Isaacs, Tracy (2011). Moral Responsibility in Collective Contexts. Oxford: Oxford University Press.
My main interest in collective action contexts is normative. I advocate a two-level theory of moral responsibility in which there is moral responsibility at both the individual and the collective levels. The normative conclusions I wish to establish rest on the claim that intentional action also operates on two levels, the individual and the collective. I understand intentional collective action as the intentional action of collective agents. In this chapter, I outline my view of collective agency by distinguishing between two types of collective agents—organizations and goal-oriented collectives—and giving an account of their respective intentional structures. My purpose is to establish that collectives may have intentions and are capable of intentional action.
Where moral agency is concerned, we can roughly distinguish between two species of views, those of the individualists and those of the collectivists.
| Individualists | Collectivists |
| Individualists believe that agency and responsibility are ascribable only to individual human agents and that claims about collective agency and collective responsibility are reducible to claims about the agency and responsibility of individuals. | Collectivists subscribe to a less reductionist and more holistic view, arguing that claims about collective agency and responsibility do not reduce to claims about the agency and responsibility of individuals. |
For both action-theoretic and normative reasons, I claim that collective agency and collective moral responsibility operate at a different and indispensable ontological and explanatory level from individual agency and individual moral responsibility. According to my view, claims about collective moral responsibility neither entail nor are derivable from claims about individual moral responsibility. Furthermore, collective intentions, from which the agency of collectives derives, are not simply collections of individual intentions, and collective actions are not simply collections of individual actions. Collective actions are the products of the intentions of collectives. My goal in this chapter is to explain my account of collective agency. I begin with a brief discussion of types of collectives.
1.1 Types of Collectives
1. organizations Highly structured collectives, which I call organizations, are the most obvious and least contestable candidates for moral agency. Corporations, nonprofit groups, nations, universities, departments of philosophy, and professional sports teams are all examples of organizations. Of the four examples I outlined in my introduction, only one—the Canadian Red Cross—clearly falls into this category. The perpetrators of genocide in Rwanda, contributors to global warming, and participants in oppression are not clearly members of organized groups. If the Rwandan genocide was perpetrated by a collective, it was not a collective we can rightly call an organization. The reason we cannot is that although there was a common goal and a coordinated effort, there was not an organizational structure of the sort possessed by the collective that perpetrated the Holocaust.
2. a goal-oriented collective A different kind of collective from an organization, one that fits the profile of the collective that perpetrated the Rwandan genocide, coalesces [더 큰 덩어리로 합치다] around action toward the achievement of a particular joint goal. I call this kind of collective a goal-oriented collective. It is important that the goal be a joint goal, the obtaining of which requires, or at least is pursued through, collective action. Goal-oriented collectives do not need to be as enormous as the one that perpetrated the Rwandan genocide. Two people going for a walk together, three people painting a house together, a thousand people doing the wave at a sports event, or tens of thousands of people “Running for the Cure” all constitute goal-oriented collectives.
3. a collective lacking a specific goal Global warming might be best construed as a harmful cumulative consequence of noncoordinated parallel actions. As such, it might well generate collective and individual obligations. I argue as much in chapter 5. But as a consequence of parallel actions, we have no moral or ontological reasons on the basis of which to attribute it to a collective entity.
In the case of oppression, it is tempting to attribute its existence to the behavior of privileged social groups, the members of which are usually beneficiaries of practices of discrimination against others. For example, some like to condemn wealthy white North Americans, especially men, for systemic injustice against the poor or women or members of nonwhite racial groups. I hesitate to identify these kinds of groups as agents. There are compelling arguments, well-represented in feminist scholarship, detailing the importance of recognizing the complex network of privilege and disadvantage at work in oppressive social contexts—we are all members of a multiplicity of social groups, some of which define our privilege, others of which yield disadvantage. Identities are, above all, intersectional. Carving out social groups on the basis of categories that typically delineate privilege and disadvantage is not, therefore, an especially helpful way of isolating collective entities. For this reason, I do not address social groups as such in my analysis of collective responsibility and collective agency.
My view does not, however, require that they not exist. It is consistent with my view that within types of collectives we might include the rough category of aggregates. Aggregates are collections of people that are grouped together because they share some common characteristic. We can think of social groups as a subcategory within aggregates. Later on, I return to this idea of aggregates when I consider what sorts of collective obligations might exist for addressing seemingly insurmountable global problems such as global warming, oppression, poverty, and starvation.
4. a mereological sum The final type of collective entity, which I mention only to set aside, is the type that is nothing more than a mereological sum. A mereological sum is the notional composite of any random collection of things, not necessarily connected in time or space or in any principled way around any purpose and not organized by any structural features or common characteristics. I, my third-grade teacher, and the queen of Denmark are an example. There are no candidate features in a mereological sum of this kind to suggest agency, though of course it is possible that I, my third-grade teacher, and the queen of Denmark could come together as a goal-oriented collective.
Organizations and goal-oriented collectives share a common feature in virtue of which they have agency: a collective intentional structure that gives rise to collective intention and collective action. Their intentional structures qualify them as moral agents insofar as the structures enable them to act intentionally. It is worth noting that the claim that moral agency rests on the capacity to act intentionally does not require that agents are responsible only when exercising that capacity. Much as rational agents have the capacity to act rationally even when acting irrationally, moral agents are capable of acting intentionally even when they are not. Thus, while the capacity for intentional action is important to account for moral agency, the scope of moral agency goes beyond those cases in which agents are acting intentionally. This distinction is important for explaining how some agents are responsible for negligence or omissions. These sorts of cases often involve behavior that is not intentional.
1.2 The Intentional Structure of Organizations
Understanding the Canadian Red Cross as an organization, we can assume structures in which the organizational roles and authority structures are outlined and the organization’s policy—including its mission and goals, as well as procedures for making organizational decisions and for the organization taking action—is articulated. The most familiar articulation of this idea may be attributed to Peter French in his widely read account of corporate internal decision structures.
These structures yield a level of intentional action that is distinct from the intentional action of the individuals who perform their organizational roles. The collective level is distinct, insofar as the organization’s actions flow from its intentions. The individuals’ actions flow from their intentions, and their intentions are not constitutive of the intentions of the organization, even if the individuals’ actions are at least partly constitutive of the organizations’ actions. In the organization case, there is a sharp disconnect between the individuals’ and the collective’s intentions (much sharper than in the goal-oriented collective case, as will shortly be explained).
French: “when the corporate [organizational] act is consistent with, an instantiation or an implementation of corporate [organizational] policy, then it is proper to describe it as having been done for corporate [organizational] reasons, as having been caused by a corporate [organizational] desire coupled with a corporate [organizational] belief and so, in other words, as corporate [organizational] intentional.”
The most significant feature of this analysis is that in order to understand the organizational intention there is no need to refer to the intentions of individuals. The content of their intentions is entirely beside the point. If organizational decisions are taken in the form of votes on motions, for example, it may well be that individuals whose roles require them to participate in the decision cast their votes on the basis of personal reasons. Nonetheless, the decision is the organization’s intention, irrespective of individuals’ reasons for voting as they do. The collective action that follows is the product of the collective’s intention.
It is theoretically possible that an organization might intentionally pursue a course of action that is not the action that anyone in the organizational structure intended that the organization pursue [???]. Compromise is frequently the rule in group decision-making, mandating courses of action that would not be pursued if any one individual were in a position to make an executive decision. Even in cases in which an individual is authorized to make executive decisions, such that this individual’s decisions are the organization’s, the decisions she or he makes qua the organizational role are not always consistent with decisions that would reflect the pursuit of her or his personal interests.
The artistic director of the Stratford Theatre Festival might, for example, have carte blanche to select the full slate of plays for the next season. Assume she has a low opinion of Broadway musicals and would not include them at all if she were at liberty to stage only what she liked. Nonetheless, the festival intends to run at least two such productions each season. They always sell out, and it is part of the festival’s mission to o er a mix of Shakespearean productions and more accessible, crowd-pleasing works, recognizing that there is not always overlap between these two categories. The director’s executive decision is constrained by the goals and mandate of the organization that is the theatre company regardless of the personal tastes of the artistic director. The decision the artistic director makes may legitimately be redescribed as the festival’s decision. However, the director’s intentions at the individual level are not redescribable as those of the festival.
Many who work in institutional settings do not have any sense of investment in what the institution is about, yet as they perform their roles their actions contribute to the actions of the organization. Requiring commitment to a shared goal overstates what is required for collective action to take place in an organization. The policies, procedures, mission, role definition, and structures of authority are what produce intentional action in these settings. Particular members might be completely alienated from the collective goals—perhaps they perform their function just for the paycheck or because they enjoy exercising the specific skills required by their job—yet their actions within their roles might be constituents of the collective actions of the organization.
1.3 The Intentional Structure of Goal-Oriented Collectives
Recall that a goal-oriented collective is a collective whose members come together around the achievement of a particular goal. That goal might be long or short term. Its achievement might involve significant planning or virtually none. The collective might have very many members or just two. The perpetrators of the Rwandan genocide were a goal-oriented collective; you and I making dinner together are a goal oriented collective. In recent years, the majority of the philosophical literature about collective, joint, or shared intention and action has had what I am calling goal-oriented collectives as its focal point. One reason for this focus is the recognition that many cases of collective action that appear to merit moral analysis do not involve highly structured organizations. Another reason for this focus is that the relevant collective actions lend themselves quite readily to an individualist analysis. A number of theorists maintain, for example, that the relevant locus of analysis for these sorts of collective endeavors—from a two-person case of going for a walk together to a many-person case of genocide—is individual intention. Some individualists propose to locate the collectivity in the content of individual intention—in the commitment to a joint goal or the individual’s intention to participate in the collective action aimed at achieving the goal. Under this analysis, a collective intention is not the intention of a collective entity but rather an individual’s intention, albeit one with collective content. If the analysis stopped there, individual intentional agency would be the only real agency there is, and any meaningful attributions of moral responsibility would remain at the level of individuals. These accounts are instructive [유이한], but I do not believe the analysis should stop there. Rather, we should understand the collective intentions of goal-oriented collectives as the collectives’ intentions.
Seumas Miller: distinguishes between “the logical priority of individual actions over joint actions” and the, at times, explanatory priority of collective ends and interests. Collective ends might contribute causally to the formation of individual ends and motivations. Thus, Miller identifies himself as being “in some sense” a holist and notes, further, that his view is “collectivist in complexion” because it acknowledges that both individual and collective activity depend in significant ways on social forms. For all that, Miller’s theory is fundamentally individualist, insofar as the central analytical concept, the collective end, is an individualist notion; a collective end is a species of individual end that “exists only in the heads of individual agents.”
Raimo Tuomela: draws a distinction between conceptual and ontological issues, noting that his theory “is conceptually rather anti-individualistic, even if ontologically it is, if not individualistic, at least interrelationistic and eschews collective agents and actions in a literal ontological sense.” He maintains that “in this account, a group is viewed as a collectively constructed agent that can have goals, beliefs, and so on and that can act, although that is not literally true.” Of the three views I have noted as hybrid, Tuomela’s instantiates the most liberal blending of seemingly irreducibly collective notions (particularly joint intentions characterized by the “we-mode”) and the assertion that none of this elaborate apparatus should be taken as literal.
Tracy Isaacs: committed to a more collectivist than hybrid view; I see no reason to deny the reality of goal-oriented collectives as agents with intentional structures resulting in collective action. They are different, however, from organizations. My purpose is to render plausible the idea that goal-oriented collectives have intentions and intentional structures that transcend the intentions and intentional structures of their members.
We should think about goal-oriented collectives as unities of a special kind. Insofar as the whole is greater than the sum of the parts, we might consider them to be organic unities or organic wholes.
1. One reason for thinking this is simple physical possibility: together we can perform actions we cannot perform alone. I cannot go for a walk together alone. I cannot play street hockey alone. I cannot perform orchestral works alone. Only collectives can be the agents of these actions. Where goal-oriented collectives are concerned, their goals are fundamentally joint. For example, doing the wave at a sports event requires the cooperation of many others, all of whom share the goal.
2. Beyond physical possibility, we need also to recognize that pooling our resources does not just create possibilities for coordination. In many cases, our interactions take us places where no one of us would even think to go alone. For example, two people writing a paper together, as a group, might well end up with a conclusion that neither would have devised on her or his own. Their collective effort yields a result quite different from what they would have come up with independently of each other.
3. A more elusive but arguably significant feature of collectives is the quality of the “chemistry” among the members. Any teacher who has taught the same course several times knows that in some years it is a great success, whereas in others it falls at. This has less to do with any particular member of the class than with how the members combine. One year the class is energetic and enthusiastic. The following year’s group is lethargic and introverted. The goal is the same, but different collectives have di erent characters or personalities.
4. In addition to these considerations, there are facts about identity and endurance through changes of membership. Many goal-oriented collectives—for example the perpetrators of the Rwandan genocide, the people who do the wave at a sports event, or the participants in the “Run for the Cure” fundraiser for breast cancer research—have the capacity to withstand signi cant changes in membership without their identities being compromised. If there is a strong wave going around the stands and I decide to sit it out when it comes to my section because I am playing with my new smart phone, there is no good argument for establishing that it is a different wave from the one it would have been if I had participated. The success of a particular instantiation of the wave phenomenon does not stand or fall with the participation of any particular contributor.
I make these points as a means of suggesting that goal-oriented collectives are in several ways more than the sum of their parts. This thesis does not require the absurd claim that the parts are irrelevant. It requires only that the manner in which the parts come together is as significant as the parts themselves. I do not believe this way of understanding this kind of collective amounts only to a useful fiction or should be thought of conceptually as merely a manner of speaking. It has metaphysical substance suficient to ground moral agency.
1.3.1 Collective Intention in Goal-Oriented Collectives
I favor an understanding of collective intention as a state of affairs consisting of a complex of appropriately constrained individual intentions, the relationships between them and to the joint goal, and the individuals’ understanding of themselves as standing in relation to others as members of a group in pursuit of a joint goal. Understood this way, the collective intention has independent standing that may, in some cases, constrain how individuals may legitimately participate in the collective activity. Michael Bratman suggests that a shared intention consists “primarily of a web of attitudes of the individual participants” and I concur with the gist of this idea. Nonetheless, our respective views rest on different sides of the individualist/collectivist divide. Bratman understands himself to be o ering a view that is “broadly individualistic in spirit.” I consider my account to be more collectivist, both in spirit and in substance.
The relations between individuals’ intentions and the way individuals understand themselves as standing in relation to others as members of a group are essential features of collective intentions, without which the collective action would not come about. Just as the collectives themselves ought to be understood along the lines of organic unities, so should their intentions be understood similarly. They flow from the collectivity of the collective. They are not simply aggregates of individual intentions. The way the individual intentions combine and the way members of groups understand themselves as members of groups give rise to collective intentions that, in turn, result in collective actions. It would be a mistake to think of this view as reductionistic or as individualistic. It recognizes the collective level as a di erent, though dependent, level of action.
Furthermore, this view does not require that the intentions of goal-oriented collectives are collective mental states or states of consciousness, any more than the intentions of organizations are these things. Some deny collective intentions on the grounds that intentions must be mental states. If that were the case, then the very idea of collective intentions would require a collective mind to house them. For example, Miller claims that intentions must exist in the heads of agents. This gives him an easy way to deny the existence of truly collective intentions: since collectives do not have heads in which intentions may exist, there are no collective intentions. Alternatively, however, we may understand collective intentions as states of affairs, identifiable in part by their functional roles. As long as they function at the collective level of action in the same way that individual intentions function at the level of individual action, then we may think of them as intentions.
The first element in the analysis of collective intentional action in goal-oriented collectives is the collective goal. The collective goal is simply the end at which the collective aims. The conception of a collective goal necessarily includes reference to its being intended, because there are no collective goals in the abstract, dissociated from the collective that intends to bring them about. For example, when we intend to make dinner together, our collective goal is to make dinner together; when the audience at the hockey game intends to do the wave, the collective goal is to do the wave; when a group of people intends to commit genocide, the collective goal is to destroy a particular ethnic, racial, or religious group because of their ethnicity, race, or religion; when we intend to dance the tango, our collective goal is to dance the tango; and so on. When we intend something collectively, there is an understanding that we, as a unity, intend to bring it about together. Some collective goals can only be brought about collectively; collectivity is inherent in our understanding of the nature of the action required to achieve them. In the foregoing list, only the goal of genocide is, potentially, not this kind of goal...The important feature of it is that the members embrace it as the collective’s goal. In so doing, they understand their contributions in the context of the collective goal.
1.3.2 Can Individual Participants Intend Collective Actions?
The goal of a collective should not be understood strictly as something any individual member of the collective can intend. The proper subject of a collective intention with a collective goal as its object is a collective agent. Individuals’ intentions oriented toward the collective’s goal are to be understood as intentions to participate in collective action. Kutz captures this idea in his notion of participatory intentions, which he explains as intentions to participate in joint action. Tuomela’s notion of a we intention is also, at bottom, the notion of an individual’s intention to participate in joint action. He describes it as “in a non-literal sense—a participant’s ‘slice’ of the joint intention, and an aim-intention based on a group reason.”
Tuomela draws a distinction between action-intentions and aim-intentions. If I have an action-intention about something, then I must think it possible for me to perform that action. Realistically, then, I can only have action-intentions about my own, individual actions—these might include contributions to collective actions, but they cannot include collective actions themselves. This condition on intention captures the intuition that what we intend must be, in some sense, under our control. I cannot, for example, intend to paint the sky blue.
Where aim-intentions are concerned, agents do not have to believe they can bring about the aim by their actions alone; “rather the agent is assumed by his actions to contribute to the aimed result.” The distinction appears useful at first, since it acknowledges the intuition that the intentions of an individual are constrained by what it is possible for that individual to achieve through her actions—such intentions would be action-intentions. Adding the notion of an aim-intention is a bonus of sorts, since it allows individuals to intend (in some sense) actions and outcomes that are not in their hands alone; they require a collective effort.
Where we-intentions are concerned, they are also informed and formed by the intentions of the collective (in Tuomela’s technical vocabulary, these are joint intentions). But stipulating a second kind of intention—an aim-intention—is not sufficient to make it the case that individual agents, even those with we-intentions, can intend group aims. Tuomela’s characterization of we-intentions as intentions to participate in joint action enables us to understand them as action-intentions with a limited scope, insofar as individuals are able to intend their contributions to collective actions.
Bratman has suggested the use of the locution “I intend that we J,” where “J” stands in for a joint action, in order to capture the intentional relationship between individual intention and a collective action. The trick of course is to explain, whether in the case of an intention to or an intention that, how an individual can intend matters that are partly up to other people. On my view, the solution to this difficulty is simple: she or he cannot. Individuals can, however, intend their own part. This is not to say that collective actions cannot be intended at all. Collectives can intend collective actions: we can intend to make dinner together; they can intend to play street hockey. My view differs from those who would stop at intentions to participate because I maintain that there is a real sense in which collectives can and do intend to do what they do. I deny that a fully reductive account of collective intentions to the intentions of individuals can be given. We should not think of them as something that distributes neatly among individual members of collectives.
Members’ intentions play an important analytical role in the intentions of the whole. When we intend to do something together, then each of us individually has a commitment to the collective goal and an intention concerning his part of what we intend to do. The intention of each is not, however, a collective intention. It is an individual intention with collective content. These individual intentions with collective content are components of the collective intention.
Another consideration in discussions of collective intention is the extent to which common knowledge of and/or mutual responsiveness to the intentions of other members is required. As the example just mentioned shows, the intentions of each in a coordinated, joint action respond to the intentions, or at least the perceived intentions, of the others. It is not enough simply for me to intend that we do the wave or even for me to intend to do my part in the wave. Rather, a good portion of those in the stands must also intend it and must be aware of and responsive to the intentions of those around them if the thing is going to happen at all.
Note that it is not the case that I cannot intend to embark on[착수하다] my part of a joint venture without your intending to do your part and without believing that our respective intentions will result in action. Clearly, we do not want to rule out that possibility. But in cases such as those, when my intention to participate in a collective action does not take into account the intentions of other participants in any way, then the conditions for collective intention have, arguably, not been met—at least by me. We may think of these conditions and the condition of common knowledge as being strongly or weakly met. Recognizing a continuum in this way allows space for degrees of collectivity. Not all collectives are equally cohesive [화합하는]. Some are “tighter” than others; this tightness will be reflected in the strength of the collective intention. With this idea in mind, I now consider the role of common knowledge in collective intention.
1.3.3 Common Knowledge
This sense of oneself as a member of a group, standing in the relation of comember or coparticipant with other members of the group, is a signi cant factor in establishing conditions of collectivity as opposed to merely parallel action (even when such parallel action has a cumulative result).
Conditions of common knowledge are frequently cited as a basis for collectivity. As Bratman points out, common knowledge provides some cognitive linkage between the participants—they are aware of the intentions of others as coparticipants in a joint venture. Accounts vary as to what the common knowledge must be knowledge of and as to how informed people must be. In Bratman’s account, each knows of the others’ intention “that we J,” where J specifies a particular collective action. On other accounts, each must be aware of others’ intentions to do their part to achieve the collective goal. Failures of awareness yield the result that individuals might be engaged in parallel, but not collective, action.
Kutz offers a more minimalist account, arguing that the only requirement for collective action is that “the members of the group overlap in the conception of the collective end to which they intentionally contribute.” He offers a normative reason for the more minimalist account, pointing out that “ethically complex cases of joint action rarely involve perfect common knowledge, wholly shared conception of the joint act, or highly responsive strategic interactions.” I agree with Kutz that full common knowledge is too stringent a demand. I am not moved to this view by Kutz’s examples, however, which are not, according to my view, examples of collective action.
The first kind of example he offers is of an agent who, as Kutz describes, begins to do what will become part of a collective action if others follow his lead. This agent may have no expectations about others’ participation. His action may go nowhere, or it may start something—either what he hopes to start or something entirely different. I maintain that actions of this sort are best described in terms of efforts to begin collective actions, not strictly parts of collective actions. Returning to my example of the wave, failed attempts to start waves at sports events abound. One or two people jump up and cheer, trying to get everyone else going, and no one joins them.
Kutz’s second example involves a spontaneous activity: two people saving a picnic from the rain. One person grabs the sandwiches and heads for the car, intending to do his part to save the picnic, and hoping that his friend will grab the drinks and the blanket. If the friend comes through, says Kutz, then it is reasonable to claim that they have jointly intentionally saved the picnic. But actions taken independently of each other in this manner do not amount to jointly intentional action, even if they resulted in saving the picnic. As I have said, an agent’s conception of herself or himself as standing in relation to others and working together toward a collective goal is an important feature of collective action. It would not even be enough if just one of the picnickers thought they were acting together. For example, one might mistakenly believe the other heard him shout “Grab the drinks and the blanket and run to the car.” In order to get collectivity into the picture something slightly di erent needs to happen.
There needs to be some mutual understanding—it could even be implicit, based on nothing more than one catching the other’s eye and gesturing toward the drinks and the blanket. But if each party acts independently, then they do not act as a unit. Despite my thinking that Kutz’s examples lack collectivity, I fully agree with him that requirements of common knowledge need to be more closely examined and ultimately relaxed.
However, in a multiperson case it can be much more di cult. Consider a multilocation fundraiser involving thousands of participants across the country, such as “The Run for the Cure” and the many other fundraisers of this kind (the MS Bike Tour, the Parkinson’s Walk, the Terry Fox Run). It is surely impossible that anyone knows all of the other participants or even who all of the other participants are. Does that mean that there is a failure of the common knowledge condition and that, therefore, no collective intention to fundraise and no collective action takes place?
I believe that it would be a mistake to think of the common knowledge condition in this way...Rather, the issue turns on the scope and content of the common knowledge claim. What are the various participants required to know in terms of who else is involved and what their intentions are? If the goal of the common knowledge requirement is to provide enough cognitive linkage to secure collective, rather than parallel, action, then it ought to be enough to know, or at least have good grounds for believing, that others embrace the collective goal, see themselves in relation to other participants as members of a collective, and act with the intention of doing their part to raise money for cancer research. It is not the case that people need to know who all the relevant others are. The relevant beliefs about other participants admit of degrees, which in turn contribute to more or less cohesiveness among groups and more or less jointness in their collective actions. In other words, some collective intentions are more collective than others. The strongest collective intentions will involve complete transparency. At the other extreme, where there is no mutual awareness of others involved, we simply have parallel activity. As I mentioned earlier, we might think of this feature of collectivity as “tightness.” Some collectives and some intentional collective actions are more tightly collective than others.
1.3.4 Tightness and Degrees of Collectivity
As described, this two-person example involves full collectivity—each participant in the collective act has the appropriate understanding of the quite specific collective goal, her respective contributions toward the goal, her intention as contributor to contribute, the intention of her coparticipant to do the same. Under these conditions, they pursue the goal as a unit. The combination of factors that brings about the collective activity is the state of affairs that constitutes the collective intention. It is out of these conditions that the collective’s actions arise.
Consider a different case, that of a book group that meets monthly to discuss a book over a potluck dinner. Though some members are known to bring the same “genre” of food contribution each month—Joan will bring a rice dish, Kathleen can be counted on to bring a cake, Nancy most often makes a salad—the group makes no e ort to coordinate the meal. The rest of the members are unpredictable. Sometimes, there is only salad. If Kathleen can’t make it, there may be no dessert.
This dinner involves moderate collectivity. The lack of specificity of the goal and lack of knowledge about what others will do loosens the collectivity. Nevertheless, collective intention is not fully absent; the collective—that is, the book group—intends to have a potluck dinner. Whenever the group ends up with a good balance of hors d’oeuvres, salads, main courses, and desserts and when the respective dishes complement one another especially well, the members pat themselves on the back for a great team effort. But of course it isn’t really a team effort; luck plays a major part in the outcome. The outcome is much more the satisfaction of a collective hope than the achievement of a collective goal. The group does not in any sense aim at or, as a group, intend, to produce a balanced meal of well-matched contributions.
Contrast this with a more coordinated potluck. This one is a themed event—the Vegetarian Indian Potluck (VIP)...Contributors must be guided by the collective intention. Those who stake out their claim early have more latitude as to what they may bring. The closer the group gets toward meeting its goal, the more constrained will be the options to contribute for the individual who intends to participate in the realization of the group’s goal. As the evening progresses, they may truly compliment themselves for a job well done. The collective intention is satis ed, the collective goal is achieved. The level of coordination in the collective’s activities toward the goal exhibits its tightness. This could be achieved with or without full common knowledge. There may be a sign-up sheet whereby each participant can see clearly what the others intend to bring. Or there may be a point person who has been delegated the task. The others know that the e ort is being coordinated but may have only a partial awareness of which dishes have been spoken for and who is bringing what. Either way, it makes sense to say the action is fully collective, particularly in contrast with the true potluck described previously
What do we gain by recognizing that collectivity comes in degrees? The greatest gain is that we have a concept of collectivity that reflects the reality of collective activity and a means of capturing a broad range of examples and articulating the differences between them with respect to their intentional character. On one end of the continuum is full collectivity and on the other end is parallel action. At and near the end where there is full collectivity, the combination of factors—much like the structures and mechanisms in organizations—produces a collective intention. No such intention arises in parallel action. The collective intention is neither a simple aggregate of individual intentions nor an individual intention with an irreducibly collective orientation. Rather, it is a state of affairs in which agents understand themselves as members of a collective and in relation to others, aiming as a group for the achievement of a collective goal, intend individually to do their part in the achievement of the collective goal, and mutually understand one another as doing the same. These conditions set the intention for the collective.
1.3.5 Bindingness
In her plural subject account of shared intention, Margaret Gilbert maintains that individuals come together to form a plural subject when they “intend as a body to pursue a joint goal.” Once the plural subject is formed and the shared intention established, two conditions take effect, captured in two criteria: the “criterion of obligation” and the “criterion of permission.” The criterion of obligation stipulates that once the plural subject’s intentions are established, individuals do not direct them. They become bound by its goals and subject to its intentions. Most theorists acknowledge that collective goals inform and shape the goals and intentions of individual participants. Gilbert is alone, however, in maintaining that the goals and intentions of the collectives bind participants by way of an obligation. This bindingness is further captured in the criterion of permission, according to which simply opting out is not an option: anyone wishing to be released from a collective intention must acquire the permission of other members.
In my view, once in e ect, a collective intention may pose constraints on individuals’ participation but should not be understood as obligating anyone to continue participating. Recall the vegetarian Indian potluck. In that example, the collective intention to have a nicely balanced vegetarian Indian meal has an organizing e ect on the participants. It guides and constrains their contributions. To the extent that given individuals share the collective goal and think of themselves as members of the group standing in relation to others committed to the same goal, they will allow their contributions to be shaped by that goal. This is not to say, however, that they cannot opt out. One individual’s opting out will not, in a multiperson example, be enough to eliminate the collective intention. It may increase the chances of its not being satis ed—if, for example, the person opting out was in charge of the naan, the meal will be incomplete in that respect and the collective’s goal will not be fully met. When there are only two people in a given collective, for example when we intend to make dinner together, one person’s withdrawing from the plan is enough to eliminate the collective intention. That it will end the collective intentional action is not a good reason to forbid it or to require that permission need be given. It may be rude to back out of plans made and impolite to dash another’s reasonable expectations, but the risk of bad manners is not a strong basis for an obligation. For this reason I believe that while a collective intention may influence, shape, and even constrain the nature of an individual’s contribution to a collective action, there is nothing built into the notion of a collective intention that makes it obligate.
1.3.6 The Collectivist Character of the View
The view I have outlined here is more collectivist than individualist. I believe that its collectivist character is a great asset. The collectivism is based in the notion that the collective intention and action operate at a di erent level from individual intention and action. Much like a supervenience theory of the mind in which mental phenomena are based in but not identical with or reducible to material or physical phenomena, I argue that collective intentions and actions are dependent on, even based in, but by no means identical with or reducible to individual intentions and actions. Indeed, the levels interact with each other, and the interaction may go both ways. Furthermore, the existence of a distinct collective level of intention, even if it depends on features of the individual level, supports the idea of collective agency. The structural features of goal-oriented collectives in action give rise to intentional actions that cannot be understood as the actions of individuals. My earlier remarks about organic unities help to lend substance to this view. An organic unity admittedly has parts, but the whole is a unity to which a simple summation of its parts cannot do justice. In addition, my discussion of the examples of making dinner together and of the two potlucks demonstrate that collectivity comes in degrees. More collectivity supports a more intentional sense of collective purpose. In cases such as those (e.g., the vegetarian Indian potluck), individual contributions are more constrained by the goals and intentions of the collective than in cases where there is less collectivity.
The analysis provided here of the collective intentional structure of goal-oriented collectives leaves room for a robust discussion of individual responsibility in these collective action scenarios without dismissing the idea that there is a distinct level of collective responsibility about which we may also legitimately speak.
1.4 Conclusion
The goal of this chapter has been to articulate the di erence in intentional structure between two kinds of collective agents: organizations and goal-oriented collectives. The accounts are meant to establish that there is a plausible story to be told about agency at the collective level in the cases of organizations and of goal-oriented collectives. The notion of collective agency has great normative merit. In particular, as I argue in the next chapter, it enables us to account for collective moral responsibility.