Why Is Sexualization Dehumanizing? The Effects of Posture Suggestiveness and Revealing Clothing on Dehumanization

Research has shown that sexualized people are perceived as possessing fewer traits of a human being. Most scholars have argued that these effects are driven by revealing clothing, with targets wearing swimsuits or lingerie being perceived as possessing less mind and less humanness in comparison with nonsexualized targets. However, revealing clothing in these studies was often confounded with other sexualizing factors, such as posture suggestiveness, and, so, the aspects which lead people to perceive women in object-like ways remain unclear. This article begins to fill this gap by examining the role of two key sexualizing factors, namely revealing clothing and posture suggestiveness, on objectification-related traits. After exposure to a picture of a woman, 223 participants were asked to indicate the extent to which this woman possessed warmth, competence, and morality. For competence and warmth, we found an interaction between revealing clothing and posture suggestiveness: Posture suggestiveness caused less attribution of warmth and competence to women wearing revealing clothing, but not for women wearing less revealing clothing. For morality, we found that women in suggestive (vs. nonsuggestive) postures were perceived as possessing less morality, regardless of the type of clothing. The implications of these findings for the field are discussed.

globally whereas they rely on a more local, piecemeal processing when viewing images of sexualized bodies and ordinary objects, and this cognitive objectification occurs at both neural (e.g., Bernard, Content, Deltenre, & Colin, 2018;Bernard, Hanoteau et al., 2018;Bernard, Rizzo et al., 2018) and behavioral levels (Bernard, Gervais, Allen, Campomizzi, & Klein, 2015;Bernard, Gervais, Allen, Campomizzi, & Klein, 2012;Bernard, Gervais, Allen, Delmée, & Klein, 2015;Civile & Obhi, 2016). In accordance with feminist scholars who posited that objectification is related to a reduction of others to their sexual body parts (e.g., Bartky, 1990), this line of research shows that sexualized bodies are cognitively reduced to their parts akin to objects (for a review, see . Moreover, when people form impressions about others, they are also more likely to perceive sexualized targets in "objectlike" ways, that is, to attribute fewer traits of a human being and more traits of an object to sexualized people than to their nonsexualized counterparts (for reviews, see Heflick & Goldenberg, 2014;Ward, 2016). In these studies, sexualization is manipulated through the amount of skin versus clothing that is visible, with sexualized targets associated with clothing that reveals a large amount of skin (e.g., swimsuit, underwear), and nonsexualized targets wearing less revealing clothing (e.g., jeans and T-shirt). A growing body of research showed that people perceived sexualized women as possessing less mind (Loughnan et al., 2010), fewer uniquely human characteristics (Vaes, Paladino, & Puvia, 2011), and less agency (Cikara, Eberhardt, & Fiske, 2011) in comparison with nonsexualized women. This subtle dehumanization predicts, in turn, the way in which people evaluate the responsibility of the perpetrator versus victim of sexual or nonsexual aggression (Bernard, Loughnan, Godart, Marchal, & Klein, 2015;Loughnan, Pina, Vasquez, & Puvia, 2013;Pacilliet al., 2017).
One may think that the latter body of research delivers a clear message: Sexualization, manipulated through revealing clothing, causes people to attribute fewer traits of a human being and more traits of an object to women. However, we suggest that this conclusion is unwarranted. It indeed appears that sexualization often has very different meanings depending on the researcher: Sexualized targets not only wear more revealing clothing in comparison with nonsexualized targets, they are often associated with greater body-to-face ratio (e.g., Bernard, Loughnan et al., 2015;Loughnan et al., 2010;Wollast, Puvia, Bernard, Tevichapong, & Klein, 2018), and more posture suggestiveness (e.g., Bernard, Content et al., 2018;Civile & Obhi, 2016;Loughnan et al., 2013) in the same investigations. The use of sexualization as an umbrella term in objectification research, therefore, makes questions of which aspects of sexualization cause people to be perceived in more object-like ways and in less human-like ways less clear. In this article, we will examine the effects of two core dimensions of target sexualization, namely revealing clothing and posture suggestiveness, and we will investigate their effects on "object-like" dehumanized perceptions.
Postural suggestiveness relates to an open body language that is sexually connoted. Generally speaking, we know that postural openness modulates the way people form impressions about others (e.g., Burgoon, 1991), with recent studies indicating that higher posture suggestiveness contributes to women being perceived as more sexually objectified. For example, Fasoli, Durante, Mari, Zogmaister, and Volpato (2018) presented to participants three types of images in line with Hatton and Trautner's (2011) sexualization coding scheme: Women wearing nonrevealing clothing with neutral poses (condition nonrevealing); women wearing revealing clothing with neutral poses (condition revealing); women wearing revealing clothing with sexual poses (condition sexualized revealing). Participants were asked to indicate the extent to which the women depicted were portrayed as a sexual object. Fasoli et al. (2018) found that the women in the sexualized revealing condition were perceived more as sexual objects in comparison with the women in the revealing condition, followed by the women in the nonrevealing condition, thereby suggesting that people perceive women as being sexual objects to a greater extent when targets are both wearing revealing clothing and displaying suggestive postures. In a related manner, another recent study manipulated revealing clothing and posture suggestiveness while assessing the neural correlates of the early visual processing of sexualized people (Bernard, Hanoteau, et al., 2018). Specifically, the authors manipulated revealing clothing and posture suggestiveness and examined whether revealing clothing, posture suggestiveness, or both cause bodies to be processed less configurally and more analytically, that is, similarly to the way most objects are perceived (Maurer, Le Grand, & Mondloch, 2002). These authors found that bodies displaying nonsuggestive postures were processed configurally regardless of the clothing type, whereas bodies with suggestive postures (regardless of the clothing type) were processed in piecemeal ways akin to ordinary objects (for similar findings, see also Bernard, Content et al., 2018;Bernard, Rizzo et al., 2018). In sum, these recent studies suggest that posture suggestiveness contributes to the perception of women as sexual objects and to visually process them as such. These studies, however, remain silent relative to the downstream consequences of revealing clothing versus posture suggestiveness on the way we form impressions about others. This article examines which specific dehumanizing characteristics are associated with clothing sexualization and which characteristics are associated with posture sexualization and whether combining clothing and posture sexualization causes increased dehumanization of women.
In this article, we will examine whether revealing clothing, posture suggestiveness, or both cause women to be perceived in object-like ways, that is, as possessing fewer traits of a human being and more traits of an object. To do so, we assessed the extent to which participants perceived a woman as possessing warmth, competence, and morality. These dimensions pertaining to the Stereotype Content Model (SCM; Fiske, Cuddy, Glick, & Xu, 2002) are indeed closely related to whether people are perceived as being human-versus object-like. For instance, people who are both low in warmth and in competence are dehumanized at a neural level (Harris & Fiske, 2006) and are also perceived as lacking in internal mental states (Harris & Fiske, 2007). In addition, research has shown that the warmth dimension includes two components: warmth (e.g., likability, kindness) and morality (e.g., trustworthiness, honesty) (Leach, Ellemers, & Barreto, 2007). Morality, just like warmth, also predicts attribution of humanness (Haslam, 2006). Furthermore, studies have considered decreased perception of competence (e.g., intelligent, capable) of sexualized targets as indicators of depersonalization (Loughnan et al., 2010) and dehumanization (Wollast et al., 2018). Importantly, it has been shown that appearance-focused participants tend to attribute less warmth, less competence, and less morality to female targets than personality-focused participants (Heflick, Goldenberg, Cooper, & Puvia, 2011), suggesting that people attribute fewer traits of a human being to women when focusing on their physical appearance rather than on their personality.
We examined two concurrent hypotheses. The first hypothesis is that revealing clothing and posture suggestiveness will be sufficient to trigger object-like attributions: Participants will attribute less warmth, less competence, and less morality to women wearing revealing clothing in comparison with women wearing less revealing clothing (H1a); and they will attribute less warmth, less competence, and less morality to the women displaying suggestive postures in comparison with the women displaying nonsuggestive postures (H1b). The second hypothesis (H2) is that women will be perceived in object-like ways only when depicted as hyper-sexualized (i.e., through a combination of revealing clothing and suggestive posture). That is, people will attribute less warmth, less competence, and less morality to the woman in revealing clothing and suggestive posture in comparison with the other targets.

Pretest
To assess the role of revealing clothing and posture suggestiveness on the attribution of warmth, competence, and morality, we have created new stimuli, that is, a woman depicted in revealing versus nonrevealing clothing and displaying suggestive versus nonsuggestive postures. Toward that end, we bought pictures of a female model wearing underwear and displaying a series of different body postures from an image bank website. We selected 12 postures and conducted a pretest to identify which postures were rated as being the more suggestive versus nonsuggestive.
Twenty-two participants (M age = 24.82; 12 men) were asked to indicate the extent to which each of the 12 postures was sexually suggestive on a 7-point Likert-type scale ranging from 1 = not at all to 7 = very much. We selected the posture having the lowest versus highest rating in posture suggestiveness. A repeated-measures analysis of variance (ANOVA) with posture suggestiveness (nonsuggestive, suggestive) and participant gender ( .05, nor the interaction between posture suggestiveness and target gender, F(1, 20) = .10, p = .76, η p 2 < .01, reached significance. To create the less revealing clothing versions of the targets, we used an image editing software to cut black clothes from a picture we took ourselves. We placed these black clothes on the two preselected pictures. By doing so, all pictures depicted the same woman; the pictures differed only as a function of posture suggestiveness and clothing. We pixelated the target's face to minimize face processing. The pictures are presented in Figure 1.

Participants and Procedure
Two hundred twenty-three U.K. participants fully completed an online survey lasting 2 to 3 min and received a monetary compensation of £0.20 ($.30; hourly wage = £6; $9). We posted the survey link on Prolific website. The survey was not visible to participants who took part in the pretest. The sample size was calculated based on Cohen's d, found in similar studies. Specifically, studies measuring the effect of appearance-focus on perception of warmth, competence, and morality found moderate interaction effects (e.g., Heflick et al., 2011). We then calculated the sample size to detect a moderate interaction effect (d = .5) with a power of .95, which translated into a sample size of 210. From the sample of 223 participants, we excluded five participants (four participants reported that the picture of the woman did not appear at all and at the beginning of the survey and one participant reported to not being fluent in English). The final sample included 218 U.K. participants (M age = 27.50; SD age = 4.81; 136 women).
Participants read the following instructions: "On the next page, a picture of a woman will appear for 15 sec. Look carefully at this woman. You will be asked to make a series of judgments about this person; so, from the picture, try to get an idea of what she is like." Participants were then assigned to one of the four conditions of the study. On the following pages, participants filled a questionnaire assessing their perceptions of this woman. Participants were asked to indicate the extent to which they think different traits were representative of the woman seen in the picture. Based on Heflick et al. (2011), who used two items to assess warmth, competence, and morality, we used two items to assess the dimensions of competence (intelligent, capable), warmth (kind, friendly), and morality (sincere, trustworthy). Internal consistency was acceptable/good for each variable (α competence = .77; α warmth = .86; α morality = .82).
Next, the participants read the following instructions: "On the next page, you will see the same woman you already saw. You will be asked to make a second and last series of judgments about this person." Participants were asked to indicate the extent to which the target wears clothes that reveal her body; displays a sexually suggestive body posture; is depicted in a sexualized way on a 7-point Likert-type scale ranging from 1 = not at all to 7 = very much. Finally, the questionnaire ended after a series of demographic questions. We agree to share anonymized data files from this research with other qualified professionals on request to confirm the conclusions of the research.

Manipulation Checks
Participants rated the targets in the revealing clothing condition as wearing clothes that reveal their bodies to a greater extent ( 14. The interaction between clothing and posture suggestiveness was not significant, p = .51. In sum, this indicates that we successfully manipulated revealing clothing and posture suggestiveness and that these two dimensions were sufficient to trigger higher sexualization self-reports.

Warmth
Inconsistent with H1a and H1b, the main effects of clothing and posture suggestiveness were not significant, ps > .10. Critically, and consistent with H2, we found a significant interaction between clothing and posture suggestiveness,

Morality
As shown in Figure 4, and consistent with H1b, the main effect of posture suggestiveness was significant, F(1, 214) = 13.15, p < .001, η p 2 = .058, with the targets displaying suggestive postures being rated as possessing less morality (M = 3.48, SD = 1.13) than the targets displaying nonsuggestive postures (M = 3.99, SD = .99). Inconsistent with H1a and H2, neither the main effect of clothing nor the interaction between clothing and posture suggestiveness was significant, ps > .21.

Discussion
During the past decade, research has found evidence that target sexualization is associated with dehumanization (for a review, see Heflick & Goldenberg, 2014). Most researchers have argued that this effect of sexualization is driven by revealing clothing, with targets wearing revealing clothing perceived as possessing less mind (Loughnan et al., 2010;Wollast et al., 2018), less agency (Cikara et al., 2011), and less human uniqueness (Vaes et al., 2011). However, because revealing clothing is often confounded with body-to-face ratio (e.g., Loughnan et al., 2010) and posture suggestiveness (e.g., Bernard, Content et al., 2018) in these investigations, what is meant by sexualization-and which facets of sexualization affect the way we perceive women-remained unclear. Stated differently, revealing clothing and posture suggestiveness often go hand in hand in the mass media (Hatton & Trautner, 2011). It matters to understand whether revealing clothing, posture suggestiveness, or both, shift the ways in which we attribute fewer traits of a human being to women. Recent studies suggest that posture suggestiveness not only increases ratings in sexualization but also causes people to see women as sexual objects (Fasoli et al., 2018) and to cognitively appraise them in a manner similar to the way ordinary objects are appraised (Bernard, Hanoteau et al., 2018).
However, it remained unclear as to what the downstream consequences of revealing clothing relative to posture suggestiveness are on the attribution of traits that people typically attribute to human beings. To address this question, we manipulated revealing clothing and posture suggestiveness and examined their effects on the attribution of traits people typically attribute to human beings, that is, warmth, competence, and morality (cf. Harris & Fiske, 2006;Heflick et al., 2011). Our findings not only corroborated the notion that revealing clothing and posture suggestiveness translate into higher sexualization (Hatton & Trautner, 2011), they also shed light on the fact that the most pronounced dehumanization occurs for hyper-sexualized targets, that is, targets wearing revealing clothing and in suggestive postures. This was especially true when considering attributions of warmth and competence. In contrast, lower attribution of morality appeared to be uniquely driven by posture suggestiveness, with women in suggestive postures perceived as possessing less morality in comparison with women in nonsuggestive postures.
Given that male participants were not equally distributed across conditions in our study, this rendered participant gender comparisons underpowered and so poorly interpretable (e.g., N = 38 vs. N = 10). This explains why participant gender was omitted in our analysis. In line with most objectification research that failed to find any moderating effect of participant gender (for a review, see Heflick & Goldenberg, 2014), it is unlikely that our findings would be moderated by participant gender. However, we encourage researchers to ensure equal distribution of male versus female participants in future studies. To date, research mostly examined the effect of body sexualization on dehumanized perceptions. This article was no exception, and we pixelated the target's face to minimize face processing. One may nonetheless argue that this methodological precaution is a dehumanizing factor per se because it heightens the focus on the target body. However, this possibility is unlikely. Research indeed showed that headless bodies trigger object-like piecemeal processing (Yovel, Pelc, & Lubetsky, 2010)-and dehumanized perceptions (Loughnan et al., 2010)-whereas masking faces does not.
Another limitation is that we employed images of only female targets. Some studies found target gender effect when assessing the effect of target sexualization on dehumanization with more pronounced effect of sexualization for female targets (e.g., Vaes et al., 2011), whereas other investigations did not observe such a target gender effect (e.g., Loughnan et al., 2010). To properly assess the moderating role of target gender on the dehumanizing effect of target sexualization, it implies to rely on male versus female targets having similar levels of target sexualization. We believe that similar results would be found when images of highly sexualized male targets are used. Consistent with this possibility, Bernard, Rizzo et al. (2018) recently used images of male and female targets that were equally rated as highly sexualized and found that male and female targets rated high in sexualization were visually processed in a manner similar to everyday objects, whereas male and female targets rated low in sexualization were processed in configural, human-like ways (for similar results, see also Bernard, Content et al., 2018;Bernard, Hanoteau, et al., 2018). In a related manner, the stimulus we used in the different experimental conditions was a young and White model. However, future research might want to reproduce our findings with other targets differing in, for example, ethnicity, age, attractiveness, or body types, to test the robustness of the findings. Future research should also examine whether the effects reported in this article can be replicated in Western versus Eastern cultures. In line with recent findings suggesting that objectification is primarily a Western phenomenon (e.g., Gervais, Bernard, & Riemer, 2015;Loughnan et al., 2015), we expect that the effects of revealing clothing and posture suggestiveness would be more pronounced in Western cultures than in Eastern cultures.
Our research contributes to a better understanding as regards the aspects of sexualization which cause people to be perceived similarly to objects, thereby suggesting potential interventions for reducing the objectification of girls and women. From content analysis of advertising, we know that women are often portrayed as wearing clothing and displaying suggestive postures (e.g., Hatton & Trautner, 2011). Our research makes a call to not rely on depictions combining revealing clothing and suggestive postures in the visual media. We also encourage researchers to further examine other target features that might contribute to the dehumanizing effect of sexualization. For instance, if dehumanization is related to target sexualization more generally, and not only because of body focus, it might be that sexualized face features (e.g., heavy make-up, flirty facial expressions) would lead to similar dehumanized perceptions. In summary, the present research contributes to a better understanding of what drives the dehumanizing effect of sexualization by showing that dehumanization occurs when women are depicted as hyper-sexualized, through a combination of revealing clothing and sexually connoted postures.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the National Fund for Scientific Research (FRS-FNRS, Belgium).