Home » Posts tagged 'bayesian'
Tag Archives: bayesian
The Big Questions: Where, how and when were monuments built, used and perceived by the people who built them and others, coming after, who re-used them?
The construction of detailed chronologies of when the monuments were built and used will not only draw on AMS 14C dating but also optically stimulated luminescence on stone structures and sediments that otherwise cannot be dated, and innovative cosmogenic nuclide dating, all within Bayesian frameworks. Whilst chronology building and refinement are immediate goals, the project will move beyond simplistic evolutionary models, to explore self-similarity in monuments through new fuzzy logic classificatory techniques, identifying monuments that are more similar to each other than to other monuments.
The geography of the Neolithic period has often been poorly served by phenomenological readings of the monuments that are based on poor or no data. In seeking to locate contemporaneous settlements and other activity areas in proximity to the monuments, this project will explore the issue of grain-size in the Neolithic exploitation of land as a subsistence, experiential and cognitive resource. This will require somewhat more sophisticated modelling than has been the norm.
The relationship between monument and horizon is thought significant. We will explore the relation between the site, the landscape and the celestial sphere from GIS-driven modelling, harnessing the enormous computing power now available to generate digital terrain map (DTM) data that render 360° panoramic landscapes, providing quantitative data about the landscape for statistical analyses as well as recreating how the world appeared to people in the Neolithic. Visibility to and from monuments will be tested in GIS reconstructions drawing on spatially precise local-scale pollen data on the density around a monument of woodland. The hypothesis that many may have been constructed in woodland and may have been invisible even from very short distances will be tested.