To say that we live in an age of nationalism is both a platitude and a provocation. It is a platitude because there are more nation-states in the world of today than at any other time in history. Decolonization gave birth to several score new nation-states in Africa and Asia from the 1940s to the 1970s. The collapse of the Soviet Union and the break-up of the Yugoslav Federation released a number of "new" old nation-states from communist imprisonment across Central Europe to Central Asia. And unsatisfied nationalist movements today seek statehood from the Basque country to Palestine.

Yet nationalism also runs counter to other trends: namely, that the internet and globalization are shaping a borderless world in which multinational corporations are more powerful than governments; that transnational bodies from the European Union to the International Criminal Court are usurping powers that once belonged to sovereign states; and that new non-governmental organizations claim to represent public opinion on international questions more faithfully than democratic governments. So nationalism looks increasingly constrained functionally even as its geographical sway extends.

To complicate matters further, there are influential intellectual trends in the advanced world that deny the legitimacy of nationalism altogether as an atavistic concept. Their adherents regard nationalism as an obstacle to human rights, international harmony and economic rationality. They accordingly seek to reduce the scope of national sovereignty in international affairs, transferring power upwards to global bodies and downwards to ethnic and other subgroups. Both President Bill Clinton and Prime Minister Tony Blair justified their Kosovo intervention in terms not simply of halting ethnic cleansing, but of defending the positive principle of multi-ethnic statehood. And General Wesley Clark stated flatly: "There is no place in modern Europe for ethnically pure states. That's a 19th century idea and we are trying to transition into the 21st century, and we are going to do it with multi-ethnic states" (emphasis added).

To those who think of nationalism as a natural loyalty to one's nation and who therefore regard the Westphalian system of nation-states as an equally natural world order, this hostility to nationalism must be something of an intellectual mystery. Is railing against nationalism not as foolish as cursing the weather or complaining that water will not run uphill? Have not nations existed since time immemorial? Is it not reasonable as well as right that a people who lack their own state, such as the Kurds, should strive to acquire one or that a people who have their own state, such as the British, should seek to protect its sovereignty against legal erosion or military attack?

These questions probably strike most people in Western countries as common sense even today. But since the Second World War--not coincidentally--there has been an immense scholarly dispute over the naturalness of nationalism. Writing in The National Interest (Fall 1997), Anatol Lieven, the historian and journalist, summed up this debate:

"One side of the scholarly debate on the origins of nationalism stems ultimately from the belief that . . . the roots of modern national allegiances lie in old and deeply felt ethnic, linguistic, religious and cultural differences. . . . In Western academia in recent decades, however, this approach has not been so much dissected as slashed to pieces by a range of scholars who have . . . sought to expose the numerous ways in which nationalisms and, indeed, national traditions were artificially created in modern times."

Natural or artificial? These are the two points of view from which we are generally asked to choose. Good authorities can be cited to justify each of them. Suppose, however, that both viewpoints mix quite different things under the name of nationalism? Might they then not both be wrong--or at least oversimplified? Let me suggest that this is the case and that three quite different sorts of political commitment have been confused under one misleadingly similar heading.

These three concepts are, first, the political doctrine that the nation is the only legitimate basis for statehood; second, the political emotion of collective loyalty that might be attached to a nation, a race, a class, a religion or even a political ideology; and, third, the sense of national identity, sometimes called patriotism, that arises from living together under the same institutions and sharing a common language and culture over time.

The concepts may seem indistinguishable at first hearing. And they are all linked, as we shall see. But they also differ fundamentally. To see exactly how, let us examine each one in detail, beginning with nationalism as a doctrine of statehood.

Elie Kedourie's classic book Nationalism opens with this simple but dramatic definition:

"Nationalism is a doctrine invented in Europe at the beginning of the 19th century. It pretends to supply a criterion for the determination of the unit of population proper to enjoy a government exclusively its own for the legitimate exercise of power in the state and for the right organization of a society of states. Briefly, the doctrine holds that humanity is naturally divided into nations; that nations are known by certain characteristics which can be ascertained; and that the only legitimate type of government is national self-government."

Those sentences are quite literally mind-altering. At least they altered my own mind when I first read them, persuading me that my own unexamined sense of nationalism as a natural emotion like love of family was simply a misconception. Once we grant Professor Kedourie's definition, moreover, we are drawn eventually to the conclusion that nationalism is not just an intellectual error, but also a destructive one.

To begin with, the doctrinal nationalism he condemns is rooted in a false and ideological conception of statehood. Its underlying error is to suppose that legitimate statehood must rest upon some universally valid principle. In fact, states are the product of history and accordingly rest upon many different foundations--dynasty, ethnicity, culture, religion, ideology, conquest and revolution.

No sensible statesman asks of a state: "Is it based upon a recognizable nation?" (Or, as Clinton and Blair might prefer, "is it multi-ethnic?") He asks rather: "Is it stable? Does it have generally accepted boundaries? Does its government really control its territory? Does it enjoy the loyalty of its citizens?" And so on. If these questions can be answered more or less positively, then it scarcely matters whether the population of the state is ethnically pure, as for instance Norway, or composed of several ethnic groups, as for instance the United Kingdom. It is likely to be a reasonably successful state and an orderly presence on the international scene.

Doctrinal nationalism is also rooted in a false account of history. As a matter of historical fact, very few states can trace their origins to ethnically distinct peoples that have remained uncontaminated by their neighbors over the centuries. That is why historians and teachers were conscripted in the 19th century--and more recently in the Balkans--to trace their present nation back to ancient times with the help of maps, poems and ancient scripts. The objection to this is not that it is patriotic in effect--so is much good history--but that it is false and invented for a political purpose.

As it happens, in Europe, where doctrinal nationalism was itself invented, very few existing states fitted the nationalist theory. In Hugh Seton-Watson's words: "Every England had its Ireland, and every Ireland had its Ulster." Nationalism was therefore divisive--not in the modern sense of stimulating debate, but because it encouraged national minorities to seek the breakup of the state in which they were allegedly imprisoned, and because it gave neighboring states a pretext to intervene on their behalf. Conflict was perhaps inherent in the circumstances of Mitteleuropa. At the very least, however, doctrinal nationalism aggravated it.

If doctrinal nationalism is easy to dismiss or condemn, does that demolish nationalism of other kinds? After all, how many people who think of themselves as nationalists because they love their country hold the opinions of the 19th century German intellectuals who largely dreamt up this theory of nationalism? Most of them would never think of asserting that nationality is the only legitimate basis for statehood. Yet they experience strong emotions of loyalty and allegiance to their country.

Recognition of this truth is the beginning of the second theory. This theory sees nationalism as a form of collective political loyalty that is usually attached to the nation but that is capable of being separated from it and re-attached to some other unit of humanity. The most famous theorist of this form of nationalism is also its most famous critic, namely George Orwell in the essay "Notes on Nationalism."

Orwell begins by specifying that nationalism is not to be confused with patriotism. In his formulation, the former is aggressive and power hungry, the latter defensive and devoted to celebrating a particular way of life. In order to justify this distinction he has to define nationalism in a singular and arguably eccentric way: namely, as "The habit of identifying oneself with a single nation or other unit, placing it beyond good and evil, and recognizing no other duty than that of advancing its interests." He then extends this definition to cover almost every political unit of humanity. Nationalism in this extended sense, he argues, covers a variety of movements, including communism, Zionism, pacifism, political Catholicism and anti-Semitism. And he observes, finally, that the devotee of some transferred nationalism, such as a Stalinist or a pan-Europeanist, is able to be much "more nationalistic, more vulgar, more silly, more malignant, more dishonest than he could ever be of his native country or of any unit of which he had a real knowledge."