1. Over-identification of instances

The first kind of over-sensitivity involves over-identifying instances. This can come about due to simple perceptual mistakes of the sort made by Kondabolu: he misheard a remark at the comedy club. But such errors are comparatively unusual, so it will help to list some other cases where the supposed over-identification of instances of unfairness or injustice instead stem from an agent’s contested counterfactual judgments.

• • A woman, Professor W, is talking to a man at a party. After introducing himself, he asks her what she does, and she explains that she’s a professor of English literature, specializing in Shakespeare. Despite having no training in the field, he launches into a detailed explanation of the meaning of his favorite Shakespeare play, King Lear, muddling several key plot points. Later, recounting this incident to her husband, Professor W rolls her eyes at her conversational partner as a sexist mansplainer who most likely wouldn’t have held forth in this way to a male English professor.Footnote ⁶ “You’re being over-sensitive,” her husband says: “He probably just jumped at the chance to talk about a play he loved.” The implication being that he would have acted the same way regardless of his interlocutor’s gender.

• • Melody takes her daughter, Ellen, 4, out to the park, where Ellen has a meltdown over getting the wrong color Popsicle out of the cooler. “What a little spitfire!” remarks a passerby. “You’ve certainly got your hands full with that one.” Melody is struck by the gendered overtones of that remark, since she doubts that a boy of the same age having a meltdown would have been labeled a “spitfire.” Over dinner, her partner disagrees: he thinks that while subtly gendered terms like that certainly exist (e.g., “sassy”), she’s being over-sensitive in this case: a boy would also have been called a spitfire for yelling “You’re ruining my life!” loudly at a parent proffering the erroneously colored cold treat.

• • Brianna, who is in a bad mood, cuts off her father-in-law abruptly during a long and boring story: “Where is this going?” she asks him, sighing audibly. He reacts to this slight with a fury that strikes her as being disproportionate and likely due to gendered norms of care—she can’t imagine him getting that mad at his son-in-law, Bryn, for not listening attentively to his ramblings. When Brianna brings this up later to her husband, John, he thinks Brianna is being over-sensitive: his father has just had a short fuse with everyone lately.

These cases, drawn from real ones recounted to me, all involve counterfactual judgments on the part of a woman: to the effect that a man or boy would likely have been treated differently than either she herself or a female counterpart (due to a gendered norm of entitlement, a distinctively gendered term, and disproportionate punishment for women, respectively).Footnote ⁷ In all three cases, a man’s disputing that counterfactual judgment furnishes him with the grounds, or at least evidence, for an accusation of over-sensitivity on her part.

I’ve chosen these three examples because the woman charged with over-sensitivity is not clearly right, and the man leveling this charge is not obviously wrong, about what actually went on in these cases. Still, his accusation of over-sensitivity toward her invites a number of more or less obvious responses that are worth distinguishing at the outset. To wit:

• • The “I was there” response: in all of these cases, the woman actually witnessed the offending behavior (the putative mansplaining, the seemingly gendered remark, and the ostensibly outsized reaction due to a gendered norm violation). She may thereby have been privy to nuances of tone and inflection, and other contextual clues, that serve as evidence for her suspicions. Some of these will be difficult or impossible to articulate after the fact: our reading of a complex social situation may be accurate but inchoate. And even if we modify these scenarios slightly, such that the man accusing the woman of over-sensitivity was also present at the time, he may have missed these features, or been paying less attention to them. After all, subtle signs of sexism and misogyny constitute important social information that she will typically be more invested in—and hence accustomed to—taking in, given its importance for surviving, let alone thriving, in a patriarchal social world.

• • The evidence for the relevant counterfactual: In some cases, the woman may be able to point to specific evidence to back up her contentious counterfactual judgment. For example, Professor W might point out that many men have mansplained Shakespeare to her in the past when given the opportunity; the same is not true, anecdotally, for her male colleagues. Or Brianna might point out that her father-in-law has a long track record of subtly sexist and misogynistic behavior in a similar vein. For example: expecting his female relatives to clear the table and not contributing himself; dominating conversations in which women have as much or more to contribute; praising his granddaughters and grandsons in different, and gendered, ways.

• • Evidence of personal reliability: In (again) some such cases, the woman may be able to point to the fact that she is typically reliable in picking up on sexism and misogyny when it is indeed present, yet doesn’t go around making these sorts of accusations willy-nilly, or baselessly, whenever there is the relevant kind of social friction. (After all, she would certainly be punished heavily for so doing.) As Kondabolu remarked in the case with which I opened: “I have never been wrong before about [racism]!” While slightly—and deliberately—hyperbolic, weaker claims in this vein are common and apt when facing charges of over-identifying instances of sexism, misogyny, racism, and so on.

All of these are potential lines of defense to the effect that the woman accused of over-sensitivity is likely to be right and the man accusing her is likely to be wrong, in a given case. They may be more or less applicable; and they may or may not be accurate (again, I am leaving it open that the woman is actually wrong in any of these cases). But there’s another point about the accusation of over-sensitivity that needs to be made here too. “Over-sensitivity” is a general accusation, typically implying the existence of a pattern: namely, that the woman is not just wrong in this case, but is too prone to “false alarms” in identifying instances of sexism or misogyny in general.Footnote ⁸ Given this, we might wonder: what about the risk of under-sensitivity in the sense of missing real instances of the relevant phenomenon? Why, given this omnipresent risk, does the prospect of over-sensitivity tend to dominate the discourse?

The comparison with other kinds of inquiry in messy, real-world contexts is instructive. When it comes to diagnostic testing, there is typically a trade-off between tests that have a higher percentage of false positives—or false alarms—and tests that have a higher percentage of false negatives—or misses. (In statistical terms, the former metric is sensitivity; the latter, specificity.) And in many endeavors, we are willing to tolerate a somewhat heightened risk of false alarms in order to avoid missing important phenomena. A rapid test for COVID that produces a higher percentage of false positives than a rival test might still be valuable if it misses a lower percentage of cases. An algorithm for detecting credit card fraud that “pings” with false alarms more frequently might still be a good one if it is also better at not letting real instances of fraud go under the radar. A sampling method for detecting spoiled food might well be useful even if it results in some perfectly good food being flagged in a higher percentage of cases, if it is less liable to miss potential food-borne pathogens.Footnote ⁹ The value of a detection mechanism will depend on a complex set of considerations: not only these trade-offs, but the relevant costs and benefits thereof, together with information about base rates and prevalence. The point is just this: a higher proportion of false alarms is not inevitably alarming. Why aren’t we similarly willing to tolerate the risk of false positives in order to detect more sexism, misogyny, and racism?

The obvious answer has to do with what we do with the relevant suspicions, and the prospect—salient to many minds—of false accusations in the latter case. Flagging food as spoiled when it isn’t is one thing; flagging a person as a bigot, quite another. But we need to back up here: in the three cases above, and many others of this kind, there is not necessarily any prospect of accusations—false or otherwise—in the offing. That is, there may be no thought of confronting the partygoer, the passerby, or the father-in-law: in part because the threshold of certainty has not been met for such action. Perhaps the woman simply wants to avoid the offending party going forward.

We should not forget that, in some cases, the stakes for not avoiding people suspected of sexism and misogyny are high. Suppose that, in addition to seemingly mansplaining Shakespeare to her, Professor W’s conversation partner at the party strikes her as being a little creepy: touching her arm and making excessively familiar comments, etc. Professor W hence decides to extricate herself quickly from the conversation and politely declines a ride home with him. If she was mistaken about the vibes, he has not been accused or wronged in any serious way;Footnote ¹⁰ if she was right to worry, she may have avoided calamity. (The spoiled food analogy is distressingly apt here.)

In many other cases, a woman will take no action whatsoever with respect to a party she suspects of sexism or misogyny. In each of the three preceding examples, she may simply want to vent her suspicions and frustrations, and note the appearance of gendered patterns that dog her and other women or girls throughout our lives, to a sympathetic audience. Who, as it turns out, is not only not sympathetic but levels an accusation in her direction: that she is being over-sensitive. It’s telling that the risk of this accusation being false isn’t as salient to many.

There is also the crucial point that, even when such accusations are leveled, they need not be of bigotry: they may be for the subtler and more common failing of channeling or perpetuating, say, sexist mores. It’s not surprising if the father-in-law, given his gender and generation, has internalized the sense that women in particular owe him a sympathetic audience. Pointing this out could be done gently and, with a conscientious person, well-taken—or at least taken under advisement.

Finally, accusations of outright bigotry will, in some cases, be justified, even on the basis of imperfect information. It’s not, after all, as if interpersonal interactions take place in a court of law: innocent until proven guilty is a legal standard, not a moral one. Imagine a case where a pattern of similar incidents unfolds with Brianna’s father-in-law, and her evidence for his sexism becomes significantly stronger (though still admittedly imperfect). Calling him out for being sexist is not a legal proceeding, and he is not about to be arrested (or “canceled”) on this basis. Still less is he being summarily thrown in prison. As in legal proceedings, a critical discussion can and should take place: is the accusation fair? Is the evidence sufficiently compelling? And, if it is, what if anything should be done on this basis? Given the preponderance of defenders and defenses a seemingly sexist or misogynistic agent will tend to have within a patriarchal society, it may again be valuable to have some others more willing to issue call-outs. We sometimes have importantly different epistemic and moral roles within our community: a socially sensitive agent may play one such.

Consider, in this vein, Olivia Bailey’s nice case of an agent, Dolores, who is exquisitely sensitive to the plight of the underdog, and the hurt of the outcast, in society.

At the same time, Dolores can be too quick to jump to conclusions: some of what she perceives as exclusion or ostracism is really just freedom of association. As Bailey points out, if Dolores undergoes a shift toward greater overall accuracy in her moral sensibility, something will have been gained, but something important will also have been lost. There is hence a case for ambivalence about this kind of moral improvement, Bailey argues, the overall assessment of which

I agree but would add the perhaps more radical point that the question of whether to count the shift toward accuracy—which Bailey glosses as greater attunement to another real and competing moral value—as moral improvement often seems to me to be open. In statistical terms: even if Dolores’s loss of sensitivity is more than made up for by gains in specificity, there are questions of base rates and the stakes of various kinds of error to consider. She may now be missing a great deal, and what she is missing may matter greatly, in an environment in which ostracism and exclusion are both common and deleterious. Moreover, as Bailey’s discussion brings out, human beings are of course not just diagnostic tests or signal-detectors: part of what we are assessing, when it comes to the epistemic and moral value of someone’s particular sensibility, is how much they can contribute to critical discussions following the alarm bells that sensitive agents are particularly adept at raising.

Moreover, such sensitivity can be a crucial survival skill—and a coping mechanism—for people facing oppression in virtue of their gender, race, and so on. It can help us advocate on our own behalf in ways that are often undermined by dominantly situated agents. In another incident Hari Kondabolu recounts, a group of three white men attacked him during his college years in Maine, yelling “What are you doing here? Why are you here?” repeatedly. Given Kondabolu’s racially marginalized status as a brown man, and the white-dominated social environment, he clocked it as the racist attack it surely was. But instead of affirming his impressions, and reflecting on the relevant counterfactual—how much less likely he would have been to be attacked in this way as another white man in that environment—the white friends whom he sought sympathy from in the aftermath raced to question his impressions. According to one of them: “You don’t actually know it was a hate crime.” “Did they use slurs?” “I don’t think it’s fair you’re accusing them of a hate crime.” Kondabolu: “I was shocked, because I didn’t know he was their defense attorney.” His friend’s reaction is a nice, negative encapsulation of many of the points I’ve been making in this section.

There’s also the background point that, such extreme incidents aside, racism is not just a collection of sins committed by bigoted individuals. Critical race theorists, following Charles Lawrence III, have long argued that racism is a kind of social disease faced by people of color and perpetuated by all of us—white people especially, given our investments in ignorance, denialism, and upholding our own privilege. (Kondabolu’s friends in the above example are a clear case in point here; and, tellingly, he said their disbelief was as terrible as the attack itself.) Racism is hence well-understood on what Lawrence calls the epidemiological model, and addressed as an urgent public health matter (Reference Lawrence1987). Sexism and misogyny can be productively understood in a similar vein, according to my own adaptation of Lawrence’s insights (see, e.g., Manne Reference Manne2017, 28). A corollary which has emerged in this section: the presence of some individuals who may be liable to a greater percentage of false positives than their less sensitive counterparts is potentially highly valuable in detecting instances of sexism, misogyny, and racism when it is occurring. To halt the progression of these social diseases, we need people who miss very little, even if they are occasionally mistaken.

Are these people sensitive, or slightly over-sensitive albeit in a defensible and valuable way? The answer to this question will depend on how short the agent falls of the threshold for “sufficient reliability” that my definition for sensitivity has built into it. We can thus imagine two ways of reading the Kondabolu case with which I opened: first, he is disposed to correctly identify racism with sufficient reliability to count as sensitive, not over-sensitive (or “too sensitive”), per his own diagnosis. He just unluckily got it wrong on this particular occasion.Footnote ¹¹ Second, and more subtly, he may indeed be slightly over-sensitive to the relevant phenomenon—disposed to identify instances of racism slightly more often than would qualify him as “just sensitive”—but in a way that is a valuable corrective to the under-sensitivity of other agents in his epistemic and moral community.

2. Over-extension of concepts

We can now move to consider the second kind of supposed over-sensitivity I want to discuss in this paper: conceptual expansion, where counting novel phenomena under headings like sexism, misogyny, and racism garner the accusation that people are becoming over-sensitive. The psychologist Nick Haslam articulates this worry in a well-known recent paper, “Concept creep: Psychology’s expanding concepts.” Haslam argues that certain, negative concepts—including abuse, trauma, bullying, and prejudice—have been expanding in both “vertical” and “horizontal” directions over the past few decades. Roughly, vertical expansion involves lowering the threshold for counting phenomena as falling under the extension of a concept—counting comparatively mild teasing or physical “joshing” as bullying, for example, or loosening the requirement that bullying behavior has to be repeated (Haslam Reference Haslam2016, 4–6). Or it may, claims Haslam, be attributable to shifting the criteria for the application of a concept from the perspective of the perpetrator to the perspective of the victim, thereby supposedly introducing an element of subjectivity into its extension. Counting emotional abuse as such requires taking into account how some form of treatment felt to the recipient, and not just what was done or how it was intended (Haslam Reference Haslam2016, 3, 11). Horizontal expansion involves enlarging the extension of a concept to include adjacent phenomena—recognizing cyber-bullying, for example. It may also, adds Haslam, count inactions or omissions in with actions or commissions—counting social exclusion as bullying, for instance, in addition to acts of violence, threats, intimidation, and spreading rumors about the target (Haslam Reference Haslam2016, 4–6). “Conceptual creep has occurred across a diverse assortment of concepts and has commonly involved an increased sensitivity to negative experience and behavior, an increased focus on harmful forms of inaction, and an increased acceptance of subjective criteria for deciding when the concepts apply,” Haslam summarizes (Reference Haslam2016, 10).

Examining Haslam’s case study of concept creep with respect to prejudice, and his worries about the drivers and results of this, will be particularly instructive given my aims in this paper. He traces the study of this topic to Gordon Allport’s Reference Allport1954 The nature of prejudice, which defined prejudice as “an antipathy based upon a faulty and inflexible generalization. It may be directed toward a group as a whole, or toward an individual because he [sic] is a member of that group.” (Allport Reference Allport1954, 9). Haslam notes that subsequent scholars were not content to leave things there, introducing concepts like aversive prejudice, where instead of harboring conscious antipathy toward a group, people are instead fearful, uncomfortable, or uneasy around its members, sometimes without any conscious awareness of being so. The related concept of implicit bias incorporates the idea of harboring unfair negative associations with a certain group, among other things, again without conscious awareness. Concepts like “benevolent sexism” crystallize the idea that some kinds of prejudice involve superficially good, but patronizing, understandings of certain groups of people: stereotyping women as caring, innocent, and lacking in aggression, say. Finally, concepts like micro-aggressions count subtler and less obvious slights toward a social group as instances of prejudice (Haslam Reference Haslam2016, 9–10).

Haslam’s notion of concept creep is ostensibly descriptive, not normative. “It is possible to define ‘prejudice’ in a coherent way that encompasses all of these variants,” he acknowledges. “My point is simply that the concept now refers to much more than it did several decades ago.” (Haslam Reference Haslam2016, 10). Moreover: “Each extension of the concept of prejudice was arguably well justified.” (Haslam Reference Haslam2016, 10). And yet the slippage to a critical mode is soon easily discernible, arguably already present in the tendentious label “creep” (as opposed to the more neutral label “expansion”) suggestive as it is of an undeliberate, insidious, ooze.

Haslam considers various explanations for concept creep, and takes the most promising to be Steven Pinker’s idea that minority groups have created various “rights revolutions,” whose campaigns “are propelled by an escalating sensitivity to new forms of harm”—some of which he is explicitly critical of as over-sensitivity—and “the slightest trace of a mindset that might lead to [them]” (Pinker Reference Pinker2011, cited in Haslam Reference Haslam2016, 13). Haslam: “A general sensitization to harm could go a long way toward accounting for concept creep.” (Reference Haslam2016, 13). More: “Psychology has played a role in the liberal agenda of sensitivity to harm and responsiveness to the harmed.” (Haslam Reference Haslam2016, 13). Evincing what he describes as a more ambivalent attitude toward the resulting concept creep, he worries that it has created a new culture of victimhood, in which people are typecast as various kinds of victims on the one hand, and moral villains on the other: abusers, traumatizers, bullies, and bigots. As the result of this, as well as the lower threshold for applying the relevant concepts (“semantic dilution”), “concept creep may release a flood of unjustified accusations and litigation, as well as excessive and disproportionate enforcement regimes” (Haslam Reference Haslam2016, 14). Moreover, “concept creep runs the risk of creating a public impression of psychology as a field that exaggerates misery … and is over-sensitive to perceived bias and discrimination.”Footnote ¹² The upshot of Haslam’s discussion is this: concept creep with respect to prejudice stems in large part from over-sensitivity. And it is liable to result in false accusations.

Haslam’s picture is in several respects seriously misleading. Allport was hardly the first theorist of prejudice in general and racism in particular. Even confining our focus to the Anglo-American context, Black scholars like Frederick Douglass had been theorizing about these matters for well over a century (Reference Douglass and Blassingame1845/1979).Footnote ¹³ Some fifty years later, in 1898, W. E. B. Du Bois analyzed racial prejudice as the conviction “that people of Negro blood should not be admitted into the group life of the nation no matter what their condition may be” (Du Bois Reference Du Bois and Chandler2015, 82), already pointing to some of the structural sources of inequality that Haslam does not reference, including economic disadvantage. In Du Bois’s famous notion of double-consciousness, one will have as a Black subject a “sense of always looking at oneself through the eyes of others, of measuring one’s soul by the tape of a world that looks on in amused contempt and pity” (Reference Du Bois2007, 3). Here we have clear precedent for the supposedly recent ideas of prejudice not always involving antipathy, and instead being patronizing or pitying.

Meanwhile, feminist theorists have been clear about the fact that sexist prejudice may involve superficially flattering stereotypes about women, which actually serve to control and disempower us, since at least Mary Wollstonecraft. In her A vindication of the rights of women (Reference Wollstonecraft1792), she wrote in opening:

Wollstonecraft is not alone here: it is hardly an exaggeration to say that noticing the insidious and patronizing quality of superficially positive stereotypes about women animated much of early feminist theorizing.Footnote ¹⁴

To state the obvious, then: this is not a matter of there once having been a narrower concept of prejudice which was enlarged, rightly or wrongly, in recent decades, due to heightened sensitivity to harm. Rather, a white man, widely treated as an authority on the topic, missed a great deal that people directly affected by prejudice—Black people and white women among them—had long been theorizing, with limited uptake in the academy. Their ideas were then gradually rediscovered, or reinvented, by subsequent scholars.

That subordinate group members would typically be the people to break ground on concepts of prejudice—which affect them the most directly and acutely—is deeply unsurprising. My point here does not rely on any particularly controversial claims within standpoint epistemology. As Lidal Dror has recently argued, there is widespread agreement with the idea that socially marginalized people, in virtue of their social position, “tend to have a superior epistemic position than non-oppressed people when it comes to knowing things about the workings of social marginalization that concern them.” This is because they “tend to have more relevant experience and motivation” to acquire such knowledge (his emphasis). Dror calls this the weak inversion thesis, and endorses it—while ultimately arguing against the stronger inversion thesis that oppressed people have this advantage in principle (Reference Dror2023, 624). I do not need such a strong claim for my purposes either. For all I have said, the epistemic advantage of people of color and marginalized genders may consist in contingent tendencies to be more sensitive to phenomena that affect us more in general, but are in principle open to people in any social position to acquire knowledge of by paying careful attention to the relevant evidence (including testimonial sources). To adapt Elizabeth Anderson’s (Reference Anderson and Edward2020) example: you don’t need to be a standpoint epistemologist to acknowledge that car mechanics, who are tasked with figuring out problems with cars and fixing them for their livelihood, are likely to be the most insightful people when it comes to car repairs and maintenance. Similarly, I think it is hard to deny that the people for whom surviving, let alone thriving, depends on understanding the contours of prejudice will typically be the most insightful when it comes to first theorizing these concepts.

What Haslam terms “concept creep” with respect to prejudice is thus far not alarming, or even an apt source of ambivalence. It is also not obviously true that wider understandings of such concepts are due to greater sensitivity to harm per se, as opposed to the specific harms of racism, sexism, misogyny, and so on. This is not obviously a problem, even when it does stem from and result in increased sensitivity. Take the concept of micro-aggressions, for example, which are, roughly speaking, slights that have the potential to accumulate and add up to something seriously harmful for their targets or victims.Footnote ¹⁵ Plausibly, this concept has resulted from an increased sensitivity to the ills of structural oppression and how they can be manifested. And it has enabled oppressed people to name harms and wrongs they thereby face that would otherwise be very difficult to identify: how do you complain of a “death by a thousand cuts” without being empowered to point out the scratches?

This leaves us to consider the central concern driving those up in arms against “concept creep”: the risk of false accusations and increased moral litigation of various harms and wrongs. But here again Haslam fails to convince: many of the ways of expanding the concept of prejudice that he himself highlights deliberately militate in the opposite direction.

Take the concept of implicit bias. The idea that many of us—those who face oppression, included—harbor implicit racial biases, over which we have no conscious awareness and subsequently little control, has long been taken to reduce the aptness of accusations of racism. “Implicit race bias does not mean a person is a racist,” write the psychologists R. Banaji, Susan Fiske, and Douglas Massey (Reference Banaji, Fiske and Massey2021).Footnote ¹⁶ Even as a concept expands within the academy, we can make important distinctions: between someone’s channeling or perpetuating racist social forces, and someone’s deserving the pejorative label “racist” per se. Similarly, in my own treatment of misogyny, I argue that there can be misogyny without misogynists, and that those agents who do deserve the label “misogynist” have to exceed a certain threshold of severity and repetition of misogynistic offenses. “Misogynist” can act as a kind of warning label, then, to highlight the worst offenders within a particular social environment (Manne Reference Manne2017, 65–66). The risk of “vertical inflation” of a kind that might give rise to false accusations is checked by design—and definition.

This leaves Haslam’s worry that defining concepts of prejudice partly from the perspective of the target or victim, rather than the perspective of the perpetrator, introduces a troubling element of subjectivity. Again, I think this worry is at least overstated. Take, again, my concept of misogyny, which makes just such a shift, and holds that misogyny is hostility and hatred girls and women face rather than something boys and men feel, at least in the first instance. I subsequently take misogyny to be, for a rough first pass, the hostile and hateful treatment girls and women face that serves to police and enforce a patriarchal social order (Manne Reference Manne2017, 33–34). I advocate this analysis of the concept for a number of key reasons: first, that the naïve conception of misogyny, according to which it comprises hostility and hatred toward all women, or even women generally, is rarely going to be instantiated in a patriarchal society in which women are socialized to please and serve men. Why would they bite (or hate) the hands that feed and serve them? (Manne Reference Manne2017, 46–48). Moreover, construing misogyny as hatred in men’s hearts makes it not only a psychological oddity but an epistemological quagmire: it is hard to know what someone feels, deep down, short of being their intimate relation or therapist (Manne Reference Manne2017, 44). The naïve concept of misogyny thus deprives women of a morally powerful tool to name the problems facing us: better to reserve the term for a more important concept, I argue.

Crucially, however, what counts on my treatment as hostile and hateful treatment depends upon the “reasonable woman” standard developed by Catharine MacKinnon in the context of sexual harassment law (MacKinnon Reference MacKinnon1979). According to this conception, not just anything goes when it comes to how a certain form of treatment “comes across” to the target or victim thereof: she must be reacting in a reasonable way given the particular details of her social position—including, crucially, her gender (Manne Reference Manne2017, 60–61). It’s not just that women feel intimidated or attacked by seeing pornographic posters pinned up in their workplace, for example; it’s that it is reasonable to feel this way as a woman for whom sexually objectifying imagery is fraught with social meaning.

The upshot of this section is that the charges of over-sensitivity with respect to expanding notions of sexism, misogyny, racism, and the like, as crystallized by the literature on “concept creep,” tend to be unwarranted. Typically, it is less a matter of these concepts being injudiciously expanded as that an artificially narrow concept, delineated by a dominant group member who was wrongly treated as an authority figure on the subject, was recognized as such. Moreover, sometimes concepts should be expanded in response to an increased sensitivity to what oppressed people are actually up against. Finally, fears about moral litigation due to “concept creep” are at least overblown: expanding the concept of racism, sexism, and misogyny in the ways that have been advocated by anti-racist and feminist scholars has characteristically led to a less punitive and individualistic approach to the subject. The “concept cops,” as we might call them, are sounding false alarm bells.

3. Overreactions in feelings

Now to consider the third kind of alleged over-sensitivity to sexism, misogyny, racism, and so on, in which people are accused not of misidentifying instances, or operating with too expansive a concept of the relevant harms, but are instead held to be reacting too strongly to them, emotionally. They are held, in other words, to be “snowflakes.”

I want to emphasize that overreactions are a real thing, and I do not aim here to exonerate—or give carte blanche to—oppressed people accused of them. Still less do I want to do so with respect to people accused of overreacting in general. Take, for example, Brianna’s father-in-law from my example in the first section, who overreacted angrily to a minor moment of conversational frisson. Or take, for a more politically charged case, those whites charged with “white fragility,” to use Robin DiAngelo’s (Reference DiAngelo2018) term. These are white people who react to any attempt to have them face the realities of racism with denialism, incredulity, and, crucially for my purposes here, tears, rage, and so on. One doesn’t have to be on board with every aspect of DiAngelo’s approach—or every exercise promoted by her and her ilk in corporate anti-racism trainings—to recognize the basic truth of her point here: white people often react to discussions of race and racism with excessively big feelings. Hence, any sensible discussion of emotional overreactions in general, and to socially charged subjects in particular, will have to be indexed to substantive normative judgments. There is no getting away from these whenever we have to make calls about which emotions are apt, fitting, warranted, and so on; what one is entitled or even obligated to feel depends, among other matters, on how good or bad things are and to what extent they ought to be or be otherwise. I proceed from the assumption—hopefully uncontroversial among my readers—that sexism, misogyny, and racism are very bad indeed and urgently need remedying.

So return to the moment with which I opened, of Kondabolu’s anger at the remark he misheard at the comedy club. Specifically, he threatened the heckler: “Why don’t you come up here so I can kick your teeth in?” Even if he had been right about this being an instance of racism, calling this an overreaction without further comment will seem natural to many people. But I think it is mistaken.

In her book-length defense of the value of rage, the philosopher Myisha Cherry points out that the insistence that anger or rage be proportionate to the wrong or offense suffered requires of oppressed people that they react perfectly—and keep reacting perfectly, with no more or less emotion than is apt—in every individual instance.Footnote ¹⁷ What is far more common is what happened in Kondabolu’s case: he took racist heckling on the chin umpteen times, and simply ignored it onstage for years, until that one fateful evening. The heckling on that occasion was the proverbial straw that broke the camel’s back. (He was subject to moral bad luck, in that this was the time he misheard the utterance.) I do not think demanding unbreakable fortitude is morally reasonable.

Again, we can distinguish here between two possibilities: one is that someone like Kondabolu is not overreacting: had someone actually said “Palestinian power!” in order to mock him, he would have been justified in feeling fury (especially after keeping his temper about other, similar incidents).Footnote ¹⁸ Alternatively, we might say that Kondabolu was overreacting, but that this overreaction is understandable and not blameworthy in light of all he had been dealing with, for many years, in many contexts.Footnote ¹⁹

I am tempted to think that both responses have it right, but regarding different intentional objects of the agent’s anger.Footnote ²⁰ Considered as a response to this particular racist incident, it is an overreaction, albeit an understandable and not blameworthy one given the social context. But considered as a response to a whole string of similar incidents, it is not an overreaction: it is in fact warranted, even if it is also partly misdirected (or, rather, concentrated on just one of the many people who the agent is ultimately angry with).Footnote ²¹ Lest I be misunderstood here, recall the object of our appraisal: the feeling, not the threatening remark (still less any action undertaken on that basis). That is, I certainly don’t believe that Kondabolu would have been justified in actually kicking someone’s teeth in, or even that he was justified in yelling at them threateningly. (This is a verdict that he himself evidently agrees with: he called his remark the worst and harshest thing he’d ever said to another human being, obviously regretful.) But demonizing his feelings seems to me wrongheaded.

As we saw in section 1, another reason to be careful of demonizing an agent’s over-sensitivity to systemic injustice is that under-sensitivity to these features of the world is a large and common problem. This holds with respect to overreactions to systemic injustice too, partly for the obvious reason that many non-oppressed people are prone to affective under-reactions to misogyny, racism, and so on. But it holds for a more surprising and interesting reason too: such under-reactions are also a predictable outcome of certain forms of social subordination faced by the people subject to oppression.

To see why, consider Kate Hamilton’s harrowing memoir, Mad wife, in which she recounts her experiences of being sexually violated for many years by her husband. She describes acquiescing night after night without true consent, without desire, certainly without pleasure. Her devil’s bargain: “You can fuck me, but you can’t kiss me and I don’t have to pretend to like it.” (Hamilton Reference Hamilton2024, 8). Here is how she describes a typical evening:

Hamilton dragged her husband, despite his extreme reluctance, to therapy. There, mustering all her courage, she confesses what she describes as her deepest, darkest truth: “When we have sex, I feel like I’m being violated.” He shrugged. “She’s always so melodramatic.” Hamilton wasn’t surprised by this response. “It was his standard reaction to my complaints about the sad state of our marriage, his way of training me to see my needs—emotional connection and communication—as excessive, and his (primarily sex) as entirely reasonable.” (Hamilton Reference Hamilton2018).

Several lessons emerge from this case and others like it. First, fairly obviously, accusations of overreacting are frequently weaponized by dominant group members against more vulnerable people, including those over whom they have a great deal of power. This feeds into gaslighting: the insidious process by means of which the target is made to feel defective for harboring beliefs, desires, needs, feelings, and other mental states to which she is entitledFootnote ²² —here, her profound aversion toward this sex she did not want, and her need for the bodily autonomy it took her years to claw her way back to. And although overreactions are of course a live possibility in any complex human domain in which emotions run high and denials go deep, under-reactions are very much a problem too, and may be at least as likely.

This isn’t just a point about other agents who undermine our warranted reactions, and gaslight us in the process. We gaslight ourselves too, I think, in coming to doubt and mistrust our own initial reactions. The popular subreddit “amIoverreacting” contains no shortage of examples of this, where the fact that a woman is not overreacting to a man’s bad behavior can be gleaned just from reading the following titles: “Am I overreacting that my partner keeps getting so drunk he wets the bed?”Footnote ²³ “My brother said I shouldn’t be mean to the man that abused us for years. Am I overreacting for being upset?”Footnote ²⁴ “Am I overreacting [that my] boyfriend keeps putting his fingers inside me as a ‘joke’?”Footnote ²⁵ No, no, and no. The fact that these posters went to reddit to ask their questions tends to suggest that they know at some level that they are not overreacting. They just needed to be emboldened by other posters to feel what they were feeling, and perhaps to take appropriate action on this basis. Sometimes we need other people to tell us that we are in the right, not only morally but emotionally; another’s anger on our behalf can jolt us out of self-doubt and vacillation.

There is also a lesson to be drawn from such cases about the precarity of sensitivity, and the ways people who are repeatedly wronged in serious ways—assaulted, threatened, violated, and so on—may become inured to their own predicament. There are already hints of the numbness Hamilton came to feel in her account of the sex itself: the dissociation in the moment, the “vacating” of her body, the becoming a “machine.” And that coerced disembodiment—a common coping mechanism for victims of sexual abuse and assault—began to suffuse her life in general. “Submitting to unwanted sex with someone who knew it was unwanted made me feel—during the sex and, over time, in general—like a thing, a vessel for my husband’s needs. A sex doll, made wholly of orifices. It made me feel like I was not a person. I no longer existed, not as a complete human being with needs, feelings, opinions, beliefs, agency.” (Hamilton Reference Hamilton2024, 10). Or, more simply: “I felt like a sex doll. I felt unselfed.” (Hamilton Reference Hamilton2018).

Hamilton describes the process of finding her way back to herself partly by getting in touch with a feeling she had long submerged: rage. “Way deep down, where I could not even see or feel it until nearly a decade later, in tandem with my revulsion grew my rage.” (Hamilton Reference Hamilton2024, 9).

Philosophy, with its emphasis on reason and rationality, has historically tended to err in a particular direction, overfocusing on constraining rage and policing other intense feelings. Many feminist theorists have recognized this point, and it has spurred a lively conversation—both inside and outside the academy—in recent years about the way anger and rage may be crucial tools for political mobilization, as well as simply apt responses to injustice and oppression.Footnote ²⁶ But what is perhaps less recognized, and deserves our attention too, is that these emotions can be a precious resource that are hard to preserve. They can become difficult to cultivate in the first place for some people.

In her recent book, To those who have confused you to be a person, the journalist Alia Dastagir recounts the stories of women threatened online, and subject to misogynistic abuse and doxing, often due to their writing and advocacy on issues of social justice. Many of these women have become inured to it, and testify to dissociation in the moment, followed by emotional numbness in the aftermath. Taylor Lorenz, for example, was subject to numerous vicious mobbings for her reporting about the tech world and right-wing sources on social media. “It’s like I cannot feel emotion. I’ve gone through so much that … I feel numb,” she told Dastagir in an interview (Dastagir Reference Dastagir2025, 116). The activist Jaclyn Friedman, a sexual assault survivor who wrote about her own experiences, spoke of the feeling of leaving her body during the online backlash. Disinformation researcher Abbie Richards recalled blacking out during the most intense attacks on her. When Dastagir interviewed me, I told her I wouldn’t be able to help much: as she writes in the section of the book, “Dissociation, desensitization, numbing,” I have almost no memory of the worst times of being subject to misogynistic abuse and vitriol (Dastagir Reference Dastagir2025, 116).

Psychologists vary in their treatment of dissociation; it is sometimes theorized as the thing our brains do to protect us at the worst of times, sometimes as a very ordinary mechanism that explains the feeling of blankness any time we go into “autopilot mode” and lose track of our own movements.Footnote ²⁷ (The sense of having driven a familiar route without really being aware of doing so is a common example of ordinary, benign dissociation in the literature.Footnote ²⁸ ) Whatever the case, it’s worth noting that, when a woman experienced sexual assault, her dissociative symptoms one month later were predictive of subsequent poor psychological outcomes; so, interestingly, was her anger (Feeny et al. Reference Feeny, Zoellner and Foa2000).Footnote ²⁹

As Dastagir points out, it’s impossible to know, without clinical evaluation as well as conceptual clarity, the extent to which dissociation was at stake in these cases and others like them, of misogynistic attacks on women who dared to be outspoken. It may be emotional numbing and cognitive avoidance too. And, as Dastagir argues, all of these responses can be useful coping mechanisms; they can also lead to harms, both short-term and long-term. In the short term, emotional numbness can compromise one’s ability to detect serious threats and dangers. Dastagir interviewed Leah Torres, an OB-GYN who provided abortions in red states, and maintained a Twitter presence to explain her work and advocacy. Torres described the experience of having been brutally maligned and attacked over the years for tweeting things like “Do dead women from forced pregnancy matter?” (This was in 2013, in response to a tweet about dead babies.)

In 2018, someone on Twitter asked Torres if, when she laid down to sleep at night, she could hear the dead babies screaming. She finally snapped: “You know fetuses can’t scream, right? I transect the cord 1st so there’s really no opportunity, if they’re even far enough along to have a larynx.” The response was vitriolic, and ultimately cost her a job in Utah and a temporary medical license in Alabama, where she had moved to practice.Footnote ³⁰ Torres: “You get numb. And then you can’t assess threats correctly … You have to keep going. In order to combat violence, you have to fight your own experiences.”Footnote ³¹ In the long term, such fighting of experience can compromise your ability to experience much of anything. Dastagir: “Emotional numbness is characterized by diminished interest, disconnection, and a lack of emotional response. It can feel like emptiness or a daze.” When you are emotionally numb, you don’t feel the worst things, “but you may not feel good things either. You may not care about things that once mattered to you. You may stop caring about yourself.” (Dastagir Reference Dastagir2025, 117). And such care, is of course, needed for surviving—again, let alone thriving and continuing to analyze, advocate, create. The poet Joy Sullivan recently put it this way: “I’m no stranger to the occasional internet hate, especially as my platform has grown and I’ve had poems take off on Instagram. Over the years, I’ve had to develop a thick skin. I realize fortitude is vital and that, in many ways, I need to toughen up. Yet, that kind of neutrality is tricky to balance against my primary task as a poet—to remain open and bruisable to this world.”Footnote ³²

So to those who worry about emotional overreactions, I would counter: a disposition to respond with dismay to systemic oppression is in fact a precious resource that merits conservation. When this disposition meets the threshold for sufficient reliability, this is just sensitivity, per my definition, which need not be perfect in order to be valuable. But even when it falls somewhat short of this mark, and makes an agent more prone to emotional overreactions than we would like (i.e., she is over-sensitive in this sense), it is also a disposition that protects her from a real risk: becoming an under-reactor due to gaslighting or numbness. And given that we typically need to feel the badness of racism and misogyny in order to fight them effectively, this risk should be guarded against with care. Recalling the opening of this section, my aim is hence not to give carte blanche to people who overreact, or to condemn those who criticize overreactions as necessarily mistaken; a great deal will depend upon the context and the stakes, as well as the frequency and the nature of the overreaction in question. My plea is a considerably more modest one: to simply bear all of this in mind and not respond to overreactions too vehemently or hastily. Ironically, this would be an overreaction.