"Knowing Terrorism as a Complex Adaptive System"
The fundamental principle of security and strategic warfare is to "know thy enemy." Since September 11, the national security community has been struggling to do just that with respect to terrorist threats. Yet the nature of the terrorist threat is that it is constantly changing. Furthermore, there is a dynamic, non-linear feedback between our responses to the threat – whether they be instantiated through military action, media coverage, social groups, or formal institutions —and its evolution. These properties are the hallmarks of a complex system. Is such a system ultimately “knowable? What does complexity science offer to help us understand the enemy we do not know in a way that can help us “win the war”? The problem of knowing what we know about terrorism was articulated recently by Jerrold Post, who stated that, with respect to the war on terrorism, our fundamental problem is that "We do not know what we know." Our society functions as a complex, interconnected "system of systems" of diverse natures— engineered structures, bureaucratic institutions, informal social groups, and natural ecosystems — developed over a wide range of timeframes and degrees of complexity to perform multiple and different functions depending on the context. Understanding terrorism demands an analysis framework that is valid within each of these diverse systems, yet can also be applied across the interconnected whole. To complicate matters, our gaze must go beyond our own borders—to systems operative across different political, cultural, socioeconomic, religious and geographic boundaries. Transcending the boundaries between these disciplines has, to date, been an insurmountable obstacle for researchers and analysts alike. This is due in no small part to the historical path of evolution of the social science disciplines, and the research methods adopted for their development. For the last half century, the social sciences have taken great pains to develop rigorous research methods that adhere to the scientific principles of investigation, yet, due to the nature of the systems they describe, these are not always satisfactory for investigating complex, cross-cultural issues. This is due not only to the diverse, evolutionary nature of the systems described within each, but also to the lack of common tools, vocabulary, and analysis paradigms that are able to translate domain specific results across domains of expertise and into a larger problem context. The tools and methods of complexity science that have evolved in the last twenty years begin to provide such capabilities. At the same time, social science research methods are putting more credence in theoretical investigations wherein the researcher is in a constant cyclical process of being guided in his/her queries by the data gathered in the field. Such methods have been termed "grounded theory." Using grounded theory as a process of social science investigation mirrors - at an abstract level - the systems engineering paradigm outlined by Garajedaghi for analyzing complex systems. So, is there hope for a meeting of minds and methods on the horizon? Some of the tools and concepts for modeling engineering and biological systems that have come out of complexity science are advancing research into social, pyschological, and cognitive phenomenon. For example, the spread of group ideas and movements may be modeled as processes of diffusion, revolution, co-evolution, epochal evolution, and punctuated equilibrium. Learning and reasoning processes can be simulated by methods of cellular automata, genetic algorithms, evolutionary algorithms, and neural networks. Rational decision-making can be modeled via decision trees, hierarchies, embedded networks, and rule-based agents. Methods for topographical analyses of networks and the dynamics that act upon them have developed to enable exploration of these systems as random, scale-free, or giant stars, and what that means about the stability, robustness and predictablilty of these structures. The beauty and power of complexity science, however, lies in our utilization of these methods to continue to generate theory and test it against what we see in reality, and finally, to work on solutions to the pressing problems our society faces – such as terrorism. Are we at a point in this field to develop a coherent capability for understanding such complex problems as the war on terrorism? If so, what might that capability look like? This paper proposes a framework for exploring this question and its solution space, based on purpose of analytic inquiry, degree of system complexity, granularity desired/required and methodologies available.