A Canonical Theory of Origins and Development of Social Complexity

ABSTRACT

The puzzle of origins and future of government and social complexity in human and social dynamics, arguably a characteristic feature of the emergence and long-term evolution of hierarchy and power in the history of civilizations, is an enduring topic that has challenged political scientists, anthropological archaeologists, and other social scientists and historians. This paper proposes a new computational theory for the emergence of social complexity that accounts for the earliest formation of systems of government (pristine polities) in prehistory and early antiquity, as well as present and future political development. This general social theory is based on a “fast process” of crisis and opportunistic decision-making through collective action, which feeds a “slow” process of political development or decay. The “fast” core iterative process is “canonical” in the sense that it undergoes variations on a recurring theme of signal detection, information-processing, problem-solving, successful adaptation and occasional failure. When a group is successful in managing or overcoming serious situational changes (stresses or opportunities, endogenous or exogenous, social or physical) a probabilistic phase transition may occur, under a specified set of conditions, yielding a long-term (slow) probabilistic accrual process of emergent sociopolitical complexity and development. A reverse process may account for decay. The canonical theory is being formally implemented through the “PoliGen” agent-based model (ABM), based on the new Multi-Agent Simulator of Networks and Neighborhoods (MASON). Empirically, the theory is testable with the datasets on polities developed by the Long-Range Analysis of War (LORANOW) Project. This paper focuses on the concepts, mechanisms, and basic formal structure that constitute the canonical theory and inform the subsequent simulation model.

ACKNOWLEDGEMENTS

Funding for this project was provided by the Center for Social Complexity, George Mason University. The first version of this paper was presented at the Workshop on “Origins of Social Complexity,” Center for Social Complexity, George Mason University, October 18–20, 2002. Revised versions were presented at the Department of Anthropology, National Museum of Natural History, Smithsonian Institution, Washington, DC, USA; the 5th International Conference on Complex Systems, Quincy- Boston, MA, USA, May 16–28, 2004; and the Third International Conference on Hierarchy and Power in the History of Civilizations, Russian Academy of Sciences, Center for Civilizational and Regional Studies, Institute for African Studies, and Russian State University for the Humanities, School of History, Political Science and Law, Moscow, Russia, June 18–21, 2004. Thanks are due to the following colleagues for comments and discussions: Guillermo Algaze, Dimitri M. Bondarenko, Mikhail Burtsev, Ray Dacey, Christian Davenport, John Ferejohn, William Fitzhugh, Alessandro Guidi, Thomas Hall, William Honeychurch, Mark I. Lichbach, Artemy Malkov, Misty Gerner, Daria A. Khalturina, Jürgen Klüver, Andrey V. Korotayev, Nikolay N. Kradin, Sean Luke, Ann Palkovich, Chris Pierce, Guido Giacomo Preparata, Dwight Read, J. Daniel Rogers, Philip Schrodt, James Snead, Christina Stoica, Alberto Trobia, Peter Turchin, David Warburton, and Tony Wilkinson.

Notes

¹Excellent recent surveys on the origins of social complexity in West Asia and East Asia are found in Feinman and Marcus (1998), Liu (2004), Loewe & Shaughnessy (1999), and Rothman (2001).

²MASON is an open-source Java simulation toolkit developed as a collaboration between the Evolutionary Computation Laboratory and the Center for Social Complexity at George Mason University. It is available at http://cs.gmu.edu/~eclab/projects/mason/.

³The relationship between the fast and the slow process resembles the relationship between ontogenesis and phylogenesis, in the sense of Klüver (Introduction: The Question of Origins), because the coherent structures in the latter emerge from realizations of the former. Thus, fast : slow :: ontogeny : phylogeny.

⁴By definition, 1 < m ≤ n. The causal situation denoted by the m-out-of-n connector is one of partial necessity (incomplete conjunction) or partial sufficiency (incomplete disjunction), depending on the value of m within the closed interval [1, n]. This occurs whenever n conditions (more than one) must be met from a broader set of n possible conditions for an occurrence, hence the term “several-among-some,” where m and n denote “some” and “several,” respectively (Cioffi-Revilla, 1998: 198–204). The m-among- n causal occurrence is produced by the binomial combination of m minimally necessary events among a larger set of n plausible causes, Cm,i = n!/[(n − i)!i!], where n! = n × (n − 1) × (n − 2) × … × 1. This causal connective reduces to (i) the pure conjunctive case as m → n, and (ii) the pure disjunctive case as m → 1.

⁵The opportunity-willingness principle is discussed in detail in Starr (1978), based on Sprout and Sprout (1969). See also Cioffi-Revilla and Starr (1996), Most and Starr (1989, ch. 2), and Russett and Starr (1996, 19–22). An equivalent perspective on the set of conditions that are minimally necessary for political action is developed by Elster (1993, 162–79), based on Tocqueville's theory of desires and opportunity, and Wendt (1987), using the structure-agent distinction.