Tradition, Genuine or Spurious

LIKE MANY SCHOLARLY CONCEPTS, is at once a commonsense and a scientific category. In its commonsense meaning, tradition refers to an inherited body of customs and beliefs. In the social sciences, an ongoing discourse has attempted to refine this understanding of tradition as it has proven empirically and theoretically inadequate. Recent efforts to clarify the concept of tradition, most notably those of Edward Shils (1971, 1981), do much to add nuance to our conventional understanding but leave unresolved a major ambiguity: does tradition refer to a core of inherited culture traits whose continuity and boundedness are analogous to that of a natural object, or must tradition be understood as a wholly symbolic construction? We will argue that the latter is the only viable understanding-a conclusion we have arrived at by comparing our independent investigations in two quite disparate ethnographic situations. In our attempts to analyze national and ethnic identification in Quebec and Hawaii we have concluded that tradition cannot be defined in terms of boundedness, givenness, or essence. Rather, tradition refers to an interpretive process that embodies both continuity and discontinuity. As a scientific concept, tradition fails when those who use it are unable to detach it from the implications of Western common sense, which presumes that an unchanging core of ideas and customs is always handed down to us from the past. As many writers have noted (e.g., Eisenstadt 1973; Rudolph and Rudolph 1967; Singer 1972; Tipps 1973), one inadequacy of the conventional understanding of tradition is that it posits a false dichotomy between tradition and modernity as fixed and mutually exclusive states. M. E. Smith (1982) has pointed out that and are interpretive rather than descriptive terms: since all cultures change ceaselessly, there can only be what is new, although what is new can take on symbolic value as traditional. Following Smith's lead, we can see that designating any part of culture as old or new, traditional or modern, has two problematic implications. First, this approach encourages us to see culture and tradition naturalistically, as bounded entities made up of constituent parts that are themselves bounded objects. Second, in this atomistic paradigm we treat culture and its constituents as entities having an essence apart from our interpretation of them; we attempt to specify, for example, which trait is old, which new, and to show how traits fit