The main agenda of this essay is that observation of and experimentation with archaeological materials cannot be separated from hypothesis building and testing, and, as a corollary, that basic research in archaeological materials is as much an integral part of archaeological question answering as the philosophical model building of Binford, Plog, and Clarke.
The two aspects of archaeological research discussed here - ethnoarchaeological and experimental archaeology - both manifest many of the attributes of traditional empirical investigation and have been spurned and belittled by the philosophers of archaeology as peripheral to the New Methodology, as unscientific, and as having only specific time-space utility. This chapter examines the methodological procedures by which these two aspects of our discipline may be transformed to play an integral part in hypothesis testing and in the formulation of probability statements in archaeological research.
We can define ethnoarchaeology as the structure for a series of observations on behavioral patterns of living societies which are designed to answer archaeologically oriented questions. "Experimental archaeology" - that is, experiments as part of archaeological investigations - on the other hand, comprises a series of observations on behavior that is artificially induced. Both may involve more or less rigorously controlled conditions and recorded results.
Both are important aspects of a materialist study of behavior, pertinent to the study not only of past behavior but also of that of the present and the future. Theoretically, both ethnoarchaeological and experimental observations should provide valuable resources for testing hypotheses concerning archaeological data. Ethnoarchaeological observations are not in competition with experimental observations in providing valuable data on behavior; their information is of a different kind, and this distinction should be made self-consciously when using these data, since it affects conclusions and the confirmation of hypotheses.
The stress in this chapter falls on the side of experiments in archaeology, since my own research has been directly concerned with this aspect of archaeological testing. But my argument will emphasize that both sources of information, although separate, increase their value in interpreting archaeological data when one recognizes and takes advantage of their close interrelationship and interdependence.