The classification of cultures into a workable number of types for descriptive or interpretative ends has occupied anthropologists since the science was born. Many kinds of data have been selected. Within the last decade Coon's (1948) subdivision of human societies into six levels on the basis of complexity of institutions, and the attempts by Strong (1948), Armillas (1948), Steward (1949), Willey and Phillips (1955) to distinguish developmental periods in the Mesoamerican and Andean archaeological sequences may be cited. Our excuse for attempting yet another formulation is that the current schemes emphasize either ethnographic criteria that are difficult or impossible to detect archaeologically, or unique features of particular cultural configurations rather than general criteria defining more universal patterns. Starting from a point of view different from those heretofore employed, we have tried to develop a classification of cultures that is usable with both ethnographical and archaeological data and that has functional and evolutionary as well as historical and descriptive significance.