All economic systems have three components: produc tion, distribution, and consumption. Archaeologists often assume that their tabulations of recovered goods constitute adequate discu sion of consumption (but see Michael Smith 1987; Brumfiel 1987; and Costin and Earle 1989 for more focused studies of consumption). Production and distribution have received more explicit investiga tion. While most archaeologists recognize the necessary link be tween the two, 1 distribution? usually discussed as" exchange" seems to receive more systematic attention in the archaeological li erature. In some instances (eg, Blackman et al. 1989; Johnson 1973; Morrow and Jeffries 1989; Torrence 1986), production is studied primarily as a way to understand the mode of exchange. The imbalance in attention is undeserved, not only because pro duction has just as important a role in society as does distribution but also because it may be that production is easier to study well. Exchange networks often cover large areas. As Torrence (1986: 139) has pointed out," It may seem obvious to the point of being trivial to emphasize that all aspects of an exchange system are unlikely to have taken place at a single one of its nodes/'Production activities, in contrast, are apt to be localized. Therefore, spatially restricted dat sets are more likely to contain the data appropriate for studying a fairly complete production system than to contain data representing a complete distribution system. Equally important, exchange event are invisible in the archaeological record, while production events often leave a clearer and more easily interpreted record in the for of debris, tools, and features, if not the products themselves. This chapter has two primary goals. The first is to review and cri tique the ways in which archaeologists have defined the organization 1