Gavan Duffy

Language Games: Dialogical Analysis of INF Negotiations

This essay proposes dialogical analysis as a method of modeling political interactions. The method combines the formal theory of dialogical disputation, a family of theories drawn from linguistic pragmatics, and formal proof procedures. By analyzing models of their explicit and implicit contents in context, the method identifies the argumentative thrust of negotiation dialogues and shows systematically how the parties signal intent and commitment to one another. The paper illustrates the method by applying it to superpower interactions in the 1980s INF negotiations. The analysis indicates that American force deployments did not motivate the Soviet retreat from their early insistence on compensation for European missiles. The change in the Soviet position is better attributed to their strategic reconceptualization of the Cold War insecurity dilemma.

Political Science: Artificial Intelligence Applications

Scholars who apply artificial mtelligence to political questions seek, most generally, to expand the scope and relevance of political model analysis. By incorporating the effects of variable human notions, traditions, and meanings, they seek to humanize political models. Most early applications of artificial intelligence in political science research address substantive issues pertaining to political decision making. Most of these works apply production-system technology to construct choice models in for eign-policy decision contexts. In recent years, political applications have begun to diver sify. Today, lively research efforts flourish in widely varied application areas, such as computational text analysis, logic programming, computer learning, and conflict sim ulation. The works reviewed here constitute the early steps of a nascent program of study. Much remains to be accomplished. Nevertheless, the efforts conducted thus far suggest many potentially fruitful research avenues. Keywords: artificial intelligence, production systems, logic programming, computational hermeneutics, belief models, political simulation, political science.

Events and versions: Reconstructing event data analysis

International event datasets generally provide singular descriptions of the stream of international interactions. Each description represents one version of events. Yet, because events may be described in several ways, each reliant upon an alternative perspective or world version, these descriptions grossly underrepresent events. In its negative moment, this paper, “deconstructs,” after a fashion, the interpretations and versions that stand behind event codes, reviewing the widely‐suspected partiality of events datasets that threatens the validity of all subsequent analyses. In its positive moment, this paper outlines a reconstruction of the event‐data project as “computational hermeneutics.” This perspective takes seriously the plurality of world versions. It would construct reproducible models of textual accounts of international events and test hypotheses regarding the inferential effects of interpretive presuppositions embodied in alternative world versions. To initiate discussion, the paper proposes ...

An Early Warning System for the United Nations: Internet or Not?

Editors Note: An ongoing discussion within international studies concerns the design of a system to provide early warning of a variety of international crises to the United Nations. In this Forum, Gavan Duffy, organizer and chair of a roundtable on the problems and prospects for early warning at the 35th Annual Convention of the International Studies Association, continues the debate on a proposal for an Internet-based early warning system. This system is designed to overcome fundamental problems arising from differing theoretical perspectives toward what constitutes crisis and disaster, especially with respect to political conflict situations. Several reactions to the proposal follow, along with a final rejoinder by Duffy himself. As Duffy notes in his rejoinder, expanding discussion on this policy issue is crucial, and the Forum itself is a means forfurthering the discussion. Anyone wishing to join in the debate over early warning or the Internet proposal itself should contact the participants: Gavan Duffy, gavan@mailbox.syredu; Ted Gurr, tgurr@umd.edu; Philip Schrodt schrodt@ukanaix. cc.ukans.edu; Gottfried Mayer-Kress, gmk@pegasos.ccsruiuc.edu; and Peter Brecke, peter brecke@inta.gatech. edu.