PEOPLE ARE COMPLICATED. We try to achieve a lot of things, and those things often conflict with each other. Our actions flow from both logic and emotion. Because of these complexities, it can be difficult to understand the behavior of other people or even our own behavior.
Political scientists who study judicial behavior generally rely on a few models of behavior that have much in common with each other. In the models that dominate the field, judges are not very complicated. As depicted in these models, judges on many courts want only one thing, good legal policy as they define it. They act in a logical, unemotional way to pursue that goal. Thus these models rule out of consideration a great deal—all of judges' emotions and most of the goals they seek to achieve.
All models of human behavior simplify reality, and even highly simplified models can be very useful. Indeed, such models facilitate the study of behavior by making analysis more manageable. The simplified models that scholars employ to study judges have done much to help us understand judicial decision making, and our understanding continues to grow.
And yet these models are not entirely satisfying, primarily because they rule out so much. It is possible that what they leave aside is not very important to judging. Perhaps most of judges' goals and all their emotions affect only their lives outside court, not their work in court. But that seems unlikely: the line between those two segments of judges' lives could hardly be that sharp. Besides, some of the judicial behavior that we observe does not fit comfortably within the simplified models that scholars employ.
For these reasons it seems worthwhile to step outside the standard models and consider judges' behavior from other vantage points. Several years ago I began to think about what I call judges' personal audiences. These are the people whose esteem judges want for its own sake, not because that esteem might help them make good legal policy or achieve some other end. In thinking about personal audiences, I focused on two questions. First, what kinds of audiences are likely to be most important to judges? Second, how might judges' interest in approval from those audiences affect their behavior as judges?
Thinking about those questions led me to a perspective on judicial behavior that is based on judges' relationships with their audiences. Not everything about judicial behavior can be understood from that perspective, but these relationships illuminate a good deal about what judges do. Consideration of judges' concern with what their audiences think of them helps in understanding some aspects of judicial behavior that the domi-