Articles by Clive Orton

Archaeology continues to generate large amounts of data, in a growing range of formats and media. Old datasets have been or are being digitised, and there is increasing emphasis on the re-use of old datasets, and on preparing new datasets with re-use in mind. That being so, surprisingly little attention has been paid to the prevention and detection of errors in archaeological data, and in acquiring or developing robust methods of analysis. The sorts of errors that can be encountered in different types of data are approached and discussed through a series of case studies, dealing with counting errors, measurement errors, and classificatory errors. They are linked to another obstacle to the re-use of data: the lack of standardised terminology between different originators. Strategies for mitigating these problems (which cannot be totally overcome) are discussed.

Point pattern analysis has been one of archaeology's quantitative approaches since at least the 1970s, and has been applied at a range of scales, from the region to the site. Various techniques have been 'borrowed' from other disciplines, notably ecology, such as quadrat analysis, nearest-neighbour analysis and kernel density analysis. There have also been 'home-grown' techniques such as Local Density Analysis, Presab and Unconstrained Clustering, as well as the use of Cluster Analysis itself. This paper reviews these developments, assessing their strengths and weaknesses. A statistical advance was made in the 1970s with the development of the K function approach. This has become embodied in the ecological statistical software package ADE-4 as the Ripley and Intertype programs. These programs were found in a search for suitable affordable software for teaching spatial analysis at post-graduate level, and have been used in this role for three years, taking as a test-bed the Danish mesolithic site of Barmose I. The outcome of this work is presented as a case study and compared with earlier analyses of this dataset. The value of ADE-4 for archaeological spatial analysis is assessed.

Classification has been an important archaeological activity for at least a century. It should not be seen as an end in its own right, but as a tool that enables archaeologist to compare and communicate. It can also be seen as mathematical activity, the study of relationships between entities in a multi-dimensional space. Comparison of these approaches leads to a set of criteria for a “good” classification. A wide and growing range of techniques is available, but more fundamental issues such as the choice of variables and the level of analysis must also be considered. The history of classification in archaeology shows a period of optimism followed by one of disappointment; the recent development of techniques more suited to archaeological needs may enable a middle view to be taken. Finally, the role of classification is assessed in the light of current trends in archaeological theory.

As well as solving two long-standing theoretical problems, this work shows great potential for the interpretation of ceramic assemblages, and has implications for the way in which pottery is catalogued. Different sorts of interpretation (functional, chronological, distributional) are possible at different levels of grouping (context, phase and site assemblages).