Conceptualizing consent: cross-national and temporal representations of sexual consent in young Africans’ creative narratives on HIV

Abstract

Sexual violence, comprising all non-consensual sexual acts, is an important driver of HIV infection in sub-Saharan Africa. Definitions of sexual violence rely on understandings of sexual consent, understood as a feeling of willingness that is communicated via shared indicators of consent. In this paper, through analysis of young authors’ narrative-based social representations, we sought to provide insight into young Africans’ sense-making around sexual consent in order to develop a conceptual framework that can guide future methodological and conceptual work. We analyzed representations of sexual consent in a sample of 291 creative narratives about HIV written for a scriptwriting competition by young Nigerians, Kenyan and Swazis in 2005, 2008 and 2014. We combined thematic data analysis and narrative-based approaches. Narratives represented consent as a feeling of wanting or being willing to have sex, or an intention to have sex, communicated via character actions, conversations or circumstances. Some narratives depicted characters not wanting but consenting to sex to avoid negative repercussions. Representations of sexual consent were fairly consistent across contexts and over time, although certain representations were more prominent in some country/year samples than others. Results are translated into a conceptual framework that can guide future prevention efforts to reframe sexual consent.

Introduction

Recent research demonstrates that sexual violence perpetrated against adolescents and children is a significant public health and human rights concern (Sumner et al., 2015) and an important driver of the HIV epidemic (Andersson et al., 2008). While sexual violence is manifested through actions, how individuals and societies define and respond to sexual violence is influenced by the social construction of sexual violence. When viewed through a social constructionist lens, no single, uncontested or universal definition exists for violence, only the ongoing struggles to define violence and frame the way we think about violence (Richardson & May, 1999). Narrow definitions of sexual violence frame sexual violence as perpetrated and experienced by a few deviant individuals, whereas broader definitions draw attention to the widespread prevalence of sexual violence, thereby requiring a challenge to the status quo (Peterson & Muehlenhard, 2004).

Historically, sexual violence has been limited in definition to situations involving penile-vaginal penetration and physical resistance by the victim (Muehlenhard & Kimes, 1999). More recently, feminist scholars and human rights activists emphasize sexual consent as a defining feature of sexual violence (Weiner, 2013), thereby widening the definition to include all non-consensual acts. However, like sexual violence, sexual consent is an ambiguous and complex concept for which multiple competing definitions exist. For example, in the United States, universities are increasingly adopting “affirmative consent” policies that conceptualize consent as “affirmative, conscious and voluntary agreement to engage in sexual activity” (California Senate Bill SB-967, 2014). These policies promote a definition of sexual consent as an ongoing process of evaluating a partner’s behavior and continuously obtaining explicit, verbal consent for each sexual activity rather than a discrete event that occurs at the beginning of a sexual encounter.

While many young people are aware of this definition of sexual consent, emic definitions often depend on notions of consent as “an internal state of willingness” (Muehlenhard et al., 2016, p. 462) that rely on shared understandings of indicators of willingness. For example, rather than verbally communicating consent, lay understandings imply that consent can be communicated by reciprocating hugs, kisses, or caresses or by general passive behaviors such as not resisting a partner’s advances or not saying no (Burkett & Hamilton, 2012). These forms of consent depend on another’s observation and interpretation of behavior and presume that all parties share similar expectations for communicating consent.

Sexual consent in sub-Saharan Africa

Sexual consent has been most extensively studied in U.S. and developed Anglophone contexts; research on sexual consent in sub-Saharan Africa (SSA) is much less common. Amongst the ethnographic and qualitative studies that have been carried out in SSA, findings show that, in line with norms that promote male sexual assertiveness and female passivity in sexual encounters, explicitly discussing sexual needs and preferences violates norms of female passivity (Burnet, 2012; Tavrow et al., 2013; Wood et al., 2007). These researchers found that their participants did not draw from explicit and verbal models of sexual consent, but instead understood willingness to have sex as communicated via behaviors upstream of the sexual encounter. Behaviors included agreeing to enter into a dating relationship (Wood et al., 2007) or a marriage (Stern & Heise, 2019), accepting gifts, money or other types of resources such as food or drink (Barnett et al., 2011), or being alone with another person (Burnet, 2012; Wood et al., 2007).

Within these models, sexual consent was a discrete event and agency to consent, particularly for young women and girls, became subsumed under male sexual entitlement following these behaviors or circumstances. While there has been considerable critique of this model of sexual consent amongst communities working to prevent sexual violence, efforts to put forth definitions of sexual consent as explicit and sequential have met with resistance. For example, Judith Singleton (2012) described how passage of South Africa’s 2007 Sexual Offences Act, which defined rape as sex without consent, was met with resistance and confusion as lay definitions conceptualized rape as occurring between strangers or perpetrated in families against children. In her ethnographic fieldwork with young people in a KwaZulu Natal province township, she observed how some participants rejected the notion of consent within relationships, thereby leaving invisible any sexual violence between acquaintances or romantic partners. J. Singleton argued that the collision between human rights rhetoric embedded in the Sexual Offenses Act and local understandings of consent made it much more difficult for young women and men to act upon the strengthened legal recourse the Act offered.

Purpose of this study

Given the burden of sexual violence in SSA and its role in the HIV epidemic (Sommarin et al., 2014; Sumner et al., 2015), there is a need for greater insight into emic understandings of sexual consent in order to inform communication and education efforts that do not elicit backlash as observed in South Africa. While existing studies have shed light on this topic, conceptual work has been constrained to single sites and populations, thereby limiting their impact.

To address this gap, we analyzed young Africans’ narrative-based representations of sexual consent in three diverse countries (Nigeria, Kenya and Eswatini, formerly Swaziland) across 10 years (2005–2014). Narratives shape our memories, knowledge and beliefs; they allow people to formulate and articulate the causes and consequences of human action and, as such, underlie social knowledge (Bruner, 1990). Creative narratives offer researchers the opportunity to explore normative issues in a way that approximates to the complexities with which they are surrounded in reality and thus provide access to richly contextualized representations (Winskell et al., 2013). In our study, comparisons of the representations of sexual consent in these narratives were analyzed across contexts (Nigeria, Kenya and Eswatini) and over time (2005–2014) in order to understand the ways in which representations are shared across space and time, and the ways in which they differ or evolve. Given that sexual violence is disproportionately perpetrated against girls and women, and how men and women often experience different levels of power, pressure and stigma in sexual situations, we restricted our analysis to representations of sexual consent by female characters. The ultimate goal of this analysis was to create a conceptual framework of sexual consent that can be used to guide methodological efforts in the future, as well as to inform sexual violence prevention efforts that seek to influence the social and symbolic contexts of risk for youth in these settings.

Method

Study sample

The research described in this paper is part of an ongoing study of representations of HIV and AIDS through creative narratives written by young Africans for scriptwriting competitions (Winskell et al., 2018). The competitions were coordinated internationally by the non-profit organization Global Dialogues (www.globaldialogues.org) as part of a multi-level participatory communication process in which young people were invited to contribute an idea for a short film about HIV. The current analysis pertains to creative narratives by young Nigerians, Kenyans and Swazis contributed in 2005, 2008 and 2014. These countries were selected based on the inclusion criteria for the larger study (Winskell et al., 2018), the varying HIV prevalence (amongst 15–49-year-olds, 27% for Eswatini, 4.5% Kenya, 1.3% Nigeria) (UNAIDS, 2020), and emerging evidence showing the burden of sexual violence amongst young people in these countries (Chiang et al., 2016).

All narratives included for analysis were contributed in English. Narratives were transcribed verbatim (complete with spelling and grammatical errors). We stratified our data by author gender, urban/rural location and age (10–14, 15–19, 20–24) and randomly selected 10 narratives from each of the 12 strata, oversampling urban/rural location if necessary to increase the likelihood that a total of 20 stories were selected for each age/gender stratum. In some countries, certain age/gender strata still contained fewer than 20 narratives; hence most country samples in 2005, 2008 and 2014 have fewer than the maximum 120 narratives (Table 1). Our sampling procedures are described in detail in Winskell et al. (2018). An overall sample of 692 texts for the three countries resulted from this sampling strategy and were subsequently analyzed.

Table 1.

Sample distribution

	Year			Totals
	2005	2008	2014
Nigeria	n = 120	n = 93	n = 88	n = 301
Kenya	n = 88	n = 25	n = 116	n = 229
Eswatini	n = 72	n = 50	n = 40	n = 162
Totals	n = 280	n = 168	n = 244	n = 692

Data processing and analysis

Our methodological approach is situated at the intersection of grounded theory (Corbin & Strauss, 2008) and thematic narrative analysis (Riessman, 2008). We used two intersecting approaches to analyze the narratives: qualitative data analysis, focusing on thematically-related text segments and memoing for emergent themes; and a narrative-based approach, focusing on plot summary and thematic keywords, as described below.

1. Qualitative data analysis: Narratives were transcribed and entered verbatim into MAXQDA 18 qualitative data analysis software (VERBI Software, 1989–2016), where they were labeled with reference to a codebook of 54 codes, including codes such as “sex description” and “chat up/proposition.”

2. Narrative-based approach: A one-paragraph summary was written for each narrative comprising key elements of plot and message. Summaries were independently double-coded with up to six out of 44 keywords, which included keywords such as “Sexual violence/coercion/child sex abuse,” “Transaction/gift-giving,” and “Abstinence.” Discrepancies in coding were resolved through dialogue.

Identifying narratives with representations of sexual consent

After applying these two strategies, we set about identifying narratives that included representations of sexual consent (including the absence of consent). We understood sexual consent as willingness to have sex, whether depicted or understood as an internal state (e.g., thoughts or feelings that imply the character is willing or unwilling to have sex) or communicated via particular actions that a character takes that can be understood (by the reader, another character or broader society) to imply willingness. Given our interest in exploring narrative representations of sexual consent and taking into consideration how nuanced and complex sense-making around sexual consent has been shown to be in other studies, we began by identifying all narratives that depicted a sexual scenario or conversations about sex. To accomplish this, we identified all narratives qualitatively coded with relevant codes such as “sex description,” “partner pressure,” “chat up/proposition,” “sexual violence/coercion/rape,” or “gift-giving/Transactional sex.” We further reviewed narratives that did not have the aforementioned qualitative codes but had a relevant keyword, such as “Within couple dialogue” or “Abstinence.”

Though rarely explicitly stated, the overwhelming majority of narratives implied sex to be penile-vaginal penetration, with only four narratives (all Nigerian) representing sex between same-sex partners, three with female same-sex partners. Given the stigma associated with same-sex attraction in many regions of SSA that could influence representations of same-sex relationships, we restricted analysis to narratives depicting heterosexual relations. Though we originally intended to analyze all narratives that depicted sexual relations (both marital and pre-marital), upon review, it became clear that expectations of sex within marriage are so pervasive that very few narratives engaged with the concept of sexual consent in marital relationships. We therefore restricted analysis to narratives with representations of sexual consent in non-marital relationships. Lastly, representations of sexual encounters generally depicted one character making a sexual request or demand in some way, whether by verbal request, seduction, invitation to be alone, force, or any number of additional strategies. Narratives depicted male characters primarily instigating sex, though a small portion of narratives depicted female characters (n = 38) inviting or instigating sex in some way (e.g., via seduction or getting a partner drunk). Those who instigated a sexual encounter or solicited consent from others were depicted as having implied their consent by virtue of their efforts to obtain sex from other characters. Narratives did not represent these characters as providing additional indicators of consent and we thus chose to exclude narratives that depicted female characters instigating sex as these narratives did not provide additional insight into sexual consent communicated by female characters.

After reviewing narratives identified via these methods, we created a sample comprised of the 291 narratives (out of 692) that provided insight into the young authors’ representations of sexual consent by female characters. Within this smaller sample (Table 2), we developed thematic codes specific to the topic of sexual consent (“fine codes”) using the same process described above. Examples of fine codes include “accepting gifts/money” and “responding to touches, kisses, etc.” We quantified codes to identify patterns across the country and time point samples, and by author demographics. This also allowed us to identify hypotheses to further explore via thematic analysis (proportions reported in Results, such as one in five narratives or 20% of narratives, refer to proportions within the sample of narratives depicting sexual consent; 291 is the denominator).

Table 2.

Distribution of narratives analyzed for representations of sexual consent

Country	Year			Author demographics*							Total
	2005	2008	2014	Male authors (%)	Female authors (%)	Rural authors (%)	Urban authors (%)	10–14 (%)	15–19 (%)	20–24 (%)
Nigeria	51	34	43	50 (39%)	78 (61%)	65 (51%)	63 (49%)	48 (38%)	54 (42%)	26 (20%)	128
Kenya	44	8	42	45 (48%)	49 (52%)	41 (44%)	53 (56%)	19 (20%)	38 (40%)	37 (39%)	94
Eswatini	30	14	25	20 (29%)	49 (71%)	46 (67%)	23 (33%)	40 (58%)	26 (38%)	3 (4%)	69
Totals	125	56	110	115 (40%)	176 (60%)	152 (52%)	139 (48%)	107 (37%)	118 (41%)	66 (23%)	291

^*Percentages may not add up to 100 due to rounding

This study, comprising the secondary analysis of existing data, was approved by Emory University Institutional Review Board. We cite the narratives verbatim, with the exception that characters’ names have been changed. Country names are abbreviated as follows: NG – Nigeria; KY- Kenya; and SW – Eswatini. Excerpts are identified by the country, contest year, sex, age, and geographic location, urban (U) or rural (R), of the author.

Results

As stated, representations of consensual and nonconsensual sexual encounters were found in 291 narratives, or approximately two out of five of the original 692 narratives (Table 2). These narratives were contributed by approximately equal proportions of rural and urban authors (52% and 48%, respectively), and by more female authors (60% of the 291 narratives). The average age of authors was 16.4 years, though this varied by country (Nigeria average age = 16.2, Eswatini = 14.5, and Kenya = 17.9). We observed some minor differences in representations of sexual consent across author genders and age ranges, which are described in the following sections; no meaningful differences were observed between rural and urban authors.

Understandings of sexual consent as indicated by female characters

The terms “sexual consent” or “consent” were not found in any narrative. Instead, representations of a character’s level of willingness or unwillingness to have sex were indicated via plot (i.e., circumstances, actions taken, or conversations had), in descriptions of characters’ intentions, and/or in descriptions of how characters feel during and/or after a sexual encounter (e.g., excited, happy, violated, angry). For example, Alice accepted gifts from a traveling trader and accepted to meet him in his house. “Alice was excited about the visit. While in the house the trader asks Alice to have sex with him. Alice accepted” (KY 2005, M 10–14 U). In contrast, Nobuhle was assaulted by her cousin Vusie after rejecting his sexual advances. He came to her bed at night, bound and blindfolded her and threatened to kill her if she did not keep quiet. Nobuhle “prayed and fought till the last drop of strength lacking, biting but in vain.” After Vusie “rapes her” and takes her virginity, “she felt pain spiritually and physically” (SW 2008, F 15–19 R).

This narrative depicted Nobuhle and Vusie’s encounter as nonconsensual via her physical resistance, deep pain afterwards, and by labeling the encounter “rape.” Other narratives were more ambiguous in their depictions of whether or not a sexual encounter was implied to be consensual. For example, Chika was tricked into attending a party where her sister brought her with the intention of Chika having sex with a male acquaintance, Chidi. Chika “got drunk and lost control” to the point of not knowing “what was happening to her.” Chidi took advantage to have sex with her. “The next morning Chika woke up and felt pains, she looked and found out that they were both naked and wept bitterly…She was so angry” (NG 2014, F 15–19 U). This narrative depicted Chika as unaware and without faculties, not intending to have sex, and then feeling angry and bitter after the fact, all of which imply she was not willing to have sex with Chidi. However, this encounter was not labeled “rape” as Nobuhle’s was but was pointed to as evidence of the type of “mistakes” that young women get into when attending parties and drinking alcohol. Chika got HIV from the encounter and commited suicide. The overarching message of the narrative implied that behaviors such as attending parties and consuming alcohol place young women in situations that can become sexual, and therefore indicate, if not willingness to have sex, blame for sex occurring.

Through our analysis, we identified patterns of circumstances and behaviors that are depicted to imply willingness to engage in sex in the narratives. These circumstances and behaviors can be somewhat distal from the sexual act (e.g., willingness to be alone with someone in their house) or can occur immediately prior to the sexual act (e.g., responding to kisses or caresses).

Distal indicators of willingness or future sexual reciprocity: Consent to the possibility of sex

Distal indicators of willingness or future sexual reciprocity included (1) willingness to be alone with someone, (2) acceptance of gifts or money, or (3) willingness to enter into a dating or marriage relationship with someone. These indicators may have occurred separately or formed part of a sequence of actions that, taken together, implied an openness to sex and sexual propositions.

1. Willingness to be alone with someone was one of the most common indicators of sexual consent (occurring in one in three of the 291 narratives), followed by accepting gifts or money (one in five narratives). Willingness to be alone was most often depicted as a male character inviting a female character to his house, a hotel, or some other private location. The female character indicates her consent by either agreeing to meet him or to move with him to a more private locale (such as moving from a party or bar to someone’s house): “One day the boy told the girl Jackie today I would like you to come and visit me at home. Jackie never wasted time and told Walter that it’s okay… They went and spend the day there and Walter did the necessary to her they slept together…” (KY 2005, F 15–19 U). Narratives represented willingness to be alone with someone as an indication of willingness to be in a situation that could become sexual.

2. Accepting gifts or money (hereafter “gifts”) is generally understood across SSA to imply consent to future sexual reciprocity (Stoebenau et al., 2016). Narratives conveyed this awareness via character interactions and/or negotiations or via narrator commentary.

   He accosted her, made promises to increase her salary, buy new clothes for her as well as take her to school. Irene was pleased which made it possible for him to lure her into his master-bedroom. He made love to her

   In some narratives, accepting gifts reflected one step in sequential consent, whereby accepting gifts was followed by being alone with a potential partner and additional indicators of consent were given. In others, accepting gifts was the only indicator of consent.

   Most narratives in this sample implied a general understanding that accepting gifts indicated willingness for future sexual reciprocity, at least amongst narrators and gift givers (the majority of whom were male characters). However, some narratives (approximately one in fourteen) depicted young naïve female characters accepting gifts or money without understanding the typically older male character’s assumptions of future sexual reciprocity. Accepting gifts without fully grasping the implications was not represented as implying the naïve female character was willing to have sex, although male gift givers interpreted it as such. In narratives where a female character had accepted gifts and later refused to consent to sex, male characters either pointed to past gift acceptance as evidence of reneging on an implied deal or used physical force to obtain sex. Female authors contributed the majority (three out of four) of these narratives.

3. Narratives depicted willingness to enter into a romantic relationship as an indicator of future sexual consent. Asking someone to be your girlfriend reflected one step in a sequence of acts that lead to sex. A common sequence of events found in the sample begins with a male character offering gifts or requesting to enter into a relationship; at times these actions occurred at the same time. Once gifts have been given and a relationship has been formalized, one partner invites another to a private locale such as their home or the couple strategizes how to be alone. For example, Obinna asked Ruby to be his girlfriend, to which she agreed. He then bought her a cell phone and called her to meet him at night. She snuck out of her room to meet him and they had sex (SW 2014, F 10–14 R). Similar to willingness indicated via gift acceptance, agreement to any of these options was implied to represent an initial willingness to have sex, although not all characters were represented as understanding this implication.

Proximal indicators of willingness: Consent at the moment of sex

The distal indicators of willingness to have sex in the future combine to facilitate the opportunity for sex to occur. In some narratives, these distal indicators served as sexual consent and no further behaviors or communication took place (for example, in narratives where a gift or an invitation to be someone’s girlfriend was accepted and the couple had sex). These representations were most often found in narratives written by younger authors and many times did not include much detail surrounding the sexual encounter. In many other narratives, however, additional indicators of willingness (or unwillingness) were communicated at the beginning of a sexual encounter. These indicators included physical and verbal responses to another’s implied or explicit request for sex.

Physical responses that indicated sexual consent included responding to kisses and caresses, whereas moving away from someone, pushing them away or more forceful types of physical resistance (e.g., face slapping) communicated unwillingness to have sex. Some narratives depicted these physical indicators of unwillingness as forms of “token resistance” (Muehlenhard & Hollabaugh, 1988), in which the female character initially refused sex out of ambivalence or a desire to protect her reputation but later consented to sex. For example, Brenda initially pushed Vusie away when he attempted to kiss her because she felt conflicted about having sex, only to later initiate sex (NG 2014, F 15–19 U). In other narratives, physical resistance indicated genuine non-consent as a character attempted to avoid sexual assault.

Verbal responses included the explicit, verbal agreements to have or not have sex, as well as the more indirect indicators of unwillingness, such as making excuses. Representations of explicit verbal agreements were found slightly more often in narratives written by younger authors (10% of narratives by 10–14 year-old authors, 8% of 15–19 year-old-authors and 6% of 20–24 year-old authors). Indirect verbal indicators such as making excuses were, like indirect physical indicators, represented in some narratives as “token resistance” that could be overcome by persistence and convincing, and genuine unwillingness in other narratives. As such, indirect indicators (both physical and verbal) could be interpreted as implying either internal willingness or unwillingness.

Issues of wanting versus consenting to sex: agency and power

A fundamental condition in etic definitions of sexual consent is agency – i.e., the freedom and capacity to express one’s willingness or unwillingness to engage in sexual acts and under what conditions (Buchhandler-Raphael, 2010). Issues of power, particularly gendered power within relationships and power over economic resources, complicated representations of consensual sex in the narratives. Many of the narratives depicted scenarios in which a female character was described as not wanting to have sex, or feeling ambivalent about sex, but showing consent in order to avoid repercussions.

Power in romantic relationships

Pressure to have sex from one’s romantic partner was represented in one in ten narratives. In roughly one-third of these narratives, this pressure was linked with gift-giving and the sense of obligation that gifts implied. Past gifts were pointed to as a reminder, at times gently and in a flirty fashion (“if you feel you love me and you owe me anything then you can repay me, you know what they say girl, actions speak louder than words” SW 2008, F 15–19 R). Other times, the reminder was more forceful and explicit. While some narratives depicted female characters wanting sex, other narratives depicted girlfriends acceding to pressure in order to avoid angering partners and thereby lose access to gifts and resources. In these narratives, female characters were represented as ultimately willing to have sex though not necessarily wanting sex.

The remaining narratives represented partner pressure without gifts, in which male characters pressured their girlfriends to have sex until the girlfriends were worn down and acceded out of fear of the relationship ending. Ambivalence often characterized these interactions. Some narratives represented female characters resisting pressure while simultaneously feeling interest in the experience of sex with a boyfriend. For example, “Michael is disturbing Jessica to please have sex with him if she love him…She doesn’t fill ready and her mother has warned her towards this…she loves Michael and she is also feeling physically. She doesn’t know what to do” (NG 2008, F 15–19 R). In these narratives, male characters wielded the power to leave their girlfriends unless their demands for sex were met. Female characters ultimately agreed to have sex (albeit ambivalently) or allowed themselves to be seduced out of fear of being abandoned, at times combined with sexual curiosity.

Power over livelihood and/or advancement

In one in eleven narratives, sex was obtained outside of a dating relationship by threatening a female character’s livelihood or potential to advance. In these narratives, male characters were employers, teachers or leaders within the community, and the female characters were either dependent on them for their livelihood (as in the case of live-in house helps) or advancement (e.g., students). Female authors contributed two-thirds of these narratives.

In some narratives, female characters either sought out or jumped at the opportunity to exchange sex for money, gifts or advancement. For example, after failing a girl with the intention of sexually blackmailing her, a lecturer “told the girl how he had been lusting for her, the girl was so glad that she didn’t hesitate, so they had a quick sex there” (NG 2005, F 15–19 U). Although refusing to have sex would bring about negative consequences, engaging in sex led to desirable outcomes. These female characters were not depicted as feeling coerced or ambivalent, but instead as willing to have sex in order to achieve personal goals. These representations were found in all countries but were most common in Nigerian narratives. Other narratives represented female characters initially refusing or resisting such advances, only to ultimately submit to sex in order to avoid or mitigate negative repercussions from the male character.

Positive deviants

A small portion of narratives depicted female characters who were similarly disempowered by poverty and partner pressure reinforced by gender and age hierarchies but were able to exercise agency and avoid unwanted sex (i.e., positive deviants). A handful of narratives (n = 7) represented female characters rejecting gifts as a way of not consenting to sex. These representations were found in all countries and time points. In these representations, female characters were aware of the implications associated with gift acceptance and rejected gifts as a way of indicating their unwillingness to have sex in the future. In some narratives, rejecting gifts sufficed as an indirect way of indicating non-consent and no further conversation took place. In others, male characters continued to persist and female characters used more forceful, direct methods of communicating non-consent, such as pushing partners away or telling them directly that they were unwilling.

Approximately one in twelve narratives depicted consent relying on HIV-prevention-related conditions being met, such as getting an HIV test, using a condom, or waiting until marriage. These narratives represented couples discussing the conditions under which one or both would be willing to have sex. In almost all of these narratives, the female character requested or demanded the conditions be met. Thus, consent was negotiated verbally until conditions were met and consensual sex could take place.

Cross-national observations

Although the three country samples represented geographically and culturally distinct populations, representations of consent across the countries had more similarities than differences. Representations of proximal and distal indicators and representations of power were found in all country samples. However, certain representations were more prominent in some countries.

Nigeria

The Nigerian sample included the highest proportion of narratives depicting female characters indicating willingness (versus unwillingness). Approximately twice as many Nigerian narratives depicted female characters going somewhere as a way of indicating consent (e.g., agreeing to leave a bar to go to someone’s room) when compared with Kenyan and Swazi narratives. Similarly, Nigerian narratives often included the most detail surrounding sexual encounters, and therefore this sample had the highest proportion of narratives containing proximal indicators of consent, particularly nonverbal indicators (for example, one in ten Nigerian narratives included depictions of responding to kisses or caresses as indicating sexual willingness, compared with one in twenty Kenyan and Swazi narratives).

Importantly, the majority of these narratives carried take-home messages that criticized female characters who had non-marital sex. The high proportion of representations of female characters indicating willingness to have sex were embedded in narratives that, via narrator commentary and character descriptions and dialogue, overtly criticized non-marital sex in general and blamed female characters for engaging in any behaviors that – however distal from the sexual encounter – could be interpreted to imply consent to future sex. In their most extreme, some narratives described “disobedient” (NG 2005, F 10–14 R) or “wayward” (NG 2014, F 15–19 U) female characters who entered (or were forced to enter) spaces where they were alone with male characters and were subsequently raped; the take-home message in these narratives was female characters were at least partially to blame for their rapes. Though extreme, these narratives exemplified how notions of blame and responsibility justify nonconsensual sex against female characters who behaved in ways that did not necessarily indicate willingness, but which facilitated an opportunity for sex to occur. Narratives with this overarching message were found in all country samples but were most prominent in the Nigerian sample.

Eswatini

Narratives from the Eswatini sample included the highest proportion of representations of consent obtained via partner pressure (comprising one in five narratives in this sample, compared with one in fifteen in the Nigerian and Kenyan samples). Partner pressure in Swazi narratives reflected that described above, where male partners pointed to past gifts as evidence of an implied sexual contract and/or leveraged their love to obtain sex from reluctant girlfriends: “Joe persuaded Thandi to have sex with him. ‘I am a virgin’, Thandi said with tears streaming down her cheeks. ‘If you really love me, you will have sex with me’, said Joe demandingly. Because she loved him, she agreed” (SW 2014, F 15–19 R).

Unlike the Nigerian narratives described above, many of these representations were empathetic in tone towards the female character, acknowledging her vulnerability to manipulation and pressure by a partner who points to distal indicators of consent as evidence of implied consent, and which overrule the indicators closer to the sexual encounter (such as, in Thandi’s case, pointing to her virginity as a reason to avoid sex). Male partners were characterized as motivated predominately by sex, rather than genuine affection for the female character.

Kenya

The Kenyan sample included the highest proportion of narratives depicting alcohol or other substances that facilitate sex (15%, compared with 5% and 6% in Nigerian and Swazi samples, respectively). The majority of these representations depicted one character either giving a female character alcohol in order to inebriate them, thus facilitating sex, or taking advantage of a character’s state of inebriation to have sex with them. Proximal indicators of consent were uncommon in these representations. Some narratives did not depict consent in any form and instead described alcohol contributing to a context in which sex happens. Other narratives described a female character either being given or consuming alcohol of their own volition, and a male character capitalizing on their state of inebriation to have sex with her:

she drunk and drunk like a abnormal woman until she fell asleep on the chair…As the woman was in a sleep she was now not hearing anything going on in the house, after a short while the man laid on her body and have sex with her

(KY 2014, F 10–14 R).

Narrative commentary and character thoughts and dialogues did not imply that alcohol consumption necessarily indicated willingness to have sex per se, in the way that entering into a relationship or accepting gifts was often depicted as mutually understood by partners to imply future sex. However, female characters who drank to excess were depicted as responsible for any outcomes that occurred from inebriated sex. For example, the woman described in the narrative above discovered blood when she comes to, rushed to the hospital and was diagnosed with HIV. Her children blamed her for making them orphans. No narrative criticized a male character for having sex with an incapacitated female character.

Temporal observations

While there were no meaningful temporal shifts in representations of sexual consent across all three countries between 2005 and 2014, there were several notable shifts in specific countries. In the Nigerian sample, representations of consent as indicated by distal indicators, specifically accepting gifts or being alone with someone, decreased over time while representations of proximal indicators like verbal consent and responding to kisses increased over time. In addition, representations of verbal consent to have sex in Nigerian narratives were all within transactional encounters in 2005 and 2008. However, in 2014, a handful of narratives depicted couples – boyfriends and girlfriends – discussing sex and engaging in verbal consent before having sex. These representations were found in several Kenyan and Swazi narratives across all time points, but were only observed in 2014 narratives in the Nigerian sample.

In the Swazi sample, representations of consent obtained via partner pressure remained constant, while consent relying on conditions being met in which female characters exercised agency to demand protection from HIV or pregnancy decreased over time. No meaningful temporal changes were observed in the Kenyan sample.

Discussion

The purpose of this analysis was to contribute to understandings of young Africans’ narrative representations of sexual consent that reflect their emic understandings of this concept. We created a conceptual framework to represent this emic understanding of youth populations from these three countries (Figure 1). Narratives in this sample represented sexual consent as either a process, in which willingness was indicated via a combination of distal (e.g., accepting to be alone with someone) and proximal (returning kisses) indicators, or as a one-time event (e.g., by accepting gifts). Consent was rarely represented as explicit and verbal; indirect indicators were more common. Representations of sexual consent were fairly consistent across contexts, although certain representations were more prominent in some country/year samples than others.

Figure 1.

Conceptual framework of emic understandings of sexual consent

Importantly, taken together representations of consent in these narratives pointed to cultural scripts that manifest the social assumptions and expectations around sexuality and sexual behavior (Simon & Gagnon, 1986). Different from interpersonal and intra-psychic scripts, which operate as the individual and interpersonal notions about expected sexual behaviors and their meaning, cultural scripts “essentially instruct the narrative requirements of specific roles” (Simon & Gagnon, 1986, p. 98) and thus, at the collective level, may prescribe sexual expectations for a given encounter based on culturally specific norms and values. In this way, cultural scripts may serve to either constrain or foster agency in sexual encounters. The narratives reflected cultural scripts that represented consent taking place within contexts influenced by broader sociocultural hierarchies such as gender and access to and control over resources, and the authority and power these factors afford. Assumptions of male sexual entitlement and female sexual passivity permeated this sample and pointed to broader cultural scripts that most often denied or constrained agency by female characters.

Existing studies on sexual consent in various settings throughout SSA echo findings found in our narratives, specifically that sexual consent is presumed upon entrance into a relationship (Jewkes & Morrell, 2012; Stern & Heise, 2019), by accepting gifts (Barnett et al., 2011), or by being alone with someone (Burnet, 2012; Wood et al., 2007), i.e. the distal indicators of sexual willingness. In these studies, proximal indicators of consent, such as saying “yes” or “no”, responding to kisses or pulling away, were argued to either be of lesser importance in sexual negotiations or be perceived to be part of a typical sexual script.

Other studies also cite additional distal factors of consent. For example, in their ethnographic study in a township of South Africa, Wood et al. (2007) found that having had prior sexual relations with a partner implied consent to future sex. Their participants understood consent as a one-time process between partners that could be presumed to extend to future sexual encounters. As well, Wood et al. observed that some male participants interpreted sexual consent from women by virtue of being at a party or bar and/or consuming alcohol, an observation found in studies from other settings in SSA (e.g., Tavrow et al., 2013).

In the present study, behavioral indicators of sexual consent, broadly categorized as distal and proximal, are represented in Figure 1. This figure represents indicators as behavioral – rather than relational or circumstantial – in order to aid future methodological work seeking to capture beliefs and norms around sexual consent and interventions promoting affirmative models of sexual consent. As shown, behaviors are represented along a continuum of willingness and unwillingness indicators. In general, agreeing to be a romantic partner and/or accepting gifts indicated willingness for future sex. Agreeing to be alone with someone or going to an isolated locale indicated willingness for future or immediate sex, and saying “yes” to a sexual request or responding to kisses or caresses indicated willingness for immediate sex. In our analysis, consuming alcohol did not necessarily indicate willingness but rather provided sexual opportunity due to decreased inhibitions or ability to react. Saying “no” or pulling away from kisses or caresses signified unwillingness in some narratives and ambivalence in others. Physically resisting, refusing to be alone with someone and/or rejecting gifts all indicated genuine unwillingness. Lastly, drawing from other studies, we included “had prior sex” and “go to party, bar, etc.” as indicators of willingness.

Importantly, this figure represents emic understandings of a continuum of behaviors that can be used to label a sexual encounter as consensual or nonconsensual and that point to broader cultural scripts about sex. Many narratives from this sample carried implicit and explicit messages about responsibility and blame for nonconsensual encounters that disproportionately tasked female characters with controlling aggressive male sexuality. These representations reflected rape myths, or “attitudes and beliefs that are generally false but widely and persistently held, and that serve to deny and justify male sexual aggression against women” (Payne et al., 1999, p. 134). Rape myths define rape narrowly, creating conceptual boundaries around what is “legitimately” perceived as rape (most often restricted to penile-vaginal penetration that is physically resisted by the victim), defining who is and is not likely to be a victim or perpetrator, where rapes do and do not occur, and what type of non-consent (e.g., physical resistance) needs to be exhibited for an encounter to qualify as rape (Peterson & Muehlenhard, 2004). For example, Wood et al. (2007) observed how their South African participants reconstructed their personal experiences and behaviors in a way that allowed young men to distance themselves and their own coercive strategies from the label “rape.” Nonconsensual experiences became reconfigured as consensual by pointing to relationship expectations, circumstances and their female partners’ behaviors that, these young men argued, indicated willingness to have sex. Some young women in Woods et al.’s South African sample similarly restructured their experiences to avoid labeling them as coercive or violence, thereby avoiding the label of “rape victim,” which carried connotations of shame, “dirtiness” and/or “damage” (p. 295). Imposing these insights onto our data, it is possible to identify behaviors, particularly distal behaviors like accepting gifts, being alone with someone or accepting a relationship, which some narratives depicted as justification for encounters implicitly represented as nonconsensual but not labeled “rape.”

Uncertainty, ambivalence and sexual consent

Many of the narratives represented female characters experiencing uncertainty or ambivalence around their decision to have or not have sex. Uncertainty and ambivalence have been found to permeate sexual encounters in other populations (Beres et al., 2014), and for this reason, sexual consent theorists argue that it is important to differentiate between wanting and consenting to sex, with consent being essential (Peterson & Muehlenhard, 2007).

Wood et al. (2007) and Barnett et al. (2011) both drew attention to the ways in which girls and young women in SSA (and beyond) avoid outright rejection of male propositioning, either out of respect due to social hierarchies and gender grooming or to avoid violent reprisals. As such, “feminine sidestepping and ambiguous responses could indicate either affirmation or rejection but neither in an absolute sense” (Wood et al., 2007, p. 286). These norms create a social environment in which initial “token resistance” is expected. In some narratives, a “no” did not necessarily reflect a genuine state of internal unwillingness but rather could be turned into a “yes” with sufficient persistence. These representations are perhaps closer to reality in SSA (and beyond) than the explicit, verbal model of sexual consent put forth in sexual violence prevention efforts in Western countries (Beres et al., 2014). However, they also contribute to an environment that can and does justify sexual assault. Addressing the negative consequences of “token resistance” requires multi-pronged efforts to address the very real threat of violence to girls and women when rejecting some male sexual offers, while simultaneously addressing sexual norms that stigmatize girls and women expressing sexual desire.

Framing consent in sexual violence prevention

Cross-national and temporal implications

Our results show representations of sexual consent that were shared across all three countries and over time. As with gendered representations (Lloyd & Duveen, 1992) and rape myths (Singleton et al., 2018), these representations appear to be hegemonic (Moscovici, 1988) given their relative uniformity and linkages with hierarchy and power. However, country and year samples reflected varying prominence of certain representations that pointed to opportunities for framing sexual consent in a manner that contributes to sexual violence prevention efforts.

Swazi narratives consistently represented more partner pressure to obtain sexual consent than other countries. These representations reflected an acknowledgement of the ways in which gendered and economic power can reduce individual-level agency to consent to sex. In part, this reflects efforts over the past 20 years to bring attention to the underlying social drivers of the feminization of HIV in a country with an extremely high HIV prevalence, drivers that include gender-based coercion, violence, and hierarchy within relationships. Eswatini has organizations that have been promoting awareness around sexual violence and coercion since the 1990s (Gardsbane & Hlatshwayo, 2012), including amongst in-school youth (Manzini-Henwood et al., 2015). While efforts like these also take place in Kenya and Nigeria, it is possible that by virtue of the size of Eswatini and gravity of its HIV epidemic, young Swazis have received a higher “dose” of awareness-raising activities that – when combined with the very real risk of HIV transmission – creates a strong sense of urgency around the topic of partner pressure. Disappointingly, while some 2005 Swazi narratives depicted young women creatively navigating partner pressure to demand that conditions for consent be met, these narratives disappeared in later years. Representations of desirable partners modeling this behavior and of female partners being willing to walk away – as depicted in the 2005 narratives – may be useful to provide young Swazis with aspirational role models to mimic in their own relationships.

Nigerian narratives in 2005 and 2008 depicted direct, verbal consent only within transactional sexual encounters; however, in 2014, verbal consent was also represented in dating relationships in which couples discussed sex and verbally consented prior to initiating sexual activity. While these representations were only found in a small portion of narratives, they nonetheless point to an opportunity to align models of sexual consent with companionate relationship models that link romantic love and joint decision-making with relationship satisfaction. Ethnographic studies in Nigeria have shown how companionate relationships, premised on modernity-linked notions of gender equality and class attainment, do not always produce genuine joint-decision making, particularly vis-à-vis sexual encounters; however, these relationships are still perceived to be highly desirable by virtue of their linkage with social mobility (Smith, 2007). This desirability provides an opportunity to frame sexual consent – particularly models of sexual consent that include dialogue within couples and proximal indicators – as an essential component of companionate relationships in Nigeria and beyond.

Kenyan narratives depicted the role of alcohol in sexual situations more than other countries. Some of these narratives depicted situations that mirror scenarios that would be labeled “date rape” in the United States, or sexual encounters in which one party has lost control of their faculties due to substance consumption such that they are unable to give genuine consent. Due to concerns over high prevalence of date rape, particularly amongst young people, Western sexual consent theorists underscore the capacity dimension of sexual consent (Cowan, 2008; Lyden, 2007) – which, in this case, refers to the cognitive capacity to understand a sexual situation and indicate willingness or unwillingness. These narratives linked alcohol consumption with sex and therefore placed the responsibility to avoid nonconsensual sex on the person who consumed alcohol rather than the person who had sex with an incapacitated individual. Our findings point to a need to promote a model of consent that acknowledges capacity to consent in all its forms, including cognitive capacity to consent following consumption of substances.

Lessons from the field

Lessons can be drawn from efforts outside of the countries of study. For example, Stern and Heise (2019) described shifts in understandings of consent amongst married couples in Rwanda as a result of participation in the Indashyikirwa program. This intervention consisted of a 20-session couples program in which participants explored a variety of topics, including gender and sexual norms that contributed to sexual coercion in relationships. Facilitated and take-home exercises allowed male and female participants to reflect on coercion in relationships, identify the benefits of consensual sex and consequences of non-consensual sex (including undermining sexual pleasure and relationship quality), and cultivate communication skills to discuss sexual preferences with spouses. The curriculum emphasized shared power within couples and included community action against violence as an intended outcome, thereby addressing sexual coercion at multiple levels of the socio-ecological model. The emphasis on shared power resonates with our findings, which point to the ways in which the young authors in all countries and time points depicted gender and economic power restricting young women’s ability to negotiate consensual sex on their terms, including within romantic relationships. Stern and Heise (2019) observed that couples felt that “open communication helped both to ensure mutual consent and enhance sexual pleasure” (p. 13), reflecting the importance of including communication skills-building in multi-level efforts to shift norms around sexual consent. A small portion of narratives in our sample did depict couples discussing sex and verbally negotiating consent, but the majority did not, pointing to a potential gap in the cultural scripts that the young authors drew from when conceptualizing sexual encounters. While the Indashyikirwa program was targeted at married couples, elements could be translated into efforts with unmarried youth, particularly strategies to facilitate reflection around coercion in relationships, build communication skills, and promote shared power within romantic relationships.

Some efforts targeted at young men in the U.S. capitalize on a norms-based approach to link sexual consent with protective masculine norms. For example, Men Can Stop Rape (MCSR) created a bystander intervention and consent campaign aimed at young men that emphasizes their role in preventing gender-based violence. This campaign uses positively framed messaging, such as “I’m the kind of guy who wants consent”, embedded in scenarios like “When we started hooking up, Liz stopped, so I stopped and asked if she was okay” (for more information, see https://mcsr.org). An evaluation found that positively framed messages that linked sexual consent with positive masculine norms were more effective at shifting injunctive norms and behavioral intentions amongst young men than negatively framed messages (Mabry & Turner, 2016). The majority of narratives in our sample depicted male characters engaging in a variety of tactics to obtain sex, including pressure, coercion and physical violence when necessary. In line with broader cultural scripts that prioritize male sexual entitlement (LeClerc-Madlala, 2009), no narrative linked prioritizing partner comfort (beyond sexual pleasure) with positive masculinity. Given the increasing prominence of companionate relationship models that include (at least the appearance of) gender equality and joint decision-making, there appears to be an opportunity to link positive masculine norms with affirmative models of sexual consent.

Limitations

This study was not without limitations. As contest participants self-selected, the data were not representative of youth populations; participants may have been better educated, and more knowledgeable and motivated about HIV than the general youth population. As a product of the same contest mechanism, however, these biases were likely to be consistent across the three countries; hence the country samples, though not representative, were comparable for our purposes. We had little demographic information about individual participants other than their sex, age, country of origin, and place of residence. The data were embedded within cultural norms of performance, discourse and persuasion (Farmer & Good, 1991), which may have been informed by rhetorical considerations specific to the scriptwriting competition, reflecting the young authors’ motivation, for example, to tell what they considered to be a good story and thereby win the contest. In light of the circumstances in which the texts were written, it is impossible to know which depicted lived versus imagined experience: we treated all as representations providing insights into youth sense-making around sexual consent.

Conclusion: Consent requires empathy

Sexual consent depends upon shared indicators of willingness and unwillingness (Beres, 2007). The results from this analysis of narrative representations revealed emic understandings of sexual consent in which indicators were located along a continuum from unwillingness to willingness in a manner that could, at times, blame female characters for nonconsensual sex. Sexual violence prevention efforts have shown that it is essential to promote a model of sexual consent that acknowledges power within relationships and that includes open communication around sexual preferences and values proximal indicators of consent (Fenner, 2017). Incorporating approaches that address sexual consent with individuals, couples, and communities, such as those described by Stern and Heise (2019) and MCSR, is ideal.

Above all, sexual violence prevention efforts must promote a model of sexual consent that underscores empathy for the comfort and wellbeing of one’s partner. This was noticeably absent from the narratives in our sample. Results from studies in other settings indicate that it is unlikely that sexual consent will always be communicated via direct, verbal indicators such as those put forth in affirmative consent campaigns (Muehlenhard et al., 2016). Realistically, a sizeable portion of sexual encounters will always depend on nonverbal indicators of sexual consent, such as physically responding or not responding to another’s behavior. Tactics that cultivate empathy, particularly amongst boys and men for the experiences of girls and women, and that emphasize prioritizing the comfort of one’s partner – at times over one’s own needs – are an important component of gender transformative approaches that seek to promote healthy relationships and sexual cultures.