Collective Intentionality,Individualism,and the Place of Children in Social Theorizing

Abstract

In Nonideal Social Ontology: The Power View, Åsa Burman argues that leading theories in social ontology fail to capture important aspects of our social world when they characterize it as paradigmatically cooperative and develops her power view to capture the centrality of power and conflict in our social lives. While Burman’s view is timely and compelling, I suggest that her arguments against the leading theories’ claim that collective intentionality is a constitutive element of our social reality drive her own view toward individualism, and that this, in turn, can lead to the continued exclusion of children from our social theorizing.

Nonideal Social Ontology: The Power View

In Nonideal Social Ontology: The Power View, Åsa Burman (2023) develops a persuasive argument that the field of social ontology is currently in the midst of a Kuhnian paradigm shift from ideal to nonideal social ontology. Burman both characterizes and advocates for this shift, claiming that it should be followed through because ideal social ontology fails to capture some important elements of our social reality. In its place, Burman develops a version of nonideal social ontology—her power view—that captures these missing elements, including largely undertheorized but critical social phenomena like economic class and non-deontic forms of social power.

In the book, Burman provides an overview of some of the major features of theories of ideal and nonideal sociology. With respect to ideal social ontology, Burman focuses on the work of social ontologists like Margaret Gilbert, John Searle, and Raimo Tuomela and shows that their theories are united by a particular conception of the social world: one which holds that fundamental social phenomena are paradigmatically cooperative. This is reflected in many of the main features of ideal social ontology, which Burman details by drawing on and extending Francesco Guala’s Standard Model of Social Ontology (SMOSO). Among these features is the methodology of ideal social ontology, which is characterized by highly idealized and abstract examples, in which small groups of adults (often, no more than two) of roughly equal status or power willingly collaborate in the pursuit of shared goals. This in turn leads to the assumption that the basic building blocks of social reality just are the features of such cases. For example, theories typically emphasize collective intentions as an essential feature of social reality, because they are thought to be necessary for social phenomena to be genuinely cooperative, and not a matter of accidental coordination among individuals.

To characterize theories of nonideal social ontology, Burman (2023, 121) draws on the work of Ásta, Johan Brännmark, Sally Haslanger, and Katherine Jenkins, noting that the latter two are examples of emancipatory social ontology, a subcategory of nonideal social ontology. Broadly speaking, theories of nonideal social ontology eschew the cooperative paradigm of ideal social ontology in order to capture aspects of social reality that are characterized by inequality, coercion, and conflict. In fact, the main dividing line between ideal and nonideal social ontology just is oppression: while ideal social ontology is silent on it, nonideal social ontology aims to describe it as a central element of social reality, and emancipatory social ontology aims to end it (Burman 2023, 155). Given this broad aim, nonideal social ontology often deploys real-world, detailed examples that involve conflict and unequal power relations.

As Burman (2023, 81) contends, the shift from ideal to nonideal social ontology is one that should take place, because ideal social ontology fails to live up to its own central aims of providing general theories of social and institutional phenomena as well as theories that can do foundational conceptual work for the social sciences. One reason for this failure is the relatively narrow focus of ideal social ontology. Burman (2023, 176) sometimes puts this in terms of power. That is, because ideal social ontology focuses on only deontic power—for example, the orderly rights and obligations that are generated by collective intentionality and institutional roles like “professor”—it systematically neglects other aspects of social power. For Burman (2023, 200), this is a mistake because social power is the “central social concept” and “nearly all the social facts in which we are interested contain one form of social power or another.”

To account for the centrality of social power in social reality, Burman (2023, 203) develops her power view, which introduces three new forms of social power: telic power, spillover power, and structural power. This new and more comprehensive taxonomy of social power is developed to expand our ability to analyze opaque social phenomena, which can exist even though no one has beliefs about them. A central example and test case for Burman is economic class, because it has been neglected by even extant theories of nonideal social ontology.

I am persuaded by many of Burman’s arguments in the book and believe that it is an important and timely contribution to the field of social ontology. Burman’s diagnoses of the various problems that plague standard theories of social ontology are also compelling. However, in what follows, I will highlight a few points of disagreement. First, I raise some concerns for Burman’s arguments about the role of collective intentionality in social reality. While I believe that the conclusion of these arguments—collective intentionality is not necessary for all social phenomena—is accurate, the arguments themselves are not fully convincing. Second, I suggest that Burman’s arguments here seem to drive her own view toward individualism. I argue that this move should be avoided because it is incompatible with some important and undertheorized social phenomena.

Collective Intentionality

Recall that, for Burman, ideal social ontology casts collective intentionality as one of the central building blocks of social reality and a necessary condition of institutions. ¹ In the second chapter of the book, Burman argues that this claim is false and aims to show that some paradigmatic institutions of social ontology like property and marriage can exist only on the basis of individual intentionality and common knowledge. I agree that it is plausible that collective intentionality, especially as it has been described in ideal social ontology, is unlikely to be an essential component of social reality. As several philosophers have recently pointed out, at least some theories of collective intentionality have the objectionable consequence of ruling out the possibility that young children engage in shared or collective action because they require the deployment of sophisticated cognitive skills (e.g., Butterfill 2012; Story 2021; Tollefsen 2005). Despite the plausibility of this claim, however, it is not clear the alternative of individual intentionality plus common knowledge fares any better in capturing social institutions or facts.

To show that the collective intentionality claim is false, Burman (2023, 92–93) first argues that “it is conceivable that institutions like private property can exist without collective intentionality.” To illustrate, Burman claims that it is conceivable that nearly everyone in Stockholm believes that all apartments are occupied by legal owners or those authorized to occupy the apartment by the legal owners. Further, it is conceivable that this belief is common knowledge among those who hold the belief. That is, in addition to having the belief, many people in Stockholm also believe that nearly everyone has the belief and believe that everyone believes that nearly everyone has the belief, and so on. Because this belief “refers to the institution of private property,” the suggestion here is that individual intentionality and common knowledge are sufficient for the institution of private property (Burman 2023, 92).

But it is not clear that this example has shown that the institution of private property does not depend on collective intentionality, as it is not clear that belief in particular a statement that refers to private property tells us very much about the institution itself. That is, this case might have shown that statements of the kind above that refer to private property need not involve collective intentionality, but presumably, the institution of private property extends far beyond what is captured by the statement, so the case does not show that the institution itself does not involve collective intentionality.

Still, even if we grant the conceivability argument shows that the institution of private property does not involve collective intentionality, it is not clear that institutions can depend on individual intentionality and common knowledge for their existence. In fact, this move seems to reintroduce some of the old problems of identifying what distinguishes social from similar but individual phenomena. As Gilbert (1996) points out, groups of individuals may have beliefs that coincide and even common knowledge about these coincident beliefs without thereby being members of a recognizably social group. To use Gilbert’s example, it might be the case that people with red hair who are over six feet tall have common knowledge about certain innocuous facts like the flammability of fire. This population is not obviously a social group, yet they meet the same criteria that Burman requires for institutions like private property. This suggests that there is some further element beyond just individual intentionality and common knowledge that is necessary for the existence of social institutions.

Second, Burman (2023, 95) argues that it is conceivable that collective intentionality is not required for institutional facts like getting married, and that such facts may only require individual intentionality and common knowledge. The key examples here are the strategic marriages of historical royals which were often more strategic than voluntary in nature. For Burman, it is conceivable that in these marriages, the queen and king would only have individual intentions to get married out of strategic interest and common knowledge of this fact.

But the move to individual intentions and common knowledge seems under-motivated. Although Burman is right to highlight the flaws of imagining that marriages are always entered into by perfectly willing participants, the case is not obviously at odds with collective intentionality. That is, while strategic marriages may not be fully voluntary, they can be voluntary to a degree that is sufficient for collective intention. To see why, we can first consider a helpful distinction discussed by Aristotle in Book III of his Nicomachean Ethics. According to Aristotle, some actions involve both a voluntary and involuntary element—they are mixed—because although they would not ordinarily be chosen, certain circumstances can make them worthy of choice, and for this reason, they are more like voluntary actions (Aristotle 2000, 1110a). With respect to the king and queen, we might therefore say that their marriage is voluntary in this qualified sense, because matters of national strategic interest make the marriage choice-worthy.

Furthermore, it is plausible that at least some theories of ideal social ontology have the resources to deal with such cases. For example, in her work, Gilbert (2009, 172) stresses that the plural subject, which is the main social concept in her view, is distinct from the individuals that come together to form it. More specifically, for Gilbert, participants in some joint endeavor may have collective intentions that do not match, or even conflict with their personal or individual attitudes. This suggests that, on Gilbert’s view, a king and queen may share a collective intention to get married (perhaps for the sake of strategic interests) even though they are individually or personally opposed to getting married, and perhaps even if they lack an individual intention to get married. Of course, this does not demonstrate that Gilbert’s argument is correct or should be accepted. However, it does seem that ideal social ontology can provide a response to this sort of objection. In addition, the earlier argument about the problems about distinguishing between social and individual phenomena is also relevant here. If individual intention and common knowledge are sufficient for institutional facts like marriage, in what sense are such facts social at all?

Burman’s third and final argument against the collective intentionality claim targets Tuomela’s claim that we-attitudes in the we-mode—Tuomela’s analysis of collective intentions—are required for institutions. ² For Tuomela (2000, 132), a token attitude like a belief or intention is in the we-mode just in case those who have the attitude have collectively accepted that the attitude is one that is for the use of the group. The we-mode also involves commitment to the group, as well as a property he calls “forgroupness.” This latter property refers to the collective availability of propositions that are correctly assertable for the group, because the group has collectively accepted them. As Burman (2023, 98) observes, it also seems to require that members of the group do things that contribute to or are in keeping with the group’s realm of concern. And yet, Burman argues, it is conceivable that there are institutions that are not beneficial or useful for its members, either because they started out that way, or because changing background conditions made them that way. And, such institutions may continue to be maintained even if participants fail to meet Tuomela’s requirements for we-attitudes in the we-mode.

But here again, it is not clear that this argument demonstrates that collective intentionality is not required for social institutions. As noted above, Tuomela’s we-attitudes involve the notion of collective acceptance, which means that the attitudes of the group are just those that the group has collectively accepted. But we can imagine that attitudes that the group has accepted can and does diverge from the individual attitudes of the members of the group. This is supported by Tuomela’s (2000, 133) claim that propositions that are correctly assertable for the group must be assessed in contexts where groups members act as group members and not private persons. This means that truth “for the group” should not be “confused with ‘objective’ truth” (Tuomela 2000, 133). Thus, we might say that the fact that an institution never was or is no longer beneficial for its members, does not show that these members do not meet Tuomela’s requirements for we-attitudes in the we-mode. The institution may continue to benefit the group in the group-mode, that is, based on what the group’s members have collectively accepted, even as it no longer objectively benefits the group’s members. This is at least an available interpretation of Tuomela’s comments, which should be addressed in order to offer a full defense of the claim that we-attitudes in the we-mode are not necessary for social institutions.

It therefore seems to me that there are several rejoinders available to those who are sympathetic to the collective intentionality claim that Burman rejects. This shows that there is at least more to be done to refute the collective intentionality claim along these lines. In the next section, I briefly highlight and discuss another kind of argument against the collective intentionality claim.

The Place of Children in Theories of Social Ontology

In a paper about the puzzling tendency of philosophers of social action (among them, Gilbert, Searle, and Tuomela) to offer individualistic analyses of social phenomena, Annette Baier (1997, 29) highlights the case of early childhood development. As Baier points out, although children are a major part of social reality, they are rarely explicitly discussed in theories of social ontology. For Baier, this exclusion impedes our understanding of the basic or paradigmatic social phenomena because much of a child’s life is paradigmatically social. In fact, children are almost entirely reliant on their adult caregivers in the first few years of life and depend on them not only for the basic elements of survival but also for their emotional well-being and ongoing social and cognitive development (e.g., Tomasello et al. 2005).

In addition, children gradually begin to acquire concepts of many of the social phenomena that both ideal and nonideal social ontologists are concerned with in early childhood. Children acquire many of these concepts through learning, a process which also seems paradigmatically social, at least in early childhood. And, while children can of course absorb information that adults are unaware they are providing, developmental psychologists believe that this, later, more individual process is scaffolded by early shared or collective intentional interactions with their adult caregivers (Tomasello and Carpenter 2007; Tomasello and Rakoczy 2003). All of this suggests that there is much to be learned about social phenomena from our acquisition of social concepts in early childhood.

I believe that there are two lessons that can be drawn from the above considerations. The first is that, as Burman argues, it is unlikely that collective intentionality is the basic building block of social reality. As noted earlier, several philosophers have recently argued that collective intentionality cannot be a necessary part of all social phenomena, because this would implausibly exclude young children from participating in those phenomena.

Here, I focus on Butterfill’s (2012) argument, as it presents the issue in succinct way. As Butterfill (2012, 24) points out, some psychologists believe that social interactions in early childhood play a crucial role in the young children’s development of theory of mind, including their understanding of concepts like belief, knowledge, and other forms of higher cognition. However, analyses of collective intentions typically presuppose the ability to use and understand these concepts. For example, Gilbert, Searle, and Tuomela all require that participants have the capacity to form beliefs about beliefs, and otherwise keep track of the mental states of one’s co-participants. This suggests that collective intentionality cannot be required for children to engage in genuine social activity in early childhood, so collective intentionality is not a basic building block of all social phenomena. Notably, Butterfill develops alternative account of collective action involving young children that does not require collective intention and also avoids individualism.

The second lesson is due to Baier (1997), and it is that the inclusion of children in theories of social ontology would discourage the development of individualistic analyses of social phenomena. This is in part because it is much harder to analyze social phenomena involving children in terms of the individual and discrete contributions of the child and the child’s co-participant. Consider the case of feeding an infant. As Baier observes, this sort of action cannot be understood in terms of discrete individual contributions neither of the individual contributions are sufficient for the act of feeding to take place. The act of feeding is interdependent, and the adult caregiver cannot feed the child without the contributions of the child, nor can the child be fed without the contributions of the adult. This suggests another kind of objection against Burman’s strategic marriage case: because marrying is a necessarily social or interdependent action, it is not clear that we can understand it in terms of individual intention and common knowledge. This is especially clear if we accept Baier’s (1997, 25) principle of “the limited sovereignty of intentions” which holds that we cannot intend what we do not take ourselves to control. And, because marriage necessarily involves two people, it seems like an individualistic understanding of even strategic marriage would run afoul of this principle. Notably, this objection also applies to Burman’s (2023, 205) later definition of social facts as “any fact involving the intentionality of two or more agents.”

I believe, however, that these points are aligned with the spirit if not the letter of many of Burman’s aims. For example, Burman’s power view may be fruitfully employed in the analysis of the fundamentally asymmetrical nature of the social interactions that young children engage in during their first few years of life. In addition, there is nothing about Burman’s arguments that seems to require the adoption of individualistic analyses of social phenomena, even if collective intentions are rejected. Instead, some other feature that can distinguish social from individual phenomena can be found.

ORCID iD

Ayana Samuel https://orcid.org/0009-0002-4480-0264