When and Why Is Social Kind Membership Morally Wrong? On the Moral Foundations of Ontic Injustice

1 Introduction

In her concise and insightful book Ontology and Oppression (Jenkins 2023), Katharine Jenkins introduces the concept of ontic injustice – a moral wrong that extends beyond familiar forms of injustice such as the unfair distribution of resources or disparate treatment. At stake is the very construction of social reality itself. Ontic injustice concerns the inherent injustice of certain social kinds, shaping how individuals are socially constructed in ways that systematically wrong them – even in the absence of specific harmful experiences or effects. In other words, it is not a matter of unequal outcomes but of unequal beings (Jenkins 2023: 3). Consider a society in which the social kind “woman” is essentially tied to sexual subordination to men. Jenkins argues that even if no actual instance of marital rape occurs in that society, the mere fact of being socially constructed as a woman constitutes an injustice. This is because under these conditions, the social kind itself imposes a burden of social constraints that impinge on moral claims concerning sexual self-determination (Jenkins 2023, 22–24).

Notably, Jenkins clarifies that her proposal does not imply that social kinds, in and of themselves, are ontically unjust. Justice does not require the complete absence of social kinds; rather, her framework accommodates a spectrum of social kinds – including morally permissible ones, such as reformed instantiations of “woman” in which wrongful constraints have been abolished, as well as those that, if they exist, would constitute cases of ontic injustice. Social reality is thus analyzed along a continuum, with some social kinds being necessarily ontically unjust (e.g., slaves), while others are only contingently so (e.g., women) (Jenkins 2023, 39–40). The inclusion of both ontically unjust and permissible – or even valuable – social kinds raises two central questions about the adequacy of Jenkins’s framework for analyzing ontic injustice: (1) Which social kinds are ontically unjust, and (2) why, in particular, are they wrong? Put differently, I am concerned with Jenkins’s account of what makes a social kind inherently unjust, even when its constraints do not manifest in concrete actions but remain merely dispositional – and how we can determine when this is the case. Addressing these questions is crucial because it clarifies the boundaries and meaning of “ontic injustice,” prevents conceptual confusion, and enables a critical assessment of the framework’s underlying assumptions.

In what follows, I critically examine Jenkins’s theory of ontic injustice, evaluating its ability to provide a robust foundation for understanding unjust social kind membership. While her framework provides valuable insights into the moral wrongs of social construction, I argue that it lacks a compelling approach for determining when and why social kinds are morally objectionable. Specifically, I contend that her account is both conceptually and normatively inadequate, as it fails to provide clear criteria for addressing these fundamental questions.

My critique is twofold: (1) Her theory lacks a precise conceptual criterion for determining when a social kind is ontically unjust, and (2) it does not offer a compelling normative criterion explaining why it is unjust. I argue that a key weakness of Jenkins’s approach lies in the tension between her implicit commitment to a form of moral essentialism and her reliance on an overly positivist ontology of social kinds. This tension is most evident in her endorsement of practice positivism – the view that social kinds are fully constituted by actual social enablements and constraints – which may conflict with the individual entitlements of moral persons. In doing so, her account fails to recognize the normative entanglement of social facts with moral claims. As an alternative, I propose moderate practice idealism, which incorporates counterfactual constitutive standards as a necessary criterion for evaluating social kinds.

Moreover, building on Jenkins’s framework, I offer a novel interpretation of her approach to the moral foundations of ontic injustice, situating it within broader philosophical discourse. Throughout the paper, I reconstruct Jenkins’s discussion into three strands, for which I introduce the labels (a) the expressive harm account, (b) the respect-based account, and (c) the improvement-claim account. Jenkins herself speaks of “moral injury,” “moral disrespect,” and “risk of harm”; I take these terms to correspond respectively to my three labels, which I use as a consistent framework to situate her analysis in a broader philosophical context. By examining these strands in relation to existing debates in moral and social philosophy, I argue that, while they offer valuable insights, they ultimately fall short of providing a sufficiently robust foundation for ontic injustice. In response to these limitations, this paper reassesses the theoretical underpinnings of ontic injustice, aiming to refine our understanding of how social kind membership can be morally problematic. This analysis deepens critical engagement with Jenkins’s work and broadens the conversation on the normative dimensions of ontic injustice. My intention is not to reject the concept of ontic injustice outright but to identify ways in which it can be more robustly grounded and further developed.

My argument is structured as follows: First, I reconstruct the core elements of Jenkins’s approach, particularly her conception of social kinds as configurations of constraints and enablements. Next, I challenge her claim that ontic injustice can be determined by comparing these actual constraints and enablements with the entitlements of a moral person. Finally, I critically examine the three grounding strategies Jenkins offers for explaining the moral foundations of ontic injustice.

2 Jenkins’s Account of Ontic Injustice

Jenkins advances the thesis that the mere existence of certain social kinds can give rise to a specific form of injustice, which she calls “ontic injustice” (Jenkins 2023). More specifically, individuals are wronged simply by being placed within these kinds. This injustice stems solely from their membership in a particular social kind, rather than from any practical consequences thereof. While some social kinds may also be morally objectionable for consequentialist reasons (such as diminished well-being) or for deontological reasons (such as exposure to arbitrary harm), Jenkins contends that – even in the absence of additional wrong-making features – mere membership in a social kind can constitute a distinct form of moral wrong: ontic injustice. This wrong, she argues, is independent of how individuals subjectively experience their membership and persists regardless of whether they suffer any actual harm to their material or psychological well-being as a result of being socially constructed in this manner. She writes:

I argue that social kinds can be such that individuals who are socially constructed as members of those kinds can be wronged by the very fact of being socially constructed in that way.… Ontic injustice is a wrong at the level of being that affects individuals simply in virtue of their membership in the social kind in question, regardless of their specific experiences. (Jenkins 2023, 3; emphasis in the original)

While Jenkins primarily focuses on social kinds associated with gender and race – treating them as paradigmatic cases of ontic oppression, a particularly severe subtype of ontic injustice – her approach clearly extends to the entire spectrum of social kinds.^[1]

At the heart of Jenkins’s approach lies a critical examination of the constraints and enablements that define social kinds. She develops this analysis through what she terms the constraints and enablements framework (CEF), which holds that “what unifies the members of such kinds is the fact of being subject to similar social constraints and enablements – or, to put it differently, the kinds are unified by the property of falling under such-and-such constraints and enablements” (Jenkins 2023, 84). By grouping social kinds based on these shared features, her method is broad enough to align with nearly all major contemporary theories of social ontology, including Searle’s constitutionalism and Ásta’s conferralism.^[2] Moreover, Jenkins contends that it is precisely these dispositional features that can violate individuals’ moral entitlements: “The thought is that in some cases, these constraints and enablements may be wrongful, in the sense that they are in contravention of the individual’s moral entitlements” (Jenkins 2023, 22). She connects justice to the recognition and protection of moral entitlements, arguing that social kinds have the potential to wrong individuals simply by virtue of membership. The proper application of ontic injustice, then, depends on whether a social kind’s constraints and enablements infringe upon the moral entitlements of moral persons. Indeed, Jenkins argues that the essence of ontic injustice “is the mere fact that an individual is a certain sort of social being, where what it means to be that sort of social being is to be subject to morally inappropriate constraints and enablements” (Jenkins 2023, 23). Within this framework, justice is fundamentally about ensuring that individuals receive what they are morally entitled to, with its core emphasis on recognizing and upholding these entitlements.

While Jenkins draws on Iris Marion Young to develop clear criteria for ontic oppression, I argue that the broader category of ontic injustice lacks sufficient conceptual and moral grounding to be established as a distinct form of injustice. In the following section, I demonstrate that Jenkins’s proposed conceptual criterion for determining when a social kind is ontically unjust is inadequate. Her approach – evaluating ontic injustice by comparing a social kind’s constraints and enablements with an individual’s moral entitlements – is fundamentally flawed. The primary issue is that the defining features of a social kind cannot be neatly separated from a person’s moral entitlements. In many instances, these entitlements arise within specific social contexts and are more deeply intertwined with social kinds than Jenkins acknowledges. This conceptual confusion undermines her argument, making it difficult to determine when a social kind is to be deemed ontically unjust.

3 When Is Social Kind Membership Morally Wrong?

As outlined in the previous section, Jenkins (2023, 37–39) appears to offer a prima facie clear conceptual criterion for applying the concept of ontic injustice: one must first establish an account of the specific entitlements of moral persons (the normative condition) and then assess these against the actual constraints and enablements that define the relevant social kind (the ontological condition).^[3] Jenkins (2023, 37) writes:

To yield detailed judgements about actual cases, [the concept of ontic injustice] needs to be combined with two further components: first, an account of the moral value of the individual(s) in question and the moral demands to which this value gives rise, and second, an account of the ontology of the social kind in question … and in particular of the constraints and enablements that partly constitute it.

As I will demonstrate shortly, the most natural way to articulate this core idea reveals a tension between two positions: practice positivism and moral essentialism.

A useful way to illustrate the essentialist implications of the normative condition is through Robert Nozick’s entitlement theory. As presented in Anarchy, State, and Utopia (Nozick 1974), Nozick’s theory provides a comprehensive account of moral entitlements within a liberal society. According to Nozick, one fundamental category concerns property rights, which are grounded in the principles of just acquisition, voluntary exchange, and rectification of past injustices. Under this account, individuals have an inviolable entitlement to their holdings, and any interference with these holdings – unless it results from a just process – constitutes a violation of their rights. Nozick’s theory offers thus a precise model for generating detailed judgments about actual cases of moral entitlement violations. From this perspective, the social kind “taxpayer” is ontically unjust because it inherently conflicts with the moral entitlements of those classified within it. Taxpayers are subjected to the coercive power of the state, which compels them to surrender their resources (taxes) to the government. This directly contradicts the moral entitlement of each individual to freely control their own property (Nozick 1974, 169).

One need not find Nozick’s assumptions about natural rights convincing to recognize that they illustrate the moral-essentialist dimension built into Jenkins’s approach. By moral essentialism, I do not refer to one specific meta-ethical position but rather to a family of views that share the assumption that moral truths hold independently of contingent social realities and are, in this sense, essential rather than conventional. These normative facts do not rely on the social contexts; instead, they are viewed as intrinsic to the context-independent nature of moral agents. Moral essentialism in this sense thus conceives of moral entitlements as existing independently of social, historical, and cultural backgrounds, rather than as being constitutively shaped by social relationships, traditions, and communal ties. Jenkins’s claim that some social kinds are inherently unjust because they impose “morally inappropriate constraints and enablements” (Jenkins 2023, 23) suggests a commitment to an essentialist perspective, as it presupposes the existence of external normative truths that must hold logically independent of social reality and be theoretically valid prior to it in order to serve as standards of evaluation.

However, regarding the ontological condition, Jenkins’s theory appears to presuppose practice positivism – the view that the social validity of the constraints and enablements defining social kinds is fundamentally conventional (Applbaum 1999, 45–60). This results in a tension because some of these constraints and enablements are normative in nature, establishing a normative status that determines appropriate behavior, delineates rights and responsibilities, defines objectives, structures competencies, and sets internal criteria of achievement. More specifically, certain constraints and enablements function as deontic elements, prescribing what individuals are permitted, obligated, or forbidden to do based on their membership in a particular social kind. These elements serve as action-guiding standards that shape the normative structure of social kinds. Searle refers to these as “deontic powers,” while others describe them as “role obligations” (Hardimon), or – in Jenkins’s case – “deontic constraints and enablements” (Jenkins 2023, 92–96). If practice positivism holds, then the action-guiding properties of social kinds do not derive from external moral principles but instead emerge through socially established and recognized norms, rules, or practices. Thus, a social kind exists and exerts normative force when its constraints and enablements are collectively acknowledged and maintained within a given social practice.

This core idea can be developed differently within social ontology. John Searle’s theory of constitutive rules emphasizes that social kinds are created through collectively accepted rules. For instance, money functions as a means of exchange precisely because it has been institutionalized through a collectively acknowledged rule such as “X counts as Y in context C” (Searle 1996; 2011). By contrast, Ásta’s conferralism provides an alternative social ontological account, in which social kinds are not constituted by rules but by acts of status conferral. In this model, social kinds exist because certain individuals are assigned a specific status by others, thereby conferring specific constraints and enablements. Despite their differing ontological formulations, both approaches share the central commitment of practice positivism: the characteristics of social kinds are derived primarily from their socially established and recognized status, rather than from any inherent moral properties.

Likewise, Jenkins’s CEF suggests that social kinds are determined by the structure of actual social practices, generating deontic constraints and enablements with normative force. Although she acknowledges the complexity of how normative constraints function within social reality, claiming that “deontic constraints need to be understood in a way that allows that the way in which entitlements, obligations, and so on operate in practice is often very different from the way in which people explicitly believe them to operate” (Jenkins 2023, 93–94), her discussion of Brännmark’s account further clarifies that these normative elements, while often rooted in implicit attitudes rather than explicit beliefs, still operate within and depend on given social environments. Thus, even though Jenkins’s framework allows for a degree of discrepancy between explicit representations and actual normative operations, it still grounds deontic constraints in the structure of social practices rather than in independent moral facts. This, however, creates a tension with her commitment to moral essentialism, which maintains that the injustice of a social kind is based on moral facts that exist independently of social reality. If we take Jenkins’s CEF at face value, it implies that social kinds gain normative significance through their role in shaping structured social reality. Yet, if ontic injustice is meant to be independent of subjective experience and social recognition (as Jenkins claims), then her account cannot rely solely on practice-based criteria for injustice. Instead, it requires a more robust moral foundation – or, as Jenkins herself puts it, “there needs to be some gap between actual moral worth and social representation of moral worth” (Jenkins 2023, 28).

However, there are at least two decisive objections to moral essentialism as a means of normatively grounding ontic injustice: the first, drawing on moral phenomenology in the spirit of Peter Strawson, and the second, concerning the misdescription of entitlements and constraints.

First, moral essentialism fails to capture an important feature of our practices of critique in relation to social kinds: namely, that some of their normativity arises entirely within practice itself. When we call someone a bad video game player, this is indeed a normative criticism – yet it is not “tracking” any obvious independent moral norm. Rather, the practice of gaming itself generates standards that can be meaningfully violated, and these violations trigger evaluative and reactive attitudes (frustration, disappointment, even blame). As Valentini (2023) argues, many socially constructed norms are content-independent sources of normativity: the fact that a norm exists can itself create obligations to comply with it, not merely because it causally furthers safety, efficiency, or fairness, but because agents are entitled to expect compliance with norms they are committed to. This normative force stems from what Valentini calls the “agency-respect requirement”: by breaching a norm, we wrong not only those directly affected, but also the community whose shared commitments sustain the norm.^[4]

The same insight becomes sharper when we look at more deeply value-laden social kinds such as “father.” Here too, the normativity cannot be explained by assuming the kind simply tracks an external moral norm. “Good father” has historically meant very different things: at one time, authority was seen as the defining normative core, whereas today, care is often considered essential. At first glance, a moral essentialist might respond that calling someone a “bad father” is normative because the social kind of fatherhood tracks an essential moral norm, such as a duty of care. But this explanation is too thin. As Jaeggi (2018) points out, the content and meaning of social kinds are not simply instantiations of timeless moral standards, but evolve historically and socially. Thus, what it means to be a good father is not a mere application of a fixed, external norm, but part of an ongoing negotiation of what the social kind itself is. The normativity here is not simply derivative from independent moral truths; it is embedded in the very social practice and its history. Changes in the content of social kinds are not just disputes about which moral standard a kind should track – they are disputes about what the kind itself is, and which expectations and entitlements are constitutive of it.^[5]

This perspective undermines moral essentialism, which assumes that moral assessment must ultimately rely on fixed, external standards – and that only from their alignment with such standards can social kinds derive their normativity. But cases like the bad video game player (or Valentini’s own examples of queuing or driving on the left) show that some forms of critique are genuinely normative even when they do not track a prior moral norm. Jaeggi’s analysis adds a further layer: it shows that not only do some practices generate their own normativity, but that the very content of such kinds can shift historically. Their authority is generated by the shared practice itself. This suggests that our judgments concerning social kinds are not merely a matter of social conformity but instead engage with the deeper normative dimensions embedded in social structures. Overstating the divide between essential moral truths and their social representation therefore risks obscuring the genuine, practice-based sources of normativity embedded within social kinds themselves.

Secondly, moral essentialism fails to adequately describe the constraints and enablements of social kinds because it misrepresents the social-ontological conditions that constitute them. In what follows, I focus on social kinds – categories such as “wife,” “lawyer,” or “convicted criminal” – which are embedded in social practices and institutions. Moral essentialism tends to treat such kinds as if their wrongness or rightness could be assessed against fixed, external standards, rather than by attending to the way these kinds are defined, constituted, and normatively loaded within social reality itself.

To see this, it is helpful to distinguish three dimensions of how social context relates to social kinds, each of which is indispensable for understanding their normative status.

1. Semantic: social contexts determine the meaning of a social kind – they define which descriptions even make sense within that practice. Whether an act counts as a “goal” in football, a “lie” in legal advocacy, or a “promise” in contract law depends on shared criteria that fix the boundaries of these terms. Calling lawyers “liars,” for instance, collapses the distinction between strategic legal argument and deception because it ignores how the practice defines what counts as lying.

2. Constitutive: social contexts constitute social kinds – they do not merely shape how we talk about these kinds, but bring them into existence in the first place. There is no “goal” without the game of football, no “lawyer” without the legal system, no “wife” without the institution of marriage. Constitutive rules and acts of status conferral make these kinds what they are – they generate the very framework within which meanings and norms can emerge.

3. Normative: social contexts determine the normative function of a social kind – the entitlements it grants and the constraints it imposes. Social kinds like “judge” or “citizen” are not neutral labels; they come with powers, permissions, and obligations. These normative profiles do not float free from meaning or constitution: they are shaped by what the kind means and how it is brought into being.

Moral essentialism struggles with all three. Even if it can concede that kinds have social meanings or are socially constituted, it tends to assume that their normative significance can be judged independently from these dimensions. But the three are deeply entangled: constitution sets the stage, semantics fixes the terms, and normativity distributes the entitlements and constraints.

From the perspective of John Searle’s theory of constitutive rules, social kinds exist because we collectively accept rules of the form “X counts as Y in context C.” These rules not only assign functions (this paper counts as money; this person counts as a judge) – they constitute the kind itself. Similarly, Ásta’s notion of status conferral shows that social kinds emerge when acts of classification or recognition grant individuals a certain standing (Ásta 2018). These theories are fundamentally accounts of constitution: they explain how kinds like “lawyer” or “wife” come into being at all. But constitution is not isolated from the other dimensions: the very rules that create kinds also set their semantic contours (what counts as “lying” in legal advocacy) and their normative profile (what a “wife” or a “citizen” is entitled or required to do). These structures are not merely descriptive: they set the boundaries of the very concepts we use, delimit what counts as consent versus coercion, and fix the constraints and enablements that guide actions within these kinds. Ignoring this framework risks crude misdescriptions: saying “lawyers are liars,” for instance, conflates strategic advocacy with deception because it strips away the internal distinctions that make such descriptions intelligible in the first place. It overlooks the constitutive rules that create the very role of the lawyer, the semantic criteria that define what counts as lying within legal practice, and the normative profile that entitles lawyers to use certain forms of argument while prohibiting others.

This matters for the debate about ontic injustice. If we assume, as moral essentialism does, that a kind’s injustice can be determined simply by contrasting its constraints with an abstract list of moral entitlements, we miss how those entitlements themselves are partly generated, specified, and limited by social kinds. Jenkins’s focus on the entitlements of moral persons, while valuable, underplays this embeddedness: it risks flattening the normative landscape into a binary of “entitlement-respected” versus “entitlement-violated.” But assessing whether a constraint is unjust requires understanding the kind itself and how it integrates into broader practices.

This becomes particularly clear if we consider Jenkins’s own case of marital rape. I fully agree that it constitutes a grave injustice. My point is not to dispute the correctness of her claim, but to show that explaining why it is unjust cannot be done simply by contrasting it with an abstract entitlement to bodily autonomy. Entitlements are rarely freestanding: they are shaped, specified, and sometimes restricted by the social kinds and practices in which they are embedded.

Consider the entitlement to freedom of movement. It would be wrong for me to lock my neighbor in my basement because this would violate his entitlement to freedom. According to Jenkins, if a social kind is structured in a way that systematically violates such an entitlement, it must be considered ontically unjust. But this entitlement does not extend to individuals belonging to the social kind “convicted criminal”: imprisoning convicted criminals does not wrong them. Merely knowing that individuals generally have an entitlement to freedom does not help us assess whether it is unjust for convicted criminals to be deprived of it.

The same issue applies to the social kind “wife.” Cynics might argue that criminals, qua criminals, are not entitled to move freely, just as wives, qua wives, are not entitled to fully autonomously decide about their sexual lives. These cynical arguments are not just rhetorical tricks – they expose a conceptual vulnerability: if we treat entitlements as isolated moral facts, we risk taking the structure of the social kind for granted. To determine the legitimacy of enablements and constraints, we must already scrutinize the social kind itself and how it integrates into the broader network of practices.

Moral essentialism, by assuming that injustice can be read off from individual entitlements alone, ignores this deeper layer. It risks flattening the normative landscape into a binary of entitlement-respected or entitlement-violated, instead of asking whether the social kind that defines the entitlement is itself justifiable. This is the crucial point: the wrongness of ontic injustice cannot be assessed solely by comparing constraints to abstract rights – it must also examine the meanings, constitutive rules, and normative functions that define those rights in the first place.

These considerations illustrate that entitlements do not emerge in isolation as purely moral claims; rather, in many cases, they arise from the broader social context. In other words, they are constitutively connected to our social kind membership and the way it relates to other social kinds within the social system. Thus, if neither practice positivism nor moral essentialism offers a sufficient framework for evaluating ontic injustice, an alternative approach is needed – one that avoids the pitfalls of both conventionalism and moral essentialism. I propose that moderate practice idealism provides such a framework by establishing constitutive rules that define the ideal characteristics of a social kind, thereby distinguishing just from unjust instances. This perspective, drawing on insights from Korsgaard, Jaeggi, and Hardimon, suggests that social kinds must be critically evaluated based on their functional roles and normative justification within social practices rather than external moral absolutes or mere social conventions.

Jenkins’s commitment to moral essentialism is particularly puzzling given her claim that her approach aligns with the critical theory tradition of the Frankfurt School (Jenkins 2023, 11). Critical theory in the Frankfurt tradition, especially in its contemporary forms, rejects essentialist moral frameworks in favor of an immanent critique that evaluates social structures based on their internal contradictions. Rahel Jaeggi’s critique of moral essentialism exemplifies this stance, emphasizing that moral criteria emerge from within historical and social contexts rather than being fixed, pre-social truths. Jaeggi (2018) argues that an essentialist approach to injustice risks being ahistorical and dogmatic, as it assumes that moral entitlements exist independently of the very social conditions that shape them. Instead, she maintains that injustices should be identified through contradictions within social practices – by showing how they fail according to their own internal logic – rather than through an external, preordained moral framework. By grounding ontic injustice in the idea that certain social kinds are inherently unjust – regardless of their historical, social, or practical manifestations – Jenkins appears to diverge from this key Frankfurt School commitment.

Granted, as “emancipatory theorists” (Jenkins 2023, 94), we do not merely accept conventionally given social kinds, constraints, and enablements; rather, we seek normative resources for their critical moral evaluation. This necessity suggests that both practice positivism and moral essentialism must be rejected in favor of a critical social ontology that incorporates evaluative standards. If ontic injustice is to be meaningfully assessed, we need a way to compare social kinds without simply treating them as isolated facts. One promising approach is to compare membership across different social kinds rather than juxtaposing a kind with a non-social individual. However, this comparison must be carefully structured. It is important to avoid comparisons between entirely unrelated kinds, since meaningful results require that the kinds being compared sufficiently overlap. To achieve conceptual clarity, the context of application must be clearly defined, so that comparisons are made under appropriate ceteris paribus conditions. In this case, the inquiry would not focus on whether being treated as a wife aligns with our entitlements as an individual moral agent, but rather on whether, when compared to another counterfactual set of constraints associated with being a wife, it morally performs worse.

To undertake a comparison between the actual social kind “wife₁” and the counterfactual kind “wife₂,” we must avoid practice positivism – which would individuate social kinds solely based on their actual constraints and enablements. While Jenkins’s moral essentialism has already been critiqued for importing external moral truths, her account here errs in the opposite direction: it remains too ontologically positivist, as if the ontology of social kinds were exhausted by their existing practices and constraints. Instead, I propose what I call moderate practice idealism: an approach that establishes counterfactual constitutive standards defining the ideal characteristics of a particular social kind. Deviations from these standards can then be treated either as disqualifying an entity from belonging to the kind entirely or as producing a defective, problematic instance of that kind. Christine Korsgaard (2009, 33) captures this distinction clearly:

Say we have two objects, call them A and B, and they are in some respect different from each other [i.e., unjust kind Wife and just kind Wife]. They have some different non-accidental properties. Now we need to distinguish two ways that A and B can be different from each other in this way: A can be a different kind of thing from B, or A can be a defective instance of the same kind of thing as B.… If properties are all that matter, then we need not – and cannot – distinguish the different from the defective: different collections of properties will just be different, and that is all.

To avoid collapsing all differences into mere variation, it is necessary to establish counterfactual ideal constitutive standards to distinguish ontically unjust from ontically just social kinds. If we assert that it is unjust for a wife, qua “wife₁,” to lack autonomous decision-making power, then we must recognize this constraint as a defective instance – or a problematic deviation – from the constitutive standards of the ideal social kind type “wife₂.” As demonstrated earlier, attempting to establish constitutive standards based solely on an individual’s entitlements is impractical. The context is too underdetermined to adequately capture the essence of a social kind in this way. Instead, a more viable approach would be to ground these standards within the broader structure of social kinds and their normative functions within a social system.

While I cannot present a comprehensive framework for establishing ideal constitutive standards for social kinds at this point, I offer here a preliminary outline. Social kinds serve specific functions within social practices and their broader ensembles. They address distinct practical challenges by structuring social interactions (Haslanger 2018; Jaeggi 2018). Rather than succumbing to moral essentialism, which assumes the pre-social worth of individuals, it is more fruitful to examine the roles these individuals play within the broader social framework, particularly as they are embedded in lived social practices. Consequently, it is crucial not to conflate an inquiry into the factual structure of a social practice with the question of how it ought to be structured. Instead, we must examine whether a given social kind successfully renders itself intelligible for those affected.

This approach aligns with Hardimon’s insight (1993, 348):

To say that a social role is reflectively acceptable is to say that one would accept it upon reflection. Determining whether a given social role is reflectively acceptable involves stepping back from that role in thought and asking whether it is a role people ought to occupy and play. Determining that a given social role is reflectively acceptable involves judging that it is (in some sense) meaningful, rational, or good.

Hardimon’s framework implicitly treats the function of a kind as normatively charged: to ask whether a social kind is “reflectively acceptable” is to ask whether its defining function can be endorsed as meaningful, rational, or good.^[6] At this point, it is important to clarify why Hardimon’s notion of reflective acceptability does not collapse into moral essentialism. The latter appeals to external, fixed moral truths to evaluate social kinds (e.g., claiming that it is wrong because it violates an independent duty of care). By contrast, Hardimon’s reflective acceptability relies on immanent critique: it asks whether a social kind can be endorsed “upon reflection” from within the practices that constitute it, rather than by appealing to some fixed list of moral truths.

Because social kinds inevitably impact the lives and actions of multiple individuals, their enablements and constraints must be intersubjectively justified within the context of a shared social practice (Stahl 2021). In this regard, it is crucial to recognize that practices are inherently tied to specific purposes, which functionally limit the scope of permissible enablements and constraints. Every context of activity has its own aims and values; every social kind exists for a purpose. Consequently, constitutive standards of social kinds cannot be constructed arbitrarily if they are to facilitate successful participation within a practice. There is, however, a familiar problem for any view that grounds justice claims in the constitutive structure of social kinds: if we simply take the existing constitutive rules at face value, we risk endorsing whatever norms a kind happens to have – a “feedback loop” that makes every entrenched function self-justifying. To avoid collapsing all differences into mere variation, moderate practice idealism builds in counterfactual standards: it asks not only whether a social kind serves a function, but whether that function could be justified under idealized, reflectively acceptable conditions.

Here, it is crucial to note that function is not a neutral, merely causal notion. Functions are intrinsically normative descriptions: to say that the function of a lawyer is to advocate for a client, or that the function of a judge is to impartially decide cases, is not simply to describe what these kinds do, but to indicate how they ought to be performed and when they fail. A “defective” lawyer is not just a lawyer who does something different; they fail to meet the internal standards implicit in the role’s function. This normative weight of function is precisely what allows moderate practice idealism to use function as a critical lens without falling back into moral essentialism. Yet, as Jaeggi (2018) emphasizes, functions do not derive their meaning from isolated practices alone: they are embedded in an interwoven nexus of related practices and social kinds. The good functioning of a lawyer, for instance, depends on adjacent roles (judges, clients, clerks) and the institutional contexts (courts, laws) that sustain the practice. Functions thus gain their normative force not only from internal standards but also from how they connect to and coordinate with surrounding practices. Crucially, moderate practice idealism also preserves the gap between descriptive functioning and normative justification: even if the social kind “slave” once “functioned” effectively within the practice of slavery, the very idea of that practice collapses under the condition of rational acceptability. Once we ask whether the function itself could be endorsed upon reflection, slavery’s “good functioning” becomes normatively meaningless – it cannot be justified, even if it appeared internally coherent in its historical setting. By asking whether the very function of a kind could be justified under reflective scrutiny, moderate practice idealism avoids the trap of merely ratifying every existing function as legitimate.

However, even with this clarification, determining the inherent logic of a practice is often contentious. Any given community will harbor multiple – and possibly conflicting – ideals. The question of which among these ideals may legitimately be subjected to critique or take precedence over another is not predetermined by any external moral essence. Still, given persistent disputes over these perspectives, it is difficult to directly discern whether a specific social kind unjustly wrongs individuals or groups – or whether such wrongness is an integral feature of its functional logic. Moreover, conflicts concerning constitutive rules are rarely resolved merely by appealing to specific purposes. As Jaeggi (2018, 151) observes, such conflicts arise due to “sharp disagreement over both the ends and the means and corresponding leeway when it comes to how they are interpreted and shaped.” Accordingly, the proper understanding of practices and their goals must remain open to debate. Thus, the criterion of aligning counterfactual ideal constitutive standards for social kinds with the necessary structural conditions of a practice is, at best, a dynamic and mediated standard for general acceptability. Even if, as is frequently debated, internally necessary conditions can be identified for a specific type of practice, this does not automatically provide sufficient justification for adhering to or engaging in that practice beyond its internal reference system. As Jaeggi (2018, 153) puts it: “For us as observers, therefore, these two things – the evaluation of a practice as meaningful and the recognition of its internal excellence – may diverge in certain cases.”

For the remainder of this article, I set aside this broader topic as it diverges too much from Jenkins’s approach. Instead, I revisit her proposal regarding the specific nature of the wrongness in ontic injustice. Until now, I have argued that Jenkins’s theory lacks adequacy for two primary reasons. First, she supports practice positivism which hinders the differentiation between ideal social kind types and their defective instances by failing to distinguish between counterfactual constitutive standards and actual constraints and enablements. Second, Jenkins appears to seek the determination of the morally wrong aspects of ontic injustice solely through the moral entitlements of pre-social individuals. This commitment to moral essentialism presents a problem because of the normative constitution of social kinds. Numerous moral entitlements rely on a broader social context that extends beyond basic individual claims. The justice of a social kind depends not only on its ability to prohibit severe violations of basic entitlements but also on whether it is inherently meaningful rather than alienating – in other words, whether it is reflectively acceptable within socially constructed reality.

4 Why Is Social Kind Membership Morally Wrong?

In what follows, I will argue that even in cases where the wrong of ontic injustice is grounded in violations of individual moral claims, Jenkins’s framework for identifying these claims has blind spots. Her explanation draws on three distinct notions – “moral injury,” “moral disrespect,” and “risk of harm.” For the purposes of analysis, I group these notions under three broader labels: (a) the objective meaning or expressive harm account, (b) the respect-based account, and (c) the improvement-claim account. Throughout the paper, I use these labels as my own organizing framework. These three accounts will be discussed in turn. I aim to show that none of these three approaches fully captures the moral basis of ontic injustice.

5 Conclusions

In summary, the critical examination of Katharine Jenkins’s theory of ontic injustice reveals a pivotal deficiency in its ability to address crucial inquiries regarding the moral dimensions of social kind membership. The paper challenges Jenkins’s adherence to practice positivism and advocates for a shift toward a more nuanced moderate practice idealism, emphasizing the importance of incorporating counterfactual constitutive standards. By scrutinizing Jenkins’s three key approaches (the expressive harm, respect-based, and improvement-claim account) the analysis underscores inadequacies in establishing a robust foundation for understanding the moral wrongness associated with ontic injustice. Jenkins’s assertion that ontic injustice stems from the violation of individual moral claims related to moral injury, disrespect, and hindrances to a fulfilling life falls short of providing a satisfactory framework. In response, this paper endeavors to contribute to the evolving discourse on ontic injustice by both critiquing existing theoretical frameworks and proposing a reevaluation of the conceptual underpinnings. This exploration aims not only to enrich the ongoing dialogue on the moral aspects of ontic injustice but also to stimulate further research and advancements in this critical area.