Masters Theses

Date of Award

3-1982

Degree Type

Thesis

Degree Name

Master of Arts

Major

Anthropology

Major Professor

Gerald F. Schroedl

Committee Members

Jefferson Chapman, Richard Jantz

Abstract

A majority of the archaeological research conducted on ceramic assemblages is directed towards delineation of assemblage associations or "relatedness" through the use of seriation and cluster algorithms. Research proposed here redefines this problem in terms of mathematical modelling procedures. The assemblage creation process is seen as stochastic and suitable for reduction to a programmable sherd production model.

Four normative groups are defined, a priori, in terms of the mean expected relative frequency of 33 whole vessel types. Thirty assemblages, each composed of 100 vessels, are developed as random variations about each of the four mean group vectors. The resulting whole vessel assemblages are then transformed into sherd assemblages according to the parameters of a random breakage model.

The whole vessel assemblages are subjected to seriation and cluster solutions to delineate the inherent assemblage associative interrelationships. The same techniques are then applied to the sherd assemblages which are defined in terms of nine traditional motif types, 26 motif/vessel-portion types, and 26 motif/vessel-portion types weighted by relative sherd size. The latter designation scheme provided the best correlation with the whole vessel solutions.

The results of this project suggest that mathematical modelling is a useful tool in artifact analysis in that it: (1) helps to organize theoretical beliefs and empirical observations, (2) improves understanding of archaeological systems, (3) suggests the need for detail in describing archaeological remains, (4) facilitates the manipulation of systems, and (5) is less expensive than comparable analyses of real systems.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS