Masculinity, power and structural constraints: Men's conceptualization of emotional abuse in Mwanza, Tanzania

Abstract

Intimate partner violence is a global problem with devastating social and health consequences to individuals and families. While some forms of intimate partner violence such as physical and sexual violence have been explored in depth, there is a lack of clarity on men's perspectives of emotional violence, particularly in low-income countries. Yet it is recognized that incorporating men's perspectives and participation is crucial for addressing intimate partner violence. We draw from in-depth interviews with 30 men and 1, 645 pictures collected through photo voice to explore men's conceptualization and experiences of emotional violence by female partners in Mwanza, Tanzania. A sub-sample of the men (n = 16) were interviewed for a second time about pictures showing different aspects of their lives. The fieldwork was conducted between April and December 2019, and the data were analyzed through a multistage inductive process. Participants described emotional violence through a narrative of ‘being hurt’ by some actions or words of their partners. These included: verbal complaints about failure to provide for family, partner's infidelity and flirting with other men, accusation of poor sexual performance, and perceived normative deviance characterized by coming home late and not carrying out domestic chores. Threatened masculinities drawing from negative communal normative ideals, and the underlying interpersonal power struggles with their partners for the control of economic provision, sexual intimacy and family matters framed men's definitions of emotional violence. Ongoing changes such as women's ownership and engagement in economic activities and the scarcity of employment opportunities for men challenges their household dominance. Multicomponent interventions should target men, couples and communal ideals reinforcing negative masculinity. Analysis of interpersonal power and structural dynamics influencing relationships must inform the design of interventions instead of the narrow focus on individual demographic attributes.

Highlights

• •Male perspectives and engagement crucial for addressing intimate partner violence.
• •Lack of insight on female to male emotional violence in low-income contexts.
• •Threatened masculinities underlie men's perspectives of emotional violence.
• •Socio-economic and policy circumstances influencing power dynamics in relationships.
• •Tackling both individual and structural factors key to addressing partner violence.

1. Introduction

1.1. Engaging men to address intimate partner violence

Intimate Partner Violence (IPV) has devastating social, economic and health consequences to individuals, families and communities (Falb et al., 2014; Gibbs et al., 2018). Men are the main perpetrators of IPV, which affects about 30% of women globally (Devries et al., 2013; Gibbs et al., 2020). Recent World Health Organisation (WHO) estimates show a higher prevalence of lifetime experience of violence (38%) reported by women in Tanzania. A study in Mwanza region reported a higher lifetime prevalence of physical and/or sexual violence (61%) against women (Kapiga et al., 2017). While it is recognized that engaging men is crucial for preventing IPV (Dunkle and Jewkes, 2007), there is lack of in-depth inquiry on their perspectives, particularly on its different forms such as sexual, economic and emotional abuse. Incorporating men's perspectives in IPV interventions will enhance their prospect for success, and in addressing the problem in society.

Emerging evidence from intervention trials and evaluation of community mobilization approaches for violence prevention show the benefits of engaging men (Abramsky et al., 2014; Dunkle et al., 2020; Ellsberg et al., 2015; Pettifor et al., 2018; Sharma et al., 2020). A trial in Ethiopia showed that an intervention package delivered in the men's arm was effective in reducing past-year perpetration and women's experience of IPV (Sharma et al., 2020). In the Indashyikirwa trial in Rwanda, couple training of men and women reduced IPV (Dunkle et al., 2020). A trial in urban South Africa evaluated community-based interventions to change social norms supporting gender-based violence. The intervention shifted men's gender attitudes and women's experience of IPV (Pettifor et al., 2018). Others have shown that engaging men increases the acceptability of IPV interventions (Kajula et al., 2016). However, questions remain on the best ways to engage men, safeguard the safety of women and sustain the impact of such interventions.

IPV occurs in all societies and across all categories of difference such as class, race and ethnicity. However, the meanings that individuals and communities attach to actions and behaviors that constitute its forms are best understood within their specific socio-cultural context. There is lack of studies investigating emotional violence in relationships between men and women, particularly in developing countries settings (Jewkes, 2010). Emotional violence takes many forms including verbal abuse, threats of violence, humiliation, jealousy, and flirting with other sexual partners (Jewkes, 2010; Kelly, 2004). It has several consequences including lowering self-esteem, causing mental health problems and suicidal thoughts (Devries et al., 2013; Gibbs et al., 2018; Jewkes, 2010). Furthermore, the prevalence and consequences of emotional abuse may be underestimated due to the different ways it is articulated in varying social contexts (Kelly, 2004). The way sufferers perceive acts of emotional abuse is important in defining and understanding its impact (Jewkes, 2010).

This analysis aimed to provide empirical evidence on how men conceptualize emotional violence from their female partners in a low-income context, and interpret the findings through a theoretically grounded framework. Specifically, we explored how the interactions between men and their partners, and the broader socio-economic and policy contexts influence these relationships including the embedded power relations between men and women.

1.2. Theoretical framework

1.2.1. Masculinity and power

The interpretation of men's perspectives of their relationships with women should consider the cultural and socio-economic influences on their roles and expectations. It is essential to ground the analysis of men's narratives and conceptualization of the different forms of IPV – including emotional abuse - in their socially defined roles within relationships, families and the wider community. In this article we draw on the concepts of masculinity and power to analyze men's conceptualization of emotional violence in their relationships with women.

Masculinity is the socio-cultural illustration of manhood (Kimmel, 1987). The concept captures both the positive and negative notions of masculinity and is useful in exploring men's dominance of their relationships with women. It explores the gendered differences – including perspectives on different matters between men and women informed by normative socio-cultural ideals (Pleck et al., 1994). In our analysis, we broaden these influences to include the socio-economic circumstances in the study setting.

Gramsci's concept of hegemony describes a mechanism of power through which dominance is attained through consensus (Scott, 2001). Through the power of persuasion, the oppressed conforms to the interests of the oppressor without direct repression or physical coercion (Scott, 2001). In some cases, the perceived possibility of the use of force by oppressor compels the oppressed to embrace consensus. Hegemonic masculinity derives from Gramsci's concept by explaining men's dominance over women through monopoly of power (Connell, 1987; Morrell et al., 2012). It incorporates both individualized and communal perspectives about violence (Jewkes et al., 2015). Hegemonic masculinity is an important concept for exploring the non-physical forms of IPV.

Expressions and forms of masculinity vary over time and across socio-contextual and geographical settings. Socio-economic and political circumstances such as poverty levels, and war and conflict affect both men and women and may influence expressions of masculinity in communities (Hughes et al., 2015; Kagaba, 2015; Mookherjee, 2004). For example, unemployment may cause men to lose income and subsequently their dominating power in relationships. Some scholars term this threatened masculinities – where overarching circumstances dwindle men's dominance (Kagaba, 2015).

1.2.2. Structural influences on relationships

Socio-political and economic changes across the globe are shaping communities and families in many ways. These structural changes are transforming the social-cultural context including norms, and the forms and expressions of masculinities (Kagaba, 2015). For instance, increased access to microcredit in low-income countries have opened up economic prospects for women, enabling them to provide for their children and families (Vyas et al., 2015). In many countries across Africa including Tanzania, Rwanda and Gambia, women are increasingly engaged in income generation activities, subsequently reducing their dependency on men (Kabaga, 2015; Schroeder, 1996; Vyas et al., 2015). Inevitably, these developments are shifting gender norms which prohibit women from engaging in entrepreneurial activities in the male-dominated public spaces. Together with other gender transformative activities and policies – such as the growing recognition of the rights of women in ownership of family property – these changes are creating opportunities for the improvement of the lives of women and addressing problems such as IPV.

Nonetheless, it is important to acknowledge that the empowerment of women and their reduced economic dependency on men puts them at risk, as men struggle to assert their dominance through violence (Krishnan et al., 2010; Morrell et al., 2012; Vyas et al., 2015). Therefore, meaningful analysis of men's perspectives on IPV must consider these structural influences on relationships, and explore both the positive and negative impacts. Exclusion of these factors omits an important aspect in investigating IPV. Besides, exploration of these transformations on interpersonal power dynamics in relationships is crucial, as historically men have exercised more power (Jewkes et al., 2015; Kabaga, 2015; Scott, 2001).

2. Methods

This study draws on in-depth interviews and photo voice (pictorial data) collected from 30 men in Mwanza city, Tanzania to explore their perceptions of IPV. The study was conducted between April and December 2019 in Ilemela and Nyamagana districts. In each district, two administrative wards with contrasting dense and sparse populations were purposively sampled. One street was then randomly selected from each ward.

To sample study participants, the research team worked with the street leaders to prepare lists of men residing in their neighborhood. The lists were used to purposively sample participants by considering of a wide range of demographic characteristics such as age, the nature of income generating activities, ethnicity and level of education. To capture relationship and age dynamics, all sampled participants were married, and aged between 22 and 61 years. All had either primary or secondary education and represented a wide range of occupations including barbers, carpenters, fishermen, small scale traders and motorcycle taxi operators (Table 1). Most of these occupations are not full-time employment, and provide sparodic income for the men as they depend on the market demands. For example, it is normal for a carpenter, casual labourer or mason to have no income generation engagement for a number of days or weeks.

Table 1.

Age, occupation, education and religion of participants.

AGE RANGE (Years)	21-31 (n = 12)	32-41 (n = 9)	42-51 (n-4)	52-61 (n = 5)
OCCUPATION
Hotel employee (n = 2)	0	1	1	0
Carpenter (n = 1)	0	0	0	1
Barber (n = 2)	1	1	0	0
Petty trader (n = 4)	0	1	1	2
Cleaner (n = 2)	1	0	0	1
Casual labourer (n = 2)	0	1	0	1
Tailor (n = 1)	0	1	0	0
Farmer (n = 2)	1	1	0	0
Mason (n = 2)	0	1	1	0
Car & Motorcycle driver (n = 4)	3	0	1	0
Religious teacher (n = 1)	0	1	0	0
Welder (n = 1)	1	0	0	0
Electrical technician (n = 1)	1	0	0	0
Shoe maker (n = 1)	1	0	0	0
Fisherman (n = 3)	2	0	1	0
Security guard (n = 1)	1	0	0	0
EDUCATION
Incomplete primary school (n = 1)	0	1	0	0
Completed primary school (n = 19)	6	7	2	4
Incomplete secondary school (n = 1)	0	1	0	0
Completed secondary school (n = 9)	6	0	2	1
RELIGION
Christian (n = 25)	10	8	3	4
Muslim (n = 4)	2	1	1	0
No religion (n = 1)	0	0	0	1

Data was collected in two phases. First, a male researcher conducted interviews with the 30 sampled participants using a flexible interview guide in Swahili (widely spoken national language). The interview guide included specific questions on how the men defined and understood emotional violence. Interviews were conducted at places convenient to participants in their homes or working places. After completion of the first round of data collection, 16 participants were invited to take part in the second phase (photo voice) which included taking pictures about their lives and being interviewed afterwards. The 16 participants were sampled to represent the different occupations and age groups, and those who had first hand and rich narratives about their experience of emotional violence from their partners in the first round of interviews. Photo voice is used in health and social research to explore a wide range of topics (Wang et al., 1996; Wang and Burris, 1997). It allows community members to participate in data collection and present their perspectives on the subject of inquiry. The 16 photo voice participants were trained for 2 days at a central location. The first day of training covered topics including the ethics of taking pictures for research, obtaining verbal and signed consent, and the importance of protecting the identity of study participants. The second day of training focused on the practicalities of taking photos. After each practical session, the facilitators and participants reviewed the pictures and discussed on how to improve.

The participants were given digital cameras to collect pictorial data for 9 days (including 2 weekends). They were requested to take pictures on four themes: about their general family life; their relationships; what was important to them; and the activities they do when outside of their homes. Throughout the data collection period, a member of the study team visited the participants to review the photos they had taken and upload the pictures in a computer. Thereafter appointments for the second interviews were arranged. Participants took a total of 1645 pictures. Each participant was requested to select 25 pictures for the second interview. The interviews explored the reasons for selecting the pictures and what they represented. One participant was interviewed on 11 pictures as they had only taken 15 pictures.

All 46 interviews were transcribed verbatim and translated into English. Two researchers (EP & DA) coded the data using the NVivo 12 (QSR International Pty Ltd, Melbourne, Australia) software. The coding captured the broad descriptions such as the activities that men do while at home and outside their homes, and the different forms of violence they experienced. In the second stage of analysis, all interviews were reviewed meticulously to capture the specific definitions of emotional violence and the described behaviours. Subsequently, detailed summaries supported by quotations were written. Three researchers (GM, DA & EP) reviewed the summaries and developed a conceptual analytical diagram from what was emerging from the data. The analytical diagram was appraised through further reviews of the data.

The pictures from the photo voice were stored in computer folders with specific identifiers for each interview. Two researchers (DM & EP) reviewed the pictures from each interview and wrote summaries. The pictorial data was used to analyze the income generating activities for men and women in public spaces in Mwanza. The research team held weekly meetings to review the findings from photos and those from the interviews and assess how they complement in the emerging overarching analytical narrative. This interactive process of review of data, hypothesis generation and further appraisal was done throughout the analysis process (Ritchie and Lewis, 2003). A review of literature guided theorization of the emerging results.

Written consent was obtained from all 30 participants after they read the study information sheet (or having it read to them). The photo voice participants obtained verbal consent from all the people they captured in the pictures. For those recognizable through the photos, signed consent was obtained. The National Health Research Ethics Committee (NatHREC) in Tanzania approved the study (NIMR/HQ/R.8a/Vol.1x/2475) and the ethics committee of the London School of Hygiene and Tropical Medicine (Ref: 11918–3).

3. Results

In presenting the men's descriptions about emotional violence, we mainly use excerpts from the interviews prior and after the photo voice exercise. These illustrate more clearly the participant's descriptions of emotional violence and their connection to the normative ideals of men identity, authority and responsibilities. We use two pictures taken by participants through the photo voice method to show the kind of income generation activities that women do in Mwanza.

3.1. It's hurting: defining emotional violence

Participants described emotional intimate partner violence through a narrative of ‘being hurt’. Men experienced emotional pain from both the words and actions of their partners particularly when in conflict. Two Swahili words were used to describe the hurt: inauma and inaumiza. The words express deep pain which is emotional in nature. Some emphasized the extent of the pain by adding the Swahili emphasis word sana, hence describing it as inauma sana (it is really painful) or inaumiza sana (it really hurts).

A woman may come … and say a word which really hurts … really hurt you to the extent that you truly feel hurt. It's possible she might tell you statements such as ‘ka kwako kibamia’ (you have a small penis). [IDI 02, 52 years, carpenter].

Further analysis of the narratives revealed that although the men explained ‘being hurt’ as a personal experience, they connected it to perceived violation of their roles as men. To them, the words and actions – even when their partners just disagreed with their opinion or perspective - were intensely painful when they seemed to question their masculinity. The perceived threat to their manhood made them react in different ways, including through violence.

If she wants to hurt you emotionally, she could tell you ‘do you think that you are a man?’[laughter] … I would ask myself, is there any difference between me and other men, what is wrong with me? [IDI 03, 37 years, hotel employee].

You know a woman has the ability to hurt her husband even by disrespecting him, by not paying attention, probably the man has his perspective and the woman does not agree with it … and when a man knows that this woman doesn't agree with my point of view then it is already part of the pain, he gets hurt. [IDI 18, 33 years, Islamic religious teacher].

3.2. Threatened masculinities: diminishing social and sexual roles

3.2.1. ‘You are just a man in trousers’: waning masculinity

The participant's narratives revealed that the emotional pain which men experience from some of the utterances and behaviours of their female partners was connected to the normative prescriptions of their role as men. In explaining this, they referred to a common Tanzanian Swahili idiom: ‘you are just a man in trousers’ – wewe ni mwananume suruali. This popular phrase is a negative reference to men who don't perform their expected responsibilities to their female partners. Such men are seen as parasites to their partners who work hard and take care of their needs – instead of them doing so. In the Tanzanian context, Mwanume suruali depicts a man who is playing the ‘woman's role’. Men find the term very offensive and are likely to respond violently, including to other men who label them so.

She would tell him: Are you even a man? Which kind of a man doesn't even have money? Eeh ? There is this thing they use, they would call you a trouser man … that you have nothing. All you can do is dress up but you don't have money. And this hurts men badly. [IDI 10, 46 years, small scale trader].

A woman provoked her man by telling him awful words: That it was a shame he could not even buy her a phone, or provide her with capital money so that she could start a business. The man told her that he was planning to do so. He told her, ‘I was looking for your business capital but you have decided to embarrass me with those words’. The man was so hurt. He even started crying because that woman didn't even consider speaking to him at a private place. She just started yelling in front of other people, saying ‘What do you have, you are just mwanaume suruali which means that he is good for nothing or hopeless … He was so tortured by her words that he started crying. [IDI 26, 22 years, fisherman].

Mwanaume suruali is a linguistic embodiment of the community's disapproval and negative expression of the changing roles between men and women. The term has emerged in recent years, in a period when the country has fully embraced a market liberal economy which has opened opportunities for women to engage in a wide range of economic activities. It encapsulates the socio-economic transformation taking place in Tanzania where men increasingly rely on their partners for survival (Vyas et al., 2015).

In their narratives, some of the participants gave examples about women who claim to have ‘married’ their male partners. In saying so, they implied that women have taken the men's dominant role in the relationship. In Tanzania, patrilineal marriages dominate, with the men paying bride price (Mwaseba and Kaarhus, 2015). Hence a woman's claim of marrying a man initiates interpersonal power struggles. It also shows disappointment in women about their partners who fail to accomplish their expected roles. Hence the participants were categorical that being called a mwanaume suruali caused deep emotional pain and constituted emotional violence.

There are women who say things like: It is like I'm the one who married you! What do you even have since I'm the one who feeds and dresses you? So don't pretend to be a man since you're just a man in trousers who has no sense of direction. You should be thankful to me. So, that shows women make statements which are unwise towards their husbands. That he has nothing. He has no money. Perhaps it's the woman who takes care of him by feeding and clothing him. [IDI 20, 55 years, fish seller].

You may find your wife treating you like a slave and ordering you around in your own family. You find your woman assigning you cooking tasks. She will be there wandering around in the streets or even take alcohol and leave you looking after the kids and cooking as if her timetable says it is supposed to be your turn. That is so disturbing and it feels like a man is living opposite to the expected marriage life. [IDI 28, 22 years, fish dam cleaner].

3.2.2. Normative deviance: disrespect from partners

Men's conceptualization of respect is linked to masculinity and power (Gibbs et al., 2014; Siu et al., 2013). In some settings, the ability of a man to work hard and take care of the needs of his family is fundamental to their respectability in the community (Siu et al., 2013; Wilson, 1969). In such contexts, men expect unwavering obedience from their wives, which denotes respect (Wilson, 1969). Participants in this study held similar views by demanding unquestionable obedience from their partners and considering the contra as disrespect.

Commensurate with hegemonic masculinity, study participants did not expect their partners to challenge their views or opinions. Neither did they expect them to change timetables of household activities without their approval. Participants expressed lack of respect from women through neglect or refusal to perform household chores or taking care of children. They described such behaviour as emotionally hurting to men, especially when a woman responded rudely when asked about it. Despite some women having to come home late as they were busy with economic activities, men still expected them to carry out ‘their’ household activities. Generally, women are not expected to spend most of their time in public spaces. Some of the pictures taken through the photo voice showed women hanging out at motorcycle stands or bus stops undertaking some income generation activities. Men's inability to secure obedience from their partners was perceived as loss of control, exposing underlying power struggles for monopoly of household affairs.

You come home from work and sit there until 10 o'clock pm in the evening. When you ask her of her whereabouts, she replies: ‘Am I a small child that you have to monitor me!’ That makes you angry. The next day she does the same … why haven't you cooked today, why did you leave the child unattended, where did you go? There is no polite or reasonable response … so you find yourself hurting inside. [IDI 01, 49, hotel worker]

Similarly, men felt disrespected and emotionally violated when their partners complained about them not making enough effort to find employment or engage in entrepreneurial activities. Comparison to other ‘successful’ men triggered emotional pain. Participants expressed a common belief that failure to subdue strong opinioned women who did not obey to their instructions showed weakness. A woman's refusal to give financial assistance to their husband or partner was also seen as disrespectful.

I remember one day I had no money for food at home, so I asked her to help me out because I had no money. I had given it all to my boss and I had not made extra money. She refused and reprimanded and insulted me verbally … she told me to stop disrespecting her. She also told me to leave her alone. The antics kept getting worse. It got to a point she would not even cook because she had eaten at work … So I complained to my parents. [IDI 17, 25 years, motorcycle taxi driver].

3.2.3. Erotic power: control of sexual intimacy

In some settings, optimum sexual performance is the core expression of manhood. Men value and make huge efforts to protect their sexual reputation in families and the wider community. In some contexts, perceived threats to men's sexual function undermined their uptake of medical interventions, to their detriment (Courtenay, 2000; Humphries et al., 2015; Mweemba et al., 2018). Besides, a woman's doubting of their partner's sexual performance leads to low self-esteem – a consequence of emotional abuse (Jewkes, 2010).

Study participants narrated about painful verbal insults men receive from their partners by ridiculing the size of their sexual organs and their sexual performance. Such accusations not only hurt men, but also raise suspicion about the faithfulness of their partners. They acknowledged that when their partners flirt or actively pursue other men it hurt them. Reportedly, women did this to either ‘revenge’ after suspecting or proving infidelity of their partners, or as a way of hurting them emotionally during conflict. To achieve maximum emotional hurt during heated verbal exchanges, some women claimed that the men did not father their children.

A person [woman] may come … and say a word which really hurts … really hurt you to the extent that you truly feel hurt. It's possible she might tell you that … you … there are phrases that people use, statements such as ‘ka kwako kibamia’ (you have a small penis). She is mocking you … or a woman might tell a man that, ‘even these children whom you see (and think are yours) are someone else's’. Words like these really hurt. [IDI 02, 52 years, carpenter].

Such cases created emotional insecurity to the men and damaged their masculine self-esteem. Men were also concerned about the negative image such a situation would portray to their relatives and others in the community. They felt to have lost control of sexual intimacy with their partners. Consequently, they would confront the woman or suspected man – risking violence.

What I know, may be letting you see that she is openly seeing another man, that can be violence. Or she has decided to do something abusive which hurts you a lot. She may not beat you or even injure you but potraying such kind of image to you, you may be heartly hurt a lot. So I consider that as huge violence. [ IDI 16, 37 years, food vendor].

Despising her husband, she might have some affairs outside the marriage and maybe her husband is aware but she takes it easy, she doesn't care. I mean if it reaches a point that she looks down at her husband … I mean she reaches a point of having some affairs outside the marriage with another man. For a man, when it reaches that point, as a man you would feel hurt. [IDI 03, 37 years, hotel employee].

3.3. Structural dynamics and disrupted monopoly of power

Socio-economic changes taking place in Mwanza and other parts of Tanzania have provided women with various income generating opportunities. It is common to see women of all ages busy in the streets and other public spaces - such as formal and informal markets - engaging in economic activities. Photo voice data established this, as participants took pictures showing their partners and other women in the community engaging in diverse income generating activities. These include tailoring, and selling clothes, vegetables, food, and fish (Fig. 1 and Fig. 2). The interviews revealed that 26 partners of the participants engage in a variety of income generation activities.

Fig. 1.

Women tailoring clothes.

Fig. 2.

Women drying sardines on a huge rock for selling.

In Tanzania, microfinance policies and institutions have made it possible for women to access funds to invest in income generation activities. Participants explained that it was the women who owned the income generation ventures captured with the photo voice method. This is profoundly different from the past when very few women owned income generation ventures. They were mostly labourers in businesses owned by their male partners or other men in the community.

In the following interview excerpt, a participant describes a picture he took of a woman preparing sardines for selling.

Interviewer: What is this photo of?

Participant: This photo is of a woman who is frying sardines. This photo reflects a good relationship in her family as when one earns income it will enable that person to improve their relationship because he or she (one of the partners) becomes independent. They don't rely on the other to be supported.

Interviewer: What caption would you like to come up with for this photo?

Participant:Creativity is required in any activity.

[IDI 03, 37 years, hotel employee, Interview about photo no 23]

The majority of our participants were happy with the contribution that their working and entrepreneurial partners make to household economics. However, they were uncomfortable when the women became confident and assertive of their own views. They were worried about portraying an image of weakness or loss of control to their relatives and the wider community.

The woman becomes the head instead of the man, the woman becomes the head of the family. So, she uses such hurtful words, and you feel you are nothing but a useless man, you are a trouser man: ‘I am taking care of you, idiot’ … That is hurtful to a man. He would feel bad … If she yells at you in any way or form in whatever surroundings, it hurts so much for a man to be yelled at. [IDI 04, 33 years, barber].

For example, she says … ’what kind of a man are you who cannot take care of his own family’. If you are mentally okay, it will begin to torment you, that now the society is aware of my situation that I cannot take care of my own family. That means the woman goes out to spill in the crowd: ‘What kind of a man are you who is taken care of by a woman, a man depending on a woman?’. [IDI 06, 61 years, carer of church premises].

4. Discussion

Our analysis shows that men in settings like Mwanza draw from the wider normative descriptions of male identity when interpreting the actions and words of their intimate partners – and hence perceiving them as hurtful, disrespectful and threatening. Also, it ascertains the influence of overarching socio-economic and policy factors in shaping men's views of emotional violence through the individual and communal expressions of masculinities in Mwanza, and likely other similar contexts. Structural factors have received less attention in the discussion of masculinities in Africa, particularly in the context of intimate partner violence. Instead, more focus is on the influence of cultural beliefs and practices (Jewkes et al., 2015; Pyke, 2020; Siu et al., 2013). While this is important, it is essential to recognize the contribution of the broader socio-economic and policy circumstances in shaping perceptions of masculinities to effectively address the problem of intimate partner violence in low-income contexts.

The microeconomic and policy environment in Tanzania has opened up more opportunities for women to own and engage in a broad range of income generation activities. Tanzania's development vision 2025 aims to achieve economic and social development, and address gender imbalances through improved livelihood and the participation of all groups in society (Ministry of Finance (Tanzania), 2000). The National Plan of Action to End Violence against Women and Children encourages women economic activities (National Plan and action to end violence against women and childrens in Tanzania, 2016). In most parts of the country, community based formal and informal organizations – such as Village Community Banks (VICOBA) - are thriving. Furthermore, women in urban and rural Tanzania engage in informal ways of raising capital among themselves – mostly through groups – commonly known in Swahili as chama. These socio-economic policies and structural developments are transforming relationships as they enable women to own income generation ventures.

These overarching circumstances enhance women's financial and decision-making autonomy within households which makes some men feel insecure as they occur in a context of deeply entrench negative patriarchal norms. Women increasingly expect their partners to complement their economic endeavours by taking on household responsibilities. However, our study showed that men draw on negative masculinity and normative beliefs to interpret their partner's actions and verbal statements as irresponsible, disrespectful and hurting, and hence being perceived as acts of emotional violence. They become anxious when their monopoly of power, decision-making hegemony and sexual imagery is challenged. Men who subscribe to traditional or religious notions of masculinity are more likely to feel emotionally violated by their partner's challenge of their masculine identity. These include older and younger generations of men who grew up in families or communities with strong cultural and religious interactions. However, popular communal expressions reinforcing negative masculine identity – such as the Swahili idiom mwanaume suruali – could pressure men with progressive attitudes to submit or feel embarrassed to challenge them.

Economic dependency erodes men's masculinity and threatens their dominance in relationships (Gibbs et al., 2014). As observed in other African countries (Kabaga, 2015; Schroeder, 1996), it highlights emerging normative expectations on married couples or partners to jointly cater for their economic needs and those of their children. Yet it conflicts with the previously cultural expectation in Tanzania and elsewhere portraying men as the main bread winners (Kabaga, 2015; Mookherjee, 2004; Schroeder, 1996). Participants in this study recounted about men's struggle for employment and income generation opportunities. This makes men feel that most of the economic empowerment efforts predominantly focus on women – as also reported in Rwanda (Kabaga, 2015).

In Tanzania, as in many other African countries, the adoption of the World Bank advocated structural adjustment programmes in the 90s led to less state control of the economy, removal of import restrictions and consequently the collapse of many state-owned industries (Noorbakhsh and Paloni, 1999; Riddell, 1992). These caused a decline of formal employment opportunities resulting in unstable income for men. In Mwanza city, most state-owned cotton processing industries and cloth factories closed down in the late 80s and early 90s and are yet to make a full recovery despite privatisation by the government in an effort to revive them. Of recent, overfishing and environmental degradation has caused reduction of fish in Lake Victoria (Balirwa et al., 2003; Ojuok et al., 2007). This has resulted in a decline of the male dominated fishing activities and the fish processing factories around the lake. All these socio-economic structural changes together with population increase through urbanisation have diminished income generation opportunities for men in Mwanza city. Such changes have dwindled men's ability to provide for their families and subsequently challenged their household power and hegemony.

While men generally welcome the economic ventures and support from their partners, most are uneasy about the resulting implications to interpersonal power dynamics in their relationships (Kabaga, 2015; Vyas et al., 2015). They seem to manage this dilemma by outwardly appearing supportive of the economic empowerment of their partners, while nursing deep emotional pain. This is clearly demonstrated by the disapproval of men and emotional pain when their partners described them as parasites through the idiom of ‘mwanaume suruali’. Consequently, they strain when their power and control is challenged and some resort to violence to assert their authority (Abramsky et al., 2020; Jewkes et al., 2015; Krishnan et al., 2010; Vyas et al., 2015).

Linguistic expressions such as ‘mwanaume suruali’ propagate negative patriarchal masculinities reinforcing men's sole role as family providers while downplaying women's achievements and support to families. Intervention messaging should challenge such expressions of entrenched patriarchal norms and the underlying negative beliefs. Emphasis should also be placed on the importance of acknowledging women's achievements and autonomy. In contexts similar to Mwanza, women's accomplishments risk a violent backlash from their male partners particularly when compared with other ‘successful men’. Interventions should discourage partners from making such comparisons. They should also include sessions to assist men to reflect and make sense of the socio-political and economic structural changes taking place in their countries and implications to the different notions of masculinity (Kabaga, 2015; Pyke, 2020).

We make the following specific and practical recommendations to IPV interventions and programmes. First, there is need to sustain interventions focusing on men. These should engage men's descriptions of the different forms of violence and the influence of negative normative masculine ideals. Inclusion of innovative, reflective and flexible sessions to enable men comprehend how such normative expectations shapes their male identity, advantage them over women, and the resulting consequences to relationships should be an essential part of interventions. It is also important to provide them with knowledge and skills on how to cope with the mental strain to meet such expectations, for example being the main bread winners for their families. Such efforts should go hand in hand with the community-based intervention activities to establish new norms about creating equitable relationships and families. Community based mental health services should be strengthened and made available to both men and women. The importance of recognizing, appreciating and supporting their partner's efforts to support families – instead of perceiving them as hurtful and emotional abusive – should be emphasized. To ensure relevance and acceptability, co-designing with both men and women should be central to developing such interventions. Secondly, interventions should continue to engage men's negative gender attitudes and behaviours – such as those prohibiting their participation in domestic chores. Instead, they should view taking care of household responsibilities as part of their obligation. Thirdly, intimate partners should be given skills on how to communicate their concerns to their partners, listening and responding to concerns of their partners, and managing differing opinions. Inhibitive masculine norms discourage men to be open about their emotional feelings compared to women (Mburu et al., 2014; Nyamhanga et al., 2013; Shuman et al., 2016). In Tanzania, men's discussion of their emotional pain is seen as a sign of weakness, resulting in reluctance to disclose about their need for medical attention for various conditions (Nyamhanga et al., 2013).

Complementary interventions should focus on couples and the wider community as female partners and community members play a role in reinforcing some negative notions of masculinity. Joint couple reflective sessions on how to identify and deal with social pressures and expectations which negatively affects their relationships are essential. In addition, priority should be given to improving couple communication and appreciation of each one's role in building a healthy and happy relationship. Couple sessions are a good platform for the conveyance of such skills – as shown in Uganda and Rwanda (Dunkle et al., 2020; Green et al., 2015). However, careful consideration should be given to avoid reinforcing men's power and dominance (Green et al., 2015), and instead aim to transform existing harmful norms. For example, sustained community wide efforts to engage negative normative notions about men and women should be implemented. These could be delivered through a variety of platforms such as the media and community gatherings. Prevailing communal negative idioms such as mwanaume suruali should be discouraged. Intervention programmes should identify and build alliances with all relevant community and national stakeholders to ensure sustainability, joint policy engagement efforts and to avoid duplication. Moreover, national level economic policies and empowerment initiatives should address the concerns of men who feel excluded.

5. Conclusion

Intimate partner violence remains a leading cause for the denial of women's rights and a barrier to the achievement of the sustainable development goals in many low-income contexts. It also a leading cause of breakdown of relationships, and families and severely affecting women and children. While there are many promising efforts to address the problem, it is crucial to incorporate the perspectives of both men and women. Addressing all aspects of the problem from the structural to the individual level should be central to such efforts. Clearly no single intervention or approach is sufficient, but a holistic package targeting the different causes – such as the underlying negative norms of masculinity - are critical, despite the conceptual, methodological and operational challenges. Analysis of interpersonal power and structural dynamics influencing relationships should inform the development of comprehensive interventions against intimate partner violence, instead of the narrow focus on individual demographic attributes.

These findings emphasize the importance of the socio-cultural context in shaping men and women's definitions of emotional violence and other forms of IPV. Gender-based violence interventions and programmes should engage such perspectives by using the evidence generated through mixed methods research. IPV interventions are likely to succeed if they engage with the concerns and perspectives of men. Undoubtedly stress resulting from poverty and lack of income puts men at risk of perpetuating violence (Krishnan et al., 2010). However, this does not mean ignoring women's perspectives and interests and sympathizing with men's justifications for violence.

Credit authors details

Conceptualization and investigation: Gerry Mshana, Saidi Kapiga and Heidi Stöckl, Data curation: Gerry Mshana, Esther Peter, Donati Malibwa, Diana Aloyce, Writing: Gerry Mshana, Review and editing: Esther Peter, Donati Malibwa, Diana Aloyce, Saidi Kapiga and Heidi Stöckl.