Starting out on the Journey to Manzikert: Agent-based modelling and Mediaeval warfare logistics
Philip Murgatroyd (Birmingham)
Digital Classicist/ICS Work in Progress Seminar, Summer 2009
Friday 12th June at 16:30, in room STB3/6, Senate House, Malet Street, London WC1E 7HU
The seminar will be followed by wine and refreshments.
The Medieval Warfare on the Grid project uses agent-based modelling to examine the logistical challenges of transporting the Byzantine army of Romanos IV Diogenes from western Anatolia to the site of the Battle of Manzikert in AD1071. This project, funded by a joint JISC/EPSRC/AHRC E-Science Programme examines the practical aspects of transporting tens of thousands of men and animals over 700 miles in the Medieval period as well as the wider impact on the areas they passed through. It uses data from a variety of different areas including transport infrastructure, taxation, agriculture and military organisation.
Although agent-based modelling has been widely used since the 1990s for applications such as analysing traffic congestion or the spread of epidemics, its use in archaeology has been limited. Agent-based modelling involves determining behavioural rules for a set of agents, in this case the Byzantine army and its support staff, that determine how the interact with each other and their environment. With these models we can look into the relationship between elements of the logistical system such as army size, army movement, taxation and rural productivity. Written sources of information on the battle have been exhausted by the historians, this project will introduce new types of evidence into the debate surround the Byzantine army’s preparations.
The relatively simple goals of the army, to reach Eastern Anatolia in a fit state to win a battle, give the model an overall simplicity lacking in more general social situations. Therefore this application is seen as a good starting point for large scale archaeological agent-based models. Archaeological and historical settings pose unique problems for agent-based modelling due to the fragmentary nature of the source data and the inconvenient formats in which it is usually found. This paper examines preliminary issues raised by modelling the practical aspects of land transport of Medieval troops and how the project aims to deal with them.