I start by emphasizing two aspects of Michael Bratman's approach to shared agency and contrast it with my own in those respects. I conclude with some related remarks on the relation of morality and joint commitment.
1. Persistence-interdependence in Bratman's construction: the problem of stability
Consider a being I shall refer to as ‘the Decider’, or ‘Dee’ for short. Dee is looking for a firm that engages in what Bratman has referred to as ‘creature construction’ (see, for instance, Bratman [2014: ch.1 sec. 6]). She wants to acquire some human-like creatures who will do such things as carry a heavy table up the steps to her house, play string quartets with aplomb, and, when at liberty to do so, play together nicely. She is considering bids from two creature-construction companies, the Bratman firm and the Gilbert firm.
The Bratman firm advertises that it can do the job in a ‘broadly individualistic’ way, ‘without appealing to an element that is fundamentally discontinuous’ from those needed to create human-like creatures with intentions of their own—individual ‘planning agents’. Dee is not sure why this point is so important, though she understands that the Bratman firm has done excellent work on individual planning agents.
The Bratman firm lays out the five elements of its proposal—a couple more elements than it invoked when making a bid to a friend of Dee's in the early 1990s.
It is one of the newer building-blocks that gets Dee's attention: each creature's intention is ‘set to persist, other things equal, if but only if the other's corresponding intention persists’. Further, that this so is ‘epistemically accessible’ to each of the involved creatures.
Dee wonders in particular at a parenthetical note maintaining that in spite of the forementioned ‘persistence-interdependence’, the creatures involved ‘will not have lost control’ over their own intentions. Now, as Dee understands them, personal intentions can not only fade, but be deliberately cancelled or changed by the person in question. So either the so-called ‘intentions’ involved here are something new, something that is not involved in the intentions of singular agents, or there is a stability problem.
What if one of the creatures is drawn to an alternative course, changes its mind, and opts out of the supposedly joint activity? Dee would like some insurance against this.
2. Giving Bratmanian creatures a moral sense
The representative of the Bratman firm suggests that it can give its creatures a moral sense, but morality is complicated and they cannot guarantee the results. Indeed, he notes, some moralists think that one is free, morally, to do whatever one wants, whenever one wants. They reference Ayn Rand.
Dee worries that if one of the firm's creatures decides to follow Ayn Rand, for instance, nothing would stop it from quitting without any ‘by your leave’, when the mood takes it. She says she will think about the Bratman proposal and contacts the Gilbert firm.
3. Bringing in joint commitment
The Gilbert firm provides Dee with its prospectus, which she reads. She notes that the firm endows its creatures with an understanding of what it calls ‘joint commitment’, and with the power to make such commitments. Evidently, the Gilbert firm does not restrict itself to ‘individualistic’ components as does the Bratman firm. Dee has no problem with that. But why bring in joint commitment?
The Gilbert firm's prospectus refers to the following aspects of any joint commitment:
• Joint commitments, like personal intentions, prevail over contrary desires and inclinations from a rational point of view: the participants ought to conform irrespective of such contrary desires and inclinations.
• Further, and unlike personal intentions, joint commitments are not unilaterally rescindable.
• Further—and this is written in boldface type—those who are jointly committed are obligated to one another to conform to the commitment.
The prospectus goes on to emphasize that this last point is not a moral one. It is equivalent to the claim that the parties have the standing to issue demands and rebukes to one another for acting contrary to the joint commitment, a standing whose basis is the joint commitment itself, and, in particular, its co-imposition by the parties.
The Gilbert firm sums up: An individual creature who understands itself to be jointly committed with others to espouse a particular goal as a body not only understands that the commitment should prevail over its own contrary desires and inclinations. It also understands that it cannot remove the commitment unilaterally, and that if it fails to act in ways inappropriate to the commitment, it risks the demands and rebukes of its fellows, to whom it owes action appropriate to the commitment. The governing joint commitments entered into by these creatures will, of course, be determined by their owner, who will tell them what she wants done, or when they are free to devise their own distractions.
Dee feels that Gilbert-style creatures will be more reliable than Bratman-style ones. She worries, however, that they will be unhappy creatures, afraid of being bombarded with one another's demands and rebukes.
The representative of the Gilbert firm assures her that these rebukes and demands do not have to be issued—and may not always be the best way to ensure that the joint commitment is fulfilled. In the case of jointly committed creatures, if helping one of them to keep up is the most efficient course, they will help it, failing any countervailing considerations. Indeed, they will be obligated to the others to help it, and hence open to rebuke if they do not. The same goes for uttering encouraging words, and similar positive enforcements.
This satisfies Dee, whose main concern is efficiency, and she signs up with the Gilbert firm.
4. Morality and joint commitment
Now imagine a world whose inhabitants regularly do things together in the sense of my account: they make joint commitments to endorse particular goals as a body and act accordingly, combining their actions in such a way that the goal is reached. Perhaps they have been manufactured to act in this way, as are the products of the Gilbert firm. Perhaps this has happened as a result of evolutionary pressures.
I have said that the concepts involved here are not specifically moral concepts. In particular, the directed obligations of joint commitment are not in themselves moral requirements. Indeed, they are not requirements at all. They are relations between the parties to the joint commitment, relations that may be expressed by the said rebukes and demands in relevant contexts.
That said, it is easy to imagine how a moral sense of things could be triggered in the context of acting together in the sense of my account. For instance, suppose five of us have just come back from hunting stag together, we sit down to a long-awaited meal, and you give everyone but me a large piece of stag-meat. I may sense that something is ‘wrong’ here, that each hunter ‘should’ have roughly the same size piece. Here we can say that a sense of fairness—or in this case, a sense of unfairness—has been triggered. I take fairness to be a moral concept, one which is not part and parcel of the conceptual scheme of shared agency (or joint commitment) itself.
In sum, there is what we might call a morality of shared agency, in which such concepts as fairness and desert come into play, along with moral concerns about coerced involvement and collective goals with specific contents. It is the role of moral theorists, broadly speaking, to uncover the details here. As Bratman says, these matters call for normative reflection, and there are hard cases.
Nothing that I want to say about shared agency denies these facts. It suggests, however, that our normative theorizing must acknowledge a realm of rights and obligations that exists independently of the moral realm. These are the rights and obligations constitutively involved in the type of shared agency involved when two people, as they would put it, go for a walk together.Footnote 1
Open access
