In what ways are these vulnerabilities inherent in the way ethnography has been harnessed to the project of anthropology? As a method, ethnography seems quaintly anachronistic. It puts its primary trust in evidence generated by the human senses, being vested in the direct empirical gaze—the naked eye, as it were. This way of seeing characterized the early-nineteenth-century biological sciences (Foucault), which served as a model for early anthropology, described by one of its founders as the "natural science of society" (Radcliffe-Brown). Yet despite occasional scientific pretensions, anthropology has remained at base a resolutely humanistic art (Evans-Pritchard, 1950) vested in insights generated by a "long conversation" between ethnographer and subject(s) (Bloch). Mastering this somewhat underspecified art is still a crucial component of recruitment to the anthropological profession. And certain canons of practice did become standard, at least as ideals: fluency in the primary field language, for example; minimal reliance on formal interviews and artificial situations; sufficient time in situ to distinguish patterns of repetition from random events; and sensitivity to the ethical implications of one's presence and one's representations of local life. The qualitative nature of such observation, coupled with the anthropological concern with demonstrating the salience and complexity of meaning and value in human action, makes writing a crucial aspect of the ethnographic enterprise: the observer must communicate his or her insights in language that conveys the expressive richness and texture of local experience, yet retains the authority of objective reportage. In the late twentieth century, a heightened sensitivity in the human sciences to the independent life—and political mediation—of texts brought the conventions of ethnographic prose under new scrutiny. More than one observer has noted that ethnography preserves features from an era of travel writing and adventure and that it is still dependent in large part on the legitimacy of first-hand experience and interpretation (Pratt).
Yet it could be argued that what some see as the greatest weaknesses of ethnography are also its major strengths. Most significantly, its practitioners refuse to put their trust in techniques that grant scientific (and above all, quantifying) methods an exaggerated sense of objectivity. They distrust instruments that rely on a priori questions and units of analysis, for instance, preferring to derive their categories of inquiry from a direct engagement with the phenomena at issue; and they are critical of investigative means that reify human action by divorcing it from its social context or claim to eliminate subjective bias. The very concept of participant observation—an oxymoron to some—implies that knowledge is inseparable from its knower. Thus, while ethnographers do seek to objectify social facts in the world, and also to make viable generalizations about them, their assertions remain clearly identified as the fruit of interaction between reporters and their subjects. But in this, ethnography is merely a marked instance of the inescapable interplay of fact and value that renders all human observation partial, imperfect, and a product of its time and place.