Farkas - State Learning and International Change

Introduction

Change, State Learning, and International Politics

A science of international politics, like any scientific enterprise, is an attempt to explain change. Whether a particular author is more interested in explaining why change occurs (e.g., Gilpin 1981), why change does not occur for a long period (e.g., Waltz 1979), or how particular changes came about (e.g., Deudney and Ikenberry 1991), accounting for variance in behavior is at the core of the study of international relations. Change in international politics occurs through alterations in the behavior of international actors. These actors, usually states, may modify their foreign policies because of changes in conditions such as the balance of power, the distribution of strategic resources, or patterns of trade. These changes, though, cannot affect foreign policy directly because states themselves cannot act. A change in foreign policy is the result of actions taken by human beings on behalf of the state. State learning, therefore, is the process that intervenes between the causeinternational changeand its effect, policy change. So, to appreciate how international changes will be reflected in foreign policies, one must understand the accuracy of and the manner through which state learning transmits information about international change. This book is an attempt to understand this process of state learning.

By focusing on state learning, this effort to understand the process of international change tries to explain how states arrive at foreign policy decisions. In so doing, it asks questions about the rationality of the state and the relationship of the state to the individuals who act on its behalf. Specifically, it tries to understand how states can appear to be rational when human policy makers do not themselves behave as if they were rational actors. Thus, one of the contributions of this project is to reconcile the research programs of rational choice and political psychology by developing a set of models grounded in assumptions consistent with psychological findings and producing behavior consistent with rational choice. To use Simon's (1976) terms, this is an effort to bridge the gap between models of substantive rationality and models of procedural rationality (see also Binmore 1987).

Although this book is about state learning, it necessarily confronts issues dealing with learning by collectivitiesthat is, groups of people striving toward a common goalmore generally. When studying international relations, it is often convenient to write as if states were thinking, analytic beings that interact with their environment and pursue their objectives. One must be very careful, however, to avoid letting this linguistic shortcut lead to unsound theorizing. States are not living beings; they are logical constructs. As such, they have no preferences, nor can they think, analyze their situations, take action, or, for that matter, learn.

Individual human beings acting as a group determine a state's goals, acquire information, analyze conditions, and take action on behalf of the state. Therefore, when studying state learning one is effectively dealing with a more generic question: how do collectivities learn? While the conclusions of this volume are applicable most directly to the behavior of states, and while the illustrative data are drawn largely from international relations, the book also deals with the broader phenomenon of collective learning. This explanation of collective learning is an attempt to ground theories of group behavior in plausible theories of individual action (see Schelling 1978). The models developed here are based on the assumption that human beings are goal-seeking actors. As such, this project is in the tradition of rational-choice modeling. However, this book is more an effort to the support the assumptions of rational-choice models, and to address some of the problems suggested by those models, than it is an effort to develop new rational-choice modeling techniques.

Learning and Rationality

The work presented here is an attempt to use evolutionary models to reconcile insights from rational choice and psychology to develop a model of state learning. For more than 50 years, rational-choice theorists have been pointing to the power, elegance, and clarity of their models (e.g., Friedman 1953). At the same time, critics have complained about the nature of the assumptions made in rational-choice models and pointed out that they do not comport with what is known of human behavior (e.g., Quattrone and Tversky 1988). This book asks how peoplewho individually do not act as if they were rationalcan form a group that does appear to act rationally.