Philosophia DOI 10.1007/s11406-013-9414-9 From Volitionalism to the Dual Aspect Theory of Action Joshua Stuchlik Received: 25 July 2012 / Revised: 19 October 2012 / Accepted: 7 January 2013 # Springer Science+Business Media Dordrecht 2013 Abstract Volitionalism is a theory of action motivated by certain shortcomings in the standard causal theory of action. However, volitionalism is vulnerable to the objection that it distorts the phenomenology of embodied agency. Arguments for volitionalism typically proceed by attempting to establish three claims: (1) that whenever an agent acts, she tries or wills to act, (2) that it is possible for volitions to occur even in the absence of bodily movement, and (3) that in cases of successful bodily actions the relation between volition and bodily movement is causal. I defend an argument for the second of these claims from an objection by Thor Grünbaum, but I show that several volitionalist arguments for the third are not compelling. I then argue that the dual aspect theory of action provides a better account of the relationship between an agent’s volition and the bodily movements she makes when she acts, insofar as it has the same advantages over the standard story as volitionalism without being open to the phenomenological objection. I also defend the dual aspect theory from an objection by A.D. Smith. Finally, I show why the dual aspect theory of action is a better alternative to volitionalism than the theory of action recently put forward by Adrian Haddock. In order to avoid the phenomenological objection Haddock suggests a disjunctive account of bodily movements. While disjunctivism should be taken seriously in the philosophy of action, on the dual aspect theory it is the category of volition, rather than bodily movement, that should receive a disjunctive analysis. Keywords Action . Volitionalism . Trying . Will . Dual aspect theory A central goal of contemporary philosophy of action is to solve what Harry Frankfurt has deemed “the problem of action”, which is to explicate “the contrast between what an agent does and what merely happens to him” (1978, 157). The most influential solution to this problem is one inspired by Donald Davidson’s essays on agency. On this view, which J. David Velleman can simply refer to as “the standard story”, J. Stuchlik (*) Department of Philosophy, University of St. Thomas, Saint Paul, MN, USA e-mail: joshua.stuchlik@gmail.com Philosophia actions are identical to bodily movements that are caused in the right way by intentions, which are themselves effects of the agent’s beliefs and desires (1992, 123).1 Of course, human agents can truly be said to do many other sorts of things besides move their body. But a proponent of the standard story does not deny this truism. She argues instead that the truism is true because an agent’s actions are also describable by reference to their effects (cf. Davidson 1971). The standard story has given rise to a vast literature. Here I will briefly rehearse just three objections to it that have been raised therein, as they will bring us to the main topic of this paper. The first is that the story seems to leave out the role that the agent plays in the production of her bodily movements. We do not simply sit passively by while our desires combine with our beliefs to produce intentions and bodily movements. As agents we reflect on what we take to be our reasons for acting, choose which to act on, and carry out our choice by producing the appropriate movements (cf. Velleman 1992, 123). Second, it seems that causation by the sorts of psychological states that figure in the standard story is insufficient to make a bodily movement into an action. The story fails to include a crucial psychological element in its account, an element that can be categorized under the headings “trying”, “willing” or “volition”.2 This additional element is necessary because even when an action can be causally explained in terms of an intention, the act in question will not be performed unless the agent, at the time of acting, tries to do what is necessary to execute her intention. As John Searle puts it, in paradigm cases of action there appears to be a causal “gap” between one’s intention and one’s activity that must be filled by an act of will (2001, ch. 3). And for any action that takes place over a period of time, the action will not continue to its completion unless the agent continues trying to do it.3 Finally, the sort of psychological etiology posited by the standard story does not appear to be a necessary condition for action either. I have in mind those actions that Brian O’Shaughnessy (1980:2) refers to as “sub-intentional”. I may be listening to a long lecture barely conscious of the fact that I am also fidgeting with my watch. The movements I make in so doing do not seem to execute any intention of mine, yet their production is still attributable to me as their agent. While none of these objections are decisive, together they make a powerful case for taking volitions seriously.4 If the first is correct the standard story fails to include the causal role the agent plays as mediator between her psychological states and her 1 In the remainder of this paper I will use “action” to refer to bodily actions. This is not to deny the reality or importance of purely mental actions, but my central topic here will be the relation between an agent’s act of will and the movements that her body makes in cases of bodily action. The topic will also be restricted to the actions of rational human agents. Non-human animals also perform actions, but this does not necessitate that the same account of agency will be applicable to both rational and non-rational animals. 2 I’ll be using these terms interchangeably, though some theorists (e.g., Lowe 2000, 248; McCann 1975, 108) distinguish between trying and willing. 3 Searle argues that there is also a gap between an agent’s beliefs and desires and her intentions, on the ground that, at least in non-pathological cases, beliefs and desires do not necessitate the formation of an intention. For a fuller description of this objection to the standard story, along with more on the role that volition might play in filling these causal gaps, see Zhu 2004, 248-55. 4 For example, Bishop (1989), Velleman (1992), and Enç (2003), are sensitive to the first objection and think it is possible to overcome it while avoiding the postulation of volitions. I cannot delve further into their proposals here, but I think they are ultimately unsuccessful. Philosophia bodily movements. But a natural thought is that the agent plays this role precisely through exercising her will, which is expressed in acts of volition. And once volition is on the scene it appears to be the essential psychological element in action. For when an action is not explainable in terms of an intention caused by appropriate beliefs and desires, as with sub-intentional actions, the movements that one makes are nonetheless voluntary and volitions can explain why this is so. And when an action is explainable in terms of an agent’s beliefs, desires, and intentions volition is still necessary to fill the causal gaps between intention and the initiation of action and between the initiation of action and its continuation. Yet if we do take volitions seriously we face two major questions. First, what is the relation that volitions bear to the bodily movements that occur when one acts? Second, how are volitions related to actions; e.g., is a volition the cause of an action, a part of it, or is an agent’s action identical to her volition? Volitionalists answer the first of these questions by maintaining that the relation between volition and bodily movement is causal. According to them whenever an agent successfully performs an action the bodily movements that occur are caused by a volition that is distinct from those movements. Volitionalism as such does not imply any single answer to the second of the two questions. However, there is a sub-set of volitionalists, which I will refer to as strong volitionalists, who claim that an agent’s actions are to be identified with the acts of volition that cause her bodily movements. Strong volitionalism has been defended by Prichard (1945), McCann (1974, 1975), Hornsby (1980), Smith (1988), and Pietroski (1998). Philosophers who accept volitionalism without endorsing strong volitionalism include Sellars (1976), Davis (1979), Armstrong (1981), McGinn (1982), Ginet (1990), Widerker (1995), Searle (2001), Zhu (2004), and Lowe (2000; 2008). Though volitionalism marks an advance over the standard story it is vulnerable to a serious objection, namely, that it entails a distorted picture of the phenomenology of embodied agency. My goal in this paper is therefore to put forward an account of action that avoids this objection while still retaining a place for the activity of the will in action. In the remainder of this paper I proceed as follows. In §1 I separate out a number of claims typically made by volitionalists and I elaborate on the phenomenological objection. In §2 I defend the claim that there can occur acts of volition in the absence of any movements of the body from a recent objection by Thor Grünbaum. The volitionalist thesis that is targeted by the phenomenological objection, however, is not that volitions can occur apart from bodily movements but that when an agent is successful in moving her body the relation between volition and bodily movement is causal. It is this latter claim that I think we should reject. Volitionalists have put forward several arguments for the causal thesis and my goal in §3 will be to explain how one can resist these arguments. An argument by A.D. Smith and Paul Pietroski receives particular attention. In §4 I then propose that a more satisfactory account of the relation between volition and bodily movement is provided by the dual aspect theory of action, according to which the relation at issue is one of whole to part and not cause to effect. I also respond to an objection to the dual aspect theory by Smith. Finally, in §5 I argue that the dual aspect theory constitutes a better alternative to volitionalism than the theory of action recently put forth by Adrian Haddock. Haddock is sensitive to the phenomenological objection to volitionalism, but he Philosophia wants to avoid it by identifying actions with bodily movements of a certain sort. Haddock’s central claim is that we should adopt a disjunctive conception of bodily movements, which parallels a disjunctive conception of perceptual experience. While I think there is room for disjunctivism in action theory, it is the category of volition rather than bodily movement that should be conceived of disjunctively. Volitionalism and the Phenomenological Objection Volitionalists claim that when agents act the bodily movements that occur are caused by acts of volition. Arguments for volitionalism typically proceed by attempting to establish three claims: Ubiquity: whenever an agent moves her body (where the moving in question is an action), she tries or wills to move it. Absence: it is possible for an agent to try or will to move her body without the occurrence of any of the bodily movements she tries or wills to make. Causation: when an agent succeeds in moving her body, the movement of her body is caused by her trying or willing to move it. If Ubiquity is true then volition is involved in every case of action. But by itself this does not yield any volitionalist conclusions. One cannot even derive from it that when an agent moves her body there ever occur acts of volition that are distinct from her bodily movements, as it is compatible with it that volitions simply are bodily movements. That is why Absence is important. It implies that volitions cannot be generally identified with bodily movements, since there are cases in which the former occur while the latter do not. This goes further towards establishing volitionalism, but it still does not entail it. The volitionalist also needs a substantive argument for the claim that in cases of successful action the relation that occurs between trying and bodily movement is causal (Causation).5 Why should we resist volitionalism? The answer is that it constitutes a distorted picture of what it is like to be an embodied agent. Haddock raises this objection against strong volitionalism in particular. He argues that strong volitionalism alienates agents from their bodies, since according to it, “Our bodies are pictured as entities whose powers are wholly distinct from our powers of agency, as entities that we can (at best) only cause to move—and in this respect they are the same as any other worldly object” (2005, 161).6 Consider, by way of analogy, what happens when someone grasps items by operating a metal claw. The agent has the power to grasp things with the claw, and she has this power in virtue of the fact that she does something else, such as manipulating a controller, which causes the claw to exercise its capacity to grasp. If Haddock’s objection is on target, then according to strong volitionalism the relation between an agent and her body is analogous to that between 5 I disagree with Grünbaum about this last point. He believes that given “standard assumptions about causality and explanation” Ubiquity and Absence entail volitionalism (2008, 68). My goal in §3 will be to show that the assumptions about causality that volitionalist arguments for Causation rely on are dubious. 6 See also Morris 1988, 485-6. Philosophia the agent and the claw. This is problematic insofar as the latter is a paradigm case of a relation between an agent and an instrument or tool, and it is phenomenologically offkey to think of our bodies as related to us in this manner. The strong volitionalist may have room for a response to Haddock’s way of stating the objection. In the example above, the agent intentionally does one thing (moving the claw) by intentionally doing another (moving her body). If this is really analogous to bodily agency on the strong volitionalist’s theory, then she would be committed to the proposition that every time someone intentionally moves her body, she does this by intentionally doing something else—presumably, by intentionally trying to move it. But while strong volitionalists propose that actions are acts of trying, it is not clear whether they are committed to saying that every time an agent moves her body her action is intentional under the description “trying to move”. Yet even if Haddock’s version of the phenomenological objection can be evaded, I believe that there is a better way of formulating it, one that is, moreover, applicable to volitionalism as such. Whether trying is done intentionally or not, all volitionalists are committed to the idea that an agent’s exercise of her will consists in an act of volition, and that this act comprises an event that is separate from, and causally related to, the movements of her body. Therefore, an agent’s bodily movements are only externally related to the actualization of the power that is the locus of her agency. This is objectionable on phenomenological grounds, though, for when we perform bodily activities we experience our agency as being, in part, in our bodily movements.7 Nor is the problem overcome by insisting that we ought not to conceive of volition as a mental trigger that precedes the movements of the body in time, but rather as a continuous activity that guides them throughout their occurrence (cf. Ginet 1990, 334; McCann 1986, 140). That the will plays an ongoing role in bodily action is certainly correct. But the objection was not that volitionalists make volition temporally prior to bodily movement but that they make it causally prior. Even if the agent’s volition guides the movements of her body throughout their duration this does not change the fact that on the volitionalist’s picture it supervises these movements, so to speak, from the outside. Prima facie it may seem that the most attractive response to volitionalism is simply to deny Ubiquity. After all, we would not normally say of someone that he tried to do something unless either he was unsuccessful or he needed, or expected that he would need, to make some special effort to do it. This argument is quite weak, however. It is a familiar Gricean point that whether it is conversationally appropriate to utter a proposition in a given context is determined by more than just its truth-condition, since some true propositions have non-semantic implications that are misleading in that context (Grice 1975). And it is plausible that if someone actually succeeds in moving his body, then he must have attempted to move it, though it may often be misleading to point this out if the agent made the movement without difficulty.8 7 I believe Brewer has something similar in mind when he complains that on the volitionalist’s picture we have a mental event of willing causing a biomechanical event of bodily movement, but neither of these is essentially an “active bodily movement” (1993, 305). 8 For a further development of this line of argument see O’Shaughnessy (1973) and Hornsby (1980, 37-8). Philosophia For this reason, I think that Ubiquity should be the least controversial of the three claims advanced by volitionalists, and I will accept it here.9 Absence and Causation have proved more contentious, and I will focus on them in the next two sections. Absence Arguments for Absence are designed to show that there exist “naked tryings”—acts of volition that occur independently of any bodily movements. Since such arguments focus on cases in which one completely fails in an attempt to move one’s body, Grünbaum has entitled them “arguments from total failure”. Here is one such argument.10 Suppose John is recovering from paralysis of the arm, and is uncertain about whether he has yet regained the use of it. His physician wants to test whether he has, so she orders John to try and raise it. It seems plausible that John can comply with the physician’s order even if, as things turn out, the limb is still completely paralyzed. If after a few moments the physician were to ask John whether he tried to raise his arm, he would answer affirmatively, and both John and his physician would learn that John’s arm was still paralyzed precisely because he tried to raise it and was unsuccessful (cf. McCann 1975, 99). This scenario appears to show that an agent can try to move his body even in the complete absence of the appropriate bodily movements. But in a recent paper Grünbaum questions how well we understand the scenario, by putting pressure on the idea that the trying in a case of paralysis is the same sort of event as the trying found in a case of successful action. In particular, he issues a challenge to the volitionalist (and others who hold Absence) to show how we can make sense of a distinction between two different sorts of mental activity in cases of paralysis, which he deems “effective trying” and “idle trying” (2008, 72). Grünbaum variously identifies idle trying with episodes of thinking about doing something, imagining doing it, or wishing to do it.11 Effective trying goes beyond all of these, and involves an actual striving to achieve some end. It is this latter sort of trying that is associated with volition. One way to accommodate the difference between idle trying and effective trying is to say that the latter only occurs when an agent’s body actually moves. But in the case of paralysis we must be able to draw the distinction between the two sorts of episode where no bodily movements occur. Grünbaum argues that his opponent could either affirm or deny that the difference between idle trying and effective trying is an introspectibly reportable one. Denying this is not an attractive option, however. After all, the argument for naked trying above depends on the paralytic’s being able to truthfully tell that he has tried but failed to raise his arm. But if the difference is introspectibly reportable then it must be possible to specify what this difference consists in, and Grünbaum doubts this can be done. The difference 9 For a more skeptical take on Ubiquity see Morris (1988, 475-6). Another argument for Absence is discussed by Hornsby (1980, 41) and Grünbaum (2008, 75-81). 11 I think it is strange to call this sort of mental episode a form of trying at all, but I will continue to use Grünbaum’s terminology. 10 Philosophia cannot consist in the representational content of the two sorts of episode, since the same representational content (whether it be propositional or imagistic) could be the content of either. Nor does Grünbaum think the distinction is explainable in terms of a difference in a non-representational “raw feel”, as it can plausibly be maintained that there is no special raw feel essentially associated with effective trying. So the proponent of Absence faces the challenge of specifying some other way in which the distinction between effective trying and idle trying can be drawn in a case of paralysis (ibid., 73). I believe that this challenge can be met. The key lies in the fact that, in addition to having a representational content, many intentional mental episodes and states also involve a certain sort of stance or attitude towards that content. Different sorts of attitude can thus serve to distinguish between different sorts of mental episodes or states that have the same content. Consider, for instance, the difference between believing that p and merely entertaining the proposition that p. Both of these sorts of states have the same representational content, the proposition that p. It is also implausible that a distinctive raw feel essentially accompanies either. The difference between them is rather that belief that p involves an element of affirmation or assent on the subject’s part that merely entertaining the proposition that p does not. Someone who believes that p is committed to p’s being true, a commitment that is constituted by the person’s affirming that what the proposition represents as so is so. Because belief involves this sort of affirmation or assent its role in one’s conscious deliberations is different than that of entertaining a proposition. But the fact that belief plays such a role in deliberation does not preclude it from being immediately reportable. In fact, belief can play the role it does in virtue of the fact that when a person affirms that p is true that fact is selfconsciously available to her.12 Like belief, volition (effective trying) involves a certain kind of affirmation or assent, and this element distinguishes it from mere wishing (idle trying). Both volition and wish can be directed at representations of possible types of action. But while a mere wish may involve desire and thought of a possible act-type as good, volition goes beyond these insofar as it also essentially involves an act of assenting to the performance of an act of the relevant type in a way that mere wish does not. Borrowing a phrase from Kant, we might say this sort of present-directed assent to act is constituted by “the summoning of all means within our power” to do A or to continue A-ing (Kant 1998, 50). In this respect, volition is similar to intention, but whereas intention is an affirmation of, or assent to, a future act, volition is directed towards one’s present activity. Some theorists distinguish between intention for the future and intention in action, and it may be thought that the introduction of intentions in action creates a problem for this account of volition. While there are different ways of conceiving of the relevant distinction, according to one such conception intentions for the future have the representational content “I shall do A at t” where t refers to some time in the 12 While there is much debate over the correct theoretical account of the self-conscious availability of beliefs, there can be little doubt that the phenomenon itself exists. For recent work on the topic see Moran (2001), Bar-On (2005), Shah and Velleman (2005), and Finkelstein (2008). Philosophia future, whereas intentions in action have the content “I shall do A now”.13 The problem now is supposed to be that present-directed assent to an action can be fully explained in terms of intention without the need to invoke anything like volition. However, I do not think that intention, thus conceived, can fully account for the sort of assent to acting that is constituted by volition. An intention in action is the subject’s assent to begin to act, and takes effect at the moment it is finished being formed. But when an agent acts or tries to act, she typically does so for some time, and the intention to perform the relevant type of act must continue to be executed by the will throughout this duration and not just at the beginning of the attempt. Here is how Ginet puts the point: “Volitions do not plan ahead, not even very slightly. They do not plan at all; they execute. I have an intention as to what course of movement my body is to make over the next few moments, and in light of that intention I go through a certain course of volitional activity, of voluntary exertion, over the whole period of the movement…” (1990, 33, italics in original). Given this characterization of volition it seems entirely possible to distinguish between it and a case of idle trying even in a case of paralysis. The paralyzed person may only wish, in the sense of think it would be a good thing, that he raise his arm, but it seems that he can also assent to raising his arm. So long as he believes that he has some means to act within his power, or at least believes that he may have some means, then he can summon them even if, as it turns out, there are actually no means to respond to the summoning. Of course, the paralytic may only form the intention in action to raise his arm now and desist immediately thereafter. But he may continue to give his assent for some time (if only for a few moments) until it becomes apparent that the arm is still immobile. And the fact that volition is a species of affirmation like belief and intention should lead us to expect that it is self-consciously available to its subject. Causation Grünbaum’s challenge to Absence can be met. Since the argument in its favor I mentioned at the beginning of the previous section is plausible, I conclude that we are provisionally justified in accepting it. However, in order to establish her position the volitionalist must also justify Causation. McCann seems to think that Causation can be inferred directly from the observation that the difference between cases in which an agent tries to act and fails and cases in which her trying succeeds is one of whether things are allowed to take their normal course (1975, 108-9). The observation is undeniably true but the conclusion McCann draws from it involves a principle of inference that begs the question. That “things 13 For this account of the relation between intention for the future and intention in action see Brandom (1998, 256-9). There are other ways of conceiving of the distinction. For example, McDowell (2011) refers to what I here call intention in action as the onset of an intention in action, and uses the term intention in action to refer to the entire activity of willing, which is what I call volition. As far as I can see this is merely a terminological difference. If one prefers this alternative terminology, then it will be true that intention can do the explanatory work I attribute to volition, but one can then re-formulate the question of this section by asking whether there can be a “naked” intention in action (and, more generally, one can understand this paper as inquiring into the nature of the relationship between intention in action and bodily movement). Philosophia taking their normal course” should be understood in terms of causation of bodily movement by an act of trying is an assumption that simply amounts to the claim that needs to be proved. Another argument for Causation begins by noting that whenever an agent successfully acts there is a true proposition asserting an explanatory relationship between the movement that she makes and a volition. For example, if Fred raises his arm, then Fred’s arm rose because he tried to raise it, and if Ed stomps his foot, his leg and foot moved because he willed them to move. The argument is that the truth of such propositions commits us to the claim that in humdrum cases of action like these acts of volition cause bodily movement, which is strong evidence for Causation.14 This argument relies on the assumption that propositions of the form: (a) S’s body moved because S tried to move it/willed to move it. imply propositions of the form: (b) There exists an event e1 of S’s trying/willing to move her body and an event e2 of S’s bodily movement and e1 causes e2. But while propositions of form (a) do invoke an explanatory relationship, it is one that holds not between two events but between two facts. The problem for the argument at issue is that, in general, one cannot assume that whenever a certain fact explains another the explanation is underwritten by the existence of two causally related events. For instance, it may be true that pairs of chromosomes in a certain cell are moving apart because the cell is undergoing mitosis. But this explanation does not imply that cell mitosis and the movement of the chromosomes are two causally related events (the explanation is true but the supposed implication is false).15 So while propositions of form (a) assert that the fact that S’s body moves is explanatorily dependent on the fact that S wills to move it, an additional step would be required to show that in the case of action propositions of form (b) capture the deep structure of propositions of form (a). A stronger argument for Causation is defended by A.D. Smith and Paul Pietroski (Smith 1988, §3 Pietroski 1998, 97-8. I focus on Smith’s fuller presentation of the argument here). Smith connects volitionalism with what he refers to as Anscombe and Davidson’s approach to action individuation, contending that strong volitionalism is its “logical conclusion” (409). I say “what he refers to as” Anscombe and Davidson’s approach because though these philosophers are often thought to speak with one voice on the topic of action individuation there is actually an important difference between them that is relevant in this context. What they share is the belief that we often refer to a single action under many specifications, and that some of these specifications describe it by reference to its effects. However, Davidson claims that descriptions of an action that do not specify it as an agent’s making bodily movements always describe it by reference to an effect, whereas Anscombe does not commit herself to any such proposition. This is noteworthy because Smith includes 14 Grünbaum (2008, 70-1) mentions this argument and attributes it to O’Shaughnessy (1973). This example is especially relevant because my own view will be that the relation between volition and bodily movement in a case of successful action is similar to the one that holds between mitosis and the movement of chromosomes. 15 Philosophia the additional Davidsonian claim in his description of the approach to action individuation he wishes to push to its logical conclusion. What is plausible is that although some of the act-specifications that are true of a certain action specify it by reference to an effect, not all do. Davidson goes beyond this, however, insofar as he takes it that there is a certain kind of act-specification that describes actions as they are in themselves, namely, the kind that describes them in terms of agents moving their bodies.16 One reason for this turns on the conceptual truth that an action is essentially an agent’s activity together with certain temporal considerations. Consider someone’s turning on the light. If her action would have continued to occur after she moved her finger, this would entail by the conceptual truth that her active contribution would have continued after she moved her finger. But after she moved her finger it was unnecessary for her to do anything else; she’d made her contribution. Whether or not the switch was flicked and the light went on was not a matter of her activity but of the obtaining of certain conditions in the mechanism of the switch and the electrical circuit. Davidson puts the point by saying that everything that happens after we move our body is “up to nature” (1971, 59). Smith argues that Davidson is wrong to think that in describing an action as an agent’s moving her body we are not describing it by reference to an effect. According to Smith what distinguishes those specifications that characterize action by reference to an effect is that their truth depends on the obtaining of a condition that is independent of the agent’s being active. A condition is “independent” in the relevant sense just in case its failure to obtain would not imply that the agent was not active (1988, 405). However, Smith also thinks that to describe an action as an agent’s moving her body is still to describe it in a way that is dependent on the obtaining of an independent condition. For instance, when I move my finger, it moves in part because certain causal connections hold between the finger and the muscles whose contractions cause it to move, and between those muscles and my nervous system. These causal connections are independent conditions because even if, unbeknownst to me, they had failed to hold I would have still been active insofar as I would have still tried to move my finger.17 So just as an act of flicking a switch might not have been an act of turning on the light if certain conditions in the circuit had failed to hold, so too my trying to move my finger might have failed to be an act of moving my finger had certain conditions in my muscular or nervous system failed to hold. Smith’s argument hangs on the following principle: Effect: Act-specifications characterize an action by reference to an effect when their truth depends on the obtaining of an independent condition, where a condition is independent if and only if its failure to obtain would not imply that the agent was not active. He believes this principle shows that Davidson is right to think that all actspecifications that go beyond describing actions as the making of bodily movements describe them by reference to effects of those movements. What Davidson failed to 16 In fact, he also makes the stronger claim that every physical action is identical with a bodily movement (1987, 102). I argue against the idea that actions are identical to bodily movements in §§4-5. 17 Smith’s argument thus relies on Absence, since it assumes that one can try to move one’s body even if no appropriate bodily movements occur. Philosophia realize, on his view, is that the same principle also entails that to specify an action as the making of a bodily movement is to specify it by reference to an effect. To parrot Davidson: all we have to do is try, the rest is up to nature. But even if we bracket Smith’s distinctive application of this principle, there are counterexamples to using it even to go as far as Davidson’s conclusion. Consider, for example, Jenny’s act of carrying a backpack. The description of Jenny’s action as that of carrying a backpack depends on the obtaining of an independent condition, namely, the pack’s being slung over her shoulders. The condition is independent in Smith’s sense because had the pack failed to be slung over Jenny’s shoulders, she still would have been active, insofar as she still would have been walking. But it is not the case that the backpack’s getting carried is an effect of Jenny’s walking. The backpack’s getting carried is not a different event from Jenny’s walking, but is what that walking amounts to when the backpack in slung over her shoulders. So the relation between the act of Jenny’s walking and the event of the backpack’s getting carried is a constitutive relationship and not a causal one.18 Therefore, Effect is false. I want to be careful about stating exactly what I think this case establishes. It does not show that whenever there is a constitutive relationship between one happening or state and another happening or state, then they are not also related causally. One of Davidson’s main arguments against opponents of the view that rationalizing explanations of action are causal explanations involved pointing out that there are cases in which one happening or state and another are related both constitutively and causally (1963, 14). We may grant Davidson’s point that such a constitutive relation between happenings and states does not entail the absence of a causal one. Still, it seems evident that in the case described above the relationship between Jenny’s carrying the backpack and the backpack’s being carried is constitutive and not causal. So, there are at least three ways in which an action and a happening or state that it is describable by reference to can be related. The relation may be (1) causal and nonconstitutive, (2) both causal and constitutive, or (3) constitutive and non-causal. What the case of Jenny’s carrying the backpack shows is that (3) is a possibility that is sometimes actualized and that it may be actualized even when the characterization of an action depends on the obtaining of an independent condition. And this is sufficient to render Smith’s argument for Causation unsound. Dual Aspect Theory In the previous section I argued that while it is true that when one acts the movement of one’s body depends on the obtaining of an independent condition, this does not show that the bodily movement is an effect of one’s action. But my reply still leaves the following question unanswered: When an agent acts what are the relationships between volition, the events that take place in her nervous and muscular systems, and her bodily movements? I think that the counterexample I raised to Effect above suggests an answer to these questions that is different from the volitionalist one and that avoids the phenomenological 18 Similar cases are discussed in Hornsby 2011. Philosophia objection from §1. On this account we agree that an action is an agent’s activity and that an agent is active insofar as she wills or tries to do something (Ubiquity). We may also accept that it is possible for an agent to try to act even if she is completely unsuccessful in moving her body (Absence). But we avoid volitionalism by asserting that when an agent does succeed in making a bodily movement, the connection between her activity and the events that occur in her muscles, her nerves, and her bodily movements is constitutive and not causal in character (therein denying Causation). The account I am describing is what is known as the dual aspect theory of action. The dual aspect theory holds that actions, which are processes that sometimes include bodily movements, can also be characterized as psychological acts of volition. Here is how O’Shaughnessy summarizes it: Each enlarging sector of the motor process, which falls already under “try”, “strive”, “do”, and “will”, proceeds to cause the next stage of the process, and finally to cause the (so to say) crowning event of (say) arm movement. Then although whatever slice of the process one likes to select falls already under the concept “willing”…this is not to say that the event of willing causes arm movement. Rather, the theory claims that the event of willing physically develops, in naturally appointed causal manner, to the point at which it incorporates the event of limb movement, and completes itself in so doing. (2003, 346) According to the dual aspect theory an action, which is an agent’s trying or willing to do something, is a process that is extended in time. As such, it is composed of various other events and processes that, having been bound together into a temporally unified whole, constitute the parts or stages of that whole. Earlier stages of this process involve the occurrence of events in the motor cortex and its later stages include events in the peripheral nervous system, contractions of muscles, and finally, the movement of certain parts of the body. It is crucial to understanding this theory that we distinguish the relation that obtains between an agent’s action and her bodily movements from the relation that obtains between the sub-personal events that occur in her nervous and muscular systems when she acts and her bodily movements. The latter relation is a causal one, but the former is not. An action is not the cause of the movement of the agent’s body. Rather, the movement is the final part or stage of the action whose earlier stages include events in her motor cortex, peripheral nerves, and muscles. A central claim of the dual aspect theory is that when an agent succeeds in acting her action is a single process that has as some of its parts events that take place in her brain and in her nervous and muscular systems. The theory therefore immediately gives rise to the question of what accounts for the unity of this process. What makes it the case that the sub-personal events that occur when one acts together compose a temporally unified whole rather than forming a mere chain or sequence of events? While I can provide only a sketch of an answer to this question here, I submit that the parts of an action are united as elements of a single teleological order. One area where cases of teleologically unified processes can readily be found is in the realm of intentional human activity.19 Suppose, for instance, that Peter is building 19 Michael Thompson has recently discussed this sort of teleological unity under the heading “naïve action explanation” (2008, 90-1). Philosophia a house, and he goes about this activity by doing things such as laying the foundation, building the frame, putting up walls, and so on. Here laying the foundation, building the frame, etc. are all things that Peter is doing and are parts of the more encompassing activity in which he is engaged, namely, building the house. Furthermore, each of these parts may themselves have parts, e.g., Peter may be laying the foundation by mixing the concrete and pouring it out in the relevant areas. So a single thing that Peter is doing is composed of certain other things that he is doing, and these activities may themselves be composed of further activities. And here it is clear that what makes it the case that a less encompassing activity is a part of a more encompassing one is that it is directed at or for the sake of the end of the more encompassing activity. While many human activities are good examples of teleologically constituted unities, they do not provide the best model for understanding the unity that obtains among the sub-personal processes that occur when someone acts. The reason is that what makes it the case that a certain activity is ordered to the end of a more encompassing one is the fact that the agent who is engaged in them represents them as being so ordered. For example, Peter’s act of pouring concrete is for the sake of the end of his building a house because he conceives of pouring concrete as a means thereto. But only in extremely rare cases does an agent represent to herself the activities that occur in her brain, nerves, and muscles when she acts. (She may not even be aware that her body possesses these parts.) There are other teleological processes that can better serve as a model, however. I am referring to certain sorts of vital processes. Digestion is one instance of the relevant type of process. It might be referred to as a “natural kind” of process in the same way that species are natural kinds of substances. Referring to it this way highlights the fact that instances of digestion are naturally unified processes just as members of species are naturally unified substances. And this is so even though it is composed of further sub-processes, such as the actions of certain enzymes upon ingested food. Moreover, as in the case of human activities, the reason that these less encompassing sub-processes are stages of the more encompassing processes of digestion is that when they occur they do so for the sake of the end of digestion, getting food into a state suitable for absorption by the body.20 I suggest that the sort of unity that obtains between the sub-personal events that constitute an action is comparable to the sort that holds between the stages of naturally unified vital processes. In fact, it seems that action is simply another instance of such a process. For just as it belongs to our nature as human beings to have a power to nourish ourselves through digestion, it is also part of this nature to have the power to respond to our environment by moving our bodies, and it is through making bodily movements that we are able to fulfill many of the ends and needs characteristic of our species. So when Jane tries to raise her arm and certain muscles contract, their contractions are no accident: they are for the sake of the end of her act of raising her arm, viz., her arm’s being in a certain position. And more generally, the events that take place in an agent’s nervous and muscular systems when she tries to move her body 20 For recent accounts of teleological processes, see Stout 1996, ch. 3 and Enç 2003, 104-11. Philosophia are directed at the state or event whose existence would make it the case that her trying is successful. But that they are so directed is not due to a representation on the part of the agent.21 Insofar as the dual aspect theory makes volition an essential element of action it shares the advantages of volitionalism over the standard story.22 But the dual aspect theory has these advantages without being vulnerable to the phenomenological objection. Recall the objection was that volitionalism cannot do justice to the fact that we experience our agency as being in the bodily movements we make when we act. It cannot do justice to this fact because according to it the activity of the will is an event that is separate from, and causally prior to, bodily movement. It should be evident why the dual aspect theory is immune to this objection. On this view, when an agent succeeds in moving her body her volition does not cause her bodily movement but includes it as a part. The movements we make when we act are internal to the exercise of our will. A.D. Smith has objected to O’Shaughnessy’s version of the dual aspect theory, though, on the ground that it has the implausible consequence that being a trying is an inessential property of events (1988, 417-8).23 Smith notes that O’Shaughnessy thinks that an action (a successful trying) is conceptually divisible into two parts: a purely physical bodily movement and a “non-autonomous psychological part” which is what remains of the action if the bodily movement is subtracted and from whence the psychological character of the action is derived (cf. O’Shaughnessy 1980:2, 21012). Smith contends that on O’Shaughnessy’s view we should conceive of the nonautonomous psychological part of the action as causing the agent’s bodily movement in the case of a successful trying. And if this is correct, then it leads to the implausible consequence mentioned above. Consider an unsuccessful trying T that fails to include a bodily movement because a certain enabling circumstance C fails to obtain. If C had obtained, then what we picked out as T in the actual world would have caused a bodily movement and thus would have been a non-autonomous psychological part of a successful trying T´ of which the bodily movement was a physical part. But since non-autonomous parts of tryings are not themselves tryings, whether some process is a trying or merely a non-autonomous psychological part of one is dependent on the circumstances. Here is a response to this objection on O’Shaughnessy’s behalf. He need not, and should not, agree to the assumption made by Smith that in a case of a successful trying the movement of the agent’s body is caused by the non-autonomous psychological part of the action. The non-autonomous psychological part, by definition, consists in the entire process that precedes the bodily movement. But the bodily movement is caused only by the stage of the process that is immediately prior to it. E.g., when I move my toe, the movement is caused by the contractions of certain muscles in my foot, not by the entire neural-muscular process that precedes it. So, we should not say that, in better circumstances, the process we pick out as an 21 The idea of a non-represented teleological process provides an interpretation of O’Shaughnessy’s idea in the quoted paragraph above that “willing physically develops, in naturally appointed causal manner”. 22 See pp. 2-3 above. 23 Moreover, given that O’Shaughnessy identifies actions and tryings, Smith thinks O’Shaughnessy’s theory also has the implausible implication that being an action is an inessential property of events. Philosophia unsuccessful trying T in the actual world would have caused a bodily movement and been a non-autonomous psychological part of an action. Rather, what we should say is that in better circumstances the final stage of T would have caused a bodily movement, and what in the actual world we picked out as an uncompleted process of (unsuccessful) trying would have been itself a completed process of (successful) trying. Nonetheless, Smith’s objection still raises an interesting question: Why should we make a conceptual division within an action between a purely physical bodily movement and a non-autonomous psychological part at all? O’Shaughnessy is led to make this division because he thinks that if a certain event is psychological then it is necessarily so. But it is not the case that a particular bodily movement that occurs when someone willingly moves her body is necessarily psychological (1980, 2:207-8). Think of a case in which an agent tries to raise her arm and dies after the nerve signals have been sent out from her brain but before her arm goes up.24 In such a case the rising of the arm (or better, the rising of the matter that made up her arm when she was alive) would not be a psychological event. O’Shaughnessy’s assumption that being psychological is an essential property of events or processes is nowhere argued for, however, and I do not see any reason to accept it. To the contrary, psychological processes are a species of psychical processes, in Aristotle’s sense of processes engaged in by living beings, and the counterpart of O’Shaughnessy’s assumption does not hold for other psychical processes. Consider again the process of digestion. It is possible that a creature should die after ingesting some food but before it has fully digested it, and that some of the chemical processes that constitute digestion in creatures of its kind should therefore take place after its death. Even so, it would be wrong to say that these chemical processes constitute digestion in the creature’s corpse. Only living things have the capacity to nourish themselves, and digestion is one of the processes by which they do so. I suggest that we take a similar line when it comes to action. When an agent successfully tries to raise her arm and her arm rises, she exercises a psychophysical power possessed by human beings. Here it is not that a part of her action is psychological and another physical; rather, the whole process is psychophysical through and through. But if she tries to raise her arm and dies before her arm goes up, then if the matter that had composed her arm rises, this rising is a purely physical event and not a psychophysical one. Haddock and Disjunctivism in the Theory of Action On the dual aspect theory an action is an act of will which is a process that includes bodily movement as its final part. In this final section I compare this theory to a different proposal recently advanced by Haddock. Haddock also thinks that a constraint on a satisfactory conception of action is that it entitles us to the thought that an agent’s activity is present in her bodily goings-on. But Haddock’s theory differs from 24 This assumes that we can make sense of her dying as a process whose duration is brief enough to allow the thought experiment to make sense. One might reasonably question whether this is valid assumption, but I will let it stand for the sake of argument. Philosophia the dual aspect one in that according to it actions are identical to bodily movements of a certain sort. I will argue that Haddock does not adequately respond to an objection to the notion that actions are identical to bodily movements. But I am also interested in Haddock’s view for another reason. He thinks that the category of bodily movements should be understood disjunctively. While I agree that there is room for disjunctivism in the theory of action, I suggest it is volition rather than bodily movement that should receive a disjunctive analysis.25 Inasmuch as he thinks actions are identical with bodily movements, Haddock shares some common ground with the standard story. But he finds the standard story’s attempt to reductively analyze action in terms of bodily movement with a certain sort of psychological etiology unsatisfactory. His alternative is an account of action modeled on John McDowell’s (1982) disjunctive conception of perceptual experience. The disjunctive conception of perceptual experience stands in contrast with the more traditional “highest common factor” conception. These two conceptions offer rival ways of accounting for the distinction between cases of genuine perception of an object (a “good” case of, e.g., seeing x) and cases of illusory experience (a “bad” case of, e.g., merely seeming to see x while not actually seeing it). On the highest common factor conception in both the good case and the bad one the subject is in a state of the same psychological type, having an appearance of an object (e.g., having a visual appearance of x, seeming to see x), which is explanatorily more basic than either perceiving x or having an illusory experience of x. The difference between the good case and the bad case, roughly, is that in the good case having an appearance of an object is caused in the right way by x while in the bad case it is not. Where disjunctivism parts ways with the highest common factor conception is that according to disjunctivism it is not the case that in both the good case and the bad one the subject is in a state of a psychological type that is explanatorily more basic than either perception or illusion. The category of appearance is rather constituted by two psychologically distinct types of states. Having an appearance of an object is either a case of genuine perception—an “appearance of reality” which is the object’s manifesting itself to the subject—or a case of illusory experience—a “mere appearance”, in which an object merely seems to be manifesting itself to the subject but is not. And these states belong to distinct types insofar as the first but not the second can constitute a conclusive reason to believe that x is located in one’s immediate environment. Haddock argues that the standard story is analogous, in an important respect, to the highest common factor conception of perceptual experience. As he understands it, in both the case of an action and the case of a non-active bodily movement the standard story posits that there occurs an event of bodily movement of a type that is explanatorily more basic than either action or non-active, mere bodily movement.26 On Haddock’s own disjunctive conception of bodily movement, by contrast, the category 25 Brewer (1993) defends a disjunctive account of willing. O’Brien (2007, 151-2) mentions the possibility of a disjunctive account of trying without endorsing it. Lowe (2000, 248) distinguishes trying from volition and seems sympathetic to a disjunctive account of the former while rejecting it for the latter. 26 The bodily movements that occur in each case are not, of course, supposed to be of the same psychological type; but perhaps they should be theorized as belonging to the same explanatorily more basic physiological or biomechanical type. Philosophia of bodily movement is constituted by two distinct types of events: a bodily movement is either an event of an agent’s moving her body (an action) or a mere bodily movement that is not an action (2005, 163-4). Haddock notes that there is a powerful objection to the claim that actions are bodily movements (of any sort). Hornsby has called attention to the fact that it is a linguistic truth that a necessary condition of the truth of propositions of the form “x V T’s y” is that x causes y to V I , where “x” and “y” denote continuants, “V” stands for a verb with both transitive and intransitive forms, and “VT ” stands for the verb’s transitive form while “VI” stands for its intransitive form (1980, 13). This linguistic truth licenses inferences that accord with the following schema: (L) If x VT’s y, then x causes y to VI . One instance of (L) relevant for our purposes here is (M) If S movesT her body, then S causes her body to moveI. Hornsby interprets (M) to mean that when an agent acts her action causes her bodily movements. And if that is so, then Haddock’s thesis that actions are bodily movements of a certain sort is false. Haddock’s responds to this argument by resisting Hornsby’s interpretation of (M). On his alternative a claim such as “when Jane movesT her body, Jane causes her body to moveI ” “simply registers (a) that an event of Jane’s movingT her body has occurred, and (b) that the fact this event occurs entails that an event of Jane’s body movingI occurs” (2005, 166-7). The word “cause”, on this view, does not mark the existence of a causal relation but merely of a logical entailment between bodily movementT and bodily movementI. If this interpretation of (M) is plausible, then there is no longer any objection to conceptualizing actions as identical with bodily movements of a certain sort. For if actions are bodily movements, then the occurrence of an action will logically entail the occurrence of a bodily movement. Is Haddock’s interpretation of (M) plausible? He writes that “cause” (or “because”) often marks relations such as logical entailment. For example, the proposition “the apple is colored because it is red” may be true even though it is not the case that there is a causal relation that holds between the apple’s being red and its being colored. Granted, but notice that in this case it is propositions that stand in the “because” relation. The proposition asserts that there is a relation of explanatory dependence of the first embedded proposition upon the second, and this dependence holds because the second embedded proposition entails the first. But in (M) it is not two propositions that are related by “cause”. On the surface, at least, what are so related are an agent and a change that her body is undergoing—a movementI.27 And agents are not the right kind of thing to stand in relations of logical entailment. Haddock’s interpretation therefore involves a distortion of (M). In this respect he has something in common with Hornsby, who interprets (M) to register a causal relation between two 27 I am taking it that the infinitival phrase “her body to moveI”, which appears in (M), denotes an event of bodily movement. This is reasonable because (M) seems to be semantically equivalent to (M*) If S movesT her body, then S causes her bodily movementI. Philosophia events rather than between an agent and an event. But I take it that, other things equal, a theory of action that does not involve such a distortion is to be preferred.28 The dual aspect theory is preferable to Haddock’s because it allows us to take (M) at face value. (M) tells us that if an agent moves her body, then she is the cause of her bodily movements. An agent is a substance, and the way that substances bring about events and states is through manifesting or exercising their causal powers.29 A central claim shared by both volitionalism and the dual aspect theory is that in the case of human agency agents act by exercising their wills. But on the dual aspect theory an agent’s causing the relevant bodily movements is constituted by the entire process that occurs when she exercises her will. This process involves sub-processes that happen in the agent’s nervous and muscular systems as well as her bodily movements. Haddock’s disjunctive theory of bodily movements is implausible. It is noteworthy, though, that there is also arguably a structural similarity between the highest common factor conception of perceptual experience and volitionalism (Brewer 1993, 305-6). In the realm of action we can distinguish between two sorts of cases as follows. In one sort of case (the “bad” case) an agent tries to move her body but her effort is a total failure, such that no bodily movements occur. But in another sort of case (the “good” case) the agent tries to move her body and she is successful. Just as the highest common factor conception of experience posits an explanatorily more basic state of appearance that underlies both cases of genuine perception and illusion, the volitionalist theorizes that in both the good case of action and the bad one there occurs a certain type of episode, a volition, that is explanatorily more basic than either successful action or failed attempt. And like the highest common factor conception, volitionalism says that the difference between the good case and the bad one is determined by the causal relations this explanatorily basic episode stands, or fails to stand, in. The good case is one in which the agent’s volition suitably causes the appropriate bodily movements, while the bad case results when her volition fails to suitably cause them. If volitionalism is analogous to the highest common factor conception of experience then the dual aspect theory I have been concerned to defend here may be seen as the natural ally of a disjunctive conception of volition. I have allowed that there can occur episodes of naked trying, and if this is right then there are cases in which an act of volition occurs in the absence of any bodily movements. But I have also explained that on the dual aspect theory when an agent succeeds in moving her body, her action is not caused by her volition—it is her volition. So on this view the category of volition should be understood disjunctively: tryings are constituted either by mere failed tryings or by successful tryings, which are bodily activities. It is not the case 28 In other words, I think a better account of action will involve the idea of agent causation. Hornsby and Haddock are both pushed into distorting interpretations of (M) because they reject the notion of agent causation without argument. However, I think that they are far too quick to dismiss it. One of the most influential arguments against agent causation is to the claim that agents cause their actions (e.g., Davidson 1971, 52). But (M) does not commit us to the idea that agents cause an action (a bodily movementT) but that they cause their bodily movementsI. I do not mean here to rule out reductive analyses of agent causation in terms of event causation (cf. Bishop 1989; Velleman 1992). A reductionist about agent causation can still admit that (M) commits us to the existence of agent causation she will just add that it is reducible to event causation of a certain sort, whereas Hornsby and Haddock want to deny even this much. 29 For a classic account of causal powers see Harré and Madden 1975. Philosophia that volitions are psychological episodes that are explanatorily more basic than either failed tryings or successful ones.30 The proponent of the dual aspect theory has a way of explaining the difference between the good case of trying and the bad one that is different from the volitionalist explanation. When an agent acts she exercises her will, and this exercise is her trying. When all goes well her trying is successful and is also describable as her moving her body. But the agent’s volition is liable to be affected by factors that interrupt or otherwise interfere with its unfolding (e.g., paralysis). If some such factor interrupts or interferes with her trying prior to bodily movement the agent will be describable as merely trying to move. The difference between the good case of volition and the bad one, then, is not a matter of whether the agent’s trying manages to suitably cause appropriate bodily movements. It is a difference of whether the process of action is allowed to reach the end it is naturally directed toward or whether it is instead thwarted from reaching its completion.31 References Armstrong, D. (1981). The nature of mind. Brighton: The Harvester Press. Bar-On, D. (2005). Speaking my mind: Expression and self-knowledge. Oxford: Oxford University Press. Bishop, J. (1989). Natural agency: An essay on the causal theory of action. Cambridge: Cambridge University Press. Brandom, R. (1998). Making it explicit: Reasoning, representing, and discursive commitment. Cambridge, MA: Harvard University Press. Brewer, B. (1993). The integration of spatial vision and action. In N. Eilan, R. McCarthy, & B. Brewer (Eds.), Spatial representation. Oxford: Oxford University Press. Davidson, D (1963). Actions, reasons, and causes. In D. Davidson (Ed.), Essays on actions and events (pp. 3–20). Oxford: Oxford University Press. Davidson, D. (1971). Agency. In D. Davidson (Ed.), Essays on actions and events (pp. 43–62). Oxford: Oxford University Press. Davidson, D. (1987). Problems in the explanation of action. In Rationality in action (pp. 101–116). Oxford: Oxford University Press, 2004 Davis, L. H. (1979). Theory of action. Englewood Cliffs, NJ: Prentice Hall. Enç, B. (2003). How we act: Causes, reasons, and intentions. Oxford: Oxford University Press. Finkelstein, D. H. (2008). Expression and the inner. Cambridge, MA: Harvard University Press. Frankfurt, H. G. (1978). The problem of action. American Philosophical Quarterly, 15(2), 157–162. Ginet, C. (1990). On action. Cambridge: Cambridge University Press. Grice, P. (1975). Logic and conversation. In P. Cole & J. Morgan (Eds.), Syntax and Semantics, 3: Speech Acts. New York: Academic Press. Grünbaum, T. (2008). Trying and the arguments from total failure. Philosophia, 36, 67–86. Haddock, A. (2005). At one with our actions but at two with our bodies: Hornsby’s account of action. Philosophical Explorations, 8(2), 157–172. Harré, R., & Madden, E. H. (1975). Causal powers: A theory of natural necessity. Oxford: Basil Blackwell. Hornsby, J. (1980). Actions. London: Routledge & Kegan Paul. 30 In §2 I argued that volition should be conceived of in terms of a present-directed and continuing assent to the performance of a certain type of action. Furthermore, I argued that this assent is present in a case of paralysis as well as in a case of successful action. This shows that there is something present in both cases that goes beyond mere wishing and may be classified as a volition. The point of the present paragraph is not to deny this but to claim that the category of volition so understood should itself be conceived of disjunctively. 31 I would like to thank two anonymous referees for their comments on an earlier draft of this paper. Philosophia Hornsby, J. (2011). Actions in their circumstances. In A. Ford, J. Hornsby, & F. Stoutland (Eds.), Essays on Anscombe’s Intention. Cambridge: Harvard University Press. Kant, I. (1998). Groundwork for the metaphysics of morals. Trans. by M. Gregor. In M. Gregor (Ed.), The Cambridge edition of the works of Immanuel Kant: Practical philosophy (pp. 37–108). Cambridge: Cambridge University Press. Lowe, E. J. (2000). An introduction to the philosophy of mind. Cambridge: Cambridge University Press. Lowe, E. J. (2008). Personal agency. In Personal agency (pp. 141–158). Oxford: Oxford University Press. Moran, R. (2001). Authority and estrangement: An essay on self-knowledge. Cambridge, MA: Harvard University Press. McCann, H.J. (1974). Volition and basic action. In H.J. McCann (Ed.), The works of agency: On human action, will and freedom (pp. 75–93). Ithaca, NY: Cornell University Press. McCann, H. J. (1975). Trying, paralysis, and volition. In H.J. McCann (Ed.), The works of agency: On human action, will and freedom (pp. 94–109). Ithaca, NY: Cornell University Press. McCann, H. J. (1986). Intrinsic intentionality. In H.J. McCann (Ed.), The works of agency: On human action, will and freedom (pp 127–146). Ithaca, NY: Cornell University Press. McDowell, J. (1982). Criteria, defeasibility, and knowledge. Proceedings of the British Academy, 68, 455– 479. McDowell, J. (2011). What is the content of an intention in action? In M. de Gaynesford (Ed.), Agents and their actions. Malden, MA: Wiley-Blackwell. McGinn, C. (1982). The character of mind. Oxford: Oxford University Press. Morris, K. J. (1988). Actions and the body: Sartre vs. Hornsby. Philosophy and Phenomenological Research, 48(3), 473–488. O’Brien, L. (2007). Self-knowing agents. Oxford: Oxford University Press. O’Shaughnessy, B. (1973). Trying (as the mental “pineal gland”). The Journal of Philosophy, 70(13), 365– 386. O’Shaughnessy, B. (1980). The will: A dual aspect theory (Vol. 2). Cambridge: Cambridge University Press. O’Shaughnessy, B. (2003). The epistemology of physical action. In J. Roessler & N. Eilan (Eds.), Agency and self-awareness. Oxford: Oxford University Press. Pietroski, P. M. (1998). Actions, adjuncts, and agency. Mind, 107(425), 73–111. Prichard, H. A. (1945). Acting, willing, and desiring. In Moral obligation. Oxford: Clarendon Press. 1949. Searle, J. (2001). Rationality in action. Cambridge, MA: The MIT Press. Sellars, W. (1976). Volitions re-affirmed. In M. Brand & D. Walton (Eds.), Action theory. Dordrecht, Holland: D. Reidel. Shah, N., & Velleman, J. D. (2005). Doxastic deliberation. Philosophical Review, 114(4), 497–534. Smith, A. D. (1988). Agency and the essence of actions. The Philosophical Quarterly, 38(153), 401–421. Stout, R. (1996). Things that happen because they should. Oxford: Oxford University Press. Thompson, M. (2008). Life and action: Elementary structures of practice and practical thought. Cambridge, MA: Harvard University Press. Velleman, J. D. (1992). What happens when someone acts? In The possibility of practical reason (pp. 123– 133). Oxford: Clarendon Press, 2000. Widerker, D. (1995). Libertarianism and Frankfurt’s attack on the principle of alternative possibilities. Philosophical Review, 104(2), 247–261. Zhu, J. (2004). Understanding volition. Philosophical Psychology, 17(2), 247–273.