Regional survey and GIS are seemingly natural bedfellows (see Gillings 2001). The distribution map has long been the standard means of presenting the results of fieldwalking; from here, it has been an easy step to the enhanced flexibility and analytical power of GIS. The bulk of all early archaeological applications of GIS concerned the analysis of regional settlement patterns (e.g. Allen et al. 1990; Lock and Stančič 1995). The adoption of GIS in the context of Mediterranean regional survey has been particularly rapid. This is well-illustrated by the complete absence of any mention of GIS at the 1988 'Roman Landscapes' conference (published as Barker and Lloyd 1991), soon followed by the dedication of a whole POPULUS colloquium to the topic in 1995 (published as Gillings et al. 1999).
Today, nearly all regional surveys routinely use GIS for the collection, integration, interrogation and display of results. Increasingly, survey data are digitally captured in the field with Global Positioning Systems and Personal Digital Assistants and are directly integrated with cartographical and existing archaeological data for instant analysis by mobile GIS during the course of fieldwork. However, from the beginning, legacy survey data have played a central role in GIS applications and they continue to do so today. Initially, desk-based applications were concerned with simple overlay functions (e.g. the distribution of sites in relation to soils); however, analysis has become increasingly sophisticated. For example, modelling studies and visibility analysis have been used to explore more contemporary theoretical concerns such as perception and embodiment.
A few examples of the use of legacy data within GIS stress the breadth of approach. The importance of movement for the social construction of space is explored through analysis of roads and paths (e.g. Bell et al. 2002) and the comparison of cost surfaces and viewsheds has been used to emphasise the contrasting physical and visual accessibility of rock-cut tombs (Belcher et al. 1999). Other GIS applications have considered 'ideational' landscapes. For example, van Hove (2004) uses GIS to consider landscape perception via Ingold's (2000) concept of 'taskscape'. Meanwhile, more 'traditional' themes such as agriculture and demography are approached in more nuanced ways. For example, Goodchild (2005) uses multi-criteria models to assess the Roman agricultural treatises; variables such as sowing ratios, labour input and yields are combined in order to assess key issues of food supply, demography and town-hinterland relations.
Finally, GIS plays a central role in the visualisation and presentation of results and data dissemination. Many regional survey projects now publish via the Internet (e.g. Davis et al. 2005) and/or deposit data with national digital archives (e.g. Given et al. 2007). The ability to share and query spatial datasets makes projects' results dynamic - surveyors' statements can be re-evaluated and new questions investigated. Online mapping tools such as GoogleEarth have made the manipulation of datasets even easier.
The importance of GIS for both legacy data and new regional surveys is unquestionable. As technologies and data become more freely available and easy to integrate, GIS will inevitably play an even greater role in the future. However, the vast majority of legacy survey data exist as paper catalogues. To realise the latent potential of these datasets and to subject them to new and innovative analysis requires digitisation. Unlike new survey projects where GIS design has become integral to the recording and analysis of data, the digitisation of legacy data requires a more flexible approach to accommodate problems such as inconsistent terminology or the variable accuracy and/or precision of geographical coordinates. Appropriate mechanisms to accommodate these problems must be found. Usually these solutions are a compromise between preserving the complexity of the original data and introducing sufficient order so that the data can be efficiently manipulated. Certainly the investment of time and effort necessary in order to digitise legacy data should not be underestimated; indeed, the preparation of data comprises a significant percentage of the duration of any GIS project, leading some to question the value of GIS all together (e.g. Sharon et al. 2004). However, as will be argued below, this process of data preparation needs to be reconceived.
© Internet Archaeology/Author(s)
URL: http://intarch.ac.uk/journal/issue24/2/2.2.html
Last updated: Mon Jun 30 2008