Justice Denied? Prosecutors and presiding officers’ reliance on evidence of previous sexual history in South African rape trials South African

This article presents data from a study conducted by the Medical Research Council of South Africa, focusing on rape attrition in South Africa at different stages in the processes (from reporting at a police station to potential conviction). The study found that of the 3 952 reported cases of rape analysed 65% were referred to prosecution, and trials commenced in 18,5% of cases. Of the total 3 952 cases reported, 8,6% resulted in a guilty verdict. Using qualitative data from a subset of trial transcripts, the article focuses specifically on the problematic views of both presiding officers and prosecutors based on rape myths and gender-stereotyping at trial


Justice Denied?
Prosecutors and presiding officers' reliance on evidence of previous sexual history in South African rape trials

South African
This article presents data from a study conducted by the Medical Research Council of South Africa, focusing on rape attrition in South Africa at different stages in the processes (from reporting at a police station to potential conviction). The study found that of the 3 952 reported cases of rape analysed 65% were referred to prosecution, and trials commenced in 18,5% of cases. Of the total 3 952 cases reported, 8,6% resulted in a guilty verdict. Using qualitative data from a subset of trial transcripts, the article focuses specifically on the problematic views of both presiding officers and prosecutors based on rape myths and gender-stereotyping at trial, and suggests that these are a factor affecting the attrition rate between cases referred to trial and those that result in a not guilty verdict.

No. 69 | 2020
Rape is not a South African invention. Nor is it distasteful sex. It is sexualised violence, a global phenomenon that exists across vast periods of human history. Rape has survived as long as it has because it works to keep patriarchy intact. The numbers presented by SAPS are not a true reflection of the number of sexual offences that actually occur in the country each year. One reason for this is that the police only record cases under the most serious crime category. For example, where sexual offences result in femicide or attempted femicide, these crimes are simply recorded as murder or attempted murder. The sexual violence that coincided with these incidents therefore remains unrecorded. 7 Furthermore, femicide and attempted femicide are not separate codified offences in South African law. These shortcomings in the recording of crime impacts the value of SAPS statistics as an accurate measure of the levels of violence against women in our country.
A second problem is that sexual violence is generally under-reported in South Africa, as it is in many other countries globally. 8 SAPS's reported statistics therefore severely underrepresent the size of the problem. A 2010 study by GenderLinks found, for example, that only 1 in 25 women raped in Gauteng reported their rape to the police. 9 We can conclude that sexual offences are pervasive in South Africa, in particular rape, with very few cases reported to the criminal justice system.
Victims of sexual violence may not come forward for many reasons. They may want to avoid the potential trauma associated with the reporting process with the police. Survivors may not be aware that what they have experienced is, in fact, a crime. For example, where the victim of the rape is a child or a person with severe psycho-social disability they may not understand or comprehend the illegal nature of the act. For child victims, the perpetrators are most commonly family members or close acquaintances, and this also can present barriers to reporting, such as coercion by the perpetrator or family members to keep the incident quiet, or the fact that the child victim is not believed by the people they disclose the crime to. 10 Women may similarly choose not to report based on coercion, often by an abusive partner. 11 This not only prevents them from reporting the rape to the police, but also prevents them from speaking out to family members and other support structures.
Problematic views and stereotypes around rape and victims of rape that permeate the criminal justice system can also frustrate reporting, investigation and prosecution of these crimes. Smythe notes that attitudes and practices by police can lead to secondary victimisation, for example not permitting women to lay charges of rape, reformulating rape as lesser offences and not providing women with private areas to recount the offence. Smythe argues that these behaviours are informed largely by sexist stereotypes, and result in minimising rape complaints and alienated complainants. 12 The reliance on stereotypes around rape and victims of rape is not isolated to the criminal justice system, but permeates society. Bonnes analysed articles about rape cases in a print newspaper and found that the publication used rape myths to blame victims for their rapes and move away from holding perpetrators accountable. 13 For example, the publication often mentioned that rape victims had consumed alcohol, placing the blame for the rape on the women for acting in an 'irresponsible' way and putting herself in a compromised position. 14 Stereotyping and victim-blaming also exists in the healthcare system. For example, a study in the United Kingdom found that nurses and trainee nurses held similar common mistaken beliefs around victims of rape as the general public. This included blaming women for their rape if they knew their assailant or placing blame on women if they were seen to be 'careless' at the time of the rape. 15 The attrition of rape cases from reporting to conviction and sentencing is partly influenced by rape stereotyping and the problematic practices that result from these attitudes. Cases may fail to move from investigation to prosecution due to the intersection of stereotypical attitudes about complainants and bureaucratic interests. 16 Smythe illustrates, for example, that police investigation into one particular victim's case appeared non-existent by the time the case was closed, in part due to the police officer's views on the complainant. He had written in his investigation diary that the suspects were apparently drunk and the crime was likely to be one of lust. 17 A study of rape case attrition in Gauteng found that very few cases (only 8%), proceeded past investigation to result in conviction. 18 The reasons cited for the failure of referral of matters for prosecution included that police did not believe the complainant, no formal complaint was made by the victim or the victim dropped charges, prosecutorial concerns of the police and apprehension of the victim around the prosecution process. 19 A study conducted by the Medical Research Council of South Africa (the MRC), that focused on rape attrition at different stages in the processes (from reporting at a police station to potential conviction), found that of the 3 952 reported cases of rape they analysed, 65% were referred to prosecution, but only 18,5% went to trial. 20 Only 8,6% of the cases resulted in a guilty verdict. 21 The high levels of attrition between cases referred to trial and those that result in a guilty verdict requires attention. This paper focuses on this attrition and suggests that problematic views of both presiding officers and prosecutors, based on rape myths and gender stereotyping, is a factor affecting successful prosecution, resulting in conviction. 'Rape myths' refers to 'prejudicial, stereotyped, or false beliefs about rape, rape victims and rapists [which creates] a climate hostile to rape victims'. 22 'Gender stereotyping' refers to prescriptions and proscriptions that are intensified by virtue of one's gender, and those that are relaxed by virtue of one's gender'. 23 To analyse these problematic attitudes, I will begin by exploring instances where prosecutors and presiding officers relied on rape myths and gender stereotyping, taken from transcripts of rape cases used in the above MRC study. I argue that problematic gendered views of presiding officers and/or prosecutors in these kinds of matters can influence the outcome of rape cases and show how the use of previous sexual history evidence can characterise the complainant as unreliable and/or jeopardise securing a conviction of the accused. 24 If South Africa seeks to secure convictions in sexual offence cases, deter other individuals from committing these offences and promote rights victims' rights, it must eradicate the reliance on rape myths and gender stereotyping in the criminal justice process. 25 I suggest that this reform should include continuous education for presiding officers and prosecutors around gender stereotyping and the problems arising from relying on rape myths. There may be a need to bring practitioners and presiding officers up to date on rules of evidence (especially around previous sexual history and the cautionary approach). Reform should also consider legal representation for complainants so that their rights are protected during the trial.

Methods
This study draws on a select set of cases from the 2017 MRC study on the investigation, prosecution and adjudication of reported rape cases in South Africa. The MRC study examined transcripts from a nationally representative sample of 3 952 cases of rape and attempted rape that had been reported to SAPS across South Africa in 2012. Cases were selected through a multistage sampling approach that involved first randomly selecting 170 police stations from 1 164 police stations in the nine provinces, using probability proportionate to size. Then, a maximum of 30 rape cases were systematically selected from the list of all reported cases at each of these police stations.
To qualify for inclusion, the trials had to have been concluded at the time of the research. This yielded a sample of 102 cases that had already resulted in a verdict. Transcripts of the entire proceedings from these trials were produced by a transcription company contracted by the South African government for this purpose.
The selected cases were then analysed using a structured tool that focused on trial procedures, including charging, pleading, application of special measures, examination and cross-examination of complainants and witnesses, accused's examination-in-chief, cross-examination of the accused, dismissals or discharging matters, closing arguments, judgment and sentencing. The tool was also used to abstract data on the application of the law of evidence, gender biases and the reliance on gender stereotypes by parties or the courts.
This article uses qualitative data in the form of direct quotes that was extracted from the transcripts of this subset of 102 cases. The article therefore explicitly focuses on the proceedings during trial, rather than an analysis of the judgments. Its analysis focuses on the themes of gender stereotyping and rape myths and how the use of these may have an effect on the ultimate outcome of the rape or attempted rape case.

Previous sexual history
South African courts (as elsewhere around the world) have historically relied on evidence gathered through the examination of a complainant around her previous sexual history. 'Previous sexual history evidence' refers to an exception to the general rule that prohibits evidence that is solely directed at establishing that the complainant is of bad character. 26 This kind of evidence therefore permits the defendant, in cases of rape or previously indecent assault, to 'adduce evidence as to the complainant's bad reputation for lack of chastity', which has been used to show that there was the mistaken belief by the accused that consent was present and/or that complainant lacks credibility and is unreliable as a witness. 27 Schwikkard notes that the principle was adopted from English law in the nineteenth century and is based on the notion that no decent woman would engage in sex before marriage. 28 Other than the problematic gender essentialism of this principle, giving evidence of previous sexual encounters can be traumatising, embarrassing and dehumanising for complainants. 29 Furthermore, having to go through this type of questioning may in fact deter individuals from coming forward to report cases in the future. 30 In South Africa, prior to decision in S v M and the subsequent enactment of the Criminal Law (Sexual Offences and Related Matters) Amendment Act ('SORMA'), 31 previous sexual history evidence was prohibited, unless a court allowed it. The only threshold for determining the admissibility of such evidence was that the court had to consider whether or not it was relevant. 32 Schwikkard points out that the consequences of this wide discretion was that the courts were in fact loathe to restrict evidence around previous sexual history. 33 In the case of S v M, before the Supreme Court of Appeal (SCA), and in reference to the Law Reform Commission paper on sexual offences, 34 the court stated: In [the Law Reform Commission's] evaluation the researchers conclude (at 501) that s 227 has to some extent failed of its purpose and that '[t]he unfettered discretion given to presiding officers to determine the admissibility of such evidence on the broad and subjective basis of relevance seems to be a large part of the problem'. Accordingly, they propose that s 227 be amended 'to clearly delineate the circumstances under which evidence of previous sexual history may be adduced'. 35 After the enactment of SORMA in 2007, which amended sections of the Criminal Procedure Act (CPA), previous sexual history evidence was prohibited in South African law, where the credibility of the complainant is being questioned. Previous sexual history evidence is also not allowed where the goal of such questioning is to show that the complainant was more likely to consent to the rape due to her previous sexual history or that she is less worthy of belief. 36 Questioning on previous sexual history is still permitted under certain circumstances, which includes cases where one of the parties makes an application to adduce such evidence and this is allowed by the presiding officer, or where such evidence has been introduced by the prosecution. 37 Despite this tightening of circumstances under which questioning on previous sexual history in rape trials are allowed, the data for this study shows that the defence continues to present this kind of evidence, without making an application for permission to do so. S v Khumalo, 38 S v Kango 39 and S v Maboe 40 provide examples of cases where reference to previous sexual history was made by the defence without application to do so, where the prosecutor did not object, and where the presiding officer did not insist on an application first being made by the defence, as required by law. 41 Thus, despite law reforms, not much has changed since S v M.
Although the defence's reliance on previous sexual history without application to do so is problematic, it is much more concerning to see the emergence of questions around, and reliance upon, previous sexual history evidence by prosecutors and presiding officers. A prosecutor's predominant aim is to secure convictions in their cases and it is nonsensical for them to rely on previous sexual history and open the gate for the defence to adduce such evidence too. 42 In S v Khumalo, S v Ngobeni 43 and S v Mqongwane, 44 prosecutors introduced lines of questioning around the complainants' previous sexual history. This kind of questioning is particularly problematic as prosecutors stand to gain little by adducing such evidence, and instead may jeopardise the state's case and violate the complaint's rights to dignity and privacy.
The relevance of the questioning around the complainant's previous sexual history was not clear in any of the cases in our sample. Although relevance is a requirement in terms of s 227(2)(b) of the CPA, this only applies where the defence attempts to adduce such evidence, rather than the prosecution. 45 Prosecutors should therefore always consider whether such evidence is in fact relevant before proceeding to introduce it. Yet, in S v Khumalo there is evidence that prosecutors led evidence of previous sexual history during trial. In one example the uncle of a 13-year-old complainant was charged with her rape. The prosecutor asked the child '[n]ow except for your uncle that put his private part in your private part did anybody else do the same thing to you?' 46 The reason for the question is not entirely clear from the transcript, although it appears that the prosecutor intended to use the testimony in order to rule out the potential rebuttal by the defence that the child had been raped by another individual and then blamed the accused. 47 In S v Khumalo, 48 the minor child was asked, by the prosecutor, if she was in a 'love relationship' with the accused, to which the minor answered in the negative. This is an example of a question that alludes to the minor's previous sexual history, but where prosecutors are pre-emptively attempting to dispel the common perception that complainants (whether women or children) falsely lay charges of rape as an act of revenge against the accused, with whom they are in a relationship.
The notion that complainants will 'blame' men or seek revenge by asserting that they were raped was also evident in S v Mqongwane, where the prosecutor asks the complainant about her previous sexual history in the following exchange: The issue with this type of questioning is that it can induce unnecessary trauma for the child who, if previously violated, would then need to delve into those past experiences of abuse. Furthermore, this type of questioning is problematic as it is steeped in gender stereotyping that depicts female victims as vindictive and untruthful.
Where prosecutors introduce previous sexual history evidence, presiding officers arguably have to take this into account. Yet, the transcripts show that presiding officers make reference to complainants' prior sexual history based on their own inferences from the evidence before them, and on arguably tainted views that are rooted in a discourse of rape myths and gender stereotyping. In such cases, presiding officers may deliver biased judgments that are misguided and reliant on myth and not fact. For example, the presiding officer in the same case stated the following when delivering his judgment: [n]ow in a lot of instances stepfathers are blamed for the abuse of their stepchildren. In a lot of cases this turns out to be true. But there are also these cases which turn out to be false allegations… Its animosity towards a stepfather is not an uncommon thing. In this case we had a 7-year-old child taken away from the mother, seeing her raise three kids with the stepfather, and she is not allowed to be a part of that … she can only visit. 50

Gender stereotyping
The reliance on previous sexual history in rape cases is a historical and universal issue which persists despite continued criticism. McGlynn, for example, notes in the United Kingdom that '[d]espite repeated legislation attempts to restrict the use of previous sexual history evidence in rape trials, it continues to be admitted in many cases'. 51 Numerous studies have found that negative views towards the complainant emerge from the introduction of previous sexual history evidence. For example, Shotland and Goodstein showed that it sets a precedent for jurors regarding future sexual interactions. 52 Shuller and Hastings found that the complainant was more likely to be viewed in a negative light by jurors where she revealed that she and the accused had an existing sexual relationship. 53 Bronstein's study of court transcripts in South Africa, undertaken in 1994 prior to the amendment of the CPA by SORMA which narrows the instances that previous sexual history evidence can be adduced, found that previous sexual history evidence 'does nothing other than access schemata, frames and scripts that imply that the complainant fits into the stereotype of a "loose woman" and is thus unworthy of the court's ambivalent protection'. 54 She argued that the South African legal system required a policy that prevents prejudicial questions being asked in the first place, yet admitted that it is difficult to imagine how this could practically be achieved. Kelly et al showed that previous sexual history evidence was frequently adduced in trials that they observed after the enactment of English laws which narrowed the instances in which such evidence could be adduced. Proper application was made to adduce such evidence in only one third of cases, and the reference to previous sexual history by the defence was predominantly aimed at 'sowing seeds of doubt' around the complainants testimony by invoking rape stereotypes. 55 Zydervelt et al observed that the defence counsel pursue very similar goals in relation to complainant cross-examination as those from 50 years ago. 56 They found that tactics leveraging rape myths are still common and defence counsel relied on previous sexual history evidence in 43% of cases, despite laws in place to prevent this. 57 The examination of female complainants in rape trials reflects our societal notions around how we stereotype women and who we deem as worthy victims. Pithey identifies an intersection between women who are deemed 'not credible' by the court and those that are seen as not 'rapeable' in society. 58 By scrutinising a woman's previous sexual history and inferring that it is likely that she would have consented, due to her past sexual behaviour, she becomes a woman who cannot be raped. Temkin et al show how rape myths shape what is considered a 'real rape', are used to discredit complainants, and emerge from a problematic interpretation of the facts of the case (including previous sexual encounters with the accused). 59 Adler identified six factors that impact a successful prosecution in rape cases: the complainant's sexual inexperience, her respectability, absence of incidents of consensual contact with the accused before the incident, resistance and early complaint. 60 Smythe further notes that a problematic restrictive interpretation of laws by judicial officers as well as masculinist interpretations of facts remains evident in the South African criminal justice system.
Spies argues that rape myths and gender stereotyping in our criminal justice system mirror the views of our communities and are deployed in gender-based violence matters to trivialise the experience of the rape victim. 61 She highlights 'how judicial attitudes in the sentencing of rape offenders can reflect, legitimise and enforce rape myths, celebrating male aggressiveness and punishing female passivity' and influence decisions around sentencing sexual offenders. 62 Focusing on judicial officers' discretion in respect of prescribed minimum sentencing, Spies shows how they rely on rape myths to give lesser sentences -focusing heavily on 'excuses' for the accused's behaviour rather than the seriousness of the offence. 63 She finds that the 'excuses' almost always correlated with gender stereotyping 64 and that the 'judge's interpretations of substantial and compelling circumstances [the criteria for increasing or decreasing a sentence] illustrate how masculine interpretations of the crime have influenced judicial reasoning'. 65 The cautionary rule is another area that gender stereotyping influences the criminal justice system by shaping the way the prosecution, defence and presiding officers see female sexual violence complainants. 66 The use of the cautionary approach was criticised for relying on various stereotypes in the case of S v Jackson, where the court stated 'the cautionary rule in sexual assault cases is based on irrational and outdated perceptions. It unjustly stereotypes complainants in sexual assault cases (overwhelmingly women) as particularly unreliable'. 67 The cautionary rule requires presiding officers to adopt caution when assessing the evidence of certain types of witnesses, such as single witnesses, child witnesses and previously, women. 68 The reliance on the cautionary rule in rape cases was based on the grounds that there were 'distinct' and 'peculiar' dangers in accepting the evidence of complainants as they may suffer from 'hysteria that can cause a neurotic victim to imagine things that did not happen', they may claim to have been raped for 'financial considerations when the complainant is pregnant' or have 'the wish to protect a friend or to implicate someone who is richer than him'. 69 Although this approach has been deemed unlawful by section 60 of SORMA, the trial transcripts examined in the study show that it is still applied (un)consciously by prosecutors, defence counsel and presiding officers, whose questioning of complainants appears to assume that women are lying about violations perpetuated against them and should be treated us untruthful until the evidence can show otherwise. In the case of S v Magagula 70 the presiding officer stated that 'it is always common that when one is sentenced to prison some members of the community will always take a chance to try and lay charges similar to rape'. 71 This paints women as opportunistic in instances where a man already has a conviction for a rape, and the presiding officer's presumption is that these are not valid or truthful accounts of rape. In these situations, the burden is shifted to the complainant to show that she is different to 'some members of the community' and is in fact believable and did experience rape.
Furthermore, in the case of S v Northander 72 the prosecutor introduced the complainant's previous sexual history in the following exchange.
Prosecutor: Did you have a boyfriend before this incident? The transcripts also reveal that prosecutors frequently fail to object to the defence's reliance on previous sexual history evidence. The prosecutor's duty to object extends to the 'unnecessary, aggressive or badgering crossexamination of state witnesses'. 74 There is a further legal duty on a prosecutor to object to this form of questioning, at least in so far as the questioning is unlawful in terms of section 227 of the CPA. Failure to do so can have a fundamental effect on the outcome of a trial and constitutes a violation of the rights of the complainant to dignity and equality. The defence's intention in this instance appears to be to establish animosity between the complainant and the accused, and to show that the complainant had falsely accused her uncle of rape as some form of retribution against him because he did not want the complainant to be in the company of the 'boys' mentioned above.
The prosecutor in question did not object to the defence's line of questioning around the complainant's previous sexual history nor insisted upon a section 227 of the CPA application by the defence. The example shows how entrenched the reliance on previous sexual history is in rape trials -so much so that it does not attract the attention of the prosecution or the magistrates in these matters, who seem to accept this as a normal part of adducing evidence, despite the fact that it is unlawful.
The next is an example of the problematic reliance on gender stereotyping in the instance of the admissibility of the complainant's previous sexual history in decisions by presiding officers.
In this example the complainant in the matter was a 17-year-old female minor who was raped by an unknown man. The presiding officer found in favour of the defence, acquitting the accused under section 174 of the CPA. Addressing the complainant's reliability, the magistrate stated the following, [s]he [the complainant] made no mentions (sic) to the court that she was seeing anybody else or that she was in a relationship with somebody but what is very interesting to note on the J88 is that the doctor asked her when was her last date of intercourse of consent it would appear that it was 2 days prior to her going to the doctors (sic). 76 The problem with this statement is that it assumes that the complainant has a duty to tell the court when she last had had (consensual) sex and that she is less reliable for not disclosing this information at trial. Furthermore, it imposes a conservative view that the complainant would need to have been 'seeing' someone in order for her to have sex. This line of questioning has undertones of 'slut shaming', where woman are prejudiced when they do not conform to the role of the 'virginal Madonna' and are consequently viewed as the antithesis of this image -a 'whore'. 77 This example illustrates how the court judges the complainant as lacking credibility because she failed to reveal to the court that she had sex two days before she went for her J88 exam. The presiding officer does not consider the complainant's rights to dignity and privacy, or the relevance of these questions, and finds her lack of disclosure problematic.
The examples highlighted above illustrate the fact that there is a systematic failure by all role-players to engage with the constitutional rights of the individual complainant when examining, cross-examining and making findings. In particular, the prosecutor and the magistrate's failure to apply the rights in the Bill of Rights is unlawful in so far as section 8(1) of the Constitution states that 'The Bill of Rights applies to all law, and binds the legislature, the executive, the judiciary and all organs of state'.

Conclusion
As shown above, and in the larger MRC study, the transcripts provide evidence of the ways that presiding officers and prosecutors problematically refer to the complainant's previous sexual history as a form of evidence in trial. The data shows that it is not only defence counsel that rely on evidence of previous sexual history: prosecutors too not only rely on this kind of evidence, but also fail to object to instances where the defence leads this evidence, and presiding officers make problematic findings based on the complainant's previous sexual history. Relying on this this type of evidence infringes the rights of the individual as set out in the Constitution.
Remedying this problem requires adherence and enforcement of the laws that are currently in place to prevent questions based on previous sexual history. If there is no willingness by prosecutors, defence counsel and presiding officers to abide by these laws then there must be sanctions for such. This could be misconduct procedures undertaken by the Magistrate Commission for presiding officers, the National Prosecuting Authority of South Africa for prosecutors and the Legal Practice Council for defense counsel. Furthermore, proper training around gender-based violence for both prosecutors and presiding officers, including on the types of questions that a complainant can and should be asked. On this Schwikkard argues that currently legislation is a poor training tool due to its reductionist approach, and instead that the any style of drafting must be accompanied by appropriate social context training. 78 A system of training which includes education around social context could assist prosecutors to understand that the kinds of gender stereotyping and the reliance on rape myths that are illustrated through this article are based in fact, and therefore should not be relied upon in rape matters. There must also be training around the types of questions and phrasing of questions that prosecutors can ask in a rape trial, and when prosecutors should object to questions posed by the defence. Prosecutors should also be provided with further training on the rules of evidence to ensure that there is a clear understanding that the reliance on the previous sexual history of the individual is unlawful in certain circumstances.
As Combrink explains, in rape cases complainants are not afforded the opportunity to challenge a court's ruling in the admissibility of previous sexual history evidence. 79 This is echoed in terms of cases where there is no formal application on behalf of the prosecution and/or the defence to proceed with such questioning. If the prosecution fails to act in objecting to this line of questioning by the defence then the complainant has no way of objecting to the questions. The Criminal Procedure Act must be developed to allow for complainants to challenge decisions by presiding officers around previous sexual history admissibility and the introduction of representation for the complainant so that objections can be made and challenges on decisions can be launched on her behalf. Artz and Smythe note that this already occurs in adversarial systems such as that of Ireland, where representations can be made by the representative of the complainant around points such as the admission of previous sexual history evidence. 80 In failing to change the discourse in South Africa's courts around gender and gender-based violence, prosecutors as well as magistrates are inadvertently biased and prejudiced as they are aligned with views that place the complainant in the default position of lacking credibility until proven otherwise.