ABSTRACT
An ethnographer cannot avoid selecting some people, objects, or events for study, thereby renouncing, for a time at least, the possibility of studying others. From a vast range of possibilities, he takes up work in a particular tribe, village, or town; questions certain respondents; employs a few informants; observes some artefacts, situations, or behavioural events, and makes observations at restricted times. If the word ‘sampling’ is used so broadly, then fieldworkers are constantly sampling the universe of people, situations, objects and behavioural events with which they are occupied. Seldom, however, do they keep track of how they drew a sample or report its composition. Even statements as general as my pseudonymous list of principal Kaska Indian informants and subjects,1 and Margaret Mead’s ‘neighbourhood maps’,2 identifying the adolescent and pre-adolescent girls she observed, are rare. An anthropologist characteristically extends his remarks beyond his sample and talks about ‘the’ Kaska Indians and Samoan girls or about child rearing, quarrels and pottery techniques in general-as though he had studied the community, category, or topic exhaustively. The usual spoken implication is that for his problem, the sample adequately represented a larger universe of actors, topics, culture patterns, techniques, or other units under study and, therefore, could provide reliable information about that universe as it existed at a particular time. A statistically conscious observer might object and point out that for a sample to be considered in a strict sense representative of the universe whence it came, it must have been selected in a suitable manner. Anthropologists are likely to respond by protesting that it is not they who decide what persons or events to use as sources of data; such decisions are practically made for them when certain individuals volunteer their help, some groups extend welcome, and some techniques happen to be accessible to observation.3 That units force themselves on a researcher’s attention, is merely a figure of speech. It overlooks the fieldworker’s readiness to respond positively or negatively to certain cues in the field situation and ignores his active involvement in deciding how to respond to environmental opportunities or when to surrender to unbreachable limitations.
