Monument sites dominate mainland Scottish Neolithic research because so few
settlements are known, less than sixty to date. Mostly found by chance, these sites are
highly fragmented, difficult to recognise and resistant to interpretation. Important
insights into monumentality have recently been achieved using geographic information
system (GIS) based landscape studies, but these approaches have not been applied to
the settlement record of the Scottish mainland.
Accordingly, this research proposes a new landscape methodology, integrating both
settlement and 'ritual' evidence to create the first GIS-based predictive models that
identify those locations most likely to have sustained Neolithic setdement over the
whole of mainland Scotland. Known settlements and monuments provide social and
environmental information, which is quantified and contrasted with a control sample.
Multivariate statistical analyses establish the variables that may have guided the
placement of known sites, forming the backbone of the predictive models. Test
samples kept separate from each model's construction were created from two case study
regions — southwest Argyll and the Moray Firthlands. Moray has yielded many known
settlements, which are wholly unknown in Argyll. This data was used to verify each
This research holds great promise for exploiting predictive modelling as a powerful tool
in Scottish archaeological research, which should be optimised for regions where limited
or no proof of archaeological remains is known.