University of Chicago Center for International Studies
and Argonne National Laboratory
The Joint Threat Anticipation Center was a collaborative project of the Center for International Studies and the Center for Complex Adaptive Agent Systems Simulation at Argonne National Laboratory. It was funded by the Advanced Systems Concepts Office (ASCO) of the Defense Threat Reduction Agency (DTRA) from 2004-08, as part of the agency's larger Threat Anticipation Program.
The foundational objectives of the center were as follows:
- Expand and champion the art and science of anticipating threats.
- Establish a center located at the University of Chicago that will become a recognized "center for excellence" of threat anticipation expertise, promotion, and products.
- Develop a repository at Argonne National Laboratory for models, publications, and institutional memory of threat anticipation.
- Provide the services that can assist ASCO with focused TAP activities and contacts that facilitate federal interagency participation.
As with any extended collaborative research project, these objectives evolved as the relationship between the two partners and funding agency grew and developed over four years. As concerns CIS, the Director and Associate Director identified and developed the relevant social science research and research product of the University of Chicago community, and built links between those researchers, Argonne, and DTRA to help advance the field in computational social science modeling and the anticipation of national security threats.
Phase I, 2004-05
In February, 2004, DTRA/ASCO agreed to provide seed funds to support existing Chicago research, in the expectation that this work would be incorporated into a future center. Thus, this initial phase of the supported two major faculty research projects at Chicago:
Professor of Political Science Robert Pape's work on the analysis of suicide terrorism. While Professor Pape has worked extensively on this topic, the objective of the portion of his research supported by DTRA was to expand the ability of the United States to anticipate future suicide terrorist threats. In the initial phase of what would be JTAC's longstanding support of this work, Professor Pape began development of a database of suicide terror attacks worldwide from 1980 to 2003. This was a labor-intensive project with a team of research assistants working on English and foreign-language sources for a twenty week period.
Professor of Economics Robert Townsend's ongoing research on inequality in Thailand, specifically the relationship of economic inequality, poverty and exclusion to violence and political unrest. In Thailand, Professor Townsend could examine these phenomena in the context of national growth and globalization. His researchers could also consider mitigating or exacerbating factors, such as the role of local networks (family, religion), the actual and potential distribution of governmental infrastructure (education, health, credit), and government transfers. The research is done with a combination of economic models, survey data, and GIS mapping.
DTRA support of Townsend's work was centered on the construction of a comprehensive database archive for model-based research. While the archive and research techniques were developed for the analysis of Thailand, it was expected that the work could serve as a prototype for other rapidly developing economically stratified countries. A secondary objective of the support was to advance capability for spatial pattern analysis, in particular to overlay databases, and thus project reliable measures of income, wealth, and other variables onto the entire country.
As mentioned above, an important part of the initial phase of DTRA/ASCO support was the expectation that an institutional center would be established to support future threat anticipation research. The principal investigators anticipated that the center (established as JTAC in early 2005) would become a recognized location of threat anticipation expertise, and serve as a repository for models of threat anticipation and social science tools to evaluate those models. To that end, during Phase I, experts at Argonne put considerable effort into research on the state of the art in terrorist modeling and the evaluation of current social science models — particularly those supported by or in use by DTRA at that time.
In April 2005, Argonne and the University of Chicago jointly held a two-day workshop, "Threat Anticipation: Social Science Methods and Models," to further consider the state of the field and other issues related to threat anticipation, the social sciences, and computational and formal models. A substantial archive of the workshop's materials, including summaries, papers, audio recordings, and presentation slides is available at the JTAC archive, http://jtac.uchicago.edu/conferences/05
Phase II, 2005-07
In the second phase of support from DTRA, the newly established Joint Threat Anticipation Center commenced a more direct working relationship with the Advanced Systems Concepts Office. In particular, JTAC set out to explore the role of cultural knowledge in reducing threats to U.S. national security and to expand the ability of the United States to anticipate future terrorist threats.
In preparing a research scope for Phase II, JTAC proposed three broad objectives:
- Development of frameworks and models for integrating cultural knowledge in meaningful ways to support policy analysis.
- Development of new social, cultural, behavioral, and economic knowledge relevant to anticipating future terrorist threats.
- Establishment of a broader dialogue within the research community and between the research and policy communities.
The first objective, focused on models, was supervised by Argonne, while the second, based on social science research, was overseen by the Center for International Studies. The aspirational third objective was where the two worked most closely together, while the partners continued an interchange to further develop the Social Science and modeling pieces of JTAC itself. A key component of this interchange was in the planning and execution of a second threat anticipation workshop (see below).
In Phase II, JTAC added its first full-time employee, Jonathan Ozik, in a post-doctoral position. Ozik spent equal time at Argonne and the University meeting with faculty and students, facilitating information exchange, and integrating research findings from the ongoing social science and threat anticipation modeling projects.
JTAC continued to support the work of Professors Pape and Townsend described above, while also issuing a call to Chicago faculty for proposals that would broaden the research scope of the center. From among the proposals, JTAC and DTRA/ASCO selected Professor John Goldsmith in Linguistics, whose ongoing work is to further develop automatic machine translation of poorly studied languages into English. Goldsmith makes use of a great deal of recent research activity in computational linguistics based on systems that learn the principles of translation of one language to another through a sample of a document that has been translated by a bilingual human being. Goldsmith's program, Linguistica, is an evolutionary step forward — it is capable of unsupervised learning: making inferences and acquiring knowledge about the structure about words or sets of words without any pre-existing information about the language from which the words are drawn.
JTAC also called for proposals from advanced graduate students with the goal of having both JTAC partners (CIS and Argonne) participate in "cross-training" graduate students already working on topics related to threat anticipation, though not necessarily with a background in computational modeling. Three advanced Ph.D candidates from Political Science were selected: Sevag Kechichian, Keven Ruby, and Mark Smith. Summaries of their work can be found at http://jtac.uchicago.edu/projects.shtml. Each student received an orientation to computational agent-based modeling at Argonne, participated in the 2nd Annual Workshop, and represented JTAC during site visits from the funding agency and other interested entities.
In addition to the research and modeling work conducted within JTAC, the center was also tasked with evaluating models provided by DTRA. Specifically, the experts at Argonne were asked to establish a framework for the validation and verification (known in the field as simply "V&V") of models submitted to or in use by DTRA from research programs nationwide. At a basic level, validation of a model is an evaluation process to determine how closely the model represents and reproduces phenomena observed in the "real world." Verification of a model is simply ascertaining if the model is programmed correctly and performs the way the designer intended.
Discussion of these V&V frameworks was central to the 2nd JTAC workshop, "What are National Security Threats", held April 3-5, 2006. The workshop was a compelling interdisciplinary effort to consider serious social science research within the context of future agent-based models that would incorporate the findings. As with the first JTAC workshop, a large archive of materials is available at http://jtac.uchicago.edu/conferences/06/.
Late Phase II / Interim Period
Following the two JTAC workshops in 2005 and 2006, many of the participants expressed concern about how closely the proposed social science models adhered to basic social science theory. These foundational obstacles, termed "Grand Challenges," were the bottlenecks in the further development of the models, as well their acceptance and widespread deployment. For example, models of social and cultural processes lacked accepted theoretical mechanisms for identity formation at the individual and group levels, key precursors to group action.
To specify and discuss these Grand Challenges, a number of informal meetings were held between July, 2007 and January, 2008. During each session, modelers from Argonne made presentations and discussed pre-circulated readings on key conceptual issues with advanced graduate students from History, Political Science, and Anthropology.
Phase III, 2007-08
In the third phase of the JTAC program, the Center for International Studies focused its efforts much more directly on providing a consulting role to Argonne and DTRA. The Grand Challenge meetings that began in Phase II were an extremely successful mode of interaction, and offered useful findings that enhanced the work of both Argonne and DTRA/ASCO. CIS thus concentrated on identifying useful participants in these meetings, selecting topics and readings, and ultimately bringing the numerous parties together.
In addition to the above, CIS also selected the work of Professor Steven Wilkinson (Political Science) for preliminary project funding. Professor Wilkinson's research considers the relationship between environmental change, scarcity, and conflict. For JTAC, Wilkinson and two graduate students assembled a detailed report on the following: (1) the existing research on environment and conflict, specifically investigating whether the increased pace of environmental change had been factored into previous studies, (2) the specifics of conflicts in the Middle East and South Asia, and how they might be affected by climate change (i.e. water scarcity, declining arable land), both within states and between existing rivals, and (3) an explicit consideration of whether environmental and climactic factors might affect our existing models of intra-state conflict and our models of nuclear escalation and interstate conflict. The report is available here (PDF).