This volume comprises nine new papers on conceptual and methodological questions of collective action. ‘Collective action’ has a more inclusive meaning in philosophy than in economics. In economics, it refers to a particular kind of market failure and ways of overcoming it. The tragedy of the commons and free riding on public goods are examples of collective action problems. In philosophy, ‘collective action’ may refer to a variety of situations ranging from shared activities (such as going for a walk together) via coordination problems to cooperation and aggregation problems, as investigated in game and social choice theory.
The volume is divided into two parts. The first part, ‘Collective Attitudes and Actions’, focuses on conceptual and ontological matters such as whether or not groups in fact perform actions and hold attitudes, such as beliefs and desires, over and above those of their individual members or whether speaking this way is a mere paraphrase. The second part, ‘Collective Rationality’, focuses on the question of how the rational properties of a group relate to those of its members. For example, is a group rational only if all of its members are? Or might the rationality of the group come at the expense of the rationality of its members? Taken on the whole, the volume is slightly skewed towards social ontology.
Both families of questions – ontology and rationality – are of interest to economists. For example, whether or not groups really have attitudes and perform actions is relevant to microeconomics, in which households, firms, and states are often considered to be agents. If, however, firms and states are merely legal or pragmatic fictions, then microeconomic models fail to accurately represent what is the case. The models may have other merits such as making predictions, illustrating, or explaining phenomena. However, being realistic is a property that seems worth having.
In this review, I summarize select contributions focusing mostly on social ontology in Sections 1 and 2. I point to some flaws in particular arguments, by which I hope to illustrate the potential of seeking synergies with related debates in the philosophy of mind. In Section 3, I observe that many contributions exhibit a significant shortcoming: crucial notions are left unanalysed, so that relevant distinctions are overlooked and some disagreements seem verbal.
1. ONTOLOGY
Most of the nine essays in this volume extend lines of argument that the authors have pursued in earlier work. For readers unfamiliar with the debate, the editors open the volume with an overview introducing the positions associated with five principal authors of the field: Michael Bratman, Raimo Tuomela, John Searle, Margaret Gilbert and Philip Pettit. This is valuable to enable outsiders and newcomers to join the debate. Though the overview is helpful, it struck me as peculiar that the natural way of introducing the positions seems to be a serial summary of the views associated with individual authors. This indicates that the field is lacking a generally accepted systematic taxonomy.
In the first chapter, Deborah Tollefsen illustrates how cognitive science can contribute to the philosophical analysis and the explanatory project arising from the question of how individuals act together. She critically examines Bratman's requirement that shared intentions ‘require that the participants be mutually responsive to the intentions of other participants’. Developmental psychology suggests that children younger than four years lack this ability. Nevertheless, it seems that ‘prima facie they do engage in joint action’ (15). The case of young children might hence be a counterexample to Bratman's proposal. Tollefsen then argues that philosophical analysis and empirical explanation of such joint actions should incorporate motor intentions in order to arrive at ‘a more psychologically informed account’ (28). Motor intentions represent specific bodily movements that an individual intends to perform in an action. These intentions would be an addition to the kinds of intentions that figure into the philosophical analysis of action. Philosophers usually distinguish between prior intentions or plans on the one side and intentions-in-action or proximal intentions on the other. Tollefsen suggests that motor intentions play an explanatory role in the coordination of bodily movements between participants in joint actions.
Subsequent chapters address the ontology of collective actions and attitudes head-on. In the third chapter, Frederick Schmitt confronts an objection to the view that groups can have beliefs of the same kind as an individual's beliefs. According to this objection, groups merely accept but do not believe propositions; and belief and acceptance are two very different cognitive attitudes characterized by different features. In response, Schmitt argues that attitudes of groups often exhibit features that are characteristic of proper beliefs: they aim at the truth, they are formed non-voluntarily, and their formation requires holding other beliefs. Hence, groups often have beliefs of the same kind as an individual's beliefs.
In the fourth chapter, Robert Rupert challenges this view. He pursues three lines of argument against group cognitive states as expressed in statements such as ‘Microsoft intends to develop a new operating system’. First, he argues that attributing cognitive states to groups is not necessary for causal explanations, because ‘there seems to be available a complete causal explanation couched in terms of the cognitive states of individuals’ (98). Second, there is no independent evidence for the existence of collectives’ cognitive states. In contrast, introspection seems to provide independent evidence for the existence of individual cognitive states. Finally, Rupert argues that there is an important difference between disposing with talk about cognitive states of groups and doing so with talk about cognitive states of individuals. If we were to eliminate individuals’ cognitive states, this talk would have to be reformulated in terms of neuroscience. Neuroscience would replace psychology. This would be a reduction between two domains.
There are several problems with these arguments. Some of the problems tie in with a larger methodological issue, which I address later. At this stage, I want to make three points. First, in response to the question of how statements about collective cognitive states, such as ‘Microsoft intends . . .’, can seem true if there are no collective cognitive states, Rupert writes: ‘I take a [cognitive] state to be the instantiation of a property by an individual’ (105, my emphasis). This forecloses a serious investigation into whether collectives instantiate those properties as well. Rupert continues his reply with an analogy. He argues that a sentence like ‘Microsoft intends . . .’ poses no greater challenge to his view than a sentence like ‘The average American family consists of 3.14 persons’. This is a bad analogy. Microsoft might be a collective agent, a statistical average could not possibly be.
Second, Rupert argues that collective intentions ‘have no causal work to do’ (98). This argument depends on a contested principle about causation. He argues that collective intentions are causally irrelevant because, if you consider alleged collective actions then ‘the occurrence of these events is due entirely to communication among individuals’ (99). Rupert's argument turns on the principle that if an event – the collective action – has a cause on the individual level, then there cannot be a distinct cause of the same event on the collective level. This sounds very much like the so-called exclusion problem for mental causation. One lesson of the vast literature on this problem is that it demands careful distinctions (see Stoljar Reference Stoljar, Hohwy and Kallestrup2008). It is a shortcoming that Rupert does not avail himself of the results of discussion of this problem in the philosophy of mind. For anyone familiar with the debate in philosophy of mind, the brisk way with which Rupert glosses over this issue is astounding.
Finally, Rupert's principle that permits reducing entities within a discipline but resists reducing entities between two disciplines is dubious. Often boundaries between academic disciplines are blurry and drawn intuitively. Why should they play such a crucial role in answering the metaphysical question of what exists?
In chapter 5, Kirk Ludwig sets out to score another point on behalf of those who deny the reality of collective entities. Unlike Rupert, who targets collective attitudes, Ludwig focuses on collective actions. He develops an argument for the familiar hypothesis that statements about collective actions are, strictly speaking, just statements about individual actions, which seems to be a paraphrase argument in the tradition of Quine. The idea is to paraphrase away reference to entities to which a statement seems ontologically committed. Ludwig, more specifically, argues for a pragmatic shift. He concludes that ‘when we assert group action sentences . . . we are asserting that certain events had multiple agents’ (128). This is an interesting development from his earlier work, in which he suggested the paraphrase to be a semantic phenomenon (Ludwig Reference Ludwig2007). At any rate, the proposal is that when we assert ‘The Supreme Court went to lunch’, we do not assert that there is a collective action of which the Supreme Court is an agent. Instead, we actually assert a statement such as ‘Each of the members of the Supreme Court went to lunch together’. Ludwig supports this view by regimenting the different action sentences in a first-order logic and by discussing the metaphysics of how each individual member in a collective action stands in the agency-relation to the same action.
2. RATIONALITY
Part two addresses questions of rationality. In chapter six, Abraham Roth poses a challenge to the view that groups hold attitudes over and above their individual members. It is generally acknowledged that an impossibility result of judgement aggregation, the so-called Discursive Dilemma (List and Pettit Reference List and Pettit2002), forms the basis for an argument for this view. Roth reconstructs two versions of this argument, which he calls the ‘explanatory’ and the ‘practical indispensability’ argument. Roth argues that a proponent of the view that there are group attitudes faces a dilemma because each version of the argument is problematic in its own way.
The explanatory indispensability argument proceeds ‘from the vantage of the social scientist’ (140). The argument is that there are group attitudes because the theory social scientists use to explain social phenomena is committed to their existence. As Roth points out, the success of the explanatory version of the argument for group attitudes depends on whether the best scientific theories indeed are so committed.
In contrast, the practical indispensability argument proceeds ‘from the point of view of the agent or deliberator’ (140). The argument is that there are group attitudes because the members of the group decide to escape the Discursive Dilemma in this way. The individuals might perform actions based on attitudes of the group, which they themselves do not accept, in order to maintain consistency with, for example, past actions of the group. In its practical version, the argument faces a different problem. It seems to require individuals to enact an outcome despite preferring a different one, which seems irrational. Proponents of this version would have to explain what individuals are doing when they act on behalf of the group, argues Roth.
In chapter eight, Paul Weirich argues that a collective action is rational if and only if the actions of each of the individuals involved is rational. This striking thesis has several apparently implausible ramifications. For example, the only collectively rational outcome of a Prisoner's Dilemma would be universal defection because defecting is the only individually rational strategy (in the sense of mutually best response). While universal defection is collectively irrational (in the sense that it is inefficient), Weirich argues that this outcome should be called collectively rational. He extends this claim to social choice theory. This commits him to the implication that inconsistencies can be rational. The Discursive Dilemma shows how under specific circumstances, a systematic aggregation of individually consistent sets of opinions results in a group having an inconsistent set of opinions. Weirich writes, ‘inconsistencies arising from [the individuals’] rational votes are not collectively irrational’. Weirich's argument is hard to follow. For example, he supports his claim with considerations in deontic logic. However, in deontic logic, the relevant deontic operator is interpreted as encoding permissibility. It is unclear how permissibility relates to rationality.
I am unable to comment on all contributions in this edition. As this selective review already suggests, they vary widely with respect to their rigour. The chapters by Tollefsen, Schmitt, Ludwig, Roth and Melinda Fagan (whose contribution on social epistemology I had to pass over to focus on social ontology) stand out by investigating cogently relevant issues, putting forward an original contribution and aiming at careful argumentation.
3. METHODOLOGICAL PROBLEMS
It is striking how frequently contributions appeal to unanalysed notions and intuitions. Mostly those intuitive judgements speak against group agency. For example, Roth registers his ‘recalcitrant metaphysical intuitions’ (139), Rupert declares that disposing with collective attitudes is the only ‘naturalistically viable option’ (108), and similarly Fagan worries that in a prominent account of group attitudes a central ‘relation is left mysterious’ (181). I see no deep problem with these judgements as such, but I do see a problem in the fact that such judgements are derived chiefly from intuitions. What are our metaphysical intuitions worth? Should intuitions guide us through this ontological terrain? Or should we rather attend to practices of social science, which contends to possess successful models of phenomena we seek to ontologise? Whether intuition or scientific practice provides appropriate data to build metaphysical theories is an important methodological question, to which many of the contributions do not pay much attention.
The good news is that these intuitions do not figure centrally in the respective arguments. Nevertheless, this may point in the direction of an important worry about the method of social ontology. In recent years, metaphysics has become increasingly self-conscious about its method and the nature of disagreements about concepts such as existence (see Chalmers et al. Reference Chalmers, Manley and Wasserman2009). Some hold the view that there is not a single concept of existence; rather, there are many. Moreover, when discussing ontology, a lot depends on what the conditions are for a thing to exist. However, among the contributions in this volume, statements of existence-conditions are conspicuously absent. It would be helpful if they were made explicit as a way of ensuring that disagreement about what exists is not due to disagreement about existence conditions.
For example, Ludwig and Roth seem to operate with the orthodox, Quinean view about what exists. According to this view, those things exist which are in the range of the quantifiers in the regimented statements of our best theories. Call this the theory-commitments view of existence. In contrast, Rupert seems to hold a truth-maker view about what exists. He suggests that we ‘take facts [about individuals] to be truth-makers’ of sentences about collective cognitive attitudes (105). According to the truth-maker view, those things exist which are needed in the world for the statements of our theories to be true. There are no collective attitudes, according to Rupert, because statements about collective attitudes are true or false in virtue of more basic facts about individuals.
Since they operate with different existence conditions, the theory-commitments view and the truth-maker view lead to different conclusions about what exists. In particular, the truth-maker view is more restrictive (Schaffer Reference Schaffer2008). Revisiting the contributions of the present volume in the light of this distinction, I am worried that the following deflationary hypothesis has some plausibility. Roth's explanatory indispensability argument (139) and Rupert's argument to explain the apparent truth of statements about collective states (105) come to different conclusions only because they implicitly appeal to different existence conditions.
Similarly, some contributions on the final topic of the volume, social rationality, would have benefited from paying greater attention to distinctions drawn in economics. The conceptions of rationality too remain implicit. For example, Weirich uses ‘rationality’ to mean Pareto-efficiency, utility maximization or consistency (i.e. being contradiction-free) depending on whether he talks about welfare analysis, decision theory or social choice. ‘Rationality’ needs to be disambiguated, especially when investigating the relation between individual and collective rationality, because here ‘individual rationality’ and ‘collective rationality’ may refer to two different senses of rationality. For example, in the case of the Prisoner's Dilemma, the worry about the relation between individual and collective rationality is that the outcome that results from the combination of mutual best response strategies (the Nash equilibrium) is inefficient in the sense that it is Pareto-inferior. Hence, the worry is not that the outcome is rational on one level and irrational on another. Instead, the worry is that the outcome is rational in one sense of rationality but not in another.
Despite these critical remarks, the present volume is to be commended as a collection of original and diverse contributions to debates on social ontology and social rationality. This book reflects the current state of these areas and provides many avenues into it for those unfamiliar with key debates.
ACKNOWLEDGEMENTS
I thank Ryan Cox, Bryce Huebner and Holly Lawford-Smith for helpful feedback on this review, which was written during a stay at the School of Philosophy of the Australian National University.
