A Simple Theory of Acting Together

Abstract I argue for an account of acting together that has a particular notion of joint commitment at its core. The account presented offers a compact explanation of four significant aspects of acting together as this is ordinarily understood: the parties have pertinent obligations to one another; each needs the concurrence of the rest with his or her untimely exit from the joint activity; an appropriate collective goal is sufficient to motivate the parties; and the parties may have personal goals contrary to the collective goal. I argue that the simplicity of the account coupled with its explanatory power argue for it in face of concerns that it takes us beyond the concepts we need to explain what it is for an individual to act alone. Indeed, there is reason to invoke joint commitment in many contexts other than acting together in which more than one person is involved. I conclude by noting that the joint commitment account of acting together plausibly accounts for the apparently transformative nature of doing things with others, citing its connection with a sense of togetherness, solidarity, and unity, as in the phrase ‘There is no I in team’.


I. The Question
In common parlance we regularly refer to ourselves as doing things together. We speak of taking walks together, painting a room together, approaching the enemy together, and so on.
My focal question in this talk can be put as a matter of semantics, broadly conceived: To what do people refer when they speak in this way? In other terms: What idea of acting together animates their so speaking? This is how I shall interpret the question: What is it to act together? I take this question to go to the heart of human life. It concerns one of the central ideas-if not the central idea-in terms of which human beings operate in society. If we wish to understand ourselves it is important to get it right. This is a lightly edited version of my Lebowitz Prize Lecture presented at the Dr. Martin R. Lebowitz and Eve Lewellis Lebowitz Prize Symposium at the  Pacific Division Meetings of the American Philosophical Association, as part of a debate with Michael Bratman on the topic 'What is it to act together?' In 'Creature Construction and the Morality of Shared Agency: Response to Bratman' (this issue), I respond briefly to Bratman's lecture, 'A Planning Theory of Acting Together', and further clarify some of the points made here. My work on acting together and closely related matters goes back more than three decades to Gilbert (: ch. ), and has continued (in, for instance, Gilbert [, , , ]). The present discussion is intended to introduce the topic, and my approach to it, to nonspecialists. I hope there is something here for specialists as well. Warm thanks to Michael Bratman for providing me with a deeply thoughtful sparring partner over many years. This point reflects the great sociologist Max Weber's idea that the scientific study of human social life requires us to adopt a distinctive posture towards the participants in that life. Weber labeled this posture Verstehen, which we may translate as understanding or, more fully, knowing how the people studied envisage their situation.
In the search for such understanding we may need to invoke something for which everyday language lacks a referring expression. That is to be expected insofar as we seek what we might refer to as the deep structure of our thought about acting together.
The discussion that follows is intended to serve as a brief introduction to my own approach to acting together for those unfamiliar with it. I have been developing it over many years. Here I shall try only to emphasize some main points about the approach.
I start with some observations about acting together that I think an adequate account should explain. I then outline my account and show how it explains the observations.

II. Four Observations
I refer to the four observations that I discuss as directed obligation, concurrence, disjunction, and motivational sufficiency. I discuss them in that order.

i. Directed Obligation
Suppose that, as José understands, he is walking to the bank with Marta. I take it that -absent any special background understanding between them-he will understand the following, as will Marta.
If José suddenly stops in his tracks before they get to the bank, Marta is in a position to call him to account, demanding an explanation of his stopping. Indeed, she is in a position to demand that he start walking.
She is also in a position to rebuke José for stopping, saying with asperity 'What are you doing? We are nowhere near the bank!' She may, of course, do none of these things, but she is in a position to do them, something that is not obviously the case with a bystander who may witness the scene. José, of course, would be in the same position vis-à-vis Marta, should she stop abruptly.
Note that to be in a position to do something is not equivalent to being justified, all things considered, in doing it. Perhaps there are gentler, equally efficacious means of getting José to walk than resorting to an outright demand, so that, all things considered, Marta should try those first. Perhaps she should start by asking José if anything is wrong, and, if there is, doing something to help him stay the course.
I take it that in addition to saying that Marta is in a position to make relevant demands of José, and so on, one can appropriately say that she has the standing to do so, where this standing is a matter of authority.
 M A R G A R E T G I L B E R T I take it, also, that the language of rights and obligations is appropriate here. There is some danger in using this language, because, to use a phrase of Shelly Kagan's, rights language is 'horrendously ambiguous' (Kagan : ).
However, there are standard understandings of the relevant terms such that the following statements are all equivalent in the sense that they refer to the same relation between Marta and José: . Marta has the standing to demand that José keep walking. . Marta has the standing to rebuke José for not continuing to walk. . José has an obligation to Marta not to stop. . José owes Marta his continuing to walk. . Marta has a right against José to his continuing to walk.
A few notes on the above list are in order. First, there may be institutional cases in which analogues of the standing to demand and the standing to rebuke come apart. Such cases are not at issue here. Second, one can think of () and () as the most perspicuous ways of referring to the relation in question, given the variety of existing uses of the words and phrases 'obligation to', 'owes', and a 'right'. Third, in some discussions, to avoid ambiguity, I have labeled rights of the kind in question here demand-rights. Here I refer to them simply as rights.
Taking the equivalence of () through () on board, I take it that an adequate account of acting together will imply that those who are acting together have the relevant standings, rights and obligations to one another. I call this point directed obligation.
Let me forestall a possible objection to it. People often engage in wrongful acts together. I take it that to say that-as they will understand-they are obligated to one another to do their parts in these acts, is neither to say nor to imply that they are morally permitted, let alone morally required or, in other terms, morally obligated, to do so.
It is equivalent to saying that they have the standing to make related demands of one another. These are demands, however, that they will neither make nor accede to, if they are to act in a morally acceptable manner.
That said, not everyone will recognize the dictates of morality in such a situation, and even those who do may still fulfill their directed obligations. Selfishly, they may want to avoid the rebukes of their fellows. More altruistically, they may prefer not to let them down. Groupishly, they may prefer not to let down the team.
ii. Concurrence Suppose now that José does not simply stop in his tracks on his way to the bank, but tells Marta, 'I'm going back home'. Marta may well feel that he is out of line. More specifically, she may well feel that he should have obtained her concurrence before settling on this plan of action.
Generalizing the point: failing special background understandings, in order to drop out of a particular joint action without fault, one must obtain the concurrence of the other participants. I call this point concurrence.
Whether and how it shows up 'on the ground' will depend on the parties' understanding of what they are doing. Compare Marta and José's incipient walk to the bank, with the following case.
Patty runs into Fred as he is walking up Fifth Avenue, and they fall into conversation, walking along together as they talk. They are, as they understand, 'walking along together as long as that suits them both'. Suppose Patty suddenly stops, and says, 'I'm stopping here'.
Fred is unlikely to react in the way Marta reacted to José's announcement, given his understanding of what he and Patty are doing together. That it is fine for Patty to stop whenever she wants is built into their understanding of what they are doing.
What might require Fred's concurrence is Patty's staying with him exactly as far as he is going, having ascertained where that is as they walk along. Their conversation might run as follows: Another point to be emphasized is parallel to one I made earlier. Though what one might call the logic of acting together requires that one obtain the concurrence of the other parties rather than simply declaring that one will no longer participate, there can be cases in which a given participant is morally permitted, if not morally required, to stop participating, even without the concurrence of the others.
A somewhat special case is that of incapacity. If I am incapable of going on, I am likely simply to say that I have to stop, rather than seeking the other party's concurrence.
Evidently the language of incapacity is a powerful tool in such situations, whether used honestly or not. People understand that saying that they cannot go on should forestall any attempts to keep them on board, and that it makes a simple acknowledgement of their situation an appropriate response.
iii. Disjunction Suppose, once more, that Marta and José are walking to the bank together. Perhaps this is what they agreed to do, and it is what each understands them to be doing.
It seems that this could be true, though their getting to the bank is neither Marta's nor José's personal goal.
It seems possible, indeed, that having established what we may call their collective goal-getting to the bank-each has privately established a contrary personal goal.
Perhaps Marta intends to provoke a nasty quarrel around the halfway mark, in order to force their collective abandonment of their collective goal. With the same purpose, José plans to complain that his knee is acting up.
In sum, the collective goal of those who are acting together need not correspond to personal goals of the participants. I call this point disjunction.
Of course, if one or more of the parties have contrary personal goals, their collective goal is less likely to be reached. That does not mean that they have no collective goal-or that this goal will not win out in the end.

iv. Motivational Sufficiency
A related point is this. Intuitively, once a collective goal is established, that suffices to rationalize the individual parties' directing their actions towards its achievement. For example, 'We are storming the Bastille' suffices to explain why I am rushing forward. In other terms, that G is our goal appears to be inherently motivating for me, even absent a corresponding personal goal.
I call this point motivational sufficiency.

III. A Challenge
I propose, then, that an adequate account of acting together will explain at least the points I have labeled directed obligation, concurrence, disjunction, and-in spite of disjunction-motivational sufficiency.
These points pose a formidable challenge to the theorist of acting together. The account I have developed meets this challenge.

IV. The Deep Structure of Acting Together
The core of my account of acting together is an account of a collective goal. After presenting this account, which uses some technical language, I explain its elements. For goal G to be the collective goal of A and B is for A and B to be jointly committed to endorse G as a body.
I need now to explain two key elements in this account: (i) What is joint commitment? (ii) What is it for A and B to endorse goal G as a body?
i. Joint Commitment I have written about joint commitment at length. I now briefly review its central features.
First, any joint commitment in my sense involves both a particular process and a particular product. The process is broadly speaking psychological, the product is normative.
a. The Process. There are two importantly different kinds of case with respect to the process of joint commitment.


In order to form a particular joint commitment in the basic case it is necessary and sufficient for the would-be parties openly to express their personal readiness for that joint commitment to be in place, in conditions of common knowledge.
In a non-basic case those parties must have openly expressed their readiness to be jointly committed with one another, within certain limits, by means of a particular process, such as the decisions of a particular one of their number, in conditions of common knowledge. Here as elsewhere in this discussion I focus on the basic case.
b. The Product. When they are jointly committed in some way, the parties are subject, as one, to an exclusionary normative constraint.
More precisely, there is something that, collectively, they ought to do, all else being equal, where personal inclinations and desires do not suffice to tip the scales in the other direction.
The ought in question is not the moral ought, though it may be morally required, all else being equal, to conform to one's joint commitments. That is a question that I set aside here.
c. The Individual Commitments of the Parties. I take it that by virtue of their joint commitment each party is, individually, committed to act in such a way that, in conjunction with the acts of the other party, the joint commitment is fulfilled. This leaves a lot open: in practice things may be filled in along one of several lines.
Evidently the individual commitments of the parties stand or fall together with the joint commitment from which they derive. In an important sense, then, they are interdependent. More fully, they are both existence-and persistence-interdependent.
I take the term persistence-interdependence from Michael Bratman, who has included in his account of 'shared intention' a clause stipulating the 'persistenceinterdependence' of the personal intentions he takes to be constitutive of shared intention. (See, for instance, Bratman [: ].) d. Rescission of a Joint Commitment. Absent special background understandings, no one party is in a position unilaterally to rescind a joint commitment. Rather, each must take part in its rescission.
This does not mean, of course, that a given party may not be justified, all things considered, in failing to conform to a particular joint commitment prior to its rescission.
ii. The Content of a Joint Commitment As I put it, for short, in any case of joint commitment the people in question are jointly committed to do something as a body, in a broad sense of do. Thus, in the case in question here-that of a collective goal-the parties are jointly committed to endorse a particular goal as a body. This is to be spelled out roughly as follows: by virtue of their several actions and utterances they are to emulate-in relevant ways-a single endorser of the goal in question.

 M A R G A R E T G I L B E R T
Conformity to such a commitment evidently requires that the actions-and intentions-of the several parties mesh in relevant ways. Suppose that our goal is for the spare room to be painted by virtue of our several actions using a single coat of paint. Then, if you have just painted this wall, it would not be appropriate for me to paint it again, and so on.
Our joint commitment thus provides us with a framework for negotiation, as necessary, about the way in which the joint commitment is to be fulfilled.
My phrasing of these last points echoes Bratman's observation that shared intention and agency provide a framework for relevant negotiation among the parties and that successful shared intention and agency requires 'meshing sub-plans' (see Bratman []).

V. Acting Together
I have now explained the technical terms in my account of a collective goal. To repeat that account: For goal G to be the collective goal of A and B is for A and B to be jointly committed to endorse G as a body.
Taking this on board the following account of acting together suggests itself: Two or more people are acting together if and only if each is acting in light of their joint commitment to endorse, as a body, a particular goal, G, in order to promote the fulfillment of G.
As I now argue, given the joint commitment at its core, an account along these lines is able to explain the four observations on acting together noted earlier.

i. A Joint Commitment Account Accommodates Directed Obligations
Consider first directed obligation. Note that to say that the parties to a joint commitment are normatively constrained in the way I have noted is not yet to say that they are obligated to one another. These are two separate claims.
That said, I take it that the parties to any joint commitment are obligated to one another to conform to the commitment-and that this is so on account of a crucial feature of the normative constraint in question. Let me explain.
Recall, first, that for me to be obligated to someone to do something is for that person to have the standing to demand that I do it. I take it that, intuitively, any one party to a joint commitment has the standing to demand of any other party conformity to the commitment.
In order to explain this intuitive judgment, one can turn to an idea suggested by Joel Feinberg: to demand an action of another, with standing, is to demand that action as in a particular sense one's own. On its own, this is not very helpful. In

A S I M P L E T H E O R Y O F AC T I N G T O G E T H E R
 the case of joint commitment, however, one can develop it in a plausible way, roughly as follows.
I am a co-author of the joint commitment in question. As such I have determined---in conjunction with the other parties-how the parties ought to behave, all else being equal. This puts me in a position to call one or more of the other parties to order, should they propose not to conform. In other terms, I have the standing to demand that they conform. (I discuss this point at greater length in Gilbert [: -].) Note that the directed obligations of joint commitment appear to have nothing to do with morality, unless one understands morality in a particularly broad way. For it seems that people can be jointly committed to accept morally suspect goals, and that, when they do, they have the standing to demand of one another that they act accordingly.
Most likely, they should not make these demands, any more than they should make the relevant joint commitments, or comply with them. That, however, does not imply that the directed obligations in question do not exist. On the contrary, the parties to the joint commitment retain the status of co-determiners of what each one ought to do, all else being equal. As such, they retain the related standings, whether or not they ought to take advantage of them.
ii. The Joint Commitment Account Accommodates Concurrence It is possible to deal more quickly with the observation that, absent special background understandings, those who are acting together need to obtain the concurrence of the other parties in order to give up on their joint project without fault.
Given its constitutive joint commitment, the parties cannot unilaterally rid themselves of their collective goal. So, without the concurrence of the other parties, they cannot cease to act in pursuit of that goal without failing to fulfill a standing joint commitment.
This point allows for a variety of background understandings relating to a given joint commitment. In particular, one party may make it clear at the outset that she is ready to rescind given only the other's expressed readiness to do so.

iii. The Joint Commitment Account Accommodates Disjunction and Motivational Sufficiency
According to disjunction, our collective goal in acting together need not correspond to our personal goals. This is clearly the case given the joint commitment account of acting together. Suppose that Marta and José are jointly committed to accept as a body a certain goal. That, then is their collective goal. It is not constructed, in whole or in part, by personal goals of either. It is therefore compatible with their lacking the corresponding personal goals, or even having contrary personal goals.
According to motivational sufficiency, a collective goal is sufficient to rationalize the parties' acting in accordance with it, absent the corresponding personal goals.
Given the joint commitment account of a collective goal, its motivational efficacy derives from several sources. Each party understands she is subject to the relevant exclusionary normative constraint-one, moreover, that she is powerless to cancel unilaterally. Further, she owes it to the other parties to act accordingly, on pain of authoritative demands and rebukes from them. ii. It Is a Simple Theory A further virtue of the account is its simplicity. Rather than a multiplicity of clauses, it has a single clause. I take it that, all else being equal, a simple theory is preferable to a complex one.

VII. Overall
iii. The Question of Theoretical Parsimony: Joint Commitment Outside Acting Together One may resist the introduction of joint commitment into an account of acting together on the grounds of theoretical parsimony: would it not be better not to go beyond the basic concepts needed for understanding what it is for an individual to act? Theoretical parsimony is a good thing if the relevant concepts can do the necessary work. For several reasons, however, it is not clear that any account of acting together that does not appeal to a joint commitment of the parties can do this. Among these reasons are the need to explain directed obligation.
Meanwhile, it may reassure the parsimonious theorist to note that there are many social phenomena other than acting together for which an explanation in terms of joint commitment is compelling. These phenomena include agreements, which often, though not always, precede joint actions, and collective beliefs, which often accompany them. (For accounts of a wide range of social phenomena in terms of joint commitment see, for instance, the essays in Gilbert [].) Given that this is so, it is not a big plus from the point of view of social theory to avoid joint commitment in the explanation of acting together. The invocation of joint commitment in an account of acting together takes us beyond the terms of which we think about individual humans with their personal goals, decisions, and intentions. I take that to be a virtue of the account for the following reason.
Acting together is transformative. It is natural to associate it with such positively valenced ideas as togetherness, solidarity, and unity. It produces such hackneyed but telling lines as 'There is no I in team'.
The joint commitment account of acting together, then, is not only simple, and apt to explain all four of the observations with which I started. It is also apt to explain why we value acting together for its own sake as well as for its often very useful consequences.
After more than a year of a worldwide pandemic, with many of us missing our usual face-to-face lives, the transformative, noninstrumental value of something as simple as taking a walk together could not be clearer.
It is important to understand precisely what we are missing when we are deprived of acting together. I propose that at least part of this is the sense of togetherness, solidarity, and unity that we experience when acting in light of the associated joint commitments.
MARGARET GILBERT UNIVERSITY OF CALIFORNIA, IRVINE mpgilber@uci.edu