People on borders, and people who cross borders, people who live on boats, and people who are nomadic may be especially difficult to classify. They are not tied to a particular territory, and do not fit the agrarian model of identity that underlies both much of Western theorizing (ethnic group plus language plus homeland) and nationalist ideologies based on peasant culture. They represent the remnants of pre-modern identities in the sense that the people concerned do not have to have a clear answer to questions such as “What is your mother-tongue?” or “What is your ethnicity?” The questions asked by the ethnographer or linguist symbolize the approaching centralizing-categorization process of the state.

Border areas between modern states are zones of transition and of complex and often bitter identity politics and territorial claims. Ethnographically, borders can seem arbitrary or artificial, and political states are often concerned to contain or deny the transitional or marginal identities found there and to emphasize their control (symbolic and actual) over territory. The role of anthropologists and linguists in this is ambiguous. On the one hand, they often recognize the “artificiality” or contingency of political borders; yet they are frequently (in some sense) the agents of colonial or nationalist states. In contrast with pre-modern polities, empires and post-colonial national states have made it their business to find out who their subjects are. They have undertaken radical labelling enterprises, literal and metaphorical mapping exercises, in which the unknown, the uncharted, and the unclassified have been measured and catalogued. Colonial and nationalist states (unlike pre-modern ones) require a much higher degree of order within their naming and labelling systems. They set up centralized education systems and devise language policies; they register their citizens and carry out censuses. Categories and classifications become a pre-requisite of policy formation and instruments of social control.