Navigazione – Mappa del sito

HomeNumeri82Solidarity and Theories of Collec...


The concept of solidarity is of central importance to the political sense of collective action. But it is a curious fact that solidarity is virtually unmentioned across the large and growing literature in philosophical collective action theory. Instead, we see discussions of collective action overwhelmingly focus on epistemic conditions and group-level correlates of individual action explanations such as collective intentions, collective beliefs, and so on. The aim of this paper is to elucidate the relationship between solidarity and collective action theory. I will try to answer two questions: (1) What is it about the structure of accounts of collective action that seems to preclude discussions of solidarity and related concepts; and (2) How can solidarity be made to play a useful role in those accounts?

Torna su

Testo integrale


  • 1 Cf. Corlett 2001: 575.

1The term “collective action” has three important senses. As it is used in political theory, it may refer to coordinated action of groups to improve some aspect of their condition; for example, workers in a factory engaging in collective action to achieve better working conditions (Klandermans 1997). It may also refer to how a group of people coordinate their activities to cope with “collective action problems” such as a Prisoner’s Dilemma or a Tragedy of the Commons situation (Ostrom 2000). In analytic philosophy, the concept is somewhat more general. Although conceptions vary from one author to the next, it generally refers to any action whose subject is a group.1 The concept is meant to be quite general. It encompasses the case of the miners who go on strike; but it also includes much simpler cases, for example, two people walking down the street together and conversing. Throughout this paper, I will be referring to the latter sense of “collective action” unless stated otherwise.

2The concept of solidarity is of central importance to the political senses of collective action. Solidarity among the workers of the factory is crucial for explaining how and why they engage in the (political) collective action in the first place, to take one example. And solidarity can be invoked to explain how a group of individuals avoids spoiling the commons in a Tragedy of the Commons situation.

3But there is a large and growing literature which I shall refer to as “analytic collective action theory”, in which solidarity is conspicuously absent as an explanation of any type of collective action. This literature centers around important and valuable work by Margaret Gilbert (1996; 2006; 2009), Raimo Tuomela (2003; 2006; 2005; 2011), and Michael Bratman (1992; 1993; 2013). It embraces David Lewis’s (1969) seminal work on convention and Wilfred Sellars’s discussion of “we-intentions” (1963). The analytic collective action theory literature overwhelmingly focuses on epistemic conditions and group-level correlates of individual action explanations such as collective intentions, collective beliefs, and so on. With one recent exception by Raimo Tuomela which will be discussed later, it does not take solidarity into account at all; and even in Tuomela’s theory, solidarity does not (and indeed, cannot) figure in an explanation of collective action because conceptual priority runs in the wrong direction, from collective action to solidarity rather than vice-versa.

  • 2 Henceforth, when I refer to “collective action theory”, I will be referring to collective action th (...)

4The aim of this paper is to elucidate the relationship between solidarity and analytic collective action theory.2 I will try to answer two questions: (1) What is it about the structure of these accounts of collective action that seems to preclude discussions of solidarity and related concepts; and (2) How can solidarity be made to play a useful role in those accounts?

1. Analyses of collective action

5Writers on collective action have generally adopted a few methodological principles that have far-reaching effects on their analyses. In this section, I will explain three of these methodological principles.

1.1 Granularity

6Analytic philosophy has avoided constructing a general theory encompassing the whole range of plausible examples of collective action. Instead, these theories are overwhelmingly concerned with carefully delineated small subsets of collective actions, and each of these subsets is given its own unique treatment. I shall refer to these theories and this approach to the study of collective action as “granular”.

7For example, a highly influential theory due to Michael Bratman is concerned with what he calls “shared cooperative activity”. A glance at the conditions for shared cooperative activity makes it obvious that this does not (nor was it intended to) encompass the full range of collective actions. His account requires that “we” engage in a shared cooperative activity only if we have a shared intention, which he defines in the following way:

We intend to J if and only if:
1. (a) I intend that we J and (b) you intend that we J.
2. I intend that we J in accordance with and because of 1a, 1b, and meshing subplans of 1a and 1b; you intend that we J in accordance with and because of 1a, 1b, and meshing subplans of 1a and 1b.
3. 1 and 2 are common knowledge between us. (Bratman 1999: 121)

8Despite the fact that Bratman’s account is intended to take in a large class of important cases, the above requirements are demanding enough to exclude many classes of collective action. We have to intend that we perform the same action J, and we must individually intend to bring this about by having “meshing” subplans and common knowledge of each other’s intentions (and therefore, enough communication to create a condition of common knowledge). For plans to “mesh”, they cannot conflict, and there must be at least an implicit commitment to mutual support if either of our subplans threaten to go awry. Furthermore, it is not enough that our subplans “mesh” or that we have common knowledge of each other’s intentions; those facts must explain why each of us carries out our individual subplans.

9To see how demanding this is, we can easily come up with cases of collective action that fail to meet Bratman’s requirements. We might successfully coordinate our behavior towards a goal, even though our individual contributions are not contingent on the other’s action. For example, I might try to lift a heavy object that is pinning someone to the ground. Even if it is too heavy for me to move, I might stubbornly persist in the attempt. You, seeing me struggling with the object, pitch in to help. Let us stipulate that either of us would continue to try to lift the heavy object even if the other person were to give up. When we successfully free the person from the object, I think it’s quite intuitive to say that we performed a collective action. But this cannot be a “shared cooperative activity” in Bratman’s sense because our common knowledge of each other’s intentions does not explain why each of us carries out our individual action.

10Note that this is not a criticism of Bratman’s account, per se. And this example is not meant as a counterexample, either. Rather, it is intended to show that Bratman’s theory is quite granular insofar as it is meant to explain only a relatively narrow subset of what we would intuitively call “collective actions”.

11Other accounts of collective action from this literature are equally granular. Margaret Gilbert’s theory focuses on group actions that carry a “normative commitment” on the part of the individuals to contribute to the group’s action (Gilbert 1990). This is motivated by Gilbert’s observation that when an individual forms an intention to perform an individual action, they thereby “commit themselves” to it. But there are clearly cases of collective action where there is no such normativity. The example above is plausibly such a case.

1.2 Strong epistemic requirements

12The second methodological assumption is something we have already seen above. Theories of collective action typically place very strong – sometimes unreasonably strong – epistemic requirements on the individuals in order for their activities to be considered “collective actions”.

13There is a reasonable motivation for these epistemic requirements. Often, the problem of analyzing the concept of “collective action” is motivated by asking what separates a truly “collective action” from a mere set of individual actions. For example, if you and I just happen to show up at the same restaurant at the same time, this is merely a coincidental alignment of our individual actions, not a full-blown collective action. But if we make plans to eat lunch together at that time in that restaurant and we coordinate our arrivals deliberately, then there is some sense in which we’ve performed a (very simple) collective action.

14The obvious feature that distinguishes between these two cases is that in the former, we had no knowledge of each other’s intentions or plans; but in the latter, we do. Furthermore, my knowledge of your intentions explains a large part of why I went to the restaurant, and vice-versa. Hence, we are led to Bratman’s requirement that our plans are “common knowledge” among us in order for us to take part in a shared cooperative activity.

  • 3 I will follow the game theory and economics literature in referring to this as “interactive knowled (...)

15It is important to note that these epistemic requirements are especially cognitively demanding on the individuals because they produce so-called “interactive knowledge”. This is knowledge of another person’s knowledge, or knowledge of the form “I believe that you believe… that p”.3 Following game theory and economics literature, we distinguish between various “degrees” or “levels” of interactive knowledge, depending on how many times the “I believe that you believe…” is repeated. That is, if you and I have interactive knowledge of the form:

“I believe that you believe that p”.

  • 4 For an overview of how the technical work on interactive knowledge plays out in the context of coll (...)

16Then we are said to have “mutual knowledge that p” or “first-degree interactive knowledge”. If the “I believe that you believe” clause is repeated N times, then we are said to have “N-degree interactive knowledge that p”. And if we have interactive knowledge to every finite degree (that is, for all N), then we are said to have “common knowledge that p”.4

17Returning to the restaurant example, we can see how the interactive knowledge requirement plays out. Suppose I happen to know about your plans to go to the restaurant; I want to talk to you, so I ambush you at the restaurant. On these accounts, we would not have thereby performed any kind of collective action, joint action, or shared cooperative activity. Intuitively, this is because you had no intention to meet me at the restaurant and we did not coordinate our plans together. The best sign of this failure is that we had no interactive knowledge of each other’s intentions, and a fortiori, no interactive knowledge played any role in producing a collective action.

18The interactive knowledge requirement, therefore, seems to get a lot of cases right; it distinguishes many cases of merely accidental coordination by a group from the ones in which the members have actually coordinated their actions to produce a truly “collective action”.

  • 5 A notable counterexample is John Searle’s (1995) theory in which no relationships among the people (...)

19We see this interactive knowledge requirement in virtually every theory of collective action.5 Margaret Gilbert’s theory is typical. It requires “joint commitment” as a necessary condition for collective action in her sense; and joint commitment requires common knowledge:

In order to create a given joint commitment of the basic kind, each party must make clear to the others that he is ready to be jointly committed in the relevant way. Thus each must express a certain condition of his will. Once these matching states of the will are mutually expressed, and this is common knowledge between the parties, the joint commitment has been created. (Gilbert 2006)

  • 6 We should note that Bratman, Gilbert, Tuomela, and other writers often do not follow the standard u (...)

20What we see from this literature is that common knowledge (or some other form of interactive knowledge) is the glue, so to speak, that binds together individuals’ plans and intentions into a group-level intention. From there, if we follow these authors in requiring some form of group-level intention, then interactive knowledge quickly becomes a necessary condition for collective action.6

1.3 The individual action model

21The third methodological assumption is what I have elsewhere called the “wash, rinse, repeat” assumption (Chant 2017). It is the assumption that a satisfactory account of collective action ought to be modeled on our best accounts of individual action.

22When we think about what makes something an action performed by an individual, we immediately think of the individual’s beliefs and intentions. For example, the reason my flipping the light switch is an action is because I had the intention to turn on the light and the belief that flipping the switch is a good way to do that. So my goals, intentions, and beliefs are intimately tied up with my actions.

23When it comes to constructing a theory of collective action, it is natural to assume that if collective action is the group-level correlate of individual action, then we might expect there to be group-level correlates of goals, intentions, and beliefs that play a similar role in explaining actions performed by groups. As Raimo Tuomela puts the point:

The notions of goal, belief, and action are linked in the case of a group to approximately the same degree as in the individual case. In the latter case, their interconnection is well established; given that the person-analogy applies to groups…, these notions apply to groups as well. (Tuomela 2011)

24In other words, because goals and beliefs are relevant to the actions of individuals, the “person-analogy” between groups and individual persons suggests that group-level goals and beliefs ought to be equally relevant to the actions of groups.

25Hence, the “wash, rinse, repeat” methodology of collective action theory:

  1. Start with your favorite theory of individual action.

  2. Identify the elements that are most valuable in formulating that theory (e.g. intention, belief).

  3. Posit the existence of group-level correlates of those elements (e.g. collective intention, collective belief).

  4. Define those group-level correlates, for example, by providing conditions for a group to have a “collective intention”.

  5. Propose those group-level correlates play the same explanatory role in explaining the actions of groups as their individual-level counterparts play in explaining the actions of individuals.

26Of the elements that explain individual action, intention is the most prominent, for it is clear that an important differentiator between an action and a mere event is that the action somehow results from the person’s having an intention. But if so, then the methodology faces a tricky problem -- intentions are mental states, and groups do not have mental states, so how are we to understand a “group-level intention”? Thus, we see a disproportionate amount of effort in the analytic collective action literature dedicated to developing theories of group-level intentions.

27We have already seen this methodology play out in the case of Bratman’s “shared cooperative activity”. Even though this is an analysis of shared cooperative activity, the bulk of the theory is dedicated to specifying the meaning of “we intend that J” – in other words, specifying the conditions for a group to have an intention. This group-level correlate of individual intention does the heavy lifting in the theory of shared cooperative activity in the sense that when group intentions are understood, the rest of the theory follows immediately Similar remarks hold for a number of other theories in the analytic collective action literature. To take another example, in her analysis of a “shared”, “joint”, or “collective” activity (Gilbert 1990: 9), the bulk of Margaret Gilbert’s analysis is on what it is for the constituent members of a plural subject to have a “shared personal goal”, which shares important features with accounts of collective intentions. Like Bratman, when this concept is in place, the rest of the theory of “collective activity” falls into place immediately.

1.4. Why not solidarity?

28We started with a puzzling observation, namely, that solidarity goes virtually unmentioned in philosophical theories of collective action. Given that solidarity plays a powerful role in the generation of collective action, its omission should strike us as odd. Fortunately, we are now in a position to say something about why solidarity is so conspicuously absent.

29If there are cases of collective action in which solidarity plays an important role, the granularity of existing theories makes it possible for writers on collective action to exclude them from consideration. As we have seen, analytic theories of collective action have focused on cases involving plans, group-level intentions, cognitively demanding epistemic requirements, and shared goals. In other words, they have focused on cases in which there is a deliberate and typically well-organized effort for a group to cohere around a shared goal, and coordination among each individual’s “meshing” subplans to achieve that goal.

30To be sure, this is an important class of cases, and it is not necessarily a mistake for a theory to focus on it. These cases would include, for example, the coordinated actions of members of a corporation or other well-structured organization. They would also include the instances of collective action that fall under Raimo Tuomela’s “Bulletin Board View” (2005). But in these cases, solidarity (whatever it turns out to be) is not an important explanatory element. It is not necessary, for example, that the employees of a business have any solidarity with each other in order to coordinate their actions on achieving a goal of the business. The same might be true of other hierarchically structured groups.

31On the other end of the spectrum from these examples of highly organized groups, we have a set of much simpler examples involving small, unstructured groups. Of these, Gilbert’s “walking together” is probably the most prominent example (1990). She takes two people walking together to be a “paradigmatic social phenomenon”, exhibiting all the crucial features of an important class of collective action. The most important feature is that when two people walk together, they make a conditional commitment to take part in the shared goal, creating a set of obligations. The presence of these obligations and commitments in the service of a shared goal is what constitutes a type of collective action. And because Gilbert takes this simple example to be “paradigmatic”, and solidarity plays no role in it, she naturally does not include solidarity as an important explanatory element in her theory.

32Thus, the cases that tend to be the focus of analysis are ones in which either (1) we assume enough coordination and structure that solidarity is not needed, or (2) the circumstances are so simple and straightforward that solidarity is not needed, either. And the fact that the theories are granular enough to justify a focus on such a narrow range means that it does not seem problematic that a concept such as solidarity is not deployed.

33A similar point can be made about the emphasis on cases with strong epistemic requirements. This class of epistemically demanding cases includes circumstances where the individual members of some group have interactive knowledge. Furthermore, this interactive knowledge centers around a specific kind of belief, namely, a belief that the members of the group have a plan or intention to do something specific. Returning to the example in which we meet at the restaurant, we noted that there is a requirement that we have some degree of interactive knowledge about our plan or intention to have a meal together at the restaurant. In the theories we see from Bratman, Gilbert, Tuomela, and so on, the proposition around which we have interactive knowledge is taken as given. The question, as it is implicitly formulated, is not, “How do we come up with the plan or the goal?”. Rather, it assumes that we are given the plan or goal antecedently, leaving us with the question of how we coordinate our actions around the problem of achieving that goal.

34But solidarity is just as important for defining what our shared goals are as it is for determining how we achieve them. For example, solidarity among the workers at a factory may be essential for those workers to determine that (e.g.) having better working conditions is a goal that they can achieve together. Of course, solidarity will also play an important part in coordinating their actions around that goal as well. But if we are taking the existence of the goal or plan as given at the outset of our analysis, then it is no wonder that we may end up ignoring the features (such as solidarity) that determine which goals or plans the group ought to have.

35Finally, the “wash, rinse, repeat” methodology also has an unintended consequence of discouraging solidarity and related concepts from appearing in these analyses of collective action. Recall that this methodology focuses attention on group-level correlates of the features of individual action. It assumes that an account of the actions of groups will have a structure similar to accounts of individual action.

36This may seem reasonable, but it has a hidden consequence. If there is a characteristic of groups that has no individual-level counterpart, then it will naturally tend to be disregarded by theories of collective action. After all, if you assume at the outset that explanations of individual and group action will be very similar to one another, then you will probably be skeptical of any theory that relies on features of groups that are only to be found in groups.

37Solidarity and related concepts such as altruism have no clear individual form. That is, I do not have solidarity with myself in any meaningful sense. And on the standard account of altruism as any behavior that benefits another at a cost to oneself (Sober and Wilson 1998), there is no sense in which a person can behave altruistically toward herself. Therefore, solidarity, charity, altruism, and so on are unlikely to find a home in any theory of collective action that employs the “wash, rinse, repeat” methodology.

1.5 Tuomela’s account of solidarity

38As I have mentioned above, there is one example of solidarity making an appearance in the analytic collective action theory literature. This is Tuomela’s discussion of solidarity in his Social Ontology. Unfortunately, Tuomela’s discussion is incapable of shedding any light on how solidarity can help us to analyze or explain collective action.

39In the final chapter of his Social Ontology, long after this theory of collective action has been completely explained (without any discussion of solidarity), Tuomela breaks down the concept of “solidarity” into several subtypes. We have solidarity (a) between persons; (b) between the members of a group toward each other; (c) at the “group level” in a “structural or explanatory sense”; and (d) toward humanity as a whole (2013: 247).

40Tuomela’s focus is entirely on “group-level solidarity” in the sense of (c). This is the type of solidarity that reveals itself in the ability of a group to coordinate around achieving group-level goals or furthering what Tuomela calls the group’s “ethos”. It has two specific requirements: (1) that it “tends to satisfy and promote its ethos”; and (2) that the members of the group “tend to function (broadly) cooperatively toward each other in matters concerning group affairs” (2013: 247). His conclusion is that “necessarily, a paradigmatic we-mode group is a solidary group” (2013: 249).

41Even without getting into any of the details of his argument, this conclusion should not come as a surprise. After all, Tuomela has focused on a small subset of cases of solidarity – namely, the ones in which a group’s cohesion is sufficient to enable it to pursue group-level goals. But this is exactly what a paradigmatic we-mode group is understood to be in the first place. As a “we-mode” group, the group has accepted a certain “ethos” (which includes goals), and the members of the group normatively ought to accept that ethos because the group has. And as a “paradigmatic” we-mode group, the group “determines its own ethos” (p. 56). Thus, we are left with the conclusion that a group that has accepted a certain set of goals, and whose members have all accepted those goals are necessarily groups that tend to promote those goals and whose members act cooperatively toward each other to do so. On a thick enough conception of “accepting” a goal, the conclusion that “a we-mode group is a solidary group” follows trivially from the stipulated definition of a solidary group. Indeed, Tuomela’s account cannot shed any light on how solidarity gives rise to collective action because his stipulated concept of “solidarity” is too narrow and ends up overlapping completely with his earlier concept of a “we-mode group”, which has already been analyzed without the help of solidarity.

2. Solidarity delineated

42In order to bridge the gap between analytic theories of collective action and solidarity, it is not enough to explain why solidarity is absent from those theories. We must also understand what concept of solidarity we are working with, and distinguish which forms of solidarity are potentially relevant to collective action theory.

  • 7 Led by the influential report of (Prainsack and Buyx 2011).

43There is a rapidly growing literature on the solidarity concept, with perhaps the fastest growth in the medical or health care arena.7 But it is safe to say that no single agreed-upon definition of solidarity has emerged in any literature. We will approach the issue of solidarity indirectly, by focusing on other concepts that are sometimes confused with solidarity, but ought to be kept separate. This will leave us with a somewhat vague, but still workable, understanding of the concept.

  • 8 See (Prainsack and Buyx 2011) for a careful analysis of the differences among these concepts. In wh (...)

44As has been noted elsewhere, solidarity is often discussed alongside a cluster of related concepts such as altruism, charity, empathy, and social cohesion. Distinguishing solidarity from these related concepts will go a long way toward a positive understanding of it.8

45Whatever “solidarity” amounts to, it sometimes manifests as a willingness to put aside one’s own interest in favor of the interests of someone else. This observation makes solidarity seem closely associated with altruism, which is often defined as any behavior that benefits someone else at a cost to oneself.

  • 9 Similar considerations show that Habermas (2017) is wrong to say that “Someone who acts in solidari (...)

46However, altruism is both too broad and too narrow to fit our understanding of solidarity. For example, it is too broad insofar as it is altruistic for me to donate money to a charitable organization that will decide which needy person or group the money will go to. Here, I have behaved altruistically because I have incurred a cost to benefit someone else. But intuitively, this cannot be an instance of solidarity because solidarity is with the members of a particular group, and I have no idea which group the recipient of my donation belongs to. And it is too narrow because one can act in solidarity with others in order to achieve an outcome that is beneficial to the entire group, as when workers go on strike to establish better working conditions. This is not altruistic because everyone comes out ahead; nobody is sacrificing solely for someone else’s benefit.9

47Solidarity is not the same as charity, either. Solidarity concerns my association with others who are similar to me in some way, or who are “in the same boat” as I am. But I can behave charitably toward people who are far less fortunate than I am, and who are definitely not in my boat. Indeed, appeals to charity typically are directed at people who do not have much in common with the people who are in need.

48Solidarity is a unique concept insofar as it connotes a sentiment, but it cannot be a mere sentiment. In this way, it is different from empathy. Although solidaristic behavior can certainly involve feelings of empathy, empathy by itself is insufficient for solidarity. Solidarity intimately involves action or a willingness to act, whereas empathy does not.

49Contrasting solidarity with these related concepts helps bring out a few key features of the concept, namely:

  1. There is no solidarity without action. It is not merely a sentiment, but rather a willingness to act.

  2. Solidaristic action may or may not benefit the person who acts.

  3. Solidarity exists among individuals who are similar or “in the same boat” with respect to some salient characteristic.

50We are therefore led to an understanding of solidarity that is compatible with the observation that:

Solidarity… involves one person giving a certain subset of the interests of another person a status in her reasoning that is analogous to the status that she gives to her own interests in her reasoning (Hussain 2018).

with the addendum that the “status that she gives” is sufficient for motivating the individual to act. However, this notion of solidarity is too broad on its own because it is compatible with altruism and various other-regarding behaviors.

51Better is the concept of solidarity from Prainsack and Buyx that:

Solidarity is an enacted commitment to carry “costs” (financial, social, emotional or otherwise) to assist others with whom a person or persons recognise similarity in a relevant respect (2017).

2.1 Examples

52This conception of solidarity is broad, encompassing a very wide range of behaviors. For example, suppose that you and I are passengers on the same plane, which is delayed at the airport. Seeing that we are “in the same boat” and “relevantly similar” because we are both inconvenienced by the plane’s delay, I offer you my cell phone so that you can make new arrangements for when we arrive at the airport. This counts as a case of individual solidarity because I have acted out of a willingness to carry a cost (albeit a very tiny cost) in order to assist you, and I’ve done this out of a recognition of a relevant similarity (being inconvenienced by the delay of our plane).

53At the other end of the spectrum, when the workers at a factory walk off their jobs at a specific time in order to protest poor working conditions, each worker incurs a cost – perhaps in the form of a risk that they will lose their job – but takes part in the walkout despite this. The workers have a commitment to carry this cost in order to assist others. And they do this from the recognition of a salient similarity with each other, namely that they all work in the same factory under the same poor conditions.

2.2 Two types of solidaristic action

54These two examples not only demonstrate the wide range of solidaristic actions; they also show that two different kinds of cases need to be distinguished.

55In the first example where you and I are passengers on the same delayed flight, solidarity manifests itself as an action by one individual toward another and for their benefit. Because I see you as relevantly similar to me, I am motivated to offer you my cell phone because it is natural for me to “put myself in your shoes” due to our being “in the same boat”. There is no coordination required, no group that needs to act together, and there is no potential benefit to me. This example may be an instance of altruism, but because it is altruism motivated by a relevant similarity between us, it is also a case of solidarity.

56But the second example is importantly different. Solidarity manifests itself among the factory workers as coordinated behavior among the group members. This coordination creates the possibility of an efficient outcome – that is, an outcome that benefits everyone and thereby maximizes the total benefit for everyone in the group.

57Let’s call the first type of solidaristic action “individual” and the second, “coordinated”. Individual solidarity does not create an opportunity for collective action; it is simply a case in which altruism has a particular type of motivation. But on the other hand, coordinated solidarity does create the opportunity for collective action. These cases have a salient action, determined and made apparent to everyone because of their similarity; when the individuals coordinate on that salient action, their collective action is one of solidarity.

3. The gap in the collective action literature

58I won’t have much to say about individual solidaristic action. But coordinated solidaristic action has a lot to show us about lacunae in the analytic collective action literature.

59The first rigorous treatment of something close to “collective action” in analytic philosophy literature is probably David Lewis’s Convention (1969). There, Lewis offered a theory of convention according to which equilibria and interactive knowledge play an important role. Specifically, conventions were to be understood as stable behaviors among the members of a population, where the behavior resulted in an equilibrium among everyone’s payoffs and was justified in part by everyone having common knowledge of each other’s potential payoffs.

60Lewis was careful to note that common knowledge was an idealizing assumption, and not necessarily required for people to coordinate their actions. Interestingly, this is a point that appears to have been lost in modern treatments of collective action. Instead of identifying other mechanisms besides interactive knowledge that could play a role in coordinating a group’s actions, as we’ve seen above, collective action is assumed to carry a heavy epistemic burden.

61But examples of coordinated behavior without strong epistemic requirements are easy to come up with. Consider a simple example of two people who are traveling together in a foreign city. They lose track of each other, but they successfully meet up together when each of them decides that going to a local landmark is the best thing to do. The landmark, to use Lewis’s terminology, is “salient”, and even without the two travelers having any plans for this contingency or working through iterations of “I believe that you believe…”, it is easy for them to find each other.

62Such cases would be excluded from analyses like those of Bratman, Gilbert, or Tuomela. They are “defined away” precisely because they do not require interactive knowledge, plans, or any other deliberate coordination mechanism. Thus, the case is not strictly speaking a counterexample to any of those theories.

63However, the deliberate exclusion of those cases is unmotivated. To see this, let’s consider two similar cases:

Organized walkout
A workers’ union at a factory organizes a walkout to protest unsafe working conditions. There is a plan for everyone to leave their workplace at noon on Friday. This plan is communicated to everyone, and each worker has a sufficient degree of interactive knowledge that everyone knows about the plan. At noon on Friday, they walk off their jobs, just according to plan.

64Clearly, this is a collective action and a solidaristic action. It is a collective action because of the plan, the interactive knowledge, and so on. It is solidaristic because the actions of each worker are motivated in part by a willingness to carry a cost for the benefit of others, and that willingness is brought about by a relevant similarity shared among the workers.

65Now consider this minor variation:

Spontaneous walkout
An egregious and dramatic safety violation occurs at a factory which is witnessed by all the workers. Outraged, the workers walk off the job. The walkout is spontaneous, unplanned, and no worker hesitates long enough to acquire any interactive knowledge about anyone else’s beliefs.

66“Spontaneous walkout” can easily be a case of collective solidaristic action, so long as the workers are willing to incur a cost to support other workers. This is a very realistic assumption to make. But even so, it is clearly not a case of shared cooperative activity, joint action, or any other type of collective action encompassed by analytic theories of collective action, whereas “Organized walkout” clearly is such a case. As such, cases resembling “Spontaneous walkout” are not the subject of any theory of collective action I am aware of in this literature.

67As I have mentioned above, it is the right of any theorist to determine the scope of the theory they are constructing. So it is perfectly permissible for writers such as Bratman, Gilbert, and Tuomela (among many others) to exclude cases like “Spontaneous walkout” from the scope of their work. Nonetheless, I find this particular demarcation between these cases to be arbitrary, unmotivated and detrimental to formulating the best theory of collective action. It treats highly similar cases as if they were different in kind from each other, and this gets in the way of constructing an appropriately general theory.

68A superior approach to understanding collective action would be to treat these two cases as being similar. Both are cases of collective action according to any common-sense notion, and both are explained by the same feature – namely, a sense of solidarity among the workers. Along the way, only one case involves careful planning and the acquisition of interactive knowledge; but the two examples, when lined up with each other, show that planning and interactive knowledge are not the crucial features that make these cases of collective action.

4. Correcting the theories of collective action

69The considerations I’ve outlined above suggest that there is something fundamentally wrong with current analytic theories of collective action. Each theory seeks to explain how the members of a group manage to work together to produce a collective action. These theories have hit upon the fact that these cases frequently involve a type of planning and execution that depends on everyone achieving a certain level of interactive knowledge about everyone else’s beliefs.

70As such, clear cases of (what ought to be considered) collective action brought about by solidarity are excluded a priori from these theories merely because they might not possess a stipulated level of interactive knowledge. But this exclusion is unmotivated and arbitrary. Theorists of collective action have fixed upon features of a subset of cases (namely, cases involving interactive knowledge and planning) and taken them to be necessary conditions, thereby arbitrarily narrowing the scope of their accounts.

71What is crucial for understanding collective action is that a group of individuals manages to coordinate their behavior in order to produce a particular, desirable outcome. How they manage this feat of coordination is a separate issue. Across an important class of cases, the coordination is brought about by deliberate planning and enough communication to ensure a level of interactive knowledge among the members of the group. But there is an equally important (and possibly more common) set of cases in which the coordination is brought about by other means. Chief among these is when a perceived commonality among the members of the group and a willingness to incur a cost make a particular action salient. These are cases of coordinated solidaristic action, and it is arbitrary and unproductive to exclude them from theoretical consideration. Other cases surely exist, as in Lewis’s example of the people who successfully find each other in a foreign city. Indeed, there is no doubt a huge range of coordination mechanisms that can give rise to collective action; analytic theories of collective action should give them their due.

Torna su


Bratman, M.E. 1992, Shared Cooperative Activity, “The philosophical review”, 101, 2: 327-341.

Bratman, M.E. 1993, Shared Intention, “Ethics”, 104, 1: 97-113.

Bratman, M.E. 2013, Shared Agency: A Planning Theory of Acting Together, Oxford, Oxford University Press.

Chant, S.R. 2017, Collective Action and Agency, in M. Jankovic, K. Ludwig (eds), The Routledge Handbook of Collective Intentionality, London, Routledge: 13-24.

Davidson, D. 2001, Essays on Actions and Events: Philosophical Essays Volume 1, Oxford, Clarendon.

Ernst, Z., Chant, S.R. 2007, Collective Action as Individual Choice, “Studia Logica”, 86, 3: 415-434.

Gilbert, M. 1990, Walking Together: A Paradigmatic Social Phenomenon, “Midwest studies in philosophy”, 15: 1-14.

Gilbert, M. 1996, Living Together: Rationality, Sociality, and Obligation, Lanham, Rowman & Littlefield.

Gilbert, M. 2006, Rationality in Collective Action, “Philosophy of the social sciences”, 36, 1: 3-17.

Gilbert, M. 2009, Shared Intention and Personal Intentions, “Philosophical studies”, 144, 1: 167-187.

Habermas, J. 2017, Why the Necessary Cooperation Does Not Happen: Introduction to a Conversation Between Emmanuel Macron and Sigmar Gabriel on Europe’s Future, “Social Europe”, 20.

Hussain, W. The Common Good, in E.N. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Spring 2018 Edition,

Klandermans, B. 1997, The Social Psychology of Protest, Oxford, Blackwell.

Lewis, D. 1969, Convention: A Philosophical Study, Cambridge (MA), Harvard University Press.

Ostrom, E. 2000, Collective Action and the Evolution of Social Norms, “Journal of economic perspectives”, 14, 3: 137-158.

Prainsack, B. and Buyx, A. 2011, Solidarity: Reflections on an Emerging Concept in Bioethics, London, Nuffield Council on Bioethics.

Prainsack, B. and Buyx, A. 2017, Solidarity in Biomedicine and Beyond, Cambridge, Cambridge University Press.

Searle, J.R. 1995, The Construction of Social Reality, New York, Simon & Schuster.

Sellars, W. 1963, Philosophy and the Scientific Image of Man, “Science, perception and reality”, 2: 35-78.

Sober, E., Wilson, D.S. 1998, Unto Others, Cambridge (MA), Harvard University Press.

Tuomela, R. 2005, We-Intentions Revisited, “Philosophical Studies”, 125, 3: 327-369.

Tuomela, R. 2006, Joint Intention, We-Mode and I-Mode, “Midwest studies in philosophy”, 30: 35-58.

Tuomela, R. 2011, Cooperation as Joint Action, “Analyse & Kritik”, 33, 1: 65-86.

Tuomela, R. 2013, Social Ontology: Collective Intentionality and Group Agents, New York, Oxford University Press.

Tuomela, R., Miller, K. 2020, We-intentions, in R. Tuomela, R. Hakli, P. Mäkelä (eds), Social Ontology in the Making, Berlin, De Gruyter: 69-88.

Torna su


1 Cf. Corlett 2001: 575.

2 Henceforth, when I refer to “collective action theory”, I will be referring to collective action theory in the sense discussed above.

3 I will follow the game theory and economics literature in referring to this as “interactive knowledge” as opposed to “interactive belief”. But we can replace “knowledge” with “belief” throughout without affecting any of our conclusions.

4 For an overview of how the technical work on interactive knowledge plays out in the context of collective action, see (Ernst and Chant 2007).

5 A notable counterexample is John Searle’s (1995) theory in which no relationships among the people in a group play any role in determining whether they are performing a collective action.

6 We should note that Bratman, Gilbert, Tuomela, and other writers often do not follow the standard usage of the term “common knowledge”. However, they do use the term to refer to some form of interactive knowledge; the difference is that they might require interactive knowledge only to a finite degree.

7 Led by the influential report of (Prainsack and Buyx 2011).

8 See (Prainsack and Buyx 2011) for a careful analysis of the differences among these concepts. In what follows, I draw upon that discussion.

9 Similar considerations show that Habermas (2017) is wrong to say that “Someone who acts in solidarity accepts certain disadvantages in his or her long-term interest in the expectation that the other will behave likewise in similar situations”. This is much closer to a description of reciprocal altruism than to solidarity.

Torna su

Per citare questo articolo

Notizia bibliografica

Sara Rachel Chant, «Solidarity and Theories of Collective Action»Rivista di estetica, 82 | 2023, 106-122.

Notizia bibliografica digitale

Sara Rachel Chant, «Solidarity and Theories of Collective Action»Rivista di estetica [Online], 82 | 2023, online dal 01 février 2024, consultato il 23 juin 2024. URL:; DOI:

Torna su

Diritti d’autore


Solamente il testo è utilizzabile con licenza CC BY-NC-ND 4.0. Salvo diversa indicazione, per tutti agli altri elementi (illustrazioni, allegati importati) la copia non è autorizzata ("Tutti i diritti riservati").

Torna su
Cerca su OpenEdition Search

Sarai reindirizzato su OpenEdition Search