Order this book
Center for Social Complexity, George Mason University
Due to the editors' objective to cover at the same time individual, organizational, and societal models, the report has an integrative character, vertically penetrating different levels. A broad range of models is treated (see the very useful table stretching from page 62 to page 83), including non-computational models ("verbal models"), system dynamics models, organizational models, cognitive models (ACT-R, Soar, etc.), game theoretical models, decision making models, social network models, agent-based models, etc. I would not consider this publication a comprehensive introductory text book; but the approach taken for discussing formal modeling is refreshing It reveals modeling's important complementary function with regard to alternative modeling techniques and non-modeling methodologies. It furthermore underlines the need to systematically come to grips with those areas that constitute spaces between the individual (micro-level), the group (meso-level), and society (macro-level). This includes conceptualizing of, exploring into, and theorizing about those mechanisms and processes that constitute the links between the next higher and/or lower levels of analysis.
That all this is done more or less implicitly against the background of a government perspective is interesting from the perspective of potential research agendas. This is even more the case as the scientific advisory board of the report consists of such authoritative figures as Steven P. Borgatti, Kathleen M. Carley, Scott E. Page, Leigh S. Tesfatsion, etc.
The report nevertheless exhibits three main weaknesses. The first one is an almost exclusive US-centric view of social simulation. Important research endeavors in Europe and Asia-Pacific are almost completely neglected. This is not to say that some of the relevant literature is not cited, but it does not get woven into the report adequately either. It would, for example, make sense to be more explicit about Moss' ideas on policy formation from first principles (Moss 2002). This is particularly the case since his writings include a distinct evidence-driven approach (as is currently very much en vogue in public health, for instance) and a suggestion of a modeling cycle explicitly taking into account policy makers and analysts instead of just informing them once model output is ready. A similar argument can be made in favor of the French school of companion modeling (e.g. Barreteau 2003) where cultural idiosyncrasies are factored in through the explicit interaction in the simulation design process between modelers and "objects of investigation", i.e. the individuals that are being studied. Needless to say that modeling for the military could profit from such inspirations too.
The second shortcoming is an obsession with numbers and prediction. In the introduction it is stated that "[t]oday scientists find themselves at the edge of what he [Isaac Asimov, author of the dystopia Foundation] imagined - working on computational mathematical models of aggregate human behavior that allows them to understand, assess, and, to a very limited extent, predict 'the reactions of human conglomerates.' This report assesses how close they have come to that vision and what still remains to be done." (p. 13) It is somehow idle to re-discuss time and again the issue of social laws that have been discarded rightly on so many accounts by numerous social scientists. As if the "truth" lies only in laws and numbers, an almost exclusive occupation with the latter constitutes a further concern. The report's editors repeatedly emphasize need for automated data extraction tools. Both the search for social laws and ignoring the power of interpretative methodologies obfuscate the elaboration of a "true" understanding of the underlying mechanisms and processes of military conflict of all sorts. There is a clear obsession with misleading precision in the form of numbers and prediction. Afghanistan and Iraq should have been good enough examples to make clear that micro-models grounded in (often qualitative) evidence are needed more than ever as complementary, but alternative modeling exercises. Recent research clearly points into this direction (Kalyvas 2003; Sambanis 2004).
The third major shortcoming of the report, finally, is a misunderstanding of the process of socio-scientific research. One gets the impression that the editors have in mind a rationalized, bureaucratically organized model of scientific work - and thus discovery. Even though most social scientists have by now accepted the fact that research projects need to be (tightly) managed, the research process as such should be given enough freedom to creatively advance knowledge. The editors should have therefore perhaps made a better effort in distinguishing between socio-scientific academic research that serves the interests of the security establishment and applied modeling that serves the same purpose.
The editors justly warn in great detail on many occasions of having too great of an expectation with regard to the usefulness of social simulation for military purposes (chap. 10). However, one could expect a more open critique in front of their client and less talk the client wants to hear so that a truly informed decision about what models can contribute to serve the interest of the defense sector can be reached. More attention could also have been given to the issue of ethical concerns in relation with individual, organizational, and societal models developed for the defense sector. Nevertheless, for those who want to get a quick overview on state of the art military-related modeling in the US, this publication is certainly recommended. It is also serviceable, particularly in Chapter 11, for those researchers who are in the progress of writing a related grant.
KALYVAS, S (2003) The Ontology of "Political Violence": Action and Identity in Civil Wars. Perspectives on Politics 1(3), pp. 475-94
MOSS, S (2002) Policy Analysis from First Principles. Proceedings of the National Academy of Sciences 99(3), pp. 7267-7274
SAMBANIS, N (2004) Using Case Studies to Expand Economic Models of Civil War. Perspectives on Politics 2(2), pp. 259-79
Return to Contents of this issue
© Copyright Journal of Artificial Societies and Social Simulation, 2009