This project started in 1994 as a regional archaeological investigation into the Roman town of Viroconium Cornoviorum. The University of Birmingham and the Leverhulme Trust jointly funded the project, with more recent funding from English Heritage. The University of Birmingham Field Archaeology Unit (BUFAU) co-ordinated the investigation with the aid of other organisations who provided the geo-physical, and other data (7). The primary aim was to study the relationship between the urban Roman town and its rural hinterland, specifically focusing on the process of Romanisation. Thus, since a feature of Romanisation would have involved spatial developments of some kind, GIS might contribute to the study.

Building on information that had already been gleaned from the area by previous strategies (8), the goal of the project was to 'study the process of 'Romanisation' through which native Britons became socially, culturally and economically integrated (or not) into the Roman Empire' (9). The literature of the project asserts that the foundation and encouragement of towns was the principal mechanism by which this could be achieved (10). By combining previous data with their own research the project was able to bring together several disciplines in order to approach their investigation in a more holistic manner.

Of course, before GIS can be of any benefit, data must be entered into the system - all kinds of survey, or any other information that might contribute to the study, must be digitised. It was recognised early in the project that the best means available to assess the amount of data was by using a Geographical Information System. This recognition stemmed from the acknowledgement that GIS can 'cope with a wide range of data provided it has a grid reference, including both point data, such as the find-spots of artefacts, and area data such as soil maps, solid and drift geology' (11).

Before continuing with an explanation of the processes that the project used to integrate their data into GIS, the computer system itself needs to be briefly discussed. Although this essay is not a technical study it is important that the rudiments of the system are understood. A Geographical Information System is a computer software package that allows several sources of information to be stored in its database. The database is retained within the 'brains' of the system and can be retrieved to the screen by asking explicit questions.

For example, if the database included the results of say three surveys - 1) the position of Roman forts in a given area; 2) the position of coins found on the surface within the same area; 3) the position of scattered potsherds found within the same area - then the database can be asked to produce in the display the relation of coins to the forts, and so on. This display is known in the jargon as a vector model. All this means is that each piece of data (a fort or a coin) is referenced to the x and y co-ordinates of the screen. An example can be seen in Fig.1.

Fig 1. This example shows the results that would appear on the screen if the operator asked the GIS to display information found in a given area. The red dots are representations of coins; the black dots are metal artefacts. The blue lines are an interpretation of an aerial photograph, and the background is the result of a geo-physical survey represented as a raster model. The figure was created by Peter Halls using data from the Cottam Project, which was directed by Julian Richards. Copyright Archaeological Data Service.

The only other display that a GIS can produce is called a raster image. In Fig.2 the black and white image is a raster model. Simply put, a raster model is one where the screen is divided into a matrix of cells allowing for any information to be allocated to its own cell. If an image created using geo-physical survey equipment is produced (provided the scale of the area is retained), all information that was collected from an area of the earth's surface, or sub-surface is accurately represented on a computer screen. A further example of a raster image is seen in Fig.2.

This image of an area at Wroxeter is a graphical representation of a fluxgate radiometer survey conducted by a team from the AML Archaeometry branch for the Historic Buildings and Monuments Commission for England. The white areas correspond to highly magnetic anomalies under the soil. The street system of the Roman city is clearly visible. Copyright © Historic Buildings & Monuments Commission for England. Image taken from: http://www.eng-h.gov.uk/Wroxeter/

Although both models show different characteristics, in each case the scale of the area is needed to give an accurate representation of the physical data. Using GIS the models can be laid over each other and integrated in any combination to produce the results of an enquiry. This system of overlaying information is, of course, not new. Transparent sheets will produce similar results. But the effectiveness of GIS is such that the information can be retrieved from remote computers, provided they have compatible software packages.

As well as the potential of easier access to information, using GIS the flexibility of the system is such that hypotheses can be tested against the data more quickly than by using transparent sheets. However, the speed of computation depends upon the available hardware. More importantly for archaeological interpretation the database of any system can be added to at any time. This allows for a continuous updating of information. This means that all of the information from a study is available from one computer terminal, thus saving valuable research time.

As has already been made explicit, all types of data to be included in GIS must be co-ordinated to the same scale. In order to reject as many uncertainties as possible, an aim of the Wroxeter Hinterland Project was to conduct a more systematic collection of data using both modern and traditional methods.

This would generate enough data to allow the site and its surroundings to be studied in a more holistic manner. It also provides an opportunity to study the results of traditional and modern methodologies.

7. English Heritage's Ancient Monuments Laboratory and Geophysical Surveys of Bradford conducted magnetometer surveys. The Centre de Recherches Geophysiques at Garchy (France) took resistivity measurements, along with other volunteers. Aerial surveys were done by the Natural Environment Research Council (NERC), who collected photographs, Airborne Thematic Mapper (ATM) and hyperspectral data. And the Nara Research Centre and the University of Miami (Japan) did Ground Penetrating Radar surveys. For more information click here.
8. Specifically, the excavations of Atkinson (1923-27);Barker (1966-1974);aerial photography of Baker (1967-8); and the post-Roman analyses of Bassett (1990, 1992).
9. Quote taken from the project's web page.
10. See (http://www.bufau.bham.ac.uk/newsite/Projects/Wh/Lever/index.html)
11. White, R. Current Archaeology 157 (1998) p.10.