A Dual Aspect Theory of Shared Intention

2.1 Bratman’s Psychological View

Bratman’s leading idea is that the phenomenon of (future-directed) intention or “planning” is at the heart of human agency, insofar as it plays some central characteristic roles in one’s practical thought and action (1987). A distinguishing role of intention, Bratman tells us, is to help organize and coordinate one’s actions both over time and interpersonally. According to Bratman, intention plays this role in virtue of involving a two-fold commitment to action which other conative attitudes, such as desire, normally lack (p. 15–18, 108–109). First, one’s intending to do something involves a disposition to “settle” or “control” – rather than merely to “potentially influence” – what one is going to do: if one intends to perform an action, and one’s intention persists until the time of action and nothing interferes, one will proceed to execute it then. Second, intending to perform an action disposes one to see the cited action as a fixed point in one’s deliberations and to reason in certain ways. In particular, it disposes one to avoid reconsidering one’s intention in the absence of new and significant information, to form further intentions about how to execute one’s intended end, and to eschew from deliberation options believed to be incompatible with one’s intention. It is one of Bratman’s fundamental ideas that such dispositions to reasoning are responsive to or guided by, associated norms of intention rationality such as intention stability, means-end coherence of intentions, and intention consistency.

But how should we understand shared intention, that is, the intention of a group of individuals? Bratman’s thesis is that we should understand it as a complex structure involving intentions of each individual participant: “a shared intention,” Bratman writes, “consists in a public, interlocking web of appropriate intentions of the individuals” (1999b, p. 9). Bratman’s argument for thinking that shared intention consists in such a structure or “web” of intentions comprises two steps. First, Bratman claims that it is distinctive of shared intention that it plays some central roles in our thought and action, and, further, that we have reason to identify shared intention with whatever it is that plays those roles. Second, Bratman argues that the central roles of shared intention are played by the cited structure of intentions of individuals. Therefore, Bratman concludes, that structure of intentions is shared intention – or, at least “one important kind” of it.^[6] Three questions arise here. First, what are, in Bratman’s view, the central roles of shared intention? Second, what exactly does the cited structure of intentions of individuals involve? And, third, how is it that such a structure plays the cited roles?

Suppose that you and I share an intention to dance the tango at the local club on Saturday. According to Bratman, our shared intention to dance the tango will typically play three interrelated roles in our lives. First, it will help us coordinate our individual actions: one of us will lead the dance, for example, and the other will follow. Second, our shared intention will help us coordinate our associated planning. For instance, if I plan to get the tickets but not to drive to the local club, I will make sure that you plan to drive. Third, our shared intention will provide a background framework that can structure forms of bargaining between us, in case we have conflicting preferences about how we will dance the tango – say, about who will lead, what tango style we will dance to, and so on. Bratman’s point is that our shared intention will typically play such trio of interrelated roles in ways that lead to our successfully dancing the tango. When an individual agent intends to perform an action, Bratman claims, she commits to future conduct in ways that help organize, coordinate, and unify her agency over time. Much in the same way, Bratman suggests, when we share an intention to act together we commit to future conduct in ways that help organize, coordinate, and unify our joint intentional agency (Bratman 1999c, p. 112).

For Bratman the complex structure of intentions constituent of shared intention involves, mainly, intentions on the part of each participant in favor of the joint activity. When you and I share an intention to dance the tango, I intend that we dance the tango and you intend that we dance the tango. In addition, such a structure involves various complex interconnections between the cited intentions of participants. The details are not central here, but Bratman speaks of “interlocking” and “reflexive” intentions (1999c, p. 118–119); of intentions in favor of “meshing subplans” (p. 120); and so on. Finally, Bratman maintains that the cited structure of intentions is common knowledge between the participants (p. 117). Bratman argues next that the cited complex structure of intentions accounts for the aforementioned trio of roles characteristic of shared intention. Basically, Bratman says, we start by noting that shared intention involves intentions of individuals, intentions with the cited special contents and interconnected in the aforementioned ways. Then, we note that these intentions are, after all, ordinary intentions of individuals, and as such, are subject to norms of intention rationality such as means-end coherence and intention consistency. Finally, we note that such norms of intention rationality exert pressure on those intentions of individuals – intentions with such contents and so interconnected – in ways that lead to the coordination of action and of planning, and to appropriate bargaining, in pursuit of the joint activity (p. 122–125).

In Section 1, I called attention to two regularities associated with shared intention, namely, that (A) shared intention normally leads to joint action, and that (B) it is normally accompanied by interpersonal obligations. Clearly, the preceding remarks constitute Bratman’s explanation of the former regularity, (A). But what does, in Bratman’s view, explain the latter regularity, (B)?

Bratman points out that we normally arrive at shared intention as a result of agreements, mutual assurances or promises, forms of reciprocally induced reliance, and so on, and that the latter “processes” normally generate moral obligations between participants (1999c, p. 125–129; 1999d). That is why, in his view, (B) holds true of shared intention. Bratman makes some further points about the connection between shared intention and interpersonal obligations that are relevant to our discussion. To begin with, Bratman points out that we may arrive at shared intention without having thereby acquired obligations to one another. He thinks that there are two kinds of circumstances where this is so. First, there are cases in which moral considerations prevent such obligations from arising. Perhaps the joint activity we intend to perform is morally impermissible; or, perhaps we arrived at shared intention through certain forms of coercion, deceit, or manipulation (1999d, p. 132–133). Second, there are cases in which we arrive at shared intention through processes that are not “obligation-generating” (1999d, p. 138–140). Perhaps I unilaterally formed my relevant intention that we act, you unilaterally formed yours, and the social context made it obvious to us that we each had such attitudes (1999d, p. 139). In addition, and more importantly, Bratman points out that the defining feature of shared intention is that it leads to joint action by playing the aforementioned coordinating and bargaining roles, and that mutual obligations are neither necessary nor sufficient for this. Such obligations are not necessary, given the remarks in the paragraph above; and they are not sufficient, since we may reciprocally incur obligations to one another by way of any of the aforementioned processes but be insincere and have no intention to comply with them (1999c, p. 111). All these points give us reason to conclude, Bratman maintains, that although “shared intention does typically bring with it associated obligations,” those obligations are “not essential to shared intention itself” (1999d, p. 132. The emphases are mine).

Bratman believes that a conception of the very nature of shared intention follows from the previous reflections. Since shared intention is a structure of intentions of individuals that typically plays some central roles in the individuals’ thought and action in pursuit of the joint activity, since interpersonal obligations are neither necessary nor sufficient to play such roles, and since shared intention does not necessarily create such obligations, Bratman concludes that shared intention is a psychological phenomenon, not a normative one. As he writes:

[S]hared intention… supports coordinated planning and action, and relevant bargaining, aimed at the joint activity and… it is typically but not necessarily accompanied by relevant… obligations. That seems… a reason to see the phenomenon captured by [the cited structure of intentions] as at the heart of the matter… [S]hared intention is primarily a psychological – rather than primarily a normative – phenomenon. The step to… obligations and entitlements is a step beyond this more basic phenomenon (Bratman 1999c, p. 128).

The just cited passage is definitely illustrative of Bratman’s conception of the nature of shared intention. However, a point needs clarification. In pointing out that “shared intention is primarily a psychological – rather than primarily a normative – phenomenon” (my emphases), Bratman is not saying that although shared intention is both a psychological phenomenon and a normative phenomenon, it is more of the former than of the latter. For Bratman shared intention is itself not a normative phenomenon at all.^[7]

2.2 Gilbert’s Normative View

Gilbert offers a picture of the nature of shared intention that contrasts sharply with Bratman’s. To begin with, the authors’ starting points of reflection are different. Whereas Bratman is impressed by the similarities – or “deep continuity” (2014, p. 4) – he sees between individual agency and shared agency, Gilbert is struck by the parallels she finds within the whole range of social phenomena. Gilbert thinks that all social phenomena – ranging from “collective mental states,” such as shared intentions and collective beliefs, to interpersonal transactions, such as promises and agreements – involve (what she calls) a “joint commitment” between the individual participants. Joint commitment, Gilbert writes, “may be regarded as the core of human sociality” (2000a, p. 4). Gilbert characterizes joint commitment in terms of several features, but three of them are central to our discussion. First, a joint commitment is not an aggregate of “personal” commitments – where a “personal” commitment is, in this view, the commitment characteristic of a “personal” decision, intention, or goal (2000c, p. 21). It is, rather, a primitive phenomenon. Second, unlike personal commitments, the joint commitment of two (or more) individuals can be created and rescinded only by those individuals together. One may rescind one’s personal commitment by changing one’s mind. But one cannot rescind a joint commitment in a similar way. Any unilateral decision to be rid of the joint commitment does not amount to its rescission, but to a violation of it (2000c, p. 21–22). Third, and most crucially, the mere existence of a joint commitment necessarily creates relevant obligations and entitlements between participants. “[T]hese obligations and entitlements,” Gilbert says, “derive from the joint commitment alone,” they “inhere in the joint commitment in the sense that once the joint commitment exists, they exist also” (2000c, p. 26. The emphasis is hers).

For Gilbert shared intention is clearly a social phenomenon, and as such, it necessarily involves a joint commitment between individual participants. It involves a joint commitment to (intend to) act together (2000c, p. 22). Since shared intention involves an underlying joint commitment, Gilbert explains, it necessarily creates reciprocal obligations and entitlements between participants. Each participant incurs an obligation to the others to conform to the shared intention – specifically, to do his or her part in the joint activity – and acquires a corresponding entitlement against the others to their conforming actions and also to rebuke them for nonconformity (2000c, p. 16–17, 25–26). Suppose that we are dancing the tango and I suddenly decide to stop and walk away without saying a word to you. In such type of case, Gilbert tells us, you are entitled to rebuke me for having failed to conform to our shared intention; you are so entitled in virtue of our joint commitment to (intend to) dance the tango. It is important to note how Gilbert conceives of such obligations. For Gilbert the obligations created by shared intention are not moral, but sui generis in nature – she also sometimes labels them as “associational” or “political” (1992, p. 411). They are grounded in the (value of the) “jointness” of its underlying joint commitment; or, as she puts it, they are “a function of the fact that the joint commitment is indeed joint” (2000d, p. 103). Moreover, Gilbert thinks that such obligations are unconditional (2000d, p. 104–105). This means in her view that “the obligations of joint commitment are present just in case a joint commitment is itself present… neither the circumstances nor the content of a joint commitment, given that it exists, affects the presence of its obligations” (2000d, p. 104. The emphases are mine). Thus, if we arrived at a shared intention to dance the tango in coercive circumstances – suppose I told you, “Dance with me or I will reveal your secret!” – we would still have incurred obligations to conform (2000d, p. 105). Similarly, if the content of our shared intention involved the performance of immoral actions – suppose we shared an intention to rob a bank together – we would still “have obligations of joint commitment to perform immoral actions” (2000d, p. 105).

The connection between shared intention, joint commitment, and obligation is central in Gilbert’s view. It is purported to explain not only why (B) shared intention is normally accompanied by interpersonal obligations,^[8] but also why (A) shared intention normally leads to joint action. Gilbert makes three central points about the explanation of joint action. The first two are negative. First, Gilbert points out that a group of individuals may share an intention to do something together, and yet lack corresponding “personal” intentions and similar attitudes of individuals (2000c, p. 17–18, 27). Second, she claims that “insofar as [shared intentions] do motivate, their motivational force does not derive from the motivational force of corresponding personal intentions” (2000c, p. 18). The third point is positive. Gilbert maintains that shared intention motivates by virtue of its ensuing obligations (2000c, p. 24, 27). Since obligations give, or constitute, (normative) reasons for action (1996b, p. 288; 2000d, p. 120, n.36), shared intention leads to joint action by giving each individual a reason to act accordingly (2000c, p. 24).

Now we are in a position to see how Gilbert arrives at her conception of the nature of shared intention. From the premises that shared intention necessarily creates interpersonal obligations, that these obligations are responsible for playing the motivational roles characteristic of shared intention independently of “personal” intentions and the like, and that such attitudes of individuals are not necessary for shared intention, Gilbert concludes that shared intention is, at its core, a normative phenomenon, as opposed to a psychological one.

4.1 A Basic Structure of Attitudes in Shared Intention

Let us begin by considering the cited basic structure of attitudes of individuals involved in shared intention. As I have argued elsewhere (2009), this structure includes in my view at least three basic elements. First, it includes intentions on the part of each individual. Here I follow Bratman (1999c) and conceive of such intentions as ordinary intentions of individuals in favor of the joint activity. When you and I share an intention to dance the tango, I intend that we dance the tango and you intend likewise. I think this conception most clearly captures the nature and extent of an individual’s commitment in shared intention, as compared to alternative proposals.^[25] Second, and this is central to my view, the cited structure includes attitudes of reliance on the part of each individual. Specifically, when we share an intention to dance the tango, I rely on your intending that we dance the tango and on your doing your part in our joint activity when the time comes; and so do you. I elaborate on this below. Finally, a third basic element in the structure is that the cited attitudes of intention and of reliance of individuals are common knowledge – that is, they are “public” – ^[26] between them.^[27]

Consider the second element of such a structure in more detail. It is a natural thought that when people do things together, they rely on each other in different ways. But why exactly are the cited forms of mutual reliance necessary for shared intention? Reliance gets into shared intention by way of its connection to the constituent intentions of individuals. It is a familiar cognitive constraint on intending in general that one may only intend what one takes one’s so intending as settling. One cannot intend to win the lottery if one does not take one’s so intending as settling the issue of one’s winning it, as being efficacious in causing that result. But how can my intention in favor of our activity satisfy the cited constraint? How can I, in intending that we act, take myself to be settling the issue of our acting?^[28] Bratman addresses this question in connection to his own account and provides an ingenious response. He argues that I can (coherently) intend our joint activity, and thus take myself to be settling the issue of our acting, as long as I can reasonably believe, among other things, that you intend (or will intend) likewise, and thus that you will be settling that issue as well (1999e, p. 148–160, 2014, p. 76). Bratman’s response, to put it in simpler terms, is that in order for me to intend that we act, I must believe that you intend (or will intend) likewise. I think that the structure of Bratman’s response is broadly along the right lines. But I also think that conceiving of the cited cognitive constraint on such intentions in terms of the attitude of belief is overly strong.^[29] Instead, we do better to conceive of it, I suggest, in terms of the attitude of reliance. But what is reliance? And how does it differ from belief?^[30]

Reliance is, like belief, a cognitive attitude, rather than a conative one. Both attitudes have it as a function to cognitively guide (or frame) one’s reasoning. Each of relying on p and of believing that p disposes one to deliberate, plan, and act on the basis of p when the conditions for such deliberation, planning, and action are satisfied. Suppose that I very much want to dance the tango with you at a party on Saturday. If under such circumstances I relied on your intending that we dance at the party, I would, for example, form my own intention that we dance, buy two tickets for the party, deliberate about how to get there, eschew alternative plans, and so on. The same would be true if in those circumstances I believed that you intend that we dance at the party. Such similarities notwithstanding, reliance differs from belief in many ways. A central difference here has to do with the norms of correctness for such attitudes. While belief is correct just in case it is true, reliance is correct just in case it cognitively guides one’s reasoning in a way that is instrumental to one’s relevant ends and values. This difference establishes a contrast with respect to the grounds on the basis of which such attitudes are justified. While belief can be justified solely by evidence for its truth, reliance can be justified on the basis of both evidential and pragmatic considerations; and while belief justification requires one to have sufficient evidence for its truth, reliance justification requires one to lack sufficient evidence for its falsity.^[31]

Since reliance is subject to a less stringent evidential constraint than belief is, conceiving of the cognitive constraint on intending the joint activity in terms of reliance allows us to make room for, and also make sense of, many ordinary instances of such intentions that are ruled out by the belief conception. It seems clear that I may lack evidence justifying the belief that you intend that we dance at the party (perhaps you are an indecisive person and have not yet responded to my invitation), but nonetheless rely on your so intending and form my own intention that we dance on the basis of such reliance. What is more, my reliance and intention may well be rational or justified in that case. After all, I may have good pragmatic reasons for so relying and intending (perhaps I very much enjoy dancing the tango with you and believe that my so relying and intending are means to securing the success of that activity) and my reliance may meet the cited evidential constraint (perhaps I lack evidence justifying the belief that you will eventually turn down my invitation). These remarks support the claim that reliance satisfies the cognitive constraint on intending the joint activity. But reliance is not only sufficient for satisfying such a constraint. It is necessary for it as well. This is because framing my intention that we dance by the belief that you intend likewise gives way to a corresponding attitude of reliance. In this context, my belief is not merely tracking the truth, but is primarily cognitively framing my intention that we dance the tango. But as soon as my belief starts playing this framing role, it gives rise to an attitude that can be appropriately assessed in terms of, and is typically guided by, its instrumentality in playing such a role; namely, it gives rise to my reliance on your intending likewise.^[32]

4.2 Shared Intention and Obligation

I have also argued in earlier work (2009) that once we look closely at the cited basic structure of attitudes of intention and of reliance of individuals, we see that shared intention necessarily involves a basis for interpersonal obligations. This is because the cited structure produces an effect of mutual reinforcement of the individuals’ attitudes of reliance and this effect helps create relevant moral obligations between such individuals. Let us consider this in some more detail. Once you and I share an intention to dance the tango, we each have the attitudes of intention and of reliance mentioned above, and these attitudes of each are public between us. This produces an important effect. In persisting in my intention that we dance in such a context of publicity about our attitudes, I thereby signal to you what I intend and will do in the future, and as a result normally reinforce your reliance on my intention and eventual actions. And so do you. The next point to note is that reinforcing reliance in others in this way is a basis for obligations. It is so for the same reasons that the more familiar activity of inducing reliance in others is such a basis. Following work by Scanlon (1998, 2003), we may say that in general others have good reasons for wanting to be protected against the losses they may incur as a result of relying on assumptions that we have reinforced or induced in them, and that those reasons normally ground moral obligations on us to prevent such losses – where these losses may include wasted time and resources, incurred opportunity costs, and so on. Given all this, it is not difficult to see why in sharing an intention to dance the tango in normal circumstances, you and I incur obligations to one another. Since in sharing such an intention we reciprocally reinforce each other’s reliance, and since in that context each of us has good reason to believe that the other would suffer reliance losses were one not to perform one’s part in the joint activity, we incur obligations to one another to prevent such losses. More specifically, we incur obligations to one another to either give the other a timely warning if one will not perform one’s part in the joint activity, perform one’s part, or (if one fails to take the previous two steps) compensate the other for reliance losses.^[33]

As anticipated, the fact that the aforementioned structure of attitudes is a basis for obligations helps us account for the normative aspect of shared intention. To begin with, this fact establishes, given the standard set in Section 3 above, that shared intention is a normative phenomenon. In addition, it offers a principled explanation of one of the regularities noted in Section 1 above, namely, that (B) shared intention normally brings with it obligations. For it is plausible to think that we normally share intentions in the absence of special circumstances, and we know that relevant obligations are created in such cases. Furthermore, the fact that the cited structure is a basis for obligations also accounts for the widespread presence in the context of shared intention of practices of holding responsible, which Gilbert has done much to emphasize.^[34] When we share an intention to act together, we are entitled, and have a corresponding tendency, to hold each other to account: I am entitled and disposed, inter alia, to demand that you comply with the obligation you have incurred to me, and to blame you if you fail to comply without excuse – where my blaming you includes typical moral reactions such as my resenting you for having violated such an obligation to me, my complaining to you for this, and my demanding an explanation from you; and similarly for you. Undoubtedly, the topic of our responsibility practices in shared intention raises large issues – including, for example, the issue of whether such practices take exactly the cited form. Nevertheless, what is important to note for our purposes is that the explanation of such responsibility practices will appeal to the fact that shared intention is a basis for the obligations for which we hold each other to account. It is precisely in supporting such responsibility practices where the significance of shared intention’s being a basis for obligations mainly resides.^[35]

In conceiving of shared intention as intimately connected to obligations and related normative phenomena, my view aligns with Gilbert’s against Bratman’s. We saw that for Bratman different bases for interpersonal obligations – or, as he calls them, different “obligation-generating processes” – are normally but yet contingently associated with shared intention. This indicates that in Bratman’s view it is a contingent fact that regularity (B) holds true of shared intention. I think, on the contrary, that it is not a contingent fact, due to the presence of some basis for obligations external to shared intention, that this phenomenon normally generates obligations. Rather, in my view a particular (though defeasible) basis for obligations is internal to, or partly constitutive of, shared intention itself.^[36] Furthermore, Bratman and I explore the question of the connection between shared intention and interpersonal obligation with an eye mainly to different issues. Bratman’s main concern is whether such obligations are responsible for playing the motivational role we commonly attribute to shared intention. I am also concerned with this issue, and will say more about this below. However, my main concern in exploring the cited question is of a more normative sort. It has to do with whether there is a basis for obligations in shared intention and, if so, with the intrinsic normative significance of this – where this includes, especially, the support of responsibility practices in this context.

Alternatively, although I join Gilbert in thinking that shared intention is non-accidentally connected to interpersonal obligations, I part company with her on that I do not think that this non-accidental connection takes the form of a necessary connection between shared intention and such obligations themselves. In my view the obligations characteristic of shared intention are created only in the absence of special circumstances, rather than necessarily. They are moral obligations, rather than “associational” ones. In these last two respects, my view aligns with Bratman’s against Gilbert’s. But there is another contrast with Gilbert’s view worth stressing here. In appealing to considerations such as the presence of attitudes of reliance of individuals in shared intention, the reciprocal reinforcement of such attitudes by those individuals, and the value to such individuals of their being protected against the losses they may incur as a result of such reliance, I have attempted to provide a simple and informative normative story of why shared intention is a basis for obligations. It is not at all obvious, however, that this is also the case with Gilbert’s view. All Gilbert seems to be offering in terms of a normative story is the aforementioned claim that the obligations characteristic of shared intention are generated by an underlying joint commitment and that they are “a function of the fact that the joint commitment is indeed joint” (2000d, p. 103). But this claim only pushes the problem one step back. For now we would want to know what it is about the “jointness” of a joint commitment that creates the relevant obligations, and about this question Gilbert seems to remain silent.

4.3 Shared Intention and the Explanation of Joint Action

I want to argue next that the cited basic structure of attitudes of intention and of reliance of individuals involves dispositions to play relevant functional roles that lead to joint action and that this fact helps us account for the psychological aspect of shared intention. First, this fact establishes, given the standard set in Section 3, that shared intention is a psychological phenomenon. Second, it helps us explain why the other regularity we noted in Section 1, namely, that (A) shared intention normally leads to joint action, is not an accidental feature of this phenomenon. In relation to this latter issue, I want to suggest, in line with Bratman’s approach, that appeal to the psychological roles played by the constituents of shared intention is an ineliminable part of the explanation of joint action. But I also want to suggest that this is not always sufficient. The moral obligations shared intention characteristically generates make a difference as to how this phenomenon leads to joint action. And I think that a theory of the basic aspects of shared intention should account for this, since the creation of such obligations is one of the two central features of paradigmatic cases of shared intention, as suggested in Section 3 above.^[37] So, when it comes down to the question of how shared intention leads to joint action, I think it is important to distinguish between two types of cases: to wit, (i) cases in which shared intention does not generate its characteristic obligations, and (ii) cases in which it does.

Consider type (i) cases. Imagine that you and I share an intention to rob a bank together. Since robbing a bank is a morally impermissible action, our having formed such a shared intention does not create relevant obligations between us. But even though this is so, I want to suggest, our shared intention may still guide our thought and action in ways that lead to our successfully robbing the bank together. My suggestion is that in cases like this our attitudes of intention and of reliance will lead to joint action much in the way Bratman has insightfully described, that is, basically, by being appropriately interconnected and responsive to their associated norms and, consequently, by inducing dispositions to play and then by typically playing characteristic coordinating and bargaining roles. Surely, more should be said as to what exactly such interconnections and norms are as well as to how they cooperate with the cited attitudes to play the aforementioned roles.^[38] But I think nonetheless that the previous remarks provide us with an outline of how the motivational story will go. That being said, let me note a difference – though, perhaps only of emphasis – with Bratman’s view. In my view, shared intention involves attitudes of reliance of individuals and these attitudes, I want to stress, occupy a central place in the explanation of how shared intention plays the aforementioned coordinating and bargaining roles. Return to our shared intention to rob the bank together. It is partly because I rely on your intending to reduce the guards and to keep an eye on the hostages that I intend to open the safe and to put the money in the bags; and it is partly because I rely on your following through with your intentions that I effectively execute mine. And vice versa. Thus, it is partly because we rely on each other in those ways that our shared intention coordinates our planning and action in ways that lead to our successfully robbing the bank together.^[39]

What to say about type (ii) cases? These, recall, are paradigmatic cases of shared intention. Here we must go beyond an appeal to the functional roles played by the attitudes partly constitutive of shared intention. In type (ii) cases, individuals incur relevant moral obligations as a result of their shared intention and this gives each of them a special reason to act accordingly, a reason they would not have in type (i) cases.^[40] Further, if such individuals accept a relevant moral principle or norm grounding the cited obligations and endorse the cited reason for action, this will typically have consequences for the way in which their shared intention coordinates their planning and action, and structures forms of bargaining between them, in pursuit of the joint activity.^[41] One of those consequences is that the individuals’ motivation will be reinforced as a result of their having incurred such obligations. Whatever their initial reasons for participating in the joint activity are, such individuals will now be moved in addition by the aforementioned moral reason. A related consequence is that individuals will stick more firmly to their shared intention than they would otherwise have, which will make their shared intention more stable as a result.^[42]

This account of the motivational role of shared intention contrasts sharply with the one offered by Gilbert. Although Gilbert and I agree that the obligations shared intention generates make a difference as to how it leads to joint action, we disagree about the difference they make. This connects to a deeper disagreement between our accounts, concerning the motivational role played by attitudes of individuals. As we saw in Section 2, Gilbert denies that an appeal to attitudes of individuals is necessary to explain how shared intention motivates. She claims that an appeal to the obligations individuals incur in shared intention is sufficient to explain how this phenomenon leads to joint action, independently of intentions of individuals, their attitudes of reliance, and so on (2000c, p. 17–18, 2009, p. 171–173). However, this claim raises many worries, one of which is that in order for such obligations to motivate individuals to act, they must somehow engage the individuals’ psychologies, and it is not obvious how this is supposed to work in Gilbert’s view. There are difficult questions here as to whether the cited obligations will induce dispositions to comply with them and whether such dispositions, if so induced, will be embedded in intentions of individuals or the like. Consequently, it is not apparent in Gilbert’s view why it is not a contingent fact that (A) shared intention normally motivates.^[43] Another worry with Gilbert’s claim is that there are cases of shared intention in which it is intuitively clear that no obligations are created between individuals, and yet such individuals are disposed to think and act in ways conducive to joint action – these we called earlier, recall, type (i) cases. Gilbert might respond to this worry by insisting that reciprocal obligations are still created in such cases of shared intention but that these obligations are “associational,” rather than moral, in kind. Our intuition that no obligations arise in such cases, Gilbert might contend, owes to the fact that in thinking about such cases we have a concept of obligation of the wrong kind in mind. Be that as it may, this response would bring us back to the normative question as to how exactly the purported “associational” obligations arise in the first place, a question for which, I suggested earlier, Gilbert fails to provide a satisfactory answer.

4.4 Centrality, Independence, and Interconnection of Explanatory Factors

I have argued in this section that shared intention involves a basic structure of attitudes of intention and of reliance of individuals, which includes both dispositions to play relevant psychological roles and a basis for interpersonal obligations. Now we can see why the included two factors have, as anticipated in Section 1, central explanatory significance for shared intention. First, such factors establish that shared intention is, respectively, a psychological and a normative phenomenon; and, second, appeal to such factors is necessary to explain, respectively, why (A) shared intention normally motivates and why (B) it is normally accompanied by obligations. Furthermore, we can see why such factors can have, as also anticipated, independent explanatory significance for shared intention as well. We may see this by reflecting on some non-paradigmatic cases of shared intention, in particular, on cases in which shared intention is successful in leading to joint action but not in creating its characteristic obligations, and vice versa. Earlier we considered non-paradigmatic cases of the former type (type (i) cases), and saw that the explanation of how shared intention leads to joint action in such cases appeals to relevant dispositions to coordination and bargaining of the constituent attitudes of individuals, which are independent of the obligations that such attitudes typically contribute to generate. Now consider the latter type of non-paradigmatic cases of shared intention – call these type (iii) cases. Imagine once again that you and I share an intention to dance the tango in a context in which the usual defeaters of moral obligation are absent: none of us has coerced the other into participating in it, what we intend to do is morally permissible, and so on. It follows from the view defended here that in such a context you and I will have incurred obligations to one another to act accordingly. It also follows from that view that each of us intends that we dance the tango and relies on the other in important ways. But now imagine, in addition, that although I have the just cited attitudes, I am not moved to act accordingly. Perhaps my intention that we dance the tango is overpowered by the force of an incompatible intention of mine, say, the intention that your friend and I dance the tango. Or perhaps I am just paralyzed by the realization that you will soon discover my poor dancing skills. In such a context, our shared intention may fail to lead us to dance the tango – since, among other things, it may fail to coordinate our actions – but I may still be obligated to you to act accordingly, and you may appropriately demand that I so act.^[44] Finally, it is worth noting that although the cited two factors – dispositions to play psychological roles and a basis for obligations – can be explanatorily independent, they also interconnect in an important way. As shown above, they interconnect in the explanation of how shared intention motivates in paradigmatic cases (type (ii) cases).

As discussed in Section 2, Bratman and Gilbert treat relevant psychological roles and interpersonal obligations as competing explanatory factors, especially when it comes down to the question of how shared intention motivates. For the authors, appeal to one of those factors, but not to the other, is necessary and sufficient for explaining how shared intention leads to joint action. I take a different approach. I see psychological roles and interpersonal obligations not as competing explanatory factors but as aimed at explaining, in the first instance, different regularities or facts about shared intention, the former how this phenomenon motivates and coordinates joint action, the latter how it obligates participants to one another and supports responsibility practices between them, all the while recognizing that such factors may interconnect in the explanation of the former regularity as well.

At this point it might be argued, however, that since what constitute a basis for interpersonal obligations in shared intention, according to the view I have proposed, are psychological features of this phenomenon – namely, a set of attitudes of individuals that normally produces an effect of reinforcement of the individuals’ attitudes of reliance – in this view the psychological aspect of shared intention is in the end more crucial to the phenomenon than its normative aspect. I think this conclusion would be unwarranted. Surely, it is correct to say that in this view the psychology of shared intention has ontological priority over its normativity. But it does not follow from this that in this view the former aspect also has explanatory priority over the latter – that is, that the former aspect is explanatorily more significant for understanding the nature of shared intention than the latter is. Once again, the parallel with promising is instructive here. Return to Scanlon’s account mentioned in Section 4. Normally, in promising you to act in a certain way, I assure you that I will so act – where this involves my having voluntarily and intentionally led you to believe that I would so act – and you desire this assurance. Once I have assured you in that way, the value to you of your being so assured is engaged and helps to ground an obligation on my part to so act. It is natural to infer from this that on Scanlon’s account the psychological aspect of promising has ontological priority over its normative aspect. However, it does not follow from this that on Scanlon’s account the former aspect has explanatory priority over the latter as well. Ontological priority need not entail explanatory priority for interpersonal transactions such as promising and shared intention. That being said, I want to stress that my main concern in this article has been to argue for the explanatory significance of both the psychological and the normative aspects of shared intention. Certainly, there is an interesting question as to whether any of those aspects is explanatorily more significant than – that is, has explanatory priority over – the other. But the view I have proposed here is silent on this latter question.