Arguments for Collective Responsibility
21.3.1 Agency-Based and Responsibility-Based Collectivist Arguments
It is worth drawing a general distinction between two types ofcollectivist argument. Responsibilitybased arguments are based on the idea that without attributing non-distributive responsibility to collectives we will have a “deficit in the accounting books” (Pettit 2007a: 194; see also French 1979: 214; Kutz 2000: 113; Copp 2006: 216; Braham & van Hees 2011; List & Pettit 2011: 214; Smith 2018, etc.).
We can end up with such a deficit because in some cases no individual seems to be responsible for some morally objectionable event or state of affairs, or even when all individual responsibility is tallied up there still appears to be some normative residue.Responsibility-based arguments aim to establish that it is not conceptually incoherent and makes much moral sense to ascribe non-distributive responsibility to collectives in some cases. Whether and in what sense such groups can be said to be agents may not be considered at all. If that question is considered, then the intuitive plausibility of the claim that responsibility entails agency may be used to show that groups can be agents (see especially Copp 2006: 216—17 for this strategy). In other words, one can argue from collective responsibility to collective agency.
However, some collectivists may not be inclined to make this last move. For one thing, one could deny the claim that responsibility implies agency altogether. Alternatively, one could argue that (pace Copp 2006) the claim does not apply to collectives in the same sense as to individuals. Thus, some may accept the general claim but try to show that whatever the collective is held responsible for can be traced back to the exercising of individual agency. In any case, the availability of these theoretical options shows that responsibility-based arguments can be made without a commitment to the possibility of (non-distributive) collective agency.5
By contrast, agency-based arguments seek to show that some collectives can be unproblemat- ically described as agents possessing the same capacities that render individual human beings “fit to be held responsible” (to use Pettit’s (2007a) apt formulation).
Why should collectives be let off the hook for the bad things they do if they have a sufficiently similar set of abilities and functions as those in virtue of which individuals are morally responsible for their deeds? Just as responsibility attributions, such ascriptions of agency are intended non-distributively. This means that it is at least sometimes wrong to understand statements about group action such as “IBM successfully conquered the Japanese market” or “the Allied bombers set Dresden on fire” to be a mere shorthand for a conjunctive list of individual actions, e.g., “Bomber pilot1's bombing1, Bomber2S bombing2, Bomber3's bombing3..., BomberN’s bombingn set Dresden on fire.”6 Rather, some such sentences refer to an “autonomous” (Pettit 2007b) or “irreducible” (McKenna 2006) collective group agent.7 Note also that such a collective agent could be autonomous in the required sense even when its actions are realized by an individual acting on behalf of the collective, e.g., when a manager signs a contract for her company (French 1979). In any case, commitment to the possibility of such agents is usually expected to be consistent with ontological individualism (about which more in section 21.6 below), i.e., the view that all group-level properties, including agential properties, supervene on properties of individual group members (and possibly non-members).As a further step then, such agency-based arguments may proceed to demonstrate that collectives are not only agents, but also morally responsible agents qua collectives (section 21.6 will discuss one way in which collectivists have tried to do this using the discursive dilemma). However, it is important to recognize that the second step in agency-based arguments is not trivial. It can be argued that while some collectives meet the conditions of agency, they do not meet the more exacting conditions of morally responsible agency either as a matter of empirical fact (see McKenna 2006) or because it is conceptually impossible for this to be the case.
21.3.2 Alternative Agency-Based and Responsibility-Based Arguments for Collective Responsibility
One type of responsibility-based collectivist argument which does not invoke the discursive dilemma has been applied to cases of collective omissions (Petersson 2008, etc.), cases of marginal individual contributions to collective harm (Kutz 2000, etc.), and cases of outcomes causally overdetermined by individual contributions (Kutz 2000, but cf. Parfit 1984). The thought is that such cases involve unjustifiable harms such that the sum total of all individual fault (if any) falls short of what would be commensurate with the harm in question. For example, each typical inhabitant of the metropolis upstream seems only to be marginally at fault for her contribution to polluting the river (after all, the individual contributions may be hardly perceptible). Yet pollution seriously wrongs inhabitants of the village downstream. If the same wrong had been caused by an individual, we would not hesitate to hold her responsible. So why not do the same in the case of the collective?
Another type of responsibility-based collectivist argument takes the Strawsonian approach to moral responsibility as its point of departure. If Strawsonian reactive attitudes such as resentment or indignation are fitting reactions to collectives, then collectives can be morally responsible (Tollefsen 2003).8 This finding can be used to show “in a roundabout manner” that certain collectives can be agents since being the fitting addressee of reactive attitudes presupposes reasons-responsiveness or some kind of normative competence (Tollefsen 2003). Alternatively, it could be argued that for collectives to be the fitting targets of reactive attitudes they do not need to be agents in a non-distributive sense.9
An influential agency-based collectivist argument, which does not appeal to the discursive dilemma (but is a precursor to arguments from the discursive dilemma), was put forward by French (1979).
The basic idea here is that we can ascribe intentions to collectives as wholes provided they instantiate certain structural features. The relevant structural features are most importantly, first, a clear and formalized division of labor, and second, a commitment to stable and long-term corporate policies (French 1979: 212). Moreover, it is argued that since we can ascribe collective intentions to corporations that exhibit such structural features, we also have good grounds to hold them collectively responsible.Empirical support for the collectivist position has been forthcoming as well. It has been found that folk psychology is at least sometimes collectivist. People in certain situations seem willing to ascribe non-distributive blame and praise as well as agency to at least certain kinds of groups (Michael & Szigeti 2018). Folk, of course, may be wrong, but these empirical findings show individualism to be more revisionist than may have been assumed.
21.4