John J. Davenport

This dissertation develops a new account of the will as a motivational power of commitment or resolve, which contrasts with volition in the 'thin' sense as a mere decision to form an intention. After tracing the historical background of these concepts in Chapter I, Chapter II argues that will involves the capacity to motivate oneself "projectively" in devotion to causes and purposes for which one did not necessarily have an antecedent "desire." "Projective motivation" differs from three types of "desire"-- inarticulate appetites and generic urges, subjective inclinations and preferences for definite objects, and evaluative desires following from judgments that something is objectively 'good' for us--all of which arise from the sense that something is lacking in the agent. Projections do not share this teleological lack-structure, and are grounded in considerations other than the agent's well-being, such as the existential meaningfulness of pursuing ends, as distinct from realizing them. ;This concept helps resolve problems with eudaimonist moral psychologies. It also helps make sense of Kant's notions of spontaneity and autonomy, although moral motivation as Kant conceives it is only one form of projective motivation. Hence the usual teleological and deontic models do not exhaust all the kinds of motivation there are. Chapter III then argues that thinkers as divergent as Scotus, Levinas, Arendt, and Frankfurt implicitly appeal to the possibility of projective motivation. ;The last two chapters focus on Harry Frankfurt's hierarchical account of volitional "identification" , comparing this notion to Heidegger's and Foucault's interpretations of 'care of the self.' Chapter IV distinguishes and reformulate five objections against Frankfurt's initial account of identification in terms of second-order desires; these objections point towards the need for a qualitative distinction between ordinary desires and the states which constitute identification. It then analyzes Frankfurt's later attempts to address these difficulties, and argues for a new account of identification in terms of 'higher-order projections.' Chapter V defends this new existential analysis against rival accounts of identification by Stump, Taylor, Watson, Velleman, Bransen, Buss, and Dennett.