Talk: Tom Brughmans (University of Konstanz), “Roman bazaar or market economy? Explaining tableware distribution processes in the Roman East through computational modelling”.
Date: Tuesday, 25 November 2014
Time: starting at 17:00 c.t. (i.e. 17:15)
Venue: DAI, Wiegandhaus, Podbielskiallee 69-71, D-14195 Berlin (map)
The study of the Roman economy is populated by a large number of sometimes conflicting models. These models are rarely formally compared, and many remain untested due to the limited use of formal hypothesis testing methods in Roman studies and the significant data requirements to enable their use. This paper illustrates how broad patterns in large archaeological datasets allow for aspects of these models to be tested, and suggests agent-based network modelling as a particularly fruitful approach for the study of the Roman economy.
One of the most robust patterns observed in the collected ceramic tableware data in the Roman East is the variability of distribution patterns of different tablewares (products characterised by a distinct clay fabric and produced in different centres). Some wares such as Eastern Sigillata A were distributed on a supra-regional scale for centuries, others were of somewhat more restricted importance (Eastern Sigillatas B, C, and D), whilst yet other wares were purely produced for local consumption (e.g. Boeotian tablewares). What were the mechanisms that led to these strong differences in the wideness of products’ distribution patterns? A number of hypotheses have been published identifying and coupling possible contributing factors, including the role of social networks in allowing for the flow of information and goods both within and between markets. Most scholars seem to agree that a complex mix of mechanisms working on multiple levels was responsible for the considerable differences in tableware distribution patterns. However, these mechanisms remain untested given the need in Roman studies for workable methods that allow for expressing and evaluating a complex mix of hypothetical processes to better understand archaeologically attested large-scale distribution patterns (Davies 2005; Morris et al. 2007).
This paper aims to evaluate aspects of two such hypotheses: Bang’s (2008) claim that differences in the distribution of tablewares can be the result of weak market integration, and Temin’s (2013) opposing claim that the markets in the Roman world were well-integrated. It presents an agent-based network model simulating the social networks which enable the flow of information and goods between traders. The model by Jin and colleagues (2001) is modified to create social networks of traders on different markets, where different degrees of market integration can be enforced by modifying the value of one variable. The results of experiments with variable degrees of market integration are subsequently compared to the tableware data collected in the ICRATES database (Bes and Poblome 2008). The results suggest that, contrary to Bang’s hypothesis, limited availability of reliable commercial information from different markets is unlikely to give rise to the large differences in the wideness of tableware distributions observed in the archaeological record. A degree of market integration is necessary (between 12-40% of all transactions according to the model). However, it also emphasises the importance of intra-market transactions (60-88% of all transactions). Moreover, tablewares produced close to large urban centres will have a much higher probability of being distributed to many sites than tablewares produced close to small urban centres. We conclude that agent-based network modelling provides scholars of Roman trade a tool for expressing aspects of their hypotheses and that future work should focus on factors driving market integration against a dominant background of local market-based trade.
This paper concludes that the study of the Roman economy would very much benefit from embracing computational modelling approaches because (i) it forces scholars to consider the comparability of descriptive models, (ii) it allows comparison of simulated outputs with archaeologically observed outputs, and (iii) it allows to map out the grey zone between extreme hypotheses and refocus our descriptive models away from hypotheses that do not compare favourably with the archaeological record.
Keywords: roman economy, ceramics, agent-based modelling, network science
Bang, P. F. (2008). The Roman bazaar, a comparative study of trade and markets in a tributary empire. Cambridge: Cambridge university press.
Bes, P. M., & Poblome, J. (2008). (Not) see the Wood for the Trees? 19,000+ Sherds of Tablewares and what we can do with them. In Rei Cretariae Romanae Fautores Acta 40 (pp. 505–514). Bonn.
Davies, J. K. (2005). Linear and nonlinear flow models for ancient economies. In J. G. Manning & I. Morris (Eds.), (pp. 127–156). Stanford.
Jin, E. M., Girvan, M., & Newman, M. E. (2001). Structure of growing social networks. Physical review. E, Statistical, nonlinear, and soft matter physics, 64(4 Pt 2), 046132.
Morris, I., Saller, R. P., & Scheidel, W. (2007). Introduction. In W. Scheidel, I. Morris, & R. P. Saller (Eds.), The Cambridge economic history of the Greco-Roman world (pp. 1–12). Cambridge: Cambridge University Press.
Temin, P. (2013). The Roman Market Economy. Princeton: Princeton University Press.