Forum ReSoc 21st March 2019

Agent-based Models in Archaeology: Are there Limits?

Featuring Edmund Chattoe-Brown & Marc Vander Linden
Panel Members Iza Romanowska & Michael Roos
Moderation Maja Gori & Frederik Schaff

Abstract Speakers Panel Programme Organisers YouTube ReSoc Main Page

Date 21st March 2019 starting 14.00h, end approx. 17.00h
Location Haus der Archäologien (Hörsaal), Am Bergbaumuserum 31, 44791 Bochum
Registration is free. Please register here to help our planning!
Organised by Maja Gori & Frederik Schaff (Ruhr-Universität Bochum)
Speakers Edmund Chattoe-Brown (University of Leicester) & Marc Vander Linden (University of Cambridge)
Panel Members Michael Roos (University Bochum), Iza Romanowska (Barcelona Supercomputing Center)

Representing archaeology’s full complexity or too many assumptions about things you cannot know?

Agent-based Modelling (ABM), a simulation method that allows modelling the interaction of individual "agents" in explicit space and time, is recently gaining momentum in Archaeology (Rogers & Cegielski 2017, Lake 2014). Proponents of ABM emphasize that it allows representing archaeology’s full complexity (Barceló & Castillo 2016). At the same time, the preconceptions that are often associated with formal modelling are amplified by ABMs (Cegielski & Rogers 2016). One may summarize these preconceptions with the claim that ABMs make too many assumptions about things you cannot know. Should this methodology be rejected, if so many ad-hoc assumptions are necessary? Quite the opposite argues Edmund Chattoe-Brown (2013, §9.1): "ABM gives us a perspective from which to unpick preconceptions (based on existing research methods)". Especially in situations where we do not have enough evidence yet – and probably never will – using ABMs can be paramount as a tool for theory building: “models are not depictions of reality, but formal thought experiments designed to test and explore specific research questions” argue Drost and Vander Linden (2018).

References

The Speakers

"ABM gives us a perspective from which to unpick preconceptions (based on existing research methods)"

Edmund Chattoe-Brown (University of Leicester)

Edmund Chattoe-Brown began training in economics, being interested in decision making, but discovered that the attractive parts of the subject (in retrospect these were the “social” parts, as when oligopolists imitate or collude) were least suited to economic tools. A recreational interest in Artificial Intelligence converged when he realised that simulation could model “properly social” systems. Since then he has tried to apply simulation to as many socially interesting decision structures as possible. These include social networks, innovation diffusion, church membership, oligopoly, social construction of choices, farmer behaviour in response to EU payments and drug use practices. Perhaps paradoxically he is also interested in modelling evolutionary processes where selection of practices by the “social environment” can lead to order without decision. He has been involved with the social simulation community since 1992 and, more recently, his main concern has been integrating simulation better with everyday social science particularly regarding to data.

“Models are not depictions of reality, but formal thought experiments designed to test and explore specific research questions”

Marc Vander Linden (University of Cambridge)

Marc Vander Linden gained his PhD from the Université Libre de Bruxelles, and moved to Britain in 2005 as a post-doctoral researcher in the McDonald Institute for Archaeological Research. He has held positions in the Cambridge Archaeological Unit, the University of Leicester and, most recently University College London where he directed an ERC-funded project on the diffusion of early farming in Europe. He joined the Department of Archaeology of the University of Cambridge in 2017 to coordinate the A2 module as part of the new Tripos in Archaeology.
Marcs research focuses on Late Prehistoric Europe. This research has focussed on spatial variation, landscape and land use, in both North-Western Europe and the Balkans and Neolithic and Bronze Age. He is currently engaged in field projects focussing on the long-term development of pastoralist economy in the Skadar Lake area (Montenegro).

Agent-based Models in Archaeology:
Are there Limits?

March 21st 2019 - Haus der Archäologien in Bochum

14:00 to 14:10

Welcome

ReSoc

14:10 to 14:30

Introduction (slides)

Maja Gori & Frederik Schaff, Ruhr-University Bochum

14:30 to 15:15

Toughening the Methodology of Agent Based Modelling by Working Across Disciplines (slides)

Edmund Chattoe-Brown, University of Leicester

15:15 to 16:00

The Fallacy of Holism: Views on Practising ABM in Archaeology (slides)

Marc Vander Linden, University of Cambridge

16:00 to 16:30

Panel Discussion

Edmund Chattoe-Brown (University of Leicester)
Marc Vander Lindern (University of Cambridge)
Iza Romanowska (Barcelona Supercomputing Center)
Michael Roos (Ruhr-University Bochum)

Moderation: Maja Gori & Frederik Schaff
(Ruhr-University Bochum)

16:30 to 17:00

Open Discussion

Questions and comments from the local audience. This part is excluded from the streaming.

after 17:00

Reception

Snack & Chat open to all participants

Discussion panel

Michael Roos
Complexity Economist
Ruhr-University Bochum

Iza Romanowska
Computational Archaeologist
Barcelona Supercomputing Center

Organising Committee

"Resources in Society" (ReSoc) is a Leibniz founded cooperative project of the Deutsches Bergbau-Museum and the Ruhr-Universität Bochum.

Maja studied prehistory in Rome. In 2007 moved to Germany to gain her PhD in Pre- and protohistory and Aegean archaeology (cotutelle de thése Heidelberg – Paris). She has worked at the Universities of Mainz, Amsterdam and Heidelberg, and at the Italian CNR (National Research Council). Since 2017 she works in the project ReSoc with a project focusing on mobility in the Balkans and Central Mediterranean.

Maja Gori, Ruhr-University Bochum

Prehistoric Archaeologist

Frederik studied economics in Trier, including classes in Sociology and Management Science. After his studies he worked as an environmental economist at University of Hagen, where he became facinated by the methodology of agent-based modelling. He received his PhD on "Pure Agent-based Computational Economics ..." in August 2016. Since June 2017 he works together with archaeologists in the project ReSoc and develops a vision of "Pure Agent-based Archaeology".

Frederik Schaff, Ruhr-University Bochum

Computational Economist

Abstract References

Barceló JA & F del Castillo (2016): Simulating the Past for Understanding the Present. A Critical Review. Chapter 1. In JA Barceló, F del Castillo (eds.): Simulating prehistoric and ancient worlds. Cham, Switzerland: Springer (Computational social sciences).

Drost CJ & M Vander Linden (2018): Toy Story. Homophily, Transmission and the Use of Simple Simulation Models for Assessing Variability in the Archaeological Record. In Journal for Archaeological Method and Theory 122 (2): 1087–1108.

Chattoe-Brown E (2013): Why Sociology Should Use Agent Based Modelling. In Sociological Research Online 18 (3): 1–11.

Cegielski, Wendy H.; Rogers, J. Daniel (2016): Rethinking the role of Agent-Based Modeling in archaeology. In Journal Of Anthropological Archaeology 41: 283–298.

Lake, M. W. (2014): Trends in Archaeological Simulation. In Journal of Archaeological Methodology and Theory 21 (2), pp. 258–287.

Rogers, J. Daniel; Cegielski, Wendy H. (2017): Opinion. Building a better past with the help of agent-based modeling. In PNAS 114 (49): 12841–12844.

Affiliations & Sponsors

Made with Mobirise