Hostname: page-component-848d4c4894-5nwft Total loading time: 0 Render date: 2024-05-27T19:14:39.763Z Has data issue: false hasContentIssue false

Communal sin, atonement, and group non-agential moral responsibility

Published online by Cambridge University Press:  09 August 2023

Harvey Cawdron*
St Mary's College, The School of Divinity, University of St Andrews, UK
Rights & Permissions [Opens in a new window]


In analytic theology, corporate and/or communal accounts of moral responsibility are gaining recognition as a useful resource in numerous debates. One of the areas to which they have been applied is the atonement. It is thought that when Christ is atoning for the human community, one evades concerns about justice because it seems permissible for a member of a group to suffer punishment for the group's actions even if they are not morally responsible for these themselves. To establish the moral responsibility of the human community, one can either adopt group agency or utilize a non-agential form of group moral responsibility. I shall explore the latter option here and shall outline the understanding of communal sin undergirding the model.

Original Article
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (, which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright © The Author(s), 2023. Published by Cambridge University Press


In the analytic theological literature, scholars have utilized corporate and/or communal models of moral responsibility, in which groups as a whole can be deemed morally responsible for their actions, to inform their accounts of key theological topics such as ecclesiology, worship, and the atonement.Footnote 1 I shall focus on the applications of these models to the atonement. Sin is an essential topic here, as one must develop an understanding of how communities, as well as individuals, can be granted and violate their duties to God.

In this article, I shall focus on a particular communal model of the atonement, namely the Communal Substitution Theory (CS) developed by Joshua Thurow. Since this model bestows duties upon the human community, one requires an account of group moral responsibility, and for this one can either argue that the group is a moral agent or that it has non-agential moral responsibility for the relevant action. In this article, I shall explore the second route. I shall not assess the plausibility of group agency and group non-agential moral responsibility themselves, but shall consign my focus to considering their fittingness in models of CS.

I shall start by outlining CS, and shall briefly point out some challenges faced by the group agency option. These concerns are not intended to render this option implausible, but instead to highlight the difficulties one will face when developing an agential account of CS. I shall then outline an alternative, non-agential model of CS, and shall spend some time developing the understanding of communal sin needed to reinforce this account. I shall then consider some potential objections. I am not going to argue for the superiority of this non-agential account over those utilizing group agency, but shall instead point to it as a way of establishing the moral responsibility of the human community if the obstacles facing agential accounts prove too great.

The communal substitution theory of the atonement

Let us begin by outlining Thurow's account of CS. Thurow develops the communal understanding of satisfaction or penal substitution models of the atonement to get around the justice worry, the concern that it is wrong to punish an innocent man, Christ, in place of the guilty.Footnote 2 Here, God has bestowed duties upon the human community, as well as upon individuals. The human community has failed to adhere to its duties, meaning that it is guilty of sin, just as the individuals within it are.

In a satisfaction account, Christ pays the moral debt owed to God by the human community, namely a life of obedience and holiness, and in a penal substitution account we can say that through his death on the cross Christ pays the punishment for humanity's sins. This communal account is said to evade the justice worry because it seems permissible for a member of a community, namely Christ, to suffer punishment or offer satisfaction on behalf of the community, even when the individual punished is innocent themselves.Footnote 3

Thurow argues that those capable of atoning for a group's wrongdoing must be appropriately causally connected to the failure of the group to fulfil the relevant responsibility. These will typically be members of the group, as usually it is the actions of members that result in the failure of the group to fulfil its responsibilities and, even in cases where there has been external involvement, some members should be involved in atonement to acknowledge the harm done by the group.Footnote 4

When it comes to the atonement of the human community, Christ can be seen as an appropriate representative because, although not sinful himself, he is causally connected to the human community's sinfulness. Thurow claims that, as the second person of the Trinity, he created humans, making him causally responsible for the existence of the community. He is also causally responsible for its maintenance due to acts like freeing the Jews from Egypt.Footnote 5 Christ, as the incarnate Word, is both creature and creator.Footnote 6 Thus, Christ is an appropriate representative because he is both a member of the group, in being incarnate, and, as noted by Thurow, causally connected to the sinfulness of the community due to his creation and maintenance of the community.

In fact, only Christ, as God incarnate and thus creator, can bear the appropriate causal connections to the sinfulness of the human community. Although individual humans and groups are partially responsible for humanity's sinfulness as a result of their own sins, they cannot be deemed causally or morally responsible for the sins of someone in a different country with whom they have not been involved, or those of someone who existed hundreds of years ago, and so on. They thus cannot be deemed appropriately causally connected to the sinfulness of the entire human community. This means that only Christ can bring about atonement, establishing the need for incarnation.Footnote 7

In this article, I shall not critically assess Thurow's model. Instead, I shall consider the options one has for developing an account of how duties can be bestowed upon the human community as a whole, a key component of CS. The obvious way of doing this is to advocate group agency and deem the human community a moral agent. Alternatively, one could endorse a model of non-agential group responsibility. Thurow points to both as potential options.Footnote 8

Some scholars have opted for the group agency route.Footnote 9 However, this comes with significant challenges. The literature on group agency is firmly focused on complex, well-integrated groups like businesses and governments, meaning that its insights are not readily transferable to unstructured groups like humanity.Footnote 10 This does not mean that such accounts are impossible, only that their development shall be complex and difficult.Footnote 11

In an approach incorporating group agency, because Christ is acting as a representative of the human community as a whole, the responsibility of the group must be sufficiently connected to that of individual members such that, in atoning for the sins of the group, Christ atones for the sins of the individual members as well. Furthermore, for Christ's sacrifice to be capable of atoning for all humans in CS, we need the human community not just to be granted agential status, but to retain its agential status through time. Christ's atoning work is for all humans throughout time, and for the moral responsibility of the human community to act as the catalyst through which their sins are atoned for, the human community must be capable of bearing moral responsibility when all humans are parts of the community. It must thus remain a group agent across time. Given the size of the human population and the unstructured nature it has, certainly in its current state, this is difficult to account for.

Models of group agency fit into two broad categories: rationalism and sentimentalism. Rationalists typically argue that rational agents are moral agents, and thus groups must be rational if they are to be moral agents.Footnote 12 Sentimentalists, on the other hand, claim that normative competence requires the ability to have moral emotions like guilt and gratitude. This entails that collective agents can only be moral agents if they have moral emotions.Footnote 13

A communal account of the atonement utilizing a rationalist form of group agency has been produced by Jonathan Rutledge. Rutledge points to God's omnirationality, and suggests that God could consider all the reasons for and against something and in the process take into consideration the inputs of individual humans. He then claims that this could form the basis for a rational decision-making procedure coordinated by the Son, granting us a way of deeming humanity a group agent.Footnote 14 He leaves open exactly what is involved in this decision-making procedure.Footnote 15

As one can see from this proposal, although providing an understanding of humanity as a group agent is perhaps possible, due to the fact that humanity does not appear to be a well-integrated group supported by some form of structure capable of generating actions and decisions at the level of the group, such an undertaking is a significant challenge. Any explanation that attempts to establish the agency of humanity is likely to be enormously complex and speculative, as Rutledge's model, although ingenious, is.

Instead of further exploring the group agency option here, I shall propose that if proponents of the view should fail to establish the idea that humanity is a group agent, or dissatisfaction is expressed towards the models they present, one can still defend the bestowal of moral responsibility on the human community in CS.Footnote 16 This can be done by opting for the non-agential route. Building on accounts of group non-agential moral responsibility that have been proposed in the philosophical literature, I shall develop an account of CS that does not require group agency in its construction and shall also outline the account of communal sin undergirding it.

The non-agential model

I shall now outline the non-agential version of CS. An example will help us understand how this works. To argue for group non-agential moral responsibility, Gunnar Björnsson came up with the case of the lake. Here, three people are all individually polluting a lake with the solvent they use when painting their boats, and it is killing the fish in the lake. At least two of them need to refrain from polluting the lake to save the fish.Footnote 17

This example allows us to sketch, in a preliminary manner, how unstructured, non-agential groups can bear responsibility. In this example, it is the group that carries the obligation to save the fish because none of the three individuals can do this on their own. This means that the group is non-distributively responsible for the death of the fish if they fail to stop polluting. This is still the case if the three boat owners do not form a group agent. They do not need to decide to refrain from polluting the lake using some decision-making procedure. They do not even have to consult each other. Each of them can decide individually to stop polluting the lake. Regardless, the obligation to save the fish and the responsibility for not doing so still applies to them as a group.Footnote 18

Developing an account of such obligations in non-agential models is crucial, and scholars working in this area have provided proposals for how one might understand these. Björnsson, for example, has provided an account of essentially shared obligations in which the three people in the boat example together have an obligation to save the fish, but may not have such an obligation individually, depending on the actions of the other members.Footnote 19 The obligation is held jointly, or together, by the members of the group in a way that is not reducible to them as individuals.Footnote 20

The immediate worry facing such proposals is the thought that moral agency is required for one to be subject to moral obligations. If one is committed to this principle, one will probably conclude that no one is obligated to save the fish in the lake example, because none of the agents involved (the individual boat owners) have the capacity to bring this about.Footnote 21 Scholars have attempted to address this problem in various ways, but I shall confine my focus to solutions that serve my purposes here in enabling the bestowal of obligations onto unstructured groups.Footnote 22 When developing his account of shared obligations, Björnsson proposes an understanding of Moral Obligation that can assist here:

P has a moral obligation to φ in C (is morally required to φ in C) if, and only if, were P in C to have the sort of motivational sensitivity that can be reasonably morally required of P, this would ensure, in normal ways, that P φ in C.Footnote 23

Björnsson claims that such an understanding can be applied in cases of unstructured groups. The sensitivities required can be those possessed by the individuals involved. The three boat owners, for instance, may individually have the sensitivities needed to move them to save the fish. It is the group, however, that has the capacity to fulfil the obligation. The sensitivities of the members just help to explain why it does so. This only requires the moral agency of individual members of the group, not that of the group itself, and, as shown in the lake example, also does not require that the group act in a coordinated way. This explains how obligations can be held at the level of the group when it is not an agent.Footnote 24

An account of how non-agential groups can have obligations is important for our purposes here. A crucial component of CS is the idea that the human community possessed, and failed to adhere to, certain duties, and such accounts of non-agential moral responsibility provide us with a way of establishing duties, or obligations, at the group level without requiring group agency.

Using this, we can provide an outline of what a non-agential form of CS would look like. We can claim that there are obligations that God has bestowed upon the members of the human community that are shared, since individual humans are incapable of fulfilling them. They have failed to fulfil these obligations, meaning that they as a group are non-distributively responsible for this failure.Footnote 25 The specific obligations that the human community has failed to fulfil shall be outlined in the next section.

For this to work as a model of CS, we need a way of understanding how an innocent person can be punished for humanity's collective failing.Footnote 26 Let us return to the lake example. Imagine that the three people polluting the lake do not own the lake themselves. The lake is actually private land, and the owner allows each person to keep their boat on the lake for a certain price. One can also imagine that the owner, although they also own a boat that they use on the lake, refrains from using the solvent and thus does not contribute to the pollution in the lake.Footnote 27

Imagine, now, that an inspection was done on the lake by an environmental agency, and the inspectors discovered that the pollution in the lake had killed a large number of the fish. One can imagine that in a scenario such as this, it would be the owner of the lake that would be punished by the agency for the death of the fish, even though they did not participate in the pollution themselves. They might say that the lake owner should have been more vigilant, or should have had strictly enforced rules about what solvents those on the lake could use on their boats. The owner would be an appropriate representative for those polluting the lake in the eyes of the environmental agency.Footnote 28

Now one can see how the atonement fits into this model. One can claim that Christ atones for humanity's collective failing. Just as in the boat example the death of the fish was caused by the individual acts of pollution performed by the boat owners, we can assume that humanity's collective failing is constituted by the sins of individuals, meaning that Christ atones for the sins of individuals in atoning for humanity's collective failing. Much like the lake owner in the above example, one can also say that it is morally permissible for Christ to act as representative here despite his state of innocence.

In the lake example, both the status of the fourth member as owner of the lake and their use of a boat on the lake are crucial for illustrating how this analogy helps us to understand the atonement. As noted earlier, key to CS is the idea that Christ is an appropriate representative for the human community, despite his innocence. The fourth member's use of a boat without polluting the lake reflects Christ's status as a non-culpable member of the human community, and their ownership of the lake in turn captures Christ's causal connection to the sinfulness of, and thus status as an appropriate representative for, the human community in atonement.

This non-agential variant of CS thus seems able to provide us with a defence of the idea that the human community can be morally responsible for a collective failing without the difficulties that come with arguing that it is a moral agent. One is still able to claim that the human community has failed as a group in its obligations to God, and that Christ's sacrifice on behalf of the community is appropriate. One may note, however, that we have pointed to some collective failing of humanity without properly specifying what that collective failing is. It is to this task that we now turn.

Communal sin

Now, we must achieve clarity on what obligation is shared, and not adhered to, by members of the human community. In a non-agential understanding, we cannot refer to the moral agency of a group to point to its capacity for collective failings. Therefore, we cannot simply thrust duties, like the duty to follow God's laws, upon the human community. We must find a collective obligation unable to be fulfilled by the members individually that the group itself failed to abide by. Then, we must consider how the group can be held responsible despite the lack of a formal structure linking its members.Footnote 29 This will provide us with an account of the communal sin of the human community.

The collective failings of humanity

Essential to this model is an account of the collective failing of the human community. In this subsection, I shall provide another example of a collective failing that may be applied to humanity as a whole that will help to form the basis of this account of sin. The account shall rest on the idea that just as one can blame groups for their actions, one can also hold them responsible for states. To return to the lake example, the boat owners can be blamed for the polluted state of the lake as well as the acts they perform to cause that pollution.

If one were to scale up a case such as this by referring to the oceans instead of a lake, one can find an example of a collective failing that applies to humanity as a whole. One can argue that humanity, or a large part of it at least, is responsible for the state of the oceans. The dumping of various materials into it, fishing that is not sufficiently restricted, and other such failings are severely damaging our oceans and their inhabitants. Although previous generations can perhaps be said to be absolved of moral responsibility due to unawareness of just how big an impact their actions would have, we now have scientific data that confirms this and the communication channels required to spread awareness.

Granted, there are groups like governments and corporations that are responsible for much of this pollution, but individual humans also contribute. However, when considered at the level of humanity as a whole, this is a result of the negligence of individual groups and persons in the sense that all humans have not conspired together as a group in order to bring about this situation. Moreover, these individual groups and persons cannot address this issue alone. Humanity must act together to address the issue, showing that we can plausibly argue that humanity as a whole can be non-distributively responsible for a negative state.

This provides us with a useful example of how there can be collective failings at the level of humanity as a whole. It enables us to conceptualize how the human community could have failed in its collective duties to God, and will prove useful in developing the understanding of sin required for this model to work.

Sin as uncleanness

Now that we have an idea of what collective failings might look like at the level of humanity as a whole, we can look for an understanding of sin to help us work out what this collective failing might look like for the purposes of the atonement. One will notice that the ocean example deemed humanity responsible for a state, namely the polluted state of the oceans. Sin itself has been construed as an ontological state by some scholars in the Christian tradition.Footnote 30 I shall argue that humanity can be deemed responsible for its sinful state on a non-agential understanding.

Before making this argument, however, it is worth further clarifying the account of the nature of sin to be used. The understanding I shall use here is that of sin as uncleanness.Footnote 31 This account of sin has been developed in varying ways by several scholars, including Rudolf Otto,Footnote 32 Marilyn McCord Adams, and Mary Douglas. It has also been used for different purposes. For example, some scholars have used it as a way of distancing the idea of sin from moral responsibility.Footnote 33 I shall not concern myself with this topic here, but shall instead outline how this understanding of sin can assist this non-agential model of CS.

Douglas's ideas shall prove particularly useful here. Douglas discusses the notion of uncleanness and suggests that it refers to things that are disordered, and that these can be both powerful and dangerous.Footnote 34 Adams has adapted this idea to formulate an understanding of sin, claiming that God's laws are put in place to maintain order, and when they are not followed, creatures become out of order, and thus unclean.Footnote 35

Douglas also points to the idea that the holy entails wholeness and completeness, and that this is exemplified by the perfection required in certain ritual elements in Leviticus. For example, when discussing rules for priests, it is said that:

The LORD said to Moses, ‘Speak to the priests, the sons of Aaron, and say to them: “A priest must not make himself ceremonially unclean for any of his people who die, except for a close relative, such as his mother or father, his son or daughter, his brother, or an unmarried sister who is dependent on him since she has no husband – for her he may make himself unclean. He must not make himself unclean for people related to him by marriage, and so defile himself.”’Footnote 36

This is also something extended to group contexts. For example, it was thought that the army of the Israelites could not win without blessing, and that those in the camp had to be especially holy to maintain this blessing. In Deuteronomy, it is instructed that ‘When you are encamped against your enemies, keep away from everything impure.’Footnote 37 To prevent defilement in the camp, anyone with bodily discharges was disqualified from entering it. Any functions or issues involving bodily discharges had to take place outside the camp.Footnote 38 This understanding of communal uncleanness will prove important in the model developed here.

The pollution of the human community

In the previous two subsections, we developed an example of a communal failing of humanity as a whole, namely the pollution of the oceans, and an account of sin as uncleanness. In this section, we can combine these two ideas to formulate the understanding of sin undergirding this non-agential model of CS.

The Deuteronomy passage in particular can assist us here. In it, we have the instruction to remain free of defilement placed upon an army, which is a group. Although an army might be something one would deem a group agent, this could also be understood in a non-agential way.Footnote 39 One can claim that the obligation to keep the camp pure is a collective obligation. For it to be fulfilled, each member must adhere to the rules imposed to prevent defilement. Keeping the camp pure is not an obligation that individual members on their own can ensure the fulfilment of, as it is possible for them to follow the rules while others do not. One can thus understand this as a collective obligation that applies to the group.

From these ideas, one can pinpoint the collective failing needed for an understanding of CS. We can claim that, just as in the case of the army, God has imposed upon humans the obligation to maintain order within humanity itself to prevent it from being made unclean by sins. However, individual humans have sinned and continue to sin, meaning that we have brought humanity into a state of uncleanness. Just as humans pollute the ocean by dumping rubbish in it, they pollute the human community by sinning.

One can deem the human community responsible for its unclean, sinful state, a state that all humans have a shared obligation to avoid.Footnote 40 This state is constituted by the individual sins of humans, and it is one for which the human community as a whole is non-distributively responsible. No individual human can ensure that the obligation to avoid this state is upheld.Footnote 41 To stop this, each person must refrain from sinning. Clearly, this is not an obligation that any one individual is capable of fulfilling, meaning that humanity as a whole is non-distributively responsible for its polluted state.

This grants us with the account of sin and collective failing needed for a non-agential understanding of CS. We can say that Christ atones for the polluted state of the human community, something for which all humans are non-distributively responsible.Footnote 42 In doing so, Christ also atones for the sins of individuals as they are ultimately what caused and preserve humanity's unclean state. This completes our outline of the non-agential version of CS.

Objections and replies

Now that we have outlined the non-agential account of CS, we can consider some objections. I shall outline and respond to numerous objections, some concerning tensions that may be pointed to within the model itself, others concerning the scope of Christ's atoning work here, and some that query the attractiveness of this model in comparison to rivals. In cases where I do not defeat the objection, I shall at least point to promising avenues for resolving the issue.

Original sin

One might object to the account of communal sin undergirding this model by pointing to the doctrine of original sin. This doctrine entails that as a result of the first sin, whether performed by Adam and Eve or a primal human community, all later humans are corrupted in such a way that they are unable to will the good, and sin themselves as a result of this. One could point to this corruption and claim that whoever committed the first sin is responsible for the sinful state of the human community. This would entail that this state is not a collective failing but is ultimately the responsibility of a member, or part, of this community.

If one were to adhere to this line of reasoning, this would render our non-agential account of CS implausible. It would entail that the sinful, unclean state of the human community is not something for which the community is non-distributively responsible, meaning that Christ cannot atone for the sins of all humans collectively by atoning for this state. This objection thus causes problems for our model of CS.

However, this problem can be avoided. Although the corruption of later humans is something that prompts them to sin, they still contribute towards the unclean state of the human community through the actual sins that they perform themselves. Therefore, we can claim that while the first sin caused the corruption of humanity, its continuing unclean state is upheld by the sins of the other members of the human community. This entails that it is not just the individual, or individuals, who performed the first sin that are responsible for the human community's polluted state. It is the human community itself, because the other members uphold its polluted state by committing actual sins.Footnote 43

I have leaned heavily on the work of Björnsson in this article, and the implications of original sin may be an element of this model he too has quarrel with. As pointed out earlier, when developing his account of shared obligations, he proposes an understanding of Moral Obligation that rests on the idea that the group has the capacity to, and thus can, fulfil the obligation.Footnote 44 However, to establish the need for atonement it is often claimed that the fall of individual humans into a state of sin through the performance of actual sins is an inevitable result of original sin. If it is inevitable that humans perform actual sins, and thus pollute the human community, can the human community really be said to have the capacity to avoid becoming unclean? If not, the shared obligation that is crucial for this model does not apply to the human community.

When considering this concern, it is important to note that this is a general issue that we must face when explaining moral responsibility in light of original sin. The claim that, as a result of one's corrupted state, it is inevitable that one will fall into a state of sin by performing actual sins itself faces difficulties with moral responsibility.Footnote 45 Moreover, scholars working in this area often adopt theories of moral responsibility that assume people are capable of preventing states of affairs for which they are morally responsible,Footnote 46 which is in line with Björnsson's claim that one's ability to perform an act is required for one to be morally obligated to do so. The challenge in the literature on original sin is reconciling these two notions.

Since this is a general issue with the doctrine of original sin, however, it will also impact accounts incorporating group agency as well as those that place no weight on communities at all. I thus take it that this issue does not reduce the credibility of this model in comparison to these alternatives.

Moreover, solutions to this issue that have been produced elsewhere may be applied to this model. W. Paul Franks has recently argued that one can establish the need for atonement while denying that it is inevitable that humans will perform actual sins. Franks uses an example to illustrate this, claiming that a victim of racist abuse from a particular family would be justified in feeling repulsed at a toddler who is a member of the racist family even if this toddler did not participate in the abuse. Franks applies this example to humanity and suggests that even those who have performed no moral wrongdoing would still be repulsive in the eyes of God due to their membership in humanity, and are in need of atonement as a result of this.Footnote 47

We can adopt this strategy in this model. We can claim that the moral responsibility of the human community is all that is required to establish the need for atonement for all its members. Those who have not contributed to the pollution of the human community are still unclean and in need of atonement in the eyes of God due to their membership in this community. Here, we only need enough humans to perform actual sins to maintain the unclean state of the human community, and, if we assume that actual sins are very likely due to the influence of this fallen world, such a claim is plausible.Footnote 48 Furthermore, since this solution does not require that actual sins are an inevitable result of original sin, it is consistent with the idea that the human community had the capacity to fulfil its obligation to avoid uncleanness.

I do not intend to imply that this is an entirely satisfactory way of addressing the tension between shared obligations and original sin. Indeed, it is typically thought that the inevitability of actual sins is a crucial component of original sin, and those in agreement with this will be dissatisfied with this solution.Footnote 49

This discussion, however, is intended to be preliminary in nature, and I am open as to whether the model is defended using the solution proposed or through the production of another solution.Footnote 50 What we have illustrated is that, first, the clash between the inevitability of actual sins and moral responsibility is a general issue faced within scholarship of original sin, and second, that solutions proposed in this area of scholarship may be applied to this communal model. There are thus prospects for solving this problem.

Morally innocent humans

Another query one might raise is whether morally innocent humans, such as babies or the severely mentally disabled, would be sufficient for atonement on this view. After all, they have not contributed to the pollution of the human community through the performance of immoral acts. Could their deaths be sufficient for atonement, instead of Christ's, and if so, would the atonement not be redundant due to the number of such humans to have sadly passed away throughout history?Footnote 51

Such a conclusion does not follow from this model. Christ, through his creation and maintenance of the human community as God incarnate, is causally connected to its sinfulness, making him an appropriate representative for the human community in atonement. Other morally innocent humans are not causally connected to the sinfulness of the human community in this way, and thus would not bear the causal connections required for them to be considered appropriate representatives. Therefore, they would not be able to bring about the atonement themselves.

However, one might point to another issue that morally innocent humans raise for this model. If they are not capable of performing morally bad acts, one may question whether they have contributed to the pollution of the human community, and thus whether they are included within the scope of Christ's atoning work. In this model, individual sins are atoned for as a result of Christ's atoning for the pollution of the human community because their individual acts contribute to this pollution. If morally innocent humans do not contribute to this pollution, it might be thought that they are not included within the scope of Christ's atoning work, which is problematic.

This objection only carries weight if we assume that the only way individuals are atoned for is through the atonement of their actual sins, and we need not deem this to be the case. As suggested in the discussion of original sin, we can claim that those who have not performed actual sins are still unclean and in need of atonement as a result of their membership in the human community. Since the pollution of the human community is ultimately the source of the uncleanness of these members, Christ, in atoning for this pollution, also atones for their state of uncleanness, providing us with an understanding of how they can be included within the scope of Christ's atoning work.

The forward-looking element of sin

One might also question whether this account fully addresses all of the issues the atonement is meant to deal with. It seemingly deals with the backward-looking element of sin, namely the sins performed by humans and their guilt for these, in addressing the pollution of the human community caused by such sins. However, it is not clear that it deals with the forward-looking element, namely human proneness to sin.Footnote 52

A response to such issues has recently been proposed by Oliver Crisp. Crisp claims that one can consider the atonement efficacious for all of one's sins across time. It cleanses one of one's sins prior to justification, and does the same for all of one's sins after that as well. We will not be without sin until we reach the eschaton, so the atonement deals with all of one's sins prior to that period.Footnote 53

There is no reason why we cannot adopt such a solution here. We can claim that since all sins, regardless of when they are committed, contribute to the uncleanness of the human community, the atonement is capable of addressing them all, including those committed by humans that are already justified.

The issue of complexity

Advocates of group agency may also have reservations about this account. Just as the complexities that came with providing an understanding of the human community as a group agent were pointed to earlier in the article, they may point to elements of this non-agential understanding of CS that also seem complex.

When arguing that there are duties or obligations bestowed upon the human community that have been unfulfilled, an essential part of any understanding of CS, advocates of models incorporating group agency can simply say that the human community is a moral agent capable of bearing such duties itself. In our non-agential understanding, we were unable to claim this, and had to develop a specific understanding of sin in order to show that there were failings for which members of the human community are non-distributively responsible. Thus, endorsing group agency seems to be the simpler route as far as this element of CS is concerned.

I acknowledge this, but in response would point to the areas where a non-agential model of CS has an advantage. Establishing the idea that members of the human community could be non-distributively responsible for certain actions was a rather simple argument to make, as the understanding of shared obligations used here does not require groups that are well-integrated and structured. The extension of the reasoning found in the literature on group non-agential moral responsibility to the human community did not come with any significant obstacles.

It is precisely this element with which those endorsing group agency struggle. Humanity is not a well-structured, coordinated group, which means that the insights raised in the literature on group agency are not readily transferable to it. Therefore, as seen in Rutledge's model, imposing such a structure on humanity is going to take a great deal of complex and speculative reasoning. The non-agential model simply comes with a different cost, that of formulating a model of sin enabling humanity to be non-distributively responsible for a collective failing of some sort. Since we have been able to address this cost in this article, I would argue that the non-agential understanding of CS remains a viable alternative.


This concludes our assessment of the non-agential form of CS. It was able to evade the first worry caused by original sin as humans contribute to the polluted state of the human community by performing actual sins. Prospects for dealing with the second worry, namely the tension between the inevitability of actual sins and shared obligations, were also outlined. Morally innocent humans are unable to bring about the atonement because they are not appropriate representatives for the human community, and their membership in the human community enables such humans to be included within the scope of Christ's atoning work. We can account for the forward-looking element of sin in deeming the atonement efficacious for all of one's sins across time. Finally, although the account of sin and collective failing offered in this non-agential understanding is more complex than that offered in accounts incorporating group agency, the non-agential model remains a viable alternative.


In this article, it has been shown that, using a non-agential form of CS, we can defend the idea that the human community is morally responsible without commitment to the idea that it is an agent. In agential forms, one is committed to the controversial and counterintuitive premise that the human community is a moral agent, and even though one might be able to argue for this, the model developed will inevitably be complex and speculative. Non-agential models, on the other hand, do not have these problems and, when supplemented with a particular understanding of sin, are plausible. They thus provide viable alternatives.

I started this article by outlining Thurow's account of CS, and highlighted some challenges for the group agency route. I then outlined a non-agential model of CS and the account of sin undergirding it using the understanding of sin as uncleanness. I then responded to some potential objections. Although this is by no means a conclusive argument for a non-agential understanding of CS, I hope that this article has shown that it can act as a way forward should the difficulties with providing an account including group agency prove insurmountable.


I am indebted to the participants of the 2023 Helsinki Analytic Theology Conference for their feedback on this article. I would also like to thank Jonathan Rutledge, Joshua Cockayne, Joanna Leidenhag, Oliver Crisp, Aku Visala, Olli-Pekka Vainio, Dani Jansen, David Worsley, Parker Haratine, and two anonymous referees for this journal for their comments on earlier drafts of this article.

Competing interest



1. See Cockayne (Reference Cockayne2019) and Cockayne (Reference Cockayne2022) for its relevance to ecclesiology, Cockayne (Reference Cockayne2018) and Cockayne (Reference Cockayne2021) for its relevance to worship and liturgy, and Leidenhag (Reference Leidenhag2020) for a model of charismatic gifts utilizing group agency.

2. See Thurow (Reference Thurow2015), 53–54. The justice worry clearly impacts the penal substitution variant, and some scholars have raised it against satisfaction models (see, for instance, Stump (Reference Stump2018), 24–25). However, Oliver Crisp has recently objected to such suggestions, claiming that they do not impact satisfaction accounts because the atonement is not achieved through the punishment of Christ (Crisp (Reference Crisp2022), 112–113). Even if this worry does not apply to satisfaction accounts, the fact that it impacts penal substitution accounts still makes this discussion relevant.

3. See Thurow (Reference Thurow2015), 52–62, for the outline of CS.

4. See Thurow (Reference Thurow2015), 51, for this argument. I shall not contest Thurow's reasoning, but note the importance of this point in establishing a satisfactory understanding of CS. If members of the group are not required in the atonement process, when applied to the case of Christ it raises questions about whether the incarnation was necessary for atonement. Moreover, if one can question the status of Christ as an appropriate representative for the group, how this model addresses the justice worry becomes unclear, as it no longer seems morally permissible for Christ to stand in for the guilty members in atonement. I thank an anonymous reviewer for raising these objections.

5. Thurow (Reference Thurow2015), 55. One might also suggest that Christ bears constitutive responsibility, which you acquire from being a part of something like the human community. This is even the case if you are a superfluous part of that thing. See Goodin (Reference Goodin2018), for more on this.

6. McFarland (Reference McFarland2014), 86.

7. Thurow (Reference Thurow2015), 57–58.

8. See Thurow (Reference Thurow2015), 62–63.

9. For a useful outline of the group agency literature, see Tollefsen (Reference Tollefsen2015). To be clear, some see talk of group agents as metaphorical. This view is eliminativism about group agents, in which only individual agents exist and they do not form a new agent when they form a group. Eliminativism is argued for in Quinton (Reference Quinton1975), 17. Moen (Reference Moen2019) challenges the idea that ascribing agency to groups helps us predict and explain their behaviour, and this can be used to argue for eliminativism. There is a difference between redundant and non-redundant group-agent realism. On the redundant realist view, group agency talk is not misconceived, but is readily reducible to individual-level talk. On the other hand, non-redundant realism holds that group agency talk is not misconceived and is not readily reducible to individual-level talk. This distinction is outlined in List and Pettit (Reference List and Pettit2011), 6. It is the non-redundant view that is of interest here.

10. Crisp seems to share this sentiment, in claiming that fallen humanity (his equivalent of the human community) is not an agent and does not act like one. See Crisp (Reference Crisp2022), 193. This is stronger than the claim I am making here, but it helps to illustrate that the idea that the human community is a group agent is at least counterintuitive. Despite this, Crisp makes good use of the social ontology literature when developing his model of the atonement in part 3 of Crisp (Reference Crisp2022).

11. Thurow also proposes narrowing the scope of the atonement to Israel, a collective entity and perhaps a group agent, as an alternative way out of this problem (Thurow (Reference Thurow2015), 63). Maintaining a view in which Jesus atones for the entire human community enables one to be more flexible with one's view of eschatology, however, as it accommodates views like universalism, in which the entire human community is saved. There are thus advantages to accommodating the entire human community.

12. Hindriks (Reference Hindriks2018), 4. For some influential rationalist accounts of group agency, see List and Pettit (Reference List and Pettit2011), French (Reference French1979), Collins (Reference Collins2019), Rovane (Reference Rovane2014), Rovane (Reference Rovane2004), and Rovane (Reference Rovane1998). Questions have been raised about whether rationalism is a satisfactory understanding of moral agency. For example, there is the problem of amoral agency. An agent can be prudentially rational without being moral. Psychopaths, for instance, are often said to be rational and amoral (Hindriks (Reference Hindriks2018), 3). The rationalist assumption that rational agents are moral agents is thus problematic.

13. Hindriks (Reference Hindriks2018), 12–13. For some influential sentimentalist accounts of group agency, see Collins (Reference Collins2018), Gilbert (Reference Gilbert2002), Gilbert (Reference Gilbert2006), and Björnsson and Hess (Reference Björnsson and Hess2017). Much as with rationalism, there are seeming issues with sentimentalism as an understanding of moral agency. It has been objected that there are people with severe emotional disorders who would be exempt from moral responsibility in such accounts, including some of the evillest people to have existed (Hitler and Ted Bundy, for example). See Tollefsen (Reference Tollefsen2015), 132–133.

14. See Rutledge (Reference Rutledge2022), ch. 6.

15. Rutledge confirmed this in private correspondence.

16. I am indebted to an anonymous reviewer for their help with framing here.

17. This example is developed and used in Björnsson (Reference Björnsson2014). I note that there are potential disanalogies between this case and Christ's relationship to humanity's sinfulness. However, this case is merely intended to help illustrate how the model works, and I am not relying directly on it. I thank an anonymous reviewer for prompting me to clarify this.

18. Here, I am drawing on Björnsson (Reference Björnsson2014) and Björnsson (Reference Björnsson, Bazargan-Forward and Tollefsen2020).

19. Björnsson (Reference Björnsson2014), 104. Such an account is further developed, along with an account of collective blameworthiness that applies in cases where collectives are not agents, in Björnsson (Reference Björnsson, Bazargan-Forward and Tollefsen2020).

20. An alternative route, taken in Wringe (Reference Wringe2010), is to deny that only moral agents can have obligations by claiming that one can have moral obligations if one has the capacity for the sorts of properties associated with moral agency. This is critiqued in Björnsson (Reference Björnsson, Bazargan-Forward and Tollefsen2020), 131.

21. Such a worry is also raised in Björnsson (Reference Björnsson2014), 111, and Björnsson (Reference Björnsson, Bazargan-Forward and Tollefsen2020), 130–131. Despite the solution offered, I note that proponents of group agency may still insist that no one is responsible here, which would be a problem for this account. I am grateful to an anonymous reviewer for pushing me to address this matter.

22. There are alternative models that are unsuitable for our purposes. For instance, Anne Schwenkenbecher has outlined an understanding of joint duties, which are ‘duties which the individual members of the unstructured group hold together’ (Schwenkenbecher (Reference Schwenkenbecher2013), 315). However, Schwenkenbecher's account rests on the idea that such groups would be capable of performing joint actions, which require them to form, and contribute towards, a joint goal. This would be difficult to establish in the case of humanity. See the remainder of Schwenkenbecher (Reference Schwenkenbecher2013) for an outline of this model. James Dempsey has also produced an argument for the idea that groups can have non-agential moral responsibility that is unsuitable here. Such groups must have a system that takes set types of input and generates certain outputs from these using a rule-based structure that can be understood mechanistically. The system must also be complex enough to give rise to group-level responsibility. See Dempsey (Reference Dempsey2013). The human community lacks a system of this type, making this understanding of group non-agential responsibility unsuitable for an understanding of CS.

23. Björnsson (Reference Björnsson2014), 116.

24. Björnsson (Reference Björnsson2014), 117–118.

25. The term ‘non-distributively responsible’ is taken from Arruda (Reference Arruda2017), 251, and it is responsibility that attaches to the group as such and not to any individual member.

26. I note that the term ‘collective’ has been used in association with groups that have ‘a rationally operated group-level decision-making procedure that can attend to moral considerations’, namely, groups that are moral agents in the rationalist sense (see Collins (Reference Collins2019), 11). When talking about collective failings and obligations, I am using the term in a looser sense that does not imply the moral agency of the group being discussed. My thanks to an anonymous reviewer for pushing me to clarify this.

27. I am indebted to David Worsley and an anonymous reviewer for helping me to refine my use of this example.

28. To be clear, one need not see Christ as negligent in the way that the lake owner is here. This element of the example is intended to emphasize the status of the lake owner as an appropriate representative for the polluters, a status obtained by Christ in the case of the human community in virtue of his creation and maintenance of humanity.

29. See Everhart (Reference Everhart2023) for a discussion of communal sin and atonement in relation to the issue of systemic sin.

30. See Timpe (Reference Timpe2021), sec. 4. For more on the nature of sin, see McCall (Reference McCall2019), ch. 5.

31. This understanding of sin is outlined in Adams (Reference Adams1991). See Couenhoven (Reference Couenhoven2009) for further discussion.

32. For Otto, this uncleanness or profaneness is experienced in contrast to the holiness and illustriousness of God. See Otto (Reference Otto1923), 52–61.

33. For instance, Mitchell (Reference Mitchell1984) deems sin a breach in man's relationship with God, and acknowledges the importance of morality while accommodating for the areas that are not accounted for in that category. A more radical separation between sin and moral wrongdoing is argued for in Ashfeld (Reference Ashfeld2021).

34. See Douglas (Reference Douglas1966), particularly chapter 3 for the theological discussion.

35. Adams (Reference Adams1991), 10–11.

36. Leviticus 21:1–4 (New International Version).

37. Deuteronomy 23:9 (New International Version).

38. Douglas (Reference Douglas1966), 52–54. These laws are outlined in Deuteronomy 23:9–14 (New International Version).

39. As pointed out by Caroline Arruda, there may be instances in which a group does not act as an agent and is non-distributively responsible for the action, but can be considered an agent across time. See Arruda (Reference Arruda2017).

40. Obligations that apply at the level of humanity are argued for in Wringe (Reference Wringe2010) and Wringe (Reference Wringe2020).

41. For an argument against the idea that non-agential groups can be responsible for an outcome for which none of the members themselves bear a correlative responsibility, see Hindriks (Reference Hindriks2019).

42. It might be thought that the collective is blameworthy because of their failure to act in the manner of a group agent, as suggested in Held (Reference Held1970). I acknowledge the possibility of this, but do not think we have to be so rigid when thinking about the pollution of the human community. I am open as to whether this is addressed in a coordinated manner or by humans individually addressing their own acts of sin. Both are consistent with this model, as placing obligations on group members to act in a coordinated way is still a case of group non-agential responsibility when that group does not have the structure required for group agency.

43. This model is also compatible with understandings of sin that deny the need for a historical fall, such as the biocultural account found in De Cruz and De Smedt (Reference De Cruz and De Smedt2023).

44. See Björnsson (Reference Björnsson2014), 114–116.

45. See the discussions in Franks (Reference Franks2012) and Vainio (Reference Vainio, Arcadi and Turner2021).

46. See, for instance, the principle of possible prevention, or MR, outlined in Rea (Reference Rea, van Inwagen and Zimmerman2007).

47. See Franks (Reference Franks2012), 368–370.

48. Franks (Reference Franks2012), 370. Swinburne similarly claims that actual sins are very likely, but not inevitable. See Swinburne (Reference Swinburne1989), 137–140.

49. A discussion of this feature of the doctrine and the solutions proposed to its apparent clash with moral responsibility can be found in Timpe (Reference Timpe2023).

50. One alternative could be a defence of the doctrine of original guilt, the idea that all humans (except Christ) are guilty from birth of the first sin and the resulting corruption, as here all humans would be guilty for the corruption that results in the pollution of the human community. I note that reconciling this doctrine with moral responsibility is also an extremely difficult task, however (see the discussions in Rea (Reference Rea, van Inwagen and Zimmerman2007) and Hudson (Reference Hudson2014), ch. 4).

51. I thank an anonymous reviewer for raising this objection.

52. This discussion is based on a critique raised in Stump (Reference Stump2018), 25–26. I am grateful to Jonathan Rutledge for pointing me to this issue.

53. Crisp (Reference Crisp2022), 113–114.


Adams, MM (1991) Sin as uncleanness. Philosophical Perspectives 5, 127.CrossRefGoogle Scholar
Arruda, CT (2017) How I learned to worry about the Spaghetti Western: collective responsibility and collective agency. Analysis 77, 249259.CrossRefGoogle Scholar
Ashfeld, M (2021) Five problems for the moral consensus about sins. International Journal for Philosophy of Religion 90, 157189.CrossRefGoogle Scholar
Björnsson, G (2014) Essentially shared obligations. Midwest Studies in Philosophy 38, 103120.CrossRefGoogle Scholar
Björnsson, G (2020) Collective responsibility and collective obligations without collective moral agents. In Bazargan-Forward, S and Tollefsen, D (eds), The Routledge Handbook of Collective Responsibility. New York: Routledge, pp. 127141.CrossRefGoogle Scholar
Björnsson, G and Hess, K (2017) Corporate crocodile tears? On the reactive attitudes of corporations. Philosophy and Phenomenological Research 94, 273298.CrossRefGoogle Scholar
Cockayne, J (2018) Inclusive worship and group liturgical action. Res Philosophica 95, 449476.CrossRefGoogle Scholar
Cockayne, J (2019) Analytic ecclesiology: the social ontology of the Church. Journal of Analytic Theology 7, 100123.CrossRefGoogle Scholar
Cockayne, J (2021) We believe: group belief and the liturgical use of creeds. European Journal for Philosophy of Religion 13.CrossRefGoogle Scholar
Cockayne, J (2022) Explorations in Analytic Ecclesiology: That They May be One. Oxford: Oxford University Press.CrossRefGoogle Scholar
Collins, S (2018) ‘The government should be ashamed’: on the possibility of organisations’ emotional duties. Political Studies 66, 813829.CrossRefGoogle Scholar
Collins, S (2019) Group Duties: Their Existence and Their Implications for Individuals. Oxford: Oxford University Press.CrossRefGoogle Scholar
Couenhoven, J (2009) What sin is: a differential analysis. Modern Theology 25, 563587.CrossRefGoogle Scholar
Crisp, OD (2022) Participation and Atonement: An Analytic and Constructive Account. Grand Rapids, MI: Baker Academic.Google Scholar
De Cruz, H and De Smedt, J (2023) Schleiermacher and the transmission of sin: a biocultural evolutionary model. TheoLogica 7, 128.Google Scholar
Dempsey, J (2013) Corporations and non-agential moral responsibility. Journal of Applied Philosophy 30, 334350.CrossRefGoogle Scholar
Douglas, M (1966) Purity and Danger: An Analysis of Concepts of Pollution and Taboo. London: Routledge.Google Scholar
Everhart, DT (2023) Communal reconciliation: corporate responsibility and opposition to systemic sin. International Journal of Systematic Theology 25, 134156.CrossRefGoogle Scholar
Franks, WP (2012) Original sin and broad free-will defense. Philosophia Christi 14, 353371.CrossRefGoogle Scholar
French, PA (1979) The corporation as a moral person. American Philosophical Quarterly 16, 207215.Google Scholar
Gilbert, M (2002) Collective guilt and collective guilt feelings. The Journal of Ethics 6, 115143.CrossRefGoogle Scholar
Gilbert, M (2006) Who's to blame? Collective moral responsibility and its implications for group members. Midwest Studies in Philosophy 30, 112113.CrossRefGoogle Scholar
Goodin, RE (2018) Constitutive responsibility: taking part, being part. Analysis 78, 4045.CrossRefGoogle Scholar
Held, V (1970) Can a random collection of individuals be morally responsible?. The Journal of Philosophy 67, 471481.CrossRefGoogle Scholar
Hindriks, F (2018) Collective agency: moral and amoral. Dialectica 72, 323.CrossRefGoogle Scholar
Hindriks, F (2019) The duty to join forces: when individuals lack control. The Monist 102, 204220.CrossRefGoogle Scholar
Hudson, H (2014) The Fall and Hypertime. Oxford: Oxford University Press.CrossRefGoogle Scholar
Leidenhag, J (2020) For we all share in one Spirit: charismatic gifts and Church unity. TheoLogica 4, 6487.CrossRefGoogle Scholar
List, C and Pettit, P (2011) Group Agency: The Possibility, Design, and Status of Corporate Agents. Oxford: Oxford University Press.CrossRefGoogle Scholar
McCall, TH (2019) Against God and Nature: The Doctrine of Sin. Wheaton, IL: Crossway.Google Scholar
McFarland, IA (2014) From Nothing: A Theology of Creation. Louisville, KY: Westminster John Knox Press.Google Scholar
Mitchell, B (1984) How is the concept of sin related to the concept of moral wrongdoing? Religious Studies 20, 165173.CrossRefGoogle Scholar
Moen, LJK (2019) Redundant group agency. Philosophy of the Social Sciences 49, 364384.CrossRefGoogle Scholar
Otto, R (1923) The Idea of the Holy: An Inquiry into the Non-Rational Factor in the Idea of the Divine and its Relation to the Rational. Harvey JW (trans.). London: Humphrey Milford.Google Scholar
Quinton, A (1975) Social objects. Proceedings of the Aristotelian Society 75, 127.Google Scholar
Rea, MC (2007) The metaphysics of original sin. In van Inwagen, P and Zimmerman, D (eds), Persons: Human and Divine. Oxford: Oxford University Press, pp. 319356.Google Scholar
Rovane, C (1998) The Bounds of Agency: An Essay in Revisionary Metaphysics. Princeton: Princeton University Press.Google Scholar
Rovane, C (2004) What is an agent? Synthese 140, 181198.CrossRefGoogle Scholar
Rovane, C (2014) Group agency and individualism. Erkenntnis 79, 16631684.CrossRefGoogle Scholar
Rutledge, JC (2022) Forgiveness and Atonement: Christ's Restorative Sacrifice. London: Routledge.CrossRefGoogle Scholar
Schwenkenbecher, A (2013) Joint duties and global moral obligations. Ratio 26, 310328.CrossRefGoogle Scholar
Stump, E (2018) Atonement. Oxford: Oxford University Press.CrossRefGoogle Scholar
Swinburne, R (1989) Responsibility and Atonement. Oxford: Oxford University Press.CrossRefGoogle Scholar
Thurow, JC (2015) Communal substitutionary atonement. Journal of Analytic Theology 3, 4769.CrossRefGoogle Scholar
Timpe, K (2021) Sin in Christian thought. In Zalta EN (ed.), Stanford Encyclopedia of Philosophy. Available at Scholar
Timpe, K (2023) The inevitability of sin. TheoLogica 7, 128.Google Scholar
Tollefsen, DP (2015) Groups as Agents. Malden, MA: Polity Press.Google Scholar
Vainio, O-P (2021) The fall and original sin. In Arcadi, JM and Turner, JT Jr. (eds), T &T Clark Handbook of Analytic Theology. London: T & T Clark, pp. 325334.Google Scholar
Wringe, B (2010) Global obligations and the agency objection. Ratio 23, 217231.CrossRefGoogle Scholar
Wringe, B (2020) Global obligations, collective capacities, and ‘ought implies can’. Philosophical Studies 177, 15231538.CrossRefGoogle Scholar