Linewalking was conducted on available ploughed fields, with further surface collections using follow-up geophysical grids. Based on previous experience elsewhere in France (e.g. Haselgrove 1996; Haselgrove and Lowther 2005; Mills 1985), we would reasonably expect to find Iron Age fabrics, even hand-made ones, on ploughed fields brought to the surface by ploughing, together with harder fired Gallo-Roman and early medieval fabrics. All non-modern pottery was collected and recorded, along with worked flints. Concentrations of tile were noted but, in general, tile was not systematically collected, apart from any obvious pieces of Roman tegula or imbrex. Any tile that was collected was sorted and diagnostic Roman pieces recorded, a sample of tegula and imbrex being retained for reference.
In practice, three significant problems were encountered. First, those fields used for fodder crops tend to be only relatively lightly ploughed most years, and are only rarely deep ploughed. Consequently there is little 'replenishment' of archaeological material in the ploughsoil much of the time, which reduces the chances of recovering more friable ceramics that degrade first. Secondly, the majority of plots are not left to weather for any length of time - far less than is needed for archaeological material to reach optimal visibility on the surface - being turned after harvesting, and then ploughed and resown in one operation not long afterwards. This is true both of the barley fields harvested in the early summer and the maize fields cut in the autumn. None of the fields in the chosen study areas was left to weather over the winter, nor was this observed anywhere else in the immediate region. Thirdly, in the Arroux valley, these difficulties are further compounded by regular use of fertilisers, which clearly accelerate the degradation of even well-made Roman wares, many of which had lost their original surfaces and were barely recognisable. Given the very poor surface conditions for recovery, a number of the more promising fields were walked on more than one occasion.
Over the period of the survey, a small number of fields changed their use from pasture to arable or vice-versa, but none in the sectors being particularly targeted by the survey. Fieldwalking was undertaken in both August and October, following the harvesting of the barley and maize crops respectively. To help counteract the problems we have outlined above, walkers were spaced at intervals of 10m, rather than 20m or more, which has proved adequate on intensively cultivated terrain elsewhere (Haselgrove 1996; Mills 1985).
Some relevant reference collections of ceramic material already existed in the region, notably for the excavations at Mont Beuvray, covering the period between the late 2nd century BC to the end of the 1st century BC, and from Autun for the Gallo-Roman period. However, due to the poor condition of the fieldwalking pottery and the different purposes for which the Mont Beuvray type series had been set up and its limited date range, we decided to construct an independent fabric series for the project, which has then been cross-referenced where possible to other reference material (Section 6.1.2). Inevitably, the variable size of the assemblages recovered and the fact that some fields were walked more than once means that the data are not strictly comparable, limiting the possibilities of inter- and intra-site quantification and analysis.
© Internet Archaeology/Author(s)
URL: http://intarch.ac.uk/journal/issue25/3/4.2.html
Last updated: Wed Oct 29 2008