One of the dangers in using new technologies is that they are exploited because they are new: innovative as a technology rather than useful as a generator of data or as a system of virtual conservation. We need to present digital models in terms of what they are able to do, not as what we expect or desire them to do (Penny 1993, 18). As Gillings suggests, we must engage with the potential of virtual models as '…a set of heuristic devices for the examination and exploration of the sensuous and physical three-dimensional world in a flexible and fundamentally three-dimensional fashion' (2002, 20). Despite the tenuous relationship with the reality of the scanned object or site, these simulated models can facilitate a detailed and useful engagement with the subject, forms of investigation that can be far more productive than those feasible in the field. This is especially true of a subject situated in an 'awkward' position, such as rock art on the ceiling of a rock overhang on a site at 2,130m in the Alps. In fact, the VR modelling of scenarios that are difficult for most people to access is possibly one of the core justifications for this approach (Fernie et al. 2003, section 2.3). Also, precise digital recording does obviate the need for the measured drawings or tracings that are often undertaken on rock art sites (Domingo et al. 2013; Cassen et al. 2014; Cassen and Robin 2010). At a general level, virtual reality models can be considered a mechanism for the transfer of knowledge, or the conversion of information about a landscape into potential knowledge that facilitates understanding of that landscape (Fernie et al. 2003, section 2.3, Witmer et al. 1996).
If we accept that the context of interaction with, and interpretation of, our rock art changes depending on whether we are in the field or assessing our scanned imagery in the office, does this mean that the hermeneutic context changes (Gadamer 2006: 390; Abadía and González-Morales 2012, 267)? In many ways this is assumed, as the process of visualisation is entirely different. Here, we take the position that both field-based and 'desk-top, digital-based' visualisation are necessary. While hermeneutics is traditionally concerned with broader issues of socialisation and the influence of cultural context on the interpreter's analysis, it is also true that micro-scale changes in context (the field during the August, and the office in December) are also important. Whereas pre-digital technologies (photographs, drawings, casts etc.) facilitated desktop assessment, it is only with the advent of pseudo-3D digital imagery that we can go some way in the re-creation of the subject/object of our interpretation.
For a virtual 3D model to have meaning today, or for prehistoric art to have had meaning in the past, tacit understandings of how we should engage with these representations are essential. While many people will have some notion of how to engage with the virtual model today, there will be variation in how we use and understand the model. The same would have been true of our prehistoric paintings in the past. Both media act as 'strings' (any phenomenon that is inscribed with patterns, whether this be air with sound-waves, or a solid medium with markings (Collins 2010, position, 173)), in the sense that meaning can be revealed via tacit knowledge rather than explicit, detailed instructions. Tacit knowledge becomes explicit within well-defined contexts that include contingent information that enhances comprehension. As archaeologists, we cannot reconstruct these contexts in their entirety. All explicit knowledge rests on tacit knowledge – whereas many people have a reasonable notion of what a traffic warning sign is signalling, the information only becomes explicit if we have read the Highway Code. Consequently, if we are to render virtual models and the prehistoric motifs comprehensible, and suggest explicit meaning or purpose, then we must develop a deeply embedded argument, inferences founded on the definition and description of a range of contexts that include the study of landscape, contemporary sites, and similar artistic representations from different chronological periods (Gill 1974; Wittgenstein et al. 2010). In this instance, the context for the use of the virtual models is defined by instructions or guidelines for intended use, along with the prior experiences that each reader has using computers and similar software. In the prehistoric past, explicit understanding or appreciation of the paintings, as well as the rationale for their production, was also contingent upon socio-economic, cultural and landscape contexts.
As authors, we can give advice on how to use the virtual models associated with this article, and most readers will (hopefully) explicitly realise their intended use; however, our interpretation of the rock art is implied by a series of advocated historical and social contingencies; therefore, comprehension is embedded. Although notions of tacit knowledge are often linked with phenomenological schools of thought, we must be clear that our understanding of tacit knowledge is not centred on the emphasis of sensory perceptions of individuals per se, but rather that which considers shared structures of knowledge and understanding. Knowledge is founded upon prior assumptions, and therefore 'tautological'; in some ways a socially acquired capacity, propensity or tendency, where an agent acts appropriately in given circumstances (Gerrans 2005, 54).
As discussed later, similar sets of 'artistic' motifs were repeated across time and space, often deployed in very different landscape types. Such repetition might be considered 'rule-following', while the meanings associated with the motifs were largely contingent upon their landscape and chronological contexts. To some extent, notions of tacit knowledge intersect with certain forms of phenomenological thinking and cognitive processual approaches. Many forms of visual culture follow accepted rules vis à vis the type and composition of motifs, and possibly had a role in the 'storage, retrieval and transmission of social knowledge and of practical information about the natural world and its exploitation' (Skeates 2005, 3), but these are then experienced and reinterpreted by other spectators. In many ways our approach fits with Skeates' characterisation of 'interpretive archaeology' (2005, 4-6).
In summary, we feel that in order to engage with and understand any representation, a form of tacit knowledge is required – an almost intuitive sense that provides the context for understanding any phenomenon, which goes some way to rendering it explicit: Landscape situation is one form of context. Context here is taken to comprise the 'natural environment' as well as the human activities that took place in the landscape. The laser-scan goes some way to facilitate the description and representation of these contexts. Although we accept that virtual models constitute artificial worlds, we feel that pseudo-3D models do facilitate and enhance our ability to engage with the detail of the art and its landscape context. An analysis of context should contribute to the interpretation of rock art; therefore the definition and representation of the topographic context at the immediate, local scale (via the Faro laser scan) was considered important for investigation of the site and its art (Figure 8).