When it is suggested that girls dress provocatively to attract men, we are often told that girls dress that way "for themselves" or "for other women". If girls dress "for themselves", or girls dress "for other women", men are supposedly not in the picture and their preferences are not being considered.
While we already have reason to suspect that this is not true due to expressed and revealed sartorial preferences across the ovulatory cycle and even more so because women judge other women who wear revealing clothes more negatively, I have found the most direct refutation of this convenient claim:
Abstract: "Women’s intrasexual competition has received significant attention only in the last decades, with even less work investigating women’s defenses against such aggression. Yet, we should expect that women can (a) grasp which perceptually-salient cues evoke same-sex aggression and (b) strategically damp the display of (some of) those cues when aggression risk is greatest, thereby avoiding the potentially high costs of victimization. Women selectively aggress against women displaying cues of sexual permissiveness (e.g., revealing dress) and/or desirability (e.g., physical attractiveness). We find that (a) women (and men) anticipate greater intrasexual aggression toward women dressed revealingly versus modestly, especially if targets are attractive. Employing behavioral and self-report measures, we also find (b) women create outfits baring less skin, select more modest clothing, and intend to dress less revealingly to encounter other women, flexibly damping permissiveness cues depending on individual features (physical attractiveness) and situational features (being a newcomer) that amplify aggression risk."
Select quotes:
"We should expect that women can (a) grasp which cues/signals evoke same-sex aggression and (b) strategically damp (some of) those cues/signals when aggression risk is heightened, thereby avoiding the potentially high costs of victimization. We test these propositions here.
Research on women’s intrasexual aggression reveals it is both effective and strategically deployed (e.g., Campbell, 2002; Vaillancourt, 2013). Women may be especially likely to employ social exclusion and indirect tactics of aggression (e.g., gossip) toward other women, and evidence suggests that these tactics may be particularly harmful to female victims (Benenson et al., 2013; Campbell, 1995, 2002; Hess & Hagen, 2006; Hess & Hagen, 2017; Vaillancourt, 2013). For example, compared to males, females may experience greater internalizing problems and suicidal ideation as a consequence of victimization (see Vaillancourt, 2013).
Importantly, women are strategic in deploying this aggression, implying that some women may face disproportionate victimization. This is particularly apparent in the context of women’s mating competition, wherein evidence strongly suggests that women who are (perceived as) sexually permissive and/or sexually desirable are considered especially threatening rivals and are disproportionately aggressed against. That is, women specifically target other women who are (perceived as) sexually permissive (e.g., revealingly dressed women; Baumeister, Reynolds, Wineguard, & Vohs, 2017; Baumeister & Vohs, 2004; Bleske & Shackelford, 2001; Gurung & Chrouser, 2007; Muggleton, Tarran, & Fincher, 2018; Pazda, Prokop, & Elliott, 2014; Reynolds, Baumeister, & Maner, 2018; Vrangalova, Bukberg, & Rieger, 2014). For example, Vaillancourt and Sharma (2011) found that women behaved more aggressively toward a female confederate when she was dressed provocatively than when that very same confederate was dressed modestly. Women also specifically target other women who are (perceived as) sexually desirable to men (e.g., physically attractive and/or fertile women; Arnocky, Sunderani, Miller, & Vaillancourt, 2012; Buunk, Zurriaga, Gonzalez-Navarro, & Monzani, 2016; Fink, Klappauf, Brewer, & Shackelford, 2014; Hurst, Alquist, & Puts, 2017; Krems, Neel, Neuberg, Kenrick, & Puts, 2016; Leenaars, Dane, & Marini, 2008; Maner & McNulty, 2013; Reynolds et al., 2018). For example, wearing red makes women more attractive to men (e.g., Elliot, Greitemeyer, & Pazda, 2013; Pazda, Elliot, & Greitemeyer, 2012), and women are more aggressive toward women in red (Pazda et al., 2014).
Are women aware that displaying certain cues—cues of sexual permissiveness (e.g., revealing clothing) and/or desirability (e.g., physical attractiveness)—increases their risk of incurring intrasexual aggression? To avoid incurring this aggression, might women strategically damp the display of such cues when the possible costs (e.g., intrasexual aggression) outweigh the possible benefits (e.g., mate attraction)? First, we test whether people anticipate that women displaying cues associated with permissiveness and/or desirability will face more intrasexual aggression. Next, because women can more easily manipulate their clothing than their physical attractiveness, we investigate whether women might dress defensively—strategically damping sartorial cues of permissiveness when the threat of intrasexual aggression is heightened—presumably to avoid the potentially high costs associated with intrasexual victimization.
Experiment 1
Women aggress more toward more revealingly dressed women; for example, an attractive female confederate elicited more intrasexual aggression when provocatively dressed than when that same confederate was modestly dressed (Vaillancourt & Sharma, 2011). To our knowledge, little work tests the straightforward prediction that people expect more intrasexual aggression to be directed toward revealingly versus modestly dressed women. We test this prediction and explore the role of target attractiveness...
Participants were asked to report their expectations of a social interaction in which Sara (the target) met Carol (a prospective new friend) and Martha (Carol’s coworker) at a Starbucks. Participants were informed that Sara was new to town and had connected with Carol through a friend-finding phone application. Carol was on the app, looking to make friends outside of work...
Because some forms of indirect aggression can occur face-to-face (e.g., acting dismissive), whereas others, such as social distancing (e.g., avoiding the target after a meeting), may be apparent only in post-meeting behavior, we divided indirect aggression into indirect aggression (at the meeting) and behavioral intentions to socially distance the target (after the meeting). Social distancing often occurs when women do not intend to welcome a target into their, their friends’, and/or their romantic partners’ social orbits (Bleske & Shackelford, 2001; Krems et al., 2016).
Participants were randomly assigned to view one of three versions of Sara, each depicting the same woman but differing in dress and body weight (Figure A1). The images of Sara are established stimuli created and used by Vaillancourt and Sharma (2011), who found that they evoked different levels of intrasexual aggression. In the modestly dressed condition, Sara wears khaki pants and a crewneck t-shirt. In the revealingly dressed condition, Sara wears a short skirt and knee-high boots with a low-cut top. In both of these conditions, Sara is of average weight. In the revealingly dressed and heavier weight condition, Sara wears the same revealing outfit, but the photo has been manipulated to make Sara appear overweight (and thus less traditionally physically attractive). We performed manipulation checks on perceptions of target sexual permissiveness and physical attractiveness (see Supplementary Materials)...
Overall, people anticipated less indirect aggression toward Sara when she was modestly (M ¼ 2.31, SE ¼ 0.18) versus revealingly dressed (average weight: M ¼ 4.12, SE ¼ 0.17, heavier weight: M ¼ 3.93, SE ¼ .17), F(1, 139) ¼ 63.89, p < .001, Zp2 ¼ .315, 95% CI ¼ [.19, .42] but did not anticipate different levels of indirect aggression toward Sara when revealingly dressed at either weight (p ¼ .416). This pattern of findings suggests that women might incur such aggression when they are perceived as intending to seek male attention, regardless of their capacities to successfully do so.
People expected Sara to be socially distanced more when she was revealingly (average weight: M ¼ 4.92, SE ¼ 0.18, heavier weight: M ¼ 4.08, SE ¼ .18) versus modestly dressed (M ¼ 3.15, SE ¼ 0.18), F(1, 139) ¼ 36.82, p < .001, Zp2 ¼ .209, 95% CI ¼ [.10, .32]. People also expected Sara to be socially distanced more when she was revealingly dressed and more traditionally attractive (i.e., average weight than heavier weight), F(1, 139) ¼ 10.94, p ¼ .001, Zp2 ¼ .073, 95% CI ¼ [.01, .17]. In contrast to the indirect aggression described above, this pattern of findings suggests that this type of aggression, which is perhaps more instrumental, may be attuned relatively less to target intentions (e.g., to seek male attention) and relatively more to target capacities (e.g., to attract male attention).
By contrast, and consistent with predictions, participants did not expect differences in direct aggression toward Sara...
Experiment 2
When revealingly dressed, women might expect costs upon encountering other women (at least in professional contexts; Howlett, Pine, Cahill, Orakc¸ıog˘lu, & Fletcher, 2015; Vaillancourt & Sharma, 2011) but might also expect benefits upon encountering men (at least in social contexts; e.g., Buss & Schmitt, 1993; Sakaguchi & Hasegawa, 2006). Therefore, we tested whether women create less revealing outfits for all-female versus mixed-sex gatherings...
In line with established methods for assessing participant-created outfits (Durante, Griskevicius, Hill, Perilloux, & Li, 2011; Haselton, Mortezaie, Pillsworth, Bleske-Recheck, & Frederick, 2007), participants received a piece of 8½ 11 in paper with two female figure outlines, depicted side-by-side, and a box of 24 crayons. Participants were instructed to draw the outfits they would wear to both of the events.
After the completion of data collection, undergraduate research assistants (RAs) who were blind to hypotheses used a graph paper transparency overlay to count the number of squares (to the nearest quarter) of skin showing versus not showing (out of all possible squares). We then calculated the proportion of skin revealed...
Figure 1. Levels of aggression anticipated to be directed toward each target. Error bars represent standard errors. Stimuli created by Vaillancourt and Sharma (2011)...
In line with predictions, women created more modest outfits for attending an all-female psychology mixer (M ¼ 20.48, SE ¼ 1.55) than a mixed-sex mixer (M ¼ 36.16, SE ¼ 1.68), baring about 16% less skin for the former versus the latter (see Figure 2).
Experiment 3
Using a complementary behavioral measure, Experiment 3 conceptually replicates Experiment 2 and further tests whether women at greater individual risk for aggression (i.e., physically attractive women) dress more modestly to encounter other women. To mitigate concerns about confounding sex composition and context, we examine context (professional, social) orthogonally...
Women were instructed to read a scenario about being new to an area, meeting another woman who seems friendly and accepting that new acquaintance’s invitation to meet a group of her friends at a small gathering. Women were asked to make clothing choices for that gathering. Women were randomly assigned to one of four small gatherings: an all-female networking event, an all-female birthday party, a mixed-sex networking event, or a mixed-sex birthday party.
Complementing the method from Experiment 2, women were asked to choose one of 12 dress or skirt outfits and then also one of 12 jeans outfits (2 outfits total). Outfits were modeled on a realistic rendering of a young adult woman. We assembled each outfit with a web application, perhaps best understood as an electronic paper doll designer (“Street Style”; created by Rinmaru Games; accessed at https://www.dolldivine.com/street-style. php). Cropping from screen captures, we made an image of each outfit suitable for incorporation into surveys (Figure C1). Each outfit was prerated on modesty by a subset of a separate sample of participants consisting of 94 women living in the United States, between the ages of 20 and 35, who indicated a straight sexual orientation. These participants were recruited via Turk-Prime and were paid 50 cents. Outfits were rated on Likert scales ranging from 1 (not at all modest) to 7 (extremely modest). Each outfit was rated by between 40 and 52 women.
Chosen outfits ranged in prerated modesty from 1.87 (lower modesty) to 6 (higher modesty). We averaged modesty ratings for the chosen dress/skirt and jeans outfits, creating one overall score of outfit modesty. (Because dress/skirt outfits vary in how much leg is bared, whereas all jeans outfits cover the entire leg, on average, jeans outfits were rated as more modest.)
We also asked women to report their beliefs about how other women viewed them (e.g., “Upon just meeting me for a few minutes, other women think . . . I’m very physically attractive”). Women responded to the focal item (physical attractiveness) embedded alongside distractor and exploratory items (e.g., “I’m very intelligent”), which were not analyzed...
Results suggest that physically attractive women show an exaggerated tendency to choose more modest outfits when meeting other women—regardless of context (see Figure 3B). We find main effects of attractiveness, t(180) ¼ 3.02, p ¼ .003, r p ¼ .22, 95% CI ¼ [.08, .35], such that more attractive women chose more modest outfits, sex composition, t(180) ¼ 2.68, p ¼ .008, r p ¼ .20, 95% CI ¼ [.33, .05], such that women chose more modest outfits to encounter other women versus men, and context, t(180) ¼ 2.20, p ¼ .029, r p ¼ .16, 95% CI ¼ [.30, .02], such that women chose more modest outfits for professional versus social contexts. These were qualified by the predicted Sex Composition Context Interaction, t(180) ¼ 3.34, p ¼ .001, r p ¼ .24, 95% CI ¼ [.37, .10]. More germane, we also found a significant Attractiveness Sex Composition interaction, t(180) ¼ 2.87, p ¼ .005, r p ¼ .21, 95% CI ¼ [.34, .06]. No other interactions were significant (ps > .351).
Probing this interaction revealed that women reporting less attractiveness (1 SD) did not differ in clothing choices for meeting all-female or mixed-sex groups (p ¼ .858). However, women reporting greater attractiveness (þ1 SD) chose more modest clothing for meeting a same-sex group (M ¼ 4.27, SE ¼ 0.10) as compared to meeting a mixed-sex group (M ¼ 4.06, SE ¼ 0.10), t(180) ¼ 3.99, p < .001, r p ¼ .28, 95% CI ¼ [.41, .14])
Figure 2. Examples of outfits created by three different participants for the (A) mixed-sex social mixer and (B) all-female professional mixer.
Experiment 4
Experiment 4 conceptually replicates the focal effect and explores boundary conditions. Women reported desired outfit provocativeness for drinks with a prospective or existing female friend, male friend, or male romantic partner. Because people may be more likely to aggress against newcomers than friends, and females are thought to be especially aggressive toward same-sex newcomers (Benenson et al., 2013; Feshbach, 1969; Feshbach & Sones, 1971), we expected that highly physically attractive women would intend to dress less provocatively to meet prospective (but not existing) female friends...
After reading the scenario, women reported how “sexy” they wanted to dress and also how much skin they wanted their outfit to reveal for drinks that night. Both items were assessed on separate sliders with respective labels (0 ¼ wear an everyday outfit, not sexy; 10 ¼ wear a very sexy outfit or 0 ¼ barely any skin showing; 10 ¼ very much skin showing). Responses were averaged to yield a provocative dress score (a ¼ .87)...
Do attractive women dress less provocatively to meet prospective versus existing female friends? In Step 2 of our regression model, we found a marginally significant Contrast Code (e) Attractiveness interaction, suggesting that relationship type (male or female friends), length, and attractiveness influence women’s sartorial intentions (see Table 2 and Figure 4).
Probing this interaction, we first explored women’s sartorial intentions for encountering other women. In line with our theorizing, we find that, with increasing physical attractiveness, women reported decreasingly intentions to dress provocatively for meeting a prospective (but not existing) female friend, t(104) ¼ 2.75, p ¼ .007, r p ¼ .26, 95% CI ¼ [.43, .07] and also that more physically attractive women (þ1 SD) reported intending to wear less provocative outfits when meeting a prospective (M ¼ 2.59, SE ¼ 0.31) versus an existing female friend (M ¼ 4.23, SE ¼ 0.44), t(104) ¼ 3.75 p < .001, r p ¼ .35, 95% CI ¼ [.16, .50]). Although not predicted, less physically attractive women (1 SD) reported intending to wear more provocative outfits when meeting a prospective (M ¼ 3.74, SE ¼ 0.28) versus an existing female friend (M ¼ 2.61, SE ¼ 0.38), t(104) ¼ 2.99, p ¼ .004, r p ¼ .28, 95% CI ¼ [.45, .09])
For encountering male friends, the effect of relationship length did not depend on attractiveness. First, more attractive women wanted to dress more provocatively whether meeting a prospective male friend, t(88) ¼ 2.60, p ¼ .011, r p ¼ .27, 95% CI ¼ [.06, .45], or an existing male friend, t(88) ¼ 3.71, p < .001, r p ¼ .37, 95% CI ¼ [.17, .54]. Second, women reported intentions to dress less provocatively when meeting a prospective (M ¼ 2.37, SE ¼ 0.27) than an existing male friend (M ¼ 3.57, SE ¼ 0.39), t(88) ¼ 3.07, p ¼ .003, r p ¼ .31, 95% CI ¼ [.11, .49], which may reflect desires to avoid unwanted attention and possible intersexual aggression from male strangers, who may be (perceived as) somewhat more likely to pose such threats than romantic partners or female friends (e.g., Blake, Bastian, & Denson, 2018; Goetz, Easton, & Buss, 2014).
In four experiments, we find (1) women (and men) expect women—and especially physically attractive women—to evoke greater indirect (but not direct) intrasexual aggression when revealingly versus modestly dressed; (2) women dress more modestly for encountering same- versus mixed-sex groups—across contexts (professional, social); (3) more physically attractive women, who may be at greater risk of incurring intrasexual aggression, demonstrate an exaggerated tendency to do this; and (4) this effect is apparent only when such women dress to meet a prospective (but not existing) female friend. Findings are consistent with theorizing that women are not only aware that certain perceptually-salient cues (e.g., revealing clothes, physical attractiveness) render them more likely targets of intrasexual aggression, but also that women might thus seek to avoid such aggression by strategically manipulating their appearance—specifically, by damping their outfit provocativeness. Importantly, this damping is flexibly engaged when aggression risk is highest: by individuals who may already be frequent targets (physically attractive women), and in situations when aggression is more likely (when women are newcomers)...
Because attracting and maintaining same-sex friends can confer numerous important benefits for women (e.g., Campbell, 2002), future work could benefit from exploring how women might manipulate their appearances to establish and maintain same-sex friendships— as well as to avoid same-sex aggression. For example, donning the baggy sorority t-shirts and short shorts common to young women on some college campuses may communicate not only the wearer’s on-campus status but also her dedication to her coalition.
Similarly, we do not assume that attracting the attention of male audiences yields only opportunities (and not also threats). Our data imply that women are attuned to the possible benefits and costs of male attention, damping provocative dress when encountering prospective male friends (but not romantic partners), whose unwanted sexual interest could be threatening. In all, findings suggest that women’s sartorial behavior is attuned to multiple factors, including women’s own attractiveness, the social context, and several audience features (e.g., audience gender composition, familiarity with their audience). Future work might further investigate these and other nuances (e.g., Glick, Larsen, Johnson, & Branstiter, 2005).
This work adds to literature suggesting that features cuing women’s intentions to attract men (e.g., provocative dress) can evoke aggression, but so too can features cuing women’s ability to attract men (e.g., physical attractiveness). This implies that some women might attract intrasexual aggression without having engaged in behavior to prompt it. That is, women might be aggressed against for being competitive (e.g., for desirable romantic partners), even if those women were not actively competing. One might speculate that, in contrast, men are unlikely to evoke intrasexual aggression without having actively provoked it"
So in conclusion, while this study does not directly and entirely disprove the claim that women dress "for themselves" (since one could come up with some complex adhoc argument about how women would dress for themselves if alone, or some nonsense about how meeting prospective male partners makes women want to feel more confident by wearing revealing clothing), it at least comprehensively refutes the claim that when they wear provocative clothing, women dress "for other women", since on multiple dimensions, we can see that other women do not welcome provocatively dressed women, and we can at least put the burden of proof on feminists who claim that women do not wear sexy clothing for the benefit of men.
The interaction of variables is particularly interesting - for example, attractive women are aware of the impacts of their dressing on both men and other women, and even on women they don't already know.
Given that all this coheres with the literature, we can be quite confident in the findings.