In her 2021 Lebowitz Prize Lecture, ‘A Simple Theory of Acting Together’, Margaret Gilbert seeks to articulate the ‘idea’ of acting together that ‘animates’ our commonsense talk about this important phenomenon. I seek a model that provides illuminating sufficient conditions for this phenomenon. As I see it, these are not quite the same project. After all, our commonsense idea and talk may well have two interrelated faces: an inchoate understanding of what the phenomenon is; and an inchoate understanding of norms about, very roughly, what those who participate in this phenomenon normally thereby owe to each other. Gilbert develops a rich and complex articulation of this second element, one according to which the interpersonal obligations in question are not in general moral obligations. Broadly speaking, her strategy is then to take this web of interpersonal obligations and the like and directly build it into her account of what the phenomenon of acting together is. This leads her to say that that phenomenon involves a non-reducible phenomenon in which A and B are ‘jointly committed to endorse G as a body’––where such joint commitments are constituted at least in part by the interpersonal obligations and the like to which our commonsense idea of acting together alludes.

In contrast, my thought is that our commonsense idea of acting together is Janus-faced, involving both an inchoate understanding of the phenomenon of acting together, and an inchoate understanding of normally associated interpersonal obligations and the like. I think that when we seek better to understand the purported obligations, we need to ground them in defensible, substantive norms of obligation. And I think we can see the cited inchoate understanding of the phenomenon of acting together as a fallible guide to a model of that phenomenon that can guide further research into this phenomenon.

We can then agree with Gilbert that an ‘adequate account’ would concern itself with both faces. Nevertheless, my proposal is that if we disentangle these two faces of our idea of acting together, we do not need to introduce as a constitutive element of the phenomenon of acting together, a nonreducible phenomenon of being ‘jointly committed to endorse G as a body’, one that directly and constitutively incorporates relevant interpersonal obligations. Instead, we can construct shared intention using building blocks from the planning theory of individual temporally extended planning agency. And that is what I try to do. This then supports the idea that our capacity for planning agency is, as I say, a core capacity at the heart of both our temporal and our small-scale social practical organization. We can then see Gilbert's articulated web of interpersonal obligations and the like as useful starting points for a theory of, roughly, what those who participate in this phenomenon of acting together normally thereby owe to each other. And we can do this while avoiding a simple and direct inference in either direction, from shared intention to those obligations, or from those obligations to shared intention.

That said, let me highlight an important agreement between us. We both think that acting together essentially involves practical interrelations between the participants, interrelations that are not merely a matter of common knowledge or the like. In this respect, we both diverge from one aspect of the approach of John Searle (Reference Searle, Cohen, Morgan and Pollack1990): Searle's appeal to irreducible ‘we-intentions’ in the heads of each leaves out appeal to practical interrelations between the participants in a shared intention. The difference between Gilbert's and my view, however, is that in my case these practical interrelations involve interlocking of intentions, intended mesh, intended and actual mutual responsiveness, and persistence interdependenceFootnote ¹––where in some cases this interdependence can be grounded in familiar moral obligations. In Gilbert's case, in contrast, these interrelations quite generally involve not-essentially-moral directed obligations of each to each.

Returning to the main thread, my proposal is to see the ideas that animate our talk about acting together as having two faces:Footnote ² one points to an explanatory sub-structure for which I offer a plan-theoretic construction; a second points to associated norms of interpersonal obligation and the like. Each is important, and they are related in complex ways. To understand these obligations, we need to engage in substantive normative reflection in defense of relevant norms of obligation. And in many cases, recognition of these obligations helps support the functioning of the explanatory substructure: this is what happens, for example, when relevant persistence interdependence is supported by assurance-based induced-reliance-based obligation. Nevertheless, as I see it, it is a mistake to reason simply and directly from those obligations to the plan-theoretic substructure, or vice versa.

Consider, in this light, Gilbert's ‘disjunction’ condition. Bracketing some subtleties, Gilbert's view is that if Marta and José have publicly (among themselves) agreed to walk together to the bank then they have a ‘collective goal’ so to act together. This is so even if each is in fact insincere in their agreement and each in fact explicitly intends not to continue walking. In such a case, I take it, neither intends that they walk to the bank. So, there is not the kind of shared intention I aim to construct, one that supports a standard explanation of the coordinated action of each. Instead, there is a structure of explicit intentions of each that is set to explain a breakdown in such coordination. To be sure, the agreement may ground relevant obligations that in turn ground criticisms of each––though to defend this we would need to defend a substantive norm that says that when folks interact in these ways, they have relevant obligations. The spirit of Gilbert's view is to read back from such obligations to a ‘collective goal’ and ‘joint commitment’. But if that is what we do, then so-called collective goals and joint commitments will not in general directly play the explanatory role associated with intention––though if the agents were to recognize the cited obligations that might lead them to associated intentions. If, as I instead propose, we disentangle the two faces of our idea of acting together, we can acknowledge the obligation, if such there be––and the way it might ground criticism of each and/or support the formation of relevant intentions––without assuming that we can read the obligation back as a constitutive element of a shared intention that directly plays basic explanatory roles.

My proposed planning theory of acting together is a philosophical experiment. If we allow for complex relations between the two faces of our idea of acting together, can we draw on the planning theory of temporally extended human action to provide broadly reductive but theoretically illuminating sufficient conditions for the phenomenon of our acting together? And can we thereby support the idea that our capacity for planning agency is at the bottom of both the cross-temporal and the small-scale social organization that is so important in our human lives. And I have argued that there is good reason, in each case, to answer: Yes.