9.1 Introduction

In legal literature, stereotypes are usually described as generalisations ‘about the attributes or characteristics possessed by, or the role that is or should be performed by, members of a particular group’.Footnote ¹ While stereotypes are often associated with inaccurate evaluations and misrepresentations of reality, their connection with evidentiary issues has not yet been fully explored in legal scholarship. Current research has especially focused on the impact of gender (and other) stereotypes on the adjudication of human rights more broadly, without attention being paid to evidence.Footnote ² Yet, many judgments, from various regional courts and bodies, demonstrate that stereotypes can prevent proper ascertainment of the facts of a case. They may lead to relevant pieces of evidence being ignored, irrelevant circumstances being given undue weight or higher standards of proof being imposed than would have otherwise been the case.

The Convention on the Elimination of Discrimination against Women (CEDAW), which was adopted in 1979,Footnote ³ has played a pivotal role in advancing women’s human rights by addressing gender stereotypes as one of the manifestations of discrimination rooted in unequal power relations, particularly through Articles 2 (f) and 5 (a). Article 2 (f) requires states to modify or abolish laws, regulations and practices that constitute discrimination against women. Article 5 (a) makes explicit reference to the stereotyped roles of men and women and prejudices or practices based on the idea of the superiority of either of the sexes. It requires states to take all necessary steps ‘to modify social and cultural patterns of conduct of men and women’ with the aim of eradicating these discriminatory practices.

As illustrated below, the Committee tasked with the implementation of this Convention – namely the CEDAW Committee – has made it clear that gender-based stereotypes may not only be conveyed through individual acts but also in laws, social structures and institutions. The Committee has shown remarkable determination in actively using its mandate to interpret and develop legal obligations on states to combat gender stereotypes. Through its general recommendations, as well as its views on individual communications submitted under the Optional Protocol to CEDAW,Footnote ⁴ the CEDAW Committee is paving the way for a deeper understanding of stereotypes in law and shedding light on the interplay between gender stereotypes and access to justice.

This chapter engages with the CEDAW Committee’s approach to stereotypes and evidence by providing an overview of the Committee’s contribution to the definition of their interaction. It does so by reviewing selected views delivered on gender-based violence, a domain where the Committee itself has expressed concern about the pervasive negative impact of stereotypes in undermining women’s access to justice.

Following this introduction (Section 9.1), a critical discussion of evidence law is first set out in Section 9.2. This aims to unpack the gendered assumptions embedded in this field and sets the stage for an examination of how gender stereotypes connect to debates about factual construction and the rational evaluation of evidence. Section 9.3 goes on to provide an overview of CEDAW’s framework for addressing gender stereotypes. It specifically focuses on the right to access to justice and how the CEDAW Committee’s General Recommendation 33 enabled a better understanding of the interplay between stereotypes and evidence. The section then delves into the Committee’s relevant jurisprudence. Section 9.4 offers an alternative approach to this interaction, which entails examining the impact of stereotypes through the concept of legal presumptions. This is followed by concluding thoughts and recommendations on the matter in Section 9.5.

9.2 Stereotypes and Evidence: An Emerging Issue

As the legal literature on stereotypes has consistently stressed, judgments should be based on law and facts, rather than myths, stereotypes and preconceived ideas.Footnote ⁵ Yet, is there a single, objective way to interpret law and facts? This question is far more complicated than the literature depicts it to be, evoking a classic debate in legal scholarship on how we conceive law, facts and the relationship between them as determinate categories. This debate assumes facts to be objective materials to be discovered and mechanically subsumed into a non-contested legal definition.

This section will delve into these ideas and their relevance for the study of stereotypes, with a twofold objective. First, it will expose how some tenets of evidence law systematically render women’s accounts unsound. Second, it will introduce the issue of stereotyping into a broader discussion about factual construction and its challenges.

Theories of factual construction have problematised assumptions about how facts enter the judicial process: how they are perceived, selected for their relevance and evaluated, with this whole process affected by the socio-cultural background of the legal actors involved in the trial.Footnote ⁶ Beyond the application of rationalist criteria of logic, scientific laws and maxims of experience, evaluating evidence is a process that involves a wide margin of discretion.Footnote ⁷

While modern legal theory has tended to altogether ignore the quaestio facti (the issue of factual construction), legal hermeneutics has shed light on the intertwined relationship between law and facts, and the way they mutually influence each other’s determination in the judicial process.Footnote ⁸ Causality, temporality, cultural dimensions and rules of language are shown to shape the interpretation of reality, thus undermining the pretence of an objective and unique factual account. The cultural dimension, in particular, concerns ethical and religious perspectives, social values, scripts and stereotypes that provide narrative frameworks to guide factual construction in a coherent way.Footnote ⁹ A narrative appraisal of the judicial process as composed of stories told by different parties is apt to reveal how factual components are likely to be selected and organised in accounts that are consistent with cultural frameworks, common sense, and regularities (id quod plerumque accidit, or common occurrences). This consistency makes the story appear plausible and coherent.

These theories have contributed to exposing the cultural contingency embedded in adjudication. Yet, they have not pushed their critique further and exposed how the rationalist tradition of evidence discourse and its assumptions about a universal and self-evident understanding of truth, rationality and justice, systematically erase the experiences of socially oppressed groups.Footnote ¹⁰

Against this background, critical legal theories have exposed the false objectivity and neutrality of legal methods and the rules of evidence. They have examined adjudication not as mere perception of the facts as raw material, but as a process that ‘purvey[s] truth’.Footnote ¹¹ Feminist legal theorists, in particular, have unpacked the assumptions underlying the dichotomic framing of procedural and substantive law, with the former being conceived as the domain of pure logic and rationality, and the latter being the place where politics belong. In exposing how method shapes substance,Footnote ¹² feminist theorists have questioned our ‘ways of knowing’ in law,Footnote ¹³ reformulating questions that bring original perspectives to the field. Instead of asking how value judgments exceptionally succeed in polluting law – portrayed as the realm of inherent objectivity – feminist discourse purports that power hierarchies are originally embedded in evidence law. This shift exposes a number of assumptions and mechanisms of the rules of evidence that systematically erase women’s voices: the preference for physical over psychological evidence, direct over indirect evidence, presence over absence, continuity over discontinuity.Footnote ¹⁴ Moreover, evidence law builds on the assumption that truth is accessible to any unbiased observer who shares with the community a set of ‘universally accepted generalisations about human behaviour’.Footnote ¹⁵ Thus, the idea of a ‘universal competence’ presents a hegemonic view of human behaviour as if it were universal, when in fact, women’s experiences and those of other subordinated groups differ from the dominant view. These experiences are thus silenced, or viewed as irrational or illegible.Footnote ¹⁶

In addition, the linear conception of truth as ‘singular, immediately apparent, and permanent’ has historically undermined the credibility of women’s accounts of sexual violence.Footnote ¹⁷ Reliance on assumptions about internal consistency, narrative coherence, hard or physical evidence and the stability of storytelling over time has contributed to shaping a notion of truth detached from women’s experiences of these violations.Footnote ¹⁸

At the core of feminist and critical approaches lies a common understanding that law in general, and evidence law in particular, is built on pre-existing social inequalities.Footnote ¹⁹ Ignoring this premise makes it difficult to understand why human rights violations perpetrated against the oppressed often remain invisible. Instead, critical theorists advocate for alternative understandings of knowledge, rationality, and truth that allow the experiences of the oppressed to make sense and regain value.Footnote ²⁰

Debates about stereotypes and stereotyping resonate with those concerning factual construction and the assumptions underlying the ascertainment of factual truth during adjudication (and beyond). Following the seminal work of Cook and Cusack on stereotyping in international human rights law,Footnote ²¹ legal research has consistently referred to stereotypes as generalised views or preconceptions about the attributes, characteristics or roles that women and men should possess or perform.Footnote ²² Their harmfulness is usually associated with making assumptions about an individual simply based on their membership in a particular social group. While anti-stereotyping theories in law date back to the 1970s,Footnote ²³ lawyers and theorists still struggle to identify the boundaries between stereotypes and other concepts such as stigma, prejudice and bias, and even generalisations, which are widely used in law.Footnote ²⁴ In addition, much scholarly work is devoted to gender and intersecting stereotypes in international human rights adjudication,Footnote ²⁵ while less research has focused on stereotypes based on other grounds of discrimination.Footnote ²⁶

Moreover, while stereotypes are often associated with misrepresentation of individual characteristics, and in general with inaccurate or false representations of reality, few contributions have paid attention to the specific impact of stereotyping on the assessment of facts and evidence in national and supranational jurisdictions.Footnote ²⁷ The first study to provide an overview of the varied ways in which stereotypes may impact probative issues was commissioned by the Office of the High Commissioner for Human Rights in 2013.Footnote ²⁸ The report focused on gender stereotypes and the adjudication of gender-based violence cases, pointing to how stereotypes affect the perception of facts. In particular, stereotypes impact perceptions of who is the victim and who is the aggressor, the culpability of the persons accused and the credibility of witnesses. They also affect the issues to be determined at trial, and may lead to the admission of irrelevant or highly prejudicial evidence, or to unjustified weight being attached to certain evidence.Footnote ²⁹

The need to improve gender-sensitive training for judges, prosecutors and other legal actors involved in the justice system has led to the development of protocols and professional manuals that provide more in-depth analysis of the interplay between stereotypes and evidence.Footnote ³⁰ There have also been important developments in how international human rights bodies understand the impact of stereotypes on access to justice – and in particular, on evidentiary matters – that demand scholarly attention. In this light, the following sections explore developments at the CEDAW Committee with regard to gender stereotypes and their impact on evidence.

9.3 Lessons from the CEDAW Committee

Within international human rights law, CEDAW provides a legal framework to combat gender stereotyping in view of achieving substantive equality. Article 2 (f) and Article 5 (a) explicitly require states parties to modify or transform gender stereotypes and eliminate wrongful stereotyping.Footnote ³¹ These have been interpreted as freestanding provisions but also as sources of an overarching and cross-cutting obligation to be read in conjunction with other human rights.Footnote ³² The CEDAW Committee has specified that the elimination of discrimination is an obligation with three dimensions: (1) ensuring there is no direct and indirect discrimination in both public and private spheres; (2) improving women’s de facto equality; and (3) addressing prevailing gender relations and the persistence of gender-based stereotypes.Footnote ³³ According to the Committee, gender-based stereotypes ‘affect women not only through individual acts by individuals but also in law, and legal and societal structures and institutions’. The obligation to address gender stereotypes binds the contracting state and its apparatus, including judges, prosecutors and law enforcement officials.Footnote ³⁴ It also extends to taking positive action against stereotyping by non-state actors and individuals (Article 2 (e)).Footnote ³⁵

This obligation was applied in connection with the right to access to justice in the leading case Vertido v. The Philippines.Footnote ³⁶ In this case, the Committee held the Philippines accountable for gender stereotyping by the judiciary, which affected the victim’s access to a fair and just trial. The Committee observed that ‘the judiciary must take caution not to create inflexible standards of what women or girls should be or what they should have done when confronted with a situation of rape based merely on preconceived notions of what defines a rape victim’.Footnote ³⁷

This principle has been repeatedly affirmed in subsequent views. It was also recalled in General Recommendation 33, which clarified the obligations to address gender stereotyping to guarantee the right to access to justice.Footnote ³⁸ In this recommendation, the CEDAW Committee considered the right to access to justice to be a bridge between substantive and procedural law and an essential component for the realisation of all the rights enshrined in the Convention. Acknowledging that access to justice in its various dimensions is currently hampered by obstacles related to structural discrimination and inequality,Footnote ³⁹ the Committee highlighted that gender stereotyping and procedural and evidentiary requirements are among the factors that constitute such obstacles.Footnote ⁴⁰ It further held that reasoning based on gender stereotyping impairs the justice system, as it ‘distorts perceptions and results in decisions based on preconceived beliefs and myths rather than relevant facts’.Footnote ⁴¹ Recalling the principle laid down in Vertido v. The Philippines, the Committee contended that by creating rigid standards of evaluation of human behaviour, gender stereotypes have a detrimental impact on the credibility given to women’s accounts as parties and witnesses.Footnote ⁴² All these elements can cause judges to misinterpret or misapply laws and compromise the impartiality and integrity of the justice system, leading to miscarriages of justice.Footnote ⁴³

Since its adoption, this General Recommendation has become a useful tool to further substantiate the Committee’s findings regarding gender stereotyping and elaborate specific recommendations to states parties. In subsequent communications, the Committee has interpreted the anti-stereotyping obligation in relation to other limbs of Article 2, which were invoked by the authors,Footnote ⁴⁴ even without a specific mention of Article 5 (a) on some occasions.Footnote ⁴⁵

While CEDAW lays itself open to some criticisms with specific reference to stereotypes and stereotyping (e.g. the absence of definitions),Footnote ⁴⁶ the Committee’s work has provided greater visibility to the negative impact of stereotyping in women’s rights adjudication. The Committee has also developed a more comprehensive approach to the issue in its recommendations.Footnote ⁴⁷ However, there remains more to be done regarding the naming of the specific stereotypes at play, exposing the inferences they trigger, and articulating the specific harms they cause.Footnote ⁴⁸ Despite this last limitation, CEDAW and the work of its Committee have deservedly become a reference in international human rights law for addressing gender stereotypes.Footnote ⁴⁹

This section will now discuss how the Committee has improved its reasoning on the impact of stereotypes on evidence. It does so by examining a set of individual communications which, submitted under the Optional Protocol, have resulted in a decision on the merits. Only communications where stereotypes were affecting the evidence and the Committee was seeing, naming and addressing this connection have been selected for analysis. In addition, only communications from the jurisprudence on gender-based violenceFootnote ⁵⁰ were sought, as this is the field where the Committee has expressed most concerns in terms of the negative impact of gender stereotypes on access to justice.

These cases illustrate the polluting effect of gender stereotypes on evidence at various stages of the legal proceedings, from investigations to the trial phase. Stereotyped reasoning often results in incomplete and biased investigations, where relevant facts and evidence are not collected. This already puts victims of gender-based violence at great disadvantage in proving their cases and seeing their right to access to justice fulfilled. At trial, stereotypes may shape how the factual hypothesis is constructed, the weight given to certain facts and evidence, and the distribution of the burden of proof.

Shortcomings in the collection of evidence to substantiate the charges stand out as particularly problematic in gender-based violence cases. While the circumstances in which certain crimes are perpetrated (usually a private setting) might partially preclude the availability of witnesses, the prompt collection of other relevant pieces of evidence is also jeopardised by a stereotyped understanding of the facts. The Committee has adjudicated several communications that displayed these issues at the investigation stage, such as the failure of the police to record incidents of violence or collect witness statements, or denials or delays in opening investigations.Footnote ⁵¹ Although earlier views did not explicitly recognise how stereotypes would have created evidentiary hurdles, they recognised that discriminatory narratives distorted the perception of the circumstances that amount to gender-based violence, including the profiles of victim and aggressor.Footnote ⁵² Thus, stereotypes hampered the investigation and collection of evidence by law enforcement officials. This nexus was made explicit in Vertido, and later in V.K. v. Bulgaria (respectively, cases of sexual assault and domestic violence). In the latter case, the Committee pointed out that ‘exclusive focus on physical violence and immediate threat to the life or health of the victim reflects a stereotyped and overly narrow concept of what constitutes domestic violence’.Footnote ⁵³

Many individual communications before the Committee also point to a biased assessment of evidence during the trial phase of domestic proceedings. This has taken place through practices such as reliance on irrelevant facts over relevant ones based on stereotyped views; selective consideration of evidence based on whether it confirms or contradicts stereotyped hypotheses; and stereotyped assessments of the author’s credibility.

The following paragraphs will examine the evolution of the Committee’s case law on stereotyped evidence. Reviewing the cases chronologically, the analysis aims at assessing whether the Committee’s reasoning has captured the impact of stereotypes on evidence in its various manifestations, and whether it has reached a rigorous and consolidated standard for its adjudication.

A first phase can be identified when the Committee was able to detect shortcomings in domestic authorities’ consideration of available evidence, but could not connect these to hints of stereotyped reasoning. In Goekce v. Austria,Footnote ⁵⁴ a case of domestic violence that ended in femicide, the Committee rightly pointed to the disconnect between abundant evidence of serious danger to the life of the author and the failure of the State to exercise due diligence to protect the victim.Footnote ⁵⁵ However, the Committee did not consider that the submissions warranted further findings on Articles 1 (anti-discrimination provision) and 5 (anti-stereotyping provision). The fact that the public prosecutor withdrew the charges based on the assumption that threats were a regular feature of the couple’s disputes, as well as the inability to ascertain which spouse started the violence, would have deserved additional analysis.Footnote ⁵⁶ The normalisation of violence within a couple and the argument of mutual conflict are recurrent discriminatory narratives that serve to downplay the severity of the violence perpetrated against women and overshadow its roots in the gender roles that relegate women to inferior social status. Another case of domestic violence that ended in a fatal stabbing, Yildirim v. Austria, was adjudicated on the same day.Footnote ⁵⁷ Similarly to Goekce, here the Committee noticed a contradiction between the dangerous situation the victim faced and the inaction of the local authorities that led to the femicide.Footnote ⁵⁸ However, it did not explore whether this contradiction could have been motivated (and normalised) by a stereotyped understanding of violence.Footnote ⁵⁹ The contradiction in the State’s reasoning is apparent: while not disputing the death threats and the ineffectiveness of the police orders and interim injunction, the State still considered it disproportionate to issue an arrest warrant, since the aggressor ‘had no criminal record and was socially integrated’.Footnote ⁶⁰ At this early stage of the Committee’s evolving approach towards stereotyped evidence, both the stereotypes of the ‘ideal victim’ in Goekce and the ‘ideal perpetrator’ in Yildirim went unnoticed, as the Committee explicitly refused to analyse the cases under Article 5.

Vertido v. The Philippines marked a shift in the Committee’s anti-stereotyping analysis and signalled a more explicit understanding of the link between stereotypes and probative issues. As Cusack and Timmer have pointed out, part of this shift is attributable to the author’s success in exposing the connection between stereotypes and the contradictions in the assessment of evidence carried out by the domestic court.Footnote ⁶¹ Following the author’s narrative, the Committee was able to retrace the link between stereotypes and assumptions that undermine the author’s credibility and impair the evidentiary process. Examples include assuming consent from lack of resistance to the assault or acquaintance with the aggressor; assuming rape accusations can be made with facility (the stereotype of fabrication); and giving less credibility to rape allegations based on the perpetrator’s age.Footnote ⁶² In this case, the author was a female executive director at the local chamber of commerce who had been raped by the president of the chamber. Her testimony was not believed by the domestic criminal court because she did not comply with the behaviour of the ‘ideal victim’. One year later, the Committee took this approach further in another key case, V.K. v. Bulgaria. Here, it explicitly found a link between gender stereotypes and the domestic court’s narrow conception of domestic violence, which led the latter to apply a higher standard of proof and place an excessive burden upon the author.Footnote ⁶³ In particular, the domestic court had refused to issue a permanent protection order against the author’s husband, based on the assumption that striking someone may amount to violence ‘only after breaking certain limits of abuse’.Footnote ⁶⁴ According to the court, the author had not demonstrated that those limits were surpassed. The Committee found that this stereotyped understanding of domestic violence had led the court to neglect the history of verbal, economic and psychological abuse documented by the author.Footnote ⁶⁵

Compounded stereotypes might pose extra obstacles for assessing a victim’s credibility. This is demonstrated in R.P.B. v. the Philippines, an individual communication submitted by a young girl who was deaf and mute. Her account had been assessed by domestic courts against the standard of an ‘ordinary Filipina female rape victim’.Footnote ⁶⁶ According to a ‘reasonable standard of human conduct’ proclaimed by the domestic court, such an ideal victim should have ‘summon[ed] every ounce of her strength and courage to thwart any attempt to besmirch her honour and blemish her purity’.Footnote ⁶⁷ Here, the Committee rightly pointed out that the guiding principles derived from judicial precedents in deciding rape cases amounted to gender stereotypes. In particular, the stereotype of the ‘ideal victim’ impaired the victim’s credibility assessment and led the court to disregard her individual circumstances (i.e. disability and age).Footnote ⁶⁸

The Committee also found a connection between stereotypes and a biased assessment of facts and evidence by domestic authorities in González Carreño v. Spain. This was a case of domestic violence that resulted in the killing of a young girl by her father during court-authorised visitation. The domestic authorities had dismissed evidence of the danger posed by the father to the safety of his daughter and ex-partner, in favour of inconclusive reports submitted by social services.Footnote ⁶⁹ The Committee found that domestic authorities’ stereotyped conception of visitation rights ‘gave clear advantages to the father despite his abusive conduct and minimized the situation of mother and daughter as victims of violence’.Footnote ⁷⁰ This was also connected to the incoherence which the Committee observed between the authorities’ actions and the evidence submitted.Footnote ⁷¹ Yet, the Committee did not point out specific gender stereotypes that could have motivated the authorities’ inaction or the selective weighing of evidence, such as the stereotypes that violence is normal between (ex-)partners or that women fabricate accusations and their accounts are not reliable. This case certainly advances a welcome line of reasoning, but following Vertido, a more explicit and articulated discussion could have been expected.

These early cases planted the seeds of a better articulation of the stereotype-evidence link. However, the Committee’s analytical approach in this regard is far from consolidated. An example of inconsistency can be seen in a case adjudicated only one year after Carreño. In X. and Y. v. Georgia,Footnote ⁷² the Committee did not delve into the stereotypes underlying the prosecutor’s decision not to open a criminal investigation into domestic violence. These stereotypes could be seen in the emphasis placed on the good reputation of the aggressor, and on his social position as a student and businessman. Both elements distanced him from the ‘ideal aggressor’ standard. The violence was also viewed as a ‘permanent conflict’ that ‘did not surpass the normal framework of attitude within the family’.Footnote ⁷³ Moreover, the minor daughter – who was a victim of violence herself – gave statements confirming the beatings, but these were put into question by a psychological assessment (considered to be produced by an authoritative source). This finding reflects the stereotype that depicts both women and minors as inherently unreliable.Footnote ⁷⁴ Additionally, the Committee remained silent on the State party’s view that the author and the grandmother ‘brainwashed’ their children, ‘causing them to testify against their father’.Footnote ⁷⁵ This could be read in conjunction with the attempt to depict the author as mentally unstable, as both elements concur in her negative credibility assessment. While the author explicitly denounced the existence of a socio-cultural pattern of conduct which ‘accord[s] greater weight to the word of a man and accept[s] a level of physical violence and sexual touching as being within the realms of acceptable parenting for a man’,Footnote ⁷⁶ the Committee did not address the issue of selective weighing of evidence. Furthermore, it did not explicitly connect the lack of objective standards for assessing prima facie evidence, or the failure to gather and present evidence impartially, to underlying gender stereotypes about women’s voices. It did, however, interpret several of these shortcomings as placing an extremely high burden of proof upon the author.Footnote ⁷⁷ Nonetheless, the Committee in this case fell short of detecting and challenging a number of visible stereotypes, and missed the opportunity to discuss their effect on evidence, even though it had done so in previous views.

Just a month after X. and Y., the Committee adopted its General Recommendation 33. It seems to be no coincidence that the views adopted afterwards mark an improvement in the analysis of the connection between gender stereotypes and evidentiary issues. This link is better delineated, for example, in L.R. v. Republic of Moldova, a case regarding the State’s failure to carry out a prompt and effective investigation into domestic violence.Footnote ⁷⁸ The author argued that the State had given insufficient weight to her statements and had not heard her witnesses or assessed her supporting documentary evidence. The Committee found that the domestic court’s reasoning, which led to the denial of protection, had been biased. This was based on the Committee’s consideration that the court had misinterpreted the violence as a civil dispute; given undue weight to the accused’s good references; interpreted the contradiction between the parties’ accounts with prejudice to the author; and disregarded medical and forensic evidence produced by her.Footnote ⁷⁹ Furthermore, the author’s attempts at seeking justice had resulted in a police officer reporting her as a family troublemaker, which undermined her credibility at trial. Moreover, prosecutorial authorities had focused on her mental state and even attempted to force her to undergo in-patient psychiatric testing – an additional sign of how women’s accounts are not deemed credible.Footnote ⁸⁰ Domestic authorities had also neglected relevant facts such as the protection order granted to the author, as well as witness statements and records of ambulance and police services which had responded when the accused attempted to strangle the author.Footnote ⁸¹ According to the Committee, all these elements – undisputed by the State – indicated that the decisions at domestic level had been based on stereotyped notions of domestic violence.Footnote ⁸² However, the Committee did not explicitly name the specific stereotypes underlying the case – such as the interpretation of women’s reactions to domestic violence as irrational, or the assumption that their claims are fabricated.

The stereotype-evidence link was finally made fully explicit in two cases in 2018. X. v. Timor-Leste concerned the criminal trial of a victim of domestic violence who claimed to have acted against her abuser in self-defence.Footnote ⁸³ The criminal proceedings had been fraught with serious negligence in the collection of evidence of domestic violence and its subsequent handling. The Committee explicitly pointed out that despite a retrial being granted to allow a due consideration of self-defence, ‘[judges] allowed gender stereotypes and bias to affect the weighing of evidence in the second trial, in particular by lending the author’s voice less credence than that of her nephew, who had not been present at all relevant times’.Footnote ⁸⁴ Similarly, in J.I. v. Finland, a case involving domestic violence and child custody proceedings, the Committee expressly pointed out how stereotypes pollute the rational assessment of evidence: by undermining the claims and evidence of the victim/survivor, while simultaneously supporting the defence arguments advanced by the alleged perpetrator.Footnote ⁸⁵ Recurrent elements appeared in this case too: from the biased attitude towards the author and her statements (she was subjected to a psychiatric assessment and depicted as a hostile mother), to the neglect of evidence of violence, with prejudicial impact on custody arrangements.Footnote ⁸⁶ In both cases, the Committee made extensive reference to General Recommendation 33 to support the stereotype-evidence link.

A preference for evidence submitted by the aggressor over that submitted by the victim was also at the core of S.T. v. Russian Federation.Footnote ⁸⁷ Yet, the anti-stereotyping reasoning in this case was not entirely rigorous and left some problematic aspects untouched. While the Committee read the shortcomings in evidentiary evaluation as products of stereotyped reasoning, it missed the opportunity to challenge some specific arguments. For example, it remained silent on the prosecutor’s request to mitigate the classification of the crime. This had been based on a psychiatric examination that excused the violent attack as an explosion of affect stemming from the loss of authority of the accused within the family.Footnote ⁸⁸ This line was followed at trial by the court and led it to dismiss the victim’s claim for compensation. Moreover, the court accused her of provoking and humiliating the aggressor through litigation and thus being responsible for the attack she suffered. In a narrative framework where hitting a woman with an axe was interpreted as a ‘nervous breakdown’,Footnote ⁸⁹ a woman’s attempts to seek justice were viewed as hints of ‘systemic amoral behaviour’ that worked to excuse the behaviour of the aggressor.Footnote ⁹⁰ A more careful analysis would have been needed here to unpack the provocation argument, and to highlight the greater probative value attached to authoritative sources – in this case, the positive description of the aggressor by the mosque administration.

The unquestioned preference for evidence produced by authoritative parties (e.g. medical staff, social services, police officers) is another specific manifestation of the selective evaluation of evidence. These parties are generally assumed to be acting in the interests of the patient/client/victim, even when other evidence disproves this assumption. In S.F.M. v. Spain, a case of obstetric violence, the Committee acknowledged that stereotyped and thus discriminatory notions led the domestic authorities to assume that the doctor should decide whether or not to perform an episiotomy, despite lack of consent and evidence of non-compliance with lex artis (the current state of medical knowledge).Footnote ⁹¹ Regrettably, the Committee did not explicitly name this stereotype in its conclusion, although the author did mention that women are stereotyped as incapable of making their own choices and are treated as passive objects with a reproductive role.Footnote ⁹² The Committee further addressed the preference for evidence by authoritative parties in the recent case of A.F. v. Italy.Footnote ⁹³ In this case, a woman alleged she was sexually assaulted by a police officer, to whom she had previously reported an assault by her ex-husband. The Committee acknowledged that the decision to overturn his conviction for lack of evidence could be explained by stereotypes that led the domestic court to give uncritical preference to the accused’s evidence.Footnote ⁹⁴ In particular, the court had re-interpreted the hospital report and the psychologist’s diagnosis showing the possibility of rape as evidence of the accused’s seductive ability. Conversely, the author’s prompt reaction in collecting evidence after the assault was interpreted to suggest a fabricated accusation.Footnote ⁹⁵ Importantly, the Committee noted the striking difference between the treatment of the author’s evidence in contrast to that of the accused, as well as their contradictions. Indeed, while the author’s revisions were interpreted as lies, ‘the accused [could not] be blamed for his divergent explanations’.Footnote ⁹⁶

As emerges from this overview, the analysis of the stereotype-evidence link has not followed a linear or consistent path. Following the adoption of General Recommendation 33, this analysis has improved and the Committee’s views have highlighted examples of selective weighing of evidence and preference for evidence submitted by the accused or produced by authoritative sources. In some cases, the link between these evidentiary hurdles and stereotypes is made explicit. In others, it emerged implicitly as a contradiction between facts presented and evidentiary findings at the domestic level. In other words, the Committee’s approach in this regard requires further consolidation.

The extent to which the authors draw attention to stereotypes and their evidentiary effect may contribute to explaining the Committee’s inconsistencies in its approach. Some submissions, like those in Vertido, R.P.B., S.F.M., and A.F., see the authors meticulously exposing and analysing the stereotypes behind each biased assessment made by domestic authorities. In other instances, the authors leave the stereotypes at issue unnamed, but invoke general gender-related myths of how women must behave,Footnote ⁹⁷ gender bias,Footnote ⁹⁸ or the Vertido formula regarding judicial creation of inflexible standards of behaviour for victims.Footnote ⁹⁹ While more articulated reasoning by authors on stereotype-related violations of CEDAW is welcome, this should not have to stand in for the Committee’s ability to carry out its own anti-stereotyping analysis. The Committee should systematically name and contest stereotypes, as a necessary step before examining the impact of stereotypes on evidentiary findings.Footnote ¹⁰⁰ Its present lack of systematicity is problematic, the more so as the Committee specialises in gender-based discrimination.Footnote ¹⁰¹ Clearer indications on its part as to what stereotypes are and the extent of the harm they produce would improve its own anti-stereotyping reasoning, as well as serving as guidance for other international bodies .

9.4 Exploring Stereotypes through Legal Presumptions

Given the widespread impact of stereotypes on the evidentiary domain, this final section seeks to explore how arguments and practices stemming from stereotypes might map onto legal mechanisms, which provide them with an appearance of soundness. Using the González Carreño communication before the CEDAW Committee as a case study, the analysis focuses on the hybrid mechanism of the legal presumption, which displays both evidentiary and substantive traits.Footnote ¹⁰² This comparison between stereotypes and legal presumptions illustrates the connections between stereotypes and evidentiary issues that the CEDAW Committee should be able to detect and challenge as forms of discrimination. In particular, the Committee should be aware that gender stereotypes may prevent the ascertainment of truth at trial by introducing ready-made narratives that resist contrary evidence, in a way similar to legal presumptions.

Presumptions are mechanisms deployed in judicial reasoning out of necessity, when it is not possible to ascertain a fact by direct evidence. De facto presumptions are mere inferences or arguments based on generalisations, usually drawn from empirical regularities or scientific knowledge. Legal presumptions, on the other hand, are characterised by foreclosing the ascertainment of truth, through substituting an unknown fact that should be ascertained with another (known) fact.Footnote ¹⁰³

Rebuttable (or relative) presumptions establish a provisional truth, whereby the presumed fact is temporarily established, until contrary evidence is submitted. Such presumptions might be established by law or by the judiciary. They do not necessarily reflect a regularity or probability, but might rather stem from social and political reasons. For example, the presumption of paternity of children born in wedlock does not necessarily reflect empirical facts, but serves the social function of protecting children’s interests. Notably, relative presumptions possess both constitutive and regulatory aspects: they contribute to shaping legal meanings or ‘constituting the presumed fact into an institutional outcome’.Footnote ¹⁰⁴ For example, the abovementioned presumption of paternity produces legal effects based on marriage between the parents without proof of paternity being required. Relative presumptions are also procedural rules which indicate the distribution of the burden of proof to the judge and the parties at trial.Footnote ¹⁰⁵ In contrast, absolute presumptions establish an irrebuttable truth by requiring a fact to be viewed as corroborated on the basis of the mere occurrence of another fact. No effort is made to verify the actual situation, and there is no possibility to submit any contrary evidence.Footnote ¹⁰⁶ Absolute presumptions do not fulfil any epistemic function and are more likely to shape legal categories and definitions.

Legal presumptions of both kinds share a constitutive feature, by which they can shape the elements of a legal definition. This highlights the pertinence of the interplay between evidence and substance. When gender stereotypes guide the selection of relevant facts, they also influence legal definitions of what amounts to, for example, violence and discrimination. The circular dynamic between the evidentiary and the substantive dimensions exposes how relevance itself is shaped by stereotyped frames, and how these frames might affect the boundaries and content of legal definitions. An analysis of the González Carreño case can be used to illustrate how stereotypes act as presumptions, with stereotypes also establishing provisional or even absolute truths, thereby allowing discriminatory views to hamper the rational evaluation of evidence.

González Carreño was one of the first cases adjudicated by the CEDAW Committee that problematised the treatment of violence in custody proceedings and called into question the primacy given to a father’s interests over the best interests of the child. As previously recounted, the case concerned the murder by a man of his daughter during an unsupervised visit authorised by the domestic court.Footnote ¹⁰⁷ In the proceedings at domestic level, it appeared that national authorities had granted more weight to accounts from social services while completely neglecting the numerous domestic violence complaints filed by the author (the girl’s mother).

This case illustrates how gender stereotypes may distort the evaluation of evidence, with evidence provided by the accused being privileged over that submitted by the victim, and inconclusive evidence being granted more weight than other consistent pieces of evidence. In Carreño, this was exemplified by the treatment of the reports from social services as opposed to the evidence provided by the author. The reports by social services contained a self-contradictory assessment of the father’s parental capacity. For example, it was mentioned that he lacked empathy; had difficulty properly interacting with his daughter; used his child to transmit messages of animosity to the mother; and showed signs of an obsessive-compulsive disorder. Nevertheless, the social services, as well as the domestic court that relied on their assessment, seemed satisfied with a minimum display of affection from the father and dismissed other evidence to the contrary. The court thus concluded that there was a need to restore the relationship between father and daughter. Contrary evidence which was ignored included around thirty complaints filed by the mother, which had resulted in a conviction for harassment, as well as protection orders issued in favour of the mother and daughter.

Such an irrational assessment of evidence can only arise from the stereotype/presumption that women are untrustworthy, as they are considered more emotional and less rational than men. This stereotype usually underlies inferences that women tend to fabricate false allegations, that they create hostility or alienate their children from the father or that they are responsible for provoking violent attacks on themselves or their children. Originating from the idea of the inferiority of women, such narratives undermine the credibility of women and the relevance of the evidence they submit.

Closely linked to the public/private division, the stereotype of women as primary caretakers underpins the very different standards against which mothers’ and fathers’ parenting capacity is measured, as also apparent in González Carreño. Mothers are subjected to pervasive scrutiny of their behaviour.Footnote ¹⁰⁸ They are expected to be collaborative and facilitate the relationship between children and father, usually by sacrificing their own physical and psychological well-being. In contrast, fathers’ behaviours are often not scrutinised. In some cases, they are not even held accountable for their negative actions, nor are they compelled to improve their parental abilities.Footnote ¹⁰⁹ In the González Carreño case, the father did not comply with his maintenance obligations, nor with protection orders. He also engaged in behaviours which were detrimental to his underage daughter. However, none of these circumstances compelled the national authorities to take action, or at least consider them as relevant elements in their assessment.

As the Committee noted in González Carreño, national courts, social services and psychological experts were moved by the purpose of normalising the relationship between father and daughter, despite the reservations expressed about the behaviour of the father.Footnote ¹¹⁰ The CEDAW Committee identified this pattern of action as a ‘stereotyped conception of visiting rights based on formal equality’, which ‘gave advantages to the father despite his abusive conduct and minimized the situation of mother and daughter as victims of violence, placing them in a vulnerable position’.Footnote ¹¹¹ In other words, the domestic authorities used the principle of joint custody not in light of the best interests of the child, but to guarantee the privileges and entitlements of the (male) head of the family. This reaffirmed the gender hierarchy by which the father’s superior role justified submission of and violence against mother and children.

Judicial practice may normalise these abusive behaviours by either neglecting or trivialising their importance, and leaving them without legal consequences. The effects are those of a legal presumption: on the level of evidence, the stereotyped reading of joint custody establishes that men – setting the standard of rationality – are presumed to be good parents. This truth is hard to rebut, since gender stereotypes usually lead to downplaying or discarding contrary evidence as irrelevant (even criminal convictions and restraining orders, as was the case in González Carreño).Footnote ¹¹² Unless there is hard, physical evidence that domestic violence occurred, or that it affected the child directly, it is presumed that a father has a right to a relationship with his child that almost nothing can curtail.

On the level of substantive law, such judicial practice shapes the legal boundaries of desirable fatherhood by portraying violence, neglect or lack of affection as tolerable features of fathers’ parenting.Footnote ¹¹³ Normalisation is reinforced by impunity and lack of consequences when parental obligations are not fulfilled, as happened in the Carreño case. This judicial presumption is so strong and pervasive that commentators have suggested that in order to counteract it, a legal presumption against contact should be instituted in cases of domestic violence.Footnote ¹¹⁴

Gender stereotypes in child custody proceedings in situations of domestic violence work to downplay women’s accounts of violence and their attempts at seeking protection. They filter and transform accounts of violence into irrational claims, through accusations of parental alienation and certain uses of joint custody that conceal male privileges. Either the credibility of allegations of violence is undermined, or the blame is shifted onto women. Overall, gender stereotypes work to protect gender norms and hierarchies within the family that relegate women to silence. This is achieved by systematically suppressing all attempts to ascertain the truth – a feature of absolute presumptions. Looking at gender stereotypes through legal presumptions sheds light on the issues of selective weighing of evidence, and might be helpful in detecting this specific feature of stereotypes. The CEDAW Committee should be more aware of these evidentiary issues, and adopt a more assertive role in examining the inconsistencies it finds and their connection to stereotyped understandings of gender roles.Footnote ¹¹⁵

9.5 Conclusion

This chapter has shed light on the way gender stereotypes can negatively affect the rational evaluation of evidence and eventually compromise the adjudication of women’s human rights. The analysis of individual communications has pinpointed different examples of how stereotypes lead to a systematic disregard of women’s statements and evidence, which manifests through contradictions between the domestic courts’ evaluation and the evidence produced. Engaging in more rigorous analysis of the factual construction (quaestio facti) is a crucial step towards ensuring that stereotyped arguments do not impair access to justice and to a fair trial.

The CEDAW Committee has made progress in its stereotype-evidence analysis, with some remarkable examples in its recent case law. That said, its anti-stereotyping reasoning needs to be further consolidated. Naming and contesting specific stereotypes is crucial to allow subsequent scrutiny of domestic judgments to ensure that any discrimination can be properly redressed.Footnote ¹¹⁶ This is still a pending task for the CEDAW Committee, which often seems to rely on the authors’ work to identify the stereotypes underlying problematic arguments, without adding to or clarifying the content and inferences attached to them. While authors’ activism is key to exposing the contradictions embedded in the evaluation of evidence, the Committee should strengthen its own independent ability to detect stereotypes and the circumstances when they pollute evidence, irrespective of the author’s characterisation. This is paramount if the Committee seeks to develop a more solid anti-stereotyping approach in the future.

A more robust and consistent anti-stereotyping reasoning is also essential for the subsequent assessment of its evidentiary effects. The Committee is not alone in this endeavour – other regional bodies and judicial authorities should also be encouraged to make efforts to identify and challenge the stereotypes underlying all phases of the proceedings. The Committee – as a specialised body – could offer them guidance. A starting point could be to pay attention to selective weighing of evidence and preference for authoritative sources. These are two evidentiary issues that could be investigated in connection with other grounds of discrimination (e.g. race and ethnicity, disability, age), and in relation to human rights violations beyond gender-based violence.

In developing a stronger anti-stereotyping approach, it may be instructive for the Committee to more closely consider how stereotypes prevent the ascertainment of truth in a way that is similar to the functioning of (mostly) absolute presumptions. This kind of comparison can shed light on the full extent of the effects that stereotypes may have, particularly in the evidentiary domain. With a better understanding of their effects and harms, a more nuanced, articulated conceptualisation of stereotypes in law can be elaborated. Without this effort, courts and judicial bodies lack an important standard to examine discriminatory practices and their impact.