There are many parallels between anthropological fieldwork and espionage. Both involve looking, listening, eavesdropping, taking notes, recording conversations, snapping photos, and establishing trusted confidants. We call it participant-observation; they call it spying. We seek informants; they seek informers. Both intend to understand and create a representation of someone else's reality. We craft these representations into an "ethnography" based on personal encounters, autobiographies, events, ritual participation, and lived experience; they craft them out of similar observational materials into an official "file."We try to understand the reality of the other, in which we participate, as ultimately having an existence independent of our own experience of it; they seek to document, distort, manipulate, change, and even fabricate that reality for their own ends. We offer interlocutors anonymity in our ethnographic texts, while they seek an ever more precise and intimate textual rendering of named individuals. At least this is how anthropologists from the postcolonial Cold War period understand the differences between fieldwork and state surveillance activities. These parallels and oppositions are changing, however, along with the meanings of self, ethnography, and the file. How might we register these changes in our efforts to understand another person or culture without succumbing to the unmooring effects of working in our era of terror politics? What kind of anticipatory reflections might help us avoid a through-the-looking-glass experience that threatens to constantly situate us in aporias about the constitution of reality or that just as easily slips into modes of conspiratorial paranoia with global reach?As both groups of experts-anthropologists and agents of state security--negotiate the empirical and ethical boundaries between what is to be revealed and what should remain secret, their projects of observation rival each other, with different effects on the observer and the observed. In the course of fieldwork, we are commonly mistaken for or accused of being spies, working as agents of the state in which we reside or of the institutions that fund our research. 1 Our method of participant-observation awakens the institutionalized paranoia of security apparatuses on all sides of the