Commitments and Collective Responsibility

Few accounts of commitments link them directly with collective responsibility.⁴⁷ Still, commitments’ roles in accounts of shared agency suggest that they should inform this debate.

By contrast, the case of SMARTPHONES revealed very little about the role of commitments in cases of individual backward-looking responsibility. Still, as I suggested in §13.1, commitments may play a significant, if underappreciated, role in collective backward-looking responsi­bility I made this suggestion on the basis of their import for views such as Gilbert’s, where commitments are necessary or sufficient conditions for shared agency. In what follows, I expand upon this suggestion. I then consider a second, more ecumenical way of using commitments to (partially) explain backward-looking collective responsibility—namely, by linking it to failures to meet the obligations that commitments create.

13.3.1 Commitments, Shared Agency, and Backward-Looking Responsibility

If joint commitments are necessary and sufficient for collectively intending an action,⁵¹ they are, by extension, at least necessary for assigning backwards-looking collective responsibility for states of affairs that, in Gilbert’s language, the plural subject has caused through its intentional actions.⁵² I say “at least necessary” because whether a group collectively intends an action is not alone necessary and sufficient for holding the group collectively responsible.

Bratman’s account of commitments provides an alternative way of linking commitments with backward-looking responsibility. Although Bratman denies that what he calls shared commitments are necessary or sufficient for sharing an intention or for shared agency more broadly, he contends that they play a substantive role in how shared agency is exercised. This is because commitments organize shared deliberation. In this regard, if commitments are neces­sary conditions for shared deliberation, then they are also relevant for the way in which groups bring about states of affairs for which they could be collectively responsible in a backward­looking sense. Recall the community garden case. Imagine that the group members deliberate about whether to dispose of their unused weed killer in the city’s drains, they know that doing so will pollute the water supply, and they decide to do so in light of a commitment they have to promote their garden’s beauty above all else. In this case, it is their commitment that plays a necessary role in their causing the water to be polluted. Insofar as deliberation is a condition for responsibility attribution (Yaffe 2009: 10; cf. Guerrero 2015: 83—4) and commitments play an essential role in shared deliberation, then it is plausible to think that they might be the source of backward-looking collective responsibility attributions.

With commitments’ two possible roles in backward-looking collective responsibility in mind, let’s consider whether commitments also help us to determine whether it is distributive or non­distributive. On first glance, it would seem that backward-looking responsibility in this context is non-distributive. Assume we adopt a robust view of commitments such as Gilbert’s, according to which commitments secure a group’s capacity to engage in shared intentional action such as maintaining the community garden. If we think that commitments play a role in unifying the participants such that they are able to act together, then the commitments are essential for explaining the way in which the community garden’s plants died. Since these commitments are jointly held, the backward-looking responsibility for failing to uphold them should, in prin­ciple, be non-distributive. But this is not necessarily the case. The specific conditions under which such failures occur will determine whether, in any particular instance, it is distributive or non­distributive. Imagine that part of the group’s plan for tending the garden involves applying com­post. Some of the new volunteers fail to learn how to apply it correctly, and those who know how to apply it do not teach the new members. In this case, there is non-distributive collective responsibility for the group’s failure to adhere to their plan and also distributive responsibility for the particular ways in which each of the individual members’ actions as group members contributed to the plants’ demise.

This discussion shows that it is natural, but ultimately wrong, to think that commitments’ role in securing shared agency entails that, insofar as they are jointly held, the kind of backward­looking collective responsibility that commitments help to explain must be fundamentally and exclusively non-distributive. First, whether it is non-distributive will depend on how the failures to uphold the commitment occurred and whether they attach to the group as such.

13.3.2 Commitments, Obligations, and Backward-Looking Responsibility

What if one does not want to grant that commitments play such a significant and fun­damental role in shared agency as that outlined in the previous section? Is there a way to link commitments with backward-looking responsibility while remaining neutral regarding their role as one of the building blocks of shared intention, shared intentional action, and the like?

One option is to return to their role in establishing the grounds of forward-looking respon­sibility.⁵³ Here the idea is that if joint commitments entail obligations to follow through on them, some failures to do so will be those for which groups are collectively responsible in the backward-looking sense. I say “some” rather than “all” failures because the way by which such a failure occurs will affect whether it is appropriate to attribute collective responsibility of any form.

Before considering this approach, four caveats are in order. First, this approach may seem inconsistent with the idea that backward- and forward-looking responsibility are different in kind (Isaacs 2014: 43-4; Rovane 2014: 12; Smiley 2014; cf. Mathiesen 2006: 254). In fact, how­ever, this approach does not deny that they are different in kind. Instead, it is focused on the kinds of backward-responsibility we assign when people fail to meet their (moral) obligations; it is mute with regard to whether forward-looking responsibility more generally (should) focus(es) on what Rovane (2014: 12) describes as “taking responsibility for what will happen in the future.” A more robust way of reading this approach is to interpret it as showing that while Rovane, Isaacs, Smiley, and others may be correct that these two forms of responsibility typic­ally come apart, the role of commitments in shared agency suggest that there is one important domain where they are intimately interconnected.

Second, it may seem that this approach inverts the typical way of understanding the rela­tionship between backward- and forward-looking responsibility. Usually, one begins from a backward-looking responsibility attribution—say, Jane’s intentionally lighting the match and knowingly throwing it on the dry leaves caused the small forest fire—and then moves to the judgment that Jane has a forward-looking responsibility to correct the harm by, say, replanting the trees. My approach in this section does not rule out this type of explanation; rather, I suggest that, in contexts where we have created obligations to follow through on certain actions by virtue of our commitments to do so, the explanation of backward-looking responsibility for failing to meet these obligations is dependent upon first establishing the grounds for the relevant forward-looking responsibility.⁵⁴

Third, this approach is not inconsistent with the standard way of understanding backward­looking responsibility attributions as those for which one has caused a harm. In fact, this approach takes the idea that antecedent forward-looking responsibility attributions, which are based on the duties we have to others, identify the context within which to determine the causal route by which we failed to meet those commitments. It is only after so doing that we can determine the propriety of attributing backward-looking responsibility.

Finally, Gilbert’s view would seem to best illustrate the approach under consideration in this section. Yet I initially characterized the approach in this section as neutral regarding whether commitments are necessary or sufficient for shared agency. Gilbert’s view is not neutral on this front. Although I treat Gilbert’s view as the paradigm in what follows, I limit my discussion to her account of the relationship between commitments and obligations. In this regard, one could easily adopt her view about this relationship while rejecting her view about the role of commitments in shared agency more broadly.

With these caveats in mind, let’s return to the original case of the community garden members. Here, joint commitments entail, first, individual backward-looking responsibility for any members of the group who fail to follow through on them. There may also be distributive, collective responsibility.⁵⁵ What does the latter depend on? Consider two modified versions of the case. In the first version, one member of the community garden misses his shift for weeding the vegetable plot, thereby causing an overgrowth of weeds in what the group had wanted to be a well-tended garden. This member is individually responsible for his part in failing to uphold the joint commitment to tend the garden, but there is no clear sense in which there is any col­lective responsibility. For the sake of comparison, imagine a second version of the case where the members are jointly committed to having a tended garden but they make no plan for how such tending will take place, thereby allowing the garden to be consumed by weeds. In this version, the group is collectively responsible in the distributive sense because each individual member, by virtue of their group membership and their failure as a group to make a plan for their garden’s care, failed to uphold their joint commitment.

What about the possibility of classifying the group as (also) non-distributively collectively responsible for failing to follow through on their joint commitments? This responsibility is non­distributive only if the responsibility for failing to follow through on the commitment to tend the garden attaches to the group as such. Can such responsibility attach to the group as such? If the failure to uphold the commitment is a product of the group failing to follow through—as in the case of the community garden members failing to make a plan to beautify their garden but nonetheless being committed to doing so—then the responsibility for failing to follow through on the commitment (and for the dead plants) is the group’s.

13.4