all. It is obvious that this view reflects the experience of a thinker with a Jewish background (or, to be precise, of a thinker upon whom Jewish identity was imposedFootnote ²). There is a sense in which he was right: the experience of Europe’s dark twentieth century has been, exactly, the creation of relatively homogeneous units through forcible assimilation, population transfer, ethnic cleansing and mass murder. I will return to Gellner later, but stress now that the main point to be made here is that his position fails in the terms he had in mind, that is, as a general theory of nationalism – based on his insistence that homogeneity derived from and was necessitated by the tectonic change brought about by industrialization.

An alternative view stresses the very varied arrangements which have allowed different ethnicities and nations to live together in peace. The key principle here is very clear, and it is best expressed in terms of Hirschman’s Exit, Voice and Loyalty (1970).Footnote ³ When a nation is denied voice, that is, when it is faced by a state denying it cultural rights and political representation, secessionist exit becomes attractive, even necessary. Allowing voice, in contrast, can produce loyalty, thereby undermining secessionist drives. This is to reverse the famous parable in Gellner’s Nations and Nationalism: the Ruritanians may indeed cause trouble but only because the Megalomanias have begun to interfere and discriminate against them so massively. The end result is clear: the nationalist principle can be honoured by the provision of cultural rights and political representation within a larger polity, for all that exclusion and discrimination politicize, thereby leading to conflict and to secession. Crafting State-Nations makes this case with brilliance and passion, as one would expect given the prominence of the team of Linz and Stepan in the field of democratization studies. One motivation for writing the book was the belief that the world has changed: cheap travel and easier forms of instant communication make multiple identities possible – or, differently put, diminish the likelihood of assimilation. And there is a second motivation with which all can agree. The thought of the rest of the world following Europe’s route to the modern world is terrifying. Prescriptively, there is everything to be said for seeing if alternative practices are possible; Crafting State-Nations suggests that hope is realistic given their description of actually existing arrangements in the key cases that concern them.

The authors make their case with power and elegance. An opening chapter suggests that we need a “state-nation” model quite as much as the received notion of the nation-state. Over half of the book that then follows is concerned with India, not surprisingly since the stability of this diverse entity does more than anything to make their case. A great deal of attention is given to the intelligent crafting of India’s founding fathers. Still more interesting are the descriptions both of moments when India was not, so to speak, true to itself, thereby causing secessionist impulses on its borders, and of occasions when its better self came to the fore – with successful accommodations of Sikhs and Tamils receiving careful and detailed attention. The authors are famous for comparative research, evident here both in accounts of the exclusionary and hence violent politics of Sri Lanka and the inclusionary politics that have allowed Ukraine to remain – literally – in one piece. The book then proposes that the politics of inclusion can work even when a full state-nation model is not possible; a federacy can grant special rights to a region, thereby gaining the loyalty of a minority. A final chapter has a negative purpose: the federalism of the United States does not provide a model to be copied. For one thing, states were only admitted to the union once white Anglo-Saxon majorities had been created; for another, the Senate is wholly malaportioned, giving blocking powers to very small numbers. And there is, of course, the little matter of the Civil War.

It is worth noting an interesting range of opinion amongst theorists who think in the alternative terms just noted. At one end of the range stand thinkers seeking an open political system so that complete assimilation can take place; at the other end are those privileging the deep diversity of a self-maintaining nation within some larger political frame. If the important work of David Laitin stands at the assimilationist end of this spectrum, middling positions are provided by Donald Horowitz and by Andreas Wimmer, both seeking reasonable accommodations but both keen to contain nationalist demands, whilst the works of Brendan O’Leary and of the authors of the volume under discussion recommend and believe sustainable the greatest degree of pluralism.Footnote ⁴ These are interesting and important differences, and they most certainly matter in terms of practical policy proposals. What is the position of Crafting State-Nations?

Much is initially made of key descriptive points. “Holding together” federations, that is, those – India, Canada, Switzerland, Belgium and Spain-based on compromise between different nations are shown to have more legitimacy than “national federations” such as the United States which base themselves on a good deal of homogeneity. The central pillar of empirical support of the whole volume is, as noted, the fact that India works as a multinational entity. The founding fathers were well aware that the imposition of homogeneity, not least given the prestige of English amongst Indian civil servants, would cause conflict, and so thought systematically about means to accommodate and include. Thanks to the brilliant empirical work of Yadav allowing for multiple entries as to identity in the “State of Democracy in South Asia Survey of 2005”, it is clear that the presence of secularism and linguistic rights have resulted in very high levels of loyalty to the Indian state.Footnote ⁵ The superlative comparison of the situation of the Tamils in Tamil Nadu and in Sri Lanka makes the case very forcibly. The secessionist nationalist drive in Tamil Nadu was contained by the granting of rights, whereas vicious state policies against the highly educated Tamils in Sri Lanka resulted in civil war.

On the basis of these findings, and of the analysis of Ukraine, policy prescriptions abound. The creation of “holding together” federations is a rare event. But various versions of federacy seem to have worked in recent years. Interesting and persuasive analyses are offered of Finland’s treatment of its Swedes, of Denmark’s treatment of the Faroes and Greenland, of the role of federacy in the Italian transition to democracy after 1945, and of the new settlement in Aceh. (I find it hard to believe, however, that Paris will use this principle with Corsica!). Finally, the authors passionately argue for non-presidential constitutions, on the grounds that a presidency, by definition a singular position, is likely to encourage zero sum politics.

One mild criticism that can be directed against Crafting State-Nations concerns the data used, and thereby a variable that is underplayed. Their survey data concerns the general population, in itself a huge achievement. But one needs also to measure the extent to which the political elite is prepared to identify, in terms of career and commitment, with the project of the larger state. The importance of this variable has been shown with characteristic brilliance by John Darwin, the historian of the British empire, when writing about the too-often-neglected imperial ethnicities.Footnote ⁶ Alternative data measure political inclusion systematically, and can be recommended as a complement to this volume.Footnote ⁷ A general consideration is relevant at this point. Attention focuses on policies derived in one way or another from federation; this is to downplay the importance of consociationalism, that is, the politics of accommodation between different elites at the centre.

A cautionary word is needed about the notion of crafting. Perhaps the most dangerous use of a phrase of recent social science has been that of “imagined communities”: it has led to a dreadfully illicit idealism and voluntarism, taking for granted that we can imagine what we will whenever we wish. It is as well to remember in this context Bertrand Russell’s attack on the social contract theory proposed by Rousseau. The idea that a set of venal and corrupt individuals would choose to make a contract turning them into virtuous citizens was ridiculous, indeed pathetic: only the virtuous would make a contract of this sort. This matters here. Wimmer’s analysis of Switzerland demonstrates that the Swiss had many transethnic institutions and a vibrant civil society before the civil war that took place in the middle of the nineteenth century.Footnote ⁸ The fact that ethnicity had not been politicized made it possible thereafter to create what Wimmer terms a multi-ethnic nation-state – in contrast to the book under discussion which classifies Switzerland as a “state-nation”. A similar point can be made about Canada: the prior presence of liberal institutions made it possible to include Quebec, albeit only in a state of bearable dissatisfaction. One might further emphasize that Indian “holding together” had as much to do with the presence of an army and bureaucracy, and the memory of a shared nationalist struggle against the British.

But this is not to say for a moment that political design cannot work in the most difficult circumstances, that is, when politicization has taken place in places bereft of rich institutional portfolios.Footnote ⁹ But crafting is hard, as another exponent of political crafting makes increasingly clear when specifying the conditions that allow it to succeed.Footnote ¹⁰ In this context one can usefully note the brilliant work of Matthew Lange on the ways in which education can enhance the possibilities for ethnic violence.Footnote ¹¹ When a large, newly educated subaltern group finds that key top professions are monopolized by a minority ethnicity, great difficulties inevitably ensue. Such difficulties can be managed. Nonetheless there was and is an essential element of truth to Gellner’s theory of nationalism, namely that the entry of the people into politics involves social and political dynamite.

A final comment on Gellner’s theory is needed, less to say something about his theory of nationalism than to specify the key variable lacking from Crafting State-Nations, and from many other treatments of nationalism. Why was it impossible for the great multinational empires in Europe to decompress, to move to more liberal arrangements that would have allowed several nations to live under a single political roof? A crucial clue can be found in the political views of Max Weber – a member of the Naval League believing that Germany needed a place in the sun and a nationalist famously worried that an excess of Poles in East Prussia would diminish the power of the state. What this amounts to is a marriage between nationalism and imperialism, in which state strength seemed to depend upon the sharing of a culture and the possession of secure sources of supply and protected markets. If Gellner was right to stress that the world from which he came had homogenizing politics at its core, he was wrong to see this wholly in terms of industrialization. What mattered most was the intensity of geopolitical competition, encouraging the great world states to behave radically, as – exactly – megalomanias, with all that this implied for consequential nationalist agitation.

We no longer live in this sort of world. Who now believes, given the fabulous success of Europe free from overseas possessions and prepared to place interdependence above geopolitical autonomy, that the retention of, say, a Chechnya, will ensure the prosperity of its metropole? This is not to say that liberal universalism has somehow triumphed. The lack of irredentist fervour among modern Hungarians, so astonishing in historical perspective, is firmly based on xenophobia: to take back Transylvania would create a minority of dissatisfied Rumanians; better to avoid a nationality problem altogether.Footnote ¹² The analytic point is simple. There may be an end to homogenizing policies in countries that do have minorities, but this has as a structural base relative geopolitical peace, at least in the North, which has diminished the necessity of states to be as unitary as they were in the past. Political design is important, but tectonic geopolitical changes lie at the back of many of its successes.

Of course, there is a final concern. What if the geopolitical order we know, but perhaps do not love, came to an end? Might size matter in a world of global warming in which interdependence faltered? One can fear that the marriage of nationalism and imperialism might yet return.