Re-thinking landscape archaeology and GIS analyses: a different way of dealing with archaeological landscapes within GIS

Our paper will start from exploring briefly the situation of GIS applications in the archaeological research fieldwork at the beginning of our century, looking at the capabilities and limits until now reached and discovered and will continue by proposing a new theoretical and practical Object-Oriented relational approach for the study of complex archaeological landscapes. Most of archaeologists have been captured finally by new computing technologies believing stili now that sophistication of powerful and expensive GIS software will be enough for high quality outputs and high levels of interpretation. For us, GIS is a set of techniques that at this stage help archaeologists to visualize and to manage huge amounts of georeferenced data and to execute some basic spatial analyses. Spatial Analysis offers several tools to allow archaeologists to move to more complex explanations. Therefore, the purpose of our paper will be to show how with well defined archaeological problems and starting from a well based theory, we can integrate some already existing tools in a GIS framework, moving in such way from beautiful images to complex analyses. Nowadays, most of GIS based archaeological projects are simple databases with a discrete representation of archaeological data in a 2D static space, with functionalities limited to primitive geometric operations used for the calculation of simple and basic relationships or for execute queries and summary descriptions between points (sites) or lines (ancient roads, streams, etc.) or areas (artifact concentrations) in a space. The result is that we have GIS used for the inputs of a huge quantities of data indiscriminately over a map, producing as final results a lot of maps but a lack of theories or hypotheses about the kind of problems archaeologists need to solve and about the relationships between spatial data. On the other hand, archaeologists are working today almost only with environmental variables (topography, lithology, hydrology, etc.) of an area forgetting the importance of social relationships and their interactions in the analyses. Our paper will finally focus on the hypothesis to introduce some elements able to develop a "theory of spatial relationships" needed to study human activities and social spaces. So, we propose a multidimensional and an ObjectOriented Relational approach in order to define and integrate in a GIS framework "activity areas" starting at a first level from the basic features found by archaeologists during the survey (some post-holes, a grave, a hearth, an artifact concentration, etc), anyway considered as "activity features ", and working on the relationships between them, in order to define other more complex levels of analyses. At the end, the purpose of our work is to demonstrate how it is possible to build a pattern of social interactions between different "activity features" (units), starting from a well defined archaeological theory, creating an Object-Oriented model, and integrating some already existing analytical tools in a GIS software (geo-statistic, intra-site spatial testing, visibility, etc.) in order, in such a way, to better define and clarify an historical reading of the archaeological landscape.