GIS and Second Palace Farmsteads
(collaborating with other KIP specialists)
Perhaps one of the richest datasets to emerge from the KIP intensive survey is the settlement landscape of the Second Palace Period (Minoan Neopalatial). GIS and spatial analysis techniques have been used to explore the way this landscape was structured, in terms of social organisation, agriculture and island demography.
We can document both a large complex community at Kastri and a rural countryside dotted with small surface scatters (ca.0.1-0.3 ha in size). For a variety of reasons, these rural scatters appear to be the permanent or semi-permanent dwellings of 1-2 families engaged in agricultural subsistence (hence 'farmsteads'). A variety of excavated buildings on Crete and Karpathos offer useful contemporary parallels for the kinds of structures involved.
GIS allows us to explore, refine and quantify intuitive or pragmatic insights about these farmsteads made by a variety of KIP researchers (fieldwalkers, ceramic analysts, geoarchaeologists) as well as to produce unexpected new perspectives. Site spacing, the distribution of 'off-site' ceramics and the location of tombs can be quantified statistically and each of these three factors suggests a threshold of Second Palace Period rural activity ca.150-200m away from a given site centre.
We can probably characterise this as a 'site catchment' within which a range of household-scale subsistence activities might occur. Geoarchaeological assessement of the shallow channels of Quaternary alluvium near these Second Palace sites suggests that they may have been terraced (cross-channel). This strategy would provide an ideal locale for harvesting rainwater run-off and practicing intensive mixed (pulse/cereal) farming. The latter agricultural regime is also consistent with our limited bioarchaeological evidence from the MB-LB Aegean. GIS allows us to explore the amount of Quaternary channel available within each site catchment - most sites have access to 1-3 ha of channel (depending on how this is measured) which is probably enough to support a nuclear family who adopted intensive farming methods.
We can also consider the question of where sites were located. GIS-led correlation of sites with specific environmental variables such as soils, water, slope and aspect is a well-established (if imperfect) technique, but we can also explore more socially-constructed priorities. A viewshed map of places that are intervisible with the contemporary peak sanctuary of Agios Geogios can be compared to the distribution of rural sites. While there is good reason for thinking that a visual link between the main site of Kastri and the peak sanctuary was an important component of the ritual landscape, there is no similarly significant correlation between rural site location and a view of the peak sanctuary.
GIS also allows us to develop more sensitive measures of topography (terrain roughness) and surface hydrology (how water behaves on the landscape). For example, we can map (at a specified scale) the shape and extent of the basins (watersheds) that contribute to each section of the drainage network. Second Palace sites appear to be located close to watershed boundaries, probably because these areas provided shallower channels for agriculture and often manifest themselves as small ridge-lines, in or up against which it was possible to build houses and tombs.
Further GIS analysis suggests probable population levels for the island in the Second Palace period and likely ways in which the rural sites interacted with each other (see Bevan 2002 for details).