In this essay we seek to investigate the way in which an agent's reasons for action explain the things he does intentionally. Our enquiry opens with an examination of recent debate on the nature of the connection between an agent's reasons for action and the actions they explain. Prominent contributors to this early debate include Anscombe, Davidson, and Melden. Accordingly, we try to recreate the intensity of a clash that emerges at this time, between supporters and opponents of the causal theory of action. The conflict is, fundamentally, about the claim that reasons for actions cause actions. Objections to the causal theory of action are examined, and we also look at responses to these objections. After considering the causal/non-causal debate at some length, we lay this issue aside temporarily, in an attempt to further our understanding of the concepts involved when an agent gives his reasons for action. To this end, we discuss the nature of reasons for action and their connection with an agent's knowledge of his intentional action. And we focus on what Anscombe refers to as a crucial distinction between an agent's non-observational and observational knowledge of his intentional action. Anscombe's pivotal claim is that an agent's reasons for action incorporate non-observational knowledge of his intentional action, as opposed to observational knowledge. One of our aims in this essay is to say what an agent's non-observational knowledge of his intentional action amounts to, and to trace the influence of Anscombe's account of such knowledge on the most recent discussion of intentional action.
Prior to pursuing the implications of Anscombe's analysis of the sort of knowledge an agent has of his intentional action, we examine Michael Bratman's planning theory of intention. Bratman rejects the theory of action supported, among others, by Anscombe, Davidson and Bishop, wherein an agent's desire to do a certain thing, forms part of his reason for action, and is held to explain the agent's intentionally performing the action. Opposing the desire-belief model of intentional action, Bratman seeks to establish a competing theory, grounded in the notion of plan-state intentions. In the planning theory of intention. Bratman claims that an agent's rational intention to do a certain thing requires consistency of related beliefs. For Bratman then, there is a normative constraint on practical reasoning, and it is notable that Bratman ties the value in doing a certain thing to an agent's beliefs, rather than his relevant desires. Our examination of Bratman s theory of intention prepares the way for an investigation of the relation between intention and belief.
Grice's attempt to defend the view that intending to do a certain thing implies believing that one will, sets the agenda for debate about the nature of the link between intention and belief. For Grice, however, there is a problem in trying to defend the above view, in as much as this view seems to entail support for the claim that beliefs implied by an agent's intending to do a certain thing, are non-evidential. That is, in so far as an agent's belief that he will do a certain thing, is implied by his intention to do that thing, such belief is independent of evidence. Hence, this belief cannot be dependent on confirmation or falsification through subsequent outcome. Yet, as Grice admits, the notion of a belief that cannot be confirmed or falsified seems incoherent. In this respect, advocates of the view that intending implies, or, in the case of Velleman, consists of believing that one will, are presented with a difficult challenge, namely how to account for the existence of non-evidential beliefs. We look at recent attempts to resolve this problem.
Having examined key areas relating to the issue of an agent's knowledge of his intentional action, we apply our increased understanding of the concepts involved in intentional action explanations to a reconsideration of the causal/non-causal debate. First, John Bishop's pro causal theory is discussed, wherein Bishop tries to show that intentional action explanations share the same status as explanations of events in the natural order. The substance of Bishop's theory lies in his claim that behavior that matches and is event-caused by an agent's reasons for action, counts as intentional action. We try to determine whether Bishop's causal theory of action is viable. Following our analysis of Bishop's view, we consider two recent non-causal theories of action, which attempt to show that adequate explanations of intentional action need to highlight the teleological nature of action. As George Wilson and Carl Ginet seek to show, an analysis of intentional action involves reference to an agent's intention in action, whereby the agent's action is accompanied by an intention with the right sort of content. On the Wilson/Ginet view, the agent's aim in performing the action cannot be carried by an intention that merely precedes his action. But if intentional action cannot be analysed in terms of intentions that precede the actions they explain, this provides grounds for a challenge to the viability of event-causal analyses of such action. We examine an attempt to defend the causal theory of action against the Wilson/Ginet non-causal view.
Finally we bring together the threads of our enquiry and, after focussing on the issue of whether reasons for action are essentially normative, we attempt to outline a brief sketch of our own view on the nature of reasons for action.