Volition, intention, and responsibility.

• Author: Audi, Robert
• Position: Symposium: Act & Crime

The philosophical theory of action is important for the law in many ways. There are difficult questions about what constitutes human action--the chief focus of both the analytical efforts of action theory and the regulatory efforts of the law. There is the perennial problem of characterizing freedom of action, with its related difficulties in classifying and appraising impairments of freedom. And with respect to both the conceptualization of action and the theory of punishment, there are questions about the nature of intention, deliberateness, premeditation, and other notions that are important for understanding and judging actions.

Michael Moore's Act and Crime(1) is a major contribution to this important and growing area of intersection between philosophy and legal theory. The book develops a theory of action in the light of a fine-grained survey of contemporary philosophy of action; it explores legal notions, such as those of an act, of intent, and of liability, in relation to both philosophical and legal literature; and it brings various elements in the theory to bear on a number of important problems in the law. The book is especially noteworthy for its wide-ranging analysis of significant positions in the literature; here, it performs a valuable service not only for legal theorists but for philosophers as well. It is among the most comprehensive and sophisticated treatises in the philosophy of action to appear in many years.

This Article will attempt three tasks: (1) to explicate key elements in an important part of Moore's theory, namely, its use of the notion of volition; (2) to appraise that part of the theory and, in the light of the appraisal, to offer an alternative approach to the same problems; and (3) to indicate how the resulting conception of action is significant for some aspects of moral responsibility and the law.

1. VOLITION AND INTENTION

   I address the theory of volition and its place in the wider theory of action in the context of a causal conception of human action and its explanation. Specifically, I assume that actions are caused by psychological elements in the agent, that these elements provide the materials for explaining the actions they cause, and that such causation may be, though it need not be, nomic, in the sense that there are laws, universal or statistical, linking the elements to the actions they explain. These assumptions are controversial, but I have defended them elsewhere.(2) They are points which, even if I am mistaken in thinking that Moore would accept them in some form, will not affect the discussion to follow. It is particularly important to consider volition in the context of a causal theory of action. The interest in volition seems to lie primarily in its promise of providing a kind of causal ground of action that roots actions in the agent in just the way appropriate to the kinds of behaviors that are paradigms of the sorts of actions the law seeks to regulate or, when necessary, to punish.

   One difficulty in approaching the topic of volition is that there is little we may presuppose in the way of a pretheoretical characterization. We might say, in line with both philosophical tradition and common speech, that volitions are acts of will; but no plausible volitional theory--which I shall call "volitionalism" for short-posits a substantive will as a kind of agent capable of its own acts.(3) If volitions are acts of any kind, as opposed to events that underlie acts, they are acts of an agent in the ordinary sense of the term "agent" as applicable to persons.

   Moore is well aware of these difficulties, and he rejects the problematic view that basic action simply is volition,(4) which implies that the overt physical deeds we call actions are not basic but are volitional acts under a behavioral description. Working within the constraints of what he terms the "mental-cause thesis," he proposes "to take "volition" to name a species of intention."(5) The broad conception he arrives at deserves full statement:

   Happily we have no need to resort to a fourth kind of mental

   state in order to think of volitions, for bare intentions fit the bill

   nicely. As we shall see, in order to execute even simple practical

   syllogisms, in order to resolve conflicts between our desires and

   between our beliefs, in order to account for the phenomenon

   known as weakness of will, and in order to save us from being

   "Sartrean persons" (where everything is always up for grabs at all

   times), there must be states of bare intention. Such states execute

   our background motivational and cognitive states into actions, do

   so even when our desires or beliefs are in conflict, account for our

   failures to execute such beliefs or desires (as in cases of akrasia, or

   weakness of will), and project our resolutions in these matters into

   the future so that we need not redecide such matters again and

   again. Volitions fit into this executory role of bare intentions very

   neatly, for they are simply such bare intentions having as their

   objects the simplest bits of bodily motion that we know how to do.

   Volitions are simply the last executors both of our more general

   intentions and of the background states of desire and belief that

   those more general intentions themselves execute.(6)

   This conception is developed by Moore both in the course of answering various objections to volitionalism and in his development of a positive characterization of volition. Among the more important objections are (1) the claim (which Moore takes to be implicit in theses of Michael Bratman's)(7) that, unlike intentions, volitions can be directed toward actions which one does not take to be possible,(8) (2) the view that, as intermediate causes, volitions are ripe candidates for elimination,(9) and (3) the psychological contention that we do not experience all the willing we should experience if volition, conceived as a kind of willing, plays the pervasive role in action that Moore assigns to it.(10)

   Let us start with the rationality objection. To Bratman's contention that intentions must satisfy rationality constraints that volitions do not, Moore replies that not all intentions satisfy the constraints in question: "I can intend to hit a target even though I know I can't hit it; and I can rationally intend to hit target I and intend to hit target 2, even though I do not intend to hit both targets I and 2."(11) As for the worry that volitions, as intermediate causes, are needless posits--which I shall call the eliminability objection--Moore responds, in part, that "it is largely the inability of states of belief and desire to rationalize our acts that lead[s] us to posit the executory state of volition."(12) And to the objection that we are insufficiently conscious of volitions to warrant giving them the pervasive role volitionalism assigns to them--which I shall call the phenomenotogical objection--Moore replies that while "[u]ndeniably we learn to string together various of our bodily movements into complex routines with such dexterity that, once we have mastered the routine, we can literally not pay attention to what we are doing (at the level of bodily movements)."(13) Still, given

   that we seek phenomenal clues to the nature of those mental states

   that execute desires into action, a good place to look for such

   clues is at the mental states we had to acquire in order to acquire

   our various skills of action. The objects of those states were bodily

   motions. It is a reasonable (although far from inevitable)

   hypothesis that those same states exist to cause those same

   motions when they occur later in life as part of speaking or

   playing, even though awareness of those states has receded.(14)

   In laying out a positive case for volitionalism, Moore makes a number of points, only some of which I propose to consider. A central one is

   that the basic thing that gives rise to the inference (that there must

   be volitions) is the existence of human actions as a natural kind.

   Our sense has long been that |something' is indeed |left over' if we

   subtract the fact that our arm rises from the fact that we raised

   our arm, to paraphrase Wittgenstein's famous observation.(15) If we call the problem Wittgenstein articulated the subtraction problem, then the point here is, in part, that this problem is not soluble without positing volitions as the crucial differentia distinguishing action from bodily movement.(16)

   A second central point is that we cannot account for the nature of practical reasoning without positing volition as its concluding element. Specifically, "we must conceptualize this concluding propositional attitude as a bare intention, not as a belief (say, that one will go downtown) and not as a desire (say, to go downtown)."(17) One might think that motivating desires, means-end beliefs, and desires to perform the action would suffice. But that hypothesis will not do. "The problem with this attempt to conceptualize concluding propositional attitudes as being the strongest competing prima-facie desires to do an act is that it leaves out the resolving function of such concluding propositional attitudes when there is (as there always is) conflict."(18) These two points are developed and defended in later sections of the book,(19) but there is one more issue that should be noted before beginning an assessment of Moore's case for volitionalism.

   It is well known that any causal theory of action must reckon with deviant causal chains--that is, strange, wayward sequences running from desires and beliefs of the kind that usually produce intentional action, but do not do so in the case in question because of the way in which they cause the relevant bodily (or, in principle, mental) motion. A machine, for instance, might cause my arm to go up just as it would have if I were signaling a question in a lecture, but only because my relevant desires and beliefs triggered the machine's manipulations, and not because I in any sense willed to raise my hand. I...