Pushing the boundaries of data? Issues in the construction of rich visual past landscapes

Pictorial, and increasingly photo-realistic, visualisation of past landscapes is underpinned by reconstructions of the past environment themselves effected through palaeoecological methods. The role played by such visualisations has traditionally been that of final-statement, or end-product, intended to convey a synthesis of the interpretation for general consumption, geared around the deceptively straightforward question ‘what did the landscape look like in the past?’. This paper seeks to provide a critical appraisal of the production of such visualisations, through exploration of the nature, spatial scale and interpretative framework of the palaeoecological datasets that underpin them. It is argued that a more sophisticated approach to the spatial interpretation of data is needed, through different approaches to spatial sampling, to facilitate what might be considered complete, and more importantly plausible, visualisations of the past. These visualisations should be considered partial at best, and certainly not full representations of the experiences of communities and individuals that engaged with them. Finally, it is argued that the role of visualisation of past landscapes should move beyond the presentation of interpretations as an end-product of research. Rather, they should form part of the interpretative framework of research that by its very nature needs to be interdisciplinary from start to finish.