Group Agency and Two Types of Preference Change

18 July 2019

Although often more in marginal remarks than at the core of their work, multiple authors have argued that preference change plays a key role in the functioning of group agency. I want to consider a minor observation I’ve recently made about this idea: Two versions of it have emerged differing in whether the change is top-down or bottom-up. Does the group change the preferences of the individual members to ensure its own functioning, or do the group preferences emerge thanks to the preference change on the individual level?

The top-down version is exemplified by remarks in List and Pettit’s Group Agency. They discuss preference change in the context of incentive-compatibility, where they want to show how group-supporting cooperative and truthful actions can be rational for the individual members of a group (List & Pettit 2011: 105-106). According to List and Pettit, one way for the group to address an incentive-compatibility problem would “be to try to transform the members’ preferences from those that generate strategic incentives into those that make cooperativeness or truthfulness incentive compatible.” (List & Pettit 2011: 126)

They are not very specific on how this preference change comes about and only point to “educational, communicative, or social measures that change the group’s informal norms and culture” (ibid.). How to describe and model this preference change is left open, but it is clear that it is top-down. The group needs to get its members in line with its mental life. It needs to make sure that it is rational for the individual members to act according to the requirements of a well-functioning group agent. Thus, preference change contributes to group agency by enabling the candidate mental states of groups to take effect in the world.[0]

In his 2014 Theory of Conditional Games (see also Tummolini & Stirling 2018), Stirling suggests that there is an opening for attributing group decisions “when the preference of stakeholders are influenced by the preferences of other stakeholders” (Stirling 2012: 76). While he also considers preference change as enabling group agency, there is a marked difference in how Stirling understands and models the dynamic compared to List and Pettit. According to Stirling, preferences are conditional on the preferences of other group members. Without going into the formal details, these conditional preferences (together with other assumptions) allow to derive a group utility. The approach is bottom-up. The conditional preferences lead to[1] the group having motivations. The group does not seek to create or impose them to ensure its own functioning.

As Brian Epstein (2017) likes to remind us, groups come in various kinds of great diversity. Thus, one should be wary to assume that either the top-down or bottom-up account holds a monopoly of describing how preference change enables group agency. The reading group and the multinational cooperation might not exhibit the same type of individual preference change. Empirical research might be required to answer questions about particular groups.

There are, however, issues that can be addressed from the philosophical arm chair at the present stage. One direction would be to compare these types of preference change with the internal functioning of human cognition. Would a similar distinction between top-down and bottom-up preference change make sense in this realm? The homuncular functionalism of Lycan and Dennett would at least provide the lower levels of agency to apply these views.

A normative question I would like to see addressed is whether these two types of preference change differ in how they affect autonomy. At least prima facie, top-down preference change might seem to be a form of paternalism at danger of infringing upon autonomy, but I wonder whether conditioning one’s preferences on other agents could be an instance of “selbstverschuldete Unmündigkeit” under some circumstances.

In any case, preference change in the context of groups has only received marginal attention so far and I believe multiple issues (group agency, extended cognition, social rationality) could benefit from future research in this direction.


Footnotes

[0] The distinction is probably not as clear cut as I suggest in this post. Given List and Pettit’s aggregation view of group mental states, the preference change on the lower level, albeit under the influcence of the group, would also affect the motivations at the higher level.

[1] I am myself not sure how to interpret this “lead to”. Perhaps it is an instance of partial grounding. It appears to be an explanatory, directed, non-causal relationship.


References

Epstein, B. (forthcoming). What are social groups? Their metaphysics and how to classify them. In Synthese: 1-34.

List, C. & Pettit, P. (2011). Group Agency: The Possibility, Design, and Status of Corporate Agents. Oxford University Press.

Stirling, W. C. (2011). Theory of Conditional Games. Cambridge University Press.

Tummolini, L. & Stirling, W.C. (2018). Coordinated Rational Choice. In Topoi. https://doi.org/10.1007/s11245-018-9589-6