in Archeology, CSS, CSS Reading Notes, English

Complexity, Social Complexity, and Modeling (Barton, 2014)

My notes on “Complexity, Social Complexity, and Modeling” article [zotpressInText item=”{22B4N3M8}”] which discusses application of agent based modeling to archaeology:

Descriptive/confirmatory statistics that dominate quantitative aspects of modern archaeological practice are not designed to deal with complex interactions and multilevel feedbacks that vary across space and time. Nor do narratives that simply state that societies are characterized by interacting agent/actors who share cultural knowledge, and whose interacting practices create emergent social-level phenomena add much to our understanding. Computational and systems dynamics modeling offer the first generation of such analytical protocols especially oriented towards the systematic study of CAS.

  • near decomposability: strong intra-component connectivity (linkages within groups) and weak inter-component  connectivity (linkages between groups). e.g. Roman Empire did not collapse into anarchy but decomposed into its administrative
    provinces (whose boundaries resemble those of modern nations that make up the European Community today).
  • nonlinear causality: CAS can absorb high perturbation and it may collapse with a minor disruption which initiates a cascades of changes that leads linkages to break. So it is difficult to predict system-level behavior from properties of the components alone.
  • emergence: CAS often exhibit novel behaviors at the system level that are very different from anything exhibited by any components. e.g. cultural knowledge construct rockets even no single person has the knowledge and skills to do so.
  • self-organization: interactions computation, and emergence that characterize CAS are not imposed by some external force, but rather develop organically as a consequence of endogamous rules that govern the behavior of individual components


  • Viewing human societies through a CAS lens entails a focus on information flow, decision-making, interactions at multiple scales of organization, and non-linear dynamics in which individual agency generates system-level emergent phenomena—all of which are invisible in the archaeological record
  • how can we systematically track and explain non-linear chains of causality that cascade? Difficult because our narratives are inherently linear and our empirical data are static.
  • how can we recognize and account for prehistoric socio-ecological CAS when key features are not preserved in the archaeological record?

Computational simulation modeling offers a valuable protocol for combining anthropological and CAS knowledge to create experimental environments in which to explore non-linear causality in complex systems and generate results that can be evaluated against the empirical archeological record.

In the rest of the paper Barton discusses the relations more concretely on an example of his early computational models where he explores the relationship between the observable archaeological record and the dynamics of past land-use and landscape change for small-scale early Puebloan societies of the upland American Southwest, swidden farming v.2 implemented in NetLogo

[zotpressInTextBib sort=”ASC”]