Discussion

Those natural features which are conducive to movement must be exploited and established in the mental map of individuals before they can truly be called routeways. While ArcGIS can effectively show us the end product and a potential routeway, NetLogo can be used to explain how routeways are established through a series of discrete actions around those natural features, acted out by individual agents over time.

In real world problem-solving, this would appear as a series of trails which impress upon any subsequent traveler. If an opportunity arises to smooth the course of a trail through cutting corners it is usually done. This can be seen in managed parklands or city green areas, where the paths laid down by the authorities are routinely ignored by walkers as they cut corners and create their own network of desire lines. The most effective path becomes the most established one both mentally, through repeated experiences, and physically, through repeated use. This can be further promoted through construction, providing a visually obvious option and usually requiring less energy expenditure. Thus, if a togher has been con­structed across a bog, travelers are unlikely to deviate from that crossing point by walking over more unpredictable bog surfaces. In this way, a stable macrostructure emerges from the interaction of local agents and feedback from microstructures (Epstein and Axtell 1996).

The process of testing each model in both directions did not yield any major differences in path. The low gradient of the landscape made novel paths unneces­sary, but where they do exist, they tend to be along slopes.

By combining Least Cost Paths with Agent-Based Modelling and conventional documentary research, it can be shown that Least Cost Paths have the capacity to emulate real human decisions as long as all the factors are considered by the program-user when producing the model. If the Least Cost Paths in these examples had been produced with only slope in mind, for example, the low slope values of the bogs would have been used to create wholly unrealistic paths which traversed the bogs in inefficient and dangerous ways. The consideration of the soil factor creates a model which is more realistic and closer to the real-world decisions which agents in the landscape are obliged to make, and while the values assigned were through subjective fieldwork, it is through the consideration of personal experience that modelling becomes its most effective, imitating real human behaviour rather that the ideal path. The Agent-Based Model is explicitly concerned with the decisions made by individuals and it produces very similar paths to those predicted by Least Cost Paths. Moreover, the documentary evidence for the Togher of Croghan, distribution of archaeological sites along both models and the close relationship between the model paths and the existing road network is persuasive evidence for the potential of both methods in predicting paths.

The models used in this research have demonstrated how attention to movement is essential in understanding the landscape of the past. The relationship between the suggested routeways and the archaeological evidence allow an enriched interpre­tation of those sites as part of an inter-connected landscape, as opposed to isolated areas of activity. Meanwhile, the difference in location between the Togher of Croghan and the trackways associated with Lough Nashade, for instance, demon­strate the different characteristics we should expect from routeways which are intended for different purposes. Where trackways deviate from the characteristics discussed in this paper, such as where the path bisects Croghan Demesne, we have an opportunity to ask questions relating to potential ritual, economic or social reasons for this deviation. In short, a movement-centric framework allows us to contextualise archaeological remains and understand landscape.

6.7