As this year’s conference season is drawing to its end, I thought I’ll share a few reflections on the very successful SPUHH (Simulating the Past to Understand Human History) meeting in Barcelona. A satellite event to the SSC (European Social Simulation Conference), it was one of the largest gatherings of archaeological modellers I’ve ever witnessed. The organisers filled three days with back-to-back presentations of archaeologically inspired simulation models intertwined with discussions, a great keynote by Tim Kohler and loads of social events (that 9 courses conference dinner will most decidedly go down in history!), and all that against the fantastic backdrop of Barcelona. In a word, it was a blast.
It was also a good round-up of all the, currently, ‘hot’ topics in archaeological simulations. So for all of those who missed it, here’s a short summary divided into the general themes linking the presented case studies. The breakdown may look familiar to some of you as many of the topics repeat from one conference to another. Follow the links if you want to learn more about the case studies presented and you can find the full conference schedule here.
- Dispersals, demic and cultural diffusions
In the classic ‘was the spread demic (people) or cultural (ideas)?’ ABM-genetic model the team from the Okayama University, Japan led by N. Matsumoto and M. Sasakura shared their results on the Jomon-Yayoi transition, while in a similar vein but using classic diffusion equations J. Fort’s team presented their newest take on the Neolithic spread. The latter topic is easily the most popular case study among modellers and a number of other presentations focused on that subject. Pérez-Losada showed a detailed sensitivity analysis highlighting the effects of of different parameter values on the diffusion, while Crema and colleagues used it as testing grounds for evaluating the advantages of the ABC (Approximate Bayesian Computation) for determining the relative probability of tested scenarios. Timm and colleagues gave a presentation about a recently launched project focusing on Pleistocene dispersal. They mostly discussed the challenges the are facing in creating such a complex, multi-scale model given the dearth of available data but it’s definitely worth watching this space as the project unravels. This topic was also pursued by myself but with a special focus on the demographic dynamics of the dispersal. Finally F. Del Castillo and J. A. Barceló’s were on the other end of the demic to cultural diffusion spectrum with their model of cultural standardisation among hunter-gatherers and agrarian societies.
- Land use and landscape mobility coupled with resource acquisition/foraging models
In this category, T. Baum gave a fantastic example of tackling simple research questions while exploring the underlying complexity of the system. His land-use simulation was build to figure out why the famous pile-dwelling settlements around Lake Constance (Germany) were so often moved from one place to another. A number of other case studies (Janssen and Hill, Oestmo et al., Lancelotti et al. and Saqalli et al., the last one being particularly worth mentioning ) were similarly focused on the land use and resource distribution over the landscape. With a more methodological focus O’Brien compared potential trackways through marshy area generated by a GIS-software with those simulated in NetLogo. Finally, Olševičová and A. Danielisová integrated land use models to drive their simulation of the rise and collapse of a Celtic oppidum (you can check out the details of their impressive model combining cellular automata, agent-based modelling and system dynamics here).
- Case studies of historical events
It is easy to notice that the younger the more detailed and less abstract the models become. T. Brughmans and J. Poblome roman trade model and Fulminante and colleagues’ urban dynamics model were presented as simple networks of interaction but younger case study, such as J. Riley Snyder and O. Dilaver’s model of gigantic aqueduct connecting Constantinople with mainland Greece, P. Murgatroyd and V. Gaffney’s simulation of the march of the Byzantine army or K. Comer and K. Comer model of emergent commercial partnerships in renaissance Italy, were developed with a lot of details.
There were also a few models more general and therefore more difficult to classify such as G. Bogle’s combination of agent-based modelling and canonical theory, H. Inoue and C. Chase-Dunn’s model investigating the evolution of global inequality or N. Gotts discussion on the role of communication technology throughout the ages.
In the introduction to the ‘Simulating Change. Archaeology into the 21st century’ Andre Costopoulos and Mark Lake complained about the weakness of the archaeological simulation and the scarcity of computational modelling practitioners. It looks like since their 2004 session at the SAAs a lot has changed, particularly in numbers but also in the breadth of the applications and techniques. The final outcome of the SPUHH conference was a long discussion on the need for an better integration of the field and more communication with the general archaeological audience in order to bring simulation into the archaeological mainstream. So watch this space, there’s some great stuff coming up soon!