Nationalism is a will-o-the-wisp. Now you see it, now you don’t. Or, more precisely, now you see something reasonably definite, clearly demarcated from other things, and now you see it all over the place. Writers talk of British or English nationalism in the eighteenth century, a passion expressing hostility to the French. But nationalism only makes sense as a criterion of the proper composition of a modern state, and nothing of the kind was involved in British feeling against the French at that time. The term ‘nationalism’ is a problem because it has been used to cover both collective self-regard expressing itself as enmity to another collectivity, and a doctrine about the criterion for a proper state. The first meaning finds instances at many times and places, the second is modern.

With some exaggeration, one may say that these ambiguities of ‘nationalism’ were decisively clarified by Elie Kedourie.footnote1 Before Kedourie, nationalism lurked in every corner of history, discoverable in any collective form of self-assertion. After Kedourie, no careful thinker would attribute it to non-modern times and places. He showed that it was a doctrine ‘invented’ in the early nineteenth century. In his article on Kedourie in New Left Review, ‘In Praise of Empires Past’, Brendan O’Leary wants to dismantle this achievement.footnote2 His is a bravura piece, occasionally crawling pedantically along, at other times featuring intrepid logic-defying leaps, as when Kedourie’s repudiation of prediction becomes a rejection of generalization tout court.

The issue here is one that needs keeping under review. For although Kedourie’s basic judgement has mostly been accepted by students of nationalism, tendencies to recidivism continue to crop up. Eric Hobsbawm, for example, more or less accepts Kedourie’s verdict, although without mentioning him, but the next moment can be found toying with eleventh century ‘proto-nationalisms’. Kedourie wrote as a historian, which meant that he was interested in the circumstantial explanation of a certain way of thinking, and its antecedents. Like others, he numbered Rousseau and Herder among the latter, and added Kant’s theory of self-determination. But he argued that nationalism was a political doctrine that only took recognizable shape in response to the universalism of the French Revolution. He thought nationalism an essentially opportunistic response to various kinds of collective grievance: in his eyes, it had no ‘essence’ (pace O’Leary). It could espouse republics at one time or monarchies at another, and might quarrel or ally with other ideologies such as socialism, as the exigencies of the moment suggested to those who picked up or perhaps developed the doctrine. Nationalism was an instrument of practical action, not an explanatory philosophical idea. That was why an understanding of it had to be essentially historical. Most of its notions, moreover, take for granted the modern sovereign state; they make no sense in the mediaeval period or in oriental empires.

O’Leary contests Kedourie’s famous first sentence: ‘Nationalism is a doctrine invented in Europe at the beginning of the nineteenth century.’ Strictly speaking, to be sure, in history no idea is ‘invented’—‘emerged might have been a safer term, and would have avoided O’Leary’s objection that an invention needs a single inventor, and Kedourie does not provide one. The real issue, however, lies in O’Leary’s attempt to turn Kedourie’s flank by asking: why can we not find nationalism in the American or French Revolutions? The answer to this lies in culture and style. The Americans had a grievance against the British government which they formulated not as the oppression of a culture but as a denial of rights. They were not even sure whether they were one nation or thirteen. The French dreamt of reviving the Roman republic; their language and attitudes appealed to the universal rather than the particular.

That nationalism emerged out of some of the enthusiasms of the eighteenth century is no doubt true. But when did it ‘crystallize’ in a vocabulary, a set of emotions, and a doctrine capable of justifying a certain sort of political involvement? Since Kedourie refers to figures like Kant, Schiller or Frederick the Great, O’Leary asks why he did not date nationalism back to them. The answer might be couched in the analogy of a photographic plate: at what point has the image actually appeared? In trying to capture the coming into being of a new idea, I would not myself talk of ‘catalysts’ subsequently ‘mushrooming’, as O’Leary does. What he means by this mixed metaphor is, I think, that nationalism suddenly turns up all over the place in—and indeed beyond—Europe, from which he draws the conclusion that the ‘real cause’ of nationalism lies in ‘specific historical developments’, presumably the arrival of modern industry. He thus rejects diffusionism in favour of what one might call ‘material causation’. The broad issue distinguishing these alternative explanations is that of history versus sociology. Kedourie’s history of ideas interprets nationalism as a response by political actors, within the crucial contexts of collective grievance and the sovereign state, to changing events. O’Leary’s sociology seeks to explain nationalism as a phenomenon ‘unavoidably’ and ‘predictably’ issuing from modernization. He personalizes this into a dispute between Kedourie and Gellner, who did on occasion discuss the issue.