Empire’s Past . . . Empire’s Future

In February , as I was writing this paper, the Financial Times reported at length on an upcoming British Museum exhibition on 16 th -century Persia: “the third in a series on great world empires,” following a “hugely popular show” featuring the terra cotta soldiers of the great Ch’in emperor, a second exhibition on Hadrian, and to be followed in turn by a show on the Aztec emperor Moctezuma II. The director, Neil MacGregor’s “overall concept for the series is to study the ‘instruments of cohesion’ that held those empires together and ask what the consequences of their dominion have been in the long term.” 1 And so by means of their artifacts, suitably displayed, empires past radiate their light downward to the present, like stars after extinction, allegedly educating us all, as Director MacGregor intends, in the interconnections of our species. Is it “interconnectedness,” as MacGregor believes, that makes empire so compelling a theme? Perhaps. A good exhibit does startle or delight us by revealing connections between remote societies and ultimately our own—whether of art, or beliefs, or commodities, alphabets and calendars, or germs and pathogens. After all, connections are the basis of one major contemporary approach to international history—that is they are the privileged subject of the contemporary effort to write “entangled histories” or what the French call histoires croisees, and the history of what many colleagues now like to call “cultural transfer.” And just as economists have developed the theory of the firm, as a way of minimiz ing the so-called transaction costs of contracts among buyers and sellers, Empires, in a sense, have been an institutional device for internalizing such cultural transfer. An empire provides a political structure for extending cultural transfer through time and space. Historians divide, as did policy makers, on the nature of that structure. For some of us it is a harsh one, built on power and domination; for others it has softer outlines and processes that we describe in terms of negotiation. Every subaltern people has agency; hence every transaction is negotiated—a flight of dubious logic, it seems to me. And for many, empire has the sentimental redolence of a wonderful and idealistic project. As a Swedish journalist explains what he termed the longest-term geopolitical project in world history, “The empire is gone, but the sun never


Charles S. Maier, Harvard University
In February, as I was writing this paper, the Financial Times reported at length on an upcoming British Museum exhibition on 16 th -century Persia: "the third in a series on great world empires," following a "hugely popular show" featuring the terra cotta soldiers of the great Ch'in emperor, a second exhibition on Hadrian, and to be followed in turn by a show on the Aztec emperor Moctezuma II. The director, Neil MacGregor's "overall concept for the series is to study the 'instruments of cohesion' that held those empires together and ask what the consequences of their dominion have been in the long term." 1 And so by means of their artifacts, suitably displayed, empires past radiate their light downward to the present, like stars after extinction, allegedly educating us all, as Director MacGregor intends, in the interconnections of our species.
Is it "interconnectedness," as MacGregor believes, that makes empire so compelling a theme? Perhaps. A good exhibit does startle or delight us by revealing connections between remote societies and ultimately our own-whether of art, or beliefs, or commodities, alphabets and calendars, or germs and pathogens. After all, connections are the basis of one major contemporary approach to international history-that is they are the privileged subject of the contemporary effort to write "entangled histories" or what the French call histoires croisées, and the history of what many colleagues now like to call "cultural transfer." And just as economists have developed the theory of the firm, as a way of minimizing the so-called transaction costs of contracts among buyers and sellers, Empires, in a sense, have been an institutional device for internalizing such cultural transfer. An empire provides a political structure for extending cultural transfer through time and space.
Historians divide, as did policy makers, on the nature of that structure. For some of us it is a harsh one, built on power and domination; for others it has softer outlines and processes that we describe in terms of negotiation. Every subaltern people has agency; hence every transaction is negotiated-a flight of dubious logic, it seems to me. And for many, empire has the sentimental redolence of a wonderful and idealistic project. As a Swedish journalist explains what he termed the longest-term geopolitical project in world history, "The empire is gone, but the sun never sets on Portuguese culture. Millions of children, on four continents, from the Far East to Africa and the Amazon, do their homework each day in the language of Luis de Camões or Fernando Pessoa." 2 Jan Morris recalls London in 1944, when she, then a he, was a soldier on leave in London: "it felt to me that we truly were part of a band of brothers-thousands upon thousands of us, from the four corners of the world, united in allegiance and in loyalty under the leadership of the most charismatically ornamental imperial chieftain of them all, Winston Churchill. Of course my sensations were manipulated [. . .] but they seemed beautiful to me then, and they remain beautiful in my memory still." 3 The sentimentalist vision is hardly analytical; but the negotiated empire sounds quite nice and in line with our aspirations for tolerance and norms. A recent analysis of the Ottoman Empire, stresses the formation of networks and connections as key to the construction of this very long-lived and extensive imperial structure. 4 On the other hand, it hardly mentions the harsh facts of military conquest and expansion. If empires rest on networks, the networks often rest on domination and domination on a grand scale. When I visit museum exhibits about empires, which I enjoy immensely, I am drawn to the large wall maps at the entrances, with their settlements and river valleys, spreading around the Mediterranean or the interior of China, or the grasslands of Parthia, or the highlands of the Andes, and their lines of land and sea communication. Whether they are brief or long-lived, it is the spatial grandeur of empire that overwhelms me. And the spatiality of empire is defined not only by size and extent, although we associate empire fundamentally with bigness. Empires work by being functionally differentiated territories. I will return to that attribute. For now I cite the exhibits because they reflect the different representations or conceptions of empire. Connections or conquest-the beautiful artifact that reveals cultural fusion, or the large map that shows the multiple sites of empire? How should we envisage empire?

I.
This reflection thus raises a critical question: Whence our current fascination with empire? We cannot deny its challenge as a historical subject-especially for those of us who have gathered at this conference. It goes beyond just mere analytical curiosity. Was it the Iraq war, which set Americans down in the sites of antiquity on a dubious mission? Was it the consequence of the brief unipolar intoxication we inherited in l991 as the Soviet Union and Eastern Bloc communism collapsed? Was it a sense of barbarians seeking to invade from outside remote frontiers that we sensed with the destruction of the World Trade Towers? Was it a cascade of cinematic themes-from Star Wars and War of the Worlds to Troy to Three Hundred? Was it the shock of recognition: the slowly dawning awareness that having lived as historians and as Americans with a narrative of secession from empire, we discovered that we had become what we thought we had revolted against? And this in two ways: first, the inner evolution of the American democracy from a clumsy set of checks and balances to an uncontrolled grant of executive prerogative, and second in our increased reliance on military force in foreign affairs. Were we not recapitulating at home-so some of us feared-the dangers of the late Roman Republic? Were not the President's lawyers developing the "imperial presidency," first in the l970s, then under the doctrines propagated by John Yoo and David Addington in the presidential administration of George W. Bush? And with respect to global politics, were not Vice President Cheney and Secretary of Defense Donald Rumsfeld celebrating the mission of the American state as a force for domination by virtue of unchallengeable military supremacy? Perhaps we were fascinated because we realized that for all the historiography of American ideology in terms of civic republicanism-that is, in terms of a particular strand of enlightenment critique of arbitrary power-in fact resistance to power seemed simply to melt away and to become irrelevant and collapse. At home, it seemed, America's leaders could not resist claims to power; in our new security doctrines they glorified power. And empire is "about" power and exerting power.
Of course, all politics is about power. But democratic politics is about the continual contestation of limited power. Imperial politics is about the placing of power beyond limits; and certainly for many of us that seemed to be the administration's objective. Still as historians, even if we are often motivated by current politics, we cannot rest content with narratives keyed to current politics. Our discipline and our scholarship never guarantee arriving at a perspective beyond politics, but we are enjoined to try. Political concerns may well have helped put "empire" on our agenda as historians. It may well have made us aware how it endows the theme or subject with a sort of aura or fascination, an aesthetic fascination with violence and inequality, irrational desires to extend territory and increase "glory"-all themes that in daily life we try to keep within safe boundaries.
But if historians begin with some of the feelings that the concept of empire arouses: sentimental memories, justifications, guilty denialsamong others, they cannot stop with feelings. But how do we do schol-arship about empire? How do we get beyond historiographical "shock and awe," to put it crudely? I have already cited the artifact and the map: tracing the impact or envisioning scale and structure as alternative approaches, each valid perhaps, but speaking to different temperaments. Can we even presuppose empire as a single abiding political state form? Indeed, does it make sense to search for the recurrent traits of empire? Is it really so continuous a state form that we can reach back through time and compare the empires of antiquity with those of the early modern era, then with those of the nineteenth-century overseas empires, and finally with the forms of domination exercised by the later Soviet Union, or in effect radiated by the contemporary United States?

II.
Historians rarely escape their eras. My own interest in empire arose from trying to understand in what ways the United States-the US that had organized a postwar alliance to fight the Cold War, that had then prevailed in that contest, and in the years thereafter became involved in wars in the Middle East-was a sort of empire. Rather than become mired in definitional disputes, it seemed sensible to establish the traits that known empires of the past had demonstrated-what I called the analog of empire. It was an effort to specify the "imperial minimum." This was the task I attempted in my own book of three years ago, Among Empires. The reader, I hope, will forgive a bit of scholarly recapitulation as I summarize those results here.
To suggest that America might be an empire antagonizes many readers. We associate empire with a process of achieving ascendancy by the exertion of power. Did America use the classic methods of conquest and intimidation? Certainly it did to fill up the continental territory. Did it then seek overseas empire, like the Portuguese, Spanish, Dutch, French, British, Spanish and Japanese? That is debatable: we kept Guantánamo, just as we shall keep a massive air base in Iraq after our withdrawal. Chalmers Johnson believes that our 700 some bases constitute an empire; I am less certain. Bases are enclaves granted by other peoples and states; they provide extra-territoriality, but they do not provide the right to rule these other places. The retention of the Philippines from 1898 to 1946 and of Puerto Rico have been classic prizes of empire and the Philippines involved a brutal war to subjugate the adversaries of our control. Generally, U.S. policy was oriented on preserving access for American business and evangelists, not administrative control.
What made sense, when I wrote my book, was to think of empire less in terms of a process of conquest, than a mechanism for security and superiority. I defined empire as a transnational cartel of elites, which gave local hierarchies security in return for their recognition of the hierarchies at the center as supreme. Empire thus becomes an institutional alternative for guaranteeing societal privilege and security. This aroused criticism, especially at the outset of my discussions. The eminent historian of international relations, Paul Schroeder, criticized me strongly for mistaking empire with hegemony. Somehow the term hegemony provided reassurances. It seemed less imperialist than the term "empire." It expressed no intention to dominate, but merely the consequence of preponderant size and power. Hegemony might arise without a program. The main distinction between hegemony and imperium that Paul Schroeder cited was that hegemony supposedly did not attempt to rule what happened within the territorial units that deferred to them.
Michael Doyle had drawn a similar distinction in a valuable work on Empires almost three decades ago, and the British diplomat and historian Richard Cooper came to the same conclusion when he spoke at a conference on empire in the summer of 2005: "The essence of empires is that they impose domestic government." 5 This distinction has thus become widespread, but I am not convinced that it is a robust one. Many imperial powers will not try to control internal policies, much less run the states whose policies they can dominate, provided-but only provided-that the client state does not try to withdraw from friendly relations. They must make sure that in their dependencies no hostile coalition takes power. If a hegemon thus intervenes to secure regime change or policy change when it fears loyalty is evaporating, it is perforce behaving as an empire.
So what criteria distinguish empires? What are the commonalities that places labeled empires have shared? What sorts of action does the US pursue that earlier empires have also engaged in? After all, what is worth knowing is not the name for the United States regime, but the elements of American behavior. The stakes are civic and practical. If we Americans can't define exactly what we are, we can say in what ways we are likely to act. Thus the task is to explore those aspects of society and politics we find throughout the history of empires.
When we seek to answer this question we have to ask about behavior at home as well as abroad. An empire is a type of regime or state, not just a pattern of conquest or expansion. The problem, of course, is that there is no one pattern or analogue. Historical interpretation remains a struggle over the appropriate basis for analogy. If empire means possess-ing populated colonies abroad, such as the British, French and Dutch did, then the term makes little sense for the United States. If empire refers not to colonization, but rather to a less formalized search for decisive control by intervening to remove governments we dislike and installing those we prefer, i.e., engaging in so-called regime change, then the U.S. should be reckoned as imperial, although most American policy discourse never describes regime change as imperial.
If empire refers to political structures at home, that is to a state where the executive is given powers of arbitrary arrest and imprisonment and the representative assembly is reduced to a rubber stamp role, then there is room for debate about whether the contemporary United States fits. Less now, happily, than was the case before 2006. Most commentators, however, probably separate temporary, wartime emergencies, where "exceptional" control may be delegated to the executive (as it has been at least since the Roman Republic), from regimes that perpetuate exceptional control in times of peace as well as war. Empires are structures that persist in peace as well as war. To pass through an episode of executive centralization and suspension of earlier civil rights does not entail becoming a different regime. Most such delegations in American history seem to have been temporary expedients if not aberrations. They have quickly ended. That is why my current answer to the question of whether we are an empire is "We came close." Of course the provisional can be indefinitely prolonged-and when the executive claims we shall be at war for an indefinitely prolonged period, the situation is rendered more fraught. But that danger seems averted.
If empire involves a historical process by which rule over an extensive territory is acquired by military expansion, then historians differ over how much a process of conquest was involved in filling the continental land mass the U.S. acquired from the late eighteenth century on. The question also remains whether a state geographically constructed, at least in part, by a process of imperial expansion must thereby remain an empire. Do the Cherokee removals and the Mexican War remain, so to speak, a sort of historical original sin?
My own preference is to define empire in functional terms, that is, in terms of what empires have sought to accomplish and how have they behaved. 6 Empire (so I suggested in my book) is best understood as a program by the elites of different national groups to stabilize their societies, and the distributive norms of their societies, by spatial as well as social hierarchy. Empires thus are about inequality across a spatial domain; call this horizontal domination. Empires are large enough to have differentiated territories that include a center and a perimeter, metropole and periphery. But empire is also about vertical domination. It helps keep certain groups wealthy and powerful, and it recruits others by birth or talent to become wealthy and powerful. And it helps assure this inequality within each territorial component. An empire is thus an arrangement, whether negotiated voluntarily, or by force, in which elites in the so-called periphery accept the ultimate control of elites in the metropole in return for securing their own local domination. The security sought can be against outside rivals and domestic subversives, or both simultaneously. Empires thus rest on collaborators, but they are not alliances of equals, but rather structures of inequality-both inside their homeland and within the imperial structure as a whole.
We can get more precise about the imperial minimum, or what might be called the imperial syndrome. First, empires tend to pursue a typical spatial dynamic. They enlarge territory or influence to confirm their own new political order-and then they must defend the boundaries they have extended to avoid endangering the expansion just attained. Territories once occupied are hard to relinquish-and sometimes for perfectly valid reasons of having taken on responsibilities toward the inhabitants who might otherwise descend into fratricidal violence. The entrepreneurs of empire need not premeditate expansion and continuing control of additional territory, although many obviously have done so. Most managers of empire rarely have a vision of their cumulative power-but like a ratchet, their acquisitions conduce to expanded commitments. Every time an expanded frontier is stabilized, threats come from just across the new frontier. Retreat or retrenchment often seems catastrophic; and there's always an unpacified and menacing site of disorder just beyond the limits already reached. Every new border, every new acquisition, every new base creates surrounding instability that often calls for further expansion. An alliance founded in 1949 to protect West Europeans from invasion now finds itself patrolling Afghanistan at the behest of its major organizer.
This means, second, that empires live with the possibility of force. They believe themselves summoned to perpetual battle: the idea of war evolves from that of a particular conflict to a generalized state of national challenge. Empires are often at war. They often arise out of war; they maintain their domains through force or the threat of force; they collapse often in conflict-in this respect the end of the Soviet Union was a striking exception, although Chechnya reveals that not every region might be relinquished easily. And finally empires leave wars behind them as a legacy-think of Ireland, Palestine, Kashmir and Nigeria in the case of Britain, or the Congo in the case of Belgium. Of course nation states are frequently at war, too, but it is more difficult for the empire, with its preoccupation with frontiers and control, to forsake the military dimension of statehood. Perhaps empires bring peace to the interior of their large domains. This was Virgil's famous description of Augustus's task: To humble the arrogant, raise the oppressed, and impose the habits of peace. But there is always combat on some frontier, someplace. For each frontier imposed usually means violence just beyond it. The state of war becomes the normal state-there are many advantages to such a conviction. It justifies an executive politics. Empires maintain decisive reservoirs of force-and control of that force is what defines the imperial executive. The advent of the nuclear age placed that power in the hands of the American president, and according to some analysts, thus decisively transformed the constitutional weight of the executive. 7 As a consequence of the tendency toward expansion, imperial regimes are preoccupied by frontiers. And politics in the empire is often made at the frontier and the consequences flow toward the center. Often the interventions are direct: I cited them in my book. Caesar returns from Gaul to descend on Rome; Bonaparte returns at the behest of his political allies from Egypt to seize power in Paris; Britain defines much of its politics under pressure from the challenges of Ireland and of India; Japanese soldiers in Manchuria drag their governments in Tokyo into ever greater ambitions for conquest; de Gaulle organizes the Fifth Republic because the Fourth cannot resolve the issue of Algeria and indeed military rebellion spreads from Algiers to Corsica. General MacArthur raises one of the rare American challenges to civilian leadership after Truman relieved him from his command in Korea; and Senator McCarthy attributes the loss of China to the machinations of Reds at home. The claims of the frontier vie with grievances at home to shape the politics of the Republic. To be sure, U.S. politics undergoes great convulsions because of slavery and economic depression-but so it does as well because of the lure of Cuba, Mexican possessions, our trade across the Pacific and the Atlantic, or our connections in China. Finally, frontiers are never simply frontiers-they are also portals across which the poorer populations of the territories controlled will stream to make a new life within the borders of the empire or, in the case of overseas colonial empires, in the metropole: whether Ostrogoths, or Pakistanis, Algerians, and Hispanics. The imperial syndrome entails a continuing dialogue, but often a violent one, between the interests at the frontier and those at the center.
Empires are thus constructed in a dialectical process with those who resist. Resistance begins where the borders end, and where the claims of rule meet the demands for autonomy. Resistance is endemic; often it seems merely bloody minded, petty, reactionary. It does not manifest itself everywhere, but at least somewhere. To be an empire is usually to confront at least one site of resistance, external or internal. That is one reason why anti-imperialists at home are often so ineffective; they do not like open resistance which is messy, uncontrollable, and requires unattractive allies at home and sometimes supporting enemies abroad. The power that empire possesses can finally be contested only from the streets; and liberals shrink from that unpredictable mobilization and thus are left often to hand-wringing after another fait accompli.
Let me admit that this description of confrontation is too stark. Since describing the dialectic of resistance in my book, one critic has usefully pointed out that for many colonial subjects of an empire (especially the overseas empires that prevailed until after the Second World War), the tactic is not resistance, but a sort of transaction or contestation (what Frederick Cooper terms "claims making")-a struggle within the norms allowed by the colonizers to achieve as much autonomy and influence as possible. The colonial subject carves out domains of relative independence in labor relations, local government, and the like, which can in fact lead to the dismantling of the colonial project. 8 So far we've cited only the dimensions of force and power. But empires expand in pursuit of some big idea: the rule of law or "citizenship," in the case of Rome and Britain; of the Catholic Church in the case of Spain; of culture, or economic growth, or paradoxically even the spread of liberty and democracy. Whether the idea motivates the advocates of empire or merely justifies their ventures gets into non-historical issues such as the nature of sincerity. Any successful empire needs a big idea. Empires thus enlist intellectuals as their justifiers. They support culture.
These great intellectual constructs often have a common structure: they propose at one level a shared interest among rulers and ruled-whether salvation, or economic advance, cultural and scientific or hygienic acquisitions-and justify the at least temporary tutelage of those in charge. At their base lies a conviction of what post-colonial writers have termed "difference." Only the most predatory empires, such as the Third Reich, have suggested that those conquered have no benefits to gain from being ruled by conquerors. Nonetheless, the more racially constructed such ideas of hierarchy are, the less reciprocity they will allow.
Empires have another potentially beneficial value; they can nurture group tolerance, granting religious pluralism or special role for diasporas; they also allow for enclaves of autonomy within their extensive spatial domains. They often welcome immigrants, especially those from the peripheries that they dominate. North Africans, Pakistanis, Latinos have flowed into the countries which have often dominated them. This does not mean that empires do not assign such migrants inferior roles or that they erase racial prejudice. And where migrants achieve high status or play key socioeconomic roles within empires, as Armenians or Jews or overseas Chinese, they sometimes face murderous backlashes from either other subject peoples or the dominant ethnicities of the empire. Still, nation-states often impose greater conformity on minority entrants. They are high on indices of belonging but potentially low on tolerance. Empires can be high on official tolerance, but low on belonging.
All this means that the imperial syndrome is built on the confidence that somehow one's own state is exceptional, cannot be called to account by the others and thus should not-that it obeys a higher law. "Trust us; we're different." American exceptionalism has had a long and venerable tradition-but we usually think of it as Tocquevillian exceptionalism, or that spelled out by Louis Hartz a half century back: the absence of feudalism, the existence of religious pluralism, welcome to immigrants, and the vast reserves of free and open land. But I am talking about a sort of less appealing exceptionalism-the belief that great power grants great rights-perhaps call it the American Sonderweg. Both the British and the Germans had this confidence before l9l4 as did most other large states. There was arrogance, to be sure. But what made the attitude even more dangerous was the inner reassurance of virtuousness, the belief that ultimately one's own country's behavior was more responsible than the others'. So, too, the conviction that whatever abuses might be uncovered-whether Herrera massacres in German Africa, or concentration camps in the Boer War, or Abu Ghraib and Guantánamo detentions-were atypical exceptions.
The imperial syndrome also involves a particular relationship of rulers to ruled, of those who govern to their own population. An imperial regime searches not for discussion and deliberation, but for approval and acclamation. It measures popularity. The media replace parliamentary debate, and if there is any symptom of empire it is the attrition of representative bodies. Perhaps they are formally kept in being. Even Hitler preserved a mock Reichstag. But even in less pathological states, government by debate loses its integrity and capacity. Granted, these tendencies afflict modern democracies in general, especially when they face complex social choice. Decisions get passed to courts as well as to legislatures. But the attrition of legislative procedures in a democracy usually arises from the complexity of issues; in a proto-imperial situation it results as a response to alleged security dangers. Parliamentary delegates accept the executive's diagnosis of danger rather than risk being seen as anti-patriotic. They pass blanket delegations of power. And even if they insist on legislation, the executive claims the right to interpret the laws that they might pass. Popularity becomes the ultimate measure. Now if those ruling fail in their enterprises they can lose popularity very quickly. But until public opinion turns adverse, acclamation, photo-ops, spectacular games, staged pageants replace debate. Now there are exceptions: rule by committee or by party can continue, as it did in the French Third Republic. But even here the issues that define empire and foreign policy are withdrawn from the arena of debate and discussion. The executive, individual or collective, reserves more and more of them. Empire, like authoritarian government more generally, involves the rule of the exception: there is always an exceptional danger that defines imperial politics; and the imperial syndrome embodies Carl Schmitt's notion that he who controls the exception in effect controls even democratic politics.
The imperial syndrome involves a rampant growth of privilege and inequality that corrupts an earlier civic spirit. This does not mean that measured by Gini coefficients or other statistical indices society is less equal as a whole. Empires can be democratic at home-the British expanded suffrage as they expanded their empire; the French third Republic was Europe's most democratic regime and it conquered Vietnam and Morocco-but empires cannot let their subject peoples share the same democratic ground rules. And even as they may extend formal equality, and even income equality toward the bottom, they give the top immense new opportunities for enrichment. This presents grave difficulties of judgment. If millions of middle-income families are each given a small tax rebate, while at the same time several thousand wealthy citizens can each reduce their bill by thousands or millions, the legislation may increase formal measures of equality because of the mass of less affluent citizens affected. But who can doubt which distribution has a greater impact on civic participation, on the control of the media, or the sense of a gulf that separates ordinary citizens from those who emerge enriched? One of the curiosities of American public discourse is that growing income inequality, which is often commented on (although so far hardly contested), is discussed solely as a domestic issue. Few commentators who are not considered on the radical fringe make a connection between the growth of inequality within the United States and the claims that the country had made for international primacy.
However, this is the transaction that the imperial syndrome usually involves: not robbing the poor to pay the rich, although the periphery may be despoiled to pay the center, but fobbing off the humble so that privilege becomes more and more spectacular. For a while public games, reality TV, philanthropy and the admirable but hardly taxing (indeed often taxexempt) charitable deeds of those enriched may counteract the emergence of populist class politics. How long that lasts is not at all clear.

III.
These criteria constitute the imperial syndrome. Can a country have the syndrome but not be an empire? That question I adjourn. Let me turn rather to the taxonomy of empires. Here I wish to make a simpler case, and I will make it briefly because it is familiar. Historians traditionally have separated landed empires from seaborne empires. The Romans, the Chinese, the Carolingians, the Turks, the Russians, the Ottomans, were land-based empires. I call them agglutinated empires. They pressed outward from their heartland and annexed territories or otherwise subjugated them. These were the empires that finally succumbed in the early twentieth century, although it could be argued that the Russian one merely changed its governing principles. Most territories were in one continuous unit although occasionally there could be separate chunks of land, and often, in the case of the empires around the Mediterranean, the shape of the territory was very irregular.
Crucial to these units was that rule was relatively homogenous. The subjects in the national capital were under the same sort of regime as those in outlying districts. Indeed in some instances the outlying districts enjoyed special collective privileges or degrees of autonomy that the subjects at the core did not.
The seaborne empires were structures of a different sort. They often involved, from the Athenians on, the planting of colonies overseas. Their landed core could be far smaller than the vast territories over the ocean, as in the case of Venice, Portugal, Britain or the Netherlands. The unstable frontiers they faced were usually the borderlands of their overseas colonies. Their major strategic and financial problems involved sea communications and raising taxes from the colonial regions. Indeed where they faced continuing warfare and struggle was at sea; for by convention and practicality the oceans could not be cordoned off into stable corridors. Instead various exclusionary principles were sought and enforced or contested, perhaps most famously the asiento or right to import slaves into the New World.
In these empires the residents at home, in the so-called metropole, often enjoyed a level of participatory government that was denied either the settlers from abroad, or the indigenous people who were subjugated. Indians did not control India the way the British controlled their home islands. Queen Victoria could never have made herself empress of Great Britain.
The purest specimen in modern times of the overseas empire was the Portuguese, which began with the conquest of Ceuta in 1415 and outlasted most all of the others, retiring from Mozambique and Angola in the mid-1970s. Essentially the Portuguese captured portals for trade and dealt with the compradors who could control the resources of the interior, whether spices or slaves. Where territories were sparsely settled-preeminently Brazil-annexation of vast interior realms might follow. The Dutch, the French in the Caribbean and the Indian Ocean, the British, eventually the Japanese and the United States, were seaborne empires that developed enclosed colonies. Still a remarkable transition occurred in the late nineteenth century. Control of the portals, so to speak, no longer guaranteed access to the interior, and colonies had to be enclosed. This development paralleled the general fixation with land masses and their permeation, expressed in continental geopolitics, the technology of the railroad, the growing rigidity of alliance systems-indeed a whole new territorial consciousness.
Alongside of this very familiar contrast, I would place what now are called nomad empires. We have tended to think of these as exceptional, but in their extent and their frequency they are certainly as distinctive a form as either the agglutinated or the seaborne empires. They have been called "shadow empires," "mirror empires," and "vulture empires," in Thomas Barlow's formulation of the Mongols' symbiotic relationships with the huge Chinese state structures they co-exist with; and historians are starting to analyze them more rigorously, as Pekka Hamalainen has done in his splendid book on the Comanches. As Hamalainen summarizes, "nomads can destroy empires, but they can also build enduring empires." 9 To do so, requires a partial transformation at least into sedentary agricultural confederations. What distinguishes them is a often a clan or tribal structure, in which descent from an original chief or charismatic founder provides a principle of ethnicity and also a concept of territory that does not require a fixed-line frontier. The Arabian tribes that exploded across the Middle East, the Maghrib and into Spain in the century after the death of Mohammed are one instance. The Mongols who surged out of the steppe both East and West, the Ottomans or descendants of Osman, who overran Anatolia and eventually the Balkans and the Arab states, and the Timurids who two centuries later controlled so much of Central Asia and might have conquered China were it not for the sudden death of Timur are another. As Hammaleine has suggested the Comanche, perhaps the other great confederacies, Lakota and Iroquois, were also variants.
What these peoples claimed was not a well bordered state, but a paramount influence over a territory with fuzzy edges, sometimes boundaries that shifted with their flocks or the migrations of the wildlife they hunted. What they developed with immense skill was the technology of rapid movement-preeminently by horseback and archery. What they found difficult over the long run was maintaining the dynamic and often brutal principles that helped their rise-whether the fratricidal succession principles of the Ottomans, or the restless drive to conquest. Here the Mongols and the Timurids were revealing. The attractions of the wealthy, sedentary and cultivated states they might conquer were often overpowering. In some instances, such as the Manchu who formed the Qing dynasty, they tried to maintain principles of ethnic separation and an ancestral homeland even as they insisted on principles of preeminence over a landed area, treating the steppe often as the seaborne powers treated the oceans. Although termed "shadow" empires when they arose in the wake of a settled megastate, or mirror empires, even vulture empires-which sometimes both extracted resources and paid tribute-they were network societies, not homogenously settled units. They absorbed migrants and captives; and as Hamalainen shows, they were vulnerable to ecological collapse or to the changed relations between the neighboring states that bordered them. We could call them fuzzy-bordered empires. In geographical extent they were the most imposing of potential state structures, but as the concept of bounded territory became more preoccupying, they found that their fluidity was no longer the advantage it had been earlier. Administration of vast spaces was also difficult, so they often partitioned their huge conquests-as did the Mongols and Timurids.

IV.
Nomadic structures, or fuzzy-edged states or confederations are generally seen as relatively primitive. But in some ways they anticipate a future in which deterritorialization makes ever greater strides. This introduces some concluding reflections. The great tribal confederations were in effect pre-bordered states. They flew in the face of the territorial concepts that emerged in the early modern period although they maintained a clear sense of their own vaguely bounded turf. In this spatial aspect, so fluid, and often so extensive, however, they are more post-modern than the imperial structures preoccupied by enclosing great swathes of territory, painting maps red or blue, and insisting on a homogenous degree of continuing unshared domination. If the United States is to be imperial, it will have to be on some of the same principles. Pre-territorial and postterritorial empires share the fact that within a given area of the earth's surface there can be overlapping regimes and communication. Such a structure seems propitious for an age in which global communication, finance, and deterritorialization has swept us along. The Islamic empires were created originally with the deterritorialized force of religious fervor. The Turkic empires (Timurids) depend upon the fluidity of boundaries. This stage often does not last long since the dynamic conquerors settle down to "farm," so to speak, the wealthy, sedentary and cultured states they have conquered.
Ultimately empires have been forms of social, political and economic organization that have relied on the control of extensive space to guarantee their elites' domination in given areas. Nation-states coveted space to assure what they believed were the conditions of sovereignty and survival for communities they felt had some long term historical mission or role to play. Empires covet space sometimes for its own sake-it is intoxicating-sometimes because horizontal extension seems to guarantee vertical ordering. Such aspirations today may be hopelessly anachronistic. Who can plausibly dream of world or continental conquest? Commentators whom I admire have claimed that the age of empires has ended. Frederick Cooper seeks, I think, to downplay the invocations of empire in order to focus on colonialism. 10 I would agree that the age of colonial empires is over, but while the concepts of empire and colonial empire may overlap, they are not, as Cooper himself explains, identical.
My own view is that a persistent residue of territoriality still clings to world politics and to American ambitions for international order. Even in the age of the internet and all the processes lumped together as globalization, global power is always contested in specific places-whether those sites be on the perimeter of control or in the heart of the metropole's cities. It is a mistake to think that globalization and all it represents has suspended the importance of bounded territory and location. Political contests are always contests over place. Without being able to control the security of town centers, or outlying districts, governments are seen to be feeble. Control of public space remains crucial, even as so many developments in the media and the economy transcend fixed territories and spill across boundaries.
The U.S. imperial zenith may have been only a brief artifact of the cold war and the collapse of rivals-first the collapse of the Japanese and German empires in 1945 and then of the Soviet imperium in l989-91. Who knows where we shall emerge after the current economic turmoil? In any case, the long-run inheritor of Japan's empire of the l930s and l940s is not the United States, but China and Indonesia or perhaps eventually some Asian regional entity. Having identified these behavioral properties, which I believe the United States reveals along with empires of the past, one can take up the question again: whether America is or is not an empire. I've attempted frankly to finesse this question, which I think must always be dependent upon definition, to focus instead on structural similarities and patterns of behavior-on analogue and syndrome. Empire-for all of its variations-has been an enduring political form, a historically compelling method by which multiple political authorities can divide up the surface of the globe on which we live, given that diverse ethnic units have had vastly different levels of development. Empires have been major components of global politics since antiquity. Certainly the United States belongs among the ranks of the powerful global actors: it illustrates comparable structural and functional features, and it has behaved in some if not all comparable ways.
At the end we have two puzzles: one about the past, the other about the future. The one about the past, which I have not attempted to answer here is not about single empires but about the prevalence of empire in general. Why are certain epochs conducive to the domination of global space by imperial forms? And why are others not so favorable?
Consider the era from 1500 to 1650, which comprises perhaps the apogee of landed and nomad empires, encompassing the Americas, much of Europe, the Mediterranean, Central Asia, South Asia, and East Asia. Consider the centuries from 1700 to 1950 as an apogee of seaborne empire. We know empires decompose and reconstitute themselves. What forces provide such widespread transformation: is it technological opportunity: the horse, artillery, the railroad and the steamship? 11 Because if it is, then the future of empire will depend upon these factors as well. In general, though, I believe that empires generate other empires-whether competitively or in the process Barsfield outlines. They tend to prosper or decline in tandem, and in rhythms that we haven't fully identified.
As for the future, if compelled to wager, I think that the most likely long-term organization of world politics will involve increasing levels of supranational association without imperial hierarchization, and the development of regional associative blocs, much like the European Union. But there is an alternative in which the functions of governance are decomposed and overlaid on each other, much like those celluloid maps that fancy atlases allowed so that the reader could superimpose ethnicity, economic resources, religious loyalties, and political organization. In such a future, sovereignty and functionality would be decomposed so that no encompassing political control existed. Territoriality would lose its meaning in that situation, there would be zones in which respectively media, communications, religious federations, financial networks, and sports federations would slide on top of each other-none claiming oldfashioned sovereignty, each claiming the right of unimpeded pursuit of its own objectives, as do market-oriented economic units today. Curiously enough, the great edgeless empires perhaps allow us to glimpse that sort of world. They remind us, too, how violent such a world might become. Hence three possible futures seem plausible: a coming global order with multiple EU-like units; or alternatively, governance organized in terms of layered functionality; or perhaps, finally, if economic difficulties persist, the world redivided into revived national protective units. Even if empire has no future, its history will remain relevant because even today spatiality remains a resource for power and for controlling future life outcomes. Empire has represented what might today be a dream and perhaps even an irrelevance: that society's life is secured by controlling land and oceans. But in undreamed of forms, perhaps the history of empire begins anew. NoTEs