ABSTRACT
Explanation, the goal toward which research should strive, stems from models that are superimposed to the evidence. These models are continually adapted to the increasingly richer knowledge produced by analyses of the data, all the more so in the framework of a multidisciplinary perspective benefiting from several different points of view. In fact, archaeological evidence is the product of choice and invention. In other words, it is a construct that is perpetually redefined. The models used to describe and explain complexity can only stress selected aspects of the underlying processes at the expense of a multitude of others, because of the vast numbers of variables involved and the lacunae in our data sets.
