Towards a virtual archaeology

There is a long tradition of recording archaeological contexts. Early accounts are typically a bald statement of the kind: 'a Roman floor was found.' Later, longer textual descriptions appear. The correctness and truth of the observation or interpretation was confirmed by the personal standing of the reporter (cf. Hodder 1989). The more cautious added weight to their view by including another equally good sort as a witness. So for example Sir Richard X would describe how he uncovered the remains of Pongo man in the presence of the Reverend Y. Gradually, the introduction of greater amounts of detail and illustrations enable comparisons to be drawn. In effect, they came to be regarded as much more objective descriptions. Photographs were then introduced as proof. Unfortunately, photographs are not always adequate to show subtle differences between, for example, light brown compact sand and loose light brown soil. Nevertheless, photographs have come to be regarded as important supporting evidence to interpretations. These methods, it must be realised, were constrained by the limitations of the available technology. That is a paper interface. The problem confronted is that of projecting aspects of a three-dimensional space on to a two-dimensional plane. This limits the effectiveness of these tools. Nevertheless, they are founded on a long tradition of convention and are useful records. Of course, scale drawings and black-andwhite photographs also have the major attraction of being comparatively cheap to mass reproduce. It is not surprising that the first computerised systems for handling and recording archaeological contexts have inherited many of the characteristics of the traditional paper interfaces (e.g. Alvey 1989. Rains 1989, StanCiC 1989, Weiss 1989, pp. 314-317). However, while the excavation plan has the merit of having a direct bearing on some naturally occurring stratigraphie interface, either the top or the bottom of some context or other, the purpose of the sectional profile is much more difficult to understand. Decisions about where to place sections are arbitrary in relation to the archaeological context. Although profile-drawings are sometimes useful for delineating the excavator's interpretation of where one context ends and another begins, if the feature is not symmetrical then a section must miss details. It is a biased and partial record, which is potentially misleading. In my experience, non-archaeologists find sectiondrawings very difficult to comprehend, often prompting the question: "Why do you do it?" 21.2 Virtual archaeology