Unsettling Culture and Rights

The project to eliminate gender-based violence is enshrined in global agreements such as the Convention on the Elimination of all Forms of Discrimination against Women (CEDAW, 1979), the Vienna Conference on Human Rights (1993), the Declaration on Elimination of Violence Against Women (DEVAW, 1993), and the Beijing Platform for Action (Beijing PoA, 1995) (Zwingel Reference Zwingel2016). The AU’s frameworks largely build on these agreements. While the significance of these frameworks for violence against queer and non-conforming groups can be questioned, these debates are outside the scope of this article. Conventionally, violence against women includes physical, sexual, verbal, emotional, and psychological abuse, threats, and coercion, as well as economic and educational deprivation. While domestic violence, rape, and sex trafficking are global problems, issues such as honor killings, dowry murders, female genital mutilation (FGM), polygamy, and child marriage are associated with local “cultures” and “traditions.”

An interpretive analysis of the AU’s gender frameworks is necessary to make sense of how this pan-African organization has translated the problem of violence against women into regionally relevant policy agenda. According to Carol Bacchi and Joan Eveline (2010), governing is guided not only by policies but also by the problematizations these policies present. Policies do not simply represent real-life issues, they produce and constitute policy problems. Epistemologically, an interpretive approach asks us to question and disrupt the content of the policy problem and to consider alternatives by problematizing power, knowledge, categorization, and seemingly everyday relations as issues that are sidelined in solution-oriented policy processes (Archibald Reference Archibald2020:14). Therefore, this analysis is explicitly normative and seeks to reveal the disadvantages of a particular way of representing problems (Bacchi Reference Bacchi2009:44). The interpretive researcher acknowledges that how problems are presented is a process that can shape lives, experiences, and opportunities.

The findings of the interpretive analysis revealed a tension between culture and rights in the AU’s approach. The “culture vs. rights” debate reached its height in the 1980s and 1990s, when the above global frameworks were negotiated. The idea of universally applicable moral principles that can be translated into collectively agreed borderless norms continues to be contested on both political and cultural grounds (Mutua Reference Mutua2002, Reference Mutua2011). In CEDAW proceedings, diplomats and committee members reiterated a view of “culture” as a monolithic and consensual whole, while activists demonized “culture” as the cause of gender inequality (Merry Reference Merry2003a). Today, governments continue to weaponize “culture” to resist and avoid accountability to internationally agreed commitments on gender issues, despite officially subscribing to a “rights-based approach” (Merry Reference Merry2003a:947). On the one hand, debates on culturalism in African studies regard “culture” as an explanation for African particularities, while others criticize it for “essentialization, determinism, homogenization, de-historicization and societization” (De Herdt & Olivier de Sardan Reference De Herdt and de Sardan2015:10). Treating “culture, tradition, language, religion, ethnicity, locality, tribe or race” as static, bounded, stable entities that are vulnerable to outside disruption is a particular, culturalist view (Cowan, Dembour, & Wilson Reference Cowan, Dembour and Wilson2001:9–10). A more dynamic understanding of culture would see it as a “critical site of social action and intervention, where power relations are both established and potentially unsettled” (Procter Reference Procter2004:1).

To address this tension, the analysis in this paper draws on postcolonial and decolonial perspectives on gender and governance. The scholarship on the colonial governance of gender, the colonization of subjectivity, and social constructions of the customary are useful in questioning the prominent role the undefined “culture” is afforded in understanding gender relations in Africa. Controlling sex and sexuality was a central aspect of colonial authority and imagination, intersecting with race and class (McClintock Reference McClintock1995). Violence against women by European men, including sexual violence, was particularly rampant and mirrored attitudes toward lower-class women in Europe (Mama Reference Mama, Alexander Jacqui and Mohanty1997). The creation of biological “genders” became the justification for dichotomous male and female gender roles (see Amadiume Reference Amadiume1997; Amadiume & Caplan Reference Amadiume and Caplan1987; Lugones Reference Lugones2007; Oyewumi Reference Oyewumi1997; Phillips Reference Phillips2005; Musisi Reference Musisi1999). Gendered identities were also tied to ideas of religious piety and purity, which were disseminated through missionary work (Pereira Reference Pereira2005). The colonizers’ moral gaze was informed by a commitment to their version of Christianity and the accompanying social norms. This led to the domestication of women’s social and political roles and the policing of gendered rituals, forms of non-monogamous partnerships, and the appropriateness of dress and behavior, among other “civilizing” body-focused policies.

The colonial image of gender and sexuality went hand in hand with its proactive construction of the categories of the colonial “other” (Said Reference Said1995). Mahmood Mamdani’s seminal work shows how the “despotic, traditional and authentic” characteristics of “customary law” were tied in with the “native” political identity (Mamdani Reference Mamdani2001:655). Categories such as “customary” or “traditional” are socially negotiated and strategically used in struggles for power. While colonial administrations played a crucial role in reifying traditions, the legacies of these processes are equally inscribed by resistance, contestation, and cooperation (Spear Reference Spear2003:5; Ranger Reference Ranger, Hobsbawm and Ranger2012). Ann Towns points out that, under colonial rule, communities with relative gender equality were considered uncivilized. For instance, matrilineal kinship structures, along with female political authority, were obliterated in the name of progress (Towns Reference Towns2009; see also Berger Reference Berger2014). The colonizers’ attention to male chiefs disregarded highly influential women such as healers and spiritual leaders. Colonial administrations prioritized institutionalized political authority, while “feminine forms of public authority were marginalized as they largely operated outside the normative ‘monarchical, authoritarian and patriarchal’ order of the colonial apparatus” (Verweijen & Van Bockhaven Reference Verweijen and Van Bockhaven2020:10).

Non-Western and postcolonial feminist scholars have pointed out the ways in which colonial knowledge production about gender and African societies has shaped the gender and development discourse. Western feminist and development discourses tend to reproduce an ahistorical and unchanging notion of “culture,” of which non-Western and non-Christian Others are victims (Mohanty Reference Mohanty1988; Abu-Lughod Reference Abu-Lughod2013). This discourse has given birth to an image of powerless “third world women” living in pathological “cultures of violence” (Abu-Lughod Reference Abu-Lughod2013; Karimakwenda Reference Karimakwenda2020; Sangari Reference Sangari2005). There is a hyper-focus on culture and social norms in women’s rights interventions, to the point that harmful traditional practices and violence against women become inseparable in policy language. Female genital mutilation (FGM) is perhaps the most outstanding example of this. Scholars have convincingly traced and exposed the lingering colonial discourses in FGM advocacy and policy (see contributions to Nnaemeka Reference Nnaemeka2005). They explore the mechanisms of knowledge production and ask whose voices are platformed, concluding that the non-African modus operandi has provoked resistance to FGM-focused interventions. Furthermore, harmful traditional practices are represented as a non-Western phenomenon, their possible existence in Western societies being left unaddressed (Winter, Thompson, & Jeffreys Reference Winter, Thompson and Jeffreys2002; Longman & Bradley Reference Longman and Bradley2015). The “traditional” and “modern” are presented as opposites on the development spectrum. By graduating to modernity, African societies are expected to eliminate these harmful practices (Arce & Long Reference Arce and Long1999). An ambiguous idea of “culture” is routinely presented as the main obstacle preventing the continent from achieving its potential.

Critics argue that interpersonal elements have sidelined “how individual factors interface with historical, structural and systemic factors to reproduce [gender-based] violence” (Amisi et al. Reference Amisi, Vanyoro, Archibald, Gould and Mulondo2021:7). In this approach, non-normative factors like institutions, (global) political economy, and politics are hardly discussed as factors that perpetuate violence against women (True Reference True2012:10; Heise & Cislaghi Reference Heise and Cislaghi2016:9; Piedalue et al. Reference Piedalue, Gilbertson, Alexeyeff and Klein2020:91). Others argue that demonizing African cultures is backfiring on the normative agenda itself, as the dominant rights-based approach fails to harness the creative and emancipatory potential of cultures in Africa (Tamale Reference Tamale2008 Footnote ¹; Mutua Reference Mutua2002). According to Sylvia Tamale and Makau Mutua, the continent’s egalitarian and uplifting traditions have failed to make it into the women’s rights toolbox, which is a major limitation of the gender equality agenda in Africa.

Contextualizing Violence against Women as a Policy Problem

Since decolonization, two relevant parallel shifts can be observed: first, how “culture” was understood to be part of the gender-equality discourse; and second, the effects of the volatile dynamics of the Cold War. The common African positions adopted ahead of the World Conferences on Women in 1985 and 1995 reflect the changing discourse around culture and gendered violence. While the language of harmful traditional practices had become part of African actors’ dialogue with the UN a whole decade before the Maputo Protocol, this language was mitigated in their regional women’s rights law.

Between 1985 and 1995, African delegates held three preparatory regional conferences ahead of the UN World Conferences on Women. Their outcome documents, the Arusha Strategies (1985), the Abuja Declaration (1989), and the African PoA in Dakar (1994), show how the culture/violence nexus emerged over those years. The Arusha Strategies take a dynamic view of culture as a system of knowledge, values, and beliefs through which society organizes itself. Culture is seen as a source of confidence and stability, and “women are [its] prime custodians” (Arusha Strategies, Article 14, p. 9). Aligned with the research of postcolonial gender scholars, the document argues that the colonial experience weakened African cultural systems in ways that enhanced gender-based discrimination. It speaks of rebuilding a culture based on “autonomy, self-reliance, and equality:”

Four years later, however, the Abuja Declaration presented “culture” as a means of “dominating” women through harmful practices and asked member states to eradicate degrading cultural practices. At the same time, it encouraged knowledge exchange on “positive aspects of African heritage” and requested further research into how traditional practices impact women (Abuja Declaration, p. 23). This links to Tamale and Mutua’s critique of human rights, as they argue that the empowering elements within cultures and traditions require further investigation in order to potentially transform the rights regime.

The African PoA in Dakar followed the UN Third World Conference on Women in Nairobi. The conference output was considered radical, but its implementation was overshadowed by the 1980s economic crises and Economic Structural Adjustment Programmes (ESAPs). The African PoA prioritized poverty in addressing women’s concerns on the continent, and criticisms of the ESAPs were included in the Beijing PoA in 1995 (Made Reference Made1996:78). Indeed, the culture/violence discourse shift paralleled an increase on donor dependency throughout the 1980s.

In Dakar 1994, the Organization for African Unity (OAU), the United Nations Economic Commission for Africa (UNECA), and the Inter-African Commission (IAC) hosted a workshop on harmful traditional practices that violate the rights of women and girls. Dakar PoA reflected this by explicitly addressing violence against women and harmful traditional practices. International attention to harmful traditional practices in Africa goes back to the colonial period. Such practices were brought to the attention of the UN by colonial administrators in the 1950s, were revived at a WHO seminar in Khartoum in 1975, and were developed further in the 1980s (Winter, Thompson, & Jeffreys Reference Winter, Thompson and Jeffreys2002; UN Office of the High Commissioner for Human Rights 1995). During this period, considerable research on the harmful effects of various culturally informed practices emerged, but the call from the Abuja Declaration in 1989 to study empowering elements within cultures in Africa remains unanswered.

Because of the sensitive nature of FGM, it was decided that a bundle of “harmful traditional practices” in which FGM was presented as a health issue would be more palatable to the communities that practiced it (IAC, online Interview, June 22, 2021). With this purpose in mind, the IAC was founded by African delegates who had taken part in a seminar of the NGO Working Group on Traditional Practices in Geneva. The work of the 1980s culminated in the adoption of DEVAW in 1993, the 1995 Plan of Action for the Elimination of Harmful Traditional Practices Affecting the Health of Women and Children, and the emphasis on harmful traditional practices in the Beijing PoA. In light of conflicting evidence regarding the potential health risks of different types of genital cutting, FGM is now primarily addressed as a human rights issue (Hernlund & Shell-Duncan Reference Hernlund and Shell-Duncan2007).

The development of “violence against women” as a policy problem can be understood in the context of a web of interconnected historical trajectories. Decolonization, nation-building, anti-racism movements, and competing Cold War ideologies may help to decipher some of the symbolic shifts in governance discourses in Africa. The rhetoric of independence views “culture” as a political issue and a question of identity, race, and class (Ajayi Reference Ajayi1961; Okech Reference Okech2019:11). Therefore, in the nationalist discourse, the integrity of “African cultures” had to be protected from criticism. Simultaneously, the advent of (African) socialisms and alliances within the global left meant the prioritization of peace and development over individualistic “rights” (Zissner Reference Zissner2002). Consequently, the World Conferences on Women, the first forum to discuss violence against women internationally, brought into conversation the post-colonial, socialist, liberal, and other competing perspectives regarding the state of women. With the end of the Cold War, the gender-based violence discourse became more homogenous and geared toward liberal human rights. As Kristen Ghodsee writes: “The critique of capitalism has been almost entirely evicted from the global feminist movement as it is currently configured, with Western governments now proposing to ‘mainstream gender’ into the very institutions that the Soviet women and their allies were trying to tear down” (Ghodsee Reference Ghodsee2010:10). While African women’s activists are actively articulating African feminism, there is an ideological shift from pursuing social cohesion to prioritizing neoliberal ideas of individuality, responsibility, and rights.

The Maputo Protocol and Nuancing Gendered Violence

The development of the Maputo Protocol came shortly after the 1995 World Conference on Women in Beijing, and the document became one of the most progressive women’s rights frameworks in international law. Adopted in 2003, the Protocol came into effect in 2005 after ratification by fifteen member states. It has been praised for its bold content, particularly on regionally significant issues such as FGM, abortion, polygamy, and child marriage (Viljoen Reference Viljoen2011). While displaying sensitivity and subtlety over the relationship between “culture” and “violence,” the Protocol generally aligns with global commitments and language. The representation of the problem of violence against women carefully delinks culture from violence, while still invoking socio-cultural factors. The Protocol’s definition of violence against women does not make any reference to culture or tradition, and significantly the document uses the phrase “harmful practices” rather than “harmful traditional practices” (see Figures 1 and 2).

Figure 1. Breakdown of references to different types of violence and harm against women in the Maputo Protocol

Figure 2. Breakdown of references to different types of violence and harm against women in all analyzed documents

Note: In figures 1 and 2 the phrase “violence against women” is a direct reference to this concept, inclusive of gender-based violence. “Harmful traditional practices” include references to this concept, FGM, child marriage, and polygamy, because these are categorized as such by the frameworks.

In 1997, the ACHPR submitted a draft Protocol on women’s rights to the OAU. Since the IAC was already drafting a Convention on Harmful Traditional Practices, the OAU requested that the two drafts be merged into one document (Murray Reference Murray2005; Wandia Reference Wandia, Kombo, Sow and Mohamed2013; Adams Reference Adams, Yacob-Haliso and Falola2020). In 2000, various women’s organizations successfully pushed for their inclusion in the drafting and consultation process and found that the document fell well below international standards (CSO participant in the Maputo Protocol negotiations, online interview, May 27, 2021). However, some of the original drafters were hesitant to resume negotiations because of controversial issues such as polygamy and reproductive health and rights, which could have resulted in backtracking on the progress made thus far (CSO participant in the Maputo Protocol negotiations, online interview, May 24, 2021). Furthermore, the women’s organizations pushed for inclusive language covering women in polygamous marriages, those married as children, and more generally all women in Africa. An additional challenge was mobilizing the delegations from the member states and ensuring their participation in the negotiations (CSO participant in the Maputo Protocol negotiations, online interview, May 24, 2021). Finally, the overlapping events of new and ambitious organizations entering the process, and the institutional transformation contributed to the tone of the Protocol:

The Maputo Protocol addresses violence against women in five of its Articles: Article 2 on the Elimination of Discrimination Against Women; Article 3 on the Right to Dignity; Article 4 on The Rights to Life, Integrity and Security of the Person; Article 5 on the Elimination of Harmful Practices, and Article 11 on the Protection of Women in Armed Conflicts. There are special articles to address violence against vulnerable women, like the elderly or women with disabilities. Before going into those provisions, the Maputo Protocol defines violence against women as:

Harmful practices are defined as:

These definitions discursively produce a policy problem, giving it “shape and meaning” (Bacchi & Eveline Reference Bacchi and Eveline2010:111). They define what is included and what is excluded from gendered violence. In both definitions, harmful practices or violence are defined without reference to culture or tradition, focusing instead more on harm. This definition resonates with that of DEVAW, which, however, lists specific actions that are to be understood as violent, including “female genital mutilation and other traditional practices harmful to women” (DEVAW, Article 2, p. 2). DEVAW’s definition was later adopted word for word in the Beijing Platform of Action. The Maputo Protocol covers the same points but spreads them out across the document and links them only loosely to violence against women. Of the five articles that address violence, two include sections that discuss the socio-cultural factors for eliminating it. For instance, Article 2.2 in the Maputo Protocol calls on State Parties to:

The aim of this is to eliminate “harmful cultural and traditional” and other practices that are based on stereotyped social roles for men as well as women. Mirroring CEDAW’s Article 5, the Maputo Protocol encourages states to “modify” socio-cultural patterns to address cultural and traditional practices that are harmful to women and that produce inequalities (Maputo Protocol, Article 2.1. b) and 2.2, p. 7.). While the Maputo Protocol addresses the “cultural” and “traditional” within violence against women, it refuses to describe it in these terms. Another example is Article 4, which addresses harmful practices and suggests that peace education and social communication should be promoted to:

While cultural factors are mentioned as potential enablers of gendered violence, the Protocol is clear that no culture or tradition is inherently violent against women. Instead, “elements” in systems of meaning may be used to justify violent practices and behavior. Additionally, the Protocol addresses the extent to which cultural and traditional matters contribute to violence against women. Article 4.2. c) requires member states to identify the “causes and consequences” of gender-based violence, pointing to a need for a deeper understanding of the issue to be able to address it (Maputo Protocol, Article 4.2. c), p. 8).

The drafters drew up an article mentioning “harmful practices” instead of “harmful traditional practices.” The phrase “harmful traditional practices” was initially included in the draft; it had inspired one of the Protocol’s initiators (namely the Inter-African Committee on Harmful Traditional Practices) and had been used since the 1980s (IAC, online interview, June 22, 2021). The article on “harmful practices” called for the prohibition of all forms of FGM but did not identify other particular practices (Maputo Protocol, Article 5, b), p. 9.). Participants explained that the women’s organizations saw the concept of ”harmful practices” as a way to include practices of unethical medical research, forced sterilization, online abuse, bullying, acid attacks, and other types of abusive behavior that cannot be linked to any tradition and may even be of “Western origin” (online interviews: UN participant in the Maputo Protocol negotiations, May 17, 2021; CSO participant in the Maputo Protocol negotiations, May 24, 2021; CSO participant in the Maputo Protocol negotiations, May 27, 2021). The trafficking of women and unethical scientific experiments are addressed in Article 4. Another interviewee explained:

The Protocol directly addresses other practices that are defined elsewhere as “harmful traditional practices,” such as child marriage and polygamy. These are not mentioned within the harmful practices section. This is vaguely addressed through a commitment to eliminating practices “which negatively affect the human rights of women and which are contrary to recognized international standards” (Maputo Protocol, Article 5, p. 9). The strong language around practices such as FGM, child marriage, polygamy, and reproductive rights indicates that these issues were discussed and agreed upon, but carefully presented as harmful rather than as matters of culture. The critics argue that the preoccupation with and “zero tolerance” toward all forms of genital cutting and medicalization, which is adopted in the Maputo Protocol, raises debates around ideas of harm, freedom of choice, and agency (Hernlund & Shell-Duncan Reference Hernlund and Shell-Duncan2007:23–26). They point out the fallacy of ignoring the central role of women in the practice and the maternalistic attitudes which suppose that the practitioners do not have a proper awareness of what they are doing. Indeed, the Office of the Legal Counsel of the AU (2002) advised against including references to the medicalization of FGM, but the CSO groups clearly won that argument.

The considerations concerning culture, tradition, and violence in the Maputo Protocol mirror the language of the Beijing PoA. Like the Protocol, the Beijing PoA uses the language of harmful practices, but links it regularly to specific practices such as FGM, child marriage, and the son preference, or custom and tradition.Footnote ⁴ The Beijing PoA also occasionally displays carefully calculated wording, for instance, “any harmful aspect of certain traditional, customary or modern practices that violates the rights of women should be prohibited and eliminated” (Beijing PoA, Article 224, p. 92). However, this quote is more of an outlier. The commitment to shifting the language of “harmful practices” away from its tradition-oriented predecessor appears half-hearted, as it links violence against women directly with tradition, custom, and culture:

And

This representation, which creates a causal relationship between tradition, culture, custom, and violence, is not present in the Maputo Protocol. The concept of “harmful practices” is used in different ways in the two documents: while the Maputo Protocol broadened the term to include various forms of gendered violence, the Beijing PoA uses it as a replacement for “harmful traditional practices.” While the document analysis does not include the UN policy frameworks after the World Conferences on Women, they tend to use the same language as the Beijing PoA. Examples could be “A framework to underpin action to prevent violence against women” (2015) and the Spotlight Initiative (2017) by the UN. The first considers violence against women as “rooted in gender inequality, discrimination and harmful cultural and social norms” (UN Women Headquarters 2015), whereas the second uses language of harmful practices to talk exclusively about child marriage and FGM.

According to the drafters of the Maputo Protocol, they knew they had to reflect the complex terrains of “culture” and “tradition” in Africa for the document to be relevant to African women. The “problem of culture” is not embedded in history and custom, but it is a political project (Bacchi Reference Bacchi2012). The interviewees reflected on this sentiment in various ways:

The Maputo Protocol presents a diversion from representations of Third World cultures that emerged from colonial epistemologies—as timeless and unchanging, where culture/tradition/religion blur into one, their practitioners being without history—and from representations of Third World countries as monolithic spaces without complexities or variations (Narayan Reference Narayan1997:45–50). One of the participants in the Maputo Protocol negotiations explained it as political and strategic:

What the document achieved was thus to shift the focus from ahistorical, apolitical, and undefined “culture,” “tradition,” and “custom” to articulate the problem of violence against women through the concept of “harm.” The Protocol insists that any behavior, attitude, or practice harming women shall be eliminated—including those that are embedded in traditions or other value systems. The Protocol thus provides a solid basis for building a regionally relevant approach to eliminating violence against women. As we shall see, this opportunity was missed in the documents that came after.

The Culture/Violence Nexus after the Protocol

International legal documents are subject to heavy negotiation and are likely to be more conservative than the policy frameworks. This may explain why the post-Maputo frameworks took a more radical stance toward “culture” and reverted to reinforcing the culture/tradition/custom and violence nexus—presumably with the idea of being progressive. The documents’ problem definition, causality, and interventions construct an ambiguous “problem of culture,” an undefined, static, ahistorical monolith that accounts for violence across member states. No political, economic, or systemic issue is given comparable weight as to culture and tradition. Nor do post-Maputo policies qualify the concept of “harmful practices,” but instead return to using “harmful traditional practices.”

The AU’s Gender Policy does not define gendered violence, but its definition of “gender systems” states: “In most communities in Africa, women are dominated by men via patriarchal power, which has been a traditional and indeed a historical privilege for men” (Gender Policy, p. 29). The quote generalizes “African communities” as a coherent category, notwithstanding the histories of patriarchy across the world and the diversity of patriarchal systems. “African communities” then come across as stereotypically “consensual” wholes, with “ahistorical, backward-looking and patrimonialist” cultures (Olivier De Sardan Reference Olivier De Sardan2005:77). This is interesting, as “gender” itself is articulated as having “changeable” and “time-bound” masculine and feminine characteristics, as well as differences that are socially and culturally constructed (Gender Policy, p. 27). The AU GEWE Strategy defines gender-based violence as interlinked with “culture” and as context-specific. It includes:

The definition of “gender equality” denotes “an element of interpretation of social justice, usually based on tradition, custom, religion or culture, which is most often to the detriment to women” (GEWE Strategy, p. 63). These definitions provide us with a vocabulary and interpretations of how we should understand the contents of the documents.

Unlike other problem representations, Figures 3 and 4 show that socio-cultural characterizations of gendered violence (GBV) are present in all of the documents, most prominently in the GEWE Strategy (2018). Additionally, the coding strategy shows that the Maputo Protocol included a comparatively wide range of causes of gendered violence. While the Solemn Declaration of 2004 included lengthy paragraphs that addressed violence against women and culture (Figure 3), such references were not particularly frequent (Figure 4). This is why the argument in this article focuses on qualitative properties found in the wording, conceptual references, and silences in those documents. For instance, the SRHR Policy states that “traditional” practices are enabled by local norms, and sexual and domestic violence are being tolerated for “socio-cultural reasons” (SRHR Policy, p. 16) The policy claims that women “suffer in silence” without reporting abuse because of normative social beliefs. Interestingly, the SRHR Policy (2006) makes only one reference to the Maputo Protocol, building mostly instead on the 1994 International Conference on Population and Development (ICPD) and the Millennium Development Goals. The GEWE Strategy reinforces the socio-cultural logic of poor reporting, claiming that such matters are private (GEWE Strategy, pp. 35, citing UNDP Africa Human Development Report 2016, p. 50). Leaning on demographic health surveys by the UN, the Strategy provides evidence of violence against women being normalized in many countries. While this may be so, other studies reveal that victims distrust the criminal system and state mechanisms, which in turn remove incentives for reporting (Medie 2017, 418). Therefore, the reality is that victims do not ascribe to a cultural mindset that accepts violence as normal; rather, they are caught between social and institutional norms, which severely restrict their options.

Figure 3. Salience of different problem representations across frameworks

Figure 4. Frequency of different problem presentation across frameworks

Note: The figures show the prominence of problem representations relative to each other across analyzed documents (Figure 1, marked by years). “Salience” refers to the volume (long and numerous sections) of codes, whereas “frequency” refers to the occurrence (however often) of codes.

The proposed interventions further reproduce a focus on culture. The SDGEA (2004) calls for public campaigns to address gender-based violence and invites its Member States to make legislative changes to “alter the attitude and behavior of the African society” (See Table 3). The culture vs. rights approach is explicit, as is the expectation that protection, criminalization, and punishment will alter the mindsets of people in “African society.” Yet, the idea that human rights can only be granted through “enforceable rights to individuals” is an element of specific political culture, one that has not explored the potential of delivering rights through already existing social institutions (Mutua Reference Mutua2002; Zwart Reference Zwart2012).Footnote ⁵ SRHR Policy (p.19) argues that, for the “largest possible adherence” to legal changes, the community needs to be included in the discussions, as “socio-cultural values are deeply rooted in the mind of people.” Here inclusion represents an attempt to convince rather than engage in a dialogue. Furthermore, states are given the responsibility to “take all necessary measures to remove socio-cultural structures and norms that promote and perpetuate gender-based inequality” (General Comments on Article 14 (2), p. 8). The references to administrative, social, and economic measures, usually the mandate of the state, are not unpacked. The focus remains on legislative and socio-cultural change.

Table 3. Examples of how violence against women is represented

The proposed interventions include awareness-raising, information provision, and other knowledge-based interventions to change social norms. The maternalistic language and approach rest on the assumption that the practicing communities do not know better. This reflects criticism of the disconnect between elite women’s organizations dominating the gender agenda and speaking “development” language, and the more organic, often rural women’s groups which engage in locally embedded practices of empowerment (Amadiume Reference Amadiume2000). More strategically, the language of saving “third world women” is a useful communication strategy used to facilitate financial and other support for specific development interventions (Cornwall Reference Cornwall2016). However, African policies require an understanding of African women in terms of more than the dichotomy between “elite women” and the “grassroots woman who is perpetually poor, powerless and pregnant” (Okech & Musindarwezo Reference Okech and Musindarwezo2019:78). The different “cultures” from which women come do not just refer to ethnic belonging but are also political cultures involving relationships with formal and informal structures and perceptions of personhood.

The policies represent gendered violence predominantly as sexual violence and harmful traditional practices. For example, the GEWE Strategy explains:

By linking gender-based violence to harmful traditional practices, gendered violence is rendered not only social, but also cultural. While harmful traditional practices are never defined in AU frameworks, they are often cited alongside FGM. Furthermore, the GEWE Strategy proposes a “continental campaign to declare ‘illiteracy a harmful traditional practice’” (p. 25). It should be questioned what is hoped to achieve by categorizing illiteracy in this way, and who is being held responsible for its persistence.

Table 3 highlights examples of the “deep-seated presuppositions and assumptions” that are internal to the policy discourse (Bacchi Reference Bacchi2015:2). The Maputo Protocol uses open-ended words such as “legitimate and exacerbate” to emphasize probability. The SDGEA example assumes that punishment will result in positive change. General Comment No. 1 reveals an assumption that collaborating with traditional and religious leaders will result in the elimination of harmful cultural and traditional practices, while the GEWE strategy presupposes a connection between violence against women, tradition, and social norms. Importantly, the meaning of culture, tradition, and custom is assumed, as they are never conceptualized or defined. Culture and tradition are presented as homogenous and as relatively static, interrelated wholes which can account for violent practices against women in all Member States.

While groups in different parts of Africa may (or may not) share common socio-cultural conceptions, these, like elsewhere in the world, are: (a) changeable, (b) not equally applicable to all people, (c) not homogeneous, (d) not necessarily integrated into “world views,” and (e) not necessarily generated by fundamental “values” (Olivier De Sardan Reference Olivier De Sardan2005:83). These issues are not reflected in the AU’s gender frameworks, which reproduce an oversimplified culture/violence nexus. Tracing the development of the AU’s gender agenda shows that the end of Cold War brought along the advent of a global liberal individual women’s rights agenda, in which activists see culture as the primary impediment to gender justice, and antagonistic to individual rights (Merry Reference Merry2003b). Around the millennium, the African women’s movements became increasingly intertwined with transnational and Western feminisms, rather than pursuing an alternative path, which is evident from their approach to eliminating violence against women. Not least, the post-Maputo policies are not legally binding, and their drafting dynamics are less politicized than the Maputo Protocol. Consequently, the contestations around culture are more visible in the implementation gaps than in the policy adoption.