Welcome to the first ever Ethical Theory and Moral Practice discussion! We’re looking at Robin Zheng‘s new article, “What is My Role in Changing the System? A New Model of Responsibility for Structural Injustice”, which can be downloaded here. Maeve McKeown kicks things off with a critical précis, which appears immediately below. Please join the discussion!

Structural injustice is ‘ordinary injustice’ (Young 2011). When individuals go about their daily lives, working, consuming, renting or buying property etc., there are unintended, cumulative outcomes that result in the oppression of certain social groups. The problem is, as Robin Zheng points out, how can individuals be held responsible for this?

On Iris Marion Young’s “social connection model” (SCM) of responsibility, all agents connected to the injustice share political responsibility to collectively organise to struggle against it. But the SCM has limitations. By addressing these limitations, Robin comes up with a new model of responsibility – the Role Ideal Model (RIM). Robin’s paper is very detailed, rich and thought-provoking, and a welcome addition to the literature on how to establish individuals’ responsibilities for structural injustice. I’ll outline Robin’s argument and then explain why, ultimately, I don’t think the RIM improves on the SCM.

The Structural Turn

Robin understands structural injustice through the lens of “the intersectionality thesis,” according to which ‘different oppressions co-constitute and mutually reinforce one another.’ To rectify one form of oppression (e.g. gender inequality), we must also fight against other forms of oppression (racism, classism, ableism etc.). So Robin understands structural injustice as ‘the sum total of oppressions, and the ways in which they interact with and compound one another, taken holistically.’

Robin identifies a structural turn in recent political theory (Young, Haslanger, Lavin). This contrasts to two previous approaches. The “aggregative approach,” which understands individuals’ responsibilities for injustice in relation to their membership of collectivities (the state, military, corporations), and considers how responsibility is distributed among group members. The “individualist approach” takes a moral principle (complicity, unjust enrichment, duties of assistance) and applies it to individuals in relation to injustice.

But structuralists point out that all agents (including powerful agents) are constrained by social structures. They focus on the structures, which, Robin argues, has two advantages: it increases the objects of moral responsibility from discrete actions by individuals or collectives to background conditions, like racism and neoliberalism; and it increases the range of responsible subjects, rejecting the guilty/non-responsible binary, and considering how all agents can be held responsible.  To do this, structuralists distinguish between different kinds of responsibility. They argue that responsibility for structural injustice is forward-looking, not grounded in desert or causation, and non-blameworthy.

Robin argues that theorists in the aggregative and individualist camps understand responsibility as attributability; as a metaphysical problem, which connects responsibility to agency, thus providing grounds for blame or punishment. Structuralists understand responsibility as accountability; as a moral or political problem whereby a community has to distribute the burdens of redress, regardless of who or what caused the injustice. For structuralists, the bar for acquiring responsibility is lowered.

The Role-Ideal Model of Accountability

Like the SCM and other structuralist approaches, the RIM does not replace the attributability model of responsibility. Instead, it establishes responsibility for structural injustice, while recognising that all agents are constrained by the structures.  How it differs from the SCM, is that it grounds that responsibility not in “connection” to structural injustice, but in our pre-existing social roles.

All people occupy multiple social roles (job, relationship, carer, citizen etc.). Each role has a set of expectations that are predictive (beliefs about how a person will act) and normative (beliefs about how they ‘should act and be’).

These expectations depend upon the relationships the role-holder has with others, which are divided into “role-segments.” In each role-segment there are associated forms of behaviour and attitude. So a teacher has relationships with students, parents, superiors, the teacher’s union, policymakers etc. In the role-segment “teacher-student”, the teacher instructs the student on how to do academic work, she feels concern when the work isn’t good, etc.

The expectations are maintained by the role-holder and by others through sanctions, which can be positive/negative, informal/formal, internal/external.

Robin draws on two sociological theories about the relationship between social roles and structure. “Structural-functionalism” seeks to understand how society holds together. On this view, social roles enable the division of labour and facilitate the smooth, continuous functioning of the whole. Social roles are maintained by socialisation (internalisation of system requirements) and sanctions. Society is a ‘boundary-maintaining system.’

“Symbolic interactionism” recognises that social roles are necessarily open to interpretation because they will be occupied by different people. On this view, social structures grow out of interpersonal interactions which are continually negotiated and changed, depending on the people filling the roles and how they interact with each other. Expectations about roles can settle over time, but they are also constantly renewed. Each person sets up a “role-ideal” – their personal interpretation of how a role can best be performed. Individuals are motivated to live up to their “role-ideals” and, when they identify with the role, find it satisfying to fulfil the expectations. This is what enables the preservation of social structure.

Robin builds up the RIM from these insights. From structural-functionalism, she takes the idea that individuals bear responsibility for fulfilling their social roles, which specify a range of duties, and will incur sanctions if they fail. Performing these roles ‘enacts’ structure. Taking from symbolic interactionism, she argues that ‘it is this simultaneous psychological and normative force of role-ideals that connects individual agency to social structure in such a way as to ground moral responsibility.’ The RIM is thus distinguished from the SCM, because the SCM grounds an individual’s responsibility for structural injustice in ‘causal’ connection to the injustice, whereas the RIM grounds it in an individual’s re-enactment of structural injustice through their social roles. The RIM also has a psychological and normative foundation for moral responsibility for structural injustice in the form of role-ideals.

Five Desiderata: The SCM vs. RIM

Robin argues that the RIM is preferable to the SCM for further reasons. It can respond to five practical-theoretical problems that the SCM struggles with.

First, how can individual actions produce structural change? On the SCM, structural change can only be enacted through collective action. But Young doesn’t say much about what that would look like in practice, except to say that it involves ‘pressuring powerful agents.’ Robin argues that this is a limited answer, because powerful agents are also constrained by structures so can only effect partial change, and it is an individualistic approach.

By drawing on the sociological theories described above, the RIM has better answers. Structural-functionalism helps explain why societal transformation is so difficult: change in one sub-system (political revolution) will be counter-acted by pressures from other sub-systems (the global economic order). It also explains the intersectionality thesis; that changing one sub-system (gender) is merely change within the system, rather than of the system. Symbolic interactionism explains that roles can be transformative, as well as constraining. Within a role there is a bundle of expectations, but these can be changed. Structural transformation becomes possible when all individuals throughout the system ‘push the boundaries of their social roles.’ This either leads to incremental change towards a ‘new equilibrium’ or can ‘prepare the way for more ruptural changes.’

Second, if I’m not doing anything wrong, why am I, as an individual, accountable for structural injustice? On the SCM, agents with differing degrees of power have more or less political responsibility, so citizens’ responsibility involves pressuring powerful agents to act on their responsibility. But Robin argues that this isn’t good enough because individuals can continue believing that it’s up to others to make changes, not themselves. On the RIM, all agents in all social roles have to push their boundaries of their role, and it’s their job to fight injustice because that’s what it means to perform the role well.

Third, what actions should an individual take? Young offers four “parameters of reasoning” for political responsibility; an individual should act depending on how much power, privilege, interest or collective ability they have in relation to injustice. From an intersectional perspective, however, an individual can be simultaneously an oppressor and oppressed, thus having different degrees of power, privilege, interest and collective ability, so it can be impossible to know how to act. On the RIM, each role has an associated range of actions, so it’s more action-guiding.

Fourth, how much can an agent be held responsible for? Young’s parameters of reasoning don’t specify how much time or resources an agent should devote to challenging structural injustice. On the RIM, the specification is that individuals must perform ‘all one’s roles with a raised consciousness.’  This is demanding, but since the individual is already performing the work, it’s manageable.

Fifth, how can individuals be held accountable? On the SCM, Young argues that individuals cannot be blamed for failing to take up political responsibility or enacting it in misguided ways; but they can be criticized without blame. Young, however, doesn’t really explain what that means. On the RIM, each role has a range of expectations and if an agent doesn’t fulfil them, they will be held accountable, by being subject to sanctions, mandates from superiors, or reminders.


Robin addresses three objections to the RIM: a) that there are unjust roles, e.g. slave-owner; b) how do you know what it means to perform roles well, e.g. does a good citizen vote for Trump or Clinton?; and c) roles can be unfairly imposed, so why should a person perform the role well. But I want to raise four different objections, by way of which I will argue that the SCM is ultimately preferable to the RIM.


  1. Causal connection vs. enactment of structure

Robin argues that ‘On the SCM, individuals are responsible for unjust outcomes because of their causal contributions to structural processes. By contrast, the RIM maintains that individuals are responsible because their role performances are what constitute unjust structures.’ Thus, on the SCM individuals can avoid responsibility by denying causal connection, but on the RIM, because the individual performing their role in society is constituting that society, they cannot avoid responsibility.

A lot hinges here on causation. But Young does not define “connection” to structural injustice exclusively as causation. I identify four forms of connection in Young’s work: existential, dependent, and causal connection, and reproduction of structures through our actions. Indeed, in Responsibility for Justice (p.59-62), Young argues that structural injustice is constantly reproduced through the actions of individuals. I argue that “connection” ought to be understood as the reproduction of structural injustice through action: e.g. when I purchase clothes in a high street shop, sweatshop labour already exists; I don’t cause it, but I reproduce it (McKeown, forthcoming). Thus, Young’s SCM can be interpreted in such a way as to avoid Robin’s objection and to already accommodate the advantage that Robin ascribes to RIM.

  1. Pressuring powerful agents vs. boundary-pushing

Robin argues that the SCM amounts to individuals pressuring the powerful to change the structures, which in effect means that people avoid their responsibility. She writes, ‘It remains far too easy for individuals to believe that other people can and should do the work of promoting structural change, and that they are morally in the clear so long as their causal contributions are not blatantly wrong. If, say, a white, middle-class, heterosexual, cisgender, able-bodied man does not have any personal stake or interest in combating these various oppressions, he may agree that these are things that are wrong with the world but not feel compelled to act on it, so long as he himself does not engage in overtly prejudiced, exploitative, or biased ways.’

Instead, on the RIM, all agents must push the boundaries of all their social roles. Robin gives the examples of a professor, who should encourage students to adopt gender-neutral language, or sign an open letter to university administration, thereby altering what can be expected of a person within that role. Or a parent , who should consider sending their child to a school in districts they wouldn’t previously have considered.

There are at least three problems here. First, Robin argues that RIM places the burden of promoting structural change ‘within a person’s role – they are burdens that she is already committed to shouldering.’ But it isn’t true that the person is already committed to the burden of challenging structural injustice. Promoting structural change places additional demands upon that role. In one’s role as professor, mother, consumer, teacher, friend, citizen etc. an individual has to add the component of researching structural injustice and the ways in which they can push the boundaries of all their various roles to promote change, and to implement that change. This is potentially a very demanding time constraint.

Moreover, why put that time into changing all of one’s roles, instead of focusing on the area where the individual can actually promote some change? This is where Young’s parameters of reasoning are useful. If the privileged male is in a position of power in relation to a particular injustice, say he works in a corporation of some sort, he can spend his time promoting gender or racial equality in the workplace or the supply chain. If the professor has collective ability as a member of the university, she can use that to push for university-wide changes. If the mother has an interest in integration in her local school, she can campaign on that. Being an activist for justice takes up time, energy, money and other resources. The RIM encourages spreading these resources thinly by trying to improve one’s performance for justice across each and every social role. The SCM encourages focusing on the social position one occupies from which one can most effectively promote structural change.

Second, what’s wrong with pressuring powerful agents to make changes? The professor signs an open letter to the university administration because the university administration can implement university-wide changes, thus implementing more structural change than the professor can as a lone individual. The parent could consider sending her child to a different school district, or the state could implement stricter rules about the distribution of school places and promote integration in the education system. Sure, the professor, the parent, or the privileged white male could shirk their responsibilities to push the powerful for change. But then they are not taking up their political responsibility.

Third, Robin argues that individuals pressuring the powerful to change is too individualised an approach. But the RIM is an entirely privatized and individualized approach to tackling structural injustice. The mother choosing to send her child to a low-income public school, instead of a typically middle-class school, is acting alone and privately. The professor encouraging students to use gender-neutral language is also acting as a private individual. Indeed, as I’ve pointed out, if an individual spends all their time and resources thinking about how they can expand all of their various social roles, they won’t have the time to engage in collective action.

But collective action is necessary. To use climate change as an example, just 100 companies are responsible for 71% of global greenhouse gas emissions. These corporations can do far more to promote environmental justice than any individuals choosing to go vegan, shop locally or use recycled products. Change won’t happen unless pressure is put on these powerful actors to change, whether it’s the corporations themselves or states implementing legislation, and pressure comes from collective action.

  1. The underlying theory of justice

The RIM relies on a tacit underlying theory of justice, which is an intersectional feminist theory. Robin argues that this is not the case; a theory of responsibility cannot tell us what our first-order duties are. Instead it forces individuals to critically reflect on what it means to be a “good X”, which will involve debating about this with others.

But as Robin notes, roles are open to interpretation. An individual’s interpretation of what it means to be a good X will depend on their understanding of justice. Take the uber-privileged man and say he works in a global garment corporation. He is a libertarian, so he thinks that sweatshops provide jobs to workers who want and need them. He thinks that being a good employee involves maximising profits and being accountable to shareholders. He thinks that he is performing his role well. Robin wants to disagree because he’s failing to challenge the intersectional exploitation of sweatshop labour. But he can simply disagree. Even after debating with others, he can hold fast to his capitalist-libertarian theory of justice. He can continue to promote his theory of justice, which will have very different outcomes to the ones Robin wants.

Young’s SCM has the same problem; how do you persuade non-leftists of their political responsibility for structural injustice? But Young has a theory of injustice to fall back on. For her, oppression is the systematic inhibition of self-development (Justice and the Politics of Difference, 1990). Robin defines oppression more vaguely as ‘the ways in which certain social groups exercise power over others.’

Also, Young separates out different structural injustices. So this garment employee might be actively participating in the perpetuation of sweatshop labour, but maybe he’s challenging some other form of structural injustice. He could be campaigning for, say, prison reform. But Robin understands structural injustice through the lens of “the intersectionality thesis,” whereby she understands structural injustice as the confluence of all oppressions. Thus, she can’t make these claims that people can be acting on political responsibility in relation to one form of structural injustice but not others. The “good X” must subscribe to intersectional feminism to fulfil the RIM.

  1. Virtue ethics over politics

Robin’s theory calls on people to strive to be the best that they can within each of their roles, with a view to challenging structural injustice. But isn’t this just virtue ethics? From a virtue ethical perspective, we should all perform our roles virtuously.

Young’s SCM generates “political responsibility.” The political dimension of “political responsibility” is certainly under-explored in Young’s account. But the call to collective action at least implies the political aspect of individuals’ responsibilities for structural injustice. By contrast, throughout Robin’s article, she discusses how the RIM generates “moral responsibility.”  The role-ideal model is a moral theory, not a political theory. This may or may not be a problem, depending on your perspective. But for me, it begs the question: Where’s the politics?





7 Replies to “Robin Zheng: “What is My Role in Changing the System? A New Model of Responsibility for Structural Injustice”. Précis by Maeve McKeown

  1. Hello! I am very excited and honored to be here, and I would like to thank Nikki and the PEA Soup blog for inviting me to this discussion. I’d also like to thank Maeve for an exceptionally clear and thorough précis and insightful commentary (along with her own work on Young, which I have found very helpful!).

    I’ll begin by saying something about these objections in turn:

    1. I completely agree with Maeve about the potential for interpreting the SCM to include both causal connection and constitution/reproduction of structures as grounds for responsibility. Insofar as the most common way of reading the SCM (including what Young herself says) is in terms of causation, I simply wanted to stress the alternative.

    2. Maeve identifies three problems here: A) that people are not committed to the burdens of challenging injustice through their roles, which is demanding, B) that powerful individuals are better able to enact structural change, and C) that the RIM is a privatized approach advocating individual rather than collective action. I hope to clarify my views on these points, because I’m largely in agreement with the motivations behind them.

    Re: A), my argument that people are already committed to challenging injustice is that this is because they are already committed to performing those roles well. The idea here is that a good X is one who is always “on the lookout” or “at the ready,” as it were, to make changes that will enable her to perform her role well. These changes are things that *always* take some time and effort to decide (e.g., in my role as teacher: should I revise my syllabus this year? update my handouts? shift my office hours to a different day? etc.), but they are very much a part of what we would intuitively consider as performing the role well. My claim, then, is just that truly performing a role well requires orienting that role toward structural change — that is, that in my (unavoidable) deliberations about how to perform the role, structural considerations need to be added into the mix of reasons on the table (e.g. by looking up pieces by underrepresented philosophers when I retool one of the units on my syllabus). In other words, doing research is already always “part of the job,” though people normally devote it to figuring out things like what would be most convenient, efficient, industry-standard, respectable, etc. My claim is just that “how and whether this contributes to rectifying structural injustice” needs to enter that deliberation which is already going to be underway.

    To be sure, it is very demanding to perform roles well, and even more so to perform multiple roles well. That’s why I think of this in terms of role-ideals, which cannot be fully attained but are important for guiding our behavior. Part of that means that we will necessarily perform some roles better than others, and here I completely agree with Maeve’s point about using Young’s parameters to narrow down which roles we should focus on. However, theoretically speaking, I still think we are in principle responsible for boundary-pushing in all our roles, even though we cannot in actuality do this, just as we are responsible for performing all our roles well even though in practice we always have to prioritize some over others. We’re expected to do the best we can.

    Re: B) and C), I should definitely clarify that I have no objections at all to pressuring the powerful, and in fact I’m 100% for collective action. I follow Young in thinking that the responsibility is born individually but best discharged collectively. Perhaps the clearest way to put this is to say that individuals are personally responsible for pushing the boundaries of their roles, but by far one of the most effective ways one can do so is through collective organizing together with others. (After all, one of Young’s four parameters is “collective ability,” which is often grounded upon the existence of some well-defined social role.) Just look at teachers’ unions, who have been mobilizing precisely on the basis of their roles as teachers and employees (who on that basis share interests with other public employees). It’s just that I don’t think this is the one and only way to contribute to social change. Importantly, even major changes brought about through pressure may not “stick” if there aren’t wider shifts in culture and ideology. (Right now, I fear, we are in a very scary moment where we see that happening.) Individuals can contribute to these wider changes through their private choices, though I heartily agree that we should always try to go for mass action aimed at the public, i.e. political action. 

    Moreover, insofar as pressuring the powerful and “collective action” narrowly conceived typically involves methods that are time- and resource-intensive, not to mention hazardous for certain groups, there are many people who simply can’t do that. But I don’t think that means they can’t or aren’t contributing in other ways. Patricia Hill Collins, for instance, has excellent work on the many ways that Black women resist injustice (e.g. through creating safe communities, and the way they raise their children) even when they may not engage in traditional protest. 

The point I was trying to make, in sum, is just that there are a variety of ways in which individuals can discharge their responsibility, and that these are all valuable for bringing about lasting structural change. Finally, I very much don’t want someone who simply won’t ever go march on the streets to be able to think that they are thereby off the hook. There are many, many other things they can be expected to do.

    3. These are really important and difficult points. First, I’d like to emphasize that I’m using intersectional feminist theory only in order to characterize “structural injustice,” and I don’t think the RIM is tied to it. Indeed, I think it’s important that a theory of responsibility be neutral between substantive conceptions of justice precisely because I don’t think the Uber-Privileged Man (UPM, hereafter) should be allowed to stop listening as soon as he hears there’s feminism involved. Yes, he can hold fast to his capitalist-libertarian theory of justice after debating, but what’s crucial here – I think – is *that he is required to engage in debate.* He is required to take on the burden of engaging with me, of taking my arguments into consideration as he deliberates on how to perform his role. Those kinds of debates are the essence of politics, and there are no shortcuts out of disagreement. (To be honest, I’m doubtful that UPM would pay any more attention to Young’s conception of the systematic inhibition of self-development.) I definitely don’t have a theory of contentious politics, but insofar as the future of our shared social world is being battled out, my view is just that everyone is responsible for taking part. When push comes to shove, I suppose I would say: *if* UPM is acting out of sincere belief in his theory of justice — rather than, as I think quite likely in real life, acting out of bad faith, motivated ignorance, or ulterior motives like greed and malice or attitudes like racism, xenophobia, and sexism — then he is discharging his political responsibility. But, he is still strongly criticizable on other grounds. He is substantively mistaken in crucial ways.

    If it turns out (however unlikely!) that UPM is an ardent campaigner for prison reform, then there I think — thanks to intersectional feminists, who have done lots of hard empirical work — that we would have very good grounds for a conversation about how he’s mistaken in thinking that prison reform can be achieved without changing exploitative working conditions. I’m not sure why I can’t say that he’s fulfilling his responsibility with respect to some dimension of structural injustice but not other dimensions. It’s been a core tenet of intersectional feminists, after all, that most individuals within the “matrix of domination” have power or privilege along some dimensions while lacking it along others.

    4. This is another great point. The RIM is indeed a theory of moral responsibility, but (though this opens up a whole different can of worms which requires much fuller defense), my own view is that in the modern world, ethics entails politics. One cannot be an ethical or virtuous individual without engaging in politics. That isn’t to deny that there is a difference, because there certainly is, as evident from the running theme of “private vs. public” actions threading through much of what we’ve been discussing.

    As for the politics, I think it arises in the clash between role-ideals that I’ve claimed we’re all responsible for engaging in. That’s most clearly exemplified, for sure, when the clash is manifest in visible collective action projected across a mass public, and in the fact that collective organization is the most effective way of orienting one’s role toward justice. But I do think politics can also take place in more local clashes on numerous different fronts all throughout the system.

    Once again, I’m very grateful to Maeve for this fantastic set of comments! And I look forward to hearing from others in the discussion as well.

    P.S. I do apologize if there is some lag, as there is a significant time difference between me (in Singapore) and people on other continents! But I will definitely make sure to answer every comment as soon as I can.

  2. Hey Robin,

    Thanks for a nice paper! I’m going to ask a different sort of question: do you think that your view is substantially different from Korsgaard’s in ‘Sources of Normativity'(1996)? I ask because it seems as though you could sub in “practical identities” for “roles” and get virtually the same theory of moral responsibility (though of course applied here to the case of structural injustice).

    In particular, this comparison really jumps out when you say: “was it really qua good NSA officer that the whistleblower acts? Is it not that the boundary-pushing behavior derives from duties attaching to pure moral agency itself, irrespective of any roles? To this I say: Yes, so it is – and that is precisely the point. The loyal whistleblower’s actions are an exercise of individual moral agency…”

    Where the requirements of “pure moral agency”, it seems, cannot be given by the requirements of the role but by some meta-role, ‘moral agent’. Because the role of NSA officer quite obviously tells you: “stay silent”, so the normative requirement here seems to come from what you call pure moral agency. Which is exactly what Korsgaard would say, I think.

  3. Hi Robin,

    Thanks for a fantastic paper! I really enjoyed reading it and found it really thought-provoking. I must confess that I share many of Maeve’s worries. I just want to press you a bit more on three points – sorry if I am a bit quick on these comments but I’m away for the weekend:

    (1) There seems to be a strong tension between (i) the need to leave flexibility in the interpretation of social roles (a flexibility that is needed both to ensure the roles do change over time and to respect human agency) and (ii) the necessity for your argument to work to include addressing structural injustice as a crucial feature of *any* role ideal. You seem to be aware of this tension; in fact, your third objection tries to deal with it. However, I found your reply not fully convincing. In a sense, you seem to underestimate the reasonable disagreement about what being, for instance, a good teacher or a good friend means. It is not clear that we should address structural injustice because it is something that we should do in virtue of our social roles. Who should what a social role is for? Who is in the position of telling me that I am not fulfilling my social role? In another sense, you seem to be too permissive in your reply that whether being a good citizen means voting for Trump or not is not something that a model of responsibility for structural injustice should tell us. How can a RIM at the same time arguing that (i) addressing structural injustice is something that we should do in virtue of our social role and (ii) claiming that being a good citizen is compatible with making certain electoral choices?

    (2) RIM seems potentially more demanding that SCM. This is because I am required – in every choice I make, i.e., in every action I take as part of my social role – to consider how it is linked to structural injustice and act on it. The parent who chooses the school for their kids are always asked to give a strong consideration to issues of structural injustice (at the expense of other legitimate reasons, e.g., sending the kids to the best school). For RIM to work and produce gradual change over time, people should almost always prioritise certain considerations over other. This is more demanding than what SCM requires (and it may be possibly unfeasible).

    (3) I agree with Maeve that RIM seems to turn responsibility for structural injustice into a virtue ethics approach. Virtue ethics seems to be unappealing for many reasons. From a structural injustice approach it is unconvincing because (i) it individualises the problem; (ii) it needs to rely on difficult judgements on what being virtuous means and (iii) it is not clear how we can make people more virtuous.

  4. Hi Robin,

    Thanks for a really great paper! And thanks also to Nikki and the Pea Soup blog for inviting me to the discussion.

    As you know, Robin, I’m completely on board with looking at structural issues through the lens of social roles. So I don’t have objections. But I do have a wish-list. Here are a couple of things I hope to hear more about as extensions or clarifications of the framework you’ve given us.

    1. The relationship between role-ideals and partiality, particularly when it comes to ‘positional goods.’

    I really liked your parents/schools example, because it touches on a very basic difficulty. A further distinction might help to make the difficulty even more vivid.

    The example as you described it is plausible to the extent that it is either permissible, appropriate, or morally required to forego certain goods for one’s children. Such limits are pretty uncontroversial within traditional roles, even if people disagree about what those limits are.

    But some goods are positional goods, viz. goods that are only goods relative to what other people have. A person’s ranking on the SAT’s is positional in this sense, because what matters is how they did relative to others. Endeavors that are competitive in nature generate positional goods. So, test scores, market shares, etc.

    Role-bearers might have role-based duties to help others achieve positional goods. Yet, the distribution of those goods may not be just. Take the professor example. Part of being a good professor is providing an education (not positional). But part of being a good professor is also to help their students compete in a job market (positional).

    It is harder for role-ideals to address this kind of injustice, it seems to me. I would love to hear more about how they might do so. That brings me to the second thing on the wish-list.

    2. The difference between non-conformance and pushing boundaries.

    If social roles are socially enforceable, then relevantly-situated participants have to be able to tell when a role-bearer is violating role-obligations and when they aren’t. People cannot impose sanctions for violations they don’t know about.

    Therefore, the core of social roles is performance. Performances are what others have access to. This is unlike reasons for acting. Unless I make my reasons known to others through some kind of performance (conscious or otherwise), my reasons are private and not accessible. Thus, having the wrong reasons when I discharge a role-duty is not something that can be subject to social sanction. Seeming as though I have the wrong reasons can be. But actually having the wrong reasons can’t be.

    Your idea of role-ideals allows for non-performative ‘layers’ on top of the performative ‘core’ of social roles. That really seems right to me because, as you say, sanctions may be imposed first-personally. We are perfectly capable of beating ourselves up about stuff that no one else knows about!

    To summarize, it seems role-ideals do the following work: i) they help specify particular performances in particular contexts and ii) they add inaccessible ‘success conditions’ for discharging role-obligations. Thus, I may fail to discharge my role-obligations even if I perform the required actions.

    This, I take it, is where the NSA example fits in. The NSA employee – let’s call her Annie – who narrowly ‘follows the rules’ may still be failing to discharge her role-duties because narrowly following the rules may contradict her NSA role-ideal. At the same time, conforming to the role-ideal may require Annie to depart from narrowly ‘following the rules.’ So Annie starts breaking the rules. I take it Annie is pushing boundaries in your sense.

    Then there’s the person – let’s call her Alex – who just doesn’t have that kind of cognitive creativity. Alex has come to believe that her role at the NSA has no legitimate moral point or purpose. She’s tried hard to come up with one but just can’t. So Alex stops ‘narrowly following the rules,’ too, in exactly the same ways that Annie did. From the outside they seem the same. Is Alex pushing boundaries, too? Or is she just not being a good NSA employee, full stop? And if the latter, what are the implications for the way that role-ideals help us address injustice?

    Well, that’s it!

  5. Hi Nick,

    Thank you for the question, and for drawing a connection I hadn’t noticed before! I find the concept of ‘practical identities’ very compelling and I can see the similarities. One’s practical identity, perhaps, is the set of role-ideals an individual has identified as core to her self.

    Where I might have to differ, I think, is that I’m not inclined to consider pure “moral agent” itself to be a *social* role. Perhaps one could make it trivially fill out the definition I’ve given, but my sense is that it simply doesn’t have enough distinctive content (its feature are not contingent enough, perhaps) such that we could think of it as being a building block of our actual social structures.

    Also, my language in that passage is a bit misleading – what my response is really meant to be is YES to both of the objector’s questions, not NO and then YES. I still *do* think that the role of NSA officer can be construed as providing the agent with reasons to whistleblow, because a natural way to understand the whole underlying purpose of the NSA is to preserve democracy (which was being undermined by the way particular agents were abusing their roles). But it is also the case that the person is exercising her own individual moral judgment — so she is exercising agency *through* that role in the sense that it is her individual judgment that allows her to see (or perhaps partially constitute, within limits) what the role requires.

    What this highlights is that I am not trying to use social roles to explain everything – in other words, I’m not defending a full-blown role ethics where all of our moral obligations stem from roles. So as it stands I’m agnostic on whether our moral requirements ultimately come from the meta-role of moral agent or elsewhere. Instead, I’m using the concept of ‘social role’ specifically to explain how an individual interfaces with a social structure.

  6. Hi Alasia,

    Thank you for pressing me on these points!

    (1) I suspect that there might be some talking across purposes here, coming from moral philosophy vs. political theory. In particular, I think political philosophers often think of responsibility in terms of substantive “responsibiliTIES,” that is, as something very akin to ordinary first-order moral obligations. Moral philosophers, by contrast, have been very concerned with responsibility in the sense (among others) of what it is that makes me liable to certain kinds of responses (e.g. blame, resentment, punishment) from others.


I think Young is addressing both. And in fact I basically agree with her on the first question that we have substantive obligations to participate in collective action (though I think that the substantive obligation that gives rise to that also allows for other kinds of action). What the RIM is primarily doing is providing an answer to the second. 

It’s NOT saying that a Trump voter is being a good citizen, because that’s not what a theory of responsibility of this sort does. However – if the voter has done due diligence, has sincerely reflected on their reasons and values, etc. – then the RIM could say that this voter has discharged their responsibility (in this very narrow instance). But there are many further things this voter should be doing, and many further substantive reasons that one can marshall (from a theory of justice, a theory of normative ethics, etc.) to criticise the voter’s action.


With respect to role-based reasons in particular: everyone else who is related to a person via their role — that is, everyone who stands in one of the relationships specified by the various role-segments — is in a position to criticise. So, for instance, all other voters and citizens would be in a position to criticize the Trump voter, e.g. on the grounds that they are ignoring how the candidate lacks the proper qualities needed to govern well, that he is likely to actually disenfranchise (at least some) voters of their electoral rights, etc. Teachers whose schools are going to be defunded are in a position to criticise the voter, because the voter’s decisions impact their working conditions. And so on.


(2) I’ll set aside the question of whether the RIM or SCM is more demanding, because I think that really depends on particular cases. (Asking workers to publicise their condition and unionise, for instance, can be extremely costly for one’s livelihood, health, or even life.) But I’ll say some more about my views on demandingness. Samuel Scheffler has said that there are four ways we can respond to a charge of demandingness: we can (1) reject a theory because it’s too demanding, (2) accommodate our personal projects within the scope what morality demands, (3) limit the scope of morality within our lives, or (4) accept that morality is just that demanding, and the people are not living up to it. I am in the last camp, at least in deeply unjust worlds. 

What I mean when I say that we are responsible for something is that there is a *claim* on me such that, in principle, it is appropriate for them to respond to my actions/omissions with a kind of “sanction”. When people suffer the way they do in this world, I do think that each of them has that kind of claim on me. However, what might soften the blow here is to recall that I don’t think the kind of sanction that is licensed by this claim is anything like blame or resentment. I am appropriately subject to the “burden” of being reminded or educated of what needs to be done, but not (necessarily) being blamed.


I should also add that I don’t think it’s incoherent – when it comes to this particular kind of moral claim – for them to be impossible to fulfil. Lisa Tessman has written books on how we can be subject to impossible moral demands. I think that is one tragic consequence of living in a deeply unjust world. 

    (3) I probably need to hear more about this worry, because to my ear (as a moral philosopher) it sounds a bit like a category error. (The alternative theories for the kind of responsibility I’m interested would be, for example, traditional theories that ground responsibility in metaphysical properties like free will, Strawsonian theories that ground it in practices inherent in human social relationships, consequentialist theories that ground it in the future good consequences, etc.) I don’t take the RIM to be at all concerned about whether people are virtuous or not – as I emphasise in the paper, the kinds of moral criticism licensed here do not involve ascribing responsibility as attributability to the agents, so there’s no appraisal of their quality as agents. 

  7. Hi Erin,

    Thanks for these interesting and challenging cases!

    1. I take the point about positional goods being more difficult. However, I think there’s a very strong case to be made here that the overall system through which these positional goods are rendered important is deeply detrimental to the professor’s ability to perform that very role well. As it happens, I have a paper (“Precarity is a Feminist Issue: Gender and Contingent Labor in the Academy”) in which I argue, among other things, that academics’ investment in institutional rankings not only contributes to the job market crisis but also erodes their own academic freedom and powers of faculty governance. What a good professor should do here, then, is work to combat the ideologies and institutional structures that give positional goods their value. (After all, it’s usually professors themselves who provide the expert opinions needed to generate ranking!). In the meantime, she can’t just abandon her student’s interests, but, she should recognize what’s at stake in the bigger picture.

    I don’t know offhand if this could work for all positional goods, but it would be something to try.

    As for 2., I’m happy to agree with this very helpful reconstruction! It does seem to me that there’s an important difference between Annie and Alex even if it’s not apparent to us on the outside, and I don’t think Alex counts as pushing the boundaries of that role.

    I can imagine someone worrying about boundary-pushing being determined subjectively by the agent’s own reasons for acting, i.e. being dependent on whether the agent takes herself to be pursuing an appropriate role-ideal. However, I think this does reflect something distinctive (which touches on some things discussed earlier) about social roles that makes evaluating them different from, say, evaluating whether some action is morally wrong. Social roles, as I briefly mention in the paper, really are highly contingent and contextually-bound. Social structures vary widely across time and cultures, and I really do think that we in our current context might not even be able to imagine what a truly just world would be like. Actually, it’s *not* hard to imagine (if we read sci-fi or critiques of the family) that even some of our most basic and cherished roles, e.g. ‘parent,’ might not exist in a more just system. We just don’t know which current roles would disappear and what new ones there would be. So the question of what a good X really, objectively would be is something that I think does need to be settled politically, by us, as we figure out what kind of world we want to live in. 

    Alex’s case of non-conformance made me think, speculatively, that maybe one way to describe what she is doing is *protesting*. She occupies this role, dis-identifies with it, and thus refuses to conform with it because she is protesting it. So too, perhaps, we might think of workers going on strike as refusing to conform with their roles as a form of protest. (Of course, as I think is more often the case in reality, the alternative is to think of striking as itself the pursuit of a certain role-ideal: a good worker is one who does what is necessary to improve the labor conditions needed for her and her colleagues to do her job well.)

    This is interesting for me to think about because it has the flavor of an objection that’s been made against structural-functionalists, i.e. that they are inherently conservative. One might make the same objection against the RIM insofar the responsibility to push the boundaries of a role — even in ways that would ultimately eliminate the role altogether — just feels kind of conservative because there is some default presumption of commitment to the role/the system. In other words, the RIM doesn’t seem to license protest which is pure utter rejection; it can only recommend protest which is grounded in *some* kind of commitment to the role. I’m not sure this is really a problem for the RIM. I myself am not trying to claim that all morally good actions are good because of social roles. So there could be other reasons that this kind of protest is morally good, even if it’s not a way of discharging the kind of responsibility one has in virtue of occupying social roles that constitute an unjust structure.


But I seem to be digressing now – many thanks again for the stimulating questions.

Comments are closed.