Capricious credibility – legal assessments of voluntariness in Swedish negligent rape judgements

ABSTRACT

A new rape legislation, premised on the requirement of voluntariness, entered into force in Sweden during 2018. This article examines the application and interpretation of voluntariness in Swedish negligent rape judgements with a unique combination of the theoretical perspectives’ sociology of emotion and feminist-legal studies. The material consists of 12 court judgements, from the period 2018–2020, analysed by means of thematic content analysis. The findings show that the judgements in negligent rape cases are based on an evaluation of credibility, supported by evidence that consists of witness statements. These evaluations entail emotions and stereotypes about gender and sexuality, pertaining to rape myths – leading to inconsistent outcomes. Yet, some fundamental assumptions about gender and sexuality are challenged, indicating that the new law may invite discursive shifts and enhanced reflexivity in legal reasoning. The findings are valuable in the work to minimize legal discrimination and secure equality before the law.

KEYWORDS:

• Credibility
• emotion
• legal assessment
• negligent rape
• voluntariness

Introduction

This article delivers the first results on the legal challenges of the new Swedish rape legislation based on voluntariness¹ (popularly consent). Consent-based rape legislations exist in European countries like Germany, Iceland, Belgium, Ireland and the UK (e.g. Temkin et al., 2018), in Australia (Burgin, 2019), and in the US (e.g. MacKinnon, 2016).² Canadian law has ‘the communicative model’ where a person is required to take ‘reasonable steps’ to make sure the partner consents (Gotell, 2010). In Swedish rape law, sexual integrity has been protected by a legislation prohibiting sexual offences associated with force, violence or exploitation (Andersson, 2001, 2004; Andersson & Edgren, 2018). Since the Swedish Criminal Code became effective in 1965, the definition of rape has been discussed regarding the criteria of force/violence and which sexual acts are included (Andersson & Wegerstad, 2016/17; Leijonhufvud, 2008). In the wake of this debate a new rape legislation,³ replacing the requirements of force, violence and exploitation with a requirement of non-voluntariness, was passed in the Parliament by an overwhelming majority (257 votes against 38) and entered into force in July 2018. As detailed below, this new legislation was explicitly designed to transform societal norms around sexual interactions. The aim of this article is to analyse the legal application of the new legislation’s key concepts voluntariness and negligence in relation to stereotypical legal reasoning identified as rape myths.⁴ Thereby, the article contributes to the continued debate on the norm changes envisaged by the law, as well as to previous research by presenting the very first qualitative study of the consent-based Swedish rape law. Understanding how legal professionals apply new legislation and its key concepts are furthermore essential for maintaining the rule of law and the legitimacy of the legal system.

The new legislation aimed to strengthen gender equality, and legal preparatory works expect a decrease of sexual violence in society and a rise of convictions of rape (Prop, 2017/18:177). The Prime Minister Stefan Löfvén announced that the legislation was designed to ‘send a clear signal to society’ that sex should be voluntary (Aftonbladet, 2017). The requirement of voluntariness was introduced to shift the view of the (most often female) victim’s body, from a body that is always available unless a sexual approach is violently resisted, to a body that is never available unless its owner expresses her will to engage in sexual interaction. Following UN recommendations, the new legislation defines voluntariness broadly as an active voluntary participation that is communicated, either through word or deed or in some other way (SCC 6:1). However, a person is never seen as participating voluntarily if: (1) participation is a result of assault, violence or threat, (2) the defendant improperly exploits that a person is in a ‘particularly vulnerable situation’ or (3) induces a person to participate by seriously abusing her/his position of dependence on the perpetrator (SCC 6:1). Additionally, influenced by Norway, the rape law was extended with a new crime; negligent rape (SCC 6:1a). This means the inclusion of acts committed by the perpetrator´s ‘gross negligence’ (grov oaktsamhet). This can be a situation where the perpetrator continues the act even if he realizes that there may be a risk that the other person does not participate voluntarily, but unintentionally fails to find out if this is the case (SCC 6:1a). In addition to establishing that everything except a yes should be legally interpreted as a no (Prop, 2017/18:177), the revision of the Swedish rape law seeks to address problems of victim blaming (Sutorius & Kaldal, 2003), secondary victimization (Törnqvist, 2018), and a high rate of acquittals in rape cases (BRÅ, 2020).

So far, the law has had success in terms of rape convictions, with a 75% increase, but out of 400 judgements from 2019, only 12 were convictions for negligent rape, which seems to go against expectations (BRÅ, 2020). The Swedish Council for crime prevention (BRÅ) suggests that the imprecise requisites of ‘gross negligence’ may explain why the new crime negligent rape is of comparatively little use. They request a clarification of what should count as an expression of voluntariness, and of the distinction between reckless intent (the minimum degree of intent, likgiltighetsuppsåt, which is required for rape) and gross negligence (grov oaktsamhet, which is required for negligent rape).

In this article, we focus on the negligent rape convictions, as exemplary cases where the new rape law is actualized in all its dimensions. Our analysis of voluntariness and negligence takes departure in two points of critique by law and emotion research and feminist legal research. Firstly, the legal system’s conventional idea about rationality and objectivity as disembodied and unemotional (Anderson, 2004; Andersson & Wegerstad, 2016; Burman, 2010; Bergman Blix & Wettergren, 2018, 2016; Lange, 2002; Maroney, 2011). This critique points out that the legal construction of objectivity and rationality makes the law blind to structural perspectives and obscures the norms and values carried by legal professionals themselves (Bergman Blix & Wettergren, 2019; Bladini, 2013). The second point concerns rape laws’ equally problematic understanding of the legal subject as rational and autonomous, always capable of purely rational will-expressions (Andersson & Wegerstad, 2016), which overlooks the complex power relations of situated interactions (Naffine, 2002; Niemi, 2010). MacKinnon (2016) argues that consent-based rape legislations tend to understand consent as an agreement between two equal thinking subjects, unaffected by gender inequality. Common to both points of critique is that the judicial system underemphasizes the role of emotions and power relations.

The article will thus also scrutinize the role of emotions in judges’ reasoning on negligent rape, and how emotion intertwines with stereotypical beliefs pertinent to rape myths. The following questions are answered: How do judges operationalize the concepts of voluntariness and negligence in judgements? In what ways are emotions and gendered stereotypes explicitly or implicitly involved? In the next sections, we present key concepts and previous research in the fields of law and emotions, and of feminist legal research. We thereafter present the method, followed by the analysis in two main sections. In the conclusion, we summarize and discuss our findings.

The role of emotions in legal reasoning

We adopt a theoretical perspective on cognition and emotion as inseparable and mutually dependent (e.g. Barbalet, 1998; Bergman Blix & Wettergren, 2018). Contrasting the assumption that emotion stands for irrationality whereas law represents pure reason, emotions are increasingly proven to be a necessary and integrated part of legal work (Lange, 2002; Maroney, 2011). Exploring the interdependencies of emotionality and rationality and the making of objectivity, Bergman Blix and Wettergren (2016, 2018), Wettergren (2019) employ the notion of background emotions to conceptualize how confidence, certainty, interest, scepticism, doubt or trust, are emotional tools orienting legal professional reasoning. For instance, if a victim behaves in line with a post-rape stereotype, this can resonate with the judge’s background emotions of trust and certainty. Background emotions are often calm and taken for granted, un-noticed as emotions, and hence disassociated from what is commonly seen as ‘being emotional’ (Barbalet, 1998, pp. 59–60). If judges do not reflect on the background emotions orienting their thinking, resonance between stereotypical assumptions and the feeling of certainty remains subconscious while informing decision-making, e.g. when assessing voluntariness, negligence and credibility. Moreover, since rape cases often lack hard evidence judges tend to be left with a situation where the credibility of the victim and defendant are central to their assessments (Anderson, 2004; Finch & Munro, 2005). Psychological studies show that emotions influence the legal evaluation of credibility; emotional testimonies are often perceived as more credible than neutral testimonies (Landström et al., 2019; Wrede, 2015). It may also affect a defendant’s sentence if he fails to show remorse (Bandes, 2016) or if he displays emotion that fit him into the ‘credible criminal’ profile (Nielsen et al., 2018). In the next section, we discuss previous research and theory about the role of stereotypical beliefs about gender and sexuality in legal reasoning.

The role of stereotypical beliefs about gender and sexuality in legal reasoning

Feminist legal research on rape laws often focus on the reproduction of gendered stereotypes and rape myths in legal reasoning (Burman, 2009; Leverick, 2020; MacKinnon, 2016; Smith & Skinner, 2017; Temkin et al., 2018). Rape myths is a concept denoting sexual stereotypical attitudes with a cultural function of a myth, used within feminist-legal research since the 1970s (Bohner et al., 1998, p. 14). Estrich (1987) analysed how some rape cases were considered as more real than others, naming it ‘the real rape stereotype’. If, for instance, there was a prior sexual relationship between the two parties, lack of resistance and absence of evidence, the rape was not considered ‘real’ (Bohner et al., 2009; Estrich, 1987; Smith & Skinner, 2017). By drawing on ‘rape myths’, the court base their judgements on assumptions and stereotypes about the female body, her ability to resist, her being intoxicated, her prior relationship to the defendant (Estrich, 1987; Temkin et al., 2018). Consequently, when a woman is raped, she must typically prove her innocence by convincing the court that she conforms to standards of conventional femininity (Renzetti, 2013, p. 40). Research in the UK and Scotland shows a prevalence of rape myths during rape trials (Smith & Skinner, 2017), e.g. in cross-examination regarding sexual history to contest consent and challenge the credibility of the victim and witnesses (Burman, 2009; Gray, 2015; Temkin et al., 2018). Similar results are found in the US, indicating that applications of consent-based rape law still face problems of rape myths and stereotyped gender roles (MacKinnon, 2016).

The application of rape laws is also researched in a Nordic feminist-legal context (Andersson et al., 2019; Bruvik Heinskou et al., 2020; Jokila & Niemi, 2020). Studies of legal processing of rape cases in Norway show how rape myths are effective to blame and silence victims (Bitsch, 2019) and that legal actors discriminate between the ideal perpetrators and victims on the basis of gender, sexuality, race and crime location (Bitsch & Klemetsen, 2017). Connecting rape myths to emotions, Bitsch (2019) contributes by using sociological theories of shame and sympathy, arguing that emotions are central in the construction of ‘the victim’ and ‘the perpetrator’ in the criminal procedure.

Method

The empirical material consists of judgements under the new rape law 2018–2020 from Swedish District Courts. Out of 391 collected cases on rape, attempted rape and negligent rape, we identified 31 judgements classified as negligent rape, either as first or second claim by the prosecutor.⁵ The first author read all judgements and put together information aboutTable 1 the year of conviction, type of court, gender, sanctions, sentences, number of defendants, verdict, acquittal and if there was an appeal. In order to compare convictions and acquittals, we selected an even number of each from the 31 judgements. We also strived for an even number of male/female judges, and a geographical spread among courts. This resulted in a sample of 12 judgements, enacted from 9 District Courts (see Table 2). All defendants were men and all the victims were women.⁶

Table 1. Analysed judgements

Judgement no.	Court	Outcome
1/A	District Court	Aquittal
2/A	District Court	Aquittal
3/A	District Court	Aquittal
4/G	District Court	Conviction
5/A	District Court	Aquittal
6/G	District Court	Conviction
7/G	District Court	Conviction
8/A	District Court	Aquittal
9/G	District Court	Conviction
10/G	District Court	Conviction
11/A	District Court	Aquittal
12/G	District Court	Conviction

Table 2. Analysis table

Judgement no.	Victim’s emotions		Defendant’s emotions		Rape myths
	Lexical level	Narrative level	Lexical level	Narrative level
1/A		Straight forward, Open, Blame			’Blaming the victim’ – Voluntarily intoxicated’ Belief that rapes only occur between strangers in public places’ – They knew each other, she had relations with other men
2/A	Panic, Fear, Anxiety	Blame	Sadness		’Blaming the victim’ – Voluntarily intoxicated
3/A	Fear Anxiety	Passivity, Blame, Shame	Anger, Fear, Sadness	’Nice guy’ Sympathy	’Blaming the victim’ – No resistance, participation’ Belief that rapes only occur between strangers in public areas’ – They knew each other
4/G	Fear Panic	Passivity
5/A	Fear Shock	Blame	Social Calm Unsuspectingly Sadness Anger	’Nice guy’ Sympathy	’Blaming the victim’- She liked the attention
6/G	Fear Anxiety Panic Shock Sadness	Fear Passivity Sympathy			’Belief that rapes only occur between strangers in public places’ – They knew each other, they have had sex before
7/G		Passivity	Sadness	Regretful ’Nice guy’ Sympathy	Challenge rape myth ‘Blaming the victim’ – Kissing isn’t consent for sex
8/A	Fear Shock Panic Sadness	Fear Antipathy	Chock Anger Calmness Surprised	Fear	’Blaming the victim’ – No resistance’ Doubt on allegation’ – Explaining victim’s reactions by other factors
9/G	Fear Panic Discomfort Shock Numbness		Joy Happiness
10/G	Fear Sadness Shock	Passivity Sympathy	Shame
11/A	Sadness	Fear			’Belief that rapes only occur between strangers in public places’ – They knew each other
12/G		Passivity Panic			Challenge rape myth ’Blaming the victim’. – Voluntarily intoxicated made her vulnerable

The written judgements are composed by the professional judge presiding the trial. They are of different length (14–30 pages), and the display various structure and detail in terms of headlines and the space given to the victim’s and the defendant’s stories. There is no jury in the Swedish judicial system, but in the District Court one professional judge is assisted by three lay judges, attending the proceedings and participating in the deliberations on the verdict and the sentence. All participant judges (lay and professional) have one vote each and in case the voting results in a draw, the most lenient side decides.

We performed a thematic qualitative content analysis, with analytical tools from narrative analysis (Braun & Clarke, 2006; Kleres, 2011). This method allowed us to deconstruct the narrative meanings, capturing the interpretative application of voluntariness in the judgements, related to emotions and gendered/sexual stereotypes. The analysis was abductive and the empirical material (the judgements) and theory enabled us to use theoretical concepts such as emotion and gender, but also to remain open to the data and discover emergent concepts such as inner and external voluntariness and credibility.

During coding, the judgements were read several times. Drawing on rape myths, stereotypes about gender and sexuality in relation to emotional expressions were coded on a lexical level (see Table 2). This enabled us to identify the meaning of the concepts of voluntariness and negligence and instances in data when legal reasoning changes or deviated from dominant norms. Rape myth was used as a broad code as we created categories such as victim being a ‘real victim’ and defendant being a ‘real rapist’ (see Bohner et al., 2009). Further, the narrative structure analysis enabled us to code background emotions, identifying how the legal evaluation of a case was given implicit emotional support in the narrative (see Kleres, 2011, p. 197) (see narrative level in Table 2). This way, we could identify, for instance, how emotions enhanced or lowered the credibility of a victim or a defendant. The table below shows the result of the coding in terms of lexical representations of emotions and expressions, narrated representations of emotions and the related evaluation in relation to the victim and the defendant.

The research has been ethically approved by The Swedish Ethical Review Authority (Dnr. 2020–02205). The judgements are public documents but to prevent identification we anonymized personal information about victims, defendants and witnesses. All excerpts have been translated from Swedish to English.⁷ In presentation of the findings, the excerpts are categorized by numbers, 1, 2 etc., followed by G for guilty and A for acquittal. In the following empirical sections, we will begin by analysing the courts’ reasoning as they apply and thus operationalize the concepts of voluntariness and negligence in the judgements. We thereafter continue with the analysis of how emotions and rape myth stereotypes inform the evaluations.

The legal inconsistency of voluntariness and negligence

The requirement of voluntariness aims at the factual conduct and not at the person’s inner attitude (inställning), while the preparatory works state that passivity can be an expression of consent. This statement means that passivity can be an expression of non-voluntary participation but also an expression of voluntary participation. The evaluation of voluntary participation shall particularly heed whether voluntariness has been expressed in words or deeds or in other way. This last rule must be considered to imply that the space for determining pure passivity as expression of a choice to participate in a sexual act is limited. The evaluation of voluntary participation or not shall be based on the situation as a whole. (Supreme court judgement, B 1200–19)

This quote comes from a Supreme Court clarification of the evaluation of voluntariness. It seemingly adds to the confusion by saying that the evaluation should be guided by ‘factual conduct’ not ‘the person’s inner attitude’ yet the same factual conduct (passivity) can be an expression of both voluntary and involuntary participation. On the other hand, it states that to determine whether passivity as a factual conduct means consent, the evaluation of voluntariness ‘shall be based on the situation as a whole’. Here they leave room for a contextual assessment previously lacking in rape cases (Andersson & Wegerstad, 2016; MacKinnon, 2016), but they also implicitly allow for inferring from the context what is in fact ‘the inner attitude’. It is thus not surprising that we found the distinction between inner and external voluntariness quite central to the evaluation of voluntary participation. Inner voluntariness denotes an inner emotional and cognitive state (the subjective meaning of will, e.g. I want sex because I feel lust⁸) (e.g. MacKinnon, 2016; Smart, 1995). However, since inner voluntariness is perceived by the defendant only if it is expressed by the victim, the court assesses the victim’s voluntariness in practice based on her external expression. This means the court does not only evaluate whether or not the victim actually acted voluntarily (her inner voluntariness), but also if the victim acted in a way that the defendant – and by extension the judge – could interpret as (external) consent-giving. This is because even if the victim did not in fact participate voluntarily, it is no reason to convict someone for rape or negligent rape. The defendant must have carried out the sexual act intentionally, or by ‘gross negligence’, towards the perceived lack of voluntariness (SCC 6:1).

Since the legislation does not define what an expression of voluntariness must entail, it is up to the judge in the particular case to decide (Prop, 2017/18:177). The judgement from case 3/A, cited below, highlights the intricate consequences of this. The judge begins the reasoning citing the Supreme Court clarification (above), and then argues for an interpretation of external voluntariness that implies, but does not explicitly mention, inner voluntariness.

Consent can also be given to bad sex, which means that lack of lust and enjoyment from one part does not mean a lack of consent. Consent does not need to be verbally expressed: rather it can be given in extremely many ways and it does not need to be dressed in words. (3/A)

The difficulties in interpreting voluntariness make the process of assessing evidence dependent on inquiries of the clarity of the victim’s external voluntariness, to assess the defendant’s perception of her will. The judge in the quote draws on the Supreme Court’s reasoning on passivity to argue that the apparent lack of lust is not in itself a sign of involuntary participation, and that there are ‘extremely many ways’ to express consent. This formulation suggests a ‘fussy’ perception of the subject, breaking with the legal perception of the purely rational subject capable of clear will-expressions (Andersson & Wegerstad, 2016; MacKinnon, 2016). However, the judgement retraces the defendant’s and the victim’s relationship to the first time they met, and what happened then and thereafter, arriving at the conclusion that it was reasonable for the accused to believe that the victim participated voluntarily at the moment in question. This is a way to heed ‘the situation as a whole’ in determining if sex without enjoyment is voluntary, but it is also a way to reconstruct the rational subject (the rape victim). Connecting her will expressions at another time with her unclear ditto at the particular time, concluding that she actually knows what she wants. The new rape law in its application thus continues to be based on the idea that a woman, in all situations, has the capacity to express her involuntariness, except in situations of assault, violence, threat, etc (SCC 6:1). In the following, we will elaborate on the dichotomy between inner and external voluntariness.

Negligence – evaluating the defendant’s interpretation of the victim’s external voluntariness

The new criteria of voluntariness make it necessary for the judges, after hearing the victim’s story, to evaluate (1) what the defendant perceived about the victim’s expressed voluntariness, and (2) how this perception influenced his behaviour (Asp et al., 2013). The line between the lower border of intent (rape) and negligence (negligent rape) is difficult for the judges to draw (BRÅ, 2020). The lowest type of intent (likgiltighetsuppsåt) requires that the defendant perceives a risk that the victim is not participating voluntarily but is indifferent to that risk (he would have done it anyhow). Whereas the highest type of negligence (medveten aktsamhet) requires that the defendant perceives a risk of her not participating voluntarily but does not think it is the case. Had he known that she did not want to participate he would have stopped (Asp et al., 2013). The judgements in negligent rape cases thus include an evaluation of the defendant’s degree of negligence in relation to the expressed voluntariness of the victim.

In case 2/A cited below, the victim said she was asleep and woke up by the defendant, whom she first believed was her boyfriend, having sex with her. The defendant admitted to the act but according to him she had consented to letting him sleep next to her, and began moving and sounding in ways that he interpreted as her wanting to have sex. The judgement concludes that the victim’s statement is credible and almost wholly reliable (supported by evidence), but notes that also the defendant’s story is credible. The judgement concludes:

Under all circumstances the victim herself has testified that when the defendant touched her, she may have expressed sounds that according to the court’s understanding could not be reasonably understood in any other way than as an expression of enjoyment and acceptance […] and a confirmation that he should continue doing what he was doing. In addition, the defendant’s statement that he asked the victim if she liked it and she answered in the affirmative is not refuted. Following the presently stated circumstances the defendant cannot be considered as having carried neither intent nor gross negligence in relation to the victim’s involuntary participation. (2/A)

As this excerpt illustrates, the judge’s evaluations of gross negligence become a matter of the defendant’s interpretations of the victim’s external voluntariness, rather than if the intercourse was involuntary in the eyes of the victim (inner voluntariness). This means that even if the victim’s claim that she believed it was her boyfriend is true, it does not change the fact that her sounds could be reasonably understood by the defendant as expressions of voluntariness.

In our material, it appears arbitrary when the judge finds that gross negligence is attained. As suggested in the preparatory work, if the victim is passive, the initiating party, the defendant, has to make sure that the other party is participating voluntarily (Prop, 2017/18:177). Although it is clearly outspoken that the defendant is the responsible party, the practice of proving his gross negligent behaviour appears difficult. Instead, as MacKinnon (2016, p. 452) argues, when evaluation of the meaning of the victim’s expressions of voluntariness becomes important, the focus of the trial shifts from the defendant to the victim (see also Anderson, 2004). Case 3/A, discussed earlier, is a clear case where the victim is passive during the sexual act itself, and the judgement begins by clarifying that this signifies non-voluntariness. However, following the Supreme Court’s rule about evaluating ‘the situation as a whole’ to understand what passivity means, the judgement concludes:

The defendant has, albeit reluctantly, respected the victim when she said ‘no’ the first time they met, and nothing indicates that he would not respect a ‘no’ again. […] The defendant had well founded reason to understand her renewed contact with him as an inquiry about sex. When they met, she implicitly asked about sex and followed him to the bedroom where she undressed and fetched a condom. […] She changed positions and never showed clearly that she no longer consented. Against this background, it cannot be said that the defendant carried any sort of intention in relation to lack of consent, which means that he cannot be sentenced for rape. […] The circumstances previously accounted for do not in any way show that the defendant has been grossly negligent in a relation to a possible lack of consent, which means that, nor can he be sentenced for negligent rape. (3/A)

We see that when contextual circumstances are weighed in, it is found reasonable that the defendant did not perceive of a risk that she did not participate voluntarily, even though she was passive during the sex. For gross negligence to occur, it must be proven that the defendant understood there was a risk but unintentionally failed to heed it. In contrast, in case 6/G, below, the defendant was found guilty on the grounds that he should have made a greater effort to find out what the victim wanted. Again, the judgement refers to the preparatory works and the Supreme Court judgement cited earlier, asserting that in cases where the victim is passive the court needs to take the ‘situation as a whole’ into account. In case 6/G the victim and the defendant had a previous sexual relationship but this time he was invited as a friend to comfort her as she was feeling sad, and she had stated that she did not want sex, as evidenced by text messages. The defendant stayed the night and initially they did have some consensual sexual interaction, but then she withdrew, and they fell asleep. During the night the defendant penetrated her several times. Like the victim, he testified that she was passive each time, but contrary to her he said she enjoyed it. Given the context, however, the court found that …

… it is proven that the defendant realized the risk that the victim did not participate voluntarily. The defendant did not do anything to make sure that she was awake and that she participated voluntarily but carried out the sexual acts in spite of that risk. This means that the defendant was consciously negligent […] It is a matter of gross conscious negligence. […] The defendant shall therefore be convicted for negligent rape. (6/G)

The similarities between cases 3/A and 6/G are that in both cases the victim was passive, and the courts then considered ‘the situation as a whole’ for clarification. In both cases, there is evidence that the victim changed her mind: she seemed to want sex but then seemed passively not to enjoy it (3/A); or did not want sex but wanted some sex but did not want sexual intercourse yet did not say or do anything (6/G). In both cases, the victims had previously been capable of clearly saying no to sex (in case 3/A the first time they met). Evaluating the expressions of external (in)voluntariness the judge decides if the defendant ought to have realized the risk that the victim was not participating voluntarily. In case 3/A, it is found that the defendant did not realize the risk since the victim ‘never showed clearly that she no longer consented’, while in case 6/G it is found that the defendant ‘did not do anything to make sure that she was awake and that she participated voluntarily’. Hence, in case 3/A the defendant’s responsibility is not emphasized by the judge because his actions are taken to be based on the assumption that she would express involuntariness in an active way (she had done that on a previous occasion). In case 6/G, the responsibility of the defendant is instead emphasized, by leaving considerations of any such assumptions aside.

Concluding this first part of the analysis, which has focused on the legal operationalization of voluntariness and negligence, these new concepts necessitate a chain of interpretation illustrated in the figure below (Figure 1). The main questions to be assessed by the court are: (1) Did she participate voluntarily? (2) How was this manifested in words, deeds or other communication? (3) How did the defendant interpret and act upon her expressed (in)voluntariness?⁹ Box 4 contains the entire reasoning required by the court. Noteworthy in the figure are the multiple layers of interpretations, indicating a dynamic, complex and intersubjective process of thinking.

short-legendFigure 1.

We have seen that the concept of voluntariness forces the courts to heed ‘fussy’ (victim) subjects who change their minds several times, contrasting a conventional rational subject, but also that legal reasoning tends to reconstruct a rational subject, by drawing on contextual circumstances. This is particularly clear in the judge’s evaluation of the defendant’s understanding in case 3/A. In contrast, in case 6/G the ‘fussy’ subject is left to be, contributing to the conclusion that the defendant, considering the circumstances, is guilty of gross negligence. In case 2/A, where the victim was active making sounds, the judge discusses her claimed involuntariness, but then leaves it, as it is seen to have no consequences for the evaluation of the defendant’s understanding. Interestingly, given the victim’s ‘expression of voluntariness’, the concluding section in case 2/A does not draw on any contextual circumstances. Our analysis indicates that while the new legislation recommends the inclusion of context, ‘the situation as a whole’ may become significant when the victim’s external expression of voluntariness, from the court’s imagined perspective of the defendant, is deemed unclear. This brings us to the next section, which will investigate the courts’ reasoning around credibility and credible stories, and the role of emotions in these.

Evaluation of evidence and credibility

The lack of technical evidence and independent witnesses is common in rape cases, necessitating focusing on the credibility of the victim, the defendant, and their witnesses (Finch & Munro, 2005; Saunders, 2018). In negligent rape cases, this lack of evidence is even more pronounced (BRÅ, 2020). Following the new legislation, the evaluation of credibility begins with the victim, whose story may be corroborated by other witnesses, typically friends or family, with whom she communicated after the event. Our findings indicate, in line with previous research, that credibility assessments leave space for interpretation influenced by stereotypical reasoning about gender and sexuality. Emotions corroborate such stereotypical reasoning both by separating the plausible from the unlikely emotional expressions, and by making this reasoning feel familiar and comfortable to the judge (De Sousa, 1987).

The credible victim

Women are traditionally encouraged to feel and express emotions of a powerless character, such as fear and sadness, while powerful emotions such as anger are encouraged in men (Shields et al., 2006). In this study, the judges’ assessment of the victim’s inner voluntariness tends to be based on interpretations of her emotional response after the rape. This in turn affects their conclusion regarding her credibility. In case 10/G quoted below, the victim and the defendant met at a party and had sex in the bathroom. They knew each other and the defendant argued that the victim was participating voluntarily. Yet, he was convicted for negligent rape with the following argument:

[The victim] was storming out of the bathroom and cried immediately after the event which strengthens the fact that she is credible and that it is not, as the defendant has argued, a voluntary action on her initiative. (10/G)

The court argues that the victim is credible based on her emotional response after the rape. The expectation that women are supposed to be sad and upset after a rape resonates with gendered stereotypes (e.g. Shields et al., 2006). As suggested by previous research, the court may also sympathize with the victim if she appears traumatized, which seems to work in favour of her credibility (Bitsch, 2019). If the victim’s emotional behaviour is in line with gendered expectations, it makes her claimed involuntariness seem authentic, and therefore the victim is deemed credible.

In contrast, when the female victim deviates from stereotypical behaviours (see Temkin et al., 2018), she risks being portrayed as responsible for the rape and not credible. In case 8/A, the victim and the defendant were colleagues and had a few beers after a working shift. The victim offered the defendant to stay the night, since he was drunk. Instead of sleeping in the guestroom, as agreed, the defendant wanted to sleep in the victim’s bed and a sexual interaction occurred. The victim said she resisted and told him that she did not want sex, but his drunkenness and violence frightened her. The defendant said she was the active party and claimed that he had to stop her because he did not want sex. After the event, the victim texted a friend that she was ‘almost’ raped, then stayed in the house and offered the defendant a glass of water. Her friend, who arrived at the house immediately following the message, testified that the victim was upset, scared and shocked. The victim also had injuries, later examined by a doctor. The defendant was acquitted. Noteworthy is the way that the judgement recounts the victim’s story as a stream of staccato sentences, spiced with frequent cursing’s, suggesting that the victim was not eloquent. The judgement lines up reasons to doubt the victim’s credibility: offering the defendant water was inconsistent; the victim had been free to leave the house, but stayed; it was the victim’s friend who called the police (against the wish of the victim); the medical examination could not determine if the sex was consensual. The judgement concludes that the victim was not more credible than the defendant and her reactions after the event could be explained by other factors than a rape:

The district court cannot ignore the possibility of an alternative explanation to the victim’s reaction and behaviour, which is that she knew her friend was coming over with her boyfriend following the text message and she knew that her friend would demand her to call the police and this demand was conflicting with her own inner wish. (8/A)

In case 8/A, the victim thus displayed some emotions and behaviours in line with the ‘real rape stereotype’ (Bohner et al., 2009), but the court’s reasoning doubts she is a real victim focusing on the deviations. The structure and reasoning of the judgement (reflecting the victim´s lack of eloquence, lining up reasons for doubt) infers background emotions (Barbalet, 2011) orienting trust and empathy depending on the impression made by the victim in court. In other words, the judgement signals lack of sympathy.

The enactments of rape myths and stereotypes about gender and sexuality likely occur in a difficult decision-making situation, in which the judges’ feelings of sympathy or antipathy orient them to take comfort in relying on these stereotypes rather than submit them to reflexive scrutiny (Barbalet, 2011; MacKinnon, 2016, see also Wettergren, 2010).

The credible rapist

Gendered stereotypes drawing on credible emotional expressions, orienting sympathy, are also crucial in relation to the defendant’s credibility, enacting the ‘real rapist stereotype’. In case 7/G cited below, the judgement discusses whether the defendant was intentional or negligent. The defendant claimed he did not know that the victim did not participate voluntarily. According to the judgement, his expression of sadness after the event supported his credibility, and that he would not have proceeded with the act, had he known the victim’s position.

After the incident the defendant became sad when he understood the victim’s attitude. This speaks against that he was indifferent to the act. The court judges his entire way of retelling the incident as if he would have refrained from having sexual intercourse with the victim if he understood that she did not want to have sex with him. (7/G)

The judge argues that sadness pointed towards a negligent, rather than intentional, behaviour. As a typically ‘female emotion’ (see Shields et al., 2006)^, sadness represented by a male seems to influence the evaluation. This corresponds to previous studies demonstrating that emotional testimonies are perceived more authentic and thus credible than neutral ones (Landström et al., 2019), and that particularly crying men gain credibility as victims of assault (Wrede & Ask, 2015). The judge’s evaluation of the defendant´s innocence seems to be based not only on the testified emotional expression after the event, but also on the defendant’s ‘authentic’ expression of sadness in court, evoking sympathy in the judge. This suggests that the fine line between rape and negligence, as marked by the defendant’s expression of sadness, enacts evaluation of the defendant’s capacity to feel remorse. Research on remorse in US courts has shown the defendant’s expression of remorse to bear on the decision about the type and length of punishment (Bandes, 2016).

Returning to case 3/A, which displays several crucial aspects when it comes to interpreting the concepts voluntariness and negligence, it involves a young victim and a 10-year older defendant, who ‘even though several witnesses portrayed him as a very peaceful and kind person, two years ago has committed a serious assault’ (3/A). Despite his previous assault conviction, the defendant is portrayed as a nice guy according to witnesses and also appears to gain sympathy from the judge, who claims that …

The parties have been unusually concordant in most things and no one is less credible or reliable than the other. Consequently, word stands against word and the victim’s account of what happened inside and outside the bedroom is not supported by any other evidence. This means that the evaluation must be based on the defendant’s account. The court thus assumes that the victim to some extent took the initiative […] that the defendant asked her if she really was sure about this, that he did not try to enter her without a condom, that she did not push him away and that he did not pull her hair particularly hard. (3/A)

As discussed previously, the victim’s stated involuntariness in case 3/A is deemed problematic in view of ‘the situation as a whole’. This brings the court to draw on a ‘he said – she said scenario’ (Binder & McNiel, 2007), which in essence gives priority to the defendant’s version of the story. The fact that the defendant confirmed having pulled her hair, that she was bleeding, and that she did not seem to enjoy it, did not lead the court to question why he continued. Instead, admitting to actions that may appear ‘self-incriminating’ is repeatedly deemed to strengthen the credibility in our selection of judgements. The court also choses to believe the defendant’s claim that he did in fact ask if she really was sure about this.

The judge’s reasoning in case 3/A considers context-based circumstances in arriving at the victim’s inner voluntariness (reconstructing the rational subject). We now see that the judge draws on context also by recognizing the defendant´s previous abusive behaviour, but contrasts this with the reasoning that he previously ‘albeit reluctantly’ respected the victim’s no, supported by the witness statements that he is a ‘nice guy’. The defendant is thereby also constructed as a rational and predictable subject. What is entirely overlooked are the contextual power relations possibly inherent in the age gap (10 years) and the defendant’s previously abusive behaviour, and how these may have shaped the victim’s actions during the intercourse where he also displayed violent tendencies. In this regard, case 3/A resembles case 8/A where violent behaviour and injuries were also disregarded as evidence supporting the victim’s version.

The analysis so far suggests that arriving at ‘he said – she said’ scenarios allows space for background sympathy/antipathy to enact comfortable ‘real rape’ stereotypes such as ‘real rape victim’ or a ‘real rapist’ (Estrich, 1987). The typical reasoning begins with scepticism towards the victim’s story, which directs interest to the defendant’s story and the potential support for his version. In case 5/A cited below the victim and the defendant met at a pub and there was some flirting. Later that night, instead of leaving her apartment he had sex with her, according to her when she had fallen asleep, according to him when she was awake. They both testified that she suddenly threw herself out of bed and shouted at him. Friends of the victim witnessed that while at the pub the victim seemed to appreciate the defendant’s attention: ‘He [defendant] perceived that the victim was playing a game with him, which is supported by witnesses saying that she liked the attention’ (5/A). However, as they were in the apartment that night, the witnesses also testified to her distressed (‘real rape victim’) reaction after she ran out of the bedroom. The court arrives again at a ‘he says – she says scenario’ giving way to the defendant’s version of the event, which they deem ‘very detailed’. Similar to case 3/A, witnesses in case 5/A supported a construction of the defendant as having a ‘calm character’ and being a respectable person. Two separate witness statements confirmed this ‘nice guy’ character, for instance, one states: ‘He [the defendant] is very social, calm and trusting and would never harm anyone’ (5/A). The court emphasizes the defendant’s personality traits when evaluating his credibility. Given the defendant’s credible story the court asserts that it is not proven that the victim was asleep. Together with the flirting game that had occurred before the event, this brings the following conclusion:

The defendant’s account that the victim suddenly changed her behaviour towards him cannot be disregarded and thus it cannot be excluded that the victim suddenly changed her mind. The defendant then interrupted what he was doing. (5/A)

In other words, even if the victim displayed a ‘real victim’ behaviour after the event, her behaviour before and during the event made the court doubt her credibility. An age gap (the victim was 15 years older than the ‘nice guy’ defendant) may in this case play a role for the sympathy the judgement displays for him. The age gap is not drawn on directly, but apparent in the judgement’s description of the witness testimonies. It seems to implicitly accentuate how much the victim enjoyed the flirting game. In this respect, the victim deviates from real victim stereotypes that inspire sympathy (see Bitsch, 2019). This reasoning comfortably aligns with the rape myth. It is obviously not in line with the shift, from inquiring the victim’s display of involuntariness to inquiring the defendant’s responsibility for securing voluntariness that the new law encourages. As argued by MacKinnon (2016), this illustrates one of the main problems with rape legislations that are based on external voluntariness; the evaluation boils down to interpretations of the victim’s external signs of voluntariness, not always reflecting her inner voluntariness.

Challenging conventional reasoning within the legal system

In contrast to the analysis so far, and more in line with the intentions of the new rape law, resistance to rape myths were also represented in the judgements, in the sense that the reasoning did not resonate with previous research on rape myths (see e.g. Temkin et al., 2018). Case 7/A is interesting in several respects. It concerns a couple who met on the internet and at the time in question they met irl for the first time, having agreed not to have sex. We discussed this case earlier because the defendant’s display of sadness freed him from intent, which speaks to ‘not a real rapist stereotype’. Instead, he was convicted for negligent rape. In relation to the assessment of the victim’s credibility, however, case 7/A displays rape-myth challenges. Firstly, by explicitly omitting descriptions of the sexual act, detailing what the parties did or did not do. It is stated that the victim was ‘fairly passive’ during the intercourse, and that there may very well have been elements of violence (as according to the victim) but that this is not proven. Secondly, it is stated that the parties separated with a hug and the prospect to meet again, but this does not evoke scepticism against the victim’s story in the judgement. Finally, the circumstance that the victim responded to the defendant’s kisses before the sex is not taken to signify her external voluntariness. Instead, the judge emphasizes the fact that the parties had explicitly agreed not to have sex before they met:

There were many circumstances that gave the defendant reason to realize the risk that the victim did not want to have sexual intercourse with him. He and the victim had agreed not to have that kind of interaction. Already that starting point requires that the defendant should have secured an explicit consent, which not even he himself claims to have done. The circumstance that the victim responded to his kisses could not lead him to conclude that she wanted sexual intercourse. […] The victim remained fairly passive during the course of the event. That should also have made him wonder what the victim really wanted. (7/G)

As we see in this excerpt, the judge emphasizes the defendant’s responsibility and shifts focus to his behaviour. The focus on the explicit agreement not to have sex is in line with the preparatory work (Prop, 2017/18:177) and, based on a differentiation made between a kiss and consent to sexual intercourse, the defendant is not excused to believe that she had changed her mind. A common rape myth in trials is to deny the distinction between consenting to some intimacy and consenting to sex (Gray, 2015; Temkin et al., 2018). Case 7/G thus resembles case 6/G analysed earlier, which also emphasized the defendant’s responsibility, despite the fact that the victim engaged in some intimate interaction. Like case 6/G, case 7/G refrains from evaluating the ‘actual’ rationality of the ‘fussy’ subject.

What sets these two cases apart from the others seems to be the evidence of an initial agreement not to have sex. In the other cases discussed, no such agreement was possible to prove, and the parties’ statements differed on this point, activating the chain of interpretations (Figure 1) regarding external and internal voluntariness. Such complex interpretations, as we have attempted to demonstrate, necessarily involves empathic perspective taking by the judge and thus the possibility of evoking background empathy or antipathy for any of the parties. An agreement seems to allow the court to abstain from such interpretations. Ironically, however, this indicates that little may have changed with regards to the laws’ continued reliance on presumed rational subjects capable of free will expressions.

Conclusion

This article has examined the legal operationalization of the concepts voluntariness and negligence, central to the new Swedish rape legislation, and found that they tie in closely with the evaluation of credibility. Credibility as we have seen is the ‘matter’ on which evaluations actualizing voluntariness and credibility are based. Our analysis shows that the evaluation of voluntariness and negligence differs between judgements in roughly similar cases, and that credibility depends at least in part on background emotions of sympathy/antipathy, doubt, scepticism and trust, which in turn intertwine with gendered stereotypes. In so far as rape myths remain unreflected, they are comfortable roads to reach conclusions when legal decisions must rely on contrasting accounts.

Feminist legal scholars have argued that rape law reforms often fail to achieve any meaningful change and risk being ‘symbolic rather than substantive’ (Burgin, 2019, p. 296). A review by BRÅ (2020) indicates that the new rape legislation enables the legal system to handle a broader range of sexual assaults and the amount of convictions for rape has increased sharply. Yet, in line with BRÅ’s results, our findings indicate that the new criminalization of negligence and the requisite of voluntariness risk (re-)producing a large degree of inconsistency in the legal evaluations of rape, especially for cases concerning negligent rape. Moreover, the new law does not necessarily redirect focus from the victim – this seems to depend on whether responsibility to ensure voluntariness is placed on the defendant or on the victim. Our findings confirm previous research about the reproduction of rape myths in countries with similar consent-based laws (e.g. MacKinnon, 2016; Smith & Skinner, 2017; Temkin et al., 2018), partly due to the range of possible interpretations of voluntariness (see Figure 1), inviting judges to stick to conventional legal reasoning that overlooks contextual and gendered power relations.

Nevertheless, the analysis also revealed efforts to heed the intention of the new legislation, and small signs of norm-changing legal reasoning. Virtually all the judgements in our material account for contextual circumstances to arrive at each case-specific understanding of voluntariness and potential negligence. This manner or reasoning allows for fussy subjects to emerge, which is an interesting deviation from conventional legal assumptions about rational subjects. However, the analysis indicated that context was often used to reconstruct the rational subject, establishing a logical link between behaviours at different points in time preceding and during the event. There is also a huge emphasis on the victim’s behaviour after the event, which is the point where stereotypical reasoning, rape myths, and background emotions are most likely to inform evaluations. We have seen that cases where the judgement seems to go against this are the ones that allow judges to almost wholly refrain from empathic atonement with the parties’ stories, due to the existence of a contract-like agreement (see also MacKinnon, 2016). This suggests that the new legislation is up against a tenacious emotive-cognitive judicial regime (Wettergren, 2019), which does not provide any tools for legal actors to engage in critically reflexive interpretations, including emotional reflexivity (Burkitt, 2012). We propose that more critically reflexive judges could make better, and thus more objective, decisions (Bergman Blix & Wettergren, 2019). Legal actors need to develop their skills and competencies in this regard, before the new rape legislation can have any substantial effect on legal reasoning.

Acknowledgements

We wish to thank the anonymous reviewers of this journal and Aida Mobarhan for their valuable and helpful comments on earlier drafts.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the Riksbankens Jubileumsfond [P19-0515:1].

Notes

1. In public debate, as well as in international criminal research, the concept of consent (samtycke) is used to describe the decisive criteria of rape. The Swedish legislator decided to use the concept of voluntariness (frivillighet) instead, in order to make a distinction from the established legal concept of consent that also functions as a general ground for freedom of responsibility (see SCC 24:7).

2. Historically, the continental countries in Europe have based the rape law on the use of force, whereas the Anglo-Saxon countries has used lack of consent. The Nordic countries have traditionally relied on use of force, but the legal construction has been debated for some time. After 2004 there has been a shift in the rape discussions towards the lack of consent-model (Jokila & Niemi, 2020).

3. Nilsson (2020, p. 116) discusses how the concepts ‘consent’ and ‘negligence’ have been heavily debated and slowly implemented in the legal discourses in Sweden. This discourse field has been developed in news reports and debates about medialized Swedish group-rape cases.

4. This article is part of the research project ‘Rape or Consent – Effects of the new rape legislation on legal reasoning and practice’ (Riksbankens Jubileumsfond, project number P19-0515-1).

5. All judgements under the new law were not available at the time of the data collection, but the 391 cases represent the vast majority. We used the criteria that the defendant was accused of negligent rape (SCC 6:1a) or ‘rape, alternatively negligent rape’ (SCC 6:1).

6. In 96% of the reported rape cases in Sweden during 2018 the victims were women and the suspects were men (BRÅ, 2019).

7. The word ‘victim’ is translated from the Swedish målsägande, literally meaning ‘the owner of the case’. In fact, the state owns the case, i.e. is the legal part of the case, but the Swedish legal tradition allows the victim status to participate throughout the proceedings as the injured part, and to give their account of the event before the turn goes to the defendant.

8. For a detailed discussion about the interpretation of voluntariness and consent outside the legal system in Sweden, see Gunnarsson, 2020.

9. Note that the first question to answer by the court is if the sexual act has been carried out at all. However, this has not been a question in the above-mentioned cases as both parties agreed on that fact.