This study is an analysis of inference as the product of archaeological subject matter. So far very little has been said about the data of archaeology except that they belong to the context of culture. According to Osgood's (1940) system for ordering data within this context, they are classified as material percepta and considered under the heading of techniculture. But this is a heading which covers all of the objects made and used by the “aggregate of human beings.” Even on the limited level of archaeology this means a tremendous coverage, far greater than is necessary for a case study. For practical purposes, therefore, some criteria of relevance are needed to define a small unit of techniculture which is appropriate for illustrating the nature of archaeological inference.