Vulnerability, Agency and the Ambivalence of Place in Narratives of Rape in Three High-Profile Swedish Cases

ABSTRACT For decades, the media have frequently been instrumental in framing rape cases by linking the deed with the place. This study demonstrates that law courts are not innocent of such social framing; on the contrary, they are significant agents. We argue that courts, by shaping the plot in rape cases, participate in an ongoing cultural production of meaning, although in a more subtle and ambivalent way than the media. In a narrative analysis of three contemporary rape cases in Sweden, we bring together feminist research on place with the concepts of vulnerability and agency. We argue that place is framed as ambivalent in relation to vulnerability and agency, and dependent on the positioning of plaintiff and defendant. In court narratives, geographical places are made relevant, including the locations where the alleged rapes took place. Court narratives of rape include highly ambivalent connotations with place in relation to vulnerability and agency, distinguished by different narratives and outcomes in the various instances. The legal and social implications of our work should include an awareness of the relevance of place in relation to rape.


Introduction
For decades, the media have frequently framed rape cases by linking the deed with the place. Although the courts' participation in framing sexual violence by shaping the plot has been discussed before (cf. Amsterdam & Bruner, 2001;Brooks, 2002;Chancer, 2005;Bumiller, 2008), it is a topic that needs to be explored further. This is especially true in relation to the meaning of place, which has not been studied before.
In this article, we study three hyper-mediatized Swedish rape cases, all of which were explicitly framed in relation to the geographical area where the crime took place. Our aim is to discuss how the meaning of place in the court narratives affects who is rendered vulnerable and who is attributed agency. As pointed out by several scholars, vulnerability and agency are often perceived as diametric opposites in law, which means, for example, that a woman who is recognized as having sexual agency is not seen as vulnerable (Andersson, 2016;Fineman, 2008;Lacey, 1997;Mardorossian, 2002Mardorossian, , 2014Stringer, 2014). We explore whether and how the courts bring these concepts into opposition and how they relate to the meaning of place. Our study is situated in a period of Swedish history during which both the general political landscape and the legal field in relation to rape were changing considerably.

The political landscape
Political (dis)agreements regarding legal regulations, the frequency of rape in various "places", such as suburbs, and the percentage of immigrants in suburban areas constitute a mixture of issues that have involved major social and political processes. Among them is gender equality, a state-sanctioned goal in Sweden during the 1990s, but a disputed and contested one. Populist debaters highlighted increasing immigration from overseas as a danger and potentially ruinous to gender equality (Martinsson, Grifflin, & Giritli Nygren, 2016). The Swedish suburbs, a result of an ambitious national housing programme that ran from 1965 to 1975, are the sites related to two of the mediatized rape cases we study. Since the 1990s, the media have portrayed the suburbs as contrasting with a secularized and gender-equal society in a way that reinforces both privileged and underprivileged positions (Ericsson, Ristilammi, & Molina, 2000;Ristilammi, 1994). The term "gang rape" was frequently linked to the suburbs during the 1990s (Bergenheim, 2005;Bredström, 2002). This was also a period of widening social gaps. Two of our rape cases are related to suburbs and one to a city-centre neighbourhood known as an upper-class hangout, whose exclusive nightclubs are a metonym for increased social gaps.

The legal field
In Swedish law, sexual integrity is protected by legislation prohibiting sexual offences, which, during the period we have studied, are defined as sexual acts associated with force or exploitation (Andersson, 2001(Andersson, , 2004. In contemporary Nordic legislation, force and exploitation are generally the decisive criteria in the provisions concerning rape. This is in contrast to British law, for example, which focuses on the victim's will, or lack of consent (Andersson, 2001;Temkin, 2002). Thus, the elements of force or exploitation set the legal boundaries of these offences. Definitions of rape focus on the perpetrator's force. In drafting the amendments in the early 2000s, the question of whether Swedish legislation should use lack of consent instead of force was discussed at length, but the Commission on Sexual Offences eventually decided to keep force as the decisive criterion. Some ambivalence was apparent, however, as the Commission also chose to place the helplessness of the victim due to such conditions as intoxication, sleep, or illness in the same category as force. The amendment was duly proposed by the Government and came into force in (Proposition 2004/05:45, cf. Proposition 2012. Despite this legal focus on force and exploitation, consent and will dominate in the courts' practice (Andersson, 2004).
In the autumn of 2016, another commission proposed an amendment with a decisive criterion that focuses on lack of consent, or rather non-voluntariness (SOU 2016:60). After this proposal was sent out for consultation, the Government presented a revised proposal to the Council on Legislation, which scrutinizes draft bills that the Government intends to submit to Parliament. The Council rejected the proposal. However, the Government went on with proposing non-voluntariness as a decisive criterion and presented a legislative bill in March 2018, entering into force in July 2018 (Proposition 2017/18:177).
In the legal definition of rape, place is not relevant. At the same time, place is always highly relevant in the criminal legal process, since the prosecution must specify when and where the act took place. This is necessary to satisfy the demands of legal certainty, so that the defendant knows what s/he is accused of (Ekelöf, Edelstam, Heuman, & Pauli, 2016). In this article, we focus on place in another way, which will be discussed in the next section.

Place, vulnerability, and agency
In the wake of ground-breaking books by Henri Lefebvre on place and space (1991), Benedict Anderson on imagined communities (1991), and Nira Yuval Davis on gender and nation (1993), feminist researchers have written extensively on how home, the community, and the nation are all places and spaces that are constituted intersectionally, touching on many different themes (e.g. Johnston & Longhurst, 2009;Johnston & Valentine, 1995;Massey, 1994;Molina, 2007;Pain, 1997;Valentine, 1992). Writings on the meaning of "home", for instance, have touched on a multitude of angles (Mallet, 2004). Such ambivalence is disclosed in the concept of home, which is not only where we dwell, cook, eat, and learn appropriate social roles. It is also invested with meanings such as belonging (Ahmed, 2000;Edgren, 2011), intimacy, fear, and danger (Johnston & Longhurst, 2009;Valentine, 1992), gender and family (Blunt & Warley, 2004), and homelessness (Thörn, 2004;Tyner, 2012). It is useful to bear this in mind when we refer to place: residential suburbs, city, and community/nation. Suburbs are not just a housing programme; public space is not just where we meet and socialize. What interests us are certain meanings relating to belonging and not belonging, being in the right place or out of place (Moran & Skeggs, 2004). The relevance of being in the wrong place in relation to rape is perfectly illustrated by Judith Butler when she quotes an attorney in a rape case: "If you're living with a man, what are you doing running around the streets getting raped?" (Butler, 1992;quote 18). The meaning of belonging and not belonging can help us to unmask narratives of rape, in which different subject positions for plaintiffs and defendants are to be found. We ask in what way the act of rape is linked to place. In line with recent research, we consider it fruitful to pay attention to the ambivalence of place (Moran & Skeggs, 2004;Stringer, 2014;Tyner, 2012) in relation to how vulnerability and agency are expressed in court narratives.
Rape law rests on the assumption of a liberal subject with individual autonomy and agency. In any individual case, this assumption may conceal structural power that leaves the subject vulnerable (see, for example, Grear, 2010;Lacey, 1997;Naffine, 2002;Niemi, 2010;Naffine & Owens, 1997). Taking issue with the liberal subject, Martha Fineman has argued that vulnerability as a human condition should instead be seen as the starting point for the legal subject (Fineman, 2008). This line of thought sees vulnerability as both embodied (in individual factors such as illness) and embedded (in the organization of societal institutions and relationships). Using this approach, we can develop a notion of vulnerability and agency that is connected, among other things, to place. For instance, individual autonomy and trauma may be related to structural factors such as patriarchal notions of female sexuality, and these in turn may manifest differently in connection to different places. In the following, we will discuss vulnerability, agency, and place in concrete cases.

Selection of cases and how we read them
The cases we study in this article are part of a larger study of 25 court cases from 1990-2013, a period that saw several changes in the Swedish law on rape. Rather than going through legal archives, as is common in traditional legal studies (Andersson, 2004;Korling & Zamboni, 2013), we looked for cases through the lens of media archives. Our aim was to find the most high-profile media cases during the relevant period and to study the legal treatment of these cases that were the subject of such intense focus by the media. We want to explore how these cases, to which the media gave so much attention, are treated by the courts. The cases received popular nicknames that in some cases were related to the offender, and in some cases the victim. Six cases stood out as explicitly related to their geographical location. From these, we have selected three to examine in this article. These cases were all known in the media by the names of the places in and around Stockholm where they occurred: in two cases, suburbs generally perceived as underprivileged, and in the third, a city-centre district seen as affluent. The cases from the suburbs reached the Supreme Court of Sweden and are thus precedential. Accordingly, they are significant not only from a media point of view but also in a stricter legal sense. Because the media profiled these cases in relation to place, we wanted to explore whether and how the courts contributed to the narration of place. The Supreme Court cases are also particularly interesting since the court dealt differently with the issue of place in relation to vulnerability and agency in each case, and seemingly without conscious reflection.
Our main material consists of the court judgements in each of the cases. The basic structure of each judgement includes the decision itself along with the grounds or justification for the decision. The grounds for the decision restate the story of each party, and the court evaluates the information provided in these statements in terms of its probative value as part of its assessment of the evidence. In its written judgements, the court mediates the stories of the defendants, the plaintiffs, and the witnesses and other various types of evidence. It is important to stress that the stories proffered in a judgement represent the court's construal of what was said in the courtroom. The court puts forward its own story and line of argument, with the aim of being logical and consistent so that its decision will stand up to potential review by the Court of Appeal. The district court judgement alone may cover as many as 30 pages. Using these stories, the court constructs the background to the case: arranging events, ordering the scene in chronological and spatial sequences, making truth claims. We refer to these as "background narratives". Some judgements also have sections specifically addressing legal issues, such as the classification of the crime, intent, or sentencing. Eventually the court makes its assessment and reaches a verdict, producing what we call an "assessment narrative". If damages are claimed, these issues are also discussed in the judgements.
Our focus here is on the court's narratives as such and the judgements as legal discourses. In reading these, we focus on how knowledge is organized and how truth claims are made in legal discourse. As Joan Wallach Scott argues, knowledge is always grounded in a historical context and always contested. Difference is an instrument of power, in that meaning is produced by contrasting and assigning categories such as male/female, us/them, and so forth; but it is also a methodological tool (Scott, 1992(Scott, , 2002. We use this tool to disclose the connection between individuals and places by exploring how legal issues are presented, contested, and managed.
A precedential case referred to in the media as "the Södertälje case" In this case from 1997 (NJA, 1997, p. 538), four young men were charged with the rape and attempted rape of a young woman. Three of the defendants had names particularly common among people of Middle Eastern descent. The fourth defendant and the plaintiff had names commonly identified as Swedish. In the course of the legal proceedings, the charge was reduced from rape to the less serious crime of sexual exploitation.
The background narrative begins, as usual, by setting out the course of events, arranging the doings and movements of the parties in chronological and spatial sequences. The plaintiff, with a female friend and some other friends, had gone by car to a restaurant in the suburb of Södertälje, where the defendants lived. Then the narrative goes on with an account of the plaintiff 's clothing: she "was wearing a short black dress, an orange seethrough blouse, shoes, bra and panties" (p. 541). Then it is argued that the plaintiff consumed a quantity of alcohol, got into an argument with her friend, and was thrown out of the restaurant. Drunk, but able to walk and talk without difficulty, she was sitting on a bench when she was approached by one of the defendants. He promised her a cigarette and the plaintiff went with him to his car, where three other men were waiting. The men, "talking together in a foreign language" (p. 542), drove the car to a car park, where three of them got out of the car. The fourth ordered the plaintiff to take off her panties, but she resisted, saying "she did not want to and that she wanted to go home" (p. 543). Then one man after another "dragged her", "pressed her down", "parted her legs", and "penetrated her". This continued in a flat in Ronna, a large housing estate on the outskirts of Södertälje, where some of the defendants lived (p. 543). Södertälje is a dormitory city south of Stockholm, and well known as a community where many new immigrants to Sweden settle. Ronna, also widely known for having many immigrant residents, sets the initial scene. Drawing on the background narrative, the assessment narrative makes truth claims as a basis for the verdict. The design of the narrative concerning the plaintiff can be understood as establishing links between morality and agency. As Peter Brooks argues, a rape narrative in the courts is also very much a cultural narrative on how a "woman is supposed to behave" (Brooks, 2002, p. 4). An example is the description of the plaintiff 's clothing, the implication being that she runs the risk of being perceived as unrespectable while out and about late at night dressed in this way. She can also be interpreted as not being an "ideal victim" (see Christie, 1986). The narratives in the judgement are dependent on the letter of the law but also need to be comprehensible to appeal courts and to audiences in a certain social context (Kohler-Riessman, 2008;Mayne, Pierce, & Laslett, 2008). The court's ability to reason and conclude by telling and recasting is dependent on rhetorical resources, as argued by Brooks (2002). Although the background narrative does not further emphasize the plaintiff 's clothing, the initial remarks still legitimize a perception of certain female behaviour as blameworthy. The same thing was emphasized by the plaintiff herself, who expressed remorse for her shameful clothing, saying that she thought what happened was "her own fault since she had been drunk and wearing a short dress" (NJA, 1997, p. 545) and out late at night at a restaurant. We understand that she interpreted this place as a dubious locality for a young girl, as was the meeting with the men who, as she put it, talked "together in a foreign language" (p. 542).
The assessment narrative seems clear on the extensive use of force, and just as clear about the plaintiff 's lack of resistance due to intoxication. Thus, it recognizes the plaintiff as vulnerable. This logic-to ignore the appearance of the plaintiff, which she herself interpreted in accordance with traditional ideas about female respectability-required another criterion: that of being in a helpless state (p. 553). In this rhetoric, rape was replaced by sexual exploitation, since it was argued that violence was not used. By introducing a new prerequisite, the plaintiff was rendered both vulnerable and bereft of agency, and as such not assigned responsibility for her clothing or being intoxicated in public. We argue that this rendering, in turn, is made possible by the image of the plaintiff as being out of place: in a place unfamiliar to her and associated with the alien background of the defendants. Multiple comments by the court explicitly situate the men within a specific cultural geography: "born in Syria but raised in Sweden together with their immigrant parents", "has had difficulty adapting to the labour market", "personality disorder", "born and raised in Sweden as the third of eight siblings in an Assyrian family where the parents originate from Turkey" (p. 555). By detailing their immigrant backgrounds, including emphasizing sibling number (which has no legal relevance and therefore can be interpreted as an allusion to a family planning foreign to Swedish culture) these men are positioned outside of Swedish culture. It has been customary to point out non-Swedishness in rape cases when the perpetrator was foreign, an immigrant or a tourist (Bergenheim, 2005, p. 386). In the present case, by drawing on the historical and cultural situatedness of the defendants-namely, their Middle Eastern background-the court narrative constructs these men as culturally alien, excluded from the image of the Swedish community. In this sense, the narrative invokes a migration context, enacting a normative understanding of culture and providing the media with information that could link the crime to that particular suburb. Hence, it could be argued that the court participated in an ongoing cultural production of place. The acknowledgement by the lower courts of the plaintiff 's vulnerability, although out of place, in a place unfamiliar to her, drunk and lightly dressed, should be understood in the context of this construction of the defendants as alien to the Swedish community.
The Court of Appeal rejected the argument that the plaintiff was in a helpless state, arguing that "she must have had several occasions to leave the men" (NJA, 1997, p. 538). The charge of rape was not tried owing to procedural circumstances. Once again, vulnerability was framed in opposition to agency, but this time the logic worked the other way around. The plaintiff was assumed to have agency; thus, she was not recognized as vulnerable: she could have left, the court argued. The significance of the fact that the parties belonged to different places, both symbolically and geographically, emphasized by the lower court, was ignored by the Court of Appeal. In sum, the ambivalence of place in relation to vulnerability and agency is demonstrated.
The acquittal by the Court of Appeal triggered a media debate (e.g. Göteborgsposten, 27 May 1997) and a parliamentary debate on the criminal justice system (e.g. Motion 1997/98:Ju703). The Supreme Court ultimately ruled that, as a matter of principle, a helpless state could arise from a combination of circumstances. In this case, it was determined by the fact that the young woman was drunk, and while not drunk enough to be helpless, she was taken "by car at night, along with four unknown men /…/ to the relatively isolated place" and could not escape (NJA, 1997, p. 565). This assessment narrative demonstrates how place connotations become important for recognizing vulnerability. In this case, a context was created that demonstrated the plaintiff 's vulnerability as a (Swedish) woman. The meaning of "helpless state" was extended; it did not have to arise from a single factor, e.g. drunkenness, but could be based on several factors, here drunkenness in combination with being in an unfamiliar place inhabited by unknown men.
A precedential case referred to in the media as "the Tumba case" In the second case, from 2004 (NJA, 2004, p. 231), the prosecution brought charges of aggravated sexual assault, specifically the sexual exploitation of a person in a helpless state due to intoxication. Four men aged between 20 and 23 were accused of exploiting a woman, 10 to 15 years older. The court's background narrative dwells on the relations among the parties involved. Place plays an important role in how the court frames the plot, in that the plaintiff 's workplace and residence both figure in the narrative. The plaintiff worked at a petrol station in the Stockholm suburb of Tumba and lived in a flat in the same suburb. As in the Södertälje case, geographical areas mediatized with immigrant associations set the scene. Including the plaintiff 's workplace in the narrative also reveals her working-class background. Further, the plaintiff knew two of the defendants through work and knew of the other two, and the narrative mentions that she had previously met one of them at a restaurant and had "a nice time with him" (p. 233). On the night of the alleged assault, the plaintiff and three of the defendants met at a restaurant. The plaintiff danced with two of the men and kissed one of them. Later, she was offered a lift home by one of them, but all three came along. After they arrived at her flat, another man turned up. The restaurant figures as a significant place that could be interpreted as something that calls into question the plaintiff 's respectability and her vulnerability: it was where "she talked to [two of the defendants] and was offered beer and danced with [one of the defendants] and kissed him" (p. 233). The background narrative further mediates the woman's account of feeling weak, disoriented, and strange but not drunk: "she felt she was being moved around" (p. 234). In contrast, in the men's version of the story, the woman was "a bit drunk but far from passed out" (p. 234), and in fact she was "the active one in the various sexual encounters" and "the one who 'guided the whole event'" (p. 234), thereby constructing the narrative in relation to place, her residence, for the reasoning below.
The assessment narrative of the District Court refers to the given evidence, including the plaintiff 's story, and concludes that her helpless state has not been proven beyond reasonable doubt. The court also gives its reasoning about her behaviour, which was "sexual, to say the least" (p. 235). The assessment narrative of the Court of Appeal similarly framed the woman as older and sexually experienced: "she was almost 15 years older" and "[the defendants] knew she had had a so-called gang bang with two of their friends on an earlier occasion" (p. 241). The higher court notes, to begin, that a certificate from the National Board of Forensic Medicine shows the plaintiff to have been extremely intoxicated. Based on this information, together with the plaintiff 's statement and the fact that she "understandably enough experienced the situation with several men, some of them unknown to her, in her own flat, as threatening" (p. 240), the Court of Appeal concludes that the plaintiff had indeed been in a helpless state. Then, and decisively for the outcome of the case, the plaintiff 's sexual history becomes the focal point regarding the intent of the defendants in relation to her helpless state (p. 241). The only time the defendants in this case are specifically portrayed as immigrants is when the plaintiff is quoted as referring to their Middle Eastern background in connection with their sexual arousal-she told her friend on the phone that she had brought "three horny Syrians" back to the apartment (p. 241). The inclusion of this detail in the narrative, along with the fact that she "voluntarily invited the three men to her apartment" (p. 241), clearly frames the plaintiff as an agent, however unrespectable, and also suggests that she was to blame for making her flat a threatening place. Next, the court engages in unusually detailed reasoning about intent. The narrative stresses that the plaintiff had in the past had sex with multiple men at once, and that she was flirtatious and active on this particular night. The implication is that it is highly understandable if the defendants, having been invited to her flat, believed that the plaintiff had consented to intercourse, given her prior experience of group sex. This line of reasoning also makes the plaintiff complicit in turning her residence into an environment in which she was vulnerable. Such reasoning clearly rests on a hotchpotch of stereotyped notions about sexuality, femininity, masculinity, ethnicity, and normality. Situating this reasoning in relation to place makes it implicitly a logical consequence of the Södertälje case. The Södertälje plaintiff was blamed for being in the wrong place late at night, but ultimately depicted as a vulnerable character in the narrative. The narrative in Tumba was rather different and, as such, a perfect illustration of the complexities of the role of agency in general and sexual agency in particular-both of which the court addressed in relation to the plaintiff. The plaintiff 's prior sexual experiences positioned her as an agent, autonomous and self-sufficient, while hindering any recognition of her vulnerability (cf. Andersson, 2016). Neither the violation of her individual integrity nor the structural aspects of that violation (in terms of the patriarchal norms surrounding women's sexuality) were recognized. Her autonomy in bringing men to her apartment made her vulnerability questionable (cf. Fineman, 2008;Mardorossian, 2014).
When the Supreme Court heard the case, its narrative did not approach the questions of helpless state or intent and focused instead on lack of proof, and the defendants were all acquitted. Legal scholars have seen this case as a confirmation of case law on the interpretation of the "helpless state" and as addressing evidentiary issues, while also clarifying the meaning of helplessness in relation to alcohol (Zeteo, 2013, ch. 6, p. 1).

A case from the Court of Appeal, referred to in the media as "the Stureplan case"
In our last case, from 2007 (CASE 3806-07), a charge of aggravated rape was tried in the District Court. The background narrative sets the scene by presenting names commonly identified as Swedish and positioning the plaintiff and the two defendants as previous sexual partners. The events leading up to the alleged assault are laid out in the ordinary way by presenting the various parties' movements and meetings. On the evening in question, the parties had met at the defendants' downtown restaurant. Later they all went to the flat of one of the defendants at a nearby address on Grev Turegatan in Stockholm. In Sweden, it is common knowledge that this address is located in the affluent district called Stureplan, which itself is frequently used as a metonym for the upper class. The parties offered differing accounts of what happened next. The plaintiff admitted having consensual sexual intercourse with both men until, as the court narrative puts it, they started "having anal sex" with her p. 8). At that point she began saying no, crying and screaming, while the defendants told her to "shut up" and called her a "whore" (p. 8). One of the men forced a remote control into her vagina and anus. The defendants denied any coercion. A mediated narrative by a witness, a female police officer, confirmed the plaintiff 's upset condition. Supporting evidence in the form of a psychologist's report was also submitted and evaluated by the court. The police officer stated that the plaintiff had been crying hard and was shaking all over when she approached her at 06.50 on Sunday morning. The plaintiff had "difficulty saying what happened", and claimed she was "afraid of not being believed" (p. 12).
As the literature points out, a woman going home with a man for sex after visiting a bar is a classic situation that signals lack of innocence (see e.g. Mardorossian, 2014). This may explain why the plaintiff in this case felt that she had to add that she was afraid of not being believed. By agreeing to sex, she had consciously exposed herself to risk and could not expect to be recognized as vulnerable according to the logic of the court narrative, which could not accommodate both vulnerability and agency. A way out of the dilemma was offered by a repertoire that confirmed mental trauma. McKenzie-Mohr argues that blaming the victim, a dominant narrative in rape cases since the 1980s in the West, has faced competition in this century from the trauma narrative (2014). A trauma narrative can eliminate guilt and shame, but at the cost of being recognized as an emotionally damaged victim. As Carine Mardorossian argues, this either/or perspective makes it difficult for "people to recognize that one may have sexual agency, even be extremely 'promiscuous'" and at the same time be vulnerable (2014, p. 35). As we will show, in the Stureplan case, the first narrative (blaming the victim) was used in the District Court and the second (trauma) in the Court of Appeal. Both play a significant role in relation to place and vulnerability.
The assessment narrative of the District Court concluded that the plaintiff 's severe intoxication reduced her credibility, and that the question was whether voluntary and "purely sexual activity" had tipped over into rape p. 16). A defence of violence as normality is visible in the narrative. The court further states that the plaintiff might "have felt raped" after the fact (p. 19); she might also "have been in love", and her "social standing was dependent on the defendants' social circle". The implication is that the plaintiff was responsible for breaking social barriers, and as such had forfeited the right to a vulnerable position. The social status of the defendants as upper-class Stockholm urbanites lent credibility to this suggestion. The historical and cultural situatedness of the place involved made it suitable for the men while denying the woman access to it, except as mediated by her male company. "Through sexual relations with the defendants she had been able to enter those circles, " the narrative continues (p. 18). In other words, her own social position did not give her access to the places and culture of affluent central Stockholm. The court's rejection of the plaintiff 's testimony led to an acquittal. The prosecution never raised the question of whether the plaintiff had been in a helpless state, even though she admitted to having been extremely drunk. In our view, the sexual willingness of the plaintiff served to underline her agency and responsibility, while vulnerability was denied to her, seemingly due to her transgression of class and place barriers.
The Court of Appeal assessed the evidence differently, especially the fact that the plaintiff 's mental reactions increased her credibility. The plaintiff 's own testimony was borne out by observations made by the responding police officer, who said the plaintiff "was upset, crying hard … and was shaking badly" (CASE 3806-07, p. 8). The court took this performance of trauma into consideration and sentenced the defendants to four years' imprisonment. In contrast to the Södertälje case, in the Stureplan case, place did not matter in claiming vulnerability. By emphasizing mental, rather than geographic, vulnerability, guilt was written off. The Stureplan case demonstrates a problematic opposition between vulnerability and agency. Invoking mental vulnerability produces a weak female position since it suggests that the problem lies within the individual and ignores patriarchal structures (see Edgren, 2016;Mardorossian, 2014;Plummer, 1995;Stringer, 2014).

Conclusions
Based on three Swedish hyper-mediatized rape cases that became explicitly linked to the geographical areas where the crimes took place, this article has discussed how and to what extent court narratives may contribute to the cultural meaning of place. We have invoked the concepts of vulnerability and agency and analysed how the meaning of place may affect who is rendered vulnerable and who is attributed agency. Our study shows that geographical place is made relevant in the courts' narration of rape, paving the way for its use in the labelling of rape cases by the media. We have demonstrated how the courts play a significant role in framing criminal acts in relation to place, in a subtle and ambivalent way. The court narratives we have studied include ambivalent place associations that are revealed in different narratives and outcomes in different judicial instances: associations of belonging and not belonging, being in the right place or out of place, all with implications for who is rendered vulnerable and who is recognized as having agency. For example, in one case, where the crime occurred in an immigrant-dense suburb, the lower court recognized the plaintiff as vulnerable and positioned the defendants as foreign, while the higher court recognized the plaintiff 's agency. In both courts, place played a significant role in the sentencing. In another case, in which the crime took place in an affluent city-centre district, the lower court attributed agency to the plaintiff and positioned her as sexually active, while the defendants were positioned as members of the upper class. The higher court recognized the same plaintiff as vulnerable, relying on a trauma narrative. In line with the results of previous research, vulnerability here-which is always set in opposition to agency-is limited to the individual level, especially when a trauma narrative is invoked. By linking our analysis to feminist research on the concept of place, our results demonstrate a new field where it could be used. To sum up, we have discussed how court narratives of rape include ambivalent place associations as they relate to vulnerability and agency, which are revealed in different narratives and outcomes in different courts. In terms of their legal and social implications, these results should increase our awareness of the relevance of place in relation to rape. and power. She is interested in the issues of sexuality and gender, and particularly power relations in regard to class, age, and ethnicity. Her research spans sexual offences-highlighting the gendered structure of legal definitions as well as the proof process-human trafficking regulations, and juvenile crime and delinquency in connection with gang activity. Andersson is currently working with Edgren on the "Rape in Sweden" project.