The concept of "scales of space and time" (Lemke 2000) has been tremendously influential in linguistic anthropology and related fields, and scalar approaches have proved useful for illuminating connections between what used to be called "micro" and "macro" phenomena. This paper disrupts "rhetorics of discontinuity" (Darnell 2001) in present-day work, arguing that many of the ideas in contemporary work on scales were already implicit in early Americanist anthropology. Through renewed engagement with the debate among the Boasians A.L. Kroeber, Edward Sapir, and Ruth Benedict on questions of individuality, culture, and social action, we can rethink "scale" and bring forth both the useful and problematic aspects of scale as a heuristic. I argue that linguistic anthropologists' reliance on scalar metaphors should be examined carefully, since the apparent scalability or nonscalability of talk and social action is, itself, a cultural phenomenon that emerges from decisions to scale our analytic vision in particular ways. Revisiting Americanist arguments about the nature of culture reveals some of the limitations of "scale," as currently conceived, but also suggests that scalar analysis might be fruitfully integrated with linguistic anthropological understandings of agency, visa -vis voice, and temporality. [scales, scalar analysis, temporality, agency, history of anthropology] S calar analyses, or those that seek to attend simultaneously to phenomena unfolding on different spatial and temporal "scales," have become increasingly prevalent in sociocultural studies of language. The concept of "scales of space and time" (Lemke 2000) has been influential in linguistic anthropology and sociolinguistics (e.g.