PALÉORIENT vol. 8/ 1 1982

ANALOGIES, ANOMALIES AND RESEARCH STRATEGY

H. CRAWFORD

irrémédiablement disparu.

Analogy, explicit or otherwise, has always been the lynch pin of archaeological interpretation. The unknown has to be interpreted in terms of the known. In common with other rather ill-defined assumptions, analogy has been brought out into the daylight, stripped of its rhetoric and required to justify its use as a conceptual tool. The rise and fall of the use of archaeological analogies has been described and put into its historical perspective in an article by Orme (1 ) and she shows that there has been a lack of rigour in their application which brought the whole technique into disrepute. Today, when analogies lie at the basis of much archaeological modelling, exactness of fit between the model and the archaeologically observed situation is of extreme importance. An analogy is essentially descriptive of a part or a whole, of form or of process; once tested for fit and incorporated in a model it can become explanatory and even in some cases predictive as well. If the fit is faulty the whole model collapses. Binford made the further suggestion that analogies could be used to test models (2) but analogies by themselves cannot test or prove anything; they indicate possibilities and probabilities; they also underline improbabilities. Proof, in the sense that the same experiment always produces the same result, is seldom available in archaeology and proving a negative is if anything more difficult than proving a positive. Analogies cannot by themselves provide this proof but they can suggest areas of research which may in the end provide data to support or condemn a model.

The analogies used in archaeology are often drawn from ethnographic material and are used in two main ways, the broadest application deals with universal processes and general principles; by using material drawn from a global spectrum of areas and periods it seeks to establish guidelines and parameters in such general areas as food gathering, trading and the interactions of groups of peoples. Examples of this sort of work are now accepted as of considerable value, for instance Sahlins' work on "Stone Age Economics" (3) opened up many new ideas for prehistoric trade, while Hodder's work on the relationship between political and social boundaries and stylistic change in material culture will be of obvious relevance in considering the spread of any prehistoric pottery style (4). In these circumstances goodness of fit is hardly an issue. However, when analogy is used as an interpretive tool for more specific problems the position is entirely different. It is proposed that for an analogy to be illuminating in these circumstances the societies to be compared should match each other in three important particulars; in ecology, that is in similar geographical settings with similar resources; in economy, that is to say they should exploit their ecology in a broadly similar fashion (pastoralists cannot be compared with agriculturists for example); and finally they should have a similar level of technological development. These three criteria have the advantage that they are relatively easy to establish objectively from the archaeological evidence which social criteria are not. In addition, Childe(5) stressed the importance of