| Land, Labor, and Difference: Elementary Structures of Race | The American Historical Review, 106.3

Forum Essay

Land, Labor, and Difference:
Elementary Structures of Race

PATRICK WOLFE

Read messages from the online discussion of this article, held Sept. 3-17, 2001, in which the author responds to postings.

In Australia, although Aboriginal people are called black, they are not ideologically credited with a natural sense of rhythm. Conversely, unlike Aborigines, black Americans have not figured as a dying race.¹ Rather, this latter condition has been the ideological preserve of red Americans. In this, as in many other respects, popular representations of black Australians and red Americans have distinctly resembled each other, while both have contrasted sharply with popular representations of black Americans.² Thus more is involved than discourses of color or nationality. 1

     In Australia and in the United States, white authorities have generally accepted—even targeted—indigenous people's physical substance (synecdochically represented as blood) for assimilation into their own stock. In both countries, indigenous people have asserted criteria other than blood quanta as bases for group membership and identity.³ When it has come to black people's physical substance, on the other hand, it has only been in the last few decades that U.S. authorities have dispensed with the most rigorous procedures for insulating the dominant stock. Moreover, with some exceptions, black groups in the United States have themselves affirmed the "one-drop rule," maintaining an inclusive membership policy that, apart from anything else, has kept up group numbers. Afro-Brazilians, by contrast, have found themselves subject to policies in many respects similar to those that have been imposed on indigenous people in Australia and the United States. In particular, Brazil's policy of "racial democracy" has sought to whiten the population by means of a combination of white immigration and officially sanctioned "miscegenation" between black and lighter-colored Brazilians. Even among Afro-Brazilians, organized opposition to this policy has not attracted widespread support.⁴ 2

     There are no grounds for assuming that such striking disparities can be reconciled under a single master category called "race." Rather, the preceding examples suggest that race is but one among various regimes of difference that have served to distinguish dominant groups from groups whom they initially encountered in colonial contexts. Focusing on the discourses of miscegenation that structure these examples does, however, enable us to say more about the particular colonial contexts concerned. For instance, American Indians and Aboriginal people in Australia share much more than the quality of attracting assimilation policies. Above all, they are both sets of peoples whose territorial expropriation was foundational to the colonial formations into which Europeans incorporated them. Thus their relationship with their colonizers—as both parties to the relationship would presumably agree—centered on land. In contrast, blacks' relationship with their colonizers—from the colonizers' point of view at least—centered on labor. In this light, the varying miscegenation policies make immediate sense, since assimilation reduces an indigenous population with rival claims to the land, while an exclusive strategy enlarges an enslaved labor force. 3

     This essay will examine certain racial regimes that Europeans have sought to impose on colonized groups, with a view to illuminating both the foundations on which Europeans established white supremacy and the changing ways in which they have sought to maintain it. To this end, we will focus on the points at which racial classifications most conspicuously come undone. For such classifications, in common with other imposed categories, apply most visibly where they are vulnerable, at the points where the divisions that they proclaim break down. Thus the essay will investigate differentiation by way of its negation. Universally, racial categories have been transgressed by sexuality, so we can approach systems of social domination comparatively by way of a cross-cultural historical survey of discourses of miscegenation.⁵ On this basis, the essay will analyze miscegenation policies that white authorities have sought to impose in Australia (on Aborigines), in the United States (on blacks and on Indians), and in Brazil (on the bulk of the population). Since the contrasts and correspondences that will concern us do not necessarily emerge in these different societies at the same times or in consistent temporal sequence, the discussion is contextual and thematic rather than chronological. 4

As I have previously argued in more detail,⁶ the Australian assimilation policy, under which Aboriginal children with a proportion of non-Aboriginal descent were taken away from their natal families by the Australian state, was integral to the logic of elimination that underpinned Australian settler colonialism. As opposed to franchise-colonial relationships (such as the British Raj, the Netherlands East Indies), settler colonialism seeks to replace the natives on their land rather than extract surplus value by mixing their labor with a colony's natural resources. Slave societies, which may be metropolitan or colonial, are different again, since—at least, in the case of the successful ones—the labor that is mixed with the land is not native but geographically alienated, the slaves being, or being descended from, outsiders removed from their original homelands. In the case of settler colonies, which may also encompass relations of slavery, the colonizers come to stay, expropriating the native owners of the soil, which they typically develop by means of a subordinated labor force (slaves, indenturees, convicts) whom they import from elsewhere.⁷ Thus the antebellum United States encompassed both settler-colonial relationships (between whites and Indians) and relationships of slavery (between whites and blacks).⁸ In the Australian case, the white convicts whom the British imported were not comparable to slaves in the United States since, no matter how long the term of their servitude, they did not transmit their juridical condition to offspring. So far as Aboriginal people were concerned, however, the British were involved in settler-colonial relationships that were substantially similar to those between whites and Indians in the United States. In either case, settler colonialism introduced a zero-sum contest over land on which conflicting modes of production could not ultimately⁹ coexist. Thus the primary logic of settler colonialism can be characterized as one of elimination.¹⁰ 5

     In Australia, the logic of elimination acquired peculiar explicitness through the doctrine of terra nullius. Although the British Privy Council did not confirm this doctrine in colonial law until 1889,¹¹ it was deeply grounded in legal and popular-cultural presupposition. In the latter, despite the whispering in a few more sensitive hearts, Aborigines were simply savages and, as such, without rights. In its more formal aspect as an item of international law, however, the doctrine emerges not so much as an enduring ethnocentric prejudice as a quintessentially modern ideology that unites the domestic and colonial aspects of European bourgeois discourse. 6

     The key concept is that of private property. In distinctively Lockean fashion, the doctrine held that property in land resulted from the mixing of one's labor with it to render it a more efficient provider of wealth than it would have been if left in its natural state, paradigmatically as an aristocratic hunting reserve. Practically, this meant settled agriculture, involving cultivation, irrigation, and enclosure. In addition to this requirement, it was also necessary that there be a properly sanctioned framework of laws to protect the property rights that the individual had acquired by dint of the application of his (sic) labor. Practically, this meant centralized governance, formal sanctions, and, again, enclosure, or fixed public boundaries. Within Europe, there could hardly have been a clearer antithesis, not to say challenge, to an unrepresentative system of hereditary landed power characterized by inefficiency and wasteful exclusions. In its colonial application, where it acquired the formality of a name (terra nullius meaning "nobody's land"), the same set of principles furnished a warrant for denying "nomadic" peoples ownership of the land they occupied.¹² 7

     In Australia, Europeans almost universally judged Aborigines to be nomadic—not in the pastoral, biblical sense but as people who merely prowled about the landscape in search of sustenance, garnering at will like so many stray animals—and as lacking anything resembling an ordered system of government. Accordingly, they had no more claim to ownership of the land than had the native fauna that grazed across its expanse. As Justice Blackburn could still conclude in 1971, in the Gove land rights case, Aborigines belonged to the land, but the land did not belong to them.¹³ Since Australia belonged to no one else, it was simply there for Europeans to take, without the requirement for contract, compensation, or other form of consideration that the acknowledgement of so-called Native Title would have imposed. 8

     This is not to suggest that the attribution by Europeans of Native Title has been of any particular benefit to colonized peoples. Indeed, Indian peoples' experiences with the treaties that were necessitated by their Native Title (or Domestic Dependent Nations status) would indicate that terra nullius had the advantage of leaving Aboriginal people's unacknowledged rights intact.¹⁴ The point is rather to indicate the deep historical anchorage of settler-colonial discourse, which is much more than some ad hoc rationalization that sprang up spontaneously in the Australian context. With this provenance recognized, comparative colonial history becomes possible. Moreover, it also enables us to move beyond the category "Europe," used in a vaguely geographical sense, to a specific social formation, capitalism, with discernible historical moments and phases. For instance, the very fact that the British colony in Australia originated as a convict settlement (the United States having been closed off after the revolution) necessarily brings us back to the bourgeois discourse of private property that underpinned the doctrine of terra nullius, in this instance by way of the enclosures that produced cities teeming with the Malthusian poor whom convict shipment was intended to siphon off. As will be noted below, such considerations did not apply to Portugal, which, remaining effectively preindustrial, lacked a surplus population, a consideration that had profound implications for the Brazilian colonial formation. 9

     With or without doctrinal explication, though, in both Australia and the United States, the logic of elimination took on a range of related historical forms. Indeed, as will emerge, the settler-colonial policies that authorities in the two countries have implemented or attempted to implement might seem to have been drawn from a common stock. With the notable exception of severalty or allotment legislation,¹⁵ the standard stage model of Indian policy—conquest, removal, reservation, allotment, assimilation, co-optation, termination, self-determination¹⁶ —elicits the ready recognition of one familiar with the history of Aboriginal-European relations in Australia. Of course, there were local variations and postponements, the details were not always just the same, and the "phases" did not always follow in strict succession—indeed, they often coincided as alternatives. Nonetheless, as an inventory of settler-colonial policy options, the list requires little substantial modification for reapplication to the historical development of Australian policies on Aborigines. 10

     Within Australia, policies on Aborigines have varied in time and place. In some regions, local factors have encouraged European reliance on Aboriginal labor. This applies particularly to the northern cattle industry, which remained crucially dependent on Aboriginal labor until relatively recently.¹⁷ Not only was it widely believed that Europeans were unfit to labor in such harsh conditions, but most of this region was not invaded until after the shipment of convicts had been terminated, so cheap white labor was unavailable. In early Tasmania (Van Diemen's Land), Aboriginal women were extensively used for sealing and oyster diving, while, later on, Torres Strait Islands men were employed as divers on pearl luggers. Aboriginal men and women were in many respects differently colonized, with women's domestic and sexual labor being valued on a different scale to their men's services. Although these and other variations are significant and need to be acknowledged, they do not alter the primacy of the dominant pattern, manifest most clearly in the south and east of the continent, where settler colonialism practically approximated its pure or theoretical form, resulting, within a short space of time, in the decimation of the Aboriginal population. 11

     Since, as in North America, frontiers moved across Australia from coastal beachheads, variously established over the century following the landing of the First Fleet in 1788, it is not possible to date the development of Australian settler colonization as an even sequence. Thus it is convenient to organize its establishment and consolidation into a typology of strategic phases, which, though overlapping, were successively introduced at different times in different parts of the continent.¹⁸ For heuristic purposes, I term these phases confrontation, carceration (which includes removal), and assimilation (biological and cultural). The first and the last represent the extremes of a historical transformation during which Aborigines' relationship to European society shifted from one of exteriority to one of interiority. Although the final phase, assimilation, was governed by a discourse of miscegenation, it was consistent with—and, in its own way, reproduced—the same logic of elimination that had underlain the first two phases. Thus the assimilation policy should be situated in the historical context of Australian colonization as a whole. 12

     During the initial phase, in which invading Europeans first confronted Aborigines defending their territory, Aboriginal resistance was subverted as a result of the combined effects of four related factors: homicide, introduced disease, starvation, and sexual abuse.¹⁹ The scenario is, of course, broadly familiar in the U.S. context. There is, however, a major difference. In the United States, depictions of black sexuality dominated miscegenation discourse to an extent that led Winthrop Jordan to claim, albeit with some exaggeration, that "the entire interracial sexual complex did not pertain to the Indian."²⁰ In Australia, on the other hand, which lacked a comparable "third race," miscegenation discourse came to focus primarily on indigenous people, whose blackness became correspondingly salient. 13

     Aborigines who survived the disaster of the first phase found themselves reduced to improvising whatever livelihoods they could in the pores of the alien new society, which generally found them repugnant. Measures were introduced to confine the surviving Aboriginal "remnant" to fixed locations, either by the lure of rations or by coercive measures.²¹ This constitutes the second, carceral phase of settler-colonial policy toward Aborigines. In keeping with Social Darwinist premises, as corroborated by Aborigines' manifest decimation, their confinement on missions and reserves was seen as a temporary measure, since they were believed to be a dying race. Although framed in philanthropic rhetoric (as in missionaries "smoothing the dying pillow"), this phase maintained the logic of elimination in that it vacated Aboriginal territory and rendered it available for pastoral settlement. 14

     Mission boundaries were not enough, however, to prevent the sexual encounters, conducted under conditions of radical inequality, that characterized relations between white men and Aboriginal women. These encounters produced offspring who, growing up as they almost invariably did with their maternal kin, identified themselves as Aboriginal. Moreover, far from dying out, this section of the Aboriginal population threatened to expand exponentially. As the nineteenth century moved to its close, the romance of extinction progressively gave way to the specter of the "half-caste menace." Aboriginal people became racialized—in the full genetic sense involved in blood quantum legislation—during the years surrounding national independence, in 1901.²² These developments coincided with the end of the frontier, an uneven process that marked the final internalization of the "Aboriginal problem." They also coincided with the introduction of the so-called White Australia Policy. Seeking to build a white man's paradise in the South Pacific, and encouraged by trade unions keen to eliminate cut-price labor, the newly federated national government in 1903 introduced legislation that curtailed non-European immigration and targeted non-white residents for deportation.²³ Since no external homeland could plausibly be assigned to Aborigines, the remedy for the challenge that they posed to white Australia was not projection without but absorption within. From around the turn of the twentieth century, a range of measures were introduced that were designed to detach individuals from Aboriginal communities, stripping them of their Aboriginal identities and incorporating them into white society. Thus the assimilation policy was a symptom of Aborigines' containment within Australian society, constituting an internal correlate to the White Australia Policy. 15

     Initially, lighter-skinned Aborigines (that is, those whose quantum of Aboriginal blood was deemed to be less than 51 percent) were expelled from reserves. They generally inhabited makeshift settlements ("fringe camps") on the margins of country towns, caught between the reserves from which they had been banned and rural white society. These settlements were subject to harassment on the part of police and local authorities, which increasingly took the form of child abduction, as predominantly lighter-skinned Aboriginal children were taken from their families and committed to youth-training institutions with a view to their insertion into the lowest echelons of white society—which is to say, with a view to their elimination from the Aboriginal reckoning.²⁴ After World War II, there was a general shift away from youth-training institutions toward a policy of having Aboriginal children adopted out to white families. This policy continued to be official practice in most Australian states until 1967, when a referendum removed the constitutional disabilities that had enabled Aborigines to be singled out in this way.²⁵ Members of the "Stolen Generations," as the victims of this policy have come to be known, have so far been unsuccessful in their attempts to secure redress from the Australian government. 16

     The conformity of this policy to the logic of elimination is too obvious to require explanation. For comparative purposes, however, we need to analyze the genetic arithmetic according to which Australian governments sought to breed Aborigines out. This will provide a basis for comparing the logic of the Australian system with the quantifications informing regimes such as the one-drop rule in the United States or the hyperelaborated set of color classifications in Brazil. In this connection, the key feature of the Australian system is that it limited the permutations of miscegenation to three descending fractions, conventionally termed "half-caste," "quadroon," and "octoroon." Crucially, a one-sixteenth category was not provided for. Rather, the category succeeding "octoroon" was full-blown whiteness. Assuming continuing miscegenation in each descending generation,²⁶ therefore, the system provided for a three-generational lap count to elimination (see Figure 1).²⁷ Though arbitrary, the fact that only three fractions were licensed is, therefore, fundamental, since it provided a cut-off point. Any greater number—let alone a Brazilian-style myriad of categories—would have extended Aborigines' anomalous persistence within white Australia on a scale of generations. 17

Figure 1. From A. O. Neville, Australia's Coloured Minority (see n. 27).

     In sum, then, the Australian solution to the problem posed for settler colonization by the recalcitrant persistence of extraneously constituted indigenous societies—who, significantly, were numerically overwhelmed—was to absorb them into the white stock. So far as race is concerned, it is important to note that, despite the Social Darwinist rhetoric of backwardness that accompanied the assimilation policy, Aborigines' physical substance cannot really have been seen as deficient, otherwise the last thing white authorities would have set out to do would have been to incorporate it into the white gene pool. In recent decades, the emphasis of assimilationist discourse has shifted from race to culture. Aborigines' day in the sun came in 1967, when a referendum removed clauses that had discriminated against them from the Australian constitution. Since then, the White Australia Policy has been abandoned in favor of multiculturalism. Positive representations of Aborigines have been a prominent feature of multiculturalist discourse. Rather than diminishing the pressure for Aborigines to assimilate, however, this has merely altered the ethnic profile of the society into which they are scheduled to blend. Thus they now find themselves represented as just another tile in the multicultural mosaic, a trivialization of their difference that effaces their status as prior owners. 18

     Using Australia as a baseline, we can now sketch out the principal comparative features of the U.S. and Brazilian systems. 19

The U.S. historian Edmund S. Morgan focused squarely on the central paradox of American liberal ideology, in which revolutionary ideals of universal freedom and the rights of man coincided with the consolidation of African slavery in colonial North America. Morgan argued that this contradictory situation had arisen as a result of the threat that the presence of an unruly white working class, the so-called "giddy multitude," had posed to seventeenth-century Virginian society.²⁸ Amid fears on the part of the governing elite that the colony of Virginia was fast becoming a "sinke to drayne England of her filth and scum"—scum who, as Bacon's Rebellion of 1676 had made alarmingly clear, could form alliances with rebellious Africans—Virginia embarked on a twin program that combined reductions in the importation of indentured English servants with a steep increase in the importation of African slaves. Since these slaves did not have indentures that expired, they would not present the threat to order that white freedmen did. Hence eighteenth-century Virginian leaders were able to extol the ideal of a free white yeomanry and profess allegiance to the rights of all Englishmen, because black slaves had taken the place of a dangerous underclass of recently freed whites. Thus Morgan's argument went much further than merely pointing to the synchronicity of slavery and libertarian ideals as if the two simply coincided by accident.²⁹ Rather, for Morgan, slavery was actually a functional prerequisite to the development of American revolutionary ideals. As he put it, slavery transformed "the Virginia of Governor Berkeley to the Virginia of Jefferson . . . [It was] slavery that made the Virginians dare to speak a political language that magnified the rights of freemen."³⁰ Developing this profoundly subversive insight on Morgan's part, David Brion Davis reconstructed the ideological progression in which blacks, who were deemed incapable of emancipating themselves, became a race apart in the late eighteenth century in a manner that excluded them from the universal category of man who was the bearer of the rights that the fathers of the revolution were so loftily enunciating. Thus race became "the central excuse for slavery."³¹ 20

     The perception that race provided an alibi for the particularity with which universalism applied in practice is by no means new. Indeed, so far as the core hypocrisy is concerned, it would be hard to improve on Alexis de Tocqueville's observation that the democratic United States exterminated Indians "without violating a single great principle of morality in the eyes of the world. It is impossible to destroy men with more respect for the laws of humanity."³² Race enabled the "men" being destroyed to be separable from the "man" in humanity. Well though this point has been made, however, its full implications have not been developed. In particular, race has not been adequately recognized as a shifter that bound together two principal but otherwise distinct strands of Enlightenment discourse, one epistemological, the other political.³³ Epistemologically, a central feature of the Enlightenment was, of course, the taxonomic imperative animating the great classificatory systems of Carl Linnaeus, George-Louis Buffon, Johann Blumenbach, and Georges Cuvier. Although the hierarchical structuring of these systems gave them obvious ideological utility in contexts of social domination, there was no necessary linkage between hierarchical taxonomies and the formal equality that hallmarked citizenship for liberal democratic theory.³⁴ As a taxonomy par excellence, however, race provided the categorical boundaries that ensured the exclusiveness of the bearers of the rights of man. This Jeffersonian fusion of bourgeois political ideology with classificatory natural science, of power with knowledge, gave race its singular epistemic purchase on Enlightenment and post-Enlightenment thought. Thus the point is not just that the prestige of science afforded an authoritative warrant to the categorical cleavage within humanity that the concept of race ordained. It is rather (or also) that race reconciled and unified two of the most formative—perhaps even the two most formative—components of the Enlightenment complex. Race, in short, is endemic to modernity. 21

     The dating of race's emergence and of its harnessing to slavery remain controversial, oceans of ink having been spilled on arguments over just when, once the first consignment of twenty Africans had been landed in Virginia in 1619, Africans became "Negroes" (as opposed to Blackamoors or heathens) and Negroes became slaves.³⁵ Nonetheless, it is at least clear that, by the last quarter of the seventeenth century, the equation of Africanness and slavery had become well established.³⁶ As the English missionary Morgan Goodwyn expressed the developing presumption in 1680, the two words "Negro" and "Slave" had "by custom grown Homogeneous and convertible."³⁷ Moreover, despite a brief period of patriliny in Maryland, slavery had become transmitted matrilineally.³⁸ In sharp contrast to the English system, wherein name and status were inherited in the paternal line, it was early on decided that, no matter who the father might be, the offspring of female slaves would themselves become slaves. Though generating some concern in cases where white women had children by black men, who may or may not have been free, this provision effectively rendered blackness equivalent to slavery, at least on the level of presumption.³⁹ With blackness denoting slavery, the consequence of black women's reproductive activity became the reverse of that which would subsequently come to be assigned to it in Australia. Whereas, in Australia, black women were to become conduits to whiteness, in the United States, black women came to augment white men's property by incubating the additional slaves whom they fathered. 22

     Even though the terms "African," "black," and "slave" became effectively interchangeable, in practice the match was by no means perfect. For not only did free blacks exist at all stages in the history of North American slavery, but, in the majority of cases, they were not fully "black" at all. Rather, these people's non-slave status was correlated on the genetic dimension by the fact that, in the main, they were of European as well as of African extraction. The contradiction between the slave as property and the slave as human being went to the core of the institution of slavery, recurrently surfacing when, for instance, slaves ran away (stole themselves?), committed crimes (with mens rea?), or converted to Christianity (property with a soul?). There could be no more tangible symptom of this contradiction than the object of property who reconciled the humanity of the master with that of the slave within the seamless compass of his or her own physical being. In some colonies and states and at some times, even up to the twentieth century, a "mulatto" category was officially acknowledged.⁴⁰ Significantly, this tended to occur when whites were demographically outnumbered and, as in the highly labor-intensive rice economy of South Carolina, relied on an intermediary or buffer population to stave off the threat of slave revolt.⁴¹ For a brief period in eighteenth-century Georgia, free blacks could even become white, though this extraordinary exception only obtained while Georgian whites were in a frontier situation and needing all the help they could get to suppress Native Americans and, to the south, the Spanish.⁴² In general, however, people who straddled the boundaries whereby the juridical opposition between slave and free was coterminal with the racial opposition between black and white presented so many problems that they were not granted the acknowledgment that a separate category would have entailed. 23

     Although the picture is, therefore, admittedly uneven, something remarkable begins to happen once the slaves are emancipated. Along with the category free black, which ceases to have any meaning when all blacks are free, the mulatto category recedes as well.⁴³ This had begun to happen before the Civil War in the northern states, where all blacks, whether mulatto or otherwise, were subjected to oppressive restrictions that in many ways anticipated the Jim Crow system that was not to be established in the South until the 1880s.⁴⁴ This is very significant. What does it mean to free slaves and at the same time to homogenize the status of blackness? Apart from anything else, it means that the boundary that previously separated a free black from a slave disappears, which is to say that, in place of the slaves, a new and more inclusive oppressed category emerges, one that, being defined by race, does not admit the awkward exceptions and contradictions manumission had entailed for the peculiar institution of slavery.⁴⁵ Emancipation, in short, canceled out the exemption—you can be an ex-slave but you can't be ex-black.⁴⁶ In common with emancipation itself, this expansion of race was initially a piecemeal process. Joanne Melish has documented how, as early as the late eighteenth century, perceptions of difference began to harden into "notions of permanent and innate hierarchy—that is, 'race.'" This development was not simply a response to the problems that revolutionary rhetoric posed for the continuation of slavery, however. Rather, the intensified discourse of race "began to emerge in the course of the first northern implementation of systematic emancipation."⁴⁷ 24

     Though born of slavery, therefore, race came into its own with the demise of slavery.⁴⁸ For all its usefulness as a justification, so long as slavery persisted, race was relatively redundant as a mode of domination. The point is, though, that the reverse applies—given race, slavery becomes redundant as a mode of domination. Since slavery was becoming redundant anyway (to put it mildly) on account of its irreconcilability with the flexible labor requirements of the emerging industrial economy, this consideration links race to industrial capitalism in a manner that cannot be expressed by means of a simple reduction of race to class. 25

     Prior to emancipation, although blacks had been spatially internal to U.S. society, their slavery had constituted a juridical barrier that insulated them from white society as decisively as physical externalization had separated Indians. Thus the consequence that emancipation had for blacks was comparable to that which containment had for Indians. Both became anomalies within. Before emancipation, the juridical barrier had not been absolute—although they were outside society, blacks could move closer in by way of manumission. It had been a social—and, accordingly, partly negotiable—barrier as well as a natural one. In the wake of emancipation, when this part social/part racial barrier gave way to an exclusively racial division between blacks and whites, it became unrelievedly natural. This occurred when, rather than merely threatening to let blacks in by the force of abstract logic, as in the case of Jeffersonian rhetoric, emancipation actually did let them in. Thus it was no accident that a concomitant response to the crisis provoked by emancipation should have been the colonization movement, which, in advocating the spatial externalization of blacks, proposed an alternative natural barrier with which to effect their separation.⁴⁹ 26

     An objection may suggest itself. If, in the absence of slavery, blacks became anomalous, like Indians inside the frontier, why did white society not seek to eliminate blacks in the same way as Indians—by assimilation? Indeed, in a passage removed from subsequent editions of the Jeffersoniad, Thomas Jefferson himself has been cited as suggesting just this solution to the problem posed by emancipation: "The course of events will likewise inevitably lead to a mixture of the whites and the blacks and as the former are about five times as numerous as the latter the blacks will ultimately be merged in the whites."⁵⁰ But five to one is not nearly as comfortable a disproportion as fifty or a hundred to one. Not that demography is an answer in itself. Nor is it simply a natural occurrence. Rather, demographic imbalance is a product of history.⁵¹ In this case, it represents the difference between one group of people who had survived a centuries-long genocidal catastrophe with correspondingly depleted numbers and another group who, as commodities, had been preserved, their reproduction constituting a singularly primitive form of accumulation for their owners. Moreover, these histories were ongoing. In large areas of the agricultural South, for instance, the ending of slavery did not mean that blacks became anomalous overnight. On the contrary, they continued to furnish a cheap source of labor.⁵² Even when unemployed, their mere presence as a hyperexploitable alternative depressed white workers' wages. Thus we need to be clear: in the wake of slavery, blacks did not become physically anomalous as labor; they became juridically anomalous as equals. In the case of Indians within, by contrast, their very presence was anomalous—as Gary Nash put it, whites "coveted Indian land but not land with Indians on it."⁵³ 27

     Since the oppression of blacks outlived emancipation, we should not allow the discontinuation of slavery to distract us from the continuities that obtain. As Frederick Cooper, Thomas Holt, and Rebecca Scott (one of them, at least) observed of histories that fail to link the slave era to the present, "Slave labor could be analyzed in economic, social, and political terms, but free labor was often defined as simply the ending of coercion, not as a structure of labor control that needed to be analyzed in its own way."⁵⁴ With the fading of the false dawn of Reconstruction, race intensified as a structure of social control,⁵⁵ its increased salience being expressed in a variety of ways, from mob barbarity to juridico-bureaucratic nicety. Miscegenation discourse encompassed the full range of race's domain, from the "black beast rapist" that animated the rhetoric of lynching to the tortuous formulations with which legislators and judges sought to locate the point where whiteness stopped and blackness began.⁵⁶ Even though some states retained legislation that technically whitened people with a blood quantum of no more than (usually) one-sixteenth African descent, the trend—at least, after the landmark 1896 case of Plessy vs. Ferguson⁵⁷ —was steadily in the direction of what came to be known as the "one-drop rule," in which any evidence of any African ancestry whatsoever, no matter how far back or remote, meant that one was classified as black.⁵⁸ There is, of course, considerable irony in the fact that the one-drop rule makes black blood immeasurably stronger than white (or, for that matter, any other) blood, even though, in white discourse, this strength consists in an unlimited power to contaminate. The corollary to—or ideological product of—the hyperpotency attributed to black blood is white racial purity. 28

     Jim Crow laws, whose obsessive quarantining of every drop of black blood reflected a wider segregation, was but one development in a series. There had been no need for segregation in the antebellum South, when slavery had precluded any doubt as to who was in power. That situation in turn was different from the one prevailing before the rise of the plantocracy in the seventeenth-century mainland colonies, when slaveowners' ranks had included blacks, Indians, and whites. So, too, with miscegenation discourse. It is a long way back from the lynch mob's black beast rapist to the Virginia of 1630, when Hugh Davis was sentenced to be "soundly whipt before an assembly of negroes&others for abusing himself to the dishonor of God and shame of Christianity by defiling his body in lying with a negro"⁵⁹ —especially since the unfortunate Hugh would probably have been whipped whoever he had been caught fornicating with. 29

     The twentieth-century intensification of the one-drop rule took place in the continuing vacuum created by the abolition of slavery and the demise of the Black Codes, as post-Reconstruction state legislatures sought new mechanisms to deliver the across-the-board system of racial control that had previously been delivered by slavery. As C. Vann Woodward remarked, the Jim Crow laws "put the authority of the state or city in the voice of the street-car conductor, the railway brakeman, the bus driver, the theater usher, and also into the voice of the hoodlum of the public parks and playgrounds."⁶⁰ The zeal with which blackness was excluded in the Jim Crow era was one aspect of a wider polarization in which whiteness was being consolidated at a time when it was multiply threatened, not only by the persistent absence of the slavery that had once served to define it but by the continually renewed immigration of people who were neither Anglo-Saxon nor Protestant. This is not to say that a concern with whiteness was by any means new to the twentieth century. The term "white" had figured in opposition to blacks and Indians since at least 1691, when the Virginia assembly had passed their much-quoted statute designed to prevent "that abominable mixture and spurious issue which hereafter may increase in this dominion, as well by negroes, mulattoes, and Indians intermarrying with English, or other white women, as by their unlawful accompanying with one another."⁶¹ During the nineteenth century, in the United States as in much of Europe, race came to be bound up in nation-building, whiteness becoming entangled with manifest destiny under the aegis of what George Fredrickson termed "white nationalism."⁶² Even though white consciousness was not, therefore, new, it was not until the twentieth century that whiteness was first defined in law, under the Virginia anti-miscegenation legislation of 1924.⁶³ Leon Higginbotham and Barbara Kopytoff strikingly summarized the accelerating polarization in which this innovation participated: 30

In the early twentieth century, Virginians made the first change in their definition of mulatto in 125 years. From the Act of 1785 to 1910, a mulatto, or "colored" person was someone who had one-fourth or more Negro blood. In 1910, that category was expanded to include anyone with one-sixteenth or more Negro blood, and many people previously classified as white became legally colored. Then, in 1924, in a statute frankly entitled "Preservation of Racial Integrity," the legislators for the first time defined "white" rather than "mulatto" or "colored." The statute, which forbade a white person to marry any non-white, defined "white" as someone who had "no trace whatsoever of any blood other than Caucasian" or no more than one-sixteenth American Indian blood. In 1930, the Virginia legislature defined "colored" in a similar, though slightly less restrictive way as any "person in whom there is ascertainable any negro blood."⁶⁴

The contrast between the one-drop rule and the Australian assimilation policy, together with the crucial opposition of land and labor, could hardly be clearer. Moreover, as already indicated, when we move to policies on Indians, we find a similar contrast. In the U.S. case, however, we not only see the contrast, we also see how these seemingly antithetical regimes of difference interact and complement each other in practice. Indeed, the following section is intended to show that, through an examination of discourses on Indians, we also come to understand more about discourses on blacks, and vice versa. Thus it is an argument of this essay that established accounts of the two histories have suffered from treating them in isolation from each other.⁶⁵

In a letter of March 1757 to his brother Moses, Peter Fountaine, a Huguenot descendant of Westover, Virginia, complained of the "many base wretches among us" who took up with Negro women, "by which means the country swarms with mulatto bastards" who, once three generations removed, would, "by the indulgent laws of the country," be allowed to intermarry with whites. As he continued: 31

Now, if, instead of this abominable practice which hath polluted the blood of so many among us, we had taken Indian wives in the first place, it would have been some compensation for their lands. They are a free people, and the offspring would not have been born in a state of slavery. We should become the rightful heirs to their lands and should not have smutted our blood, for the Indian children when born are as white as the Spaniards or Portuguese, and were it not for the practice of going naked in summer and besmearing themselves with bears grease, etc., they would continue white.⁶⁶

Peter was nothing if not succinct. This short passage bristles with themes that would animate American racial discourse for centuries to come: the recruitment of Indians to furnish their usurpers with sovereign rights to the soil, the hereditary nature of the stain of slavery, the vulnerable purity of white blood,⁶⁷ and the environmental determination of Indians' physical characteristics. For our purposes, however, the key theme is the stark contrast between Indian women's acceptability as marriage partners and the taboo on African blood.

     Few Australians would be pragmatic enough openly to advocate miscegenation as a solution to the ideological conundrum presented by the fact that civic institutions rested on the seizure of Aboriginal lands. Nevertheless, the perception that indigenous people's physical substance was assimilable into the European stock is common to the two national histories. In both cases, this perception occurred in contexts that also encompassed other eliminatory strategies, notably homicide, removal, confinement to reservations, child abduction, and a range of procedures intended to bring about cultural assimilation, from missions and boarding schools for the young to legislation that sought to transform Native Title into a scattering of alienable private lots. Moreover, all but the first of these strategies could unite hard-line native-haters and philanthropists who saw in assimilation an opportunity for native uplift. Noting how some assimilationists began to propose miscegenation as a solution to the "Indian problem" at around the same time as state legislatures began to introduce laws against Indian/white intermarriage, for instance, Alden Vaughan stressed the consistency of outcome: "But whatever the solution—miscegenation, allotment of farmlands in the East, removal to the West, or education in white-controlled boarding schools—the Indian was marked for gradual extinction by the uneasy coalition of his friends and foes."⁶⁸ In combination, the attributes of marriageability and cultural malleability provided for Indians' difference to be erased either physically, culturally, or both. This was, of course, in complete contrast to the regime imposed on blacks, whose difference was made absolute, essential, and refractory.⁶⁹ The disparity between blacks' and Indians' respective eligibility for assimilation reproduced a deeper distinction between the complementary discourses of land and labor on which U.S. society was ultimately predicated. As Ronald Takaki summarized this complementarity, "[I]n order to make way for white settlement and the expansion of both cotton cultivation and the market, some 70,000 Choctaws, Creeks, Cherokees, Seminoles, and Chickasaws were uprooted and deprived of their lands, and hundreds of thousands of blacks were moved into the Southwest to work the soil as slaves."⁷⁰ 32

     The fact that Europeans generally found Indians to be less viable objects of enslavement than Africans was not a consequence of Indian (let alone of African) biology.⁷¹ Rather, it was first and foremost a consequence of the fact that successful regimes of slavery characteristically involve alienation—not just the natal alienation that Orlando Patterson has made famous,⁷² since indigenous societies in both Australia and the United States have also had their children taken away, but the spatial alienation that slave transportation effected. There was nothing inherently unenslaveable about Indians. Rather, there is an inherent tension between enslavement and indigeneity. Thus the problems attending the enslavement of Indians largely ceased to apply if they were exchanged for Africans from Caribbean plantations—indeed, as Jean-Baptiste Le Moyne de Bienville, founder of New Orleans, urged upon the French crown early in the eighteenth century, such exchanges would prevent either group of slaves from escaping, since Indians could hardly run away from the islands, while Africans in Louisiana would be contained by fear of the surrounding Indians.⁷³ When we look behind the unstable distinctions of race and color to the historical relationships that they encode and reproduce, it is entirely consistent that the racial regimes to which Aboriginal people in Australia, for all their blackness, have been subjected should be antithetical to those that have been imposed on blacks in the United States but closely parallel to those that have been imposed on Indians. 33

     In formal terms, there is, of course, a major discrepancy between the colonization of North American Indians and the colonization of Aboriginal people in Australia. No matter how often treaties were broken, their mere existence presupposed the acknowledgement of at least some form of native title, even sovereignty.⁷⁴ In practice, however, there is less to this difference than meets the eye. For instance, sovereignty emerges as at best a mixed blessing when it makes relations international enough to allow the army to be used against you. In general, as the regularity of their breakage indicates, we should see treaties as tactical advances, designed to pacify, not to compensate, rather than as concessions. Moreover, when we look beyond straightforward treaty violation, a practice whose conformity with the logic of elimination hardly needs elaborating, we encounter a string of ostensibly respectable policy innovations whose practical outcomes led to the same end. With post-treaty Indians contained within the settler-colonial nation-state, however, a specifically racial component, blood quanta, becomes a key feature of policy, converting international relations into identity politics. Genetic admixture serves to detach individuals from the tribe, rendering them assimilable (as in the Australian case) to mainstream society.⁷⁵ After internalization, in other words, the racialization of Indians presents a mirror image of the post-emancipation racialization of U.S. blacks that we have already observed. 34

     Despite this fundamental discrepancy between the racialization of Indians and of blacks, in either case we find race intensifying when social space becomes, or threatens to become, shared.⁷⁶ To this extent, the Indian experience and the black experience are comparable after all. The analogy ends here, however, since, though their sharing of social space with whites led to both being (re-)racialized, the forms that the respective racializations took were diametrically opposed, in a manner that reflected and preserved the foundational distinction between land and labor. For, whereas race for black people became an indelible trait that would survive any amount of admixture, race for Indians became an inherently descending quantity that was terminally susceptible to dilution. Mixed-bloodedness became the post-frontier version of the vanishing Indian. The point is worth stressing because it enables us to see how race is a manifold regime that, in conducing to a plurality of outcomes, is at once both unitary and heterogeneous. In that it marked Indians and blacks out for diametrically separate destinies, race was heterogeneous. Unless this heterogeneity is understood, various aspects of the two groups' histories become paradoxical in relation to each other. For instance, whereas Indians who assimilated conformed to the requirements of this regime, blacks who passed as whites frustrated it. In a similar vein, without an appreciation of the antithetical but complementary histories involved, the 160-acre allotments with which the Bureau of Indian Affairs sought to break up Indian society could seem to represent a much better deal than the forty acres and a mule with which black people hoped to establish an independent social basis for themselves after the Civil War. This essential heterogeneity notwithstanding, the separate destinies that race inscribed harmoniously reproduced the foundational structures of U.S. society, simultaneously providing for both the elimination of Indians and the exclusion of blacks. As such, the two disparate racializations together served a unitary end. 35

     On the basis of these considerations, we can now turn to post–treaty era federal Indian policy with a focus on the problems presented by Indians' interiority. As I hope to show, this focus brings out the strategic continuity linking a series of measures that might otherwise seem discontinuous. Although the details are complex, and the differences between the various policy initiatives substantial, each of them combined a rhetoric of Indian improvement with assimilative measures that sought to reshape Indian institutions into conformity with dominant-society models in a manner that facilitated the transfer of Indian resources into white hands. 36

     In the case of the 1887 General Allotment (or Dawes Severalty) Act and associated legislation, tribal ownership of land was to be broken down into individual allotments.⁷⁷ In addition to substituting the propertied individual for the clan or tribe as the basic unit of Indian society, this legislation provided for the alienation of surplus reservation land left over after the allotments had been parceled out. Ostensibly, this surplus would arise because, as John Locke had pointed out, agriculture is more efficient—it takes up less space—than hunting.⁷⁸ But there was another, less remarked but crucial, factor in the generation of surpluses: restrictions that prevented "full bloods" from selling their allotments (to prevent them from being duped) did not apply in the case of the presumedly more civilized "mixed bloods."⁷⁹ In the event, Indians lost about two-thirds of their land (down from around 155 million acres to around 52 million acres) between the early 1880s and 1934.⁸⁰ In addition, the "checkerboarding," or interpenetration of reservation land with white-owned allotments that resulted from the sales of surplus, contributed to further Indian/white intermarriage and, thus, to an increased number of ineligible heirs. Allotment was brought to an end by the New Deal reforms associated with John Collier's dynamic stint as commissioner of Indian Affairs, which sanctified the principle of tribal self-government. Nonetheless, tribes that reorganized under the 1934 act found themselves adopting a distinctly Anglo-Saxon form of governance by way of the Bureau of Indian Affair's model constitution, which usually specified blood quantum–based membership criteria and included the phrase "subject to the approval of the Secretary of the Interior," whereby elected tribal authorities did not have the final say over expenditure or land use.⁸¹ As many have complained, this form of co-optation facilitated concessionary arrangements in which Indian resources came to be made available to non-Indian interests at inequitable rates. The post–World War II drive to termination, or withdrawal, represented a thoroughgoing rejection of, and diametrical departure from, Collier's policies. In combination with the policy of relocation, which provided subsidies for reservation Indians, particularly mixed bloods, to move to cities, assimilating to mainstream urban culture and leaving a predominantly full-blooded traditional remnant to wither away on reservations, termination provided for Indians to be freed from federal controls and granted the ordinary rights and duties of citizenship.⁸² The legislation sought to remove a burden on federal (as opposed to state) expenditure while facilitating private land sales on the part of individual Indians who opted out of tribal management plans in a manner that recalled severalty as it had operated until 1934.⁸³ Thus it is not surprising that the post-1960s policy of self-determination, a term initially coined by President Richard Nixon in a context of American Indian Movement militancy, should hark back in significant ways to the Collier reforms. In particular, even though tribal governments acquired the power to contract a number of government services, not only was their capacity to do so tightly bureaucratically circumscribed, but official permission was required before they could enter into contracts, which were subject to the political vicissitudes governing congressional appropriations. To a considerable extent, self-determination has involved Indians implementing federal policies rather than deciding how to run their own lives—as Joyotpaul Chaudhuri has put it, "The 'self' in self-determination remains in large part non-Indian."⁸⁴ Moreover, through all the above policy shifts, mixed bloodedness has operated as a synonym for—or at least a conduit to—a wider cultural and political assimilation whose achievement would amount to a dissolution of Indianness, a process that Annette Jaimes has termed "statistical extermination."⁸⁵ 37

     Like severalty itself, this principle has deeper roots than the Dawes legislation, or even the 1871 act that brought treatymaking to an end. In the 1858 treaty between the Ponca and the U.S. government, for instance: 38

The Ponca being desirous of making provision for their half-breed relatives, it is agreed that those who prefer and elect to reside among them shall be permitted to do so, and be entitled to and enjoy all the rights and privileges of members of the tribe, but to those who have chosen and left the tribe to reside among the whites and follow the pursuits of civilized life . . . there shall be issued scrip for one hundred and sixty acres of land each, which shall be receivable at the United States land-offices in the same manner, and be subject to the same rules and regulations as military bounty-land warrants.⁸⁶

For those left this (white) side of the frontier—which is to say, those who become internalized—the essential features of the post-treaty, Dawes-style assimilation program are already in place here. The tribe goes. In its stead, individual mergers into white society are effected by means of allotments of land. (Even the extent, 160 acres, anticipates the post-1887 agricultural standard.) Mixed-bloodedness is a key operator. Thus it is important not to see the different modalities of the logic of elimination as a tidy chronological sequence. Here, geographical removal and socio-cultural assimilation are two sides of the same coin. Either way, the Ponca tribe ceases to obstruct white access to its territory.

     One aspect of the Ponca treaty does, however, stand out as contrasting with the post-Dawes regime. Although mixed-bloodedness is an operator (in that it denotes those eligible for assimilation), it has no implications for tribal membership. Here, the relationship between blood quantum discourse and the internalization of Indian societies is particularly clear. For the Poncas whose mixed-bloodedness is without consequence are those who remain external by virtue of consenting to removal. Externally, the U.S. government's Indian problem was a tribal one. Assimilating individual members would not make tribal territory—which was collectively held—available. Moreover, for treaty purposes, it was in the U.S. interest for tribes to be as composite as possible. Breaking them down into smaller units would only necessitate additional treaties. Prior to internalization, in other words, the federal government depended on the very tribal governments that it would subsequently seek to dismantle.⁸⁷ Once a tribe was internalized, however, its government formed an intervening layer that obstructed the U.S. government's access to individual Indians. The impediment to assimilating tribes into the body politic was not simply that they were collective entities, since the United States encompassed other collectivities—in particular, of course, the states. Rather, tribes were unassimilable because they were separately and independently constituted entities whose organizing principles were discordant with those that governed the structurally regular institutions of U.S. society, which were uniformly organized around the centrality of private property. Thus the obstacle to the Indian Territory's admission to statehood was not its collective constitution but its commitment to collective ownership. Failing the allotment of tribal land, as Vine Deloria and Clifford Lytle explained, 39

It was inconceivable to the federal officials that a state could be admitted to the Union that did not provide for free commerce with other states, and the communal holding of land struck directly at the personal land tenure system already entrenched in the other states. If an Indian could not sell a tract of land within a state, how could the other states have equal status with the newly admitted Indian state and how could commerce proceed when the best that white citizens might ever achieve within the new Indian state might be the leasing of lands?⁸⁸

     Allotment, in sum, had two inseparable aspects: the end of tribal government and the production of the propertied individual. As a means of converting tribal membership into a fragile form of property ownership on an individual basis, blood quantum discourse became central to this transformation. Thus it is important to stress that the Dawes legislation invented neither severalty nor the accelerated dispossession/assimilation of mixed bloods.⁸⁹ Rather, it set the seal on a deeper and more diffuse historical tendency that derived its logic from the most elementary premise of the settler-colonial project, the requirement for undisputed (or "quiet") possession of territory.⁹⁰ The resultant maximization of mixed bloods' access to the dubious privilege of assimilation antithetically complements the increasingly rigorous exclusion of the once-enslaved, whose involuntary contribution to the North American colonial formation had been one of labor rather than of land. 40

     Summarized thus, the situation is bound to appear more regular than it actually was. In practice, there were all sorts of exceptions to the general rule. The example of enslaved Indians has already been noted, to which we might add the phenomenon of slaves being owned and traded by Indians and by other blacks.⁹¹ In some respects, blacks could be treated like Indians, as in the colonization movement. It provided for blacks' removal to geographically remote places, while, at various points in the nineteenth century, the fear was expressed by some whites that blacks could die out.⁹² Yet these two counter-tendencies both surfaced as part and parcel of their proponents' programs for post-slavery U.S. society. The exceptions are, therefore, arguable. Nonetheless, they should not be dismissed. This is because, sharing a language of race, rhetoric and policies on Indians and blacks did tend to bleed into one another, producing occasions when a common vocabulary of difference could apply indiscriminately to either or both. Thus I would not want to be taken to be legislating for every individual situation or event. On the level of the whole, though, there is no doubt that eliminatory policies such as warfare, removal, and assimilation were characteristically applied to Indians and not to blacks. Such regularities are empirically real and should not be particularized out of historical description. On the same basis, so far as discourses of miscegenation are concerned, some states' opposition to intermarriage between Indians and Europeans notwithstanding,⁹³ the overall picture is one of a tolerance extending to encouragement, whose contrast with policies toward blacks could hardly be more marked. Even the Virginia anti-miscegenation legislation of 1924, notorious as a high-water mark in Jim Crow codification, conceded the so-called Pocohontas exception, in which certain categories of Indian-European unions were specifically exempted from the statute's otherwise draconian catalog of proscriptions.⁹⁴ Given the manifest continuity between this legislation and the regularities that we have noted, it would be misleading to dismiss it as an idiosyncracy of the legislature concerned or to attribute it to the cultural half-life of John Rolfe.⁹⁵ While such factors may well have played a part, a different level of analysis is also required, lest we fail to recognize the deeper historical motivation to which the legislation conformed, and which it thereby reproduced.⁹⁶ 41

     Just as the assimilation of Indians into the white population was consistent with the logic of elimination, so was the assimilation of whites into the Indian population anathema to it ("squaw men," etc.).⁹⁷ If this much is obvious, it should also be noted that Indian-black unions became problematic to the extent that they threatened to produce part-black Indians.⁹⁸ Part-Indian blacks did not occasion comparable concern (a contradiction that became incarnate in the person classified red when on the reservation and black when off it).⁹⁹ It was stated above that discourses on Indians and on blacks should be situated in relation to each other. Even the one-drop rule, apparently so specifically targeted at African descent, emerges in a fuller light once it is recognized as not only sanitizing the white population¹⁰⁰ but as simultaneously eliminating the Indian population through its assimilation of red-black people to the black category. A logic that once made slaves of "mustees" today makes blacks of Indians by excluding mixed categories from census forms (a situation that compounds the irony of Zora Neale Hurston's "I am the only Negro in the United States whose grandfather on the mother's side was not an Indian chief").¹⁰¹ Providing as it does for any color so long as it's black, the one-drop rule makes black unhyphenable. Accordingly (and pace Jack Forbes), there is no such category as red-black people—indeed, no such category as anything-black people. There are only black people. Thus even in Louisiana, where creole classification attains a complexity that might seem to confound the rigid binarism of the one-drop rule¹⁰² , one category—the white one—stands out as monolithically undivided, leaving the rest as so many permutations of black. 42

     Thus the key factor in colonial and "post"-colonial race relations is not, as some have argued, simple demographic numbers,¹⁰³ since populations have to be differentiated before they can be counted. Difference, it cannot be stressed enough, is not simply given. It is the outcome of differentiation, which is an intensely conflictual process. If a one-drop rule applied in Australia, for instance, the Aboriginal population would escalate overnight. Hence the incendiary effect of a Queensland bumper sticker, the display of which was truly for none but the brave, which proclaimed an "Aboriginal family reunion—invite your white relatives."¹⁰⁴ Rather than simple counting, demography involves the most complex and tortuous contestation, as in native Virginians' century-long struggle to refuse categorization as "colored," a struggle that was waged, as Forbes remarked, "with uneven success and . . . which served to poison African-American Indian relations as well as to split communities, churches, and even families."¹⁰⁵ Miscegenation discourse is about holding the line when it comes to power, privilege, and access to resources. As such, it is at the material core of identity politics, which should not be discounted as merely aesthetic or superstructural. 43

     In the context of capitalist property relations, there is no tension between the antithetical discourses of race to which blacks and Indians have been subject. This is because the simple undifferentiated product of the encounter between African labor and Indian land was European property. In this connection, obvious though it may seem, it is surely significant that the most durable names that have been applied to the two groups, Negro and Indian, refer respectively to a bodily characteristic and a territorial designation.¹⁰⁶ As observed in relation to the doctrine of terra nullius, the mixture of labor and land was central to the ideological edifice of private property. As slaves, blacks were chattels or instruments comparable to plows or horses, whose application to vacant land realized its value and converted it from wilderness into property. Chattel slavery, one's appropriation of another's body, presupposed a prior alienation. One was appropriated, the other expropriated. The twin dimensions of this Janus-faced procedure cannot be appreciated separately. 44

     Having reached this point, we are in a position to appreciate the distinctiveness of the Brazilian situation. 45

Compare the rigorous binarism of the one-drop rule or the simple unidirectionality of the Australian assimilation policy with the extravagance of the Brazilian¹⁰⁷ system of color classification (see Figure 2).¹⁰⁸ If ever there was a worthy challenge to class analysis, this system surely offers it. Truly baroque in its excess, it establishes an apparently unassailable limit to racial classification. What was it about Brazilian race relations that could have produced such a scheme? 46

     The fact that this system so spectacularly abjures a hard and fast binarism has sustained a misleading historiography of slavery and race relations in Brazil. The inspiration for this stems principally from the Brazilian historian Gilberto Freyre, whose Casa-grande&senzala, originally published in Portuguese in 1933, appearing in English translation as The Masters and the Slaves in 1946, fostered the myth that the Portuguese had been relatively benign slavers. Freyre's claims inspired a number of American historians—in particular Frank Tannenbaum, Stanley Elkins, and even, to a lesser extent, Winthrop Jordan—to assert that Brazilian slavery had been milder than the North American variant because, whereas Iberians in the New World were familiar with slavery, which had been codified and regulated in a reasonably humane manner derived from Roman Law and filtered through the Catholic Church, slavery had not been institutionalized in England for centuries, so there existed no rules to regulate the practice and restrain its excesses.¹⁰⁹ The complex Brazilian system of color classification testified to this comparative mildness, since it indicated that, rather than a rigorously polarized society that ruthlessly distinguished between master and slave, the Portuguese had presided over an integrated polity in which manumission had been commonplace and people could move up and down the hierarchy with relative ease.¹¹⁰ So far as the Portuguese are concerned—and without holding any brief for Anglo-American slavers—the notion of a mild form of slavery could hardly be further from the truth. 47

     In comparison to North American slaves, whose lives represented a valuable commodity to their masters and were accordingly carefully—albeit not kindly—preserved, Africans who were enslaved into the Brazilian sugar industry could expect brutally truncated lives unless they were manumitted.¹¹¹ Although estimates continue to diverge as to the precise numbers involved, a negative demographic regime in which high death and low fertility rates were accompanied by a major gender imbalance in favor of male imports necessitated constant resupplies from Africa. Few would now quarrel with Stuart Schwartz's characterization of the adult mortality and general fertility rates among slaves in eighteenth-century Brazil as "staggering . . . far worse than recorded in other slave regimes."¹¹² 48

     Even though the importation of slaves into the United States declined substantially from the 1790s on, this did not greatly affect the system of slavery there, since the slaves could and did reproduce themselves, albeit with a little help from their masters, so that slave numbers actually grew in the nineteenth-century South.¹¹³ In Brazil, on the other hand, there was no pretense of natural increase. Rather, the lives of slaves were simply used up, whereupon they were replaced with what Thomas Nelson, a British surgeon resident in Rio during the 1840s, referred to as "the shoals of doomed Africans who are annually drawn from the opposite shore to supply the defects."¹¹⁴ The importation continued apace until 1851, when the British finally terminated it by means of a naval blockade, and Brazilian slavery began to break down. Although Brazilian slaves were not to be finally emancipated until 1888, the combination of the curtailment of the trade from Africa, which had ceased altogether by 1853, and the effects of the Rio Branco (or "free womb") law of 1871, in which the condition of slavery was no longer transmitted to offspring once they had reached their majority, meant, in the words of Richard Graham, that "no new slaves would be available either from Africa or from procreation."¹¹⁵ The predominantly Brazilian-born group of slaves who remained after the 1870s were well equipped to resist their subjection in a variety of ways, culminating in the mass flights from slavery that took place in the 1880s, often along the very railroad system that had been installed to improve plantations' export efficiency.¹¹⁶ 49

     One can discern all sorts of reasons for this apparently profligate waste of human resources on the part of the Portuguese and their Brazilian successors. The immediate financial returns on putting an African to work on a Brazilian sugar plantation were much higher than in the case of a Virginian tobacco plantation.¹¹⁷ Apart from anything else, the distance involved was considerably—almost 50 percent—shorter, so it was much cheaper to ship Africans to Brazil.¹¹⁸ The Portuguese controlled both ends of the Brazil slave trade, so there were fewer middlemen and levies to be encompassed in the price paid by their ultimate owners. Moreover, since the Portuguese crown imposed per capita levies on both Angolan exports and Brazilian imports, high volumes of trade were encouraged.¹¹⁹ Above all, from the outset, Portugal's domestic economy had been basically agricultural, producing little for export, a situation that led to a circulation economy, in which a complex range of commodities, including slaves ("black ivory"), indirectly complemented one another at the level of the seaborne empire as a whole. The sugar plantations on the islands of São Tomé and Fernando Po, for instance, from which the Brazilian industry was to be developed, were initially established to provide an outlet for surplus slaves whom the Portuguese had acquired in the course of opening up African markets.¹²⁰ Thus slaves were not simply a means to the end of sugar production. Rather, trading in them generated systemic value in its own right as one of the primary links in a global chain of commodity exchanges.¹²¹ 50

     Concerning the question of internal complements, whereby local Brazilian factors converged with these systemic or empire-wide conditions, it is clear, firstly, that constantly replacing slaves with fresh imports from Africa militates against the development of a culture of resistance among them.¹²² Moreover—and crucially—the much-vaunted high rate of manumission in Brazil conduced to the same end. In other words, manumission is not inconsistent with the practice of working slaves into the ground only to replace them with fresh consignments. Rather than being mutually inconsistent moral options, the two commonly contribute to the discouragement of slave revolts.¹²³ In place of a sterile alternation between Lusitanian apologists who stress the manumissions while discounting the mortality rates and critics who reciprocally discount the manumissions in favor of the deaths, we should see the two features of the Brazilian system as harmoniously preempting the threat that slave solidarity would pose to the reproduction of a social system based on slavery. To put this another way, manumission and high mortality together subtended a high turnover strategy that, internally, prevented slaves from developing the consciousness of a social group capable of acting for itself at the same time as, externally, it conformed to the empire-wide requirement for an expanded flow of trade. 51

     What, then, was the internal context in which this rationale came to prevail? Simply put, in the areas of Brazil that the Portuguese colonized, the natives were exterminated, assimilated, or marginalized, while the millions of Africans who were imported over the centuries heavily outnumbered the Portuguese, whether peninsular or creole. Beset by perennial fears of underpopulation, Portugal—which, as observed, lacked England's early industrial experience of enclosure and urban drift—did not feel a comparable need to export an unwanted surplus population to the New World, with the result that the Portuguese stock in Brazil was proportionately much smaller than the white component of Britain's Australian or American colonies. Thus the reason why the Portuguese should have gone to such apparently uneconomical lengths to prevent the development of solidarity among its African slaves and creolized slaves and ex-slaves is only too obvious. The combined group greatly outnumbered the Portuguese, and would have had little trouble in overthrowing them if ever (or, perhaps, if only) they had set their collective mind to it. That they have not done so is a matter of record. Accordingly, throughout Brazilian history, blacks have consistently dominated the lowest, most impoverished and exploited positions in society. To this day, though it is a Brazilian cliché that money whitens, since there are some well-off blacks and a larger number of poor whites, the great majority of blacks are poor and the great majority of the elite are white.¹²⁴ 52

     Slave creolization is inherently ambivalent. On the one hand, a shared language and a degree of shared cultural experience contribute to diasporan solidarities, while, on the other hand, the same qualifications also provide avenues for the maintenance of planter control. As Schwartz has shown, this ambivalence manifested itself in Brazil as a division between two schools of slaveowners, "those who thought that permitting slaves to maintain their African cultures was a positive way of stimulating differences among them and thus an effective social control, and those who thought that such cultural persistence stimulated rebellion."¹²⁵ Dividing around such questions, the planters were not, however, divided to an extent comparable to the divisions that fragmented those of African descent. In times of crisis, planters of all persuasions knew only too well where their collective interest lay, and seamlessly closed ranks. 53

     In postcolonial Brazil's nineteenth-century imperial era, the significance of the divisions that plantation society strove to impose on slaves, ex-slaves, their successors and descendants lies in the simple fact of division itself. Juridically, Afro-Brazilians were divided into slave versus free, with freed slaves in turn being divided into those born free versus those who had been manumitted (libertos) and, after about 1830, those who had been released from slave ships intercepted by the British (emancipados). After the passing of the free womb law of 1871, children born to slave mothers but destined for freedom became known as ingénuos. In addition to these formal juridical distinctions, a range of informal social distinctions obtained: African versus creole; black versus mulatto; Indian versus black, mulatto, and caboclo (Indian/white); together with a locally various range of phenotypical oppositions. None of these divisions were coterminal. On the contrary, they cut across each other, broke each other up, and multiplied the fragmentation. In combination, they secured and maintained the dominance of the ruling group, the only group whose divisions could be relied on to dissolve when collective interests were threatened.¹²⁶ The baroque system of nomenclature grew out of and compounded this fragmentation, on whose reproduction across time Brazilian society depended. This is why creole homogeneity presented such a threat, and why, therefore, the number of creole slaves had to be kept in check—which, in turn, is why the number of fresh imports and manumissions had to be kept up. Creolization was a matter of degree—slaves who had been around for a while, like Ira Berlin's Atlantic creoles, posed more of a threat than those who had not.¹²⁷ Every year on Brazilian soil was another year of creolization. Thus we can understand how it should be that a plural but not unduly Gordian set of social classifications should begin to complexify in the nineteenth century, during the very period when Brazilian slavery was finally and decisively shifting, through a period of creole majority, to the impossible situation of an entire slave population that was only distinguishable from the surrounding population by virtue of its slavery. 54

     Since Afro-Brazilian deprivation survived the emancipation decree of 1888, it is important not to exaggerate the significance of that decree.¹²⁸ Nonetheless, along with the juridical condition of slavery, emancipation automatically abolished the juridical distinctions that had previously served to divide Afro-Brazilians. This left the unofficial distinctions, which were predominantly couched in terms of color. In the twentieth century, these informal distinctions effloresced into the full extravagance of the Brazilian baroque.¹²⁹ Clearly, there are significant correspondences between this phenomenon and the post-emancipation racialization of U.S. blacks. In particular, both emancipations signaled an intensification of miscegenation discourse. Yet the outcomes of this intensification could hardly have differed more profoundly between the two societies. Whereas, in the United States, the one-drop rule enforced the most thoroughgoing of racial polarities, the Brazilian baroque obscured an empirical polarity in which African extraction overwhelmingly correlated with deprivation. Where one promoted solidarity among a white majority, the other promoted fragmentation among an Afro-Brazilian majority. In fragmenting, of course, it also contained. 55

     In a very straightforward way, the contrast between the Brazilian and U.S. regimes of difference reflects the bare demographies involved. In Brazil, a one-drop rule would have resulted in instant engulfment for the elite. This consideration casts light on a post-emancipation complement to the color-classification system, the energetic program of white immigration that gathered impetus from the 1880s on.¹³⁰ In terms of conventional economic rationality, the logic of this program is hardly less baroque than that of the color terminology, since the Brazilian economy already had millions of laborers on hand and available to it without the trouble and expense of importing foreigners. Yet this labor force had hardly evinced docility in the run up to emancipation.¹³¹ As an impoverished potential alternative, however, the beneficiaries of emancipation served, by their very presence, to discipline the newcomers.¹³² Indeed, when the immigrants showed signs of having imported disruptive European ideologies such as socialism or syndicalism, they could find themselves passed over in favor of Afro-Brazilians.¹³³ Such situations were, however, limited exceptions to the rule. In general, Brazilian authorities acted as if, along with slavery, they had also dispensed with the slaves. Ex-slaves in Brazil were much more marginalized than blacks in the U.S. South, whose labor continued to be exploited under changed forms of control.¹³⁴ Thus it is consistent that, unlike U.S. blacks, but like the anomalous Indians and Aborigines within, blacks in Brazil should have been targeted for assimilation under the policy of whitening (branqueamento). Even the iconography anticipated the Australian three-generation lap-count to whiteness (see Figure 3).¹³⁵ 56

branco
mulato
mulato claro
moreno claro
negro
escuro
claro
roxo
sarará escuro
preto claro
cor de cinza
caboclo escuro
branco sarará
branco caboclado
mulato sarará
cor de cinza clara
louro
mambebe
roxo de cabelo bom

preto
moreno
mulato escuro
moreno escuro
caboclo
cabo verde
araçuaba
car de canela
roxo claro
vermelho
pardo
mambebe
moreno escuro
negro
creole
vermelho
pelé
preto escuro

Figure 2 : Brazilian Color Classifications

     In sum, the Brazilian system of color classification performs a socially reproductive function complementary to that which, in the slave era, was also performed by the combination of manumission, high mortality, and juridical heterogeneity. These factors operated to prevent a hyperexploited Afro-Brazilian majority from realizing its community. Following the protracted (1851–1888) build-up to emancipation, distinctions of color acquired increased salience as the other factors became obsolete. Subsequently, in combination with the program of white immigration, the elaboration of the color-classification system helped prolong the oppression of Afro-Brazilians into the post-slavery era. Rather than elucidating this situation, historians such as Tannenbaum and Elkins recapitulated Brazilian racial ideology. On the other hand, although Carl Degler's famous "mulatto escape hatch" was not well received by Brazilian critics at the time, we should recognize the value of the insight that the categories intervening between black and white function to discourage non-whites from electing to be the majority.¹³⁶ We should, however, amend Degler's thesis, since the reprieve that counts is not opened up for individual mulattos but for the dominant group as a whole. The Brazilian baroque is a ruling-class escape hatch. 57

It follows that the Brazilian categories are not racial in the sense that applies to the racialization of black people in the United States—after all, full siblings in Brazil can be assigned different classifications, and even these vary with the judgments of individuals.¹³⁷ This is the reason for referring to the Brazilian system as one of color rather than of race. For all the stark differences between the systems, however, race in the United States (or, for that matter, in Australia) and color in Brazil are alike in being regimes of difference that reproduce relations of power, whether by maintaining social divisions that would otherwise be incoherent or by effacing social divisions that would otherwise be coherent. To focus on miscegenation discourse is to focus on that incoherence at the point where it is most conspicuous, which is the point at which the ideological resources of a colonial or "post"-colonial society that is premised on distinguishing between colonizer and colonized are most intensely summoned. At that point, those resources become maximally visible. The primary issue is not, therefore, race or sexuality per se but the maintenance of social divisions, an imperative that requires difference to be configured and reconfigured in highly contextual manners. By the same token, I am presenting an approach or methodology, not a model. Accordingly, if we were to turn our attention to, say, the differentiation of Loyalist from Republican in Northern Ireland, or of Mizrahi from Palestinian in Israel, or of Anglo- from French-Canadian in Quebec, we should not expect to find differences configured in the ways we have seen them configured in Brazil, the Unit