Rethinking methods and paradigms of ceramic chronology.

Abstract

Methods of ceramic chronology building are based on certain assumptions concerning the pattern of stylistic change in ceramics. These assumptions are, however, not necessarily identical in different methods. Also, the general applicability of the assumptions in each method is not endorsed by solid empirical observations of stylistic change in ceramics and theoretical considerations concerning processes producing stylistic change in ceramics. The inapplicability of assumptions of a method undermines the reliability of ceramic chronology created by the method. In order to evaluate the reliability of existing ceramic chronologies, (1) theoretical considerations were made concerning processes producing stylistic change in ceramics and (2) empirical observations were made concerning aspects of stylistic change in ceramics in a well-controlled archaeological setting, i.e., stylistic change of Tusayan White and Gray Wares in the American Southwest between A.D. 850 and 1150 where tree-ring dating is available as an independent means of temporal control. As a result, it was revealed that (1) substantial temporal overlap can be present in the manufacture of successive styles of ceramics, (2) continuity criteria of the typological method are not necessarily applicable to stylistic change in ceramics even in a continuous population, and (3) significantly large time lags can be present in the diffusion of manufacturing frequencies of styles even within an area in which the styles are shared. In light of these findings, the typological method cannot be accepted as a method of ceramic chronology building. Occurrence and frequency seriations are, on the other hand, acceptable methods. However, for reliable chronological seriation attention must be paid to potential errors caused by contemporaneous variation of stylistic compositions among assemblages due to time lags in diffusion and variation in generational composition of individuals who produced assemblages.