Empathy, Shared Intentionality, and Motivation by Moral Reasons
- First Online:
- Cite this article as:
- Hourdequin, M. Ethic Theory Moral Prac (2012) 15: 403. doi:10.1007/s10677-011-9288-5
- 311 Views
Internalists about reasons generally insist that if a putative reason, R, is to count as a genuine normative reason for a particular agent to do something, then R must make a rational connection to some desire or interest of the agent in question. If internalism is true, but moral reasons purport to apply to agents independently of the particular desires, interests, and commitments they have, then we may be forced to conclude that moral reasons are incoherent. Richard Joyce (2001) develops an argument along these lines. Against this view, I argue that we can make sense of moral reasons as reasons that apply to, and are capable of motivating, agents independently of their prior interests and desires. More specifically, I argue that moral agents, in virtue of their capacities for empathy and shared intentionality, are sensitive to reasons that do not directly link up with their pre-existing ends. In particular, they are sensitive to, and hence can be motivated by, reasons grounded in the desires, projects, commitments, concerns, and interests of others. Moral reasons are a subset of this class of reasons to which moral agents are sensitive. Thus, moral agents can be motivated by moral reasons, even where such reasons fail to link up to their own pre-existing ends.
KeywordsEmpathyShared intentionalityMoral reasonsMoral motivationInternalismExternalism
The debate between internalists and externalists about normative reasons centers on the question of motivation. Internalists generally insist that if a putative reason, R, is to count as a genuine normative reason for an agent, A, to do something, then R must make a rational connection with some element of the agent’s “subjective motivational set” (Williams 1981). Put in other words, if R is a reason for A to φ, then A must have “some motive which will be served or furthered by his φ -ing” (Williams 1981). Externalists disagree. They hold that the existence of normative reasons is independent of the conative particularities of individual agents. Normative reasons need not be tied to motivational sets of the agents to whom they are addressed. The appeal of internalism is that it helps to explain how normative reasons motivate, while externalism seems to leave this a mystery. The appeal of externalism is that it avoids relativizing normative reasons to particular agents, while internalism’s agent-relativism seems to leave no room for normative reasons—such as moral reasons—that apply to agents regardless of their particular ends.
Debates about internal and external reasons have deep significance in understanding the nature of morality. In particular, if internalism is true, but moral reasons purport to apply to agents independently of the particular desires, interests, and commitments they have, then we may be forced to conclude that moral reasons are incoherent: such reasons presuppose conditions of applicability that simply make no sense. Richard Joyce (2001) develops an argument along these lines, concluding, in the end, that morality as we know it is “a myth.” Against this view, I argue that we can make sense of moral reasons as reasons that apply to agents independently of their pre-existing ends, and at the same time, have the potential to motivate. The trick in squaring these two aspects of moral reasons—their applicability to moral agents independently of those agents’ ends (what one might call “the intersubjectivity of reasons”) and the potential for these reasons to motivate (or what I shall call “the motivational constraint on reasons”)—is to explain how it is that moral agents can be motivated by normative reasons that fail to connect to the pre-existing ends that each of them has.
My view is that an instrumentalist conception of motivation by reasons—where reasons can motivate only if they properly link up with a pre-existing desire or goal of the agent in question—is too narrow. In particular, it is too narrow because it excludes the possibility that the desires and goals of others can provide normative reasons for agents to act while meeting the motivational constraint. What I aim to show is that moral agents, in virtue of their capacities for empathy (understood as a particular kind of attunement and responsiveness to others’ emotions) and shared intentionality (understood as a particular kind of attunement and responsiveness to others’ goals), are sensitive to reasons that do not directly link up with their pre-existing ends. More specifically, they are sensitive to, and hence can be motivated by, reasons grounded in the desires, projects, commitments, concerns, and interests1 of others. Moral reasons are a subset of this class of reasons to which moral agents are sensitive. Thus, moral agents can be motivated by moral reasons, even where such reasons fail to link up to their own pre-existing ends.
1 An Argument Against Moral Reasons
Moral reasons purport to apply to agents independently of their pre-existing ends.
To count as legitimate or genuine normative reasons, reasons must have the potential to motivate those to whom they apply (the motivational constraint).3
A normative reason has the potential to motivate only if it rationally connects to some pre-existing end(s) of the agent to whom it applies (the instrumentality of normative reasons).
A normative reason rationally connects to an agent’s end(s) only if acting on that reason would further one of A’s ends.
Not all moral reasons point to actions that would further some end(s) of the agent(s) to whom they purport to apply.
Not all agents have ends to which moral reasons rationally connect.
Therefore, not all agents have the potential to be motivated by moral reasons.
Hence, moral reasons (qua reasons that apply to agents independently of their pre-existing ends) are not legitimate or genuine normative reasons.
For the sake of argument, I grant premises 1 and 2. My argument in this paper centers on rejection of the third premise, the instrumentality of normative reasons. My central aim is to provide an account that shows how agents can be motivated by reasons grounded in others’ ends, not merely by reasons grounded in their own, pre-existing ends. In doing so, I show how agents can be motivated by moral reasons, which, I have suggested, are a subset of reasons grounded in the ends of valuing subjects.
2 Motivation by Reasons
It is worth noting at the outset that in order to show that agents can be motivated by reasons grounded in others’ ends, it won’t be sufficient to show that agents who lack the requisite rationally-connected end(s) are sometimes motivated to φ as a result of being offered a putative reason to do it. For this phenomenon might be explained by the fact that under certain circumstances an agent might feel coerced, threatened, or pressured to act in accordance with a putative reason even though s/he was not compelled by the reason itself. What we want, then, is an account that shows how agents can be motivated by reasons that do not further any of their ends, and this account in turn will depend on some conception of what it is to be motivated by a reason.
I cannot give here a full account of what it is to be motivated by a reason, but I will proceed under the assumption that being motivated by a reason requires that the agent take the putative reason as a consideration that counts in favor of a particular action for her and not just a consideration that counts in favor of a particular action from the perspective of someone else who (for example) has power over her in ways that give her reason to act in accordance with that person’s wishes or reasons.4 My claim, then, is this: even if an agent lacks pre-existing ends to which moral reasons rationally connect, that agent can have moral reasons to do things, and (as required by the motivational constraint) potentially be motivated by those reasons.
My position is one that straddles the divide between internal and external reasons. On the one hand, I am sympathetic to the concern, expressed by internalists, that external reasons are “spooky” or metaphysically queer (Mackie 1991). If there are external reasons, internalists ask, then how do these reasons get their motivational grip? Though this is legitimate worry, I am dissatisfied with the internalist conclusion that normative reasons must be relativized to the ends of the agent(s) to whom they apply in order to meet the motivational contraint.
The strategy I favor is to relativize reasons to ends,5 while at the same time insisting that it is not just an agent’s own ends that can be reason-giving, but anyone’s ends.6 This straddles the divide between internalism and externalism, because normative reasons are internal in the sense that all reasons are grounded in the desires and interests of valuing subjects, and also in the sense that they have the potential to motivate moral agents, but external in the sense that they can apply to and motivate agents who lack pre-existing ends that would be promoted by acting on these reasons.7
If we relativize reasons to ends in the broad sense I have suggested, then moral reasons will have to be grounded in a subset of the ends that valuing subjects have, because clearly others’ nefarious ends do not give us moral reasons to promote them. However, my argument does not depend on the specification of this subset; we need only a rough sense of what such a specification might involve. One way of going is to say that others’ ends provide prima facie reasons for us to promote them, but since there are innumerable ends of this kind, and because the promotion of some ends would compromise the realization of others, we need to disqualify some ends as reason giving and somehow prioritize the rest. I leave open for now the criterion one might employ here, though my own view is that a naturalistically-informed account of morality that specifies the function of morality in human life might provide some clue.8
So let us suppose that all normative reasons, including moral reasons, are reasons grounded in ends, broadly construed, and therefore that there are no reasons built into the fabric of the world, independent of beings that have ends (Goldman 2005). Now, to counter the internalist, I need not deny that reasons can and often do motivate by rationally connecting to an end of the agent(s) to whom they are addressed. However, as I argue in the next section, reasons can also motivate by connecting with a general human capacity to recognize and be motivated by the ends of others, and to see others’ ends as providing reasons to act. The next section defends this claim by outlining two key psychological mechanisms that undergird this capacity. Although I am not the first to argue that others’ ends can be reason giving, the provision of a psychologically realistic account of motivation by such reasons advances the case by indicating more clearly how such motivation might work.
3 The Capacity to be Motivated by Reasons Based in Others’ Ends
One way to defend the view that an agent can be motivated by reasons that are independent of her own ends is, as I have suggested, to appeal to the relevant phenomenology. When we teach a young child to share, for example, we are not only teaching him that sharing advances some standing desire of his (e.g., to be treated well by other children), we are also pointing out that the desires, wants, and needs of other children are things that can and should count as reasons for him, regardless of his own subjective desires. In our moral teaching and conversation, we frequently presuppose that others’ desires and interests can be sources of reasons. When we tell a person that they owe us the truth, we are not, in general, appealing exclusively to that person’s interests or pointing out how telling the truth will best satisfy certain ends that person already holds. Instead we are pointing out that our ends count—or should count—in their decisions about what to say and do.
Now, if we accept as a working hypothesis the proposal that all normative reasons are grounded in ends (though not necessarily the ends of the agents to whom they apply)—then the critical question is this: how can and do such reasons motivate? Here, we might consider the possibility that each of us has, as one of our ends, a standing desire to promote others’ ends. Yet this is implausible: there are surely many people who lack such a desire.9 Fortunately, there is an alternative explanation: the capacity to be motivated by reasons grounded in others’ ends does not require a standing desire, because there is a general human capacity to recognize and be motivated by the ends of others, and to see others’ ends as reasons to act, and this can and frequently does underpin moral motivation.10 The critical challenge is to defend the existence and explain the workings of this capacity. I do this by identifying two key elements that undergird it: attunement to others’ emotions through empathy, and attunement to others’ ends through joint attention and shared intentionality.
4 Empathy, Shared Intentionality, and Intersubjective Perspective Taking
Shared Emotion and Emotional Perspective Taking: Empathy
Empathy, in the most basic sense, involves the capacity to share emotions with others, either by sharing another individual’s actual emotions or by anticipating the way an individual would be likely to feel in a particular situation (Hoffman 2000).11 In recent work, Frans de Waal (2008, 281) defines empathy as “the capacity to (a) be affected by and share the emotional state of another, (b) assess the reasons for the other’s state, and (c) identify with the other, adopting his or her perspective.” Thus, empathy typically involves shared emotional experience, but in all but its most basic forms, it goes beyond emotional mirroring. Empathy involves a particular orientation toward the other, in which agents are attuned and responsive to the affective states of those with whom they empathize.
The capacity to share emotions through empathy appears early in infancy. Empathy at this stage often works through emotional contagion; empathic responses are “primitive, automatic, and…involuntary” and involve the triggering of an emotional response in the observer that mirrors that of the observed, or is a response to the situation of the observed (Hoffman 2000, 36). One baby’s cries, for example, can trigger an empathic response in another through mirroring, or what Hoffman calls “mimicry.”
More advanced modes of empathy involve sophisticated cognitive and verbal abilities. In mediated association, an individual can be prompted to feel empathy on the basis of a written or verbal description of another person’s situation. In such cases, “language is the mediator or link between the model’s feeling and the observer’s experience” (Hoffman 2000, 49). In role taking, empathy is still more complex, because it requires that the observer take the perspective of another, imagining the emotions appropriate to the other’s situation.
There is an important feature of empathetic arousal that all modes—primitive and advanced—share in common. Although advanced modes may involve greater conscious control, the empathic reaction, in general, reflects responsiveness to others’ emotions and needs.12 Once children are old enough to differentiate self from other, empathetic arousal typically initiates sympathetic concern and motivation to alleviate the other’s distress (Hoffman 2000, 87).
How, then, should we understand empathy in relation to the question of whether individuals can be motivated by reasons independent of their particular ends? Research on empathy generally, and Hoffman’s work in particular, indicate that empathic responses generate motivation to alleviate others’ distress, and such motivation is typically triggered not by the individual’s own pre-existing desires, but rather as a response to something external: encounters with others and their emotional states. Empathy allows us to see others’ distress as a reason to act to alleviate it, and we need not have a standing desire to alleviate others’ distress in order to be moved by the distress of others that we happen to encounter in the world.
In its more mature forms, empathy involves a general orientation of concern for others, or what one might classify as an affective attitude. Yet even in this mature form, empathy itself is not a desire, goal, commitment, or project. Empathy is not an end. That is, empathy is not a desire to alleviate the suffering of others; it is an orientation toward others’ emotions that makes these emotions salient and enables others’ distress to get a grip on us, to matter or to count.13
Roberta may have had no desire prior to viewing the film that explains her decision to join the boycott. And whatever desire she does have after the film seems itself to be the result of her becoming aware, in a particularly vivid way, of considerations that motivate here desire and that she takes as reasons for her decision: the unjustifiable suffering of the workers.
Darwall’s key point here is that prior to seeing the film, it need not have been the case that Roberta had a general desire to relieve suffering “or even a sensitivity to suffering in the sense that she was likely to notice it without its being brought to her attention” (1983, 40). Roberta’s response to the film, her perception of a reason to act, need not be explained by reference to a standing desire of hers. My view of the case is that it was Roberta’s empathic response to the textile workers that enabled her to see their suffering as a reason to act.
So empathy itself is not a desire, though it may trigger desires. For example, seeing pictures of victims of the 2010 earthquake in Haiti or the 2011 earthquake and tsunami Japan may trigger a desire to help not only those particular individuals, but those affected by these crises more generally. Further, exposure to humanitarian crises may prompt the development of a desire to promote emergency preparedness and to change the social, institutional, and infrastructural conditions that exacerbate the effects of natural disasters. In this way, experiences of empathy may reveal to us conditions in the world that we then desire to change.
“Shared Intentionality,” Shared Conation, and Conative Perspective-Taking
One of the crucial features of empathy, with respect to moral motivation, is that it enables us to see others’ emotions—particularly others’ distress—as reason giving. Our capacity to share conations with others can work analogously: in this case, we see others’ goals as reason giving. The literature that most closely explores this possibility is work in developmental psychology on joint attention and “shared intentionality.” I describe this work below, explaining how what I call “conative perspective-taking” provides a second means by which agents’ can be motivated by reasons based in others’ ends.
Some of the most interesting and relevant work on shared intentionality is that of Michael Tomasello. According to Tomasello, shared intentionality, or “we” intentionality, involves “collaborative interactions in which participants share psychological states with one another” (Tomasello and Carpenter 2007; see also Gilbert 1989; Searle 1995).15 Shared intentionality begins to develop during the first year of life. It involves responding to and sharing others’ emotions, identifying the objects of others’ attention, discerning and sharing others’ goals, cooperative communication, and coordinated social interaction. Joint attention is a key element of shared intentionality and involves two individuals intentionally looking at the same thing (object, scene, person, etc.) with a shared awareness of one another’s attention to that thing (Tomasello and Carpenter 2007). Non-human primates seem to lack this capacity for joint attention. Chimpanzees, for example, will follow another’s gaze to an object or location, however, they don’t seem to engage in joint attention (Tomasello and Carpenter 2007). In contrast, human infants not only follow the gaze of others, but “attempt to share attention with others”: before one year of age, human children not only share similar experiences (in virtue of attending to the same object, for example) with an adult, but are aware of and motivated to seek these shared experiences (Tomasello and Carpenter 2007). Tomasello and Carpenter (2007) identify a number of domains in which human children seek to “[share] psychological states with others,” in contrast to non-human primates, who remain focused on individual goals. For example, young human beings share information with others by pointing and gesturing, and this information sharing occurs in cases where the child is not trying to “get something” from another person, but simply seeking “to share experiences and information with others” (Tomasello and Carpenter 2007, 122).
Young children, therefore, unlike non-human primates, have general propensities to share interest, attention, and goals with others. These pro-social propensities begin in early infancy with “protoconversations” between human infants and their caregivers, in which the adult and infant often exchange expressions of similar emotions in dyadic, face-to-face interaction (Tomasello et al. 2005). Later, 9–12 month olds employ joint attentional capacities in triadic engagement: for example, an adult and child may take turns in an effort to accomplish together a joint task, such as building a tower of blocks (Tomasello et al. 2005). As children pass the one year mark, their involvement in these kinds of joint activities grows more sophisticated, as they begin to identify and take up distinct roles in joint tasks, and even assist or prompt others to take up their roles (Tomasello et al. 2005). In contrast, ape infants do not engage in protoconversations with their mothers, nor do they participate in triadic interactions with conspecifics with a shared goal (Tomasello et al. 2005).
Work by Tomasello and others thus suggests that from a very young age, humans guide one another’s attention—through gaze following as well as through more directed means such as pointing. Furthermore, young children are able to discern not only what someone else is looking at, but also the nature of another person’s interest the objects to which they are attending and with which they interact. In other words, human beings seem oriented to interpret others’ behavior as goal directed, and to use attentional, cognitive, and affective skills to discern the nature of others’ goals.
Tomasello frequently suggests that shared intentionality involves not just discernment of others’ goals, but adoption of those goals, and the philosophical literature on shared intentionality emphasizes joint intentions and joint goals.16 Tomasello’s view is supported by his finding that young children are eager to join in goal-oriented games and tasks with others, such that once involved, they prompt the initiator of the game or task to continue if they begin to withdraw from the game. In such cases, it appears that children adopt a joint goal with their partner, and persist in their allegiance to that goal even when the partner’s commitment to it wavers.
However, although Tomasello is often unclear about how we should understand “shared intentionality” and whether shared intentionality necessarily involves the adoption of a joint, or shared, goal, the evidence suggests that shared conation—averted to in Tomasello’s description of “[shared] psychological states”—is possible in the absence of a joint goal. To understand how this is possible requires a distinction analogous to one we can make in relation to empathy. With empathy, it is possible to share emotions with another while also continuing to mark a difference between one’s own emotions and those of others. This distinction is important, because it allows the empathizing individual to distinguish her own distress—marking, for example, the frustration of some goal or a response to injury of hers—from empathic distress, which draws her attention to the situation of the other. Conative perspective-taking seems to allow something quite similar, with respect to goals: one can “share goals” through conative perspective-taking in the same way one shares emotions through empathy, while maintaining a distinction between those goals (or emotions) that belong to oneself and those that one experiences vicariously on behalf of others. Of course, this is not to say that conative-perspective taking never involves the adoption of another’s goals as one’s own: one can be moved to adopt the goals of others as one’s own goals in this way, but this doesn’t seem to be a necessary consequence of conative perspective-taking. One might, instead, simply see others’ goals as providing reasons to promote them, independently of whether one adopts such goals as one’s own. This is important, because I want to argue that adoption of joint goal is not necessary to enable a person to be motivated by reasons grounded in others’ ends. Instead, the kind of conative perspective-taking described above will suffice.
The as if preference…is, of course, a devised construction and need not have any intuitive plausibility seen as preference. A morally exacting choice constraint can lead to an outcome that the person does not, in any sense, “desire,” but which simply mimics the effect of his self-restraining constraint…To illustrate, there has been a good deal of discussion recently on the alleged tendency of many Japanese workers to work extraordinarily hard… The tendency to do one’s “duty” to the point of severely damaging one’s health…is easier to explain as the consequence of adhering to a deontological obligation rather than as an outcome that is actually “preferred” by the hapless worker. Social psychology can be important here. The as if preference works well enough formally, but the sociology of the phenomenon calls for something more than the establishment of formal equivalences. (Sen 1997, 770, emphases in original)
Sen’s suggestion that identification with others may be key to choices that respond to others’ goals is consistent with the kind of conative perspective-taking described above. Though Sen provides limited details, and seems to tie this process of identification to a sense of identity that encompasses membership in groups of various sizes and descriptions,18 the central idea bears similarity to the view I defend. My own view differs, however, in that I believe that the ability to discern and see as reason-giving the goals of others is fairly basic: it is present in young children and requires no sophisticated sense of identity. Nevertheless, the idea that motivation by others’ ends is grounded in a kind of identification or perspective-taking seems to me exactly right. Work by Tomasello and others shows that we engage in conative perspective taking—recognizing, understanding, and responding to the goals of others—just as we engage in affective perspective taking through empathy. What’s more, Sen too notes that there is an important distinction to be made between acting in pursuit of one’s own goals and acting to promote the goals of others, even though both kinds of goals can directly motivate (or in Sen’s terms, guide choice). On my view, conative and affective perspective-taking allow agents to identify with the goals and emotions of others, and to see those goals and emotions as sources of reasons to act, while at the same time maintaining a separation between self and other that enables the agent to distinguish her own emotions from the empathically-shared emotions and her own goals from those experienced through conative attunement with others.
A person’s concept of his own welfare can be influenced by the position of others in ways that may go well beyond “sympathizing” with others and may actually involve identifying with them… [P]erhaps most important in the context of the present discussion, the pursuit of private goals may well be compromised by the consideration of the goals of others in the group with whom the person has some sense of identity.
Implications for Motivation by Reasons Grounded in Others’ Ends
Empathy and shared intentionality, individually and jointly, facilitate the interpersonal sharing of outlooks and standpoints, and in doing so, they generate responsiveness to reasons grounded in others’ interests and ends. Empathy emphasizes the sharing of emotions, with particular attention to distress, and what I am suggesting is that empathy is what allows us to see another’s distress as a reason to act. Shared intentionality, on the other hand, emphasizes the sharing of perceptions, intentions, and goals. It grounds reasons in a way analogous to empathy, but rather than emphasizing others’ distress, shared intentionality allows us to see others’ goals and projects as reasons to act. Both empathy and shared intentionality involve motivational orientations that facilitate the recognition of others’ ends as reasons to act, and they often may work together, with empathy focusing particular attention on reasons in favor of actions that alleviate the pain and suffering of others. That is, while shared intentionality, as described in the literature, appears to be relatively content-neutral19; empathy helps pick out others’ interests in avoiding suffering and distress as grounding particularly strong reasons to act.
One might, at this point, wonder whether the model of motivation on offer depends on an agent’s coming to believe that she has certain reasons that she previously did not believe she had. In fact, it does not. In offering an account of this kind, I am not claiming, as some externalists about reasons do, that motivation by reasons grounded in others’ ends works primarily by the agent’s coming to believe that such reasons exist. The kind of motivational capacities I’ve discussed begin to work in early childhood and it is doubtful that their working depends on beliefs about reasons per se. The account is more pragmatic than doxastic. There is, however, something to the claim that we can perceive external reasons and be motivated accordingly (McDowell 1998). Recognizing another’s goal and perceiving that goal as a consideration that favors action to promote it does fit the phenomena involved with shared intentionality in early development. Thus, it may be helpful to understand the perspective taking capacities described above as propensities to see or take others’ ends and others’ emotions (particularly others’ distress) as reason giving.
At this point, another objection may arise: Can’t the capacities I’ve described above be understood simply as species of desire? If so, the model of moral motivation described here merely enlarges the set of desires that agents are assumed to possess, and my disagreement with the internalist will turn on whether I am right that agents have these desires (to promote others’ goals and to alleviate others distress) or whether internalists are right that they do not.
As I have indicated already, however, empathy and shared intentionality are not desires. Their action-guidingness depends on features external to the agent’s own desires and ends, and this dependency does not fit a standard belief-desire model of motivation (where it is a standing desire that provides the motivational force for action, and world-guided beliefs merely point to ways to realize that standing desire). Rather, in encountering the ends and distress of others, our capacities for shared intentionality and empathy are not only activated, but their intentional objects are given and shaped (Wong 2006, 2007). The empathic or conatively-attuned agent is not simply someone who desires to alleviate the suffering of others or to promote their ends; rather, he is someone who is responsive to the suffering and ends of others. Empathy and shared intentionality thus enable a robustly world-guided kind of motivation by reasons. This motivation is world-guided in that the source of reasons is external to the agent’s own desires. In the case of internal reasons, agents seek ways in which the world can conform to their interests; in the external case, agents are receptive and motivated to conform to the needs, interests, and desires of other valuing subjects in the world.
In a related vein, and as I have already suggested above, it is not the case that the model developed here can be assimilated to an instrumentalist model of motivation by arguing that empathy and shared intentionality promote the adoption of others’ ends, and that it is these newly-adopted ends that give the agent reasons to promote them. Empathy and shared intentionality enable an agent to see others’ goals as reason-giving regardless of whether they are adopted by her. If Sally observes a dog threatening Bill and sees the terror on his face, empathy allows her to see Bill’s terror as a reason to stop the dog from threatening Bill; she need not first adopt as one of her ends “alleviating Bill’s distress” or “preventing Bill’s terror” or any other end that stopping the dog from threatening Bill would promote. Similarly, the perspective-sharing dimensions of shared intentionality allow an agent to see others’ goals as reasonable or worthy, and hence as reason-giving, even without adopting these goals. Imagine that a child comes to your door selling cookies to support a class trip to the local art museum. You aren’t especially interested in cookies, as you are trying to reduce the amount of sugar in your diet. In addition, you have no particular interest in or appreciation for art. Yet you recognize that art enriches many people’s lives (even if it does not enrich your own) and that the trip will likely enrich and expand the students’ understanding and appreciation for art, potentially enriching their lives. So you buy some cookies. You see the child’s project as worthy of support and as reason giving, even as you do not take on the promotion of the arts (or more generally, the enrichment of others’ lives) as one of your ends.
Although I cannot provide a full argument for it here, the orientation bequeathed to us by empathy and shared intentionality seems to be one in which the default position is to see others’ ends as reason-giving. Thus, we have to rationalize the decision not to give these reasons weight. Imagine that you and I walk by a homeless person. Though we may have no standing desire to help the homeless20 (by hypothesis, let us assume we do not), we nevertheless feel the pull of the person’s request for help: the default position is that this person’s desire for help counts for each of us as a reason to act. Our dismissal of the request therefore requires justification—we must try to rationalize, to provide some reason or excuse for not helping, and this is typically what many of us do (sometimes rightly, sometimes wrongly). As this example suggests, it is not my claim that others’ ends are always a source of overriding reasons, either normative or motivating. Nevertheless, these examples show that empathy and shared intentionality are the basis for human motivational orientations in which others’ ends count as reasons for us to act, regardless of whether we adopt these ends as our own.
5 Moral Motivation and Moral Agency
If what I have argued thus far is correct, then there are at least two mechanisms by which we can come to see others’ ends as reasons for action, regardless of what our own individual ends happen to be. This drives a wedge in Joyce’s argument from internalism about reasons to the incoherence of moral reasons. If we allow that others’ ends are a source of reasons to promote them, and that empathy and shared intentionality enable us to recognize and potentially be motivated by these reasons, then it no longer seems reasonable to disqualify moral reasons as legitimate normative reasons on the grounds that they fail to meet the motivational constraint.
Here, it is crucial to see that the argument I have given does not depend on a universal human concern for others to ground morality, some shared desire that every moral agent possesses; instead, it depends only on the basic human capacity to be motivated by others’ interests and ends, which appears extremely widespread.
The class of individuals who lack the capacity to be motivated by others’ interests and ends (or in whom this capacity is severely impaired) is relatively small, and what’s more, it seems to correspond reasonably well to the class of individuals classified as having antisocial personality disorders (ASPDs), in whom moral agency is compromised or lacking. Recent studies suggest that individuals with ASPDs lack not the ability to infer others’ intentions or to understand others’ emotions cognitively, but rather the motivational and emotional dimensions of empathy and shared intentionality.21 If this is right, then those to whom moral reasons do not appropriately apply according to my account of moral motivation should correspond well with those with ASPDs, who often are treated as examples of “amoralists” or as deficient or lacking in the capacities required for moral agency (for discussion, see Blair 1995; Kennett 2002; Krahn and Fenton 2009). So unlike a theory that makes moral motivation dependent on an “active concern” for the community and risks classifying large numbers of individuals who lack such concern as sociopathic, my view preserves an alignment between those we typically think of as moral agents and those to whom moral reasons apply.
My suggestion, then, is that empathy22 and shared intentionality are extremely widespread in human beings and that these capacities are (among the) necessary conditions for moral agency. If this is right, then all moral agents, in virtue of having these capacities, have the potential to be motivated by moral reasons. Now, if the capacity for shared intentionality is robust, and if, indeed, its development is essential to the development of language in human beings—as Tomasello (2001, 2008) argues—then we may conclude that any competent human language speaker also possesses (at least to some extent) developed shared intentionality. And if this is right, then these individuals have the capacity to respond to reasons grounded in others’ interests and ends, and even in the absence of robust empathic capacities, such individuals have the potential to respond to moral reasons, insofar as such reasons are tied to others’ ends.23
What’s more, even if Tomasello is wrong, and shared intentionality is not a prerequisite for language learning, shared intentionality in humans does appear to be extremely widespread, and the although the “self-centered person” may not have an altruistic orientation or altruistic ends, this in itself does not establish her insensitivity to moral reasons or the inappropriateness of directing moral reasons to her. “Self-centered” persons, so long as they possess the basic capacities for shared intentionality, have the potential to be motivated by moral reasons—and there is good reason to think that the vast majority of self-centered persons do have these capacities.
The argument for this point goes as follows: Even selfish individuals need to cooperate with others to achieve their ends and to engage in joint projects, and success in these endeavors requires shared intentionality, so we can reasonably infer that selfish individuals who successfully engage in cooperative projects have the capacity for shared intentionality, and indeed, that they frequently exercise this capacity. Take, for example, an opportunistic professional whose primary goal is to get ahead, with little concern for the broader social good. Such a person may give little weight to moral reasons, and for example, fail to see poverty as a reason to help others. Nevertheless, this professional is good at what he does, in part because he possesses certain fundamental social capacities. To collaborate effectively with his coworkers, colleagues, clients, and others, he must be capable of discerning their ends (whether these ends are expressed explicitly or implicitly) and of (at least sometimes) taking their ends as reasons to act. In contrast, imagine an individual who is completely unable to take others’ interests as reasons for action: this person, arguably, is not merely a garden-variety self-centered individual. Rather, she is profoundly socially dysfunctional: when she and her colleagues are working late and decide to order out for pizza, for example, her inability to see others’ desires as reasons leads her to insist on her own preferences despite the vociferous protests of her colleagues. Even if she eventually concedes and compromises on the pizza toppings, she does so not because she sees their preferences as providing reasons, but rather because her own interests provide reasons to placate her colleagues. (Perhaps she knows that if she does not placate them, they will treat her badly in the coming weeks, that she will not be able to secure a promotion, or that she will be passed over for a raise.) But I think that it is not hard to see that the person who fails to see others’ preferences in this case as themselves providing reasons (however weak) is not only unusual, but worrisomely so: a person who simply cannot see others’ interests as sources of reasons is, arguably, a person incapable of moral agency.
My point is not that individuals with the capacity for shared intentionality are necessarily moral persons, in the robust, substantive sense that they regularly act on moral reasons or give such reasons priority in their deliberations.24 Yet they are moral persons in a weaker sense: individuals who have the basic ability to recognize and be motivated by others’ ends and interests have the capacity to be motivated by moral reasons, or at least to see the pull of such reasons. Selfish agents may be criticized for selecting an overly narrow group of persons with whom to cooperate or whose ends are taken to be important reasons for action, or for putting too great a weight on their own narrow preferences and too little weight on the preferences of others, but insofar as they possess the capacity to see others’ interests as reason giving, they have the potential to be motivated by moral reasons.
One virtue of this account is that it offers a way of explaining how an agent who we might describe as “undermotivated” by moral reasons—i.e., a person who fails to give such reasons sufficient weight—nevertheless can be understood as susceptible to such reasons, or as having the capacity to recognize and act on such reasons. The strong parallel between the kinds of reasons the agent does recognize and moral reasons is the source of this susceptibility. The account thus seems to allow for the right kind of distance between what a person does and what she should do; it gives moral reasons their normative bite by showing how they have the potential to explain an agent’s actions even in cases where they don’t in fact do so.
I conclude, then, that an instrumentalist model of motivation by reasons is inadequate, and that shared intentionality and empathy enable agents to be motivated by others’ ends, independently of their prior motives. This expanded account of motivation undermines Richard Joyce’s argument against moral reasons by showing how an agent can have a reason to act, even if that reason would not further any of her own interests and ends. Insofar as moral reasons are grounded in others’ ends, they can’t be rejected on the grounds that they fail to meet the motivational constraint. Although not everyone has the capacity to be motivated by reasons grounded in the ends of others’, those who are incapable of such motivation typically have deficits in empathy and shared intentionality, and it is these deficits that seem to impair moral agency.
There is, of course, more work to be done. Others’ ends aren’t always the source of moral reasons—as noted earlier—so moral reasons must be a subset of reasons grounded in the ends that valuing subjects have. I am inclined toward a view that grounds moral reasons in others’ ends, but constrains moral reasons through the recognition of the need to harmonize these ends in order to facilitate cooperation, suppress selfishness, and make social life possible (Haidt and Kesebir 2010, p. 800). A theory along these lines clearly needs fleshing out, but it should not be rejected on the basis of its commitment to normative reasons that apply to agents independently of their particular individual ends.
I shall use “ends” as shorthand to refer to such desires, concerns, interests, and commitments. I intend “ends” broadly so as to include the desires and interests of sentient, non-human organisms.
For sake of clarity and contextualization, I have reconstructed the argument Joyce (2001, p. 77) offers, filling in some key premises from earlier in the text and rephrasing the argument to emphasize Joyce’s claims about reasons (while omitting his related claims about ‘ought’ statements and moral obligations).
Joyce distinguishes between subjective rational reasons and objective rational reasons, where the former count as reasons from the agent’s limited epistemic perspective (e.g., if the agent is hungry, and believes that the food she’s been offered will satisfy her hunger, then she has a subjective rational reason to eat the food, even if she lacks an objective rational reason to eat it, because the “food” she’s been offered is actually made of plastic.) Moral reasons are presumably reasons of the latter—i.e., objective—kind. To meet the motivational constraint, it must therefore be the case that an agent could be motivated by a moral reason R, assuming she had the relevant true background beliefs. An agent’s failure to be motivated by a moral reason due to certain false beliefs is not sufficient to disqualify the reason: if the agent, rational and fully informed, could be motivated by the reason in question, the reason meets the motivational constraint.
On my view, being motivated by a reason doesn’t require conscious awareness of the reason on which one acts. On this point, I find Nomy Arpaly’s (2003, ch. 2) arguments particularly convincing.
For simplicity, I allow that ‘ends’ can include subjective ends (desires and subjective interests) and—if there are any—“objective ends,” such as survival, that count as ends independently of whether the subject desires them or not. This broad account of ends can accommodate cases in which a person lacks the desire/end of preserving her health, although it is in her interest to do so. Interesting and difficult issues arise surrounding cases where subjective and objective ends come apart. The inclusive view I adopt here can help make sense of cases where an individual lacks “ends” in a robust sense or lacks the capability to form ends, but nevertheless has interests, the promotion of which might enable her to gain or regain the capacity to form and act on her own (and others’) ends. The overall point about the possibility of motivation by others’ ends can stand, however, even if one denies the existence of objective ends.
This strategy is not new; many philosophers have tried to highlight the symmetry among agents to suggest that if one’s own ends are the source of reasons, then others’ ends must be too. Kant’s ethics is the seminal source for arguments of this kind; influential twentieth century works include Thomas Nagel’s The Possibility of Altruism (1970) and Alan Gewirth’s Reason and Morality (1978). In recent work, Stephen Finlay (2006) defends an “end-relational” theory of normative reasons, which grounds reasons in ends (and not necessarily in the ends of the agent to whom they are addressed) and allows for the existence of external reasons. However, Finlay maintains a broadly internalist model of moral motivation, where external reasons fail to motivate.
David Wong (2006, ch. 7) makes a related claim. Wong “[affirms] externalism with respect to the individual’s motivation, but…affirms internalism with respect to human motivation” (183).
A number of authors have developed naturalistic accounts of morality tied to the function of morality in human life. Allan Gibbard (1990, 26) , for example, understands morality as growing out of humans’ evolved social nature and emphasizes the role of morality in coordination and cooperation. David Wong (2007, 39–44 and ch. 2 more generally) describes the interpersonal function of morality as “facilitating social cooperation” and the intrapersonal function as “promoting a psychological order within the individual” by “specifying what is worthwhile for the individual to become and pursue.” Wong argues that these functions developed through the course of human biological and cultural evolution and that these functions constrain the content of moralities and moral reasons, yet allow for a plurality of acceptable moral systems. I have argued that the central function of morality is to harmonize the diverse ends people hold (Hourdequin 2005). Psychologist Jonathan Haidt argues that the function of morality is to facilitate cooperation, suppress selfishness, and make social life possible (according to Haidt and Kesebir 2010, 800); this function bears significant resemblance to those articulated by Gibbard, Wong, and myself, and to the extent that this function is accepted as authoritative, it will constrain hat count as legitimate moral reasons.
On this point, I agree with Joyce (2001). Joyce actually puts this in terms of a desire for the good of the community; but the general point holds. The question of whether a standing desire of this kind can ground moral reasons is discussed further below.
In this regard, my view resembles that of David Wong (2006, 2007, 2008, and 2009). The key features our views share in common are these: 1) Both allow that moral reasons can motivate without connecting with a standing desire of the agent, and 2) the existence of a moral reason for an agent to do some action X is not dependent on the particular motivations of the agent in question, though 3) there is a sense in which (as Wong 2009, 345 puts it) “moral reasons are constrained by human psychology.” Insofar as internalism about reasons holds that the existence of a reason for A to Φ depends on the presence of some desire, interest, or end of A’s that would be served or furthered by Φ-ing, neither my view nor Wong’s counts as internalist.
Although ‘empathy’ is sometimes used interchangeably with ‘sympathy,’ it is helpful to distinguish the two: empathy necessarily involves feeling what the other feels or what one might expect that he or she feels, it is a kind of “feeling with”; whereas sympathy involves “feeling for” another person—e.g., feeling badly for another’s misfortune—without necessarily sharing the feelings of that person.
Following Martin Hoffman (2000), I am understanding empathy as involving both cognitive and emotional elements. Some authors (e.g., Blair 2008) distinguish “cognitive empathy” (involving the ability to discern/know what others are feeling) from “emotional empathy” (involving concern for what others are feeling), and it appears that “cognitive empathy” can function in the absence of the kind of responsiveness described here—i.e., some individuals (e.g., those with antisocial personality disorders) appear to have normal capacities for cognitive empathy paired with significant deficiencies in emotional empathy. For details, see Blair 2008.
Empathy can perhaps be helpfully be understood as grounding an attitude or way of seeing others that is analogous in important ways to the attitude toward or way of seeing ourselves that gives internal reasons their motivational grip. On this view, my own desires ground reasons for me to act in virtue of a general orientation of concern for myself. One way of understanding the failure to be motivated by one’s own desires is thus through a kind of alienation that weakens self-concern: in such cases, it no longer matters to me that I have certain ends, the reason-giving power of my ends is attenuated. My own distress, even, may cease to be strongly reason-giving in certain circumstances.
I thank David Wong for helping me appreciate the salience of this case. Wong (2006, ch. 7) also discusses Darwall’s example.
Elsewhere, Tomasello et al. (2005, 680) describe shared intentionality as “collaborative interactions in which participants have a shared goal…and coordinated action roles for pursuing that shared goal”, but this definition is too narrow to capture the range of phenomena described in Tomasello’s work on shared intentionality. In addition, for reasons described below, I don’t believe that the attention to and motivation by others’ ends requires goal sharing per se.
See, for example, the work of Raimo Tuomela, Margaret Gilbert, and Michael Bratman.
I am indebted to Schmid (2005) for offering a very helpful interpretation and commentary on Sen’s views of self-goal choice, which has informed the discussion below.
Bernhard Schmid, in a recent commentary on Sen’s work, emphasizes the importance of group identity in grounding action motivated by the goals of others. Schmid (2005, 57–58), for example, argues, “If identification with a group lies at the heart of the structure of commitment, and agent does not have to perform the paradoxical task of choosing someone else’s goal without making it his own in order to qualify as truly committed…[I]n committed action, the goals in question are not individual goals, but shared goals.” Although this may be the right way to analyze certain kinds of commitment, Schmid’s analysis still seems to require that an agent’s action be explained by a pre-existing goal—in this case, a shared or common goal rather than a private one. Schmid’s analysis is therefore inadequate to explain the phenomenon that concerns me here, in which others’ goals can ground reasons for action, even if the agent does not adopt those goals as her own. I thus disagree with Schmid in the sense that I try to show here how one can—unparadoxically—be motivated by another’s goal without making it one’s own.
But see Vaish, Carpenter, and Tomasello (2010) for discussion of how children’s observation of an agent’s harmful intention reduces helping behavior toward that person.
Or any other interest or desire that would be served by helping the homeless.
Dolan and Fullam (2004), for example, found that subjects with ASPDs performed no worse than subjects in the control group on standard theory of mind tasks, and they were not significantly impaired in their detection of social faux pas in a story they were told, but subjects in the ASPD group were less successful in understanding the feelings of the speaker and listener following the faux pas. Dolan and Fullam (2004, 1100) explain this finding as a possible result of a failure to “truly [empathize] with the characters in the stories or an indifference to the impact of the faux pas on the speaker or listener.” They therefore conclude that “the key deficits [in APSDs] appear to relate more to their lack of concern about the impact on potential victims than the inability to take a victim perspective” (2004, 1093). Relatedly, James Blair (2008, 158) has argued that psychopaths “have no impairment in ‘cognitive’ empathy [or Theory of Mind] but marked, and selective, impairment in ‘emotional’ empathy.” Interestingly, Blair argues that autistic individuals have the reverse problem: intact “emotional” empathy and deficient “cognitive” empathy. Although the precise relationship between shared intentionality and empathy has not been well worked out, and the terminology can be unclear, these findings suggest that deficiencies on the motivational side of empathy and shared intentionality may be central in ASPDs.
My claim here is that emotional empathy is a necessary condition for moral agency; cognitive empathy alone is insufficient.
This point raises some complex issues regarding the relationship between empathy and shared intentionality. My (albeit speculative) view is that a lack of robust empathic capacities compromises moral agency in certain ways; but insofar as shared intentionality has built into it a pro-social orientation, such a deficiency does not undermine moral agency altogether.
Because the motivational propensities involved in empathy and shared intentionality are very general, they may not lead to specifically moral action: not all actions which promote the ends of others can be considered moral. Moral education is needed precisely because these diffuse propensities need to be given an appropriate shape: through moral education we learn which of others’ (many) ends provide moral reasons, which do not, which moral reasons deserve the greatest weight, and ultimately (perhaps), why.
Acknowledgements: I wish to thank Rick Furtak, Leonard Kahn, Ivan Mayerhofer, Bill Rottschaefer, David Wong, the Springs Philosophy Discussion Group, and an anonymous reviewer for this journal for helpful comments on earlier versions of this paper.