dissenting in part and concurring in the judgment in part:
Over the objections of Lisa Ocheltree’s counsel, the district court directed a jury of four men and four women to answer a detailed set of special interrogatories about Ocheltree’s sexual harassment claim against Scollon Productions, Inc. The jury answered every question in Ocheltree’s favor and awarded her compensatory damages of $7,280 and punitive damages of $400,000. (The district court later reduced the punitive damages to $42,720, bringing the total judgment against the company in line with the $50,000 cap imposed by 42 U.S.C. § 1981a(b)(3)(A).) Today, the majority reverses the entire judgment. It concludes that Ocheltree suffered (at most) only two incidents of harassment that were sex based: the vulgar song and the incident when Ocheltree’s coworker performed simulated oral sex on a mannequin. It then concludes that, as a matter of law, these two incidents spaced over the eighteen-month period when Ocheltree worked at Scollon Productions were insufficient to create a hostile or abusive work environment. I agree that, under our precedents, the jury’s verdict could not stand if the two incidents identified by the majority were the only evidence of sex-based harassment in the record. I believe, however, that the majority’s conclusion rests on both an overly narrow conception of when harassment is “because of sex” and a failure to read the trial evidence in the light most favorable to Ocheltree. I would uphold the jury’s decision that Ocheltree was subjected to a hostile work environment be*367cause of her sex. Because I would also hold that Scollon Productions had constructive knowledge of the harassment and failed to take effective remedial action, I respectfully dissent from the majority’s decision to reverse the award of compensatory damages. I concur (with some reluctance) in the majority’s judgment that the punitive damages award must be reversed.
I.
Title VII does not protect workers against all forms of verbal and physical harassment in the workplace. It protects only against conduct that is (1) unwelcome, (2) because of sex, and (3) “sufficiently severe or pervasive to alter the plaintiffs conditions of employment and to create an abusive work environment.” Anderson v. G.D.C., Inc., 281 F.3d 452, 458 (4th Cir.2002) (internal quotation marks and citation omitted). In addition, a plaintiff who establishes that her work environment was abusive can only recover if there is some basis on which responsibility for the abusive environment can be imputed to her employer.
In reviewing the district court’s denial of Scollon Productions’s motion for judgment as a matter of law, the first question is how much of the conduct complained of by Ocheltree could be seen by a reasonable jury as “because of sex.” Once the truly sex-based conduct has been identified, the second question is whether that conduct could be seen by a reasonable jury as sufficiently severe or pervasive to create an abusive work environment. My problem with the majority’s analysis is its answer to the first question. I believe that a reasonable jury could identify a much greater amount of the alleged harassment as sex based than the majority would allow. Once the first question is answered properly, a reasonable jury could regard the sex-based conduct complained of by Ocheltree as sufficiently severe or pervasive to create an abusive work environment.
Because I believe that a more detailed and explicit account of the trial evidence is needed to determine how much of the conduct complained of by Ocheltree could reasonably be seen as sex based, I begin my analysis by supplementing the majority’s account of the evidence concerning Ochel-tree’s work environment. Much of what I add is raw, but without it the evidence is not presented in the light most favorable to Ocheltree.
Lisa Ocheltree began working in the production shop at Scollon Productions in February 1994. She was the only female employee in the shop, working alongside “ten or eleven” men. J.A. 110. In contrast, at least twenty women were employed in the “sewing room production area.” J.A. 312.1 Ocheltree testified that the atmosphere in the production shop was “fun” and “friendly” when she first began to work there, but that over the course of the first year of her employment, sexual banter and sexual conduct of other sorts began to occur with increasing frequency. J.A. 111, 113. Her testimony was seconded by Brian Hodge, who started working in the production shop several months after Ocheltree began working there. Hodge testified that the atmosphere seemed okay to him initially but that over time the work environment became increasingly coarse. J.A. 199-200.2 Ochel-*368tree recounted several specific incidents of harassment, the three most prominent of which (the vulgar song, the oral-sex-on-the-mannequin incident, and the body-piercing book incident) are summarized by the majority.
The three incidents mentioned by the majority bear recounting here as part of the whole picture facing Ocheltree. On one occasion a male coworker went up to Ocheltree and sang the following song to her “like he was in the opera”: “Come to me, oh, baby come to me, your breath smells like c[o]m[e] to me.” J.A. 115. To Ocheltree’s chagrin the men in the production shop expressed their enjoyment of the incident with much laughter. Id. On another occasion when Ocheltree arrived at work and proceeded to the time clock, two coworkers were positioned at a nearby mannequin. One was pinching the mannequin’s nipples, and the other was on his knees simulating oral sex on the mannequin. Ocheltree said to the men, “You guys are disgusting. This needs to stop.” As she turned to leave the room, she heard laughter in the background. J.A. 115-17. On the third occasion Ocheltree was seated at her work station, and some of her eoworkers were looking at a book that contained pictures of men with pierced genitalia. One coworker took the book, approached Ocheltree, and opened the book to the centerfold photograph showing a man’s crotch area. The scrotum was pierced with hoops, and there were chains running up to the top of the penis. The coworker, with his male colleagues looking on, said, “Lisa, what do you think about this?” Again, this generated laughter from the men in the shop. J.A. 117-18.
Ocheltree also presented evidence, not discussed by the majority, of other incidents that could reasonably be seen as sex-based harassment. Brian Hodge testified that some of the men who worked in the production shop would “often fondle” the mannequin as they walked by, J.A. 200, and that “anytime [Ocheltree] was walking by just about they would do something sexual to the mannequin in front of her just because they knew it bothered her,” J.A. 202. In addition, Ocheltree testified that Ellery Locklear, vice president of Scollon Productions, berated her for using the phone to check in on her son, who was at home recovering after breaking his tail bone in two places. (Apparently, Ochel-tree’s call violated the company’s telephone usage policy.) According to Ochel-tree, Locklear told her, “I don’t care if someone is dying in your family, you are not to be on the phone and you must be here at work.” J.A. 129. Ocheltree claims that Locklear then said that if she did not like that rule, she ought to go home and be a housewife because she was not cut out for her work at Scollon Productions. Id. A jury could reasonably see this as another piece of evidence suggesting that Ochel-tree experienced harassment because of her sex. Cf. Hartsell v. Duplex Prods., Inc., 123 F.3d 766, 773 (4th Cir.1997) (characterizing question to plaintiff about “whether she would be a ‘mini van driving mommy’ or ‘be a salesperson and play with the big boys’ ” and statement that the plaintiff should “ ‘go home and fetch [her] husband’s slippers like a good little wife’ ” *369as “logically attributable” to the plaintiffs gender).3
Then there is the matter of the sexual banter that Ocheltree claims occurred on a daily basis. The majority describes this banter in only the most general terms, but the majority’s reticence blunts the force of Ocheltree’s case. I will therefore present, in the light most favorable to Ocheltree, an account of the running sexual commentary in the production shop.
Ocheltree presented evidence of several different kinds of inappropriate and unprofessional sexual remarks in her workplace that she views as evidence of sex-based harassment. One category is simply the extensive use of profanity, with many of the words sexually tinged: motherfucker, fuck, faggot, dickhead, pussy, ass, and the like. A second category involves Ochel-tree’s male coworkers’ use of explicitly sexual insults to needle each other. For example, Ocheltree testified that “[gjuys would make hand gestures down at then-private parts and tell other guys to suck it.” J.A. 113. Coworkers sometimes suggested that various male employees were involved in homosexual relationships and that one employee was having sex with a dog. A third category, and the one that strikes me as most significant, is Ochel-tree’s evidence that her coworkers constantly discussed their sexual exploits with their wives and girlfriends in extremely graphic terms. Ocheltree testified that her coworkers would regularly talk about their sexual experiences of the night before “as to that she swallowed, she gave good head, that I fucked her all night long,” etc. J.A. 118. One employee related that his girlfriend “gave good head and that she likes to swallow, that she liked it from behind, that she would do it anywhere with him.” J.A. 120. He further said that she “could suck a golf [ball] through a garden hose.” J.A. 120. In his testimony, Brian Hodge recounted how one of the employees in the production area often “would speak of [his wife] sucking his dick and swallowing and letting it run down the side of her face and stuff.” J.A. 200. Ocheltree testified that she heard remarks along these lines “every day.” J.A. 120. Hodge also testified that he heard such remarks on a daily basis. J.A. 204. Finally, Ocheltree testified that on one occasion, shop supervisor Harold Hirsch said that he was interested in and enjoyed having sex with young boys and that Hirseh’s comments were “purposefully said in front of [her] because [Hirsch and two other production shop employees] enjoyed looking at [her] and seeing [her] reaction.” J.A. 119.
Having set out more fully the evidence concerning Ocheltree’s workplace environment in the light most favorable to her, I turn to the question of whether and to what extent the objectionable conduct was sex based. Under the Supreme Court’s Oncale decision the proper question is “'whether members of one sex are exposed to disadvantageous terms or conditions of employment to which members of the other sex are not exposed.’” Oncale v. Sundowner Offshore Servs., Inc., 523 U.S. 75, 80, 118 S.Ct. 998, 140 L.Ed.2d 201 (1998) (quoting Harris v. Forklift Sys., *370Inc., 510 U.S. 17, 25, 114 S.Ct. 367, 126 L.Ed.2d 295 (1993) (Ginsburg, J., concurring)). The majority acknowledges, at least for the sake of argument, that a reasonable jury could regard the vulgar song and the incident involving simulated oral sex on the mannequin as sex based because they could be seen as “particularly demeaning towards women or as veiled sexual propositions.” Ante at 359. It then reasonably concludes that two incidents over the space of eighteen months are insufficient as a matter of law to constitute an abusive work environment. I suspect that the majority would hold that even if a reasonable jury could regard the body-piercing book incident, the various mannequin fondling incidents described by Hodge, and Locklear’s comment that Ocheltree should go home and be a housewife as sex based, the sum total of these incidents would still be insufficiently severe or pervasive as a matter of law. Though I have my doubts about that conclusion, I am willing to accept it for purposes of argument. To my mind, the majority’s primary mistake is its conclusion that a reasonable jury could not regard any of the day-to-day sexual banter complained of by Ocheltree as sex based. The majority offers several reasons for this conclusion. First, it portrays Ocheltree as a mere bystander to discussions between the men in the production shop and claims that the same kind of talk would have occurred regardless of Ocheltree’s presence there. In addition, it argues that even if some of the sexual banter was a reaction to Ocheltree’s presence in the workplace or her complaints, the alleged harassment was still not sex based because “even if the alleged harassers were intending to bother Ocheltree, there is no evidence that those participating in the offensive conduct were attempting to bother her because of her gender.” Ante at 359. The majority also observes that even though the banter was “sexually explicit” and “generally degrading, humiliating, and even insulting,” it was not “aimed solely at females in any way.” Ante at 358. I find these reasons unconvincing.
In my view, there are two ways in which a reasonable jury could find that much of the sexual banter complained of by Ochel-tree satisfied the “because of sex” prong. First, a reasonable jury could find that much of the banter was “directed at” Ocheltree in the sense that it was intentionally said in her presence in order to make her uncomfortable and self-conscious about her status as the only woman in the production shop. Second, a reasonable jury could find that even if very few of the sexual remarks were made in response to Ocheltree’s presence in the production shop, her male eoworkers’ relentless, graphic descriptions of their sex lives count as sex-based harassment because they portray women as sexually subordinate to men. Ocheltree’s coworkers made her uncomplaining submission to an atmosphere suffused with degrading images of female sexuality an implicit condition of her employment, and this harassment was “because of sex” in the sense that it made the workplace more hostile to Ocheltree precisely because she was a woman.
A.
As to my first point, Ocheltree concedes that most of the offensive conduct was not aimed exclusively at her. Instead, she claims that her coworkers knowingly made sexual remarks and engaged in other behavior with a sexual content in her presence with the intention of making her uncomfortable. Ocheltree presented the following evidence in support of this theory. Ocheltree and Hodge both testified that Ocheltree’s coworkers enjoyed offending her and laughing at her reactions. *371When asked whether the incident involving simulated oral sex on the mannequin was intended to provoke a reaction from her, Ocheltree testified: “Why else would two guys be doing that at the time when I was supposed to be in there going to work ... and then whenever I say all this is disgusting and everyone laughs, who would it be directed to? I was the only female there.” J.A. 117. She also testified regarding Hirsch’s remark about having sex with little boys that “[i]t was purposefully said in front of me because they enjoyed looking at me and seeing my reaction.” J.A. 119. Brian Hodge testified that “[ajnytime [Ocheltree] was walking by just about they would do something sexual to the mannequin in front of her just because they knew it bothered her.” J.A. 202. In addition, both Ocheltree and Hodge testified that sexual talk and conduct in the production area first became a problem after Ocheltree began working at Scollon Productions. Hodge further testified that the amount of sexual conduct and talk escalated considerably after Ocheltree complained about it at a safety meeting for production shop employees. J.A. 202-OS. A reasonable jury could infer from this evidence that a considerable amount of the sexual talk and behavior in the production shop was intended, at least in part, to make Ocheltree uncomfortable and to provoke reactions from her.4
The majority also suggests that even if Ocheltree’s coworkers intended to bother her, “there is no evidence that those participating in the offensive conduct were attempting to bother her because of her gender.” Ante at 359. The majority does not elaborate, but I take its argument to be that Ocheltree’s evidence could at most show that she was harassed not because she was a woman, but simply because she was offended by and objected to working in an environment saturated with sexually explicit remarks. Cf. Hopkins v. Baltimore Gas & Elec. Co., 77 F.3d 745, 751 (4th Cir.1996) (“Title VII does not reach discrimination based on ... [an employee’s] sexual behavior, prudery, or vulnerability.”). Perhaps the majority is correct to suggest that Ocheltree’s argument fails if her coworkers harassed every employee — male or female — who was offended by or objected to the sexual talk and antics in the production shop. Such proof would suggest that Ocheltree was targeted not because of her sex, but simply because of her “prudery,” her sensitivity to the sexual remarks.5 The question, though, is how much proof Ocheltree has to produce before a reasonable jury could conclude that she was harassed because she was a prude and a woman, not simply because she was *372a prude. Here, Ocheltree was the only woman in a working environment with ten or eleven males who engaged in sexual talk and behavior in order to make her uncomfortable. Further, the majority observes that men sometimes complained about the sexual tenor of the workplace, but the jury heal’d no evidence that any conduct subsequent to these complaints was intended to bother the complaining men. Nor was there evidence that their objections or reactions were the subject of derisive laughter in the way that Ocheltree’s were. In light of these points, I think a reasonable jury could see the harassment as rooted in male resentment of Ocheltree’s intrusion into “their” workplace and in resentment of her demands that they clean up their act. Cf. Kathryn Abrams, The New Jurisprudence of Sexual Harassment, 83 Cornell L.Rev. 1169, 1211 (1998) (observing that when women enter a predominantly male workplace, male workers often seek to reaffirm the dominance of masculine norms in the workplace by “engag[ing] more intensely ... in talk that sexualizes or derogates women”). In other words, a reasonable jury could conclude that Ochel-tree was harassed not simply because she found the sexual behavior in the workplace offensive, but because she was a woman who found that behavior offensive. In sum, I believe a reasonable jury could conclude that much of the sexual banter in Ocheltree’s workplace was directed at her because of her sex.
B.
What I have just said in part I.A is sufficient by itself to support the affir-mance of Ocheltree’s compensatory damages award. There is, however, another reason why a reasonable jury could conclude that a large part of the sexual banter in the workplace satisfied the “because of sex” prong. A reasonable jury could find that the content of much of the banter was “particularly demeaning towards women,” ante at 359, and therefore made Ochel-tree’s working environment more hostile to her as a woman, regardless of whether the banter was intended to bother her or was directed at her in any other way. This means that even if I agreed with the majority that a reasonable jury would have to find that Ocheltree was a mere bystander to the sexual remarks in her workplace, I would still conclude that many of the remarks could reasonably be seen as satisfying the “because of sex” prong.
Title VII’s because-of-sex requirement is most obviously satisfied in two common scenarios: when the plaintiff is the object of unwanted sexual advances, see, e.g., Harris v. L & L Wings, Inc., 132 F.3d 978 (4th Cir.1997), and when the plaintiff is the target of open hostility because of her (or his) sex, see, e.g., Smith v. First Union Nat'l Bank, 202 F.3d 234 (4th Cir.2000). We have never held, however, that these two scenarios exhaust the field of sex-based harassment. Courts have also recognized that harassing conduct can be “because of sex” even when the conduct “is not directed at a particular individual or group of individuals, but is disproportionately more offensive or demeaning to one sex.” Robinson v. Jacksonville Shipyards, Inc., 760 F.Supp. 1486, 1522-23 (M.D.Fla.1991). See also Andrews v. City of Philadelphia, 895 F.2d 1469, 1486 (3d Cir.1990) (stating that “we do not consider it an unfair burden of an employer of both genders to take measures to prevent an atmosphere of sexism ... [from pervading] the workplace”). This category of sex-based harassment “describes behavior that creates a barrier to the progress of women in the workplace because it conveys the message that they do not belong, that they are welcome in the workplace only if they will subvert their identities to *373the sexual stereotypes prevalent in that environment. That Title VII outlaws such conduct is beyond perad-venture.” Robinson, 760 F.Supp. at 1523. In Robinson the court held that a workplace plastered with pictures of nude and partially nude women (often in sexually submissive postures) was a hostile environment even though the posting of the pictures “did not originate with the intent of offending women in the workplace (because no women worked in the jobs when the behavior began).” Id. It was enough that the pictures had a “disproportionately demeaning impact on the women now working” in the same environment. Id.
Robinson and Andrews suggest that the majority is too quick to conclude that none of the sexual banter in the workplace constitutes sex-based harassment because Ocheltree “would have been exposed to the same atmosphere had she been male.” Ante at 357. Even if true, the majority’s point would still leave the question of “whether members of one sex are exposed to disadvantageous terms or conditions of employment to which members of the other sex are not exposed.” Harris, 510 U.S. at 25, 114 S.Ct. 367 (Ginsburg, J., concurring). In other words, there would still be a question of whether Ocheltree’s environment was more hostile to her because of her sex than it would have been to a man.
Although our circuit has not yet decided a case using reasoning similar to that in Robinson and Andrews, such reasoning is a logical extension of well-established Title VII principles. Consider, for example, a traditionally all-male workplace in which women are routinely referred to as “bitches,” “whores,” and “cunts.” Surely it is uncontroversial that a woman employee subjected to that environment would have a claim for sex-based harassment even if the employer could establish that the male workers would have spoken the same way regardless of her presence and even if the offensive words were never said directly to her. The reason is that such “unambiguous gender epithets,” ante at 359, signal hostility to the presence of women in the workplace and create an atmosphere that is inhospitable to women because of their sex. Whether or not plastering pornographic and quasi-pornographic pictures on the wall can be said to display outright hostility toward women in the workplace, it can certainly be said to evince and to perpetuate attitudes that make the workplace hostile to women because of their sex:
Pornography on an employer’s wall or desk communicates a message about the way he views women, a view strikingly at odds with the way women wish to be viewed in the workplace. Depending upon the material in question, it may communicate that women should be the objects of sexual aggression, that they are submissive slaves to male desires, or that their most salient and desirable attributes are sexual. Any of these images may communicate to male coworkers that it is acceptable to view women in a predominantly sexual way. All of the views to some extent detract from the image most women in the workplace would like to project: that of the professional, credible coworker.
Robinson, 760 F.Supp. at 1526 (quoting Kathryn Abrams, Gender Discrimination and the Transformation of Workplace Norms, 42 Vand. L.Rev. 1183, 1212 n. 118 (1989)). See also Andrews, 895 F.2d at 1485-86 (“Obscene language and pornography quite possibly could be regarded as highly offensive to a woman who seeks to deal with her fellow employees and clients with professional dignity and without the barrier of sexual differentiation and abuse” (internal quotation and citation omitted)). If use of unambiguous gender epithets can constitute sex-based harassment because it *374creates a workplace atmosphere suffused with hostility to women, there is no principled reason why a workplace suffused with depictions of women as sexual objects could not also constitute sex-based harassment. Thus, I conclude that a workplace environment could be abusive “because of’ a plaintiffs sex even if the environment was essentially the same both before and after the plaintiff entered the workplace. Further, I can see no reason why the reasoning of Robinson and Andrews should be limited to sexually explicit photographs of women in the workplace. Sexual banter in the workplace that is the aural equivalent of pornography should surely be just as actionable as pornographic images. Here, the question is whether a reasonable jury could conclude that any of the sexual banter in Ocheltree’s workplace was so disproportionately demeaning to women that it should count as harassment “because of sex.” This question can only be answered by considering the content and context of the banter in question.
Although my reasoning is different, I agree with the majority that much of the sexual talk Ocheltree heard — vulgar though it may be — cannot be seen as disproportionately demeaning to women. Frequent use of such words as “fuck,” “dickhead,” and “ass” may be unprofessional and offensive, but I cannot say that a working environment permeated by such language is more offensive to women because of their sex than it is to men. Similarly, frequent exchanges of mock homosexual taunting (for example, male coworkers pointing to their genitals and telling other males to “suck it,” J.A. 113) would be distasteful to most women (and, I believe, to most men), but they would not necessarily make the work environment more hostile to women because of their sex.6 Ocheltree’s coworkers’ constant descriptions of their sexual exploits, including their near-obsessive interest in discussions of oral sex, are another matter altogether. Obviously, discussions of sexual matters (including discussions of oral sex) are not automatically demeaning to women. See Vicki Schultz, Reconceptualizing Sexual Harassment, 107 Yale L.J. 1683, 1795 (1998) (citing research suggesting that “where men and women work alongside each other in balanced numbers .... sexual talk and joking occurs with frequency, but is not experienced as harassment”). But the tone of the discussions in the production shop was hardly one of mutuality and respect. See supra at 369 (recounting daily remarks by Ocheltree’s coworkers along the lines of “she gave good head,” “she likes to swallow,” “she likes it from behind,” etc.). Comments like these portray women as sexually subordinate to men; indeed, it is not too strong to say that the overall ten- or of the workplace banter conveyed the message that women exist primarily to gratify male desires for oral sex. A reasonable woman would find this message offensive, to say the least. Further, the comments were far too graphic and frequent to be dismissed as “occasional vulgar banter, tinged with sexual innuendo.” Perry v. Harris Chernin, Inc., 126 F.3d 1010, 1013 (7th Cir.1997) (internal quotation marks and citation omitted). Instead, they are instances of sexual harassment because they express and reinforce a regime of gender hierarchy in which men are portrayed as sexual subjects while women are portrayed as sexual objects. See generally Abrams, The New Jurisprudence, supra at 1205-25; Katherine M. Franke, What’s Wrong with Sexual *375Harassment?, 49 Stanford L.Rev. 691, 762-72 (1997). It is true, as Scollon Productions points out, that Ocheltree’s coworkers did not proposition her or speculate about her sexual habits. But this is hardly dispositive. See, e.g., Jackson v. Quanex Corp., 191 F.3d 647, 660 (6th Cir.1999) (stating that “offensive comments need not be directed at a plaintiff in order to constitute conduct violating Title VII”). When a workplace is suffused with representations of women as sexual objects, a woman in that workplace would doubtless wonder whether the primary questions about her in the minds of her coworkers involved such matters as whether she “swallows” or whether she could “suck a golf ball through a garden hose.” The demeaning portrayals of women as sexual objects in Ocheltree’s workplace constituted sex-based harassment because they made the working environment more hostile to Ocheltree than to her coworkers precisely because she was a woman. That is enough to satisfy Title VIPs “because of sex” prong, for the “critical issue ... is whether members of one sex are exposed to disadvantageous terms or conditions of employment to which members of the other sex are not exposed.” Harris, 510 U.S. at 25, 114 S.Ct. 367 (Ginsburg, J., concurring).
I would hold, then, that a reasonable jury could conclude that the pervasive workplace comments depicting women as sexually subordinate to men constitute harassment “because of sex.” These comments are indistinguishable in principle from the harassment discussed in Robinson because they are “disproportionately ... offensive or demeaning to one sex.” Robinson, 760 F.Supp. at 1522-23.
C.
The majority mounts a strident attack on this last conclusion, and the attack boils down to three basic arguments. First, the majority argues that my position is incompatible with Supreme Court and circuit precedent regarding the “because of sex” requirement. Second, the majority asserts that my arguments depend on the untenable proposition that all sexual comments in the workplace are “disproportionately demeaning to women.” Third, the majority argues that under our decision in Lack v. Wal-Mart Stores, Inc., 240 F.3d 255 (4th Cir.2001), Ocheltree did not experience sex-based harassment because the production shop environment was also offensive to men. I will address each argument in turn.
First, according to the majority, it is well established that harassing conduct can only be “because of sex” if the plaintiffs gender is the “but for” cause of the harassment or the harassment is motivated by the plaintiffs gender. The majority therefore concludes that the comments and conduct on which I rely cannot constitute harassment “because of sex” because “Ocheltree would have been exposed to the same atmosphere had she been male.” Ante at 357. I acknowledge that courts regularly explain the “because of sex” requirement by using the formulations favored by the majority, and in many contexts these formulations are helpful analytically. I believe, however, that the “because of sex” requirement allows for more interpretive flexibility than the majority recognizes. Cf. David S. Schwartz, When Is Sex Because of Sex? The Causation Problem in Sexual Harassment Law, 150 U. Pa. L.Rev. 1697, 1781 (2002) (suggesting that “ ‘because of does not necessarily mean ‘motivated by’ ”). Indeed, I believe this flexibility is inherent in Oncale’s formulation of the “because of sex” requirement as turning on the question of “whether members of one sex are exposed to disadvantageous *376terms or conditions of employment to which members of the other sex are not exposed.” Oncale, 523 U.S. at 80, 118 S.Ct. 998 (quoting Harris, 510 U.S. at 25, 114 S.Ct. 367 (Ginsburg, J., concurring)). As the majority would no doubt observe, there is a sense in which both male and female workers were exposed to the same environment at Scollon Productions because the explicit discussions of oral sex and similar matters were heard by both Ocheltree and her male coworkers. Yet there is an equally obvious sense in which women in an atmosphere saturated with remarks demeaning to women are “exposed to disadvantageous terms or conditions of employment to which members of the other sex are not exposed” precisely because the language heard by both women and men is more demeaning to women than to men. It is this understanding that I rely upon in suggesting that the workplace comments portraying women as sexually subordinate to men qualify as harassment “because of sex.”
By reasoning that harassment cannot be “because of sex” if the plaintiff “would have been exposed to the same atmosphere had she been male,” the majority appears to commit this court to a course that is clearly wrong. Suppose, for example, that an African-American plaintiff brings a race discrimination claim alleging a hostile work environment due to his coworkers’ daily use of the meanest racial slur against African-Americans. Suppose further that the workplace had previously been all white and that the pattern of racial slurs was the same both before and after the plaintiffs arrival. The majority’s reasoning suggests that if the employer could show that none of the racial slurs were directed at the plaintiff and that he would have been exposed to exactly the same language if he had been white, the harassment in this example could not be “because of race.” Yet I find it difficult to believe that any court would fail to find race-based harassment on these facts. If the right to be free from a racially hostile work environment means anything at all, surely it includes the right to be free from a workplace environment permeated by racial slurs. See Spriggs v. Diamond Auto Glass, 242 F.3d 179, 184 (4th Cir.2001) (“Evidence of a general work atmosphere therefore — as well as evidence of specific hostility directed toward the plaintiff is — • an important factor in evaluating [a hostile environment] claim.” (quoting Hicks v. Gates Rubber Co., 833 F.2d 1406, 1415 (10th Cir.1987))). I do not see how Oncale could compel a different conclusion. My explanation for this is simple: the plaintiff in my example suffers discrimination “because of race”- because he is exposed to disadvantageous conditions of employment to which his white coworkers are not exposed. The workplace is therefore more hostile to him precisely because he is black. My example shows, I believe, that harassment can be “because of race” even if the plaintiff would have been exposed to the same atmosphere had he been white. If this is true, harassment can also be “because of sex” even if the plaintiff would have been exposed to the same atmosphere had she been male.
In further support of its first argument, the majority makes the interesting suggestion that my approach to the “because of sex” requirement wrongly attempts to apply disparate impact models of proof to sexual harassment claims. It is important to see why this is not so. As Professor Schwartz points out, disparate impact theory is designed to address “facially neutral employment practices that have a disparate impact (such as testing instruments or height-weight requirements) but that might be defensible under some degree of business necessity.” Schwartz, supra at 1773. Workplace conversations depicting *377women as sexually subordinate to men, on the other hand, are not facially neutral and they have no business justification. When women are characterized in this way, it is wrong to suggest that offense to women in the workplace is a “potential unintended effects” of the characterizations. Ante at 360. Instead, such characterizations constitute sexual harassment because they convey the message that women are, and should be, subordinate to men.
The majority’s second argument is that under my proposed definition of sex-based harassment, any discussion of sex in the workplace constitutes sexual harassment because such discussions are necessarily more demeaning to women than to men. See ante at 362 (stating that my reasoning would mean that “conversations between males about their heterosexual activities in the presence of a female virtually always constitute sex-based harassment because, according to [my] characterization, these conversations depict women as ‘sexually subordinate to men’ ”); ante at 363 n. 12 (stating that I propose “a rule whereby a reviewing court would affirm a finding of gender-based discrimination once a certain number of conversations about hetero-sexual behavior occur in the workplace in the presence of a female”). The majority observes that the rule it attributes to me would stand in some tension with the Supreme Court’s observation in Oncale that it has “never held that workplace harassment, even harassment between men and women, is automatically discrimination because of sex merely because the words used have sexual content or connotations.” Oncale, 523 U.S. at 80, 118 S.Ct. 998. Further, the majority implies that I am ignorant of the “feminist literature” suggesting that “working women can be, and usually are, as comfortable as are men with sexually explicit conduct and conversations.” Ante at 364. Indeed, my views are “paternalistic,” ante at 364 and 365, rely on “outdated stereotypes,” ante at 364, and attempt to transform Title VII into a “neo-Victorian chivalry code designed to protect what [I] imaginef ] to be the tender sensitivities of contemporary women,” ante at 365. See also ante at 363 n. 12 (stating that under my approach “Title VII would become a workplace code for ‘gentlemanly conduct’ in the presence of women”).
My first response to all of this upbraiding is that if I actually held the views attributed to me by the majority, the substance of its criticisms would be largely justified. It would, of course, be absurd for me to contend that all discussions of sex — regardless of their specific content— are automatically degrading and offensive to women. I am well aware of feminist criticisms of the idea that sex equals sexism in the workplace. See, e.g., Franke, supra at 714-25. To make the point once more, I do not claim that all discussions of sexual matters in the workplace are automatically demeaning to women or that work-place sexual discussions are always more offensive to women than to men. Instead, I claim that Ocheltree’s coworkers’ constant, graphic descriptions of oral sex and other sexual practices could be seen by a reasonable jury as sex-based harassment because they portray women as sexually subordinate to men and thereby serve to perpetuate gender hierarchies in the workplace. It is the specific content of these descriptions (“she gave good head,” “she swallows,” etc.), and not simply their sexual nature, that qualifies them as harassment “because of sex.” This is one of the reasons why I have found it necessary to set out Ocheltree’s evidence about the atmosphere in the production *378shop in explicit detail.7 In short, the majority’s second argument simply attacks a straw man.8
The majority’s third argument is that under our decision in Lack, Ocheltree’s sexual harassment claim must fail because some men complained about the work environment in the production shop. See ante at 358 (quoting Lack’s statement that the plaintiffs claim was “undercut[ ] ... to a substantial extent” by the “fact that female employees ... lodged similar complaints” against the male plaintiffs supervisor).9 Lack should not be read as broadly as the majority suggests. The statement the majority takes from Lack was made in the context of examining whether the male plaintiff had offered direct comparative evidence about how his alleged harasser (a male supervisor) treated both men and women in the workplace. See Lack, 240 F.3d at 262; cf. Oncale, 523 U.S. at 80-81, 118 S.Ct. 998 (observing that harassing conduct that would not otherwise qualify as sex based can be shown to constitute discrimination “because of sex” through the use of “direct comparative evidence about how the alleged harasser treated members of both sexes in a mixed-sex workplace”). My argument here does not rely on comparative evidence. Instead, I am claiming that workplace comments portraying women as sexually subordinate to men satisfy the “because of sex” requirement because they express and reinforce gender hierarchy in the workplace. Such comments are, in this respect, analogous to Oncale’s “sex-specific and derogatory terms” that signal “hostility to the presence of women in the work-place.” Oncale, 523 U.S. at 80, 118 S.Ct. 998. If this theory is correct — and I recognize, of course, that the majority rejects it — I do not see how Ocheltree’s claims can be defeated simply by observing that some men at Scollon Productions also complained about the sexual conduct in the production area.10 The Second Circuit has explained *379that harassment of a woman can be sex based if she is “abused in ways that cannot be explained without reference to her sex, notwithstanding the fact that a man received treatment at least as harsh, though for other — non-sexual—reasons.” See Brown, 257 F.3d at 254. Along the same lines, our analysis in Lack seems to indicate that the sexual conduct complained of by some of Ocheltree’s male coworkers here would not constitute harassment of them “because of sex.” See Lack, 240 F.3d at 261 n. 8 (noting that “when expressions such as ‘fuck me’ ... are used by men speaking to other men, often’ their use has no connection whatsoever with the sexual acts to’ ” (quoting Johnson v. Hondo, Inc., 125 F.3d 408, 412 (7th Cir.1997))). It follows that if I am right in my basic contention that the workplace comments depicting women as sexually subordinate to men could be regarded by a reasonable jury as discrimination “because of sex,” the fact that some male coworkers were also offended by the comments does not undermine Ocheltree’s case.
D.
Because I would hold that much of the sexual discussion in Ocheltree’s workplace constituted sex-based harassment, my approach to the “severe or pervasive” prong of the hostile environment analysis is quite different than the majority’s. For me, the proper question is whether a reasonable jury could find that, taken together, the vulgar song, the various mannequin incidents, and the workplace banter involving degrading descriptions of female sexuality were sufficiently severe or pervasive to alter the conditions of Ocheltree’s employment and to create an abusive work environment. There is no question that Ochel-tree subjectively perceived the production shop environment as hostile. In deciding whether a jury could also reasonably conclude that the environment was objectively hostile (hostile to “a reasonable person in the plaintiffs position,” Oncale, 523 U.S. at 82, 118 S.Ct. 998), we are required to consider “all the circumstances,” including “the frequency of the discriminatory conduct; its severity; whether it is physically threatening or humiliating, or a mere offensive utterance; and whether it unreasonably interferes with an employee’s work performance.” Harris, 510 U.S. at 23, 114 S.Ct. 367. The Supreme Court has emphasized that an abusive work environment need not “lead[ ] to a nervous breakdown.” Id. at 22, 114 S.Ct. 367. It is enough if the environment “detract[s] from employees’ job performance, discouragefs] employees from remaining on the job, or keep[s] them from advancing in their careers.” Id. Applying these standards here, I readily conclude that a reasonable jury that fully credited Ocheltree’s evidence could find that a reasonable person in her circumstances would experience the work environment as abusive. Ocheltree and Hodge both testified that the explicit and degrading discussions of oral sex and related matters took place on a daily basis. While the relentless sexual remarks were not physically threatening, they were humiliating to women generally and highly offensive to Ocheltree. They surely made it more difficult for her to do her job.
E.
My primary objection to the majority opinion is that it has turned the “because of sex” requirement into an obstacle where it had not been an obstacle before, thereby making it more difficult to establish a sex*380ual harassment claim. I have an additional point, however. There appears to be a profound difference in our respective approaches to reviewing a jury verdict. Again and again, the majority characterizes the evidence in a light more favorable to Scollon Productions than to Ocheltree. This reaches its apex when the majority suggests that there is “substantial evidence” that Ocheltree’s own offensive conduct contributed to the coarse atmosphere in the production shop. See ante at 364-65. Ocheltree specifically denied the vulgar activity attributed to her by the majority, J.A. 147-48, and the jury believed her. Indeed, it appears that the jury believed virtually everything that Ocheltree and her witnesses said and dismissed the testimony of Scollon Productions’ witnesses as unworthy of belief. That was the jury’s prerogative, and we are bound to respect it.
For all of these reasons, I respectfully dissent from the majority’s holding that the evidence presented at trial was insufficient to support the jury’s decision that Ocheltree was subjected to a hostile environment because of her sex.
II.
This conclusion makes it necessary for me to indicate, albeit briefly, how I would resolve the questions in this case that the majority has no need to reach. On the question of whether liability for the hostile work environment could be properly imputed to Scollon Productions, I conclude that a reasonable jury could find that the company had constructive knowledge of the harassment because it failed to establish adequate procedures for receiving sexual harassment complaints. Accordingly, I would affirm the jury’s modest award of compensatory damages to Ocheltree. I concur in the majority’s judgment reversing the award of punitive damages because Ocheltree failed to produce evidence that would allow a reasonable jury to conclude that any Scollon Productions employee “discriminated in the face of a known risk that his conduct [would] violate federal law.” Anderson, 281 F.3d at 460.
. The record does not indicate whether other women worked in the production shop either before or after the period when Ocheltree worked there.
. The majority points to uncontested evidence that the atmosphere in the production shop was essentially the same before Ocheltree began working there as it was at the beginning of her employment. Ante at 357. This is irrel*368evant. The crucial point is this: Ocheltree’s evidence established that the behavior she complained of worsened at some point after her arrival, and it became still more objectionable in the wake of her complaints. This evidence supports Ocheltree's claim that the harassing behavior was targeted at her as the only female in the production shop, thus belying the majority's assertion that the jury heard uncontested evidence that Ocheltree "would have been exposed to the same atmosphere had she been male.” Ante at 357.
. The majority criticizes my reliance on this incident on the ground that the special interrogatories submitted to the jury were limited to questions about the conduct of Harold Hirsch, the shop supervisor, and Ocheltree’s coworkers. Ante at 356-57 n. 3. Yet Locklear was the vice-president of the company, and his actions are relevant to the ultimate question of whether Scollon Productions can be held liable to Ocheltree for the creation of a hostile work environment. See Andrade v. Mayfair Mgmt., Inc., 88 F.3d 258, 261 (4th Cir.1996) (stating that harassing behavior by a corporate officer will be deemed that of the employer).
. The majority understates the evidence that the harassing behavior was intended to bother Ocheltree by mentioning only Hodge's testimony that Ocheltree’s coworkers would "do something sexual to the mannequin'' whenever Ocheltree walked by "just because they knew it bothered her.” J.A. 202. The majority then dismisses this testimony because Hodge stated that he was only speculating about the motives behind the mannequin incidents. The majority places more weight on Hodge’s characterization than it will bear. A reasonable jury could conclude that by characterizing his testimony as "speculation," Hodge simply acknowledged that he was making an inference when he said that Ochel-tree’s coworkers fondled the mannequin in order to bother her. Evidence in the form of inferences about the motives of other people is a common feature of Title VII cases, and there is nothing problematic about Hodge's testimony in this regard. See, e.g., Gossett v. Oklahoma ex rel. Bd. of Regents for Langston Univ., 245 F.3d 1172, 1179 (10th Cir.2001).
. I note that the word “prudery” seems misplaced here. Prudery connotes an artificially refined sense of delicacy about sexual matters. A person would hardly need to be prudish to find the atmosphere in Ocheltree’s workplace offensive.
. I would add, however, that such language could be sex-based harassment in the proper context — if, for example, it was used as part of a general campaign to create a sexually explicit atmosphere in the work-place in order to drive out women employees.
. In further response to the majority’s charge that I have relied "upon the shock value of the salacious conduct” of Ocheltree's coworkers, ante at 360, I would add this: if we are going to overturn a jury verdict that so strongly favored Ocheltree, we ought at least to present the evidence in a way that gives some insight into why the jury might have reached the conclusions it did.
. I also add two comments about the majority's efforts to tag me as paternalistic and neo-Victorian. First, if one of the majority's goals is to avoid paternalism in the Title VII field, that goal would be better served by focusing on whether the sexual talk was welcome or unwelcome than by adopting an unduly narrow reading of the "because of sex” requirement. Cf. Schwartz, supra at 1756-58. Second, if there is anything neo-Victorian lurking in the opinions released today, it is the majority's suggestion that Ocheltree's claim should fail because "the offensive conduct [here] took place in a costume production shop where public access is controlled, not a church office, retail shop, bank, or professional office.” Ante at 363. In other words, women in blue collar jobs must put up with conduct that women who work in banks or professional offices need not tolerate. Cf. Williams v. Gen. Motors Corp., 187 F.3d 553, 564 (6th Cir.1999) ("Surely women working in the trades do not deserve less protection from the law than women working in a courthouse.”).
. The majority claims that Ocheltree "conceded that the [sexual] conduct [in the production shop] was equally offensive both to men and women.” Ante at 358. I do not believe the record supports this characterization. So far as I can tell, Ocheltree simply acknowledged that some of her male coworkers also voiced some complaints about sexual conduct in the production shop. This does not amount to a concession that men were equally offended. Cf. Brown v. Henderson, 257 F.3d 246, 254 (2nd Cir.2001) (stating that discrimination can be sex based if "a coworker or supervisor treats both men and women badly, but women worse”).
. The majority is not specific about the subject of the men’s complaints, but I assume the majority’s point is that some men in the pro*379duction shop were offended by taunting from male coworkers that included explicit or implicit references to homosexuality (men pointing to their genitals and telling other men to "suck it,” etc.).