Modelling Prehistoric Mining Kerstin Kowarik*. Hans Reschreiter* Gabriel P. X. Wurzer** *Natural History Museum Vienna, 1010 Vienna, Austria (e-mail: kerstin.kowarik|hans.reschreiter@nhm-wien.ac.at). **Vienna University of Technology, 1040 Vienna, Austria (e-mail: gabriel.wurzer@tuwien.ac.at) Abstract: Mining structures are among the most complex economic systems in prehistory. Until recently, research into prehistoric production processes has strongly focused on technological reconstruction. The complexity of production processes, their interconnectedness with the surrounding socioeconomic network and issues of quantification have, quite regrettably, been addressed to a much lesser extent. Simulation can contribute important insights into the latter problem areas, however, the lack of consistent methodological discussion on data collection, model building and comparability of results still represents a significant gap in research, which we address in our work. Keywords: Archaeology, Prehistoric Mining, Agent-Based Simulation, System Dynamics, Process Simulation. 1. INTRODUCTION Mining areas are not only characterized as centres of production, but also as centres of consumption - yielding high demands with respect to workforce, means of production (mining tools, raw materials) and means of consumption (e.g. food, clothing). The necessity of expert knowledge, intra- and superregional communication, traffic and trade networks further adds to the complexity. All these interdependent conditions demand an analytic approach combining different levels of observation, both spatially and in context of the model used. We argue that a consistent and methodical analysis of prehistoric mining structures can best be done by the combination of several different simulation techniques in connection with a multidisciplinary database drawing on experimental archaeology, ethnography and historical records. This combinational method, which we will demonstrate by using data gained through the modelling of the Bronze Age salt mining complex of Hallstatt/Austria (1458-1245 BC), lies at the core of our efforts. In more detail, our contribution is broken down into the following parts: First, we give a background over previous uses of simulation in the field of archaeological research, listing formalization efforts targeted at ensuring comparability of the models used (see “Related work”, Section 2). We further deal in some length with experimental archaeology and ethnographic analogy, which are methods of data collection specifically suited for dealing with data gaps, which are omnipresent in the archaeological record (see “Data Collection”, Section 3). We then proceed to the archaeological basis concerning prehistoric mining in Hallstatt, looking specifically at input data and research questions that are to be answered by applying simulation (see “Archaeological background”, Section 4). We then report on the types of simulation models that have been used so far (see “Simulating prehistoric mining”, Section 5). 2. RELATED WORK The idea of using computer simulations in archaeological research has been around for nearly half a century. The 1970s saw considerable enthusiasm which was then thwarted by the limitations of contemporaneous computer technology and the lack of a sufficiently sophisticated theoretical framework. The developments in computer technology and scientific theory (complex systems theory) in the 1990s have given the application of computer-based modelling to archaeological research a considerable new boost (Kohler and van der Leeuw 2007, pp.1-12; Costopoulos, Lake and Gupta 2010). Especially Agent-based Modelling (ABM) has been popular with the scientific community since the late 1990s. It has been applied to a multitude of research topics, from the development of social complexity, decision-making, culture change, and spatial processes (Doran et al. 1994; Dean at al. 2000; Bentley and Maschner 2003; Beekman and Baden 2005; Premo 2006; Clark and Hagmeister 2006) to the exploration of civil violence in the Roman World (Graham 2009) and the work flow analysis in prehistoric mines (Kowarik et al.