Use this URL to cite or link to this record in EThOS:
Title: Validating classical multivariate models in archaeology : English medieval bellfounding as a case study
Author: Bayliss, Alexandra Louise
Awarding Body: University of London
Current Institution: University College London (University of London)
Date of Award: 2006
Availability of Full Text:
Access through EThOS:
Access through Institution:
The principal aim of this study is to apply various methods of numeric classification and ordination (commonly used by archaeologists) to the incidence matrix of stamps occurring on medieval bells from England, and to compare the results with what is known independently about these data from documentary sources. The incidence matrix records the presence of 1116 stamps on 3390 bells. Recorded bells have been assigned to 89 different founders, 51 of whom have bells appearing in the incidence matrix. Three varieties of cluster analysis and correspondence analysis have been applied to this matrix. These analyses reveal clusters of bells and stamps relating to particular founders and foundries, and the relative chronological sequence in which the bells were cast and the stamps used. The success of each technique in defining these clusters and sequences accurately has been tested quantitatively by comparing the results of each analysis with the documentary record. For this to be valid, it is vital that the link between the documentary evidence and the surviving bells is rigorous and explicit. The criteria which have been used to link these two types of data are discussed in Chapter 2. The results of the different mathematical approaches are given in Chapters 3 and 4. Only k-means cluster analysis provides results which are consistently in disagreement with the documentary evidence. The other techniques allocate between two-thirds and three quarters of bells or stamps accurately to clusters which relate to particular founders or foundries. Correspondence analysis has proven particularly successful at identifying clusters of bells which relate to foundries. The techniques have been less successful at identifying accurate chronological series of bells or stamps, with other sources of variation predominating. Finally, some indication of the potential for such analyses to illuminate our understanding of the English medieval bellfounding industry is provided.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available