Rethinking the International System as a Complex Adaptive System
interactions within the components of the system , without any central control .” In other words , self-organization creates stable macroscopic patterns arising from the local interaction of agents with limited information and computational power ( Epstein & Axtell , 1996 , p . 35 ). One generic pattern of self-organization is the one of “ self-organized criticality ,” which implies “ that from any initial condition , the system tends to move toward a critical state , and stay there , without external control ” ( Bak , Tang , & Wiesenfeld , 1987 , p . 381 ; Downey , 2012 , p . 81 ). Self-organization , as it will be further explained later , requires the system to signal and process information .
If we want to define the international system as a CAS , we can say that : the international system is composed of many diverse , interconnected , and interdependent agents that iterate nonlinear relationships from which multilevel behavior evolves and emerges . Because of non-linearity , lack of central coordination , and the presence of lever points — the international system should not be studied with traditional positivist methodologies that assume linearity ( Holland 2013 , Chapter 3 ). Positivist theories are built inductively or deductively , with the previous discovering patterns in empirical data , and the latter specifying a set of axioms and testing them ( Harrison , 2006b , p . 139 ).
Complexity theory follows a third way of doing science between induction and deduction . 6 By relying on computational simulation , it deductively sets axioms and generates data that can be studied inductively ( Harrison , 2006b , p . 139 ). Since CASs are irreducible , scholars ultimately need the assistance of simulation to be able to explain those ( Earnest & Rosenau , 2006 , p . 145 ).
To conclude , studying complexity does not require a paradigm shift in the way IR is studied . However , it does require some change in the criteria that scholars use to observe the world and build theories . When scholars model complexity , they need to shift from continuous to discrete , from linear to nonlinear , from deterministic to stochastic , from abstract to detailed and from homogeneous to composite ( Downey , 2012 , p . 4 ). In addition , also their purposes in research have to change ; studies should be explanatory and not necessarily predictive , models should be instrumental and not realist , and theories should be holistic rather than reductionist ( Downey , 2012 , p . 4 ).
For a New Understanding of International Relations
A New Grammar and Taxonomy
Ontology and epistemology dictate what can be classified in a taxonomy ,
6 “ But unlike deduction , simulation does not prove theorems with generality . Instead , simulation generates data suitable for analysis by induction . Nevertheless , unlike typical induction , the simulated data come from a rigorously specified set of assumptions regarding an actual or proposed system of interest rather than direct measurements of the real world . Consequently , simulation differs from standard deduction and induction in both its implementation and its goals . Simulation permits increased understanding of systems through controlled computational experiments ” ( Axelrod , 1997 , p . 4 ).
119