Chapter One of this thesis distinguishes two views of the metaphysics of attention. According to the first of these views - referred to as the 'process-first' view - attention is a process and what it is for a subject to pay attention is for that process to be executed by or in the subject. According to the second view - referred to as the adverbial view - what it is for a subject to pay attention is for the subject's activities, whichever processes constitute them, to be executed in a certain manner. The distinction between these two approaches is explained in terms of a more primitive distinction between properties of objects and properties of events.
The process-first view is then shown to be incorrect by an argument drawing on empirical research into unilateral neglect syndrome, and into the processes of feature binding.
Chapter Two presents an adverbial analysis of attention according to which attention to a task is the absence of irrelevant activity among the cognitive resources that the agent can, with understanding, bring to bear in the performance of that task. This analysis, which is explained in detail, can be understood as saying that attention is not any one particular process, but is unison among various cognitive processes. It is referred to as 'the cognitive unison view'.
Chapter Three examines the implications of the cognitive unison view for the science of psychology, showing that the view leads to the rejection of some of the questions and methods found in the scientific literature. Recent work on the recognition of facially expressed emotions is found to be especially problematic.
Chapter Four defends the cognitive unison view against three objections which accuse it of being unable to account for attention's causal efficacy on the grounds of cognitive unison's extrinsicness, its privativeness, and its higher-order nature.
Chapter Five reviews the history of philosophical approaches to attention, suggests that attention has an important role in explaining why aesthetic value has a moral dimension, and speculates about ways in which the cognitive unison view might account for this role.
