CAA Conference has ended

View analytic
Tuesday, March 27 • 4:00pm - 6:00pm
Poster - Digging out uncertainty from the ground

Sign up or log in to save this to your schedule and see who's attending!

The problem concerning the use of data collected on the surface as reliable indication of sub-surface remains represents a crucial issue for the practice of archaeological surveys, at both regional and site level. Any sort of surface investigation postulates a correlation between surface and sub-surface remains by over-estimating the value of the data collected on the surface. Nevertheless, there are clear evidences (e. g. geological post-depositional factors, past and actual human activity, etc.) showing the unreliability of the surface remains as data to trust for detecting spatial patterns and archaeological features both on regional and intra-site level. The surface collection may offer a biased understanding of the original population for a series of causes: 1) local movements of finds can merge clusters that were originally separated; 2) surface finds are likely to move down a slope; 3) collection rates can be affected by alluvial and aeolian natural processes; 4) the burial of a site by sediments moved by gravity (Colluviation); 5) collection rates depending on the skills of individual surveyors; 6) tillage can damage artefacts and confuse the spatial patterns of finds collected on the surface. Therefore, the present paper aims to acknowledge uncertainty as a quantitative estimation of error present in data collected by surface investigation and to show how all measurements contain some degree of uncertainty generated through systematic error and/or random error. I will show, by using different case studies, how the existing kinds of probabilistic sampling techniques (e. g. systematic sampling, random sampling, and judgmental sampling) can cope with the problem represented by the missing data. Finally, I will offer a range of possible solutions (e.g. Bayesian statistics, further archaeological excavations, remote sensing techniques) addressed to reduce uncertainty by correcting for systematic error and minimizing random error and to assess how the patterns characteristics of the surface remains approximate to those of the parent population.


Alessio Palmisano

UCL (University College London)

Tuesday March 27, 2012 4:00pm - 6:00pm
Building 65, South Corridor

Attendees (2)