Although “you are what you eat” is a well-worn cliché, personal identity does not figure prominently in many debates about the ethics of eating interventions. These debates are typically framed as a clash between respect for autonomy and the effective promotion of healthy eating (Barnhill et al. 2014) or some other benefit. Within this framing, if identity has any ethical relevance at all, it is as a factor influencing our food choices, habits, and preferences, and little more.

But recent philosophical work has begun to theorize the relationship between identity and eating and draw attention to some of the complex ethical issues that identity raises for eating interventions. For example, Rebecca Kukla (2018) argues that through a process called interpellation, food messaging in interventions can reinforce inaccurate and harmful stereotypes about certain group identities, shaping the self-understandings and treatment of members of those groups. In a paper surveying empirical work on the link between identity and eating, Daniel Kelly and Nicolae Morar contend that this link means that “attempts to alter some eating habits will also be attempts to alter something significant in people’s identities” (Kelly and Morar 2018, 652). This possibility raises the ethical stakes of eating interventions alongside questions about the ethics of altering others’ identities.

While these contributions offer a welcome correction to the way identity has been overlooked in food ethics literature, they paint only a partial picture of the ways identity matters for eating interventions. In this paper, I explore another way that identity raises ethical issues for eating by offering an account of how “identity-policing,” a key mechanism for the social constitution and maintenance of identity, applies to eating, and tracing its ethical implications for eating interventions.

I begin with an account of the relationship between eating and identity. One way we construct identity is by holding ourselves and others to “identity-congruent” actions, including ways of eating.Footnote 1 When we fail to eat in identity-congruent ways, we are subject to “identity-policing,” punitive attitudes and punishments that motivate compliance with identity norms. While this mechanism is not necessarily bad, it can have bad effects, including the harms of the punitive attitudes and punishments themselves. I outline two ways in which eating interventions can subject eaters to these harms. While they may be outweighed by the benefits of the diet intervention being promoted, these harms should nonetheless be acknowledged and accounted for. To aid in these evaluations, I discuss several factors that may modulate the presence and severity of identity-policing and discuss strategies for developing less harmful eating interventions. In the final section of the paper, I consider how identity-policing interacts with and complicates the possibility of identity-loss through long-term diet change, one of the central identity-related ethical concerns raised by Daniel Kelly and Nicolae Morar. By offering a conceptual framework for incorporating identity-policing’s effects into ethical evaluations, and tools to support the development of less harmful eating interventions, this paper contributes to the centring of identity in conversations about food ethics and to a more comprehensive picture of identity’s ethical importance for eating interventions.

Identity and Eating

“You are what you eat” may be an anglicized mistranslation of Jean Anthelme Brillat-Savarin’s francophone imperative to “tell me what you eat and I will tell you who you are” (cited in Division of Rare and Manuscript Collections, Cornell University Library 2002). Taken together, though, these phrases indicate a reciprocal relationship between eating and identity: what you eat might, in part, make you who you are, but what you eat also reflects or expresses who you are. Such a relationship is also implied by contemporary theories of identity and agency, like the one presented by Hilde Lindemann (2014; 2001).

According to Lindemann, identity—who we take ourselves to be and who others take us to be—is normative, or action-guiding; it informs how we act. In their recent paper on the moral psychology of identity and eating, Daniel Kelly and Nicolae Morar explain that this normative aspect of identity can be understood through the concepts of “social role” and “social norm” (Kelly and Morar 2018, 640). A social role is an identity category like mother, Latina, nerd, or athlete. Many social roles are conferred on us without us ever seeking them out, like familial roles (son, sibling), racial or ethnic designations, and, often, gender roles (Lindemann 2014). Other social roles we may actively pursue, like being a teacher or a soccer player, or aspire to have conferred on us, like being popular. Whether chosen by us or conferred on us by others (or something in between), each social role comes with social norms—expectations to behave in ways appropriate to that role. These identity-related social norms can be more or less explicit; they may be encoded in written rules or laws, or remain unspoken expectations (Kelly and Morar 2018, 643). These norms may change over time, and as individuals, we can tweak or negotiate our relationship to them, but, as a rule, they are never fully within any individual’s control.

Eating norms are a subset of social norms. So, social roles can come with a set of expectations about what, how much, where, with whom, and how people inhabiting those roles will eat.Footnote 2 For example, “real men” eat meat (Adams 2015; Rothgerber 2013; Ruby and Heine 2011), while “foodies” take food “seriously” (Solier 2013). In certain families, being a “member of this family” comes with the expectation that you clean your plate. Even fairly thin or temporary social roles like “guest” come with eating norms; a lively comment section on a blog post originally titled “The Most Difficult Dinner Guest Ever and Five Delicious Meals to Feed Them” reveals expectations that guests should eat what is offered, and not make a fuss about dietary preferences or constraints (allergies excepted—at least sometimes) (Durand 2012).

Like other social norms, individuals internalize identity-related eating norms through various psychological processes. Once internalized, these norms inform our actions, intuitions, and judgments in non-conscious, emotionally-laden, and immediate ways (Kelly and Morar 2018, 643–644). We do not have to consciously think, “I am a foodie and foodies drink nitro kombucha so therefore I drink it,” though we may sometimes reason that way. Instead, it simply feels natural and right to eat in identity-congruent ways, and unnatural, transgressive, and wrong to do otherwise. The affective pull of internalized norms is strong, and it can be difficult to detach oneself from affective relation to these norms, even if, upon reflection, we intentionally and forcefully disavow them.

Kelly and Morar suggest that when we internalize norms, we acquire “intrinsic motivation” to comply with the norm and to “punish, or at least direct punitive attitudes towards, those who violate it” (Kelly and Morar 2018, 643). We may shame, scold, call out, or direct disgust, disdain, or revulsion toward those who violate identity-related expectations. We may also shun, exclude, or otherwise punish them. This array of punitive attitudes and punishments are forms of what I call identity-policing: affects and behaviours which work to motivate and secure compliance with identity norms.

Identity-policing is not limited to norms that apply to ourselves and others like us; we also police norms we take to apply to others and not to ourselves. As Sandra Bartky points out in her analysis of weight-loss dieting as “feminine” eating, it is not only women who police women’s eating—it is everyone and anyone (Bartky 1990). And, in addition to policing others for their norm violations, we also self-police. We may feel shame, guilt, and self-disgust at our own norm transgressions, or even the mere thought of them (Luo 2004). These affects work to keep us in line with identity-norms, even when no one else is around to hold us to them.

By holding ourselves and each other to eating norms through identity-policing, we motivate eaters to comply with those norms and reinforce identities. As Lindemann would put it, I hold you in your identity by expecting you to eat in ways that are in line with that identity and by punishing you or directing disapprobation toward you when you fail to do so. This is because eating not only reflects who you are, but also makes you who you are, in some sense. As described above, my identity may inform how I eat, but how I eat also informs and constrains who I take myself to be and who others take me to be. The link between identities and expected behaviour means that some identities are conferred, or de-conferred, on the basis of our actions (Lindemann Nelson 2001, 51), including our eating. For some identities, if you do not eat in identity-congruent ways, you cannot hold or maintain that identity. By holding people to identity-related eating norms, then, we can hold people in those identities; we can help them construct and maintain those identities.

In this sense, identity-policing is not necessarily a bad thing. Holding ourselves and each other to behaviours we endorse and identities we value is an indispensable function of loved ones and communities. But identity-policing can also be used to reinforce harmful behaviours, like disordered eating or extreme dieting, and to hold people in oppressive identities, like hegemonic masculinity. And as I suggest in the next section, the punitive attitudes and punishments that constitute identity-policing may be harmful in themselves, whether the identities or behaviours they encourage are good or bad.

Intervening on Eating Through Identity-Policing

Though a full accounting of why we eat as we do would require consideration of physiological, environmental, structural, and other factors, identity is clearly a powerful influence on the way we eat. This is not just because of the “intrinsic motivation” provided by internalized norms, but because of the ways we police eating.

Recent studies of people with coeliac disease provide a striking example of the motivational power of identity-policing. When faced with identity-policing—or even the anticipated possibility of it—some participants in the studies ate glutinous or potentially contaminated foods, even though they knew that this could cause significant physical damage and have painful and uncomfortable physiological side effects (Schroeder and Mowen 2014, 466). The participants reported that they did so to appear “normal,” to avoid “embarrassment,” social exclusion, or being “judged” by family and friends (Schroeder and Mowen 2014, 466–467). Participants reported being subject to ridicule, derision, criticism, and disbelief from others for refusing or avoiding potentially contaminated foods or making special food requests when eating out, even from family members and friends who were generally supportive and understanding (Schroeder and Mowen 2014, 466; Olsson et al. 2009). In short, the participants with coeliac did not want to eat in a way that would lead others to police them for transgressing what they took to be “normal” eating norms and were concerned that others would confer devalued identities on them (i.e., “abnormal,” odd, anti-social, rude) if they followed their medically necessary diet.

Some public health practitioners and food activists have tried to harness the motivational power of identity to change people’s eating. One way people have tried to do this is by directly associating “bad” foods or ways of eating with identities that the target population devalues, rejects, or otherwise does not want to be associated with. In other words, they attempt to cultivate the kind of situation that can motivate coeliacs to eat gluten, but with the aim of promoting beneficial rather than harmful eating. For example, Jonah Berger and Lindsay Rand recommend that public health practitioners reduce “risky” eating (i.e., unhealthy eating) in a target population by associating that eating with undesirable identities. The authors tested the idea through a study that found that undergraduates were less likely to choose unhealthy food after exposure to a campaign associating eating “junk food” with nerdy grad students (Berger and Rand 2008).

Other strategies associate foods and ways of eating with unwanted identities through the use of identity-policing. For example, Chloë Taylor (2010) expresses concern that arguments in favour of vegetarianism fail to motivate people to give up meat because of the ways gender, ethnic, and racial identities are linked to meat eating. It is not only that meat eating is positively associated with certain identities, like masculinity, but that eating vegetarian food is associated with the (ostensibly) incompatible and socially devalued identity of femininity, as well as whiteness, which means that eating vegetarian may show up as a rejection of other racial and ethnic identities. To overcome these barriers, Taylor suggests that vegetarian activists deploy campaigns to cultivate disgust toward the consumption of animal flesh. This disgust is not just a way of making eating meat undesirable, but, I suggest, is itself a form of identity-policing that effectively associates eating meat with a subhuman identity.

Drawing on research from Paul Rozin and others, Kelly and Morar argue that disgust is one of the “most prevalent” emotions used to police food norms (2018, 645–647). They also contend that disgust is intrinsically dehumanizing; in other words, when a food or way of eating is deemed disgusting, it is associated with being subhuman (see also Lupton 2015). A disgusting food is not something a human should or would eat, so eating disgusting foods violates the eating norms associated with human identity. While “human identity” might seem rather vague, it is widely understood as the grounding for moral standing, a basis for dignity and respect. Taylor’s strategy, then, associates eating animals with subhuman identity, and so places the humanity and moral standing of animal eaters in question. The hope seems to be to tap into an identity that is more fundamental than the gender and racial identities acting as barriers to vegetarianism and, in so doing, overcome those barriers to change the way people eat.

Certain “anti-obesity” campaigns deploy a similar strategy, cultivating disgust to deter people from eating innutritious foods. Consider, for example, the 2009 New York City Department of Health and Human Hygiene campaign to reduce soft drink intake. The campaign’s print ad “depicts globs of human fat gushing from a soda bottle” (Chan 2009), while an accompanying video shows a man drinking a tall glass full of rendered fat. The effect is grotesque. The advertisement not only makes soda itself look disgusting, but links soda drinking with being disgusting, and therefore, subhuman (Lupton 2015).

These intervention strategies attempt to directly or indirectly associate “bad” ways of eating with unwanted identities and to encourage people to police themselves and others accordingly. They may create new associations where none exist or tweak or amplify extant associations. Those deploying these strategies believe they are promoting beneficial ways of eating, so the identity-policing they cultivate may not be harmful on that account (unlike the coeliac case). But these interventions may have other harmful effects.

A common critique of anti-obesity interventions like the one mentioned above is that the interventions do not just strategically threaten people with dehumanization to encourage them to eat a certain way but actually result in the dehumanization of certain groups (Kukla 2018; Kelly and Morar 2018; Lupton 2015; see also Abu-Odeh 2014). This is not to say that the people in question actually lose their humanity (an impossibility if you think this is an essential or inherent trait), but that they are recognized and treated as less than human. In deontological terms, this is morally unacceptable because it is a violation of essential human dignity. As Kelly and Morar point out, those who are perceived as subhuman may also be treated terribly, so this conferral not only violates human dignity but enables other harms (2018, 650).

Importantly, if associations like the one promoted by the NYC campaign are successful—that is, soda drinking or other “junk food” consumption becomes disgusting—then not only those who do but also those who are assumed to consume those foods may be dehumanized. Fatness is widely assumed to be caused by bad eating habits (Reiheld 2015).Footnote 3 Indeed, this is an essential assumption behind the anti-soda campaign; otherwise, it would not make sense as an anti-obesity campaign. Regardless of their actual eating and drinking habits, fat people may be conferred the identity of “soda drinker,” perhaps as part of a more general conferral of “unhealthy eater,” and may thereby be subject to dehumanization and its harms. In this way, such interventions may contribute to the widespread stigmatization of fat people, which is associated with a variety of significant physical, psychological, social, and material harms (Pearl 2018; Vartanian and Smyth 2013).

Similar concerns apply to Taylor’s vegetarian activist strategy. If successful, Taylor’s strategy might not only lead to the dehumanization of those who eat meat, but also those who are assumed to eat meat. Identities associated with meat eating are varied but include groups that are already often dehumanized and otherwise devalued in white supremacist contexts like the United States, such as Black and indigenous people (Bailey 2007; Harper 2010; Williams-Forson 2006). In this way, Taylor’s strategy might contribute to and reinforce forms of racist oppression.

Using disgust to police people’s eating is an extreme case of how identity-policing can be harmful. But other forms of identity-policing—using different attitudes and behaviours and policing different identities—can also have ethically significant effects. Importantly, the policing itself can be harmful or lead to harm, regardless of the identity being policed. For example, the coeliac studies mentioned earlier report that even when policing did not lead to eating harmful foods, it produced social friction, emotional distress, and—due to avoidance of situations in which policing could occur—material losses, because business is often done over lunch (Schroeder and Mowen 2014, 466; Olsson et al. 2009).

Harmful outcomes seem possible in any situation where identity-policing of eating is at play. Being shamed, mocked, or embarrassed for what one eats—especially when it throws one’s identity into question, even if that identity is just “cool undergrad”—can cause significant emotional distress. Being policed by others can strain relationships, including important familial, friend, or work relationships, in part by leading eaters to avoid situations where food will be consumed or discussed (Kukla 2018). And if the policing, or threat of policing, occurs regularly, repeated emotional and psychological distress could lead to the sorts of mental, physical, and social harms associated with stigmatization, which I will discuss in the next section.

I do not mean to overstate the harm done by teasing or dirty looks, or to suggest that the effects of policing are always properly characterized as harms. The existence and extent of the harm caused by identity-policing depends on various factors, as I discuss later. What I am suggesting is that the possibility of harm is present when identity-policing is at play, even when the identities being policed are not harmful themselves and the diets being promoted are beneficial. Ethically speaking, the benefits of any eating intervention must outweigh the harms, but this calculation will not be accurate if all the harms are not taken into account. Those who deploy strategies that intentionally tap into identity and encourage identity-policing in the hopes of altering eating should therefore acknowledge and account for the possibility of harms caused by identity-policing. But, as I explain in the next section, even if those trying to change others’ eating do not intentionally use identity-policing, they may trigger it and its harms nonetheless.

Inadvertent Identity-Policing and Eating Interventions

Even when interveners are not deliberately intervening on identities to change people’s eating, they still need to take the possibility of harmful identity-policing into account. This is because we are always already embedded in webs of identity-related eating expectations. We risk violating these expectations whenever we eat, or do not eat, as the coeliac case illustrates. Interventions that encourage diets requiring violations of identity-related eating norms may subject eaters to identity-policing, and this may harm them, even if the intervener did not intend for the diet to have any identity-related effects.

An example of this phenomenon can be found in a 2011 study of teens in the United Kingdom. Martine Stead and colleagues found that at school, healthy eating was taken to reflect being “geeky” or “nerdy,” and eating off-brand food taken to indicate one was “poor and cheap.” Being “cool” required eating brand-name junk foods for lunch (2011, 1135). Violating this norm triggered teasing, bullying, marginalization, and stigmatization by the eater’s peers. And although it is well-recognized that not everyone in high school is cool or well-off, popularity norms are in force for almost everyone within this social space, even those not identified (first- or third-personally) as popular. In this context, being “popular” is an identity nearly everyone is expected to aspire to, as being “normal” might be in other contexts.

Being the subject of bullying has been found to lead to significant mental health problems, self-harm, and suicidal ideation (Fisher et al. 2012; Arseneault, Bowes, and Shakoor 2010) and can have long-lasting effects on relationships, economic well-being, and perceived quality of life (Takizawa, Maughan, and Arseneault 2014). Marginalization is a process that excludes or “peripheralizes” individuals or groups from “dominant, central experiences” in ways that leave those marginalized at “increased safety, health, social, and political risk” (Hall and Carlson 2016, 202). And stigmatization can lead to “strained and uncomfortable social interactions, more constricted social networks, a compromised quality of life, low self-esteem, depressive symptoms” and poor material outcomes like loss of employment opportunities or income (Link and Phelan 2006, 528). Given the possibility of these significant and long-term outcomes, Stead and colleagues suggest that healthy eating might be bad for these teens’ health. As a result, it might be unethical, not to mention counterproductive, for public health practitioners to encourage these students to eat healthier at school.

Stead and colleagues’ study details some of the harms that can result from policing eating and shows that healthy eating can violate identity norms. While the eating norms associated with “being cool” vary from context to context (and, thankfully, high school is not forever), healthy eating may also be associated with femininity (Gough 2007; Reiheld 2015), middle-class whiteness (Oyserman, Smith, and Elmore 2014, 212), “otherness” (O’Neill, Rebane, and Lester 2004), and rejection of family and community (Mulvaney-Day and Womack 2009). Healthy eating may therefore violate gender, ethnic, race, or class identities, or identity as a member of a community or family.

The diets promoted by food activists may violate identity norms in a similar way. As mentioned earlier, eating vegetarian may constitute a violation of hegemonic masculinity norms, and, because of its association with whiteness, eating norms linked to other races and ethnicities. These associations are more than just a practical barrier to getting people to eat in the desired way. In these contexts, interventions may make their target populations vulnerable to harmful identity-policing by pushing them to eat beyond the bounds of what others will hold them to. Those evaluating or planning these interventions should take this into account.

Minimizing Harm from Identity-Policing

By using a strategy or promoting an intervention that might harm those affected by it, those who intervene on others’ eating may violate the bioethical principle of non-maleficence. But non-maleficence can be overridden by other moral principles or considerations. It is possible that the harms attending identity-policing could be outweighed by the benefits of the intervention itself, and the use or inadvertent triggering of identity-policing could be ethically justifiable. For example, in addition to the health benefits of a nutritious diet, eating healthily is highly lauded in many parts of contemporary society. Perhaps the social approbation that comes with eating healthily would outweigh any negative consequences of identity-policing caused by a healthy eating intervention.

Given the significant moral harms associated with dehumanization, I agree with critics who argue that interventions that cultivate disgust are ethically unjustifiable (Kelly and Morar 2018; Kukla 2018). Otherwise, whether the benefits of an intervention that uses or causes identity-policing will outweigh the harms caused by that identity-policing cannot be decided in advance. This needs to be determined with attention to the context and circumstances of the situation on a case by case basis. However, there are some considerations that may modulate the presence and extent of harms caused by identity-policing and which can help people evaluate and plan ethical interventions.

First, whether or not a given intervention will trigger identity-policing will depend on the eating norms associated with the identity in question. Some identities have highly flexible or permissive eating norms, and those with such identities may change their eating practices to a great degree without risk of violation (Bisogni et al. 2002, 134). Other identities may have narrow and inflexible eating norms. In these cases, what is central is the relationship between the content of the diet being promoted and the identity’s eating norms; does the diet cross the boundaries or not?

Second, whether or not identity-policing is triggered may also depend on the salience of the relevant identity in the eating context. Some identities may have nearly global salience, like gender or race, while other identities are more salient in some contexts than in others (Oyserman, Smith, and Elmore 2014). Even if healthy eating violates an individual’s identity-related eating norms, if she is eating in a context where that identity is not salient, harms resulting from the violation may be minimal. In a school cafeteria, social status amongst peers is extremely salient and so eating healthily is highly risky, but it may be less so at home with one’s family.

Third, the extent of harm from identity-policing depends, in part, on who is doing the identity-policing. The costs of stigmatization, for instance, can depend on the relative power and influence of the group doing the stigmatizing. While high school peers may not have the capacity to, say, prevent the employment of those they consider to be uncool, they can certainly contribute to significant social, psychological, and emotional harms. It is important to note, though, that since we internalize norms and police ourselves, eating alone or away from others belonging to the relevant identity group does not guarantee that certain harms, especially emotional and psychological ones, will be avoided.

Fourth, the extent of harm may also depend on the combination of identities held by the individuals in question. Every individual has multiple identities (as I have been using the term): for instance, we each have a race, gender, class, sexuality, relationship status, social status within different communities, and so on. Some individuals have multiple identities that ground or enable significant social, emotional, and material support, so that if they are ostracized from one group, they can turn to another for resources. However, others may be very reliant on one identity for their material, social, and emotional well-being (Bisogni et al. 2002, 136). As a result, violating the norms relating to that identity would pose a greater risk.

Consider, for example, Leslie Kaufman and Adam Karpati’s study of a Latinx community in Bushwick, Brooklyn. The authors found that sharing food amongst family, friends, and neighbors was a way of expressing identity with the community and to show love and care for others. Often the food offered to a family by, say, a grandmother, would be innutritious, but Kaufman and Karpati argue that to reject the offered food would not only be “socially unacceptable” but, given the context of food insecurity in that community, “unrealistic” (Kaufman and Karpati 2007, 2184). In that case, there would be real material costs to violating eating norms: the family might not eat. On the other hand, there may also be individuals whose identities are so secure (perhaps because of engagement in other identity-confirming practices) that violating identity-related eating norms would not be enough to provoke punishments or punitive attitudes from others, or the eaters may be impervious to any punishments or punitive attitudes that came their way (Stead et al. 2011, 1135).

A final, and related, factor is the relative importance of an identity to an individual. Norah Mulvaney Day and Catherine Womack (2009) suggest that the effect of an identity on our behaviour is not simply determined by the fact that we hold that identity or that others take it to hold; it is also determined by how we value that identity. The more deeply we care about an identity or the more central it is to our self-understanding and way of being in the world, the more harm that might come from violating related norms. Those identities that are central to one’s self-understanding, one’s social, emotional, and material well-being, and one’s experiences in ways that may be opaque to the individual—like some racial identities, particularly whiteness (Alcoff 2005)—may also make for “high risk” violations, compared to identities that are less fundamental to oneself.

These considerations suggest three guidelines for avoiding or minimizing harms caused by identity-policing. First, interventions should promote dietary changes that work within their target populations’ identity boundaries. If we are trying to promote a healthier diet, we could try to get men who have internalized “real men eat meat” (or who are surrounded by others who have) to eat less red meat rather than not eat red meat at all. Second, when desired changes do violate an individual’s identity norms, interventions should encourage people to make those changes in contexts where the relevant identity is less salient, or they are less likely to be policed by others. For example, high schoolers might be encouraged to eat more fruits and veggies at breakfast or when they eat dinner at home. If there are good reasons to target school food, and there may be (perhaps that is the only location where interveners can influence eating), then this higher risk of harm will need to be accounted for. Third, interveners should be especially cautious when recommending changes that violate identities that are central to individuals’ social, emotional, or material well-being. The risk of harm is high here.

Implementing these guidelines will require that those planning and evaluating interventions be knowledgeable about the target population’s identities and related eating norms, so gathering this information will be key. Resources from food studies and related fields can be helpful here, though specific contextual knowledge will be indispensable—a point in favour of developing interventions alongside and with members of the target population.

It is important to recognize that by avoiding or minimizing harms from identity-policing we are at the same time protecting identities that people already hold. But some identities may be ethically problematic, including identities that are central to an individual’s sense of self and their emotional, social, material well-being. For instance, hegemonic masculinity not only perpetuates sexist, heterosexist, and patriarchal systems of oppression but also has measurably bad effects on men’s health (Courtenay 2000). Encouraging men to eat in ways that violate masculine eating norms might therefore seem like a good thing—not only because they would be eating healthier, but because violating these norms may undermine and corrode this harmful form of masculinity. It may also seem like a good idea to try to change the content of the eating norms associated with hegemonic masculinity; we can look at “plant-based” activist Rip Esselstyn’s insistence that “real men” eat plants (Trachtenberg 2009) as an example of such an attempt.

Such cases raise questions about when and under what circumstances it is ethically appropriate for public health practitioners or food activists to meddle with the content of an identity. I will discuss this further in a moment, but I want to emphasize that even if we have good reasons to reject or undermine the content of certain identities, we should not discount the harms that violating this identity through eating may cause. Even when punitive attitudes or punishments are motivated by policing of a problematic identity, the harm they can do is real and should be incorporated into our ethical evaluations.

Diet Change, Identity-Loss, and Identity-Policing

As mentioned earlier, one concern raised by Kelly and Morar is that the link between eating and identity means that “attempts to alter some eating habits will also be attempts to alter something significant in people’s identities” (Kelly and Morar 2018, 652). This concern could be about one of two things: altering eating habits could change the content of an identity, that is, change the eating norms linked to the identity (as “real men eat veg” tries to do). Or, it could mean that changing an individual’s eating habits could alter that individual’s identities. Consider, for instance, Taylor’s suggestion that many people do not give up eating meat because doing so might change who they are as an individual.

This second possibility brings us back to the reciprocal relationship between identity and eating. Recall that identity is partially constituted by how others identify you, and eating in certain ways can lead others to confer certain identities on you. Consistent, long-term failure to eat in identity-congruent ways may lead others to de-confer the related identity on you, making it difficult if not impossible to maintain that identity. A vegetarian who eats meat regularly is not a vegetarian; a foodie who no longer takes food “seriously” is no longer a foodie. If you do not eat in the ways associated with an identity, you are no longer recognized by others as having that identity. Since most interventions aim at long-term change—public health practitioners do not want kids to eat just one healthy lunch, but to have a healthier diet overall, and most vegetarian activists want people to give up meat altogether, not just to eat one veggie wrap or go meatless only on Mondays—this loss of identity may be a real possibility.

Considering this possibility in relation to identity-policing sheds some new light on its ethical implications. Given the potential harmfulness of identity-policing, we might think that identity-loss is a good thing; in many cases, if others do not recognize you as having an identity, they will not expect you to conform to related eating norms. They will therefore have no reason to police you in relation to that identity, and we need not worry about those sorts of harms. Admittedly, there may be some harm during the transition process (the period during which it is unclear whether the identity holds or not), but once others stop attributing that identity to you, you are in the clear.

That said, just because others no longer attribute a certain identity to you does not mean that you do not attribute that identity to yourself or hold yourself to related norms. A lapsed vegetarian taken by everyone else to be an omnivore may continue to feel ashamed, guilty, or disgusted at his own norm violations even if no one else takes those norms to apply to him. And even if you explicitly disavow a norm or the related identity, it can be hard to shake off your affective relationship to norms. Identity-loss in this sense is therefore not always, or obviously, a way to avoid the harms of identity-policing.Footnote 4

But say that someone did manage to fully detach herself from an identity and no one else took that identity to hold, even aspirationally. It does seem like this would sidestep the possible harms of identity-policing, at least with regard to that particular identity. Nonetheless, there are reasons to be cautious about precipitating this outcome. We may have some identities that we are not particularly attached to and which play only minor roles in our lives. But, as noted earlier, certain identities are central not only to our self-understanding, but our emotional, psychological, social, and material well-being. To lose such an identity could upend a life; it would require a significant reshaping of relationships, practices, and self-understandings. This is not to say that losing an identity is necessarily a bad thing. Our identities can and do change over time, and as with the case of hegemonic masculinity, it may be that we would be better off leaving certain identities behind. But identities are important grounding features of individuals’ lives, and we must acknowledge that intervening on someone’s eating in a way that causes them to lose an identity is ethically significant and should be given serious consideration and weight.

This concern about loss of identity may not be very compelling to some. An objector could point out that on the account of identity given here, eating that is considered a failure vis-à-vis a given identity may be expressive of a different identity. The failure to be a popular kid is “success” at being a dork or a loser, for instance. If someone has lost or “failed” at an identity because of public health interventions on their eating, presumably they are now conforming to the norms of a “health-conscious” identity. Work by John Coveney (2006) and Gyorgy Scrinis (2013) suggests that many people in contemporary Western societies already hold something like this identity, to greater and lesser degrees of salience and first-personal importance. This identity is associated with many positive attributes, such as being a good citizen, a responsible person, and a rational agent (Biltekoff 2013; Crawford 1994; Coveney 2006). Individuals who lose an identity because of public health interventions might embrace and centre a health-conscious identity. Being a health-conscious person has social cachet in many circles, and there are social groups and communities (virtual and otherwise) that could provide social and emotional support based on such an identity. Perhaps these could fill in any gaps left by the lost identity.

Although this may be a possibility for some individuals, the people targeted by eating interventions already have identities and are often embedded in communities that are built up around and reinforce those identities. Losing such identities is significant, and even if in some cases this loss may be mitigated or compensated for through the development of a new identity, this may not always be the case. Importantly for public health interventions, a health-centred identity is itself one associated with middle-class white people (Oyserman, Smith, and Elmore 2014), and women in particular (Reiheld 2015). Thus, embracing and centring an identity as a health-conscious person is a more viable alternative for some individuals than others. Similar concerns obtain with regard to food activist identities (Guthman 2008).

Even if we do agree that causing someone to lose an identity is ethically significant and should not be taken lightly, an objector might downplay the likelihood of identity-loss through diet change. Sarah Conly offers a version of this objection: “having a culture, or a personal identity, doesn’t depend on continuing to ingest what you’ve typically eaten in the past and thought well of, or on eating what lots of people in your culture habitually eat, either generally or on specific occasions” (2018, 464). Few identities are entirely constituted of eating practices; therefore the risk of losing an identity through intervention-led eating change is minimal. A vegetarian who shifts to eating meat will lose her identity as vegetarian, but surely a Latina whose identity as such is central to her life would not lose that identity by continually refusing to eat culturally traditional meals (say, because she has gone vegetarian). Identities that are often conferred “involuntarily,” like racial, ethnic, or gender identities, might be particularly resilient in this respect. If the way we act is as important to identity as Lindemann’s account suggests, perhaps a person could violate eating norms but double down on other identity-related practices to secure their identity. In this context, it seems like a total loss of an identity due to change in eating practices would be rare.

It is important to recognize that for some people, identity as vegetarian, foodie, or other similarly food-centred identity that could plausibly be lost through repeated eating norm violations does play a central role in their lives and the loss of such identities should not be minimized. Conly is right to suggest that the analysis with respect to identities that are not wholly centred on food is more complicated. It may be true that a Latina could not fully lose her identity as such by consistently violating related eating norms, or that someone would no longer be “part of this family” even if they refused to “clean their plate” ever again. But even if diet change alone is insufficient to cause the loss of these identities, it is worth bearing in mind that identity could be at risk if the eater no longer engages in the same activities or holds the same relationships due to social friction, avoidance, or exclusion spurred on by changes in diet and associated identity-policing.

Moreover, in cases where identity-loss is not a real possibility and the eater’s identity remains in place, concerns about identity-policing still hold. Long-term, consistent failure to conform to identity-congruent eating norms could create a great deal of social friction, tension, and could seriously harm relationships. Individuals in this situation may be subjected to repeated punishment and punitive attitudes from themselves and/or others, and, as with all harms caused by identity-policing, those evaluating and developing eating interventions should take these harms into account.Footnote 5

Conclusion

Though identity has not played a central role in debates about the ethics of eating interventions, the relationship between eating and identity raises various ethical issues for those trying to change others’ eating. This paper aims to contribute to a fuller theoretical account of the relationship between identity and eating and its ethical implications for eating interventions through a discussion of identity-policing. I have suggested that while identity-policing can be a mechanism for holding ourselves and others in identities that we value and endorse, the ways we police identity by policing eating can also harm eaters. I have argued that eating interventions can subject people to these harms in at least two ways: through the strategic use of identity-policing to change eating, or by encouraging diets that lead eaters to violate eating norms that others will hold them to. With the exception of the use of disgust to police eating, the potential harms from identity-policing should be weighed along with other harms against the benefits of an intervention to determine whether the intervention is ethical. I have also discussed identity-loss as a possible and ethically significant outcome of eating interventions and highlighted some of the ways identity-policing and identity-loss interact. By demonstrating the moral relevance of identity-policing for eating and underlining the need to think carefully about the ethical implications of identity for eating interventions, this paper contributes to more ethically comprehensive debates and discussions within food ethics and about eating interventions in particular.