Towards understanding the Nordic paradox: A review of qualitative interview studies on intimate partner violence against women (IPVAW) in Sweden

Sociology Compass. 2019;e12699. https://doi.org/10.1111/soc4.12699 Abstract Intimate partner violence against women (IPVAW) is a major public health problem in countries around the world, including the Nordic region. Contrary to what would be expected, as the Nordic countries are rated among the most gender equal in the world, survey data suggests that the lifetime prevalence rates of IPVAW in Sweden and neighboring Nordic countries are among the highest in the EU. This phenomenon, which has been termed the Nordic paradox, requires elucidation. The aim of this review is to explore what previous qualitative studies, based on interviews with or field observations of victims, perpetrators, or professionals working in the area, primarily in Sweden, have to teach us about the nature of this paradox. Three interrelated themes are discussed. The first is how gender equality is perceived and addressed as linked, or not, to IPVAW in Swedish research and practice. The second is how violence tends to be othered, for example, through assumptions that violence is perpetrated by foreign rather than Swedish men. The third is how causal mechanisms of IPVAW have been described, typically with reference to norms and dynamics of masculinity, femininity, and heterosexual interaction.

SocINDEX, and PsychINFO, using the terms "intimate partner violence" and "violence against women," in combination with "qualitative" and "Nordic countries OR Scandinavia OR Sweden OR Denmark OR Iceland OR Norway OR Finland." A total of 710 studies in Swedish and 1,200 studies in English were identified. Abstracts and studies were assessed, and studies that did not fit the inclusion criteria were removed while others were added through following citations and reviewing authors' publication lists. Finally, 116 studies were included in the review, 56 written in Swedish and 60 in English (10 PhD theses, 60 articles, 33 reports, 9 book chapters, and 4 books). The large majority are focused on Sweden, as many studies from other Nordic countries were written in the respective vernacular languages.
All studies except one were based on qualitative interviews, with individuals (85%) or couples or groups (15%).
The number of informants ranged from 2 (Wiklund, Malmgren-Olsson, Bengs, & Öhman, 2010) to 44 (Gottzén, 2013b). A minority (around 6%) of the studies included material gathered through observation. In some cases, the same interview material was used in more than one study. Typically, results were based on thematic, narrative, or discourse analysis.
A complexity of the IPVAW research field is the wide variety of concepts used, which relate to different ways of understanding the phenomenon and which have been used to varying degree over time (Hearn, 2012b;Holmberg, Enander, & Lindgren, 2015). Other search terms, such as domestic violence, domestic abuse, gendered violence, family violence, kvinnovåld, kvinnomisshandel, kvinnofrid, partnervåld, and familjevåld could thus have been used. The selected search terms-intimate partner violence and (men's) violence against women-were however judged to be sufficiently wide in scope and common in use to provide adequate coverage of the subject area. While a choice of different search terms might have slightly altered the selection of studies, the included sources, encompassing those identified through following citations, are therefore deemed to sufficiently underpin the themes emphasized and arguments made in this review. It should be noted, meanwhile, that through disregarding text analyses and quantitative research, the review only grasps a limited section of Nordic IPVAW research.

| Analysis
In search for material that might elucidate the Nordic paradox, the studies were read with focus on gender equality, on juxtaposition of violence perpetrated in Sweden with violence in other countries or cultures, and on explanations of the origin and dynamics of IPVAW. As this material was synthesized, three themes emerged. These are how gender equality is perceived and addressed as linked, or not, to IPVAW in Nordic research and practice, how violence tends to be othered, for example through assumptions that violence is perpetrated by foreign rather than Nordic men, and how causal mechanisms of IPVAW have been described, often pointing to norms of masculinity, femininity, and heterosexual interaction. A common thread discernible throughout these themes is that the strong gender equality discourse present in Sweden may create particular challenges for efforts to address and curb IPVAW.

| IPVAW AND GENDER INEQUALITY: CONNECTED OR NOT
Researchers point to a clash between Swedish discourses of gender equality and realities of IPVAW (e.g., Gottzén & Jonsson, 2012;Wiklund et al., 2010). This perception of a clash is based on the typical understanding of IPVAW as related to gender in/equality, which has been affirmed since the 1970s (Enander, 2010a) and is now well established in Swedish public debate (Eriksson & Berg, 2010;SOU, 1995:60). In Sweden, IPVAW is thus largely considered as a societal problem requiring collective solutions, rather than as a problem of single individuals (Ekström, 2011). Still, the gender perspective remains contested (Eriksson & Berg, 2010) and coexists with other ways of interpreting IPVAW (Helmersson, 2017;SOU, 2004:121). This coexistence of, or "discursive battlefield" (Steen, 2003) between, modes of understanding IPVAW is a reiterated theme in the studies under review. Repeatedly, a gender perspective is positioned against explanatory models centered on factors related to the individual, such as psychological pathology or socioeconomic vulnerability, or to dynamics of specific relationships (e.g., Agevall, 2012;Helmersson, 2017;Holmberg & Bender, 2003;SOU, 2004:121). Enander (2010b) thus posits feminist advocacy and theory against understandings of IPVAW as perpetrated by or against pathologically or otherwise deviant individuals, while Mattson (2013) traces a line between structural feminist understandings and relational therapeutic perspectives.
A main difference traced by Enander (2010b), between feminist research and approaches more focused on individually or socially oriented explanations, is that the latter tend to see IPVAW as a result (of upbringing, psychological disposition, socioeconomic status, etc.), while the former perceives violence as a means (to maintain or achieve power) within the present societal order. As interpreted by feminist scholars, research looking toward pathology, substance abuse, or upbringing as explanatory factors (e.g., Bergman, 1989;Lennéer-Axelson, 1996) tends to perceive IPVAW as an obstruction or deviation from existing societal norms, while research focused on gender rather perceives IPVAW as a way of upholding such norms, as an extreme end of a continuum (Kelly, 1988) of male violence and dominance which finds various expressions in different spheres of society (Enander, 2010b;Lundgren, 2012).
Feminist researchers add that explanations centered on individual or social factors, in the Nordic context, tend to take gender equality more or less for granted (Holmberg & Enander, 2005;Mattson, 2013) while largely perceiving IPVAW as a problem and responsibility of individuals (e.g., Brännvall, 2016;Wiklund et al., 2010).
The coexistence of these broad ways of understanding IPVAW ties into the common replacement of the concept of men's violence against women (mäns våld mot kvinnor) with the more gender-neutral IPV (våld i nära relationer) and with the development of research on male IPV victims and on violence in LGTBQ relationships.
While these developments enable more inclusive perspectives on IPV (e.g., Hellgren et al., 2015), researchers warn that they have tended to come alongside a shift from a more structural understanding of violence toward focus on specific groups or relationships (Brännvall, 2016;Ekström, 2011Ekström, , 2018Helmersson, 2017). Research on male IPV victims relates, further, to the gender symmetry debate (e.g., Nybergh et al., 2015), in which it is posited that men and women are comparably violent in intimate relationships. This notion has been debated, for example with reference to Johnson's (2011) typology of different forms of IPV (intimate terrorism, violent resistance, and situational couple violence; Nybergh et al., 2015), through studies showing that male victims do not describe levels of fear or loss of control comparable to those of female victims (Boethius, 2015;Hellgren et al., 2015;Nybergh et al., 2015), suggesting that men reporting IPV exposure have sometimes also perpetrated (more severe) IPV (Boethius, 2015;Nybergh et al., 2015;Uhnoo, 2011), and arguing that women generally do not have the same ability as men to reinforce societal inequalities for example through economic or sexualized violence. While male victimhood must of course be taken seriously, the mentioned research suggests that, in general, women's and men's experiences of IPV tend to differ in opposite-sex relationships (BRÅ, 2010;Nybergh et al., 2015). Still, the debate, and the associated gender neutrality of concepts, implies that in the realm of violence, women and men are indeed actors on a more or less equal-level playing field.
While Nordic qualitative IPVAW research is quite dominated by feminist perspectives, considering gender inequalities (Agevall, 2012;Brännvall, 2016;Helmersson, 2017;Holmberg & Enander, 2005;Hydén, 1992;Lundgren, 1992) and masculinity norms (Enander, 2010b;Gottzén, 2013bGottzén, , 2014, many researchers blend a feminist approach with other more socially or individually oriented perspectives (Edin et al., 2008;Jansson, 2018;Nybergh et al., 2015), often with reference to the ecological model developed by Heise (1998) and recommended by WHO (Krug, Mercy, Dahlberg, & Zwi, 2002;Edin et al., 2009;Fernbrant, Essen, & Cantor-Graae, 2013;Rejmer et al., 2010;Scheffer Lindgren & Renck, 2008a). Thus, IPVAW is typically perceived as a complex problem which cannot be explained or addressed unidimensionally. The coexistence of different understandings of IPVAW, or of understanding various facets of IPVAW, is likely not a problem in itself. Highlighted by many studies, however, are specific ways in which perspectives coexist and are negotiated in narratives and practices of professionals, survivors, and perpetrators, in their efforts to explain and address IPVAW. Researchers hereby point toward tendencies through which assumptions of gender equality can validate individual-oriented explanations (Enander, 2010b;Wiklund et al., 2010) and toward gaps between principle and practice (SOU, 2004:121) enabling gender perspectives to exist in principle but fall away in practice.

| IPVAW and gender inequality according to IPV professionals
In Sweden, the responsibility for supporting IPVAW survivors has in recent decades largely shifted from the women's shelter movement to publicly funded social services (Ekström, 2018), frequently targeting victims alongside perpetrators and children. Studies of such units (Keskinen, 2005;Mattson, 2011Mattson, , 2013SOU, 2004:121) note that while a gender perspective was seen by professionals to be relevant on a structural level, the individual level was assumed to be more complex (Edin et al., 2008). A gender perspective was thus presented as a self-evident basis of activities but tended to disappear in concrete practice, where dimensions of gender and power were often not addressed (Mattson, 2013), as IPVAW was largely framed as a problem of individuals (Keskinen, 2005). This relative absence of a gender perspective, related to an associated assumption of gender equality (Mattson, 2013), was found to have potentially problematic consequences, for example in cases of custody conflict and of placing substantial responsibility on female victims (Bruno, 2018;Helmersson, 2017;Mattson, 2013;SOU, 2004:121). Alongside this gap between a structural gender perspective and more individually or relationally oriented practice, Mattson (2011) saw tension between an emphasis on gender neutrality and gender symmetry among (male) professionals, often referring to violent relationships as consisting of two violent parties, and the reality observed during field-work during which only women and children sought help due to IPV. Mattson (2011Mattson ( , 2013 further notes that gender perspectives were regarded by professionals as partial and subjective, while the more gender-neutral, relational orientation was invested with higher professional authority (also Helmersson, 2017;Keskinen, 2005;SOU, 2004:121). Meanwhile, the stated theoretical starting points for treatment of male IPVAW perpetrators do not always include gender (Bergström & Rudqvist, 2006), while IPV project managers have emphasized the importance of downplaying gendered terms to not provoke or appear too political (Eriksson & Berg, 2010). In sum, the gender perspective on IPVAW shows tendencies towards being deprioritized.
Within the Swedish judicial system, while government propositions and legal texts reflect a gender perspective (Prop, 1997:55), this appears to not always be mirrored in the practices and assumptions of professionals. While gender perspectives have been present among prosecutors (Törnqvist, 2017) and police (Weinehall, 2011), the individual characteristics and dispositions of persons involved have been of importance not only to professionals' explanations of violence (Eriksson, 1995;Weinehall, 2011) but also to the victims' chances of gaining the judicial system's support (Agevall, 2012;Augustsson & Kuno, 2006;Brännvall, 2016;Eriksson, 1995;Jarnkvist & Brännström, 2016b;Lundberg, 2001;Nikupeteri, 2016), which in turn may affect the continuation or ceasing of IPVAW (Brännvall, 2016;Nikupeteri, 2016). Individualization of the problem, through focus on the woman's behavior, here aligns with placing responsibility on the victim (BRÅ, 2010;Ekström, 2011;Eriksson, 1995).
Within health care, despite overarching understandings of links between IPVAW and gender inequality, individually or socially oriented understandings of IPVAW have been observed among professionals (Bengtsson-Tops et al., 2009;Gillå et al., 2018;Henriksen et al., 2017;Husso et al., 2012;Mørk et al., 2014;Pratt-Eriksson et al., 2014;Socialstyrelsen, 2002;Sundborg et al., 2015), alongside tendencies to blame the victims for violence and for not leaving the relationship (Bengtsson-Tops et al., 2009;Leppakoski et al., 2011;Örmon et al., 2014a;Pratt-Eriksson et al., 2014). Among professionals working at schools, Bruno (2012) found an understanding of IPVAW as consisting of conflict between essentially equal parties, associated with socioeconomic vulnerability or foreign descent. This had consequences for the professionals' ways of handling situations, including their placing of large responsibility on female victims.
A gap between perspectives and levels was further found by Holmberg and Bender (2003), in their study of municipal decision makers and professionals, which found no reflection of the nationally formulated prioritization of IPVAW, encompassing a gender perspective, in local-level discussion.

| Narratives of male perpetrators
In contrast to studies from other countries expressing high levels of acceptance of IPVAW among men (Boethius, 2015;Gottzén, 2013b;Gottzén & Korkmaz, 2013), the reviewed studies point to low or no acceptance of IPVAW among interviewed male perpetrators (Gottzén & Jonsson, 2012;Uhnoo, 2011). Thus, nonacceptance of IPVAW coexists, alongside the strong Swedish gender equality discourse (e.g., Gottzén & Jonsson, 2012;Wiklund et al., 2010), with actual exertion of IPVAW. In men's explanations of their violence, individually or socially oriented perspectives are prominent, together with emphasis on the reciprocity of violence. Gender perspectives are uncommon.
Despite gendered perspectives on IPVAW in Swedish public discussion, then, among male perpetrators, individually or socially oriented explanations are common, alongside victim-blaming and emphasis on gender neutrality or symmetry.

| Narratives of female survivors
Blends of gender perspectives and other explanations are further found in the narratives of female victims, where structural explanations (Andersson & Lundberg, 2000;Augustsson & Kuno, 2006) and references to types of masculinities (Gottzén & Korkmaz, 2013;Uhnoo, 2011) coexist with explanations focused on the pathology, social situation, or substance abuse of the violent partner (e.g., Agevall, 2012;Almqvist & Broberg, 2004;Gottzén & Korkmaz, 2013). The latter is particularly common when the woman is still in the relationship (Hydén, 1992) or recently separated (Brännvall, 2016;Enander & Holmberg, 2008). Some women speak of control and power, but without connecting it to broader societal patterns (Enander, 2010b;Uhnoo, 2011). In Enander's (2010b) study of women's views on IPVAW, explanation based on pathology or deviance clearly dominated over gender perspectives. Thus, in women's narratives, IPVAW is largely perceived as a personal rather than a structural problem (also Uhnoo, 2011).
Women self-identifying as strong and independent have, accordingly, been found to experience greater feelings of shame and guilt for IPVAW exposure (Scheffer Lindgren & Renck, 2008b). Feelings of shame, and relatedly of stupidity (e.g., Edin et al., 2010;Enander, 2010a;Hydén, 2005;Örmon et al., 2014b), are thus connected to unfulfilled expectations of assumed gender equality, and of cultural expectations of "positive" Swedish femininity denoting strength and self-sufficiency (Enander, 2010a). Such feelings have hindered women from seeking help (Gottzén, 2013b; and may discourage victims from leaving violent relationships (Enander, 2010b). As assumptions of gender equality relate to perceptions of IPVAW as an individual problem, this corresponds with noted tendencies towards seeing victims as responsible for their choices of partners and for stopping the violence, and towards framing IPVAW as the responsibility and the personal failure of the women themselves (Gottzén & Korkmaz, 2013;Wiklund et al., 2010).

| IPVAW and gender inequality-Linked or not?
In sum, gender inequality is sometimes linked to IPVAW by professionals and victims, if not by perpetrators, and sometimes not. While this may bear testament to the importance of diverse causative factors of IPVAW, it has also been linked to gaps between principles and practices (SOU, 2004:121) and to a disjunction between the assumed Swedish gender equality and existing power relations between the genders (Holmberg & Bender, 2003). As presupposed absence of gender inequality aligns with and may validate individual-oriented explanations, gender neutrality creates particular conditions and challenges for efforts to handle and curb IPV (Enander, 2010b).

| OTHERING OF IPV
The concept of othering has been defined as a processes serving "to mark and name those thought to be different from oneself" (Weis, 1995, p. 18), in construction of difference from the self or the mainstream (Johnson et al., 2004). Such processes may involve projection of aspects of self onto others (Griffin, 1981) and can perpetuate positions of domination and subordination in society (Hall, 1991;Johnson et al., 2004). Several sources under review have noted tendencies towards making IPVAW into something of an Other, in Sweden and neighboring Nordic countries. The Swedish discourse of gender equality, central to the national self-image (Gottzén, 2016), aligns with views on IPVAW as committed by individuals and groups deviating from "ordinary" men in "normal" heterosexual relationships, within the gender-equal majority culture (Agevall, 2012;Brännvall, 2016;Eliassi, 2015;Rejmer et al., 2010). Gottzén (2013b) points to a dichotomous relationship between the "regular man" and the "woman abuser" in Sweden, as IPVAW is "primarily seen as conducted by men in other cultures and in other times" (Gottzén, 2016, p. 165).

| The foreign Other
Distinctions are commonly made, in public discussion, between gender-equal and nonviolent Swedish culture, within which individuals may use violence due to pathology or vulnerability, and patriarchal and violent foreign cultures (Brännvall, 2016;Keskinen, 2011;Lundgren, 1992;SOU, 2004:121). Gender equality has thus been used as a marker of dichotomous difference between allegedly "gender-equal Nordics" and "patriarchal immigrants" (Keskinen, 2011).
Not only is the violent man taken to be the gender-equal man's Other, but as ordinary Swedish men are assumed to be gender equal, he becomes Other of the Swedish man (Gottzén & Jonsson, 2012).
While studies focused on immigrated women are rare within this review (Andersson & Lundberg, 2000;Eriksson & Berg, 2010;Fernbrant et al., 2013), included studies clarify that a large number of IPVAW perpetrators and survivors are indeed of native Nordic origin (Ekström, 2016;Hydén, 1992;Olsson & Bergman, 2018). One study (Georgsson et al., 2007) finds foreign background to be more common among victims than among perpetrators, while another (Andersson & Lundberg, 2000) suggests that women arriving in Sweden through partnership with a Swedish or foreign man are subjected to comparatively severe abuse.
Some interviewed professionals make distinctions between "ordinary" IPVAW and honor-related violence (HRV) perpetrated by immigrants (Finnbogadottir & Dykes, 2012;Helmersson, 2017;Olsson & Bergman, 2018). Others cannot see systematic differences between women with Swedish and foreign backgrounds, regarding the nature and effects of IPVAW (Alsaker et al., 2016;Andersson & Lundberg, 2000;Mattson, 2011;Scheffer Lindgren & Renck, 2008a), the dynamics of marital projects (Hydén, 1992), or the fear of reporting or leaving violent partners (Jarnkvist & Brännström, 2016b). Agevall (2012) argues that jealousy, a commonly reported trigger of IPVAW in Sweden, is not essentially different from concepts of honor or masculinity associated with foreign cultures. Distinctions between IPVAW and HRV point, further, to the importance of family members as supporters of violence in the latter. Studies show that while supportive social networks can be crucial for survival from IPVAW in Sweden (e.g., Ekström, 2011;Flinck et al., 2005;Jarnkvist, 2015;Scheffer Lindgren & Renck, 2008a), social responses are diverse (Hydén, 2015;Sandberg, 2016) and include contribution to IPVAW by parents and other social contacts through downplaying violence and supporting perpetrators (Alsaker et al., 2016;Edin et al., 2010;Engnes & Lundgren, 2013;Gottzén & Korkmaz, 2013;Green, 2002;Hopstadius, 2014;Münger, 2009;Rejmer et al., 2010;Wiklund et al., 2010). Disclosure or reporting of IPVAW has been seen to bring shame or upheaval to families and social circles (Agevall, 2012;Edin et al., 2010;Green, 2002), while reminiscent of renderings of HRV, young Swedish IPVAW victims describe processes of "dwindling life space" due to violation (Wiklund et al., 2010). Pointing to ways in which the Swedish welfare state has indirectly legitimized IPV, Holmberg and Enander (2005) highlight that arguments with which men from Sweden can use violence against their partners differ from those of men more traditional patriarchal societies more in terms of form than content.
Simultaneously, studies emphasize that while immigrated women form a heterogeneous group, factors related to migration, economic insecurity, social isolation, and discrimination may increase vulnerability to IPVAW . Differences between gender structures in Sweden and countries of origin may create friction (Andersson & Lundberg, 2000;Fernbrant et al., 2013). Immigrated women may be met with racism (Pratt- Eriksson et al., 2014) or have needs that the Nordic welfare institutions are not organized to meet Helmersson, 2017;Olsson & Bergman, 2018), while the perceptions among professionals that violence is accepted in some cultures may also lead to nonintervention (Holmberg & Bender, 2003;Keskinen, 2011;Sundborg et al., 2015). Keskinen (2011) emphasizes the dangers both of culturalizing explanations of violence as resulting from static foreign cultures, and of universalizing assumptions that dynamics of gendered violence are the same everywhere. While avoiding potentially racist categorization of difference, the latter may impede the addressing of IPVAW occurring among different individuals and groups in society.
In conclusion, while Keskinen (2011) points to the importance of not bypassing IPVAW perpetrated or suffered by persons with migration backgrounds, the studies briefly referenced above show that a simple dichotomy of genderequal nonviolent Swedish citizens and violent patriarchal foreigners does not hold up to scrutiny. Meanwhile, the framing of IPVAW as something that only occurs in non-Swedish cultures contributes to rendering inequality and IPVAW in Sweden invisible (SOU, 2004:121), while simultaneously feeding into xenophobic anti-immigration discourses (Eliassi, 2015).
Resistance to victimhood has been tied to the image of IPVAW victims as different to "normal" women, related to individually or socially oriented IPVAW explanations emphasizing pathology or vulnerability, and to the image of the "ideal" crime victim (Christie, 1986) as helpless and unresisting while violated by a deviant unknown perpetrator (Brännvall, 2016;Enander & Holmberg, 2008). Resistance to IPVAW victimhood is thus associated with unwillingness to deviate from positive femininity in gender-equal Sweden (Brännvall, 2016); a deviation which Holmberg and Enander (2005) relate to as "standing in discursive head-wind" (p. 179). While making it difficult for women to conflate IPVAW victimhood with self, preconceptions of IPVAW victims, alongside assumptions or expectations of gender equality, can thus create obstacles for identifying and reporting IPVAW in Sweden (Brännvall, 2016;Enander & Holmberg, 2008).
IPVAW victimhood is further complicated by women's experiences of failing to receive legal support or redress because they did not fit the image of the battered woman (Nikupeteri, 2016), due to being confident, well-dressed, married to a wealthy partner (Jarnkvist, 2015), or not being sufficiently innocent due to resistance, substance abuse, or socioeconomic vulnerability Münger, 2009). Women who act in accordance with the helpless ideal victim have had better chances in legal proceedings (Lundberg, 2001), while women who show agency and initiative risk having their victim status undone (Agevall, 2012;Brännvall, 2016;Eriksson, 1995). Wiklund et al. (2010) states that women are often dichotomized as either being strong or a potential IPVAW victim. Thus, the "ideal" IPVAW victim and the ideal, strong, and gender-equal Nordic woman do not appear to mix well.
Dissociation is effected through differentiating between the self and attributes of those assumed to be "real" IPVAW perpetrators (Boethius, 2015;Gottzén, 2013b) and between the "real" self and the act/s of violence (Edin & Nilsson, 2014). The self is said to not be recognizable in the violent acts (Edin & Nilsson, 2014), which were caused by "someone else," due to alcohol, drugs, neurochemistry, blackouts, or loss of control (Boethius, 2015;Gottzén & Jonsson, 2012;Hydén, 1992). Hydén (1992) speaks of an absence of the self as agential subject in men's IPVAW accounts, while Gottzén and Jonsson's (2012) informants speak of violence as forgotten or repressed. Håland et al. (2016) note that perpetrators seem to experience cognitive dissonance between their different behaviors.
Fear of being pointed out as an IPVAW perpetrator-of becoming the Other-has hindered men from disclosing their violence (Gottzén & Jonsson, 2012;Gottzén & Korkmaz, 2013). Simultaneously, experienced gaps between selfimage and actions have been described as a motive for help-seeking (Gottzén & Jonsson, 2012). In their description of men enrolled at IPV units, Gottzén and Jonsson (2012) explore how they "try to avoid the shadowlike figure that is pointed out as the Other of gender equality, while not necessarily trying to become gender equal men" (p. 150).
Images of "real" or "ideal" perpetrators come into play in the legal arena, as men convicted for IPVAW have held lower social positions with regard to class and ethnicity, while those cleared were more "ordinary" in terms of Swedish ethnicity or class privilege (Agevall, 2012;Brännvall, 2016). This may reproduce assumptions that IPVAW is perpetrated by men who deviate from a white middle-class norm (Agevall, 2012;Brännvall, 2016). Thus, the "ideal" IPVAW perpetrator does not appear to blend easily with the "ordinary" Nordic man.

| Violence as Other
IPVAW is made Other through disappearing or being bypassed, due to resistance toward or difficulties with recognizing, speaking about, or taking responsibility for it. Such rendering of violence as invisible is commonly noted in feminist IPVAW research (Törnqvist, 2017).
Violence has been noted to "disappear" in public discussion and practice, through representation of IPVAW as infrequent, as a side effect of other problems, or as a problem only relevant to Other groups. Such "disappearance" has been related to a willingness to raise the issue of IPVAW in theory but not in practice, due to assumptions of prevailing gender equality (SOU, 2004:121). Municipal professionals interviewed by Holmberg and Bender (2003) often referred to IPVAW as a problem of "others," such as tourists, immigrants, city dwellers, or substance abusers. Holmberg and Bender (2003) therefore call IPVAW an "escaped problem," while referring to gender inequality, also found to often be a non-issue on the municipal level, as "untouchable." Within health care, violence has been bypassed through midwives "beating about the bush" and affected women "keeping up a front" (Edin et al., 2010) and through health professionals fixing injuries or focusing on diagnoses while bypassing violence as their cause (Husso et al., 2012;Örmon et al., 2014a;Örmon & Hörberg, 2016). Women report having disclosed violence in health care settings without getting any response (Leppakoski et al., 2011) or being disbelieved or blamed and thus feeling invisible (Örmon et al., 2014a). Health care workers, in turn, report a lack of support structures for asking about and responding to IPVAW (Finnbogadottir & Dykes, 2012). In her study of a social service IPV-unit, Mattson (2011Mattson ( , 2013 argues that the therapeutic relational approach often had an effect of rendering violence invisible, for the benefit of other aspects of the relationship or situation of parties involved, particularly by male professionals and when focus was aimed towards men. Professionals further found difficulties in speaking about violence; about how its origin and dynamics were understood, and how it related to gender and power (Mattson, 2011(Mattson, , 2013. Violence has also tended to be disregarded in Nordic custody conflicts (Bruno, 2018;Enander, 2011a;Häggblom & Möller, 2007).
In Brännvall's (2016) interviews with women who had not reported IPVAW to the police, they commonly pointed to ways in which they thought the violence they were subjected to differed from "real" IPVAW. Häggblom and Möller (2007) similarly quote a woman who was almost killed by her husband, saying it "was not real violence against women that I experienced" (p. 172). Meanwhile, a woman stated that for a long time, she "had no words for the abuse" (Flinck et al., 2005, p. 388). Other survivors (Edin & Nilsson, 2013;Scheffer Lindgren & Renck, 2008b) and perpetrators (Agevall, 2012;Hydén, 1992) describe violence disappearing through coping mechanisms of forgetting or repressing memories. Accordingly, breaking free from IPVAW is related to remembering and reminiscing (Flinck et al., 2005); to recognizing the violence (Enander & Holmberg, 2008), telling one's story (Hydén, Gadd, & Wade, 2016), and being acknowledged (Örmon et al., 2014a;Örmon & Hörberg, 2016). Dealing with committed IPVAW has similarly been described by Håland et al. (2016) as moving from denial to recognition of oneself as a perpetrator. Such encounters with the self as IPVAW perpetrator "can be understood as an encounter with the Other as one's own mirror image" (Gottzén & Jonsson, 2012, p. 156). As efforts to address IPVAW in Sweden require that the problem be recognized and made visible, on personal and societal levels, then, it appears that work remains to be done.

| TOWARDS EXPLAINING VIOLENCE
Every reported case of IPVAW in Sweden, writes Hydén (1995a), is a sign that Swedish efforts toward gender equality have not been sufficiently successful. How, she asks, "can these basically impossible actions be possible" (p. 7)?
Researchers have attempted to answer this question through pointing to dynamics of normalization, marital acts, and norms of masculinity, femininity, and heterosexual interaction. Lundgren's (1992Lundgren's ( , 2012 theory of normalization has been influential in IPVAW research and in the women's shelter movement. In brief, it denotes a process through which a relationship gradually becomes more violent and controlling, while the woman progressively adapts to her partner's demands, internalizing his perceptions of the violence and of her. Central to the pattern is the man's oscillation between violent and loving behavior, and the woman's social isolation. This process of control and isolation, as well as internalization and breakdown of the woman's self-esteem, can cause her, to some degree, to accept and even defend the violence. This process constitutes gender, as the man typically strives to not only control the woman but to discipline her towards a notion of ideal femininity, and as violence, partially through increasing sexualization, contributes towards constitution of masculinity. IPVAW should thus, according to Lundgren (1992), be seen as a product of gender norms persisting in Nordic countries despite movements towards gender equality. While Lundgren's concept of normalization has been critiqued, for awarding limited space for resistance, complexity, and change (Hydén, 1992), its significance for research (e.g., Enander & Holmberg, 2008;Häggblom & Möller, 2007;Mørk et al., 2014;Örmon & Hörberg, 2016;Scheffer Lindgren & Renck, 2008a), and for social and legal professionals (Törnqvist, 2017), is clear. Normalization offers explanation to noted tendencies among women to redefine violence and blame themselves. The concept of normalization has been applied to women's stories about violence increasing step by step, while their life space diminished (Gottzén & Korkmaz, 2013;Holmberg & Enander, 2005;Wiklund et al., 2010) and they adapted (Brännvall, 2016;Enander & Holmberg, 2008), and about how it became difficult to understand what happened, leave the relationship (Jarnkvist, 2015), or seek help (Hopstadius, 2014). Lundgren's concepts can thus mediate understanding of why women perceiving themselves as strong individuals can remain in violent relationships (Gottzén & Korkmaz, 2013;Scheffer Lindgren & Renck, 2008b). Further, studies point to normalization of IPVAW occurring in the legal arena (Agevall, 2012;Brännvall, 2016).

| Normalization and marital acts
Another central contribution comes from Hydén's (1992Hydén's ( , 1995aHydén's ( , 1995b focus on interaction within marriage and on successive stages of violent acts. The prehistory of violence is here located in verbal fights, typically based on discording perceptions of the marital agreement, and of whether it is being followed. According to Hydén, such verbal fighting serves to maintain a hierarchical social order, which is heightened through the violence. The man's purpose of using violence is thus to enforce subordination, through influencing the woman to act differently. After the violent event, continuation of the marriage typically requires mutual narratives changing the content or meaning of the violence and releasing the man from responsibility. Thus, when the marriage is maintained, women's stories typically become similar to the men's, downplaying violence and emphasizing its mutuality. This neutralization of the man's responsibility further contributes to the possibility of violence recurring (Hydén, 1992(Hydén, , 1994. While highly impacting later research, Hydén has also been critiqued primarily for an insufficiently strong focus on gendered power structures (Agevall, 2012).
Under influence from Lundgren and Hydén, Enander and Holmberg (2008) speak of normalization in terms of mechanisms creating space for interpretation and renegotiation of violence. Some of these relate to what they term the traumatic bond between perpetrator and victim (Holmberg & Enander, 2005). Distinguishing between aspects of leaving violent relationships, Holmberg and Enander point to the importance of recognizing events as IPVAW, as interviewed women typically had to undergo a mental paradigm shift to see themselves as IPVAW victims. This often happened after they had left (also Brännvall, 2016;Jarnkvist, 2015). Lundgren (1992) and Hydén (1994) both link IPVAW to gender norms in Nordic society. Lundgren (1992) relates violence to norms connecting masculinity to virility, power, and violence, while correlating the shrinking life space of victims with the limited space awarded to, and subordination expected of, women in general. Hydén (1992) argues that IPVAW should not be regarded as an expression of individual men's personal characteristics or impulses, but as a product of Swedish norms pertaining to masculinity and marriage. While romantic ideals do not celebrate violent control per se, they emphasize male strength, independence, and ability to provide for women, in exchange for free access to their bodies.

| Masculinity and femininity
Along related lines, and influenced by seminal theorists like Hearn (1998Hearn ( , 2012a, a range of researchers and informants relate IPVAW to norms of masculinity, femininity, and heterosexual relationships. Violence has thus been seen to reaffirm masculine identity (Edin & Nilsson, 2014;Jansson, 2018;Pratt-Eriksson et al., 2014). In a society where men are allowed or expected to be more violent than women, and violence may be excused with reference to provocation or alcohol (Römpötti, 2009), men learn to use legitimate violence through military training and are expected to dominate to gain power and control. Men are further expected to hold back empathy for purposes of competitiveness, while aggression is perhaps the only emotion accepted as part of masculine expression (Edin et al., 2008;Edin et al., 2009). Although Swedish men reject the acceptability of IPVAW (e.g., Edin et al., 2009), studies suggest that some violent men do assert the right to dominate (Rämgård, 2016;Römpötti, 2009) and have unlimited sexual access to (Augustsson & Kuno, 2006;Flinck et al., 2005) their partner, as well as to influence her towards femininity norms (Augustsson & Kuno, 2006;Münger, 2009).
Simultaneously, studies point to tensions arising from the coexistence of ambiguous or paradoxical masculinity norms (Edin et al., 2008;Jansson, 2018) and to notions of "good" versus "bad" violence (Agevall, 2012). While the power of "being a man" may be confirmed through IPVAW (Edin, 2006), other norms pose that IPVAW is not a legitimate feature of a "real man," who should practice self-control, care for his family (Edin & Nilsson, 2014;Gottzén, 2013b), and be gender equal (Gottzén, 2016). Protecting women from other men's violence is constructed as heroic masculinity (Gottzén & Korkmaz, 2013), as masculinity is framed as both problem and solution (Sandberg, 2016). The assumed acceptability of male violence has further depended on the femininity of the woman, on whether she was weaker than the man or strong and thus "beatable" (Gottzén & Korkmaz, 2013). IPVAW has been described as being triggered by women abandoning feminine characteristics, through being bossy or pushy, or by making the partner "less of a man" by claiming power in the relationship (Edin & Nilsson, 2014). Thus, as noted, while perpetration connects with masculinity norms, victimhood ties into notions of femininity. A master narrative of femininity in contemporary Sweden, Jarnkvist (2015) argues, corresponds with the image of the "ideal victim" (Christie, 1986); passive, helpless, and subordinate. Correspondingly, interviewed social service professionals saw female victims as trapped by empathy, raised to care and take responsible for the well-being of others and to adjust to aggressive male behavior (Edin et al., 2008;Edin et al., 2009). One informant spoke of how her violent partner educated her to become more appropriately feminine, meaning not being active or taking space (Münger, 2009), while other described strategies for avoiding IPVAW follow a similar pattern: getting out of the way, adapting, not expressing personality traits, and obeying unquestionably (Edin & Nilsson, 2014).
Ambiguity is present, as another mentioned and central notion of positive femininity in the Nordic countries lies in independence and gender equality. Thus, women aligning closely with "ideal victims" have been most likely to receive requested help from IPVAW support systems, but also least likely to ask for help or leave the relationship (Jarnkvist, 2015;Jarnkvist & Brännström, 2016a). Women positioning themselves as strong and independent are meanwhile more likely to leave violent relationships but less likely to receive support. While notions of femininity may help maintain women in abusive relationships (Jarnkvist, 2015;Jarnkvist & Brännström, 2016a), then, existing IPVAW support systems appear to be geared toward passive subordinate femininity rather than toward women claiming strength, agency, and gender equality, even though the latter is largely expected of women in Nordic countries (Agevall, 2012;Brännvall, 2016;Lundberg, 2001).

| Norms of heterosexual interaction
Studies point to ways in which gender roles played out in Swedish heterosexual relationships contribute towards IPVAW, and towards maintaining women in violent relationships (Enander, 2011a). Norms of, and gender inequalities found within, "normal" heterosexual relationships can thus contribute towards normalization of IPVAW (Brännvall, 2016).
IPV professionals (Edin et al., 2008) see an origin of IPVAW in men's romantic expectations of perpetually understanding and responsible partners, and in their generalization of women into "the whore and the Madonna." Other notions of romantic love that may spill over into violence (Agevall, 2012;Green, 2002;Häggblom & Möller, 2007) include expectations on men to take sexual initiatives and be driving and on women to respond and "give in" (Brännvall, 2016), as well as notions that violence might end if the woman loves her partner more (Edin et al., 2010;Enander, 2011a). Accordingly, Green (2002) quotes a woman speaking about having mistaken her partner's jealous and later violent behavior for love.
In her study of young Swedish people's perception of rape, Jeffner (1998) shows how heterosexual relationship norms create space for renegotiation of violence. While all of Jeffner's informants agreed on the unacceptability of rape, ample space for negotiation existed, all of which decreased the freedom of movement for girls and increased it for boys. Both female and male informants related to the perceptions and expectations of boys, in accordance with hegemonic norms and, one might add, in alignment with principles underlying internalization of violence. A genderneutral ideal of equal opportunities was thus found to stand at odds with gendered practice, as "norms of the ideal image render the hegemonic norms invisible" (Jeffner, 1998, p. 288). Referring to Jeffner, Brännvall (2016) uses the concept of normality markers to convey how women's stories about unreported sexual IPVAW suggested that violence could be interpreted as "normal" sex, even though it was against the women's will. Similar renegotiations, rendering power inequalities between men and women invisible, can be seen within the legal system (Brännvall, 2016). Agevall (2012), also relating to Jeffner, further points how norms of heterosexual coupledom enable sliding definitions of violence, for example through framing jealousy, protectivity, or idealization as expressions of control or love.
Coupledom norms include adaptation to the other and creating a mutual world together. As women are expected to love and adapt more, the balance typically tips in practice (Agevall, 2012). This, in turn, aligns with concepts of internalization and normalization. Edin et al. (2010) similarly speak of women's descriptions of gendered adjustments, comparable to those made in nonviolent relationships. These include reinterpreting subordination as love and caring, and downplaying problems to save face for both. Again, IPVAW is seen as an extreme version of prevailing gender positions in heterosexual relationships, enabling fluid limits between violence and nonviolence, as gendered behaviors may gradually accelerate. Jeffner (1998) and Edin et al. (2010) point to potential conflict between resisting IPVAW and adhering to femininity norms. Stating a clear no to nonconsensual sex follows the rules in relation to rape but breaks norms of positive heterosexual femininity (Jeffner, 1998). Leaving a violent relationships follows IPVAW rules, whereas maintaining positive femininity through showing compassion and keeping things together entails staying (Edin et al., 2010). Similarly, Uhnoo (2011) notes that women who do not report IPVAW because of love, or to not be "mean," align with ideal feminine understanding and acceptance. Hydén's (1994) observation that norms of heterosexual relationships place Nordic women and men in paradoxical situations appears, then, to have merit.

| Gender roles in conflict
Ambiguities and tensions between norms of femininity and masculinity in Nordic countries can be related to demands for gender equality, which can in turn may provoke IPVAW, as suggested by the backlash hypothesis forwarded with regard to international (Eriksson & Mazerolle, 2013) and Nordic (Bjelland, 2014;Caman, Kristiansson, Granath, & Sturup, 2017; Johnsson-Latham, 2014) contexts. Professionals working with IPVAW perpetrators speak along these lines, while observing that societal messages about desired male behavior are indeed ambiguous and that tensions may occur when traditional gender norms are thrown into confusion due to demands for equality (Edin et al., 2008). Violence may be a way of reaffirming masculinity when it is perceived as threatened (Scheffer Lindgren & Renck, 2008a), while young men are noted to experience difficulties knowing how to interact in heterosexual relationships (SOU, 2004:121). Researchers further point to violence as originating in conflict between norms in the context of changing (Flinck et al., 2005) or coexisting (Fernbrant et al., 2013;Gottzén & Korkmaz, 2013) gender roles in contemporary Scandinavia. Potentially violent consequences of female insurgence is further seen by social workers noting that when IPVAW survivors regain strength, they may challenge their partners which in turn can lead to more violence (Helmersson, 2017). This corresponds with evidence that the risk of lethal IPVAW is highest at times of separation (e.g., Ekbrand, 2006). This view on violence due to gender role conflict is further supported by Hydén's (1992) finding that violence often originates in differing interpretations of the marital agreement, and of whether it is being followed. Discord may occur, Hydén argues (1992), regardless of whether the relationship is organized according to a traditional model, with complementary roles for men and women, or along quasi-symmetrical lines idealizing equality. Similarly, Agevall (2012) points to romantic versus confluent love relationships (Giddens, 2013), arguing that while the latter rests on an assumption of gender equality, in reality inequalities are proven to persist. While ideals of symmetrical or confluent love relationships exist alongside gender equality discourses in contemporary Sweden, then, obscured struggles for agreement on gender positions within the context of existing inequalities, potentially enforced by violence, persist. Reminiscent of previous sections, gaps exist between principles and expectations of gender-equal relationships and practices and experiences of gendered inequality and violence.

| Layers of dominance
Much of the above can, rightly so, be critiqued for its heteronormative bias and for disregarding male IPV victims.
While the presence of male and nonheterosexual IPV victims can challenge gender perspectives on IPV (Hellgren et al., 2015), studies of these groups can also provide valuable contributions to gender analysis (Agevall, 2012).
Norms of masculinity, and of heterosexuality, appear in studies on male IPV victims. Abused men have had trouble defining themselves (Hellgren et al., 2015;Jarnkvist & Brännström, 2016b;Simmons et al., 2016), and being supported as victims (e.g., BRÅ, 2010; Kullberg et al., 2015;Simmons et al., 2016;Hellgren et al., 2015), as exposure to violence from female partners is not included in normative understandings of masculinity, according to which men are expected to be stronger and exert more power in heterosexual relationships (Jarnkvist, 2015). Conforming to hegemonic masculinity has been identified as an influence towards IPV nondisclosure in health care, while help is often sought from female professionals (Simmons et al., 2016). Victimized men have further described being scorned with sexist or homophobic language, by violent partners, for not fulfilling masculinity norms (Nybergh et al., 2015). IPV against men has also been interpreted by social workers as being triggered by a lack of masculine traits (Kullberg et al., 2015). Gender norms thereby do not appear to be irrelevant for IPV against males.
IPV against men further actualizes the fact that not all men use violence but that IPV can rather be seen as a way of constituting masculinity (Enander, 2010b), through practices that people of any gender may reproduce (Agevall, 2012).
IPV can thus be understood in terms of practices associated with masculinity, of exerting dominant control, decreasing freedom of movement, and placing responsibility on the other, within a relationship characterized by a hierarchical social order of dominant/subordinate positions of power (Hydén, 1992). This perspective aligns with experiences of revictimization by institutions. With reference to Kandaswamy's (2010) metaphor of abused women in need of welfare having to trade in a man (i.e., an individual) for "the man" (i.e., an institution), Ulmestig and Panican (2015) point to parallels between women's descriptions of IPVAW and their feelings of being economically controlled by social workers.
Similar descriptions appear in studies relating experiences of revictimization by legal, health care, or social welfare system (e.g., Häggblom & Möller, 2007;Hellgren et al., 2015;Nikupeteri, 2016;Örmon & Hörberg, 2016;Pratt-Eriksson et al., 2014), or of such institutions being felt to do the perpetrator's bidding (Bruno, 2018;Jarnkvist, 2015;Jarnkvist & Brännström, 2016b). Women's experiences of the legal system, for example, have been described as having beaten again (Brännvall, 2016) or as a new assault (Pratt-Eriksson et al., 2014). Meanwhile, midwives report feeling resistant to addressing IPVAW due to feeling unprotected toward the perpetrator (Finnbogadottir & Dykes, 2012). Potentially violent relationships of power and domination, then, do not only exist between individual males and females. In their analysis, however, power should not be left out of the picture.

| CONCLUSION AND DISCUSSION
The review does not provide a resolution to the Nordic paradox. It does, however, show that although the gender perspective has had a dominant position in Swedish research and public discussion on IPVAW since the 1970s (Hydén, 2000), it coexists with and in practice often falls away under other more individually or relationally oriented explanatory models. The review has pointed to how, in the context of a strong Nordic gender equality discourse, IPVAW is often assumed to be perpetuated by and against Others, while violence itself tends to disappear from discussion and practice, even though the images of the IPVAW victim as a weak, helpless, and unrelated to the genderequal Swedish woman, and of the perpetrator as a deviant or foreign Other to the Swedish man, do not fit with realities of IPVAW in Sweden. In addition, the review has discussed norms and dynamics of normalization, masculinity, femininity, and heterosexual interaction as contributing towards IPVAW in Sweden. While there is no reason to downplay the progress that has been made in Sweden and neighboring Nordic countries in efforts to address IPVAW (e.g., Nikupeteri, 2016;Olsson & Bergman, 2018), we conclude that challenges remain. In the face of achievements towards gender equality, norms and dynamics of gendered power in intimate relationships appear to continue to contribute towards IPVAW, which in itself feeds into increased gender inequality (Alsaker et al., 2016). Such norms and dynamics of power are further still reflected in legal, social, and health care institutions.
As noted, the fact that gender inequality is sometimes linked to IPVAW and sometimes not can result from the existence of diverse causative factors of IPVAW. The ecological model recommended by WHO for IPV research (Heise, 1998;Krug et al., 2002) and mentioned by several studies under review (Edin et al., 2008;Rejmer et al., 2010;Scheffer Lindgren & Renck, 2008a) integrates structural-level causative factors with individual-level ones.
Efforts to grasp both gender perspectives and other more individually or socially oriented perspectives are obviously ongoing and complex. Regarding the Nordic context, this review suggests that the existing discourse of gender equality may create particular challenges, as assumptions of gender equality can obscure IPVAW and related gendered dynamics of power, while validating focus on individual-oriented IPV explanations and also making it difficult for men and women to recognize themselves as IPV victims and perpetrators and to seek and receive help. Thus, a challenge for future research and practice on IPVAW in the Nordic countries lies in making sure that the importance of gendered relationships of power for IPV is not disregarded.
The study of male victims of IPV and of IPV in LGBTQ relationships, alongside tendencies towards gender neutrality in IPV research and practice, challenges static dichotomies of male perpetrators and female victims of violence.
Simultaneously, such studies and tendencies may strengthen assumptions of existing gender equality and divert focus from structural relationships of power. In the interest of enabling focus on diverse relationships of dominance and violence without losing structural perspectives, intersectional approaches (de los Reyes & Mulinari, 2010) on IPV have been recommended and applied in several studies under review (Agevall, 2012;Brännvall, 2016;Gottzén & Korkmaz, 2013;Keskinen, 2011;Sandberg, 2016;Törnqvist, 2017;. In efforts to elucidate individually or socially oriented causative factors, as well as structural ones, intersectional perspectives should also be applied in quantitative IPVAW research. Categorical (McCall, 2005) intersectional approaches can shine light on particularly vulnerable groups, while an anti-categorical (McCall, 2005) stance can place the relevance of socio-demographic groupings in question, (Merlo, 2018;Wemrell, Mulinari, & Merlo, 2017), in relation to IPVAW. The latter is of importance due to the tendencies towards othering of violence described above.