an extended abstract of a paper delivered at a sociology conference
The explosion of scientific interest in chaos and nonlinear dynamics has brought with it a number of attempts to draw broad implications for areas usually considered far removed from the physical sciences. Philosophical analysis can play a role in the important task of critically evaluating these interpretations, extensions, and appropriations of nonlinear dynamics. Such analysis can help us sort out the broad conceptual implications of scientific results, and to clarify the possible connections with other fields, including the study of human societies.
In this paper, I examine three ways researchers in the social sciences have used chaos theory. The uses I consider include straightforward application of mathematical techniques as well as more speculative uses to which I have given the name "metaphorical extension." I also consider some attempts to draw normative conclusions from results in the physical sciences. These examinations are part of an ongoing project to understand what chaos theory means and what it does not mean, as a way to investigate the general question of how research in the physical sciences is and ought to be translated across disciplines.
These reflections will concentrate on chaos theory narrowly construed: that is, the application of the techniques of nonlinear dynamical systems theory to bounded, aperiodic, unstable deterministic systems. The first type of use of chaos theory is the application of mathematical techniques. In the physical sciences, a typical experimental situation will yield a long series of measurements, from which a "strange attractor" can be sought in reconstructed phase space. Social systems present a genuine challenge to such attempts, though perhaps not an impossible one. Under what circumstances can the mathematical techniques of nonlinear dynamics and chaos theory be applied to social systems?
In general, the application of techniques from nonlinear dynamics to a system requires the following: the ability to identify and quantify the relevant variables of the system, the availability of time-series data of sufficient length and precision, and the plausibility of using a simple model that does not change in time. These requirements raise questions about the roles of quantification, precision, and human intensionality in the social sciences. Ultimately, we must ask whether some social systems may not be "dynamical" at all. That is, it may not be appropriate to consider some social systems to be governed by any simple mathematical equation whatsoever.
By metaphorical extension, the second kind of use of chaos theory, I mean the use of nonlinear dynamics as a new conceptual resource for the social sciences. To illustrate this kind of use, I will examine some of the actual metaphorical uses to which chaos theory has been put -- the implications some have drawn from the study of chaotic dynamics for our conception of the understanding of human history and social systems in general. My hope is that examining the fruitfulness of these uses will reveal some patterns that allow us to analyze and critique them.
In critiquing some uses of chaos theory in the study of human society, I focus on the notions of instability (including the technical notion of "sensitive dependence on initial conditions") and pattern (including the technical notion of "strange attractor"). While chaos theory can be misinterpreted and misused, I argue that a metaphorical application of chaos theory may serve as a useful antidote for a previous importation of conceptions from linear dynamics. The insights of linear Newtonian physics have sometimes proven useful for conceptualizing human social change. If these uses are legitimate, then reconceptions of the physical world can legitimately lead us to rethink our picture of the social world.
The third kind of use of chaos theory is to draw normative conclusions from analogies with nonlinear dynamics, such as the contention that chaos theory proves the optimality of laissez- faire capitalism or liberal democracy. I argue that such attempts often rely on mistaken interpretations of chaos theory. But while it is important to "get the science right," the mathematical structure of a theory does not fix its broader cultural meaning; scrupulous attention to scientific correctness may cause us to miss the very real meanings attached to scientific results.
To the extent that our widely accepted frameworks -- even frameworks for thinking about human society -- have been influenced by the physics of regular linear behavior, those frameworks are indeed challenged by chaos theory. But there are a number of dangers in attempting to spell out those challenges. Interaction between the natural and social sciences requires mutual respect and careful attention to the broader implications of our research.
