9

Data Interpretation: Methods and Tools

Hypothesizing Southampton in 1454: A Three-dimensional Model of the Medieval Town

by Matt Jones

ABSTRACT
Three-dimensional modelling has an important role to play in the archaeological process. Computer Applications and Quantitative Methods in Archaeology (CAA) conferences and British Archaeological Reports (BAR) prove that a section of the archaeological community is validating the use of such models based on archaeological data. Barcélo claims that models should be embraced as they offer the means to represent features of concrete or abstract entities as opposed to mere illustrations that have been previously relied upon.¹ Indeed 3D modelling offers a level of control and an ability to reconstruct past, built environments in an entirely new fashion. Inexpensive fully 3D environments can be built based upon factual data. However, their reception by academics is often frosty. One set of scholars claims ‘the appearance of an image or animation in a digital format out of context is nearly always misleading’.²

A lack of adequate documentation, auxiliary data and intellectual transparency lie at the root of this distrust of computer models. It is indeed misleading to present a model without providing the audience with the tools fully to understand the model and its limitations. This paper looks into the process of creating a 3D model of medieval Southampton based on disparate data sources, such as historical documents, map data and excavation data. It includes a discussion of the importance of being intellectually transparent, honest and informative at all stages of the model’s creation. When working with different data sources and, as is often the case with archaeology, an incomplete dataset, it is constantly necessary to make decisions. These decisions have a direct impact on the resulting hypothesis; in this case a computer-generated model of a medieval town of which only a fragment has survived. The methods employed in an attempt to add credibility to the model are presented and evaluated. These include how missing data and contrasting data were dealt with, recorded and presented alongside the actual model. It is important to make such intellectual transparency the rule rather than the exception to allow 3D digital modelling techniques to be fully embraced by the archaeological discipline.

Notes
1. Juan A. Barcélo, ‘Introduction’, in Juan A. Barceló, Maurizio Forte and Donald H. Sanders (eds), Virtual Reality in Archaeology, BAR International Series 843, (Oxford: Archaeopress, 2000): 9.
2. Geeske Bakker, Frans Meulenberg and Jan de Rode, ‘Truth and credibility as a double ambition: reconstruction of the built past, experiences and dilemmas’, Journal of Visualization and Computer Animation, 14/3 (2003): 159.

Plate 9.1 Left: plan of medieval Southampton based on the Southampton Terrier of 1454 (Burgess, 1976). Right: Plan derived from combined sources: 1454 Terrier (green), archaeology data (red); cartography data (blue) and inferred (magenta) © Matt Jones, 2008.

Plate 9.2 Medieval Southampton. 3D model © Matt Jones, 2008.