Feelings argues for the counter-intuitive idea that feelings do not cause behavior, but rather follow from behavior, and are, in fact, the way that we know about our own bodily states and behaviors. This point of view, often associated with William James, is called self-perception theory.Self-perception theory can be empirically tested by manipulating bodily states and behaviors in order to see if the corresponding feelings are produced. In this volume, James D. Laird presents hundreds of studies, all demonstrating that feelings do indeed follow from behavior. Behaviors that have been manipulated include facial expressions of emotion, autonomic arousal, actions, gaze, and postures. The feelings that have been induced includehappiness, anger, fear, romantic love, liking, disliking, hunger, and feelings of familiarity. These feelings do not feel like knowledge because they are knowledge-by-acquaintance, such as the knowledge we have of how an apple tastes, rather than verbal, knowledge-by-description, such as theknowledge that apples are red, round, and edible. Many professional theories of human behavior, as well as common sense, explain actions by an appeal to feelings as causes. Laird argues to the contrary that if feelings are information about behaviors that are already ongoing, feelings cannot be causes and that the whole mechanistic model of humanbehavior as "caused" in this sense seems mistaken. He proposes an alternative, cybernetic model, involving hierarchically stacked control systems. In this model, feelings provide feedback to the control systems, and in a further elaboration, this model suggests that the stack of control systemsmatches a similar stack of levels of organization of the world. An original contribution to the study of the relationship between feelings and behavior, the volume will be of interest to social, emotional, and cognitive psychologists.