Estimating the Historical Timespan of Archaeological Phenomena

Al-Ruzaiza, Abdulrahman Sulaiman (1994) Estimating the Historical Timespan of Archaeological Phenomena. PhD thesis, University of Glasgow.

Abstract

When a reasonably representative selection of material can be collected from an archaeological event (eg. a particular site or culture), it is often of interest to archaeologists to derive a credible historical range, based on radiocarbon measurements, for the event. Using the concept of the floruit of a culture suggested by Ottaway (1973), this thesis provides a contribution in this sphere, by which a set of calibrated dates is used to provide a calendar range (i.e. floruit estimate) for the archaeological event. The thesis consists of three subsequent parts, each corresponding to a different aspect of this problem. Part one (chapter 2) deals with the basic calibration problem that uses a radiocarbon date to predict the corresponding calendar date. Part two (chapters 3-5 inclusive) considers the development and applications of five methods for estimating the floruit of an archaeological event. Part three (chapter 6) considers the question of temporal overlap between two archaeological cultures. Chapter 1 is a general introduction to this thesis and gives a brief background to archaeological contexts and to radiocarbon dating. Definitions of the problems involved in the thesis are given and the pursued aims outlined. Chapter 2 is concerned with the calibration problem and the fitting of the calibration curve. A non-parametric regression is used to fit the calibration curve and a Kernel method based on the regression estimator of Watson (1964) is employed. The technique of calibration of a single radiocarbon date is reviewed and the provision of point and interval estimates for calibrated dates is discussed with consideration of the problem of multiple point estimates. In Chapter 3, the definition of the floruit is given and the idea of a quartile interval is considered and employed in development of five methods of floruit estimation. Three of these methods require only pencil, paper and the calibration curve to be applied, while the other two seek to estimate a frequency distribution underlying the archaeological event. A non-parametric kernel density estimation is used to construct an estimate for this frequency distribution. A cross-valldation method is used to choose the smoothing parameter but, as it is not entirely satisfactory, alternative weighting methods are considered. Chapter 4 involves a simulation study to investigate the performance of the floruit interval estimation methods presented in the previous chapter over different parts of the calibration curve. Data are simulated from Normal, Skewed and Bimodal distribution functions for different combinations of sample sizes and Inter-Quartile Ranges (I.Q.R.) of the underlying frequency distribution. The performance of the five methods are compared according to four statistical criteria. Simulation results indicate that density estimation methods are superior to the others. In Chapter 5, the performance of the five methods are compared by applying them to a selection of groups of real radiocarbon dates collected from different archaeological cultures. The effect of several factors on the floruit estimates are considered in these applications and the results for each are summarised and discussed. Chapter 6 propounds two different approaches to quantify the degree of temporal overlap between two archaeological cultures on the basis of a set of radiocarbon dates from each culture. One is based on the proportional overlap between their frequency distributions, while the other is based on the time overlap between their floruits. Applications of these methods on real data are presented and discussed. Finally, Chapter 7 outlines the conclusions from the preceding chapters and suggests some ideas for further work.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Additional Information: Adviser: Tom C Aitchison
• Keywords: Archaeology
• Date of Award: 1994
• Depositing User: Enlighten Team
• Unique ID: glathesis:1994-76437
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 19 Nov 2019 14:20
• Last Modified: 19 Nov 2019 14:20
• URI: https://theses.gla.ac.uk/id/eprint/76437