Collective Action and Collective Responsibility: Bad Concepts and Practical Dangers


    In short letters to Nestor Makhno and concerning the Platform (1927-30), Errico Malatesta asserts that the notion of “collective responsibility” is “a moral absurdity in theory” because only individuals can be held morally responsible for their actions. Moreover, he considers use of the concept by anarchists to be “a general irresponsibility in practice.” For it threatens an executive committee forcing individuals to act a certain way, since all associated anarchists will prospectively be held responsible for the actions of some, or forcing individuals to take responsibility for the actions that others have committed regardless of the individual’s lack of involvement or disapproval. Thus, the concept of collective responsibility is at odds with anarchist principles of individual autonomy and free initiative.


    We will substantiate Malatesta’s rejection of the concept of collective responsibility by rejecting the concept of collective action to begin with.  Upon reviewing key theories of action, we will claim that collectives can act only if they can have minds, and that this thesis is difficult if not impossible to defend. If there are no collective actions, we argue, then there can be no collective responsibility since the latter concept depends on the former as its premise: only agents who are able to act intentionally can be considered responsible. However, we recognise with Malatesta that individuals can coordinate their activities in so-called “common actions” and thereby work to achieve common purposes. Having rejected the concept of collective responsibility and affirmed only individual responsibility for individual actions, Malatesta lacks a concept of responsibility that is appropriate to ‘common actions.’ To rectify this, we will outline a conceptual framework consisting of the concepts of ‘cooperative actions,’ ‘common actions,’ and finally ‘shared responsibility.’ Our concept of “shared responsibility” shall finally describe how an individual can be held responsible to varying degrees relative to the contribution of their cooperative actions to the common action in which they are or were associated.


    Why and how would one deny that collectives can act?

    From an action theory perspective, talk of collective actions seems mysterious: most theories of action are based on the existence of mental states.

    In her book Actions (1980), Jennifer Hornsby claims that actions (defined as intentional doings) are mental events that cause bodily movements. Kent Bach rejects the thesis that actions are events in his 1980 article “Actions are not events”. Here, he claims that when someone acts, their mental states cause events in the world, but the action is the causing: actions are defined as “instances of a certain relation, the relation of bringing about (or making happen), whose terms are agents and events.” In other words, one acts when one’s mental episode “of the right sort” causes an event “(in the right way)” (Bach 1980, 114, 120). Maria Alvarez and John Hyman have instead argued for a kind of agent causation. Like Bach, they argue that actions are not the causes of events but disagree with him concerning the role of mental events: an action is a causing by an agent (and not by a mental event) of an event (Alvarez and Hyman 1998, 220). They define agents as entities with causal powers, so their theory allows for inanimate agents (Alvarez and Hyman 1998, 233, 245). But not all agents can act intentionally: for an agent to be able to act intentionally, they need to have reasons (knowledge or supposed knowledge of facts), wants (e.g., desires) or emotions (Alvarez 2009, 300–1). Their view would allow one to recognize collectives as agents of intentional actions if they can be proven to not only have causal powers, but also knowledge, wants, and emotions, i.e., mental states. In what follows, when we talk of ‘action,’ we refer to the notion of action used by Bach (actions are the causing of events by mental states) and Hornsby (actions are mental states that cause events), and Alvarez’s notion of “intentional action” (actions are causings of events by an agent with reasons, wants or emotions). We will come back to Alvarez and Hyman’s notion of action tout court (actions are causings of events by an agent) towards the end.

    Now, we start from the presupposition that when we talk of ‘collective action’ and ‘theories of action,’ the term ‘action’ refers or aims to refer to the same thing. If this is the case, and if the theories of action presented above are in fact the least uncontroversial theories, then it seems that we have a problem: a collective mind problem. Despite their many disagreements, the aforesaid theories of action all basically share an inclusion of mental events or states in their account of action. If this is correct (i.e., if mental states and events are necessary for action), then it seems that there need to be mental states or events for collectives as well in order for them to be able to act.

    The following is a quick proposal of how one might explain collective minds. What does it mean for a collective to have mental states?

    1. There are certain mental states (e.g. beliefs, desires, intentions…) that cannot be attributed to the individuals (members of the collective) themselves, but rather to the collective.
    2. Each individual member of the collective has the same mental states (e.g. a particular belief about the next protest to organise). In this case, the individuals would need to have the same mental states, if the collective is to have that mental state.

    However, these two answers seem problematic. If one were to accept the first, then one would have to explain where these mental states come from. Would they derive from the mental states of some of the individuals? If so, which individuals? The majority? But in that case, is the minority who has other mental states still acting as part of the collective? (If not, then the first option seems to be reducible to the second, where the “collective mental states” are the mental states of each individual of the collective) Or do those mental states arise independently of the individuals? How could that be?

    Pettit (2009) seems to have defended such a view of an independent collective mind. He recognizes one type of group who satisfies the criteria of agency (criteria that do not need to be expanded upon here), a group he calls “straw-vote assembly.” The judgments made by this type of group are rationally consistent: if the group has to vote on a certain issue, and it has already made a judgment on a related issue, then the group has to make sure that the two judgments are consistent; if the latest judgment is problematic (i.e. it is incompatible with the previous), then the group decides whether to abandon the previous judgment or the latest one (Pettit 2009, 80–81). If the group decides to abandon the latest one, this means that the current votes of the individuals are dismissed out of coherence. Thus, the group judgment might be different from the individual judgment, and tends to remain internally consistent, thus remaining independent from the individuals (Pettit 2009, 87). We agree with him that some groups function in this way, but disagree on his observation that this is a group whose judgments are “radically discontinuous” from those of the individuals (Pettit 2009, 87). The ‘discontinuity’ can be explained in terms other than ‘independence’: individuals as members of a group decide differently than when they are deciding for themselves. These individuals might value group principles (i.e. their previous judgments) above their own (i.e. their current judgments) in that context, so even if they do not personally agree with the decision, once they realize a potential conflict they might decide to take a different course of action because it’s better, more appropriate or more coherent for the group. In other words, the individuals’ judgment will always, on some level, be consistent with the group’s judgment. Hence, it is not the group as an independent entity that acts, but rather that the individuals in this group act in their roles as individual members of the group.

    One main problem arises with (2) (the option that every member has the same mental states): if (2) is true, then why do we need to add the notion of a collective at all? With such an account of collective mind, it seems that what we mean to say when we say “The group thinks that p” is “Member A, member B, … member n think that p”. “The group thinks that p” seems therefore to be a shortcut, an attribution that is reducible to the individuals’ minds.

    One way of solving this issue is to suggest that collectives have minds differently than how individuals (human or non-human) have minds. We do not believe that it is very charitable to reject the notion of collective minds outright simply because collectives do not have brains ‘over and above’ the brains of the individuals. Although human beings seem to need brains to have mental states, this might not be the case for collectives. However, we are not sure if, and how, we can, at this stage, make sense of mental states as arising independently from “traditional” minds (i.e. brains).

    Now, if one agrees with these objections, or in general with the difficulty of explaining collective minds, and if one agrees with the above theories of action (at least in their general form), then it seems that the notion of collective action is in a precarious position. If minds are necessary for action, and collectives cannot have minds (at least not any minds that we could recognize as such), then collectives cannot act intentionally. (As I mentioned at the beginning, the actions that Hornsby discusses are always intentional, and Alvarez’s intentional actions also require some mental states). However, collectives could still act (non-intentionally) if one follows Alvarez and Hyman’s theory, although one would first have to prove that collectives have causal powers (i.e. can cause events).

    We can now approach the problem of responsibility. According to at least some theories of responsibility, one is only responsible for one’s own actions, or for not having acted in a way that they should or could have (Talbert 2022). The question now is the following: is the notion of “action” here intentional or non-intentional? Could a collective who acts non-intentionally (or fails to act non-intentionally) be responsible for that action (or omission of action)? It seems that we do consider some individuals who have not acted intentionally to be legally and morally responsible for what they have done non-intentionally (e.g., manslaughter). However, we could explain this in the following way: we consider responsible for their non-intentional actions those individuals who are capable of acting intentionally, or those who through their intentional actions or their omissions led to what they have done non-intentionally. We do not consider volcanoes or plants to be morally, legally or politically responsible, which are the examples of inanimate agents of Alvarez and Hyman’s theory, because they are unable to act otherwise, according to any sort of plan, or to have good or evil intentions.

    So, in order to be responsible, collectives need to be able to act intentionally, i.e., to be agents capable of having mental states. If they are not, then they cannot be considered responsible.


    Common actions and shared responsibility

    We thus affirm Malatesta’s position that it is problematic for anarchists to use the concept of collective responsibility. But we substantiate this from the perspective of action theory, not present in Malatesta’s short letters, that this is because there are no such things as collective minds and their collective actions to which collective responsibility can be attributed. This is why the concept of collective responsibility is a “moral absurdity in theory,” as Malatesta puts it. Consequently, any attribution of collective responsibility would in fact entail wrongly attributing the actions of some members of a group and the consequences of those actions to all members of the group. To do so is an “irresponsibility in practice,” in Malatesta’s words again. For if anarchists use the concept of collective responsibility, then this presents the threat of an executive committee forming that either forces individuals to act a certain way since all associated anarchists will prospectively have to take responsibility for the actions of some. Or the executive committee might force individuals to take responsibility for the actions that others have already committed regardless of the individual’s lack of involvement or disapproval. Thus, Malatesta states, the concept of collective responsibility is at odds with anarchist principles of individual autonomy and free initiative. These principles better rely on a notion of individual responsibility, both in the sense that there is then no issue of relying on a dubious concept of collective action and consequently no danger of domination occuring when the concept is put into practical use.

    However, we do not wish to imply that individual actions occur in isolation of socio-cultural conditions and without any relation to the actions of other individuals, whether those individuals are freely associated in performing those actions or not. Moreover, it is the case that anarchists could only agree to freely associate and organize their individual activities between themselves on the basis of sharing the principles of individual autonomy, free initiative and thus individual responsibility if at the same time they each have some sense of solidarity and common purpose as well as some concept of cooperation or coordination. In his letters, Malatesta accordingly speaks of “the accord and solidarity that must exist among the members of an association.” He also mentions that anarchists must “coordinate [their] forces in a common action.”

    We can now elucidate  Malatesta’s remarks  by defining the concepts of ‘common actions’ and ‘cooperative actions’ in relation to one another. Anarchist individuals are freely associated with one another in collectives such as local activist groups, propaganda publication networks, communes, etc., because they recognise that they each respectively hold the principles of individual autonomy, free initiatives and individual responsibility. Recognising this, they have feelings of accord and solidarity towards one another. However, these principles and the feelings of accord and solidarity are of course continually developed in the course of them associating for whatever common purpose.

    Now, being associated, anarchist individuals reach agreements upon common purposes and the requisite actions to achieve these that they wish to undertake. They do this  either formally through mechanisms such as consensus democracy or more casually and implicitly through conversation or social and bodily cues in the course of actions occurring between associated individuals. The requisite actions at the most general level can be called ‘common actions.’ For example, if they all agree that they want a society without a government, then the common action entailed is ‘overthrowing the government.’ If they all agree on wanting everyone at a squat to be well fed, then the common action entailed may be ‘cooking a communal meal.’ And if two people agree that others should be persuaded that ‘collective responsibility is a bad concept,’ then the common action entailed is ‘co-authoring an essay on collective responsibility.’

    But, crucially, common actions are not collective actions in the sense of being performed by a collective with an independent mind. They are rather the interdependent occurrence of particular cooperative actions that are performed by particular associated individuals. Common actions are ‘interdependent occurrences’ in the following sense. They are no more than the individual actions that comprise it, and those individual actions are defined by their very being coordinated with the actions of other individuals associated in the common action. Co-operative actions are thus understood as individual actions that are coordinated with the actions of other associated individuals with respect to a common purpose. Co-operative actions are enabled by and enable other co-operative actions, as well as being constantly calibrated by each other, and they are all oriented towards a common purpose. People interact.

    A way of stating all this more plainly is that a ‘common action’ is just a summary of all the co-operative actions that are being performed by associated individuals in the course of achieving their common purpose. In his essay “Anarchy” (1891), Malatesta uses the concept of “social action” in a manner that is nearly similar to how we have defined common actions. He states that “social action [...] is not the negation, nor the complement of individual initiative, but it is the sum total of the initiatives, thoughts and actions of all the individuals composing society.” (Malatesta 2014, 132–33). But, to speak more carefully than Malatesta here, we do not conceive of common actions (with ‘social actions’ referring to common actions cooperated in by numbers of individuals on a societal scale, like climate change) as being the mere ‘sum total’ of individual actions, as if the latter were merely added together as independent units to achieve this result. We instead conceive of cooperative actions as mutually enabling and calibrating actions that constitute common actions as interdependent occurrences. In other words, people interact.

    Now, with the notions of common actions and co-operative actions in hand, we can finally introduce a concept that is absent from Malatesta’s letters in which he rejects collective responsibility solely in favor of individual responsibility. Only individuals can act, not collectives, and thus only individuals can be held responsible for what they have respectively done. However, co-operative actions are interdependent with the actions of all individuals in the collective who are associated in a common action, and thus individual responsibility in these cases must still be understood in terms of collectivity somehow.

    What is required is a concept of responsibility that accounts for both these facts. We accordingly introduce a certain concept of shared responsibility that stipulates the following. An individual is responsible for the consequences of their co-operative actions in a common action if they have intentionally and freely associated themselves in that common action. Subsequently, the responsibility for the overall consequences of the common action are shared to varying degrees among associated individuals relative to the contributions made by their cooperative actions in the common action (and, we might add, by reference to how relevant others cooperated with them and in view of mitigating circumstances such as broader socio-economic or personal circumstances). Ergo, everyone who cooperates in a common action shares responsibility for it. But each individual only shares responsibility for the cooperative actions that they individually contributed therein. Accordingly, any member of a collective who did not participate in a given common action shares no responsibility for it - though they may benefit or suffer from it by virtue of their association in the collective in general.



    In our talk, we have built on Malatesta’s writings to support his claims against the legitimacy of collective responsibility as a concept and its use by anarchists to frame their activities. We showed that collective minds are necessary for collective actions, and that there can be no responsibility without actions. By rejecting the collective mind hypothesis, we reject the concept of collective responsibility altogether. Nevertheless, we do not want to propose an individualistic account of human life. Human beings can and do persistently act together. We call these ‘common actions,’ understood as the summary of the co-operative actions of individuals doing things interdependently to achieve a common purpose. It is in this framework that one can talk of ‘shared responsibility,’ which is the responsibility shared in varying degrees between the individuals who participated in the common action according to the relative contributions of their own cooperative actions therein. We hope to have thereby offered an account that includes what theorists of collective actions want to explain (cooperation between individuals) but without the drawbacks (the difficult problem of collective minds), and that furthermore respects the principles of anarchism (individual autonomy and free association).



    Alvarez, Maria. “Acting Intentionally and Acting for a Reason.” Inquiry 52, no. 3 (2009): 293–305.

    Alvarez, Maria, and John Hyman. “Agents and Their Actions.” Philosophy 73, no. 284 (1998): 219–45.

    Bach, Kent. “Actions Are Not Events.” Mind 89, no. 353 (1980): 114–20.

    Hornsby, Jennifer. Actions. London: Routledge & Kegan Paul, 1980.

    Malatesta, Errico. “A Project of Anarchist Organisation.” The Anarchist Library, 20 July 2020,, 1927.

    Malatesta, Errico. “Reply to Makhno: In Reply to About the Platform.” The Anarchist Library, 20 July 2020,, 1929.

    Malatesta, Errico. “On Collective Responsibility.” The Anarchist Library, 20 July 2020,, 1930.

    Malatesta, Errico. “Anarchy.” In The Method of Freedom, edited by Davide Turcato, 109–48. AK Press: Oakland and Edinburgh, 2014.

    Pettit, Philip. “The Reality of Group Agents.” In Philosophy of the Social Sciences: Philosophical Theory and Scientific Practice, edited by Chysostomos Mantzavinos, 67–91. Cambridge: Cambridge University Press, 2009.

    Talbert, Matthew. “Moral Responsibility.” In The Stanford Encyclopedia of Philosophy, edited by Edward N. Zalta and Uri Nodelman, Fall 2022. Stanford: Metaphysics Research Lab, Stanford University, 2022.