Abstract:

Intensive surveying is restricted to limited areas. As it is a line-walking method, there is a smearing effect on the results from averaging over larger stretches of land, compounded by archaeological visibility problems, the definition of diagnostics and walkers idiosyncrasies. Numerical correction cannot produce qualitative differentiation, however, and thus reliability (replicability) of the results is compromised. The present paper discusses a field methodology in which systematically arranged (gridded) points of two square metres are cleared of vegetation, and all non-soil objects in the points taken in for further analysis. That way, the definition of -diagnostics-, most of the time conceived of as practical/logistic in nature, but in the final count a reliability problem (e.g. Bintliff et al. 1999), is obviated and additional fabric analysis considerably enhances the database. Numerical analysis of the results from nine field seasons of the Riu Mannu Survey Project in Sardinia shows that visibility effects are reduced far below customary confidence levels. Illustrations are provided by a general description of the finds in one transect and a more detailed analysis of the finds from a Punic farm site.

Author: Pieter van de Velde