Defensiveness and Identity

Abstract

Criticism can sometimes provoke defensive reactions, particularly when it implicates identities people hold dear. For instance, feminists told they are upholding rape culture might become angry or upset because the criticism conflicts with an identity that is important to them. These kinds of defensive reactions are a primary focus of this paper. What is it to be defensive in this way, and why do some kinds of criticism or implied criticism tend to provoke this kind of response? What are the connections between defensiveness, identity, and active ignorance? What are the social, political, and epistemic consequences of the tendency to defensiveness? Are there ways to improve the situation?

1. On Defensiveness

You're not perfect. This, we hope, is not surprising or threatening news. We're not perfect either: we are not perfectly virtuous or rational or perfect academics or perfect anti-oppression activists. And we hope you will not take it personally if we say: and neither are you. This is technically a criticism, but it is not a very threatening one. Pointing out imperfection in the abstract typically feels more like a banal truism than an attack.

But sometimes criticism provokes quite different responses—especially when it implicates identities people hold dear. For example, most people at least nominally hold antiracist ideals. If you tell someone that what they just did or said is racist, you are likely to provoke a defensive reaction. They might become angry or upset and refuse to keep talking to you. Such defensive reactions are the primary focus of this paper.

In this paper, we will articulate a functionalist characterization of defensiveness and illustrate harmful ways in which defensiveness can shut down important inquiries. We connect this characterization to the idea of ‘active ignorance’ and discuss why some kinds of criticism or implied criticism tend to provoke this kind of response. The dynamic is particularly interesting in cases where people with anti-oppressive identities fail to live up to their own ideals. Our central case studies involve ‘gender-critical’ feminism and rape culture; to help us analyze them, we will draw on bell hooks's (1984) discussion of defensiveness and White feminism. We aim to illuminate the social, political, and epistemic consequences of defensiveness, and we will give some suggestions as to how, depending on one's social position, one might improve the situation.

We identify defensiveness with the tendency to employ defensive strategies. We will give some examples of such strategies below, but the characteristic aim of a defensive strategy is to protect someone's own views or self-conception from perceived or genuine attack.

Here are two important clarifications about our treatment. First, this is a functionalist, as opposed to an affective, approach to defensiveness. We emphasize what one does in a conversation, not how one feels about it. There are, we think, characteristic feelings of defensiveness that may play significant roles in moral psychology. But our project here is more a project in social epistemology: we wish to explore the epistemic effects of defensiveness on conversation and social understanding. The functionalist characterization is the more useful one for this purpose. We will not set emotional responses aside entirely, but we talk about them primarily as motivators of defensive responses rather than as characterizing defensiveness in the first place.

Second, our characterization of defensiveness is normatively neutral. One is defensive to the degree to which one tends to employ defensive strategies. While we think defensiveness is often the result of cognitive dissonance, this is not always the case. Sometimes, for instance, defensiveness is a reaction to accurately perceived unfair attacks. In such cases, it may well be all-things-considered best to take a defensive attitude. Protecting oneself is often appropriate! Yet, even when defensiveness is justified, we will argue, it carries significant social costs.

Our interest in defensiveness is primarily about some of the mechanics and motivations for its production and the harms that it can create, especially when unwarranted. An interesting project in normative ethics—not ours—would be to demarcate the conditions under which defensiveness is justified. Our contribution is a deeper understanding of the harms of defensiveness and some strategies for ameliorating them.

2. Defensive Strategies

Defensiveness contrasts with openness. When one is open, one is looking to learn; when one is defensive, one is looking to reject. One way to reject an idea is to articulate arguments against it. But the suite of defensive strategies is much more diverse.

Defensive strategies are often performative. We do not mean that they are insincere—just that their expression has strategic use by centering the interests and feelings of the person being criticized or possibly by being used to forestall additional criticism. This is consistent with thinking of such responses as reactive attitudes, following P. F. Strawson (2008: 7–8), as the person who is being defensive is responding to a perceived moral injury. But these issues are gendered and racialized in ways that prompt careful contextual treatments of individual cases (see Jaggar 1989; Maclachlan 2010; Hamad 2019; Táíwò 2020 for much more on these aspects). In the face of such anticipated responses, potential critics might engage in what Kristie Dotson (2014) calls testimonial smothering. They might even be (or fear being) preemptively attacked, for instance, by being accused of hypocrisy.

Here is an example illustrating what we take to be a familiar dynamic, recounted in Choi (2019). In February 2019, at a US House of Representatives committee meeting, Republican Congressman Mark Meadows used a Black woman as a literal prop in an argument against accusations of racism against then-President Donald Trump. He asked Lynne Patton, a federal employee, to stand up. He did not invite her to speak, instead merely gesturing at her visible Blackness, and spoke on her behalf (denying the possibility of Trump's racism) as she stood silently behind him.

Democratic Representative Rashida Tlaib criticized Meadows's action, saying (rightly) that it ‘in itself it is a racist act’ (Choi 2019). But her criticism provoked an extreme defensive reaction from Meadows. He complained to the Black Democratic chair of the committee, Elijah Cummings, about her harsh words; Cummings focused his energy on placating Meadows: ‘I can see and I feel your pain, and I don't think Ms. Tlaib intended to cause you that, that kind of pain’ (Choi 2019). Tlaib eventually clarified that she did not intend to call Meadows personally racist and apologized for upsetting him. Meadows's defensive strategies ended up having a dramatic and disruptive effect on that discourse.

This example illustrates several common defensive strategies: it involves a performative show of anger and woundedness, and reorients subsequent discourse around Meadows's feelings and standing. These strategies fall neatly under Jennifer Freyd's concept of ‘DARVO’. The acronym stands for ‘Deny, Attack, and Reverse Victim and Offender’ (see Freyd 1997; Harsey, Zurbriggen, and Freyd 2017). It describes a characteristic way in which a subject of criticism may lash out, casting those advancing criticisms as the ones doing wrong.

Meadows's example also involves another common defensive strategy in the form of catastrophizing exaggeration. One might hear a complaint about one's behavior and interpret it as a statement about one's character—as when the discourse around Meadows focused on whether he himself was racist, as opposed to whether his action was a racist one.

Due in part to the tendency of terms like ‘racist’ to provoke such defensive reactions, some — e.g., Anderson (2010: 49) — have advocated to limit their use to only particularly egregious examples. As we will explain in section 7, we do not entirely agree. As Liao and Hansen (2022) carefully argue, ‘racist’ and other such oppression-condemning expressions function linguistically as unexceptional gradable notions, perfectly capable of expressing various degrees or respects in which something is contributing to oppression.

There is an enticing inference from actions to characters: they did something racist, and so they are racist. And another equally slippery inference from there to an even more incendiary idea: they are racist, so they are a racist. When we assume that one can chain these inferences together so that doing something racist implies that one is a racist, the stakes are high; nobody wants to be painted with that monstrous brush.

Both of these inferential steps are invalid. While racist people might perform racist actions and hold racist beliefs, so might other people. Most people at least sometimes perform actions or hold beliefs that are out of step with their overall character. Generally patient and considerate people sometimes snap at others for small offenses, and people who are generally committed to being antiracist might sometimes do racist things out of carelessness or genuine ignorance. Treating criticisms of behavior as attacks on one's personal character is a common defensive reaction.

Recall that for us, defensiveness is a matter of what one does, not of how one feels. Responding to these criticisms in this way is not merely a disruptive, unpleasant, and unfair instance of DARVO. It also shuts down conversations and provides obstacles to knowledge—both broader social knowledge and self-knowledge. And it often does this even when the expressions of frustration are sincere.

That said, we think there are some typical motivations for defensiveness when it comes to things like accusations of racism. Getting clearer about them can help us understand some of the practical implications of defensiveness in lived reality.

3. Active Ignorance

We think that many defensive strategies support active ignorance. To be ignorant about something is to fail to know it. One might be ignorant of something simply because one lacks evidence; you are likely ignorant about our dogs’ names, for instance, if you do not know us personally. But ignorance can also be active and motivated.

Sometimes there are things we feel like we need not to know about ourselves in order to maintain our self-concept. What this means is that sometimes having a stable self-understanding can actually require us to be ignorant, either of some facts about ourselves or some facts about the world around us. As José Medina writes,

Social agents can be ignorant in many ways. As we just saw, sometimes there is ignorance out of luxury—when one does not need to know. But sometimes there is also ignorance out of necessity—when one needs not to know. (Medina 2013: 34)

Not needing to know is ubiquitous. You might not need to know the bus fare in Victoria because you prefer to drive around Victoria—this need not be pernicious ignorance, although it may reflect your economic privilege. Or you might not need to know the bus fare in Victoria for the simple reason that you never spend time in Victoria. What you need to know—and indirectly, what you do know—will depend on your social location in a variety of ways.

Our main interest here lies not in not needing to know, but in needing not to know. Here, one's ignorance is actively maintained to protect oneself from potentially disruptive knowledge. Sometimes the maintenance of ignorance is the result of oppressive social structures like White supremacy or colonialism. Such ignorance, as Charles Mills (2007: 13) characterizes it, is ‘an ignorance that resists’, and ‘fights back’. For instance, in order to maintain the image of Canada as a peaceful settler state, rather than one founded on genocide and dispossession, a great deal of history (and current practice) needs to be concealed (Regan 2010). Settler Canadians who had been miseducated about the truth of residential schools need not be operating in bad faith; rather, one can easily form mistaken beliefs because of the ‘social suppression of the pertinent knowledge’ (Mills 2007: 21). In such cases, counteracting these practices of active ignorance requires concerted and coordinated social action and educational endeavors, such as the work of the Truth and Reconciliation Commission (see the final report of Canada's Truth and Reconciliation Commission here: https://nctr.ca/records/reports/).

But such work often generates significant friction, in part because individuals often also contribute to the maintenance of their own ignorance. Following Medina, when we speak of ‘needing not to know’, we mean a motivation not to know in order to protect a particular interest. We do not mean to suggest that it really is best or necessary that settler Canadians not know about the genocidal history of Canada; on the contrary, we think it is important that all Canadians learn this important history. But they ‘need’ not to know it in order to preserve a particular self-concept they value. It is, after all, very uncomfortable to learn that you are complicit in oppression.

We do not say this is the only motivation for active ignorance. For example, Kinney and Bright (2021) argue that, given certain independently motivated assumptions about preference and risk, dominantly situated people may even be rational in preferring not to have further information about their privilege even if they have no interest in perpetuating oppression. But we think defensiveness is often an important part of the story.

Gaile Pohlhaus Jr.'s notion of willful hermeneutical ignorance also helps explain how ignorance is preserved. Pohlhaus focuses on concepts and other hermeneutical resources that have been developed in marginalized communities but that are actively resisted by members of dominant groups even when they are introduced to them (Pohlhaus [2012], following Fricker [2007]; see also Frost-Arnold's [2016: 6] ‘defensive ignorance’). Racism is a case in point—there is plenty of mainstream discourse about racism's complex structural power. Anyone willing to listen can easily be disabused of the cartoon picture of racism, where it is perpetuated exclusively by evil people with self-aware discriminatory intentions and beliefs. But many people find it easier to avoid the uncomfortable conversations that disabusing them of this ignorance would allow. The example of Mark Meadows given above fits this mold perfectly. Tlaib called his action racist, and, in a show of performative woundedness and anger, Meadows refocused the conversation as one about how he himself is not personally racist. (And as a result, the criticism of his racist action was practically forgotten, and he was not held responsible for it.)

There are also broader social epistemic harms. When it is hard for individuals, say, teachers, to gain insight about whether they might be perpetuating racist tropes in their classrooms, it will likewise be hard for social groups like schools to gain such insights, thereby perpetuating the ignorance in question. (And of course, in this particular example, there are downstream epistemic and social harms inflicted on students, who absorb those tropes and are subjected to a less inclusive learning environment.) There is always a complex interplay between the motivated ignorance of individuals and structural factors that suppress information.

4. Marginalization, Identity, and Defensiveness

Defensiveness will often work alongside other strategies for maintaining active ignorance. One common strategy for maintaining active ignorance is to seek out, or remain inside, an epistemic bubble from which alternative perspectives are excluded or in an echo chamber in which alternative perspectives might be heard but are consistently undermined (Nguyen 2020). Such spaces can feel ‘safe’, as serious challenges to one's active ignorance rarely reach through.

But in an increasingly interconnected world, it is hard to remain in a bubble at all times even for those motivated by active ignorance. Sometimes a challenge to one's more comfortable picture of the world penetrates. And even those in echo chambers can know when others take criticisms of their views particularly seriously. It is here that the defensive stance described in section 1 and the particular defensive strategies we discussed are among the primary ways to satisfy the need for ignorance. People resist taking ideological challenges seriously, instead treating them as to be refuted or countered.

Defensiveness in general can manifest as an epistemic vice, but there is a more specific vice illustrated in the cases we are considering. People do not necessarily exhibit epistemic virtues and vices across the board; while some people might just generally be more closed-minded or open-minded, it is also common to be closed-minded about some things and open-minded about others. And we are not necessarily good judges of our own tendencies. In what follows, we consider several cases where people's image of themselves as marginalized, open-minded, and/or progressive exacerbates their defensiveness.

Many systems of oppression are interconnected, which means that sometimes people who are somewhat knowledgeable about one form of oppression may not know as much about it as they think they do, if they do not understand the ways in which it connects to other forms of oppression. For example, White feminists may know a great deal about gendered oppression, but can still overestimate their knowledge if they have never learned how gendered oppression is experienced differently by women of color. bell hooks (1984) describes experiences that illustrate this well. She points out that criticisms by Black women would go unheard or be dismissed if they did not echo the views of the White women in the group—and worse, that Black women who spoke out would often be branded as angry or aggressive.

In addition to the racist association of Black people with irrationality and aggressiveness, part of what is going on here is identity-based defensiveness. Here is a telling anecdote from bell hooks:

In 1981, I enrolled in a graduate class on feminist theory where we were given a course reading list that had writings by white women and men, one black man, but no material by or about black, Native American Indian, Hispanic, or Asian women. When I criticized this oversight, white women directed an anger and hostility at me that was so intense I found it difficult to attend the class. When I suggested that the purpose of this collective anger was to create an atmosphere in which it would be psychologically unbearable for me to speak in class discussions or even attend class, I was told that they were not angry. I was the one who was angry.

. . . Often in situations where white feminists aggressively attacked individual black women, they saw themselves as the ones who were under attack, who were the victims. During a heated discussion with another white female student in a racially mixed women's group I had organized, I was told that she had heard how I had “wiped out” people in the feminist theory class, that she was afraid of being “wiped out” too. (hooks 1984: 13)

We note two things about these exchanges. First, they provide clear examples of people who have (correctly) taken on identities as members of oppressed social groups, and, perhaps partly for that reason, have failed to recognize their own complicity in related oppression. Second, the strategy invoked is an example of the kind of catastrophizing exaggeration we discussed in section 2. Initially, hooks criticized the syllabus for its omission of all women of color, and by the time the story got around, hooks was notorious for ‘wiping out’ her interlocutors in the seminar.

(The ‘wiped out’ vocabulary is zeitgemässig, but contemporary readers may recognize a parallel with the comically lax way in which talk of ‘cancellation’ is thrown around today. If some feminists on social media point out that someone did a microaggression, centrist and right-wing discourse may look at it and tut-tut at how everyone gets ‘cancelled’ at the drop of a hat these days. Quite often, of course, they who were ‘cancelled’ weather a few days’ criticism and get on with their careers; two years later, few will remember that they were ever ‘cancelled’. The same, we are confident, goes for many of the White women that hooks allegedly ‘wiped out’ in that 1981 seminar.)

The White feminists hooks discusses exhibit epistemic vice: their use of defensive strategies derived from a too-rigid attachment to their identity as oppressed. We do not have to think that these White feminists are epistemically vicious generally to think that they are being epistemically vicious in this particular instance. There is every reason to expect that the women enrolled in graduate courses on feminist theory in 1981 were relatively open-minded across many fields of inquiry, especially compared to their peers who did not study feminist theory.

Indeed, we are sure many of these White feminists had experienced unfair and oppressive attacks on the part of patriarchy. Challenging gendered oppression is genuinely anti-oppressive work, and such work creates backlash. Consequently, many of them would often have been in exactly the kinds of situations we described in section 1 in which defensive strategies are appropriate. They correctly recognized that antifeminists were attacking them unfairly and engaged defensive strategies in the best possible traditions of self-defense. The problem is that, from the inside, reasonable critiques and unfair attacks can feel very similar, and so very naturally tend to produce similar responses. But becoming accustomed to defensiveness can easily lead someone to overapply these strategies in cases where they are a mistake. True virtue here, as in so many other cases, lies between the two extremes, and can be a life's work to develop.

Thus, vicious epistemic character in general might not be the best way to understand the resistance to knowledge on display here. Instead, we will argue that we can understand the active ignorance of these feminists as a kind of identity maintenance, analogous to the kind of work that needs to be done to maintain an image of a settler colonial state as peaceable and legitimate. (Think of people who say ‘this is not who we are’ after the discovery of yet another colonial atrocity.) We think that this identity maintenance can contribute to the realms of epistemic errors that Medina identifies:

not just areas of epistemic neglect, but areas of an intense but negative cognitive attention, areas of epistemic hiding—experiences, perspectives, or aspects of social life that require an enormous effort to be hidden and ignored. Ignorance in these cases functions as a defense mechanism that is used to preserve privilege. (Medina 2013: 34)

As in so many cases, recognizing intersectionality is key to understanding the privilege referenced here.

Sometimes unwanted knowledge is too difficult to ignore—for instance, when a classmate or fellow activist persists in calling attention to it. In these cases, one will feel a need for some reason to dismiss the unwanted knowledge to prevent it from interfering with one's self-concept. This is exactly the kind of reasoning one of hooks's classmates describes in an apology letter from a fellow student who came to regret her behavior. The White student writes,

In class after a while I noticed myself, that I would always be the one to respond to whatever you said. And usually it was to contradict. Not that the argument was always about racism by any means. But I think the hidden logic was that if I could prove you wrong about one thing, then you might not be right about anything at all. (hooks 1984: 13)

For this student, discrediting hooks served to protect herself from hooks's critiques. Upon coming to terms with her own bad actions and epistemic limitations, hooks's classmate also found herself at odds with other White colleagues; speaking of a former close friend, she conjectures that ‘the possibility that we were not the best spokespeople for all women made her fear for her self-worth and for her Ph.D.’ (hooks 1984:13). These students’ identities as feminist thinkers and activists were clearly important to them. When those identities were threatened in class by hooks's arguments, the response was to reject the arguments by discrediting the speaker, rather than to reflect on how their conclusions might mandate a change in thinking or behavior. After all, if they accepted criticisms like hooks's about the exclusion of people of color from mainstream feminism, that would imply that they had not really been working against gendered oppression generally—only against the kinds of gendered oppression primarily faced by White women. Accepting that you are not such a great feminist after all, when being a feminist is critical to your personal and professional identity, is not an easy thing to do.

Not all identities feel so important. It might be very important to someone that they are a skilled writer, a loving sibling, and a hard worker, but fairly unimportant to them that they are a good guitar player and a resident of Burnaby, even if they believe all of those things about themselves. They might take it completely in their stride if they find out that their guitar playing is actually substandard but might become upset if they find out that their coworkers see them as lazy. (Were they to become defensive about their guitar skills, that might be a clue that guitar skills were actually more important to them than they had realized.)

Thus, it is very understandable that many of our paradigm cases of defensiveness are situations in which people face the idea that they are contributing to racism, sexism, or other forms of oppression. For many of us, it is quite important to see ourselves as basically good people. As we pointed out earlier, when someone's racist actions are called out, they often react as though they were told that they are actually a racist monster. This is only amplified when someone is also deeply invested in their identity as someone decent. This is especially true of self-styled activists, who hold anti-oppression identities close to their cores. We think the so-called gender-critical feminist movement illustrates this dynamic well.

5. Gender-Critical Feminism

People who identify as gender-critical feminists typically argue that affirming trans people's (especially trans women's) gender is a threat to women's rights. Two frequent but contested claims that they make are (1) that they are (as feminists) concerned with women's well-being and (2) that they are being silenced for speaking up for women's rights. The political aims of many gender-critical people coincide with the political aims of many right-wing groups who seek to restrict the freedom, autonomy, and inclusion of LGBTQ+ people more generally. Yet while it is sometimes difficult to separate those who support trans-exclusionary policies on ostensibly feminist grounds from those who support such policies on the basis of straightforward conservatism, it is important for our purposes to do so here. This is because of the role identity maintenance plays in the defensiveness we are considering.

Pointing out that trans-exclusionary policies and views are antifeminist is no challenge to conservative skeptics about trans rights with no affinity for feminism. But the situation is quite different for self-identified gender-critical feminists who argue for trans-exclusionary views on feminist grounds. They need not to know that their exclusionary views and their commitment to feminism are in conflict with each other, exemplifying defensive ignorance. Their reactions have clear similarities to those of the White feminists in hooks's earlier account.

People's beliefs and values need not be in perfect internal harmony. But sites of tension should merit careful consideration, not denial. For example, plenty of people are environmentalists and also meat eaters. Many of them acknowledge that it might be better for their environmentalism if they did not eat meat. Others might connect their environmentalism to sustainable farming or hunting practices. But if someone's devotion to factory-farmed meat is supposedly based on environmental grounds and it is very important to them that they are both an environmentalist and an eater of factory-farmed meat, then there might be things they would need not to know, say, about methane production and deforestation, to maintain those identities simultaneously. Defensiveness—maybe accusing their interlocutor of hypocrisy for not being vegan—seems likeliest in this last case, where it becomes a way to sustain a person's active ignorance.

We are interested in a similar kind of mechanism of active ignorance in those who argue for trans-exclusion on feminist grounds. One form that argument takes requires the background view that the rights of trans women (say, to have access to women's spaces and services) are in conflict with the rights of cis women (say, to use those spaces and services unimpeded). Many feminists have already pointed out that this is a nonproblem. For instance, there is often no practical way to maintain trans-exclusionary practices, and gendered spaces have been trans-inclusive for many years in many countries. Because you cannot tell whether someone is trans or cis just by looking at them, in order to enforce a policy requiring people to use the facilities matching their birth-assigned sex, you would need to check whether people are following the rules. Gender nonconforming women (including many cis women) are especially likely to face such challenges, which compromise their ability to use public spaces. This happens to gender-nonconforming cis women fairly often (for one high-profile case, see https://www.nytimes.com/2008/05/14/nyregion/14gender.html). Thus, someone might need defensive ignorance to maintain a gender-critical worldview as well as a general commitment to feminism. This means that, even granting the gender-critical framing assumptions, enforcement mechanisms for any such policies at least conflict with the legitimate interests of some cis women. More simply, we agree with Talia Bettcher (2018) and Hay (2020: 10) that gender-critical views are in fundamental tension with feminism.

Like White feminism, the gender-critical worldview at most supports the well-being of a narrower class of privileged women—in this case, gender-conforming cis women. As such, both versions of feminism can actively contribute to gendered oppression in some cases. While this might initially seem counterintuitive, recall that there is a structural element to ignorance as well, in which factors that shape one person's life—for instance, ways that they are oppressed by racism or transphobia—might not be apparent to someone else who is not oppressed in that way. This is how feminism that centers on the interests of White women can neglect the ways in which non-White women's oppression can differ (for much more on the issue of intersectionality and oppression, see Crenshaw 1993). Yet, when faced with the criticism that gender-critical views might in fact reinforce gendered oppression, for instance, for non-gender-conforming cis women or intersex people, defensiveness is a frequent response.

In many cases, the defensiveness of gender-critical people takes the form of claims that they are being silenced and portrays their speaking up as an act of feminist courage on their part. The writer Julie Bindel describes herself as the target of a campaign of targeted harassment by trans activists, resulting in student protests at her speaking engagements, some of her other events being canceled, and her failure to win a journalism award from the LGBTQ+ organization Stonewall UK. Nevertheless, she writes, she will continue to work as a ‘tireless, radical feminist’ (Bindel 2018). Similarly, after the philosopher Kathleen Stock was awarded an OBE (Order of the British Empire), several academics, including ourselves, wrote an open letter critical of that decision on the basis of her writing about gender (https://sites.google.com/view/trans-phil-letter/). In response, Stock wrote an article in the newspaper The Spectator, in which she characterizes that letter as attempting to intimidate and silence those who criticize a prevailing orthodoxy (Stock 2021).

The strategy of catastrophizing exaggeration is apparent here. The open letter critical of Stock explicitly affirmed her academic freedom and her right to express the trans-exclusionary views in question—it simply expressed disagreement and disapproval of those views and criticized the British government's decision to award her an OBE. Her response to the letter explicitly paints it as ‘a document which wouldn't look out of place in the Salem Witch Trial archive’ (Stock 2021). And, as in the case of Mark Meadows, the criticism about the effect Stock's writing has on discourse around trans people became reinterpreted as a claim about her character. (She writes: ‘The authors of this letter clearly believed they could see into my soul’.)

In the wake of student protests at the University of Sussex, Stock ultimately resigned her position there. Bindel wrote an article portraying Stock as the victim of a ‘witch hunt mentality’ that eventually pushed her out of her job (Bindel 2021). Recall Freyd's (1997) acronym, DARVO (Deny, Attack, Reverse Victim and Offender) and Frost-Arnold's (2016) defensive ignorance, both describing cases in which someone, when criticized for doing harm, portrays their critic as the ‘real’ wrongdoer. In this case, the LGBTQ+ students at Sussex who protested perceived transphobia are portrayed by Stock and Bindel as aggressive and abusive—as the real offenders. Stock denies being transphobic; whether or not she is correct to do so, the defensive reaction on her part is certainly bound up with her identity as a woman and self-identified feminist. It is precisely because she identifies as both gender critical and a feminist that the defensiveness takes the shape it does, namely, accusations that her critics are attempting to silence women for speaking up.

6. Rape Culture

The Rape, Abuse & Incest National Network (RAINN) estimates that slightly over 10 percent of all American adults have suffered an actual or attempted rape at some point in their lives (https://www.rainn.org/statistics/scope-problem, citing the 1998 National Violence Against Women Survey.). Although there is no controversy as to whether rape is acceptable, feminist theorists have described the world we live in as embodying a ‘rape culture’—our society includes norms and practices conducive to the perpetuation of sexual violence and men's sexual entitlement to women (see, e.g., Brownmiller 1975; Burt 1980; Buchwald, Fletcher, and Roth 2005; Crewe and Ichikawa 2021).

Rape culture is a structural phenomenon, but like all structural phenomena, it is reinforced by individual actions. (Whether an audience is loud is not a feature of, or typically attributable to, any one individual, but individual actions contribute to the broader social fact in obvious ways.) Contributions to rape culture are not always deliberate, but they are harmful. People who personally think that rape is a horrible act might nevertheless contribute to rape culture: they might discount a sexual assault allegation against an acquaintance on the grounds that their interactions with that person had previously always been pleasant, or they might hypothesize that a woman complaining about being groped was probably inviting the attention, or they could support policies that make it more difficult for people to raise complaints safely. Indeed, as Yap (2017) and Tilton (2022) have emphasized, many common ways of expressing opposition to rape actually contribute to rape culture by perpetuating harmful ‘monster myths’ about rape and rapists. We will discuss this dynamic in more detail in section 7.

Discussion of rape culture often triggers defensiveness. Despite widespread theoretical discussion of rape culture as a structural phenomenon, people who are told that they are contributing to rape culture can easily hear it as an accusation that they are literally in favor of rape. Recall the defensive strategy from section 2 in which someone exaggerates a criticism so as to raise the evidential bar.

This tendency toward exaggeration is not always purposeful. Sometimes, it might be the result of a misinterpretation of the label ‘rape culture’. ‘X culture’ labels sometimes signify a culture where X is literally celebrated—consider, for example, ‘gun culture’. Someone with limited exposure to feminist thought might be forgiven for thinking personal opposition to rape inconsistent with contributing to rape culture. If they hear that they are doing so, their likely defensiveness is understandable, but regrettable.

Although defensiveness is a common response when someone is accused of contributing to rape culture, it manifests differently in different cases. Consider generally well-intentioned men. For men with feminist values, it is a comforting picture to draw a clean line between the bad men who perpetuate sexual violence and the good guys (like them) who condemn it and treat the women around them with respect, even serving to protect them from the bad ones. Even this much, while often well-intentioned, arguably reinforces oppressive norms (see, e.g. Brownmiller 1975: 16–17; Card 1991). When a ‘nice guy’ aligned with progressive values is accused of sexual misconduct, defensive reactions (from him or on his behalf) are common. When the (pseudonymous) ‘Grace’ wrote about the comedian Aziz Ansari pressuring her for sex, she faced vitriolic responses (the original account is here: https://babe.net/2018/01/13/aziz-ansari-28355, and some backlash against her can be found here: https://www.buzzfeednews.com/article/tanyachen/ashleigh-banfield-open-letter-to-aziz-ansaris-accuser). It is much easier to align oneself with the #MeToo movement by engaging in strong public criticism of high-profile offenders like Harvey Weinstein than it is to reflect on how one's own actions might contribute to the problem. Recognizing one's own contributions to rape culture—whether by victim-blaming, or exhibiting objectification and entitlement to women around them, or by ignoring sexual harassment against women they do not find personally attractive, or in other ways—can threaten one's self-concept as a progressive and decent person. The obstacles that obscure men's recognition of their contributions to rape culture have also been a common theme among feminist theorists and activists (see Yap 2017; Tilton 2022).

But we are especially interested in a different issue, namely, the interaction between rape culture, defensiveness, and feminist identities, especially identities attached to feminist women. Self-identifying feminists, even more than other people, have identities that depend for their maintenance on opposition to sexual violence, and therefore we feminists are particularly likely to reach for defensive strategies to resist the idea that we contribute to rape culture. But feminists—including feminist women—do sometimes contribute to rape culture. When they do, their identities as feminists give rise to special challenges.

Here is an example. In November 2016, Canadian feminist author Margaret Atwood signed an open letter criticizing the University of British Columbia (UBC) for its dismissal of Stephen Galloway, who had been accused by a former student of sexual harassment and assault. The letter's concern was mainly procedural, but it was criticized by several feminist activists for centering Galloway's interests and for its failure to challenge the status quo of rape culture and an environment that too often silences victims of sexual violence.¹

Although she was one of several dozen signatories, Atwood's feminist reputation made her a particular focus of the controversy. Antirape activists who had grown up thinking of her as a feminist icon felt betrayed and disappointed by her position. We fully acknowledge and agree with the importance of due process in response to sexual misconduct allegations. Our point is not to defend the university's treatment of Galloway, but to point out how Atwood's response to her critics employed some of the defensive strategies we outlined above.

One of Atwood's moves was to exaggerate the criticisms into ‘the position that the members of a group called “women” are always right and never lie—demonstrably not true—and that members of a group called “accused men” are always guilty’ (https://thewalrus.ca/margaret-atwood-on-the-galloway-affair/). A position like this is a caricature of antirape activism; Atwood's rejection of it is no challenge to the claims her critics actually made. This exaggeration also allowed for a display of performative woundedness. Atwood portrayed herself as a victim of her critics, insinuating that she is being held ‘responsible for all ills’, for ‘failing the world on gender equity’, and that perhaps she should ‘stop trying’ (https://twitter.com/MargaretAtwood/status/952583123157516288). She put a person-label front and center in a (2018) Globe and Mail op-ed entitled, ‘Am I a Bad Feminist?’. In it, she defends her comparison of UBC's response to Galloway with the Salem witch trials and adds comparisons to ‘the French Revolution, Stalin's purges in the USSR, the Red Guard period in China, the reign of the Generals in Argentina and the early days of the Iranian Revolution’.

Recall from our discussion of bell hooks's classmates that unfair patriarchal attacks might establish justifiable and proper defensive habits, which can then be misapplied. Atwood's case matches the same pattern, we think. As a high-profile feminist author, Atwood has considerable experience as a righteous victim of unfair misogynistic attacks; we have no doubt that, by necessity, she has developed something of a resistance to criticism. But as in the case of White feminism, it is all too easy for genuine feminist criticism to feel like an unfair attack and a call for defensiveness. Atwood's op-ed explicitly connects complaints about her defense of Galloway to such antifeminist backlash:

It seems that I am a “Bad Feminist.” I can add that to the other things I've been accused of since 1972, such as climbing to fame up a pyramid of decapitated men's heads (a leftie journal), of being a dominatrix bent on the subjugation of men (a rightie one, complete with an illustration of me in leather boots and a whip) and of being an awful person who can annihilate—with her magic White Witch powers—anyone critical of her at Toronto dinner tables. I'm so scary! And now, it seems, I am conducting a War on Women, like the misogynistic, rape-enabling Bad Feminist that I am. (Atwood 2018)

As understandable as this error was, Atwood was not listening sympathetically to her critics. If she had been, she would know that none of them said that women are never mistaken or dishonest or that accused men are always guilty. What they said is that they were disheartened by her prioritization of her literary colleague's interests over those of the women who had spoken out against him and by her invocation of rape culture tropes in defense of doing so. Portraying her critics in this light had the function of sparing Atwood the necessity of questioning the tension between her feminist commitments and her actions. As with the gender-critical feminists discussed in section 5, these tensions are sometimes things that one needs not to know about. It is easy to understand how someone could respond to this perceived attack by reaching for defensive strategies, including exaggerating and lashing back against their critics.

Another related phenomenon has to do with the role of anti-oppressive identities in making harmful ideas palatable. We think this happens when, for example, one invokes one's feminist identity as part of one's efforts to discredit someone's sexual assault complaint. Kipnis (2017) is, in our view, an example of such a project. Some of the same tropes that would be rightly dismissed as toxic (and tired) contributions to rape culture, coming from dude-bro ‘Men's Rights Activists’, were treated as a fresh heterodox perspective, coming from within the feminist camp. (See, for example, Jill Filipovic's New York Times [2017] review, which compares Kipnis's book to a similar book by two men. Filipovic rightly recognized the latter's contributions to rape culture but wrote favorably about Kipnis's book, despite its trading in the same one-sided rape culture tropes.) While we do not see defensiveness as such at play here, this is an instance of a broader phenomenon, whereby anti-oppressive identities serve to mask one's contributions to oppression.

7. Living with Imperfection

Defensive strategies quite generally can give rise to epistemic and social problems. But as these examples illustrate, there are particular problems that arise when defensiveness serves to protect a person's strongly held identity. For feminists, a tendency toward defensiveness can sometimes make us epistemically worse off with respect to those very commitments. For instance, as feminists, it is in our interest not to uphold rape culture. And yet, defensiveness can result in our failure to know—or accept—that we may be doing so when we are. After all, we all fail in our commitments sometimes. We are also liable to misidentify the cases that do and do not call for defensiveness. None of us are perfect.

Given that we are all living with imperfection, there are difficult and important conversations that defensiveness might prevent us from having. Many approaches that, in our view, perpetuate oppression—including gender-critical rhetoric, descriptions of #MeToo as having gone ‘too far’, etc.— do so with the stated intention to open space for ‘important conversations’. These are critical social questions that need to be talked about, their proponents say, in the course of their complaints that less progressive perspectives are being crowded out. We entirely agree that these are important matters that need to be talked about. We need to talk about how to balance due process with supporting survivors. We need to talk about how to satisfy the needs of trans people to exist in society alongside cis people who are uncomfortable with them. But those conversations are impossible without a serious reckoning with the possibility that one is perpetuating harm. As discussed above, hooks's fellow students were also in such a position, where their focus on their own discomfort meant that they did not consider the possibility that they might be perpetuating oppression.

In section 3 we discussed the ways active ignorance and defensiveness can derail sensitive conversations. Discussion of racism, sexism, transphobia, ableism, and other forms of oppression can all raise the conversational temperature and create hostile and defensive conversations. In all of these cases, serious people—professional philosophers, often—sometimes say that describing actions as problematic in the relevant way (as transphobic, say) is incendiary and unprofessional.

We have not seen this claim advanced in peer-reviewed publications, but we encounter it often on philosophy blogs and social media (for example, in some of the comments to a 2018 blog post on philosophy norms of discourse here: https://dailynous.com/2018/06/02/a-note-on-making-discussions-here-better/ as well as in a blog post about that discussion here: https://theelectricagora.com/2018/06/05/epithets-in-philosophy/). Elizabeth Anderson (2010: 48) articulates a more moderate version of this idea.

Such taboos against criticism represent dramatic obstacles, both political and epistemic. Although we do not think conversation is always good—we are convinced by arguments that sometimes it is better not to engage (see, e.g., Fantl 2018)—we do think that productive conversations are quite often an important prerequisite of positive social change.

Being aware that defensiveness can be triggered by the perception of an attack on a closely held identity might help a potential interlocutor approach a sensitive discussion in a less threatening mode. For example, we think there is often good reason to offer criticism constructively, choose less incendiary words or tones, and focus more on descriptions of actions than on negative labels for people. We are interested in producing better conversations and a better culture; less defensive conversations further that goal better than more defensive ones do. Though as we have argued throughout this paper—and seen in our personal lives—even carefully worded criticisms can still trigger defensive responses.

However, people vary in their spaces of options for sensitive conversations. We ourselves—the cis trans-inclusive feminist authors of this paper—have reason to approach cis gender-critical people who are uncomfortable with trans women with at least a degree of empathy and charity to make a productive conversation more likely. If our interlocutor wants to have a good faith discussion with us about the subject, then we ought to come to the conversation with openness. We ourselves should try not to be unwarrantedly defensive. This will often require listening seriously to criticism from the perspective of ideological stances we do not hold. And if we reject those criticisms, it should be for considered reasons, not a defensive commitment to find fault with the attack. And we certainly should not respond to good faith engagement with anything along the lines of ‘how dare you call me a terrible person?’ or ‘why are you trying to get me fired?’.

Julia Galef (2021) suggests that people tend to react to argument, discussion, or information in either a ‘soldier mindset’ or a ‘scout mindset’. While we do not wish to rest on any particular empirical commitments of mindset psychology, we do see an affinity between Galef's soldier mindset and defensiveness, in our sense. A soldier seeks ways to defend their beliefs—must I accept this information, or can I find an excuse for doubting it? A scout seeks ways to correct their misconceptions and augment their knowledge—can I accept this? Are there good grounds for it? Galef says that too much of the time, most people tend to be too much soldier and not enough scout. We agree—especially when it comes to certain kinds of normative criticism directed against people from relatively privileged perspectives.

Of course not everyone has such privilege. In particular, people whose fundamental rights are challenged by gender-critical stances—trans people, for instance—are being attacked in these conversations in a way that we personally are not. So they may well have good reason to be defensive, to react in anger, and so on. Given the bad set of choices available to many victims of oppression, defensiveness and anger may well be fitting responses to attacks on their basic human rights. Such responses can also serve a political end, underscoring the importance of the issue (see, for instance, Lorde [1981] or more recently Srinivasan [2018] and Cherry [2021]). As such, we are not attempting here to give a general argument against defensive reactions or claim that they are never justified. Instead, we simply point out that there is often something epistemically unfortunate about them even in those cases where they are emotionally proper.

Moreover, as we indicated in our discussions of White feminism and rape culture, the more often one employs defensive strategies, even if they are appropriate, the likelier it is that one will misapply them in cases where they are not appropriate. We think this is what is going on when gender-critical cis women justify their defensive stances on the grounds that their rights are similarly being assaulted. While we acknowledge that they feel that way, feeling it does not make it so. This is not the place to argue for it, but our view is that trans-inclusion simply does not threaten cis women's basic rights the way that trans-exclusion threatens trans people's. Gender-critical people see trans-inclusion as continuous with the patriarchy and would likely disagree with their behavior being included as examples of undue defensiveness. However, even if one disagrees with the ways we have illustrated the phenomenon, the potential harms of defensiveness still stand. And the normative question of when exactly defensive strategies are justified is not our focus here. We simply explicate a particular kind of harm that often happens when they are employed.

Human relationships are complicated and difficult, and well-intentioned people sometimes make harmful mistakes. That means negotiating accountability is also frequently complicated and difficult. It was natural, for example, for Atwood to want to defend a member of her community. After all, feminists are in a difficult position if we do not believe allegations against our male friends and colleagues, and we do not mean to suggest that there are easy answers. Solutions to rape culture in general are not at all straightforward—and a commitment to anti-oppression does not pull in a single direction. As Angela Davis (1981: 184) describes, Anti-Black racism is deeply ingrained in the history of the public imagination of rape (see also Hamad [2019: ch. 3] and Tilton [2022]).

But as Atwood's critics pointed out, we uphold rape culture if we think only of the potential harms to the men accused and stay silent on the harms to the people who make claims against them. After all, part of dismantling rape culture is the understanding that sexual assault is not the exclusive domain of monstrous people, and a tendency toward defensiveness serves to obscure this in several ways.

First, the defensiveness leads to epistemic costs in the form of false beliefs. Just as the catastrophizing description of a racist calls to mind an extreme exemplar—one that can provide powerful reasons to deny the label to someone who does not match it, even if they are engaging in racist actions—so, too, does the stereotype of a rapist call up an extreme image of a moral monster. Brownlee (2016) argues that using the person-label ‘rapist’ itself wrongs its referents by essentializing them to their worst actions. We do not go so far as Brownlee in this regard—in at least some contexts, we think it is important, for instance, for survivors of sexual violence to be able to identify ‘their rapists’ as such—but we agree that there is a moral cost to this language that should not be ignored. When the stereotype manifestly does not match the potential perpetrator, it is easy to conclude, mistakenly, that he must not after all be guilty of the accused conduct: ‘only a monster would do that’ (see Yap 2017; Tilton 2022; Falbo 2022). This is a kind of modus tollens response, but a modus ponens response is also possible when the evidence of someone's guilt is overwhelming. If it is only monsters who commit sexual assault, then accepting someone's guilt will entail accepting that they are a monster.

This epistemic error arguably leads to several social harms. Believing that someone is a monster carries with it ideas about irredeemability and allows us to write them off, morally speaking. We oppose rape culture; rape is an incredibly harmful act. But we also recognize that many perpetrators of rape are ordinary complex human beings with many different life experiences, interests, and impacts on the world around them. Taking rape culture seriously means recognizing the ways in which it desensitizes men to the harms of sexual violence and portrays domination as desirable. Treating sexual assault as something people do just because they are monsters lets culture—and our shared responsibility for it—off the hook.

Now, what accountability looks like more specifically is complicated, but we want to resist what Elizabeth Bernstein (2007) has called ‘carceral feminism’. Carceral feminism relies on the criminal justice system to help end gendered violence despite its poor track record at prosecuting sexual assaults (Bernstein 2007; see also Crewe and Ichikawa 2021: 253–54). And we also do not think that sending offenders to prison (where they themselves face increased risk of sexual assault and other trauma [see Alcoff 2018: 235]) is a particularly good solution. Luckily, there is a lot of work out there on alternatives to carceral feminism, for instance, in restorative justice and on feminist prison abolition more generally (for feminist arguments in favor of prison abolition see, for instance, Davis et al. 2022; Kaba 2021).

Nor do we think the kinds of harsh but noncriminal sanctions one sometimes sees—working-class people becoming the Twitter meme of the day and being fired over old racist Facebook posts, for instance—are the way forward. On the contrary, a culture that leaves no room for the toleration of even mild contributions to oppression exacerbates the problem. It further motivates the kind of defensiveness we oppose and fails to acknowledge that we have all done something wrong at some point in our lives.

Moreover, people change their minds. For example, people who no longer align themselves with gender-critical views sometimes describe coming to realize ways in which their past actions and stances were harmful. Sometimes this is because they see their former allies using unfair defensive strategies. For instance, Beau Dyess became disillusioned with the gender-critical movement when their lesbian friend was attacked—in a textbook DARVO moment—for criticizing prominent gender-critical figures’ homophobia (Dyess 2020). Sometimes such people find themselves uniquely well-positioned to have productive conversations with others about these issues (see, for instance, Dalwood's [2021] profile of Alicia Hendley).

Regardless of how we choose which conversations to have, we need to recognize the potential effects of defensive strategies. Shutting down conversations altogether is often a way to stagnate social progress, not to advance it. Defensiveness may often be understandable—or even the best available option—as an emotional response, especially when it comes to already marginalized people facing unjust attacks. But the use of defensive strategies nevertheless carries significant epistemic costs. There is no simple or guaranteed way to avoid triggering defensive responses in others or to ensure that we are only defending ourselves in ways that are warranted, but being aware of the costs of defensiveness can perhaps lead us to be attentive to the argumentative tactics we use and the interlocutors with whom we choose to engage.²