Legitimizing Intimate Partner Violence: The Role of Romantic Love and the Mediating Effect of Patriarchal Ideologies

Romantic love in heterosexual relationships is recognized as an important aspect to be considered in relation to the psychosocial mechanisms associated with the persistence of intimate partner violence and the help-seeking barriers faced by female victims. However, few studies have explored the processes underlying the relationship between this form of love and attitudes toward this kind of violence. To do this, the current study aims to analyze the relationship between the adherence to romantic love and the legitimization of intimate partner violence (i.e., perceived severity of violence, victim blame, and exoneration of the perpetrator). It was also to test the mediating effect of patriarchal ideologies (i.e., ambivalent sexism and domestic violence myths) on this relationship. Two hundred thirty-five French adults (51.1% women) were surveyed. The data were analyzed with structural equation models to study the mediations between the variables considered. Consistent with our hypotheses, the results showed that the more the participants adhered to romantic love, the more they blamed the victim and exonerated the perpetrator. They also showed a positive link between romantic love, ambivalent sexism, and domestic violence myths. Finally, the results demonstrated that the relationship between romantic love and the legitimization of violence is mediatized by ambivalent sexism and domestic violence myths. The findings illustrate the need to deconstruct romantic love ideology and the psychosocial logics underlying the legitimization of intimate partner violence.


Introduction
Intimate partner violence (IPV) is a worldwide social problem affecting between 10% and 71% of women in their lifetimes (Garcia-Moreno, Jansen, Ellsberg, Heise, & Watts, 2006). This phenomenon has been acknowledged as one of the leading causes of injuries for women (Hague & Malos, 2005). However, this common violence remains rarely denounced and prosecuted (see Lelaurain, Graziani, & Lo Monaco, 2017, for a review on help-seeking barriers). Thus, one of the most important scientific issues in this field concerns the understanding of the psychological and social mechanisms underlying the persistence of this violence and the help-seeking barriers faced by victims. Patriarchal ideologies defining socio-symbolic relationships between genders affect judgments toward survivors and contribute to the nonrecognition or acceptance of IPV (Baldry & Pagliaro, 2014;Lelaurain et al., 2018). Beyond simply underlining these ideologies, Peterson del Mar (1996) proposes to "examine our most cherished values and most powerful cultural traits" (p. 174) that could contribute to the acceptance of IPV. Among these ideologies, romantic love seems to be one of the most widespread and socially accepted. For example, the above-mentioned French newspaper headlines illustrate the spontaneous relationship that people make between IPV and romantic love in the social field. However, romantic love has often been considered as a peripheral aspect of the IPV phenomenon in literature (Power, Koch, Kralik, & Jackson, 2006). Thus, our objective in the present study was to tackle the question of the effect of adherence to romantic love, ambivalent sexism, and domestic violence myths on IPV legitimization.

Romantic Love and Acceptance of IPV
Romantic love is a set of beliefs about the power of love and the perfection of romance including, for example, idealization of the partner, mysticism, complete involvement, and exclusiveness or jealousy (see Knox & Sporakowski, 1968;Sprecher & Metts, 1989). It also refers to the nature of love and more specifically the characteristics or criteria for an ideal relationship (Bartell, 2009) as to how it should develop, function, and be maintained. Romantic love was established as a reference model for heterosexual love relationships in the Western culture of the 19th and 20th centuries, valuing sexual and amorous exclusivity, predestination of the encounter and conjugal perpetuity (Marquet, 2009).
Nowadays, the postmodern love relationship is not only built on the basis of romantic love, but also incorporates representations based on the independence and autonomy of individuals (Chaumier, 1992). We thus find several representations of love which coexist, and which are not the object of the same social valorization depending on the sociocultural context (Apostolidis & Deschamps, 2003). As we can see in cultural productions for youth, romantic love remains one of the most socially valued representations of love (Bonomi, Altenburger, & Walton, 2013;Borgia, 2014;Hefner & Wilson, 2013). For instance, a study showed that almost 80% of Disney animated classics, at the time of the study, portrayed a couple that fell in love at first sight, got married, and "lived happily ever after" (Tanner, Haddock, Zimmerman, & Lund, 2003). These love themed media products can generate idealistic expectations and practices of romantic love (Segrin & Nabi, 2002). Such romantic narratives shape experiences, actions, and understandings of love and relationships, and in turn men and women learn what is love through them (Jackson, 1993).
Several studies have demonstrated the role of adherence to romantic love in the legitimization of IPV. For instance, the narratives of love and passion are very strong in the news media in framing intimate partner homicide, to the detriment of the structural and social dimensions underlying it (DeShong & Haynes, 2016;Gius & Lalli, 2014). In particular, Gius and Lalli (2014) highlighted that "the frame of 'romantic love' is used in order to legitimate and justify violence against women" (p. 68). Indeed, men who are violent and murderous toward their wives are often described as being under the influence of uncontrollable emotions, such as having "too much pain" or being "too much in love," but rarely as aggressors (Gius & Lalli, 2014;Romito, 2006). Thus, adhesion to romantic love can help to make IPV socially acceptable in perceiving violence as a male form of amorous expression, a signs of intimacy or affection rather than as something problematic (Bonomi et al., 2013;Bonomi et al., 2014;Lelaurain et al., 2018). However, romantic love is not only a veil that diminishes the recognition of violence within a couple but also one of the main barriers to the victims leaving the violent partner (Fanslow & Robinson, 2010;Jaspard, 2007;Lelaurain et al., 2017). Love is also a "weapon" for IPV because male dominance often depends on the partner's emotional dependence (Jaspard, 2007).

The Role of Ambivalent Sexism and Domestic Violence Myths
According to feminist perspectives, the social valorization of romantic love also helps to maintain gender inequalities through the naturalization of a culture of male dominance (see Monckton-Smith, 2012). Indeed, within many Western cultures, romantic heterosexual love is often used to maintain traditional gender relationships and to justify the order of women's subordination to men (Chung, 2005;Singh, 2013;Wood, 2001). For instance, the romance narrative portrays women as needing to be rescued by man and as needing to be complete and fulfilled, while men are portrayed as naturally strong, full of power, and sure of themselves (Vincent & McEwen, 2004;Wood, 2001). An important aspect of romantic love is therefore to contribute to the definition of gender roles in couple relationships (Rudman & Glick, 2008), in particular, by reducing women's identity to their feelings of love as well as making love into a destiny they must achieve (Chung, 2005;Power et al., 2006).
According to this literature, ambivalent sexism and domestic violence myths are two socio-cognitive constructs that should be theoretically linked to romantic love insofar as they contribute to defining traditional gender roles legitimizing male domination. The first construct brings together two coexisting forms of sexism whose common function consists in maintaining women in their gender role and legitimizing gender inequalities (Glick & Fiske, 2001). On one hand, benevolent sexism is defined as a set of attitudes that lead to seeing women through perceived positive stereotypes (Glick & Fiske, 2001). On the other hand, hostile sexism is based on the idea of domination and male superiority (Dardenne, Delacollette, Grégoire, & Lecocq, 2006). Domestic violence myths were defined by Peters (2008) as stereotypical beliefs that are widely and persistently held to minimize or justify IPV against women. According to this author, the function of these myths is to support the patriarchal organization of society insofar as they are positively correlated to negative attitudes toward women and to certain restrictive visions of social gender roles.
Thus, not only do ambivalent sexism and domestic violence myths help to define the traditional gender roles from which representations of romantic love are drawn, but these socio-cognitive constructs are also important predictors in the evaluation of IPV. This suggests that, behind the adherence to romantic love, the legitimization of IPV is likely to be determined by the internalization of patriarchal ideologies defining gender-symbolic roles and justifying male domination. Consequently, the potential effect of romantic love on the attitudes of individuals who face situations of IPV is likely to be mediated by the adherence to patriarchal ideologies. To our knowledge, this mediation has never been empirically tested. Yet it is an important issue to the extent that romantic love is a relevant analytical framework for improving our understanding of ideologies underlying the acceptance of IPV.

The Present Study
In line with the literature, the first objective of the present study was to verify whether adherence to romantic love predicts different evaluations, such as victim blame, exoneration of the perpetrator, and perceived violence severity. The second objective was then to examine how adherence to romantic love accounts for these evaluations that tend to legitimize IPV. As we have seen, several authors highlight the fact that romantic love may help to maintain gender inequalities (Monckton-Smith, 2012;Singh, 2013;Wood, 2001). This suggests that adherence to patriarchal ideologies are mediators in the relationship between romantic love and the evaluation of IPV situations. We therefore hypothesized as follows: Hypothesis 1 (H1): The more individuals adhere to romantic love, the more they will blame the female victim for the violence she suffered. Hypothesis 2 (H2): The more individuals adhere to romantic love, the more they will exonerate the male perpetrator of the violence. Hypothesis 3 (H3): The more individuals adhere to romantic love, the less they will perceive the male violence suffered by the female victim as severe. Hypothesis 4 (H4): The effect of romantic love on the evaluation of IPV will be mediated by the adherence to ambivalent sexism and domestic violence myths.

Procedure
Participants were approached in various public places such as university libraries, train stations, and main streets of a city in the south of France. The research was presented to participants as a study on couple relationships and as anonymous, and confidential. It was made clear that they could refuse to participate and that they could stop at any time without any explanation and without any consequences. A questionnaire was given to participants only after receiving their verbal consent. They completed it in the public places where they were recruited. A debriefing was offered to them after completion.

Data Collection
Evaluation of IPV. Based on the procedure of Giger et al. (2017), participants were asked to evaluate an IPV situation presented in a short scenario (see the appendix) through a 12-item questionnaire measuring three subscales: the perceived responsibility of the female character (Victim Blame), the perceived exoneration of the male character (Exoneration of the Perpetrator), and the perceived severity of the violence perpetrated by the male character on the female character (Perceived Severity of Violence). Responses were given on a 7-point Likert-type scale, and overall scores could range from 1 to 7, with higher scores representing higher victim responsibility, aggressor exoneration, and perceived violence severity.
Romantic love. We used the Attitude Toward Love Scale (Knox, 1970; see also Knox & Sporakowski, 1968). This is a 30-item scale measuring the degree to which one holds romantic versus realistic attitudes toward love. Participants answered items on a 5-point Likert-type scale, and overall score could range from 1 to 5 with higher scores representing a greater degree of romantic love. Items were averaged so that higher scores reflected stronger romantic attitudes toward love. Internal consistency (α = .87) in the present study was good.
Ambivalent sexism. Sexism was measured with the Ambivalent Sexism Inventory (Dardenne et al., 2006;Glick & Fiske, 1996). This 22-item questionnaire is composed of two dimensions: hostile sexism, which is composed of negative feelings toward women, and benevolent sexism, which reflects a chivalrous ideology of being sympathetic and protective toward women who agree with conventional gender roles. As allowed by the different validations of this scale, we used its unidimensional version that groups together the two forms of sexism. Responses were given on a 6-point Likert-type scale, and overall scores could range from 0 to 5, with higher scores representing a greater degree of ambivalent sexism. Internal consistency (α = .94) in the present study was better to those reported by Glick and Fiske (2001;α = .81 in Study 1; α = .83 in Study 2).
Domestic violence myths. We used the Domestic Violence Myth Acceptance Scale (Lelaurain, Fonte, Graziani, & Lo Monaco, 2019;Peters, 2008). This 18-item questionnaire measures stereotypical attitudes and beliefs contributing to the minimizing, denial, and even justification of IPV. Responses were given on a 7-point Likert-type scale, and overall score could range from 1 to 7 with higher scores representing greater endorsement of myths surrounding IPV acceptance. Internal consistency (α = .88) in the present study was similar to those reported by Peters (2008;α = .81 in Study 1; α = .88 in Study 2) and by Giger (2017;α = .86 in Study 1; α = .90 in Study 2).

Results
Descriptive statistics and correlations for the relevant measures are reported in Table 1. Pearson correlations were performed to examine whether Romantic Love, Ambivalent Sexism, Domestic Violence Myths, and evaluations of IPV were correlated. These analyses revealed that all correlations were significant (all ps < .05), except for the relationship between Romantic Love and Perceived Severity of Violence. Then, linear regression analyses were performed to examine the predictive ability of romantic love on the evaluation of IPV. Results showed that the more participants adhered to Romantic Love, the more they blamed the victim, β = .31, t(1, 233) = 4.94, p < .001. Thus, our hypothesis H1 was not rejected. Then, the more participants adhered to Romantic Love, the more they exonerated the perpetrator, β = .29, t(1, 233) = 4.66, p < .001. Thus, our hypothesis H2 was not rejected. However, romantic love was not a predictor of Perceived Severity of Violence, β = -.06, t(1, 233) = -.87, p = .39. Our hypothesis H3 was rejected.
Two successive structural equation models were performed from a confirmatory perspective to test our theoretical model according to which patriarchal ideologies mediate the effect of Romantic Love on the evaluation of IPV. Structural equation modeling was used because it allows complex mediation models to be tested in a single analysis (Gunzler, Chen, Wu, & Zhang, 2013;MacKinnon, 2008). In the first step, we tested a model in which Romantic Love, Ambivalent Sexism, and Domestic Violence Myths were related to Victim Blame, Perceived Severity of Violence, and Exoneration of the Perpetrator (Figure 1). This analysis revealed that all standardized path coefficients were significant (all ps < .01), except for the relationship between Domestic Violence Myths and the Perceived Severity of Violence (β = -.09, p = .284). A nonsignificant path was also revealed for the relationship that Romantic Love maintains with Victim Blame (β = .08, p = .155), Perceived Severity of Violence (β = .11, p = .121), and Exoneration of the Perpetrator (β = .07, p = .229). To check for the significance of the indirect mediated effect of these nonsignificant direct effects, bootstrap tests were performed (mean of 5,000 bootstrap samples with bias-corrected 95% confidence intervals [CIs]). These analyses revealed a significant indirect effect for the relationship that Romantic Love maintains with Victim Blame (β = .08, p < .001, 95% CI = [.17, .30]), Perceived Severity of Violence (β = .10, p < .001, 95% CI = [-.22, -.11]), and Exoneration of the Perpetrator (β = .07, p < .001, 95% CI = [.16, .29]). This suggests that ideologies legitimizing patriarchy are significant mediators in the relationship between Romantic Love and the evaluation of IPV. The goodness-of-fit tests yielded a significant chi-square, χ 2 (3) = 29.02, p < .001, which means that the model did not fit the observed data. In the second step, we rectified our theoretical model to decrease the chisquare index: The nonsignificant direct effects were withdrawn and, on the basis of the modification indices, a covariance was added between Perceived Severity of Violence and Victim Blame, and between Victim Blame and Exoneration of the Perpetrator (Figure 2). The analysis of this new model revealed that all standardized path coefficients were significant (all ps < .01).  Goodness-of-fit tests yielded a nonsignificant chi-square, χ 2 (5) = 7.15, p = .210, which meant that the new model fitted the observed data. The model's baseline fit was adequate (RMSEA = .043, SRMR = .030, CFI = .995, TLI = .985). Comparing the models using Δχ 2 revealed a significant difference between the final model and the initial model, Δχ 2 (Δdf = 2) = 21.87, p < .001, which indicates that the model fitted the observed data better when Romantic Love was only related to the evaluation of IPV via patriarchal ideologies. Thus, our hypothesis H4 was not rejected.

Discussion
This research provides an original contribution to the understanding of the mechanism by which an IPV situation could be legitimized. The findings showed that romantic love was associated with evaluations that tend to legitimize IPV, confirming the relationship suggested by some qualitative research (see Gius & Lalli, 2014;Wood, 2001). More specifically, the adherence to this ideology predicts significantly victim blame and exoneration of perpetrator of a case of IPV. They also showed that patriarchal ideologies mediated the relationship between romantic love and the evaluation of IPV, which is consistent with the assumption that romantic love is underpinned by stereotypes legitimizing gender inequalities and violence against women (see Chung, 2005;Singh, 2013;Wood, 2001). Thus, IPV is legitimized insofar as the adherence of individuals to romantic love also leads them to adhere to ambivalent sexism and domestic violence myths; in return, the adhesion to these two ideologies leads individuals to perceive the male aggressor as less accountable and the female victim as more accountable. In others words, this suggests that, behind the adherence to romantic love, the legitimization of IPV is determined by the internalization of patriarchal ideologies defining gender-symbolic roles and justifying male domination. These results stand out from those of other studies exploring the relationships between romantic love and IPV, which have focused more on love as a reason to stay and not seeking help (see Anderson et al., 2003;Dziegielewski, Campbell, & Turnage, 2005;Towns & Adams, 2000) or to justify early signs of IPV (Frías & Agoff, 2015;Power et al., 2006;Pyles, Katie, Mariame, Suzette, & DeChiro, 2012). Indeed, we shed light on another aspect of adherence to romantic love ideology: how romantic love favors the acceptance of IPV by reinforcing the adherence to patriarchal ideologies.
Our findings led us to conduct an ideological and macro-social (Doise, 1986) level analysis to question the social functions of romantic love in Western cultures. Specifically, this would allow us to understand the ambivalence between the strong condemnation of violence in society, perceived as "intolerable" (Herman, 2016), and its persistence as well as its social acceptance. While romantic love depicted in popular culture is conflated with the idea of control (Bonomi et al., 2013;Deller & Smith, 2013), this type of love remains highly socially valued and seems harmless on the surface (Papp, Liss, Erchull, Godfrey, & Waaland-Kreutzer, 2017). Such a conception of love is therefore symptomatic of a culture of male dominance perceived as natural and unproblematic (Dobash & Dobash, 1980). This ideology thus appears as a hegemonic construct that could help to facilitate the negotiation of social norms outlawing IPV against women. A recent study aiming at exploring the representations that emerge in communications when French individuals discuss an IPV situation illustrates this strategy (Lelaurain et al., 2018). Romantic love has been shown to be a salient conditional parameter for legitimizing the deviation from the norm of "one doesn't slap a woman" to which all participants spontaneously refer. Specifically, they discussed different aspects of a situation of male violence against a woman to estimate its severity or its justification. The violence thus appeared as more justified and less severe if the couple were in a long-term relationship or if they were in love.
These considerations finally invite us to take a greater interest in consensual and apparently insignificant cultural objects but which nevertheless play an essential role in the legitimization of IPV against women. Such cultural objects therefore present a particular challenge for researchers who are in close axiological proximity between their own values and those of the society in which they live. The latter always face the risk of being "blinded" by ideologies whose dominant character lies precisely in not being perceived as such. Bourdieu's (2001) considerations about love in the postscriptum of his book on male domination are illustrative of this risk. Although this work is part of a perspective of systematically distancing and deconstructing common sense evidence on gender relations, his specific reflections on heterosexual love mark a break with this line of conduct. Love is therefore thought of by Bourdieu as a feeling that suspends male domination, and not as a vector of this domination as the feminists of his time already conceptualized it.

Limitations and Future Directions
This work has one main limitation concerning the correlational analyses used. Therefore, it invites us to be cautious when interpreting and generalizing the findings as further research is needed to provide causal evidence for our hypotheses. From this perspective, it could be relevant to use experimental procedures to deepen our knowledge of the effect of romantic love on IPV evaluation. One perspective of testing the causal effect of romantic love on the adherence to ideologies legitimizing male dominance and the evaluation of IPV situations could be to use the autobiographical recall technique (see Galinsky, Magee, Gruenfeld, Whitson, & Liljenquist, 2008). Specifically, this technique would consist in asking participants to write about a life event in which they felt in love and then to measure the variables mentioned above. We finally suggest that the development of this kind of research would be useful for enhancing sex education programs, especially for adolescents and young adults, for whom love being essential and eternal are dominant ideas (Singh, 2013). Thus, instead of just trying to change individuals' attitudes and behaviors, these programs could also be directed toward the deconstruction of the dominant representation of love that contributes to the perpetuation of hierarchical gender constructions and to the acceptation of IPV. The knowledge produced by gender studies is an important matter here on which such programs should be based. They are conceptual tools that allow us to initiate a reflection on the problems stemming from gender relationships and to denaturalize certain shared evidence on these relationships, but also to raise young people's awareness of the existence of all the diversity of possible relational ways between men and women. I hung up, so we wouldn't be bothered. John asked who it was. I replied that it was nobody important. He picked up my mobile phone and began to check my calls and he saw that it was Peter. He got furious and threw the mobile phone against the wall. The phone broke into pieces. I told him he had no right to do so. He grabbed my wrist with full force, insisting that I tell him who Peter was. I told him he was hurting me, but he didn't let go of my wrist. I replied that it was my new colleague, with whom I was working on an important project and I had already spoken about him. He looked at me angrily and said: 'Why are you doing everything to make things go badly for us and make me angry?' When I didn't answer him, he continued, saying, 'If I get angry it's because I love you so much,' and he hugged me."

Research Involving Human Participants and/or Animals
The study was carried out in accordance with the 1964 Helsinki declaration and its later amendments, the ethical principles of the French Code of Ethics for Psychologists, and the World Health Organization ethical recommendations for research on violence against women.

Authors' Note
Jean-Christophe Giger is now affiliated with Centre for Research in Psychology -CIP-UAL, Lisbon, Portugal.

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 paper is financed by the National Funds provided by FCT -Foundation for Science and Technology through project CIP-PSI/04345/2013 and by the region of Provence-Alpes-Côte d'Azur (PACA).