Skip to main content Skip to navigation

Statistical methods applied to Anglo-Saxon archaeological data

Interest has increased in statistical methods in archaeological studies of Anglo-Saxon England over the last five years. Accumulation of measured plans of archaeogical digs and Anglo-Saxon placename location has led to historians formulating rather specific and statistically testable hypotheses (Blair, 2013) which have led to definite statistical responses (Kendall, 2013, Zanella; 2016), producing very positive response from historians. A number of other statistically testable questions have arisen from the Leverhulme-funded project “Planning in the Early Medieval Landscape” (WSK served on the Advisory Board for this project).
The question addressed in Kendall (2013) concerns evidence for a common unit of measurement (the “Anglo-Saxon Perch”) in construction of large buildings (typically churches) in Anglo-Saxon Mercia over the seventh and eighth centuries. Adapting an ancestral method (Kendall, 1974), it was shown that measurements do indeed supply definite evidence for this use. Barnes (2015) has considered the statistical assessment of evidence for gridding in Anglo-Saxon building plans.
A similar question has now been raised concerning data arising from sites in Wessex and Northern France, where it is thought a rather different unit of measurement was in use. The challenge here is that the archaeological data (postholes rather than walls) requires pre-processing and cluster analysis in order to identify the probable walls between which relevant measurements exist, and this must be automated in such a way as to facilitate simulation-based inference.
This initial project would lead naturally to full PhD projects arising from a GIS dataset for all of the UK which was developed by the “Planning in the Early Medieval Landscape” project. This has stimulated ambitious hypotheses concerning the presence or absence of Anglo-Saxon grid-planning, and the impact this has had on the present landscape.
While this particular project is focussed on questions of Anglo-Saxon history, the resulting development of methodology will have much broader impact on a wide range of concerns in science and society in which GIS datasets find essential use.
A prerequisite for this project is experience with data analysis and programming in R or similar scientific computing environment (eg Matlab, or Python + Numpy).

Selected References

  1. Blair, J. (2013). Grid-Planning in Anglo-Saxon Settlements: the Short Perch and the Four-Perch Module. Anglo-Saxon Studies in Archaeology & History, 18, 55-57.
  2. Barnes, C. (2015). Statistics in Anglo-Saxon Archaeology, MSc Dissertation, University of Warwick.
  3. Kendall, D. G. (1974). Hunting Quanta. Philosophical Transactions of the Royal Society, London, Series A, 276(1257), 231-266.
  4. Kendall, W. S. (2013). Modules for Anglo-Saxon constructions: Appendix to Blair (2014).
  5. Zanella, G. (2016). Bayesian Complementary Clustering, MCMC, and Anglo-Saxon Placenames. Annals of Applied Statistics 9.4 1792-1822. See also Arxiv, 1409.6994.
  6. Planning in the Early Medieval Landscape