This essay is concerned with the construction of colonial categories and national identities and with those people who ambiguously straddled, crossed, and threatened these imperial divides.1 It begins with a story about métissage (interracial unions) and the sorts of progeny to which it gave rise (referred to as métis, mixed bloods) in French Indochina at the turn of the century. It is a story with multiple versions about people whose cultural sensibilities, physical being, and political sentiments called into question the distinctions of difference which maintained the neat boundaries of colonial rule. Its plot and resolution defy the treatment of European nationalist impulses and colonial racist policies as discrete projects, since here it was in the conflation of racial category, sexual morality, cultural competence and national identity that the case was contested and politically charged. In a broader sense, it allows me to address one of the tensions of empire which this essay only begins to sketch: the relationship between the discourses of inclusion, humanitarianism, and equality which informed liberal policy at the turn of the century in colonial Southeast Asia and the exclusionary, discriminatory practices which were reactive to, coexistent with, and perhaps inherent in liberalism itself.2

The object of this essay has been to help examine spatiotemporal variation in literacy. The research reported here centered on the Valley of Oaxaca, an agricultural region in southern Mexico, during the period from 1890 to 1980. The data consist of a systematic compilation of tax and voting lists from the nineteenth century, census responses from 1890 to 1980, community ethnographies, published histories and biographies, and government reports. Attending to both the spatial and the temporal scales of events and causes was methodologically important for this research.

In the late nineteenth and early twentieth centuries, Argentina and Canada experienced unprecedented economic growth. In the period stretching from 1890 to 1914, Argentina and Canada played host to millions of migrating Europeans and became the largest borrowers on the world's capital markets. The infusion of foreign labour and capital helped to convert the empty grasslands into bread baskets for the world.

The expansion was propelled by crops in the export sector, mainly cereals cultivated on the Argentine pampas and the Canadian prairies. By the early years of this century, wheat became the premier export for both countries, and eventually ranked among the world's top cereal exporters. After World War I, both countries combined to supply around 60 percent of the world's total wheat export trade.1 Argentina and Canada exemplified what was beneficial about export-led development.

Recent dramatic changes in the social and political organization of Eastern Europe and what was the Soviet Union have led to a widespread reformulation of certain generic terms that have long plagued comparative scholarship. Similarly, with the destruction of the monolithic Berlin Wall has come the imperative to deconstruct monolithic terms, such as communism or traditionalism, which have often obfuscated difference and negated geographical and historical specificity. In this essay, and in the spirit of laying to rest the ghost of the ideal type, I compare the work and authority relations in Chinese and Soviet factories in the 1960s and 1970s. When the differing variables that coalesced to form distinctive patterns of labor relations in these two countries during those years are more clearly understood, it will be possible to discuss the current patterns of change with greater accuracy. In addition, when the essentially structuralist constraints of an overarching communist type are loosened, it is also possible to reintroduce actors into the dialectic and to enrich the comparison with finer social and historical detail.

Over the last decade, studies of ‘third world’ histories and cultures have come to draw to a very considerable extent upon the theoretical perspectives provided by poststructuralism and postmodernism. With the publication in 1978 of Edward Said's work, Orientalism, these perspectives—now fused and extended into a distinctive amalgam of cultural critique, Foucauldian approaches to power, engaged ‘politics of difference,’ and postmodernist emphases on the decentered and the heterogeneous—began to be appropriated in a major way for the study of non-European histories and cultures. Certainly in our own field of Indian colonial history, Said's characteristic blending of these themes has now become virtually a paradigm for a new generation of historians and anthropologists. These directions have been most recently and sharply endorsed in Gyan Prakash's discussion, ‘Writing Post-Orientalist Histories of the Third World: Perspectives from Indian Historiography.’

When George Balandier proposed his theoretical approach to a colonial situation, the colonization of language was not an issue that piqued the interest of scholars in history, sociology, economics, or anthropology, which were the primary disciplines targeted in his article. When some fifteen years later Michel Foucault underlined the social and historical significance of language (‘l'énoncé*’) and discursive formation, the colonization of language was still not an issue to those attentive to the archaeology of knowledge. Such an archaeology, founded on the paradigmatic example generally understood as the Western tradition, overlooked the case history in which an archaeology of discursive formation would have led to the very root of the massive colonization of language which began in the sixteenth century with the expansion of the Spanish and Portuguese empires.

The problem with Prakash, O'Hanlon and Washbrook conclude, is that he tries to ride two horses at once—one Marxist, the other poststructuralist deconstructionist. ‘But one of these may not be a horse that brooks inconstant riders. …’ So, they say we must choose only one to ride on, not both because the two, in their view, have opposing trajectories. One advances historical understanding and progressive change, the other denies history and perpetuates a retrogressive status quo. Posed in this manner, the choices involve more than a dispute over which paradigm provides a better understanding of the histories of the third world and India. At stake is the writing of history as political practice, and the only safe bet, from their point of view, is Marxism (of their kind), not the endless deferral and nihilism of deconstruction and postmodernism. Having set up this opposition, O'Hanlon and Washbrook's either/or logic has no place for the productive tension that the combination of Marxist and deconstructive approaches generates. They are uncomfortable with those recent writings that employ Marxist categories to analyze patterns of inequalities and exploitation while also using deconstructive approaches to contend that Marxism is part of the history that institutionalized capitalist dominance—approaches which argue that although Marxism can rightfully claim that it historicizes the emergence of capitalism as a world force, it cannot disavow its history as a nineteenth-century European discourse that universalized the mode-of-production narrative.

Since the publication of Edward Said's Orientalism in 1978, it has been difficult for anthropology to avoid the fact that its own discourse is ever entangled in a whole Western archive. What became clear, of course, was that the categories through which anthropology constructs descriptions and analyses of the social discourses and practices of non-Western peoples are themselves participants in a network of relations of knowledge and power. Interestingly enough, however, whereas the general import of this Foucauldian thesis has now been quickly assimilated, its challenge has hardly been taken up in terms of tracing out the lines of formation of specific anthropological, or, let us say, anthropologized, concepts.

In one of the twenty lines it allocates to a description of Hungary, the nearly 300-page edition in 1877 of A Satchel Guide for the Vacation Tourist in Europe summarizes the architectural and aesthetic worth of the country's capital city for sightseeing American visitors by pronouncing that, in Budapest, “the churches and the public buildings are of no particular interest” (Satchel 1877:194). Twenty years later, the 1897 edition of that same guidebook takes a more amiable but scarcely enthusiastic pitch, allowing that “some of the new public buildings are elegant in their way” (Satchel 1897:184). Twenty-seven years later—following a world war, two revolutions, and a foreign military occupation resulting in the dismemberment of the Austro-Hungarian Dual Monarchy—the presence of what has remained of Hungary is noted by an increase to eighty-one lines (Satchel 1924). Except for a one-sentence reference to a Danubian steamboat trip downstream from Pressburg (Bratislava, Pozsony), the entire description remains restricted to Budapest. Nearly two generations after the pronouncement of the disparaging opinion above, the 1924 text notices that Budapest's “picture at sunset is one of the most striking in Europe” (Satchel 1924:272) and that “it is not only the most considerable city of Hungary, but is probably to be numbered among the four most beautiful capitals of Europe” (1924:273).

Stamp collecting and industrial capitalism in the United States emerged simultaneously in the mid-nineteenth century. England issued the first government postage stamp in 1840, and other nations quickly adopted the idea. The United States printed its first official stamp in 1847, although it was preceded by the provisionals issued by local postmasters. Postage stamps were a product of the industrial revolution. The adoption of the prepaid penny post in England, while opposed by the General Post Office, was widely supported by large merchants who understood that a low-cost, single-rate system was vital to the communication demanded by an increasingly national market. The adhesive label was originally conceived by the English postal reformer, Rowland Hill, as a convenience for illiterates who would not be able to write addresses on the official envelopes that he preferred as proof of prepayment. Within a decade, almost every major nation in the world had borrowed this device, which became a symbol of the economic transformation of the nineteenth century. I would argue that the collecting of these tokens was a microcosmic performance of the system that created them. Stamp collectors took on many of the key roles of actors in the market economy and played out various conflicts embodied in the larger society in the philatelic arena.

Colonial discourse, sometimes referred to in the singular, seems unmanageably vast and heterogeneous, for it must encompass not only the broad field of colonialism's relations and representations which constitutes or arises from the business of official rule, including administrative reports and censuses, but also the works of metropolitan literature and other forms of high culture which deploy images of the exotic or the primitive, paintings of unfamiliar landscapes, tourist guides, anthropological studies, and Oriental fabric designs. Colonial discourse includes chinoiserie, Kim, the Victoria and Albert Museum, Camus' Algerian stories, Frans Post, and Indiana Jones, as well as the Vital Statistics of the Native Population for the Year 1887 and the annual reports from wherever.