WHAT'S IN A NAME?

It is rare in popular usage to distinguish between imperialism and colonialism. They are treated for most intents and purposes as synonyms. The same is true of many scholarly accounts, which move freely between imperialism and colonialism without apparently feeling any discomfort or need to explain themselves. So, for instance, Dane Kennedy defines colonialism as “the imposition by foreign power of direct rule over another people” (Reference Kennedy2016: 1), which for most people would do very well as a definition of empire, or imperialism. Moreover, he comments that “decolonization did not necessarily involve the rejection or negation of imperialism, nor did it cause empires to entirely disappear from the scene” (ibid.: 6, my emphasis).

The great scholar of empire, Ronald Robinson, was similarly cavalier about the distinction between imperialism and colonialism. In one of his most influential articles, “The Eccentric Idea of Imperialism, with or without Empire,” he moves effortlessly and as it were unconcernedly between imperialism and colonialism, so that “free trade imperialism” and “informal empire” jostle “colonial empires” and “indirect types of colonialism,” while “the paradox of imperialism after empire” can be investigated as a matter of “post-colonial issues” in the “post-colonial period” (Reference Robinson, Mommsen and Osterhammel1986: 267, 273–74, 279).

Jürgen Osterhammel, in his monumental work The Transformation of the World: A Global History of the Nineteenth Century, is somewhat more circumspect. He sees colonies as one variety of the “peripheries” that define all empires, and warns that “not all imperial peripheries were colonies,” giving as examples the autonomous white dominions such as Canada in the British Empire, and Finland, a “semi-autonomous grand duchy” in the Russian Empire. Though dependent, they were not colonies like Jamaica or Turkestan. At the same time, Osterhammel feels no need to make any kind of typological distinction between imperialism and colonialism. Colonies are simply a particular kind of sub-species of empire, and “colonialism is but one aspect of nineteenth-century imperial history” (Reference Osterhammel and Camiller2014: 431). Empire and colony do not stand as independent or alternative possibilities.

One last example is D. K. Fieldhouse's well-regarded The Colonial Empires (Reference Fieldhouse1982), which, though it eschews definitions of both imperialism and colonialism, is explicitly concerned with European overseas colonization and the construction of new kinds of empires based on the possession of colonies. He is aware of other types of empires, mostly land empires of the kind found in the East and in older times. For him, the European colonial empires are different, and eventually, in the nineteenth century, come to dominate the world. But Fieldhouse sees no need to distinguish colonialism from imperialism, and indeed states as the purpose of his book “to distill … the essential features of modern European imperialism” (ibid.: x). Throughout the book he writes of “empires” and “imperialism” as if colonial empires—empires formed through colonization—require no special designation or analysis. They are in that sense, as for Hans Kohn (Reference Kohn, Strausz-Hupé and Hazard1958:3), simply a later instalment of the venerable empire story, empire “by other means,” not the introduction of a new principle.Footnote ¹

While more recent students of empire have apparently felt little need to distinguish between imperialism and colonialism, it is important to recall that in one very influential tradition of the study of the British Empire, in particular, this distinction was very much insisted upon. In his famous book The Expansion of England (Reference Seeley1971[1883]), Sir John Seeley distinguished between the colonies formed by the settlement of mainly British people—the “white dominions” of Canada, Australia, New Zealand, and South Africa—and the other parts of the British Empire where Europeans formed a small minority, such as British India. In fact, for Seeley, “properly understood,” the British Empire was “not an Empire at all in the ordinary sense of the word. It does not consist of a congeries of nations held together by force, but in the main of one nation, as much as if it were no Empire but an ordinary state” (ibid.: 44). The “British Empire” for Seeley was thus constituted principally by the colonies of settlement, the white dominions. The British Empire, that is, was tantamount to what Sir Charles Dilke had called “Greater Britain,” an extension of British nationality, united to the home country by the ties of blood and held together, like all states, by a community of race, religion, and interest. Seen in this way, it was not so much an Empire as traditionally understood but “a vast English nation” (ibid.: 63; see also Hobson Reference Hobson1988[1902]: 6–7; Bell Reference Bell2007).

Despite the forthrightness of these assertions, Seeley in much of his book wrote of the British Empire as a whole, in terms not very different from those of his contemporaries. He did distinguish the “Colonial Empire,” the white settlement colonies that together with Britain made up “Greater Britain,” from the “Indian Empire,” whose subject peoples are for the most part “of alien race and religion, and are bound to us only by the tie of conquest” (Reference Seeley1971: 14–15). These two parts of the empire had different destinies, the first to cement ties even more strongly with the mother country, the other eventually to be granted independence. But Seeley did not take the final step, implicit in his analysis, of separating colony from empire, colonialism from imperialism. Some of his followers, such as the famous classicist and prominent intellectual Gilbert Murray did do so when, agreeing that the “British Empire” is a misleading term, Murray wrote in 1900: “Empire’ is the rule of one nation over other nations. We hold empire over India, over Soudan; we do not hold empire over Canada or Australia” (quoted Kumar Reference Kumar2017a: 337). This more or less direct echo of Seeley conformed to his precept but went further in finally separating the “Empire” from the settlement colonies.Footnote ²

This usage was challenged vigorously at the time by those, such as the poet Alfred Lord Tennyson and even Sir Charles Dilke, who could not imagine the British Empire without India, “the jewel in the crown.” And as we have seen it was a distinction that has found little favor with later students of empire. But one at least, another famous classicist, Moses Finley (Reference Finley1976), attempted systematically to distinguish colony from empire, colonialism from imperialism. Finley mentions Seeley only in passing, and on an unrelated point, but much of what he has to say is in the same spirit. Finley's discussion, moreover, is unusually comprehensive, covering many examples of empire both ancient and modern. It raises many of the questions commonly asked about imperial rule and the relations between rulers and ruled in empires. It allows us to ask what might be gained, and what lost, if we accept the distinction between colonialism and imperialism. In what ways are colonies different from empires, colonialism from imperialism? Do they operate according to different principles? How should that affect our study of empires? This paper will focus mainly on Finley's account since his is one of the few attempts to deal systematically with this neglected area.Footnote ³

A GLOSSARY OF TERMS

We need first to establish some primary meanings of our terms, to understand their origins and common uses. Our discussion will relate mainly to the Western empires, though we shall see that it has important implications for how we view other, non-Western empires.

For all Western cultures, the terms empire/imperialism and colony/colonialism derive from Rome and the Latin language. The first come from imperium, the second from colonia. Imperium for the Romans meant more or less absolute rule, originally in the military, later in the state. Military commanders and rulers of empires—imperatores—have imperium. Before long, the state or territory could itself be described as an empire, as for instance the imperium Romanun—the Roman Empire. With that use, imperium acquired what at that time and later became the accepted meaning of empire, namely, a large state exercising authority over a multiplicity of peoples and territories (Koebner Reference Koebner1961: 11–16; Kumar Reference Kumar2017a: 7–13).

While the actual term in most of the Western languages (English empire, French empire, Italian impero, Spanish imperio) clearly derives from this Latin root and Roman usage, it has not proved too difficult to apply it to non-Western cases.Footnote ⁴ Thus scholars speak freely of the Chinese Empire, the Mongol Empire, the Arab empires of the Umayyads and Abbasids, and other Islamic empires such as the Mughal, the Ottoman, and the Safavid. So too it has been possible to apply the term to ancient, pre-Roman cases, to speak of the Akkadian, Assyrian and Babylonian empires, the Egyptian Empire, the Persian Empire, and the empire of Alexander the Great. “Empire,” most people seem to agree, applies to not just the Roman Empire but also, through the concept of translatio imperii—the passing on or handing over of empire—to other European empires, and also to ancient and non-Western ones. It is true that the applications of the Western term “empire” to these diverse cases is a complex process that can be traced in detail. For instance, only in the late nineteenth century was the two thousand-year-old Chinese state designated an empire, mainly through a borrowing from Japan, which had itself taken the concept from the West. We also must be sensitive to the different inflections of “empire” in different languages, within different cultural traditions. But it is clear nonetheless that we cannot do comparative history or historical sociology without this all-encompassing understanding of empire.Footnote ⁵

“Imperialism” is a different matter. “Imperialism,” says Michael Doyle matter-of-factly, “is simply the process or policy of establishing or maintaining an empire” (Reference Doyle1986: 45). So, we can and do speak of “Roman imperialism,” and even “Athenian imperialism” (e.g., Champion Reference Champion2004). Neither the Romans nor the Greeks had a word for imperialism, as opposed to “empire,” which suggests that there may be an interesting and important distinction to be made. Might it be that a political entity can be an empire—ruling over a multiplicity of peoples—without pursuing imperialism? Whatever the case with Rome or Athens, it seems that “empire without imperialism” might be one way of characterizing the Chinese Empire before the period of the Qing dynasty (1644–1912). China was clearly an empire from the time of its unification in 221 BCE under the Chin “First Emperor” (Shi Huangdi), but for a long time it was not expansionist and made little effort to push its boundaries beyond those established by the Qin and Han dynasties (221 BCE–220 CE). Only with the Manchu dynasty of the Qing did what seems to be a definite policy of imperialism appear, resulting in the conquest of Tibet, Mongolia, and Xinjiang (Perdue Reference Perdue2005).

It is significant that when Christopher Bayly defines imperialism as “the ruthless drive for dominance” (Reference Bayly1998: 28), or Thomas August calls it “an insatiable quest for territory” (Reference August1986: 85), both authors clearly link it to the overseas conquests of the Europeans from the fifteenth century onward. Only from about that time do they think it appropriate to characterize policy by that term. It does seem as if “imperialism,” as opposed to “empire,” was late in coming, as if there was no felt need for it as a concept for a long time. According to the detailed study by Richard Koebner and Helmut Schmidt, the word “imperialism” first made its appearance only in the mid-nineteenth century, and then with the highly negative connotation that has marked its use for much of the time since. It seems first to have been used by the British to express their revulsion at the “despotic” regime established by Louis Bonaparte and his Second Empire following the coup d’état of 1851. Imperialism, seen by the British as an unfortunately persistent French trait, here recalled the equally despotic regime of the First Empire created by Napoleon Bonaparte, Louis's uncle (Reference Koebner and Schmidt1965: 1–26).

Since both Bonapartes followed adventurist and expansionist policies, imperialism at its first appearance connoted not just populist authoritarianism and Caesarism—“Bonapartism”—but also the effort to “export” these principles abroad, to extend the French Empire by force. It was in this sense that it was used by liberal critics such as William Gladstone to attack Benjamin Disraeli's policies of “imperialism” in the 1870s, seen as threatening peace abroad and freedom at home (hence the opposition to making Queen Victoria Empress of the whole British Empire—in the end she was Empress of India only, and for other British territories she was simply their Queen). “[Imperialism] first became a popular word in the English language as an anti-Disraeli slogan” (ibid.: 134).

“Imperialism” for much of the twentieth and twenty-first centuries has come trailing the clouds of opprobrium in which it was first wreathed.Footnote ⁶ A widely accepted starting point of this career is J. A. Hobson's Imperialism (Reference Hobson1988[1902]), whose influential criticism was taken up by Rudolf Hilferding, Rosa Luxembourg, Lenin, and other Marxists (Etherington Reference Etherington1982; Wolfe Reference Wolfe1997). In the lexicon of Third World Marxism it became a standard term of abuse, mostly applied to Western powers, including the United States. Here it was often conjoined with “colonialism” as a kind of synonym, though increasingly colonialism supplanted imperialism (Howe Reference Howe2002: 26–27). The implications of that shift are significant.

To turn then to colonia, the parent of “colony” and “colonialism.” The Latin word colonia stems from the verb colere, “to cultivate, to farm.” Colonia came from colonus, who was “a tiller, cultivator, a farmer, a planter,” and also “a settler in a new country,” an “outsettler, a colonist.” The Oxford English Dictionary (OED), noting this etymology, continues: “Latin colonia had thus the sense of ‘farm,’ ‘landed estate,’ ‘settlement,’ and was especially the proper term for a public settlement of Roman citizens in a hostile or newly conquered country, where they, retaining their Roman citizenship, received lands, and acted as a garrison.… Hence it was applied to the place so occupied, or to towns which were raised to the same rank and privileges” (OED 1989: s.v. “colony”).

While colony in this sense is relatively old, “colonialism,” like imperialism, is relatively new. According to the OED, colonialism, as “the practice or manner of things colonial,” was first used in the second half of the nineteenth century, and examples are given from 1864 and 1883. It was not, though, until the 1880s that colonialism was used to mean “the colonial system or principle.” Interestingly, unlike imperialism, colonialism did not initially have negative connotations, being used in a fairly neutral way. It took twentieth-century developments to give it its largely negative charge. The OED, giving examples from the period 1949–1957, notes that colonialism is “now frequently used in the derogatory sense of an alleged policy of exploitation of backward or weak peoples by a large power” (ibid.: s.v. “colonialism”). It was in that pejorative sense that colonialism came to be the preferred term, supplanting imperialism, for many critics of the Western empires (Howe Reference Howe2002: 25).Footnote ⁷

MOSES FINLEY: DEFINING AND SITUATING COLONIES

It is very much this etymology and this history that Moses Finley draws upon in attempting to give “technical” meaning to the concepts of colony and colonialism, and to distinguish them from empire and imperialism. He notes, again following the OED, that the word “colony,” in the Roman sense, first appeared in the modern European languages in the fourteenth century, and by the sixteenth century it was being applied to the “planting” of settlements in newly discovered lands. It was only then, he says, that “colony” and “plantation” came to be synonyms, the one standing for the other (in Roman usage, he says, it was rare for “plantation,” plantatio, to be used of people; usually it indicated plants or crops). An early, and fully developed, use of colony in this sense occurs in Thomas More's Utopia (1516), and Francis Bacon's celebrated essay “Of Plantations” (1625) is another classic example. By the mid-eighteenth century it was so common that Adam Smith, in The Wealth of Nations (1776), could begin the chapter “Of Colonies” with the statement: “The Latin word (Colonia) signifies simply a plantation” (Finley Reference Finley1976: 170–71, 179).

This dating of terms makes it very clear that the early-modern understanding of colonies as plantations was linked to the first wave of European settlements overseas—mostly in the Americas—in the sixteenth and seventeenth centuries. “Colonization”—the practice of planting colonies—also dates from this time. It may have taken much longer for the felt need for a term, “colonialism,” to describe the principle or philosophy of colonization. But from the sixteenth century onward Europeans were energetically engaged in planting themselves in all the corners of the world, creating “Neo-Europes” everywhere (Crosby Reference Crosby1986: 3).

Drawing upon this understanding, Finley lays out two main features of colonies. First, “For more than three hundred years, however much disagreement there may have been about the objectives of colonization or about the ways of governing colonies, there was complete agreement that a colony was a plantation of men, to which men emigrated and settled. Colon in French, Siedler in German, make the same point” (Reference Finley1976: 171).

Secondly, “There was also, in those three hundred years, complete agreement that a colony was not only a plantation but also a dependency of the country from which the emigration was initiated.” Colonization therefore cannot be equated with just any emigration, with, say, the Chinese diaspora in Southeast Asia, or South Asians in East Africa or the West Indies. Nor can it include enforced emigration, even where the migrants come to be the majority, as with the African slaves in the Caribbean or Brazil, so long as the migrants are not under the control of the country of origin. The settlements must be and remain dependencies of the mother country, even where there may be considerable numbers of migrants from other countries (“a complication,” says Finley, “which I cannot discuss”) (ibid.; see similarly Anderson Reference Anderson2006: 188–91).

Finley therefore argues that “the history of colonies is surely the history of the ways in which the power, prestige and profits of some countries were enhanced (or so they hoped) by external dependencies of migrant settlers” (Reference Finley1976: 174). This is clearly a very mercantilist view of colonies, reinforcing the point that this concept of colony was elaborated in the very particular period of the sixteenth to the eighteenth centuries when mercantilism dominated in the thinking of European rulers. This does not prevent its being applied more generally, in time and place, but it might indicate some of the problems associated with it, as we shall see.

Finley is aware that his definition of colonies is not new, nor even particularly problematic (see, e.g., Horvath Reference Horvath1972). Indeed, he is at pains to stress its antiquity. But he is concerned that, at some point in the late nineteenth century, “colony” lost its specificity and became simply a synonym for any type of “dependency,” including such as are to be found in most instances of empire (Reference Finley1976: 170). A valuable distinction was hence lost. What is particularly important for him is what his “technical” concept of colony excludes—where, despite both popular and scholarly practice, it is improper and misleading to apply the term.

Thus, contrary to much common usage, Finley denies that we can speak of “colonies” in the ancient Greek world. That is because what are usually referred to as colonies in that context were not dependencies but independent communities. “The so-called Greek and Phoenician colonies of the eight, seventh and sixth centuries B.C., extending from the coasts of the Black Sea to Marseilles and Carthage, were more peaceful enterprises in some instances, less in others, but what is essential is that they were all, from the start, independent city-states, not colonies” (ibid.: 173–74).Footnote ⁸

Further, since “there can be no colonization without colonies,” Finley rejects as colonization “the extensive Macedonian and Greek migration into the territories of the Persian empire conquered by Alexander the Great.” This is because, after Alexander's death, these communities became independent kingdoms, the Hellenistic kingdoms of Egypt, Syria, and others (ibid.: 173). But there seems no reason, on his definition, to deny that these were indeed Greek colonies while Alexander still ruled his empire, though admittedly his early death meant this was a short-lived empire. Finley seems to be on firmer ground when he denies that the barbarian invasions of the Western Roman Empire, or the Norman conquest of England and Sicily, can be held to be colonization, since in these cases independent kingdoms were set up from the start (though there could be some quibble in the case of the Norman conquest of England).

On the same grounds—absence of dependency—Finley dismisses the concept of “internal colonization,” or “internal colonialism.” “No one,” he says, “speaks of the colonization of the midwest and west of the United States” (ibid.: 173). Finley must have been aware that, particularly in the context of the U.S. intervention in Vietnam, there had been much discussion of the “American Empire,” but since colonialism was not for him imperialism the westward movement of white Americans—clearing away the indigenous people as they went—did not constitute a case of true colonization. Conquest, per se, is not colonization.Footnote ⁹ The settlers created communities that became Territories, then States of the Union, not dependent colonies.

Michael Hechter's Internal Colonialism: The Celtic Fringe in British National Development, first published in Reference Hechter1975, and which did much to popularize the concept, perhaps had appeared too recently for Finley to be aware of it (he does not refer to it). But he would have felt reasonably confident in resisting the term colonization, if not conquest, for describing the process of English domination of the British Isles that Hechter examines. For, except perhaps in Ireland, it was not mainly through the planting of English settlers that English rule was established. Wales, Scotland, and even Ireland were not English colonies.Footnote ¹⁰ There may be good grounds for speaking of an “English Empire” (e.g., Davies Reference Davies2000), but it was not an empire that was acquired through colonization. Similar objections, by Finley, would presumably apply to Alexander Etkind's Internal Colonization: Russia's Imperial Experience (Reference Etkind2011). Etkind tells a story, as with England, of Russian expansion through empire, but it was not mainly through the establishment of dependent settlements of ethnic Russians in different parts of the empire (Siberia partially excepted).

Finley excludes as colonies a number of other celebrated examples, mostly from the medieval period. These include the Crusader States in Palestine, German expansion east of the Elbe, and the Norman conquest of Sicily. For Robert Bartlett (Reference Bartlett1994), these are indeed all examples of both “conquest and colonization.” But not so for Finley, since the kingdoms and principalities that resulted from these movements “were never, not even in inception, subordinate to anyone in the territories from which the migrants came” (Reference Finley1976: 176). Once again, no dependency, no colony.

If the exclusion of the overseas Greek settlements as colonies is one eyebrow-raising proposition, then the exclusion as such of most of the European possessions in Sub-Saharan Africa is probably the most deliberate provocation. Hundreds of books exist with titles such as Colonialism in Africa, 1870–1960 (Gann and Duignan Reference Gann and Duignan1969). For many people, when they think of European colonialism, it is Africa that they principally have in mind. Perhaps because of the influence of such books as Joseph Conrad's The Heart of Darkness (1902), or the many popular treatments of “the scramble for Africa” in the 1880s, or innumerable Hollywood films of the 1930s and 1940s set in “darkest Africa,” Africa has stamped itself, as an emblem of colonialism, more firmly on the European imagination than any other part of the world.

Finley will have none of this. It follows directly from his definition of colony that most of Africa, though ruled by Europeans, was not colonized by them. That was because in the majority of cases few Europeans settled in the African territories. There were, that is, few European “plantations” in Sub-Saharan Africa, hence few European colonies. Finley is prepared to accept that that one might designate as colonies those few areas where a significant number of Europeans were to be found, even if they were not, as they were not anywhere, a majority. South Africa is one such case, with its large numbers of Boers and British, so too are Kenya and South Rhodesia, with their relatively large British populations. So, “in my categorization, Kenya was a colony, Uganda and the Gold Coast were not. Nor were the Congo, Senegal and the Ivory Coast.… The struggle for Africa was not, or at least not in large part, a struggle for colonies” (Reference Finley1976: 184). The British had a few colonies in Africa, in the south and east; but none of the French, Portuguese, Belgian, or German dependencies in Sub-Saharan Africa can be called colonies.

The important qualification, in which a significant number of Europeans ruling over a majority of non-Europeans can be considered as forming a colony, allows Finley to designate as colonies a number of European settlements in North Africa, most prominently French Algeria. This is in one way problematic, since, as Finley knows, the French themselves did not consider Algeria a colony but, since 1848, as part of metropolitan France. Administratively it was in this sense similar to the position, in the United Kingdom, of Ireland, Scotland, and Wales, which Finley declares emphatically were not colonies but sub-divisions of the kingdom.

But, remarking airily that “administrative definitions are essentially unhelpful,” Finley goes on to say that “Algeria was a fully incorporated department of metropolitan France, yet it was indubitably a colony.” This is because “in every respect other than the administrative, [the non-European] Algerians in the overwhelming majority still considered themselves, more than a century after the conquest, to be the exploited subjects, not so much of the metropolis as of the settlers backed by the coercive power of the metropolis” (ibid.: 187). Finley here adds a crucial dimension to his idea of colony. It is now not so much, or not only—as in the case of the British in North America, Australia, or New Zealand—a matter of the absolute number of Europeans, replacing or subjugating a diminishing number of indigenous peoples and making themselves in time the majority. Now it becomes also or instead a matter of a certain relation between settlers and the indigenous people, as well as between the settlers and the metropolitan power. For Finley, the “paramount distinction” among European overseas possessions “centres around the extent to which the settlers have both reasons and the power to determine policy, not only against the indigenous population but, even more important, against the metropolis” (ibid.: 186). Thus, by this criterion, though not by the “majority rule,” South Africa, Rhodesia, Kenya, and Algeria, with their powerful settler communities, are colonies. So too—though he refers only glancingly to them—might be Portuguese Angola and Mozambique, with their significant populations of Portuguese (ibid.: 170).

Finley here introduces considerations that might seem to threaten to undermine the old, early-modern concept of colony that he wishes to bring back. It might be very difficult to determine that the power of the colons in Algeria and their relation to the indigenous people differed categorically from the clout of the admittedly thinner stratum of Dutch residents in the Dutch East Indies and their relation to the indigenous Indonesians. Yet one is regarded as a colony, and the other not. On the other hand, it is true that the French in Algeria, like the British in Kenya or Southern Rhodesia, regarded themselves as permanent settlers in their respective countries. They felt themselves to be as Maghribi, or as African, as were the non-European populations, if in somewhat different ways. They and their families were there to stay. It was this consciousness that was expressed so powerfully in Algerian-born Albert Camus's posthumous novel The First Man (Reference Camus and Hapgood1996), his passionate proclamation of his Algerian identity. As the Algerian civil war of the 1950s dragged on, with atrocities on both sides, an anguished Camus insisted that France could not just leave Algeria to its fate in the hands of the Muslim majority. “She cannot, because she could never agree to throw one million, two hundred thousand Frenchmen into the sea” (in Messud Reference Messud2013: 56; see also Zaretsky Reference Zaretsky2013: 66–67).

Probably, for Finley, each case has to be decided on its own merits, depending on the character of settler rule, and in particular perhaps the extent to which the settler community is long-standing, extending over several generations. For such European populations, the settled territory is home, as Algeria was for Camus, and as Kenya and Southern Rhodesia were for many British settlers (see, e.g., Kennedy Reference Kennedy1987; Lonsdale Reference Lonsdale and Bickers2014; Lowry Reference Lowry and Bickers2014). Such was not the case for the majority of British in India, or French in Indochina, or Dutch in Indonesia. Not only were they mainly a thin stratum of Europeans—administrators, soldiers, traders, engineers, missionaries, educators—sitting atop a population of millions of Asians. Most of them did not regard their stay in the East as permanent. They were there to do a job, to make some money, to get some experience, perhaps just to have an adventure. Unlike the old English couple in Paul Scott's novel Staying On (1977), who knew that they were exceptional in wishing to remain in India even after independence, most Europeans felt that they were just passing through. Home was Britain, France, or the Netherlands. Their children were usually sent to the home country for their education. After a certain time abroad in the empire, they too expected to return to the homeland, to rejoin their families and to continue a familiar pattern of life. That they were often disappointed in this—finding that their home societies had changed in their absence—did not change their sense of who they were, and where they belonged.Footnote ¹¹

For Finley, these people, and the territories they inhabited, did not form colonies but were components of empire. One senses that the main polemical thrust of Finley's contribution is in fact to draw a sharp line between colony and empire, colonialism, and imperialism (see also Horvath Reference Horvath1972; Ferguson Reference Ferguson2005: 169). He wishes to stop people speaking of the British Raj as a colony, or French Indochina as an example of French colonialism. The British, like the French and Dutch, and the Spanish and Portuguese before them, constructed empires, some of the largest in the world. But only some portions of those empires can be considered colonies. All empires are composed of dependencies, but only some dependencies, mostly a minority of them, are colonies. In the British Empire, India was a resplendent part of empire but, according to Finley, it was never a colony.

Finley does not, at least in his article analyzed here, concern himself with the properties of empires.Footnote ¹² His interest is in defining and delimiting colonies. What is left out, what many people mistakenly call colonies, he is happy to call empire, but more by implication than by detailed analysis. Part of the reason for that self-limiting restriction comes in the observation that “the semantics of colonial terminology have not been systematically investigated (unlike ‘empire’ and its cognates)” (Reference Finley1976: 168). In other words, we know pretty well what empires are; it is colonies we are confused about.

That may be so, but for all the investigation of empire and imperialism we remain quite some way from agreement about them.Footnote ¹³ The continuing disputes about “formal” and “informal” empire, and the differences between them, is just one example of an unfinished scholarly controversy (see, e.g., Colás Reference Colás2007: 71–115). Still, and focusing on Finley's main concern—colonies and colonization—we might ask what is gained, and lost, by accepting his restrictive definition of colony? Does it, as he claims, help us to establish a more precise and helpful typology? Does it clarify what we might mean by empire, by specifying its difference from colony? Can it allow us the range of cases, over time and space, that have been the strength of some of the best accounts of empire, whether or not they include colonialism as a sub-category?

ASSESSING FINLEY: SETTLER COLONIALISM, EUROPEAN EXCEPTIONALISM, AND EMPIRE IN GLOBAL CONTEXT

Though Finley's was a relatively lone voice in the 1970s, and the distinctions that he wished to stress did not really catch on in the scholarly, let alone popular, discourse of empire, one strain of thinking since then has gone very much in his direction. This has mostly taken the form of analysis of the particular properties of “settler colonialism,” as opposed to other forms of colonialism and imperialism. It is unfortunate that scholars involved in this enterprise have tended to see settler colonialism as different from other forms of colonialism, rather than from types of imperialism (e.g., Veracini Reference Veracini2010). It would have sharpened the distinction they wished to make, and certainly lessened terminological confusion, if by settler colonialism they had indicated its identity with what Finley simply called colonialism, and if they had allowed other forms to go under the more capacious rubric of imperialism. The latter, on inspection, is often what they seem to intend.Footnote ¹⁴ Like Finley himself, in their concentration on the peculiar qualities of settler colonialism they have seemed vague or incurious about other forms of what clearly to other scholars come under imperial rule, of which colonialism is a sub-type.

That said, this new literature on settler colonialism had made a valuable contribution to understanding the specificities of colonial life, and by suggesting at least by implication how it differs from life under imperial rule. There is the stress on the fact that, as Patrick Wolfe says, “settler colonialism is first and foremost a territorial project” (Reference Wolfe and Moses2010: 103). Or, as Finley puts it, following Adam Smith, “land is the element round which to construct a typology of colonies” (Reference Finley1976: 178). Doctrines of original “discovery,” of terrra nullius, of the argument from use and exploitation of land otherwise left uncultivated and unproductive by “wild” and “savage” people, all were part of the conceptual armory drawn upon for the justification of the colonists’ appropriations of and settlements on the lands ancestrally lived on by indigenous peoples, whether in the Americas or Australia.Footnote ¹⁵ With empire, by contrast, as with the British in India or the French in Indochina, labor might be exploited in various ways, and heavy land taxes imposed, but the land generally remained in native hands.Footnote ¹⁶ Empire is rule over peoples; colonialism is the acquisition of territory for the purposes of settlement and cultivation.

Again, another striking feature of settler colonialism that distinguishes it from empire is the strenuous attempts by the colonists to “indigenize” themselves. On the face of it, they are the interlopers, perhaps illegitimate takers of other people's long-held lands. They have come from elsewhere, and in fact part of their self-justification is that they are bringing European civilization to savages. But at the same time they have come to stay; there is no future for them beyond the colony (unlike Europeans in imperial settings who have every hope and expectation of returning home). Hence, as Veracini insists, there can be no such thing as “settler decolonization,” unlike the decolonization process that was such a striking feature of the post-Second World War European empires (Veracini Reference Veracini2010: 99–100).Footnote ¹⁷ The colonists possess a dual identity as Europeans and as settlers, but because they see themselves as permanent residents in the new lands they increasingly seek to throw off their Europeanness and to stress their identities as “new men,” formed by the novel conditions of the “Outback,” the frontier, the “brousse” (ibid.: 21, 123). And, because they do not regard the native populations as legitimately possessing the land, they come to see themselves, as Camus saw the pieds noirs in Algeria, as the “first men,” the original creators of the societies that they have formed in the wilderness or on unused lands (Reference Camus and Hapgood1996: 279). They, whether the French in North Africa, the English or French in North America, or the British in Australia, are the true indigenes, not the native people (Veracini Reference Veracini2010: 36, 46–48; see also Lorcin Reference Lorcin2013).Footnote ¹⁸

A linked feature of this settler claim to indigeneity is the propensity to appropriate indigenous symbols once the indigenous people have been removed, physically or culturally, or as part of the struggle against metropolitan power. “Australian public buildings and official symbolism, along with the national airlines, film industry, sports teams and the like, are distinguished by the ostentatious borrowing of Aboriginal motifs” (Wolfe Reference Wolfe2006: 389). Once the Indian population of North America had been suppressed, the new United States appropriated their identities, so that now more than half of the fifty American states bear Indian names, and having at least a portion of Indian blood is regarded by some as a matter of pride. In South America, in the Viceroyalty of New Spain, the Spanish criollos who fought for their independence from metropolitan Spain in the early nineteenth century renamed their country “Mexico,” in homage to the Mexica from whom their Spanish ancestors had seized the territory. Likewise, in the early twentieth century the famous Mexican muralists Diego Rivera, David Siqueiros, and Jose Orozco painted the Spanish conquistadores as cruel and sadistic conquerors, and the Indians as noble victims (Rochfort Reference Rochfort1993). In settler representations, settlers have become natives, or at least identified with natives, in atonement for past sins or as common victims of a metropolitan power.

Renaming countries and cities in homage to indigenous peoples and cultures and as retrievals from the precolonial past occurs also in imperial settings, but with the fundamental difference that it is the indigenous peoples, not the European settlers, who do the renaming. Thus it was native Indians who renamed Bombay Mumbai and colonial Madras Chennai, native Burmese renamed Burma Myanmar, and native Africans renamed Southern Rhodesia Zimbabwe and Upper Volta Burkino Faso. It is also natives, not settlers—who sometimes dress up in native garb (Veracini Reference Veracini2010: 47)—who revive past forms of dress, as with Gandhi's dhoti and Nehru's kurta. Settlers might pay homage to Indians and Aborigines, but they are also anxious to stress their own indigeneity, their own status as pioneers and “new men,” and hence they cannot summon up the precolonial past in the way that is natural to native people achieving their independence from their imperial rulers.

One last way might be mentioned of illustrating the difference between colony and empire: they have different structures of rule. Colonial rule involves a tripartite or triangulated set of relations between the metropolis, the settler community, and the indigenous peoples (ibid.: 16–32; Fradera Reference Fradera and Mackay2018: 22–52). Settlers are caught between two challenges and pressures, one from the demands of the metropolitan authorities, the other from the resistance of indigenous peoples. Their response is usually to attempt to throw off the rule of the former and to suppress the latter.

In empires, on the other hand, rule is more direct, not mediated by a settler community. The British in India or in Africa might practice “indirect rule,” employing native rulers and chiefs as agents of rule. But that is merely a cheaper and more efficient means of gaining stronger control over the native population. There is no intervening power that they have to deal with, no European community employing the weapons of European technology as well as European ideology in resisting metropolitan authority (though natives might also borrow ideas and techniques from their European rulers).Footnote ¹⁹

These accounts of settler colonialism, implicitly supporting as they do Finley's argument, are additionally valuable in that, like Finley, they draw attention to the distinctive nature of European colonialism from about the sixteenth century onwards.Footnote ²⁰ General treatments of empire often incorporate modern European colonialism in their accounts without necessarily stressing how different it was from earlier episodes, and how consequential that difference has been (e.g., Doyle Reference Doyle1986: 104). Modern European colonialism opened a new chapter in world history, one whose impact can still be seen and felt the world over. European conquests certainly involved older elements of empire, and indeed the aspiration to be a “new Rome” was common to nearly all the European empires (Kumar Reference Kumar2017a: 37–44). But the new, unprecedented thing was the scale and significance of the movement of Europeans to all the corners of the world, the creation, in other words, of colonies as Finley understands that term.

This began as a trickle in the sixteenth and early seventeenth centuries, as relatively small numbers of Portuguese, Spanish, Dutch, French, and British crossed the Atlantic to settle in the New World, the territories of which had nearly all been claimed by the European powers. The numbers increased in the later seventeenth and eighteenth centuries, with Asia now being added to the points of destination. In the nineteenth century the stream of European settlers grew into a flood. Between 1820 and 1930 more than fifty million Europeans—about one-fifth of the entire population of Europe at the beginning of that period—migrated to other lands. Many went to an ex-colony—the newly-independent United States—but many also swelled the numbers in the European colonies in other parts of the Americas, as well as in Australasia and Africa. Smaller numbers migrated to Asia and the Middle East (Cipolla Reference Cipolla1974: 115–16; Crosby Reference Crosby1986: 2–5; Livi-Bacci Reference Livi-Bacci and Ipsen1992: 123–24).

No other subdivision of the human species has occupied so many parts of the earth as Europeans have. The result has been the establishment of a large number of “neo-Europes” around the globe. People of European descent—European creoles—make up over 90 percent of the population of Australia and New Zealand, over 80 percent of North America's, and over three-quarters of South America's. There are or were substantial communities of Europeans in North and South Africa, and smaller pockets in Asia (Belich Reference Belich2009: 25–42). A UN report of 1953 commented that “the great exodus from Europe has been the most important migratory movement of the modern era, and perhaps the largest in all human history” (in Cipolla Reference Cipolla1974: 115).

The established settler communities—in the Americas and Australasia, and parts of Africa—developed, as we have seen, a distinctive pattern of life that differed in many respects from the attitudes and practices of the usually smaller groups of Europeans in other parts of the European empires. There were different policies, often of a savagely exclusionary kind, toward indigenous peoples, leading to their expulsion from their homelands and sometimes to their near extinction. This is part of what Michael Mann, noting the relative egalitarianism that prevailed among the settler communities, has called “the dark side of democracy” (Reference Mann2005: 70–98). This compares with, say, the attitudes of the British community in India or the French in Indochina, where awareness of the antiquity of the civilization, together with the relatively small size of the European communities, compelled generally a more cautious and often respectful attitude towards the native peoples.

There were differences also in the attitudes toward the homeland. The small European communities in most of Asia and sub-Saharan Africa continued to keep their metropolitan identities, as British, French, or German, often stressing them even more than their counterparts at home. That was to a good extent true for much of the time also of the larger settler communities, as has been clear in the extensive work that has now been done, in the case of the British Empire, on the “British World” of the white dominions (e.g., Bridge and Fedorowich Reference Bridge and Fedorowich2003; Buckner and Francis Reference Buckner and Francis2005; Bickers Reference Bickers2014). The American colonists insisted on their Britishness and claimed the rights of “true-born Englishmen” until their break with Britain. Australians, Canadians, New Zealanders, and British South Africans did similarly. But, like the Americans, when the time came, they did not find it difficult to declare their own identities, different from the metropolitan British, as Australians, Canadians, and so forth. In the established settler communities, where the residents saw themselves as there to stay, along with their descendants, it was inevitable that a new sense of the self would emerge, in environments and climates often wildly different from those of their home countries. This often led to clashes with the metropolitan authority, even before independence or autonomy was attained by settlers. This was less easy for those Europeans who had to keep themselves in readiness, sooner or later, for the return home.Footnote ²¹

The term “colonial empire,” so ubiquitous in the literature, by eliding the two entities hides the distinctiveness of the modern European empires that, unlike most of the ancient or non-Western empires, laid down their presence in the world by settling, or “planting,” large numbers of their own peoples on virtually every continent. The value of Finley's exercise in classification lies above all in stressing that. The European imperial experience was in that respect unique, and it has had profound effects on the legacy of empire (the power and influence in the world of the “Anglo-world” or “Anglo-sphere” being one such obvious product; see, e.g., Mead Reference Mead2007; Belich Reference Belich2009).

But what might be lost in drawing so hard and fast a line between colony and empire? Does it prevent us from generalizing about the British or French empires, which contained both colonial and non-colonial elements? Is there no place for ideologies of empire, and concepts such the “civilizing mission” that not only found individual forms in the various empires but can also be held to have been a general European justification of empire (Conrad Reference Conrad and O'Hagan2012: 137)? That so many scholars, as well as other commentators, feel no need to distinguish between colony and empire suggests that in discussing empire in general we may unduly constrain ourselves if we insist too strongly on the distinction. The search for commonalities in the imperial experience is as important as the stress on differences. The European settler colony no doubt needs to be noticed for its special qualities. But it was also part of a wider movement of empire, one that saw itself in a particular tradition of empire, with its own hallmarks. For European rulers it was important to be included in the “family of empires,” to show that they understood what it was to be an empire.

In the European case, the hold of Rome was paramount (Kumar Reference Kumar2017a: 37–73). For Europeans, Rome had invented empire. They studied it through the classical authors—Sallust, Vergil, Cicero, Plutarch, Polybius, Tacitus—that formed the bedrock of the education of the European ruling class. Rome had wrestled with the problem of imperial citizenship, coming up with categories that were thought to be highly applicable to the European overseas empires, incorporating as the Roman Empire did “barbarous” peoples such as the Gauls and Lombards. The Roman Empire, too, had incubated Christianity, eventually adopting it as the state religion and thus ensuring its future and its widespread diffusion. Charlemagne's Holy Roman Empire, in self-consciously reviving the Roman tradition, had confirmed this union of religion and empire. All European empires adopted this heritage, even after the division of Christianity into Catholic and Orthodox, Catholic and Protestant. Moscow might in the sixteenth century proclaim itself “the third Rome,” in succession to Rome and Byzantium, but whatever the differences, the promotion of Christianity was a common feature of all the European empires. At certain moments of threat, as from the Muslim Ottoman Empire of the sixteenth and seventeenth centuries, it might even get them to act in concert rather than, as so often, in competition and conflict.

The concept of translatio imperii, the translation or handing on of empire, was one influential way in which successive empires in the West were bound to Rome. First elaborated in relation to Charlemagne's assumption of the Roman title in 800 CE, it proved serviceable to all succeeding empires. It was a key term in the idea of a tradition of empire, one that emphasized not just succession but universality, just as Rome had ruled the orbis terrarum, the whole known world (Wilson Reference Wilson2016: 38). Imitation of Rome, however, was not, and could not be, slavish. Time mattered. The Habsburg ruler Charles V, elected Holy Roman Emperor in 1519, was a passionate admirer of Rome but also felt the need to proclaim his distance and difference from Rome. The Romans had emblazoned across the two “Pillars of Hercules” on either side of the Straits of Gibraltar the motto “Non Plus Ultra” (“thus far and no further”). Charles turned this into “Plus Ultra” (“still further”) (Elliott Reference Elliott1989: 8). The Atlantic beckoned. It was a new age. Within a century the Spanish had conquered and colonized a good part of the world totally unknown to the Romans. Tradition was important. The Spanish had no doubt that they ruled in succession to Rome. But tradition could only be maintained by adapting to change. European empires engaged in dialogue with Rome, but in the wake of the “voyages of discovery,” the scientific revolution of the seventeenth century, and the industrial revolution of the eighteenth, they knew that they lived in a world vastly different from Rome's (Kumar Reference Kumar2017b).

There are other problems with Finley's attempt to erect a boundary between colony and empire. One concerns the concept of “internal colonialism,” which many have found illuminating in considering the development of certain societies, whether or not they had formal empires, and whether or not they had colonies (as Finley understands them). Thus, the westward movement of the American colonists after independence may not have led to the establishment of colonies, but there certainly seems something “imperial” in the expansion to the Pacific, as noted earlier. The debate about the “American Empire” has many facets, but whatever its global aspects, a strong case can be made for considering America as at least a land empire, in this case perhaps comparable to Russia's eastward expansion to the Pacific in the construction of its empire (Kiernan Reference Kiernan2005; Khodarkovsky Reference Khodarkovsky2002). The United Kingdom has also already been mentioned as a land empire formed through internal colonialism. We might also consider the fruitfulness of the concept in relation to large multi-ethnic states such as India and Nigeria (e.g., Devji Reference Devji2020). Too strong an emphasis on the distinction between colony and empire, and between colonialism and imperialism, inhibits the search for comparisons and parallels that can often be highly instructive.

This points to the need not too overemphasize, as is common in much of the literature, the distinction between land and overseas empires. Some countries—Britain, Spain, France—in any case had both (Kumar Reference Kumar2010: 124–28). The two types of empire have their differences, no doubt, and some of these can be very important, but there are also continuities and similarities between them which should not be overlooked. This links to a similar observation that can be made about the equally common distinction between “ancient” and “modern” empires. Most ancient empires were land empires, though some, like the Roman Empire, had significant, though “near-abroad,” overseas possessions.Footnote ²² Moreover, few possessed (Finley-type) colonies. There were settlements on occupied land, but they were usually incorporated directly into the state, as with the provinces of the Roman Empire (Finley Reference Finley1976: 185). For Finley, the ancient world is replete with empires and imperialism, but conspicuously lacking in colonies and colonialism. Not only are the so-called colonies of the Greek world not true colonies; neither are the dependent settlements of people from the home territory, where and when they occurred (as for instance in the empire of Alexander the Great). There is no colonialism in the ancient world.

If colonialism is the distinctive hallmark of modern empires, this view of settlements in the ancient world tends to make the empires of that world something very different from modern empires and requiring perhaps separate treatment. At the very least that would vastly diminish the field for purposes of comparative analysis. Michael Adas has noted that the term “colonization” has “come to refer almost exclusively to historical processes involving western Europeans, or their settler progeny.… Colonialism is deemed to be one of the global forces that has defined the modern age; empires are seen as modes of state expansion with an ancient lineage, increasingly anachronistic in an age of industrialization and high technology” (Reference Adas1998: 371).

One consequence of this persistent “Western-centrism,” Adas further observes, is to relegate many non-Western examples, such as China and Mughal India, to the archaic realm of empire. Western colonialism is world-spanning; “empires” are more restrictive cases of largely regional expansion, involving the conquest of neighboring peoples and states (i.e., “empires” are almost exclusively land empires; overseas empires are “colonial”). The East had empire in abundance (“Oriental Despotism”), but it lacked colonialism. This limits its role in world history. “Colonialism suggests processes that are global in scope, of relevance to human societies everywhere. By contrast, empires, whatever their size and influence, are regional, or at best intracontinental entities, whose history is assumed to merit serious concern from only specific and more restricted portions of humanity” (ibid.). Adas suggests that this restrictiveness of approach does not just hide important points of similarity between Western and non-Western empires but, equally, interesting differences (some of which are contained in the contrast between land empires—both Eastern and Western—and overseas empires).

China is certainly a case in point. It has become another one of those “academic parlour games,” to use John Darwin's expression, to speculate what might have happened if the famous seven voyages (1405–1433) of the Chinese Admiral Zheng He in the early Ming period had resulted in Chinese overseas colonization. Certainly, the Chinese possessed the means, in ships, men, money, and military technology, to have done so, in an area stretching from China to East Africa. Had they done so, they would have had a head start on the Europeans, whose own colonization did not begin until much later in the fifteenth century. But, for reasons much discussed, the Chinese did not engage in conquest and colonization, and indeed, it is generally agreed, they had no such intention in mind in sponsoring Zheng He's expeditions (Brook Reference Brook2010: 93–94). The Chinese Empire remained a land empire, one of the largest in history.

But that should not of course exclude China from any large-scale, comparative account of empires, including the European overseas empires. This is particularly so if we do not treat the Chinese Empire as some sort of timeless, unchanging entity with an uninterrupted, two-millennia lifespan from 221 BCE to 1912 CE. There were several marked discontinuities, most notably the Mongol conquest that established the Yuan Dynasty (1279–1368), and the Manchu conquest that of the Qing (1644–1912). During the Qing dynasty, especially in the seventeenth and eighteenth centuries, China expanded massively in Inner Asia, annexing Tibet, Mongolia, and Xinjiang. This more than doubled the size of the country. It also strengthened its hold over the south, and it occupied the nearby island of Taiwan. This was, in the eyes of many scholars today, “Manchu colonialism” (International History Review 1998; Perdue Reference Perdue2005).

Nicola di Cosmo (Reference Di Cosmo1998) has made a strong case for considering Qing China—at least in its Inner Asian sector—as a colonial empire very similar in many ways to the European colonial empires. Consolidation in the south of the country, and in the conquest of Taiwan, did involve the settlement of quite large groups of Han Chinese. But, as with the Roman Empire in Finley's account, these parts of the empire were directly incorporated as provinces in the central state administration, and policies of assimilation—sinicization—were actively pursued there. Towards their possessions in Inner Asia, however, the Chinese emperors adopted a very different attitude. Immigration by Han Chinese was discouraged; there was greater respect for Mongol, Tibetan, and Islamic cultures; and administration was put largely in the hands of Manchu and Mongol—not Han—officials, who did not feel the need to press for sinicization. In addition, these administrators relied heavily on local Mongol, Tibetan, and Turkestani elites to carry out much of the day-to-day business of imperial rule. In other words, “Qing rule in Tibet, Mongolia, and Xinjiang did not differ in principle from the European penetration of overseas dependencies: the Qing, too, developed original systems of administration for newly acquired territories by relying on native elites and separating the dependencies from metropolitan China. The number of Chinese troops and settlers in relation to the natives was kept small, and religion and local custom were used to induce submission” (ibid.: 306).

Of course, “Manchu colonialism” is not colonialism in Finley's sense. The Inner Asian territories were undoubtedly dependencies, but they did not involve plantations of significant numbers of Chinese from the homeland. But it is here that the narrowness of Finley's concept of colony most glaringly reveals itself; or, perhaps more importantly, where the attempt to separate colony from empire most displays its limitations. One can, if one wishes, focus on the distinctiveness of settler life and the settler community. That is the strength of Finley's approach, and the main justification for trying to revive the old—that is, the early-modern—meaning of the term colony. But it seems perverse to restrict colonialism to simply the establishment of these dependent settler communities, when, as the Chinese case clearly shows, empires usually can and do have both this type of colonialism (e.g., in Taiwan) and colonies in the wider and, today, more conventional sense, as in the Inner Asian dependencies. It is not just that, as di Cosmo insists, colonies do not have to be overseas. Nor do they have to be composed of a majority of members of the metropolitan power. What is more important is the fact of dependency, and, more importantly, that they are of a “separate, geographical, ethnic, and cultural nature”—that they are in some important sense not a normal part of the metropolis. “Under the Qing, the geographical and cultural differences between the peoples of the fan (outer regions [such as the Inner Asian dependencies]) and the peoples of China were not only maintained but also legislated and adopted as a criterion for the establishment of native self-governing communities” (ibid.: 307).

This is something like the “rule of colonial difference” in the European empires that Partha Chatterjee famously enunciated, “the preservation of the alienness of the ruling group” (Chatterjee Reference Chatterjee1993: 10). Josep Fradera (Reference Fradera and Mackay2018) too has recently seen the doctrine of “specialité,” codified in the French constitution of 1799 and laying out different legal regimes for the colonies and the metropolis, as characteristic of a wide range of empires—British, French, Spanish, and American.Footnote ²³ But we must be careful not to exaggerate this feature of empires. Many empires moved toward a relatively high degree of integration without giving up on the imperial principle. That was true of the Roman Empire, as symbolized especially by the decree of 212 CE in which the emperor Caracalla conferred citizenship on virtually all the free subjects of the empire. The French Empire too moved in this direction, and in both the Roman and French empires there were conscious policies of assimilation. So, the difference between the two forms of colonization in China identified by di Cosmo should not be regarded as rigid or absolute.

What comes out in this discussion is the danger of seeking too precise a definition of colony, as in Finley's attempt, and trying then to pry apart colony from empire. Colonies and colonization come in a variety of forms. Some take the form identified by Finley—plantations of large numbers of settlers. But others might involve a relatively small group of metropolitan people ruling over diverse groups, and yet living in ways that are clearly distinguishable from those of the metropolitan society. That would apply not just to China's Inner Asian colonies but also to, say, the settlements of Russians in Central Asia and the Far East in the eastward drive of the tsarist Russian Empire (see e.g., Khodarkovsky Reference Khodarkovsky2002). We have already considered America's westward movement as equally a kind of colonization, even though the settlers did not establish separate, autonomous communities.

Finley's intervention was stimulating and productive, though it has not yet received the attention it warrants. It forces us to think more clearly about what we mean by many terms—colony, colonialism, colonial empires—that are common in the literature but are often loosely used as synonyms for empire and imperialism. There is something distinctive about colonies as Finley understands the term. They do seem, as Fieldhouse (Reference Fieldhouse1982) and others suggest, to mark a new departure in the imperial story, from about 1600 onward. European imperialism from that time, for all its continuities with past forms of empire, does seem to incorporate a new element, in the creation of dependent yet largely self-governing communities of settlers. The analysis and description of those communities, and how they differ from rule in other regions of empire, continues to be an important and valuable field of study, as the newer work on settler communities amply shows.

At the same time, it seems dangerously restrictive to carve out a separate intellectual domain of “colonies and colonialism,” as something demanding wholly different treatment from that we give to empire and imperialism. Colonies, of different kinds, are a part of empire. They only have existence as manifestations of an imperial drive. They do not form a self-sufficient universe. The very fact of dependency means that they exist in structures that go beyond them, that incorporate many diverse elements, and that have purposes and intentions that often offend one or more of their component parts. Empires are agglomerations, often untidy and unwieldy, but they also have visions and ideologies that give meaning to their existence in the world and justify their continued presence (Kumar Reference Kumar2017a). The colonies that are part of them draw on these ideologies, even as they sometimes use them against the imperial centers themselves. We should beware of treating colonies as distinct from empires; they are part and parcel of them and their story is an imperial story. Even when they separate from them, they carry marks of their former membership well into their futures.