Viewshed analysis is starting to become so routine that Aldenderfer has suggested it is now a viable alternative to thinking about the past (Aldenderfer 1996). It aims to explore substantively the relationships of visibility and intervisibility between particular archaeological locations in the landscape. Contrary to popular belief, this kind of analysis pre-dates GIS by at least two decades, with both formal studies such as Renfrew and Fraser's work on Orkney (Fraser 1983; Renfrew 1979) and anecdotal discussions about visual impressiveness or placement of monuments. The widespread availability of digital tools for modelling and quantifying visibility provided significant impetus to this research area providing, for the first time, the ability to calculate indices such as area-of-view and angle-of-view rapidly for many sites. As such, 'viewshed analysis' need not fear the accusation, often levelled at it at conferences and implied in Aldenderfer's remark, that it is a form of technological determinism with no archaeological research agenda.
The formal analysis of visibility presents as many — if not more — methodological obstacles as predictive modelling. These have been discussed elsewhere (Gillings and Wheatley 2001; Wheatley and Gillings 2000) but a non-exhaustive list of significant methodological problems would include issues of reciprocity (Loots 1997), the quality of DEM data (and its influence on visibility), the effects of vegetation, the modelling of probable and fuzzy viewsheds (Fisher 1991; 1992; 1994; 1995; 1996) and the availability of suitable hardware and software to calculate the desired indices of visibility or 'inherent visibility' (Lake et al. 1998).
The key reason why visibility analysis can be argued to be a significant positive contribution to archaeology is that it is a form of analysis that begins from a model of the field of vision of human actors within a landscape. From this human-scale component, generalisations and quantifications are built and methodologies developed. Of course, computer models are not perfect representations of either human perception or the landscape, and visibility is not the only sense implicated in the human experience: sound, smell, feeling or even taste might also be amenable to substantive analysis (Gillings and Goodrick 1996), although only sound has so far been significantly investigated (Watson 1999). But pointing out that visibility analysis is not complete is not the same as arguing that it is inherently defective. Viewshed analyses begin from the human-scale experience of existing in the physical and social world. As such, they contribute to the formulation of substantive approaches to issues as diverse as cognitive perception, culture/nature dichotomy, visualism and sensory primacy, temporality and directionality. This is why viewshed analysis has so effectively resonated with many of the criticisms of unreformed functional-processual spatial archaeology, while at the same time offering a real, substantive alternative to relativist positions such as phenomenology.
There has undoubtedly been work that seems to be aimlessly calculating viewsheds for no obvious reason (other than that there is a button available to do GIS-based visibility analysis). But this area of GIS applications has also produced interesting findings in understanding the patterns of relationships between Neolithic and Bronze Age monuments in Wessex (Wheatley 1995; Wheatley 1996b) and in Scotland (Fisher et al. 1997), the locations of petroglyphs (Bradley 1991), symbolic astronomy (Ruggles and Medyckyj-Scott 1996; Ruggles et al. 1993) and Hellenistic defence systems (Loots et al. 1999).
Visibility analysis, and similar human-centred forms of analysis, need to mount a more vigorous defence of their rationale than has been available to date. Not only can visibility analysis claim to be significantly rooted in recent intellectual debates within archaeology and related disciplines, but it can also point to methodological developments and archaeological results that set it apart as the most promising current area of application for GIS technologies.
© Internet Archaeology
Last updated: Wed 28 Jan 2004