********** My fieldwork is in Macedonia because jarring juxtapositions fascinate me. Picasso did it with legs and eyes; the Washington Post does it with bra ads and stories about human-rights abuses; and the Balkan Peninsula did it with cultural iconography and people--people now at war once shared villages, markets, even houses. Picasso doesn't do it any longer because he's dead; the Washington Post does it daily; it isn't done on the Balkan Peninsula any longer because ethnic cleansing and separatist movements have all but eliminated any signs of cultural difference from the once-shared spaces. Why is it that in art, in newspapers, or, indeed, in language--all of which are nonplaces-differences can be juxtaposed but that in reality, on the ground, they must be spatialized and distanced? Standing on the Balkan Peninsula, any geographer would be struck by a multitude of questions--about limestone caves, about linguistic diversity, about religious differences, about transhumance, about war. My questions were and are about the maintenance of ethnic diversity and cultural identity amid difference. Why study such a question there? Because in Macedonia, until this century, people shared spaces without assimilation; they shared places without acculturation. How they did so has been lost.' I imagined my question as much like that of an ethnobotanist trying to resurrect endangered or lost knowledge about plants, but mine was about space. I hoped to learn how so many groups, distinct in identity, had shared the same space and preserved their distinction amid such difference until the late twentieth century. How had people on the Balkan Peninsula conceptualized their space and their identity within that space? A METHOD? Establishing a fieldwork methodology was an enormous problem. I was searching for evidence of an alternative form of order, an alternative episteme that enabled a landscape of heterogeneous and overlapping spaces and public identities to be created and maintained. How could I conceptualize and understand that alternative sense of order if I didn't know what I was looking for? The methodological problem is actually greater than "How do I collect data?" for it lies at a far more fundamental level, which asks, "What are data?" I did know that the ethnospatial categories into which the peacemakers and ethnic cleansers chose to classify the people of the Balkan Peninsula may have caused once-integrated or coexistent societies to disintegrate. If these categories were, as I believed, the wrong categories, categories imported from afar, then what would locally produced categories look like? Even more precisely, because the people of the Balkan Peninsula had adopted the ethnospatial categories in the late nineteenth century, what did locally produced categories look like in the early nineteenth century and before? How did people categorize themselves then? I had no shortage of methodological models from which to draw. The best examples were historical surveys into periods of change, documenting peculiar points in time at which a variety of forces produced tremendous change: the fourteenth century, for example, or the early twentieth century. Historians, especially, have been doing work like this for the past thirty years. Their data have consisted of a variety of sources, including art (Edgerton 1975), architecture (Bourdieu 1977), science (Shapin and Schaffer 1989), representations (Alpers 1983), the use of language (Auerbach 1953), categorizations (Foucault 1994), and religion (Koyre 1957). The second set of models from which I drew a methodology was that of ethnobotanists. They sought to work backward, much as I did. Their work was historical, like mine. And they tried to piece together a system of order that once created a world. I set out to do the same. I devised a methodology mirroring similar studies in the United States and Western Europe. My underlying assumption was that there is difference in the world. …