Abstract
Raimo Tuomela is one of the pioneers of social action theory and has done as much as anyone over the last 30 years to advance the study of social action and collective intentionality. Social Ontology: Collective Intentionality and Group Agents (2013) presents the latest version of his theory and applications to a range of important social phenomena. The book covers so much ground, and so many important topics in detailed discussions, that it would impossible in a short space to do it even partial justice. In this brief note, I will concentrate on a single, though important, theme in the book, namely, the claim that we must give up methodological individualism in the social sciences and embrace instead irreducibly group notions. I wish to defend methodological individualism as up to the theoretical tasks of the social sciences while acknowledging what is distinctive about the social world and collective intentional action.
This is a preview of subscription content, log in via an institution to check access.
Notes
- 1.
Of course, what follows is not a full defense of methodological individualism but a partial defense of it against the main arguments in Social Ontology.
- 2.
Given the definition of ‘we mode intention’ (WMI) on page 68, it is clear that the I-mode notions on page 70 are defined so as to exclude the we-mode, that is to say, to act or intend in the I-mode definitionally excludes acting or intending in the we-mode. Thus, there appears to be no hope of reducing acting in the we-mode to acting in the I-mode.
- 3.
Tuomela discusses Bratman’s account explicitly, but mischaracterizes it: “In his [Bratman’s] account the basic cooperative intention has the form “I-intend that we perform joint action X,” where X instantiates an individually shared goal G (recall (CIM)). His account does not make use of the constitutive feature of collective acceptance nor the other central elements of the we-mode framework. Accordingly, he deals with a weaker notion of cooperation than my notion of we-mode cooperation. His account seems to deal rather with “pro-group I-mode” cooperation and would seem to be a special case of my (CIM) account” (pp. 158–9). But, first, the characterization of the basic cooperative intention is incorrect. For Bratman, for one to we-intend, one must intend (at least) that the group J by way of the intentions of each that the group J by way of meshing subplans of their intentions that they J (in his (2014) account this is secured in modest sociality by the mutual responsiveness condition). That is a much more powerful condition than simply that I intend that we perform joint action X. This is what secures that the collectivity condition is met. Furthermore, one cannot object here that the collective acceptance condition is not met unless one can otherwise object to the account. For if the account of shared intention that rests on this account of the content of a participatory intention at least provides sufficient conditions for joint intention, since joint intention entails collective commitment, he has thereby secured that the collective commitment condition is met. And for similar reasons one cannot object that Bratman’s account is not adequate to the idea of authoritative group reasons without further argument. Tuomela raises a question about whether Bratman’s account (and the same question can raised about my account) would be circular if we instantiate ‘J’ to an essentially intentional collective action type, for example, playing chess, or shaking hands. But if these types of actions can be analyzed into a component that is neutral with respect to whether it is being instantiated intentionally and the requirement that it be instantiated jointly intentionally, then there is no harm in embedding the concepts in the content of joint intentions, for then the participants know that to execute the intention to do something intentionally together, they need merely we-intend that they instantiate the neutral component. See (Ludwig 2014, n. 11).
- 4.
I think what I say below shows how the response would go to arguments in chapter 6.
- 5.
- 6.
One would not expect that being a team reasoner, as opposed to not, would leave one’s individual preferences unaffected. Team reasoners are still individual agents, and they still have their own preferences. So the hypothesis that members of a group are team reasoners as opposed to individual strategic reasoners should leads us to reassess their preferences in putative cases of PD and Hi-Lo. If team reasoners have a commitment to maximizing or optimizing group utility, then that corresponds to a preference ranking in which maximizing or optimizing group utility per se is given a high value in each agent’s preference ranking.
- 7.
Hi-Lo makes it clear how this works. Team reasoning ranks choice profiles by Pareto efficiency. One profile is more efficient than another if one agent’s payoff can be increased without decreasing that of any other. In Hi-Lo, HiHi is the only Pareto optimal choice. Thus, in team reasoning, one choses Hi as one’s part in HiHi. The team benefactor, whose ranking mirrors the group ranking, still reasons as an individual, and so, armed with only the resources of classical game theory, is at an impasse. Additional principles have to be added to resolve PD because there are three Pareto optimal choice profiles. Bacharach assumes the players in team reasoning prefer (strongly enough) both cooperating to one free riding (Bacharach 2006, pp. 168–9).
- 8.
It is not clear that, in these cases, we should think of we-mode we-reasoning as aiming at maximizing group utility rather than, say, aiming at the Pareto optimal solution if there is a unique one, or a solution which maximizes group utility relative to the requirement of certain minima for all participants, or aims for the greatest least inequitable distribution of goods. Maximization of group utility per se would potentially require participants to find it rational to make any personal sacrifice (giving his or her organs to save five other members of the group, e.g.) as his or her part in maximizing group utility. But if we, for example, say that the group is aiming at a Pareto optimal solution, and if there is more than one, then the one with the highest least inequitable distribution, it is clear that even in pro-group I-mode we-reasoning there is only one unique choice for Hi-Lo (where there is one Pareto-optimal solution) and for PD (where, while there are three Pareto optimal solutions (CC, DC, CD), one yields the highest least inequitable distribution (CC)).
- 9.
What about PD? Let’s take pro-group I-mode we-reasoning first. There doesn’t seem to be any question about which combination of choices maximizes group utility—in our example. CC yields 6, DC and CD yield 5, and DD yields 4. Thus:
-
1.
You and I intend to maximize group utility.
-
2.
Choosing CC maximizes group utility.
Therefore,
-
3.
I will perform my component in HiHi, that is, Hi.
A fortiori, we-mode we-reasoning gets the right result, but without a contrast.
-
1.
- 10.
This is how Bacharach is thinking of team benefactors, who still reason like strategic individuals.
- 11.
I say ‘justifiably’ because if their beliefs are irrational or unjustified, even though they are able to reach the right result, we have not shown that they can do so by acting rationally. We-mode we-reasoning requires that members of the group believe the others will do their parts, but doesn’t prima facie require their beliefs to be justified. We-mode we-reasoning could result in the right solution without the members of the group being rational in doing so, if their admittedly true beliefs that the others are participating and all the conditions for success are in place are not rational. So what is needed is not just we-mode we-reasoning (and the same goes for pro-group I-mode we-reasoning) but we-mode we-reasoning in which the participants are justified in engaging in that form of we-reasoning. To put it another way: if they are not justified in believing the premises of the arguments above, they are not justified in the practical conclusion they reach on its basis. So for we-mode we-reasoning (or pro-group I-mode reasoning) to be a rational approach to solving a collective action dilemma, the group members must have reason to think we-reasoning will lead to success. What reason do they have for this if all the information they have is that given by the standard game-theoretic payoffs? While we-reasoning leads in Hi-Lo to what we think of as the best choice, it looks reasonable only if one has reason to think the other is we-reasoning as well. In my example involve the two restaurants, this secured by a prior agreement to eat lunch together. In the case of situations involving no prior agreements or strangers, we face a problem. The empirical problem might be solved by saying that people we-reason by default. The conceptual problem is solved only if we add that they reasonably expect others to we-reason. Perhaps this is reasonable, but if so, it is so because we have a broadly inductive assurance that this is a default reasoning mode, which can be undermined in particular circumstances—which may explain why in PD we see both C and D chosen by many people.
References
Bacharach, M. (1999). Interactive team reasoning: A contribution to the theory of cooperation. Research in Economics, 53(2), 117–47.
Bacharach, M. (2006). Beyond individual choice: Teams and frames in game theory. Princeton: Princeton University Press.
Bernheim, B. D. (1984). Rationalizable strategic behavior. Econometrica, 52(4), 1007–1028.
Bratman, M. (1992). Shared cooperative activity. The Philosophical Review, 101(2), 327–341.
Bratman, M. (1999). I intend that We J. In Faces of intention: Selected essays on intention and agency. Cambridge: Cambridge University Press.
Bratman, M. (2014). Shared agency: A planning theory of acting together. Oxford: Oxford University Press.
Gilbert, M. (2006). A theory of political obligation: Membership, commitment, and the bonds of society. Oxford: Clarendon.
Ludwig, K. (2007). Collective intentional behavior from the standpoint of semantics. Noûs, 41(3), 355–393.
Ludwig, K. (2014). Proxy agency in collective action. Noûs, 48(1), 75–105.
Pearce, D. G. (1984). Rationalizable strategic behavior and the problem of perfection. Econometrica, 52(4), 1029–1050.
Risse, M. (2000). What is rational about nash equilibria? Synthese, 124(3), 361–384.
Tuomela, R. (2013). Social ontology: Collective intentionality and group agents. New York: Oxford University Press.
Acknowledgement
I wish to thank Raul Hakli, Kaarlo Miller, and Maj Tuomela for good critical comments related to the topic of this response.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2017 Springer International Publishing Switzerland
About this chapter
Cite this chapter
Ludwig, K. (2017). Methodological Individualism, The We-mode, and Team Reasoning. In: Preyer, G., Peter, G. (eds) Social Ontology and Collective Intentionality. Studies in the Philosophy of Sociality, vol 8. Springer, Cham. https://doi.org/10.1007/978-3-319-33236-9_1
Download citation
DOI: https://doi.org/10.1007/978-3-319-33236-9_1
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-33235-2
Online ISBN: 978-3-319-33236-9
eBook Packages: Religion and PhilosophyPhilosophy and Religion (R0)