Abstract
Introduction
Across the last decades, states have demonstrated increased recognition, at least rhetorically, that they should assume regulatory responsibility for the wide-ranging harms that are generated by gender violence. Commonly, institutionalization of this responsibility involves increased criminalization of gender violence along with improved state funding to finance public awareness campaigns and victim-survivor support programmes. The distressing insistence of his phenomenon, which continues to impact the lives of 1 in 3 women globally (World Health Organisation [WHO], 2021), puts the effectiveness of these measures in question. The impacts of this challenge remain evident across stable democracies with established ‘rule of law’ systems, within pluralized regulatory contexts, and within postcolonial or still decolonizing contexts where state-building may be ongoing.
While there are well-documented challenges evident in the way state authorities confront gendered violence (Douglas, 2021; Whittier, 2016), it is also true that women are often positioned as passive recipients of national-scale projects designed to ‘regulate’ gender violence. This discounts the possibility that women in everyday 1 settings have repertoires of knowledge on gender and violence that they may draw on to resist insecurity.
The research discussed in this article investigates this possibility. It develops a vernacular security approach to examine situated everyday understandings of the gendered factors that fuel familial violence and women’s perspectives on how this can be managed in localized contexts. The study was conducted with women in three Pacific Island jurisdictions: Fiji, Bougainville, and with Indigenous Kanak women in New Caledonia. More detailed discussion below will show that gender violence is perpetrated at extreme rates in each site. While efforts have been made to improve state regulatory responses, these reforms are frequently implemented by Indigenous elites in ways that can diminish the standing of women and undermine their ability to achieve security.
Some analysis conducted in these contexts concludes that women are failed by both state and Indigenous customary authorities if they are exposed to familial violence and therefore have scant resources at hand to manage or resist resulting insecurity (Jewkes et al., 2014; Salomon, 2021; Salomon and Hamelin, 2007). Other assessments contend that because state law and justice institutions tend to be poor guarantors of women’s rights, women as complainants may be better served if customary mediation processes are more fully integrated into state justice systems (Forsyth, 2011). While these studies provide important background, new ways of approaching this challenge can be revealed if we ask women themselves about the everyday factors that they understand to be generative of violence and how these can be managed to advance their security.
In this article, I develop a vernacular security enquiry to answer these questions and draw from data generated through photo elicitation (PE) interviews conducted with women from everyday communities in each case study site. This approach has affinity with other vernacular security research that has deployed innovative methods to uncover ‘security speak that has been subjugated or hidden’ (Jarvis et al., 2025, see Atakav et al., 2020; Downing, 2020). The research I discuss here contributes further to that objective, but was also motived by ethical concerns to limit participant trauma and increase participant engagement, as I explain in later sections. 2 Interviews involved discussions of original photographic images depicting tensions in familial settings. Study participants were asked to reflect on the cause of the tensions depicted, who might be responsible if violence was to ensue, and the sorts of resources the woman in each image might draw upon the achieve security.
The discussion of these findings demonstrates important variance in women’s perspectives. Many respondents explained gendered familial violence as caused by a failure to act rightfully because insufficient deference is given to customary gendered hierarchies in familial and social contexts. Other respondents discussed custom differently. They emphasized not the rightfulness of gendered hierarchy in custom but the inappropriateness of it, and nominated principles of reciprocity and balance as customary ideas that should protect women from familial violence. These perspectives offer novel insights into the knowledge and expertise that everyday women hold about the relationship between gendered violence, security and customary practice.
Some of the findings I discuss align with feminist security studies research that has revealed the generalized gender subordination and oppression that is generated by hegemonic state security projects. The vernacular lens I develop here also draws more specific attention to the ways in which respondents identified principles of gendered rightfulness that, in their view, should provide foundations for their security. As I will show, this perspective challenges the hegemonic power of customary knowledge as it is pronounced by male leaders and opens the way for understanding how women’s customary knowledge on familial relations provides a basis for resisting gendered oppression.
The article proceeds in five sections. The first explains the vernacular security lens I develop for this enquiry and how this can extend feminist study of gendered security. The second explains the research context and the contested ways in which gender, custom and state-building intersect in each case study jurisdiction. The third explains the research methodology and the motivations for using PE in interviews with study participants. The fourth discusses findings from the photo elicitation study. The concluding section offers final reflections on the relationship between gender, custom and security in each case study site and reflects on the value of innovative vernacular security approaches for research on gendered security.
Vernacular perspectives on security
Bubandt’s (2005) seminal article defining the concept of vernacular security focused critical attention on ‘bottom-up’ perspectives of security as these are understood by individuals and communities in an everyday sense. Bubandt’s contention was that security is contingent. Accordingly, to come to grips with the ‘problem of security’ we must consider how it is configured differently at different scales (global, national and local). On this theme, he observed ontologies of ‘danger, risk and uncertainty’ to be socially and politically defined at distinct scalar and temporal points (pp. 276, 277). Hence, he proposed vernacular security as a ‘convenient shorthand’ to capture how discourses on security respond to ‘idioms of uncertainty, order and fear’ as well as the desire for ‘social control’ as that is perceived in a particular context (p. 277).
As many working with this concept have since observed, vernacular security approaches enable exploration of the everyday impacts of security policy initiatives for ‘ordinary people’ in their ‘daily existence’ (Jarvis, 2019: 109). These enquiries may demonstrate the security dividends that individuals or communities understand to have accrued to them as a result of security policy. They may also demonstrate how individuals or communities are made to feel more vulnerable and insecure as a result of authoritarian or ‘anti-democratic’ state security interventions (Kurylo, 2022: 17).
Vernacular security approaches also allow for the concept of security to be stretched conceptually beyond the narrow focus of security as ‘survival’ to security as citizen entitlement, or as a requirement for human flourishing and dignity (Jarvis, 2019: 113; Luckham and Kirk, 2013: 341). Beyond the universalizing and, as some have alleged, Eurocentric definitions of human security frames (Jarvis, 2019: 113), vernacular security approaches focus on how meanings attached to security can be ‘very different . . . for different people’ (Jarvis and Lister, 2013: 168). This also opens the way for understanding where and how people may engage with a wide range of actors, beyond those associated with state security agencies, as legitimate security guarantors including community leaders, elected officials, chiefly and religious leaders, amongst others (Baker and Lekunze, 2019; Jarvis, 2019).
Whose security matters?
Vernacular security approaches are also productive for research examining how security operates as a modality of power that can produce subordinating impacts for particular groups. This lens has amplified voices that may be ‘hidden, inaccessible or silenced’ (Jarvis et al., 2025: 507). In this regard, vernacular security research has affinity with critical feminist security studies that explore the gendered impacts of security interventions (Jarvis, 2019). For example, feminist perspectives on security have focused critical attention on the way hegemonic security projects such as the Global War on Terror were framed by racialized ‘rescue’ motifs targeting local women. Meanwhile, the gendered subordination, violence and insecurity that local women experienced as a consequence of these militarized strategic interventions were downplayed (Shepherd, 2006).
Feminist security scholarship has also shed light on the anti-emancipatory outcomes experienced by some women when societies are perceived to be under security threat and logics of masculine protection come to the fore (Young, 2003). In these conditions, women who are dutiful, dependent and sufficiently engaged in the social reproduction work that is typified as essential to the preservation of culture and identity are rewarded with protection (de Alwis et al., 2013: 178). Gender dissidents who openly flout expectations of gender-appropriate behaviour may, on the contrary, be far more insecure because they are positioned beyond the pale of masculine protection (Young, 2003: 14), although these logics can also be reversed in some contexts too (see Aharoni and Sa’ar, 2025).
Despite the implicit synergies between feminist research and vernacular security, it has been relatively rare to see feminist research on gender and security integrate insights from vernacular security studies. Notable exceptions include studies examining how everyday women experience the gender-marginalizing security logics that have been part of peacebuilding interventions in the Pacific Islands (George, 2018, 2017), how security as a concept is expressed by Jewish-Israeli women living on the southern border of the Gaza Strip (Aharoni et al., 2022: 446), how feelings of security are expressed between women across the lines of the Israel–Palestine conflict (Aharoni and Sa’ar, 2025) and how masculinized norms of protection are reoriented in Nigeria in response to vernacular security practices in areas under the influence of Boko Haram (Oyawale, 2025).
In the following sections, I extend on this work and develop a vernacular security approach to examine the varying ways women in Pacific Island settings narrate the causes of gender violence and imagine achieving security from this violence. The vernacular security perspective gives prominence to situated rather than generalized narratives and understandings (Jarvis et al., 2025) of the relationship between principles of gendered rightfulness and gender security and, from here, women’s perspectives on their ability rightfully resist gendered violence and insecurity. The analysis demonstrates that it is not custom itself that generates insecurity for women, but rather the erasure of women’s everyday knowledge and authority on gendered rightfulness. This is a commonly overlooked perspective in debate on where and how Pacific women’s vulnerability to gender violence might be enhanced, which tends to remain resolutely focused on the reform of state regulatory and carceral responses. To provide context to support this argument, I next develop a brief history of state-building, regulation and gendered security in each of the Pacific Island sites where the research was undertaken.
Gender, regulation and state-building in the Western Pacific
Across all the case studies, and the Western Pacific more generally, postcolonial state-building occurred later than in many other regions of the world (Demian, 2015) and is incomplete in Bougainville and New Caledonia, where Indigenous sovereignty claims are yet to be fully recognized.
Fiji achieved independence in 1970 after close to 100 years of British colonial rule. The country’s postcolonial history has been shaped by political tensions between the i-Taukei or Indigenous population and Indian Fijians, who are descended from indentured labourers brought to the country under imperial rule, as well as later waves of Indian free-settlers (Firth, 1997). Indigenous nationalists proclaim anxiety about any development that they perceive to challenge their claims to political paramountcy (Madraiwiwi, 2001 cited in Tubman, 2008: 23; Tomlinson, 2009). This also informs their well-rehearsed refrains that liberal democratic values, including human rights, are ‘foreign flowers’ for which Fiji had the ‘wrong soil’ (Tabakaucoro, cited in George, 2012: 112).
Fiji is a highly militarized country and its postcolonial history has been dominated by the influence of military-strong men who have often acted to defend the i-Taukei interests described above. The military has had direct or indirect involvement in the four coups that have occurred in the country since 1987 and has actively governed the country for extended periods, most recently between 2006 and 2014.
Gender advocates in Fiji who oppose these assaults on democracy and who advocate for the rights of women have frequently been chastised by political leaders, who contend their advocacy is too confrontational and challenges the norms of ‘quiet diplomacy’ that are said to align with i-Taukei cultural expectations (George, 2025). Sometimes the sanctioning has been much harder with some gender activists, women union leaders and even female politicians exposed to state security force surveillance and arbitrary state detention.
State efforts to address gender violence have evolved in ways that reflect authoritarian policing norms. For example, zero-tolerance and community policing approaches institutionalized in 2009 elevated state security force attention to gender violence (Government of Fiji, 2009). In practice, however, these policies also allowed the surveillance capacities of the state to be scaled down to sites of community and village life (George, 2017).
In the settler-colonial context of New Caledonia, anxieties about the survival of the Indigenous Kanak people have fuelled outbreaks of violence since the territory came under the French flag in 1853. Indigenous uprising against ongoing French control continued until the late 1980s, when a political settlement was agreed that brought peace to the territory for the next 20 years but did not result in self-determination as many Kanak leaders imagined it would. Since May 2024, the territory has again been rocked by serious violence and an ensuing heavy French state security response, as Kanak groups have resisted French government efforts to legislate electoral provisions that have the potential to weaken the political influence of pro-Independence groups (Durel, 2024).
Kanak women such as Déwé Gorodé and Susanna Ounei have played critical roles in the pro-Independence movement since the 1970s, although their efforts to expose gendered oppression was not always welcomed by male nationalist leaders. Jean Marie Tjibaou argued, for example, that women’s contributions to Kanak sovereignty were most valuable when they were ‘solidly anchored within their tribes’ (Tauru, 2011: 10). Kanak women were often harshly denounced as enacting an ‘anti-Kanak’ sabotage of the decolonization agenda if they protested against the constraining impacts of ‘familialist’ discourse or drew attention to the insecurities experienced by women within their families (Wamytan, cited Salomon, 2017: 61).
Similar gendered tensions are evident in Bougainville, the third case study site discussed in this article. This territory was devastated by a deadly secessionist war in the 1990s, fuelled by Bougainvilleans’ demands for independence from Papua New Guinea (PNG). A peace process, whose early stages were driven by the country’s women leaders, resulted in a political settlement in 2001 which established political autonomy for the territory as part of the PNG state (Braithwaite et al., 2010).
Although matrilineal principles are described by Bougainvilleans as a defining feature of their culture and identity, the political climate for women’s rights activists remains challenging. The legacies of Bougainville’s war have seen the cultural and religious logics that emphasized veneration for women replaced by a logic of force, whereby those with the greatest physical or material means expect to also hold the greatest power (DVA and AusAID, 2010). In this environment women who challenge national narratives that emphasize maternal virtue (Sirivi and Havini, 2004) by drawing attention to familial violence or advocating for women’s reproductive rights can be subjected to slurs on their character, threats and sometimes physical violence (Eves, 2015; Jewkes et al., 2014).
Despite these shared challenges, gender advocacy continues in all three case study sites where women leaders campaign to advance women’s standing and their security from gender violence. Much of their activity nominates state security and justice sectors as critical sites for the promotion of gender just reforms (Fiji Women’s Crisis Centre [FWCC], 2013; Putt and Dinnen, 2020; Salomon, 2017). Some assessments of this advocacy describe it as ‘liberal evangelism’ that spreads a ‘“gospel” of gender equality’ relevant to middle-class, urban-based elites and policy bureaucrats (Demian and Rousseau, 2019: 321). There may be some element of truth to this. Yet it is also the case that Pacific women live with, and are concerned by, extremely high rates of physical insecurity and familial violence.
For example, a 2013 study conducted in Fiji found that 64% of the ever-partnered women surveyed had experienced physical or sexual violence during their lifetime, and 24% experienced physical violence in the previous 12-month period. Within the general population, 72% of surveyed women reported exposure to physical, sexual or emotional violence in their lifetime (FWCC, 2013: 35). Experiences of gendered violence were reported by all ethnic groups but were recorded more frequently by i-Taukei or Fijian women (FWCC, 2013: 35). In Bougainville, available studies show that between 52% and 75% per cent of women report exposure to violence in their lifetimes (Dyer et al., 2019: 11; Jewkes et al., 2014). Some observers contend that this violence was normalized in the wake of Bougainville’s war (Braithwaite, 2006), a period where high levels of sexual and gender-based violence were perpetrated with impunity (George, 2025; Zale, 2004). Studies conducted in New Caledonia show that 1 in 3 Kanak women have been exposed to violence (Salomon and Hamelin, 2010: 206). Kanak women are 4 times more likely to experience gender violence than women from other communities in New Caledonia and young economically dependent Kanak women are living with the highest levels of vulnerability (Solomon, 2021: 85).
Policymakers at the national level in all jurisdictions have recognized their obligation to address this insecurity and institutionalized a range of policing and policy reform in the last two decades. These include legal frameworks that give specific recognition to gender violence as a criminal activity, that require police to protect women from perpetrators through enforcement of restraining orders, and repeals of older gender discriminatory legal statues that protected perpetrators of gendered violence crimes. 3
These reforms hold promise but are poorly implemented in practice. In Fiji, 200 statements were collected between 2014 and 2018 from women who had failed to convince police officers to act on their complaints of familial violence (Parliament of Fiji, 2019). In Bougainville, the territorial police force is poorly resourced and tends not to prioritize gender violence amongst a range of competing responsibilities. Where cases do progress through the judicial system, police investigations often produce inconsistent or low-quality evidence that reduces the quality of prosecution briefs and, consequently, the prospect that perpetrators face conviction (Putt and Dinnen, 2020: 4, 25). In New Caledonia, it is estimated that only 2% of Kanak women bring gender violence incidents to the attention of state authorities (George, 2025: 69), a figure described as ‘the tip of the iceberg’ (Salomon and Hamelin, 2010: 206). Ongoing projects to increase formal state recognition of Kanak customary civil law since the 2000s have elevated the regulatory authority of Kanak customary assessors, who are usually drawn from the male chiefly elite. 4 While described as a decolonization of law in New Caledonia (Lafargue, 2014: 188), the practical operation of customary law can produce discriminatory rulings for Kanak women when family violence matters are regulated under customary civil law (Salomon, 2018: Trépied, 2016).
This effort to integrate Kanak customary regulation into judicial practice is emblematic of Pacific Island leaders’ efforts to ensure Indigenous sovereignty is safeguarded within the broader state-building enterprise. Ironically, while this project is defended as an assertion of, and protection for, Indigenous culture and identity (Lafargue, 2014, 188), it is also informed by colonial praxis. As Demian shows, since colonial times state authorities have sought to identify ‘a combination of elements’ from a diverse field of customary regulatory practice that had ‘enough features in common’ that allowed them to be combined into ‘a unified if novel form’ (Demian, 2015: 104). These pronouncements of customary regulation were, and are still today, informed by a desire to put a stamp of fixity and consistency on custom so that it is legible to state authorities.
But these efforts may sit uncomfortably with the ways custom is lived in an everyday sense because they erase the distinctiveness of local place-based practice, as this is understood and valued by everyday people (Demian and Rousseau, 2019). This has led some observers to contend that it is possible to differentiate between ‘custom’ as a national-scale political project and ‘kastom’ as it is understood in specific contexts and shaped by the contingencies of time and place. Benedicta Rousseau defines kastom in this sense as a measure of the correct or appropriate way of being. Centred around the values of respect, unity and harmony, in contemporary everyday life kastom operates as a critical tool in determinations of the propriety and legitimacy of behaviour, personality, relationship and intent. (Rousseau, 2008:16)
The contention here is that the political project of custom as it is yoked to postcolonial state-building and the security of Indigenous identity in the Pacific might be differentiated from the kastom that is lived in an everyday sense by Pacific people as a more fluid and negotiated phenomenon.
This insight has important implications for postcolonial perspectives on security which have, rightly, critiqued the hegemonic imposition of ‘top-down’ systems of regulation in state-building contexts that simultaneously diminish the legitimacy and everyday relevance of local-level systems of authority (see Baker and Sheye, 2007). To remedy, and bring security ‘close to the people’, Baker and Sheye (2007) contend, for example, that regulatory authority should be devolved to recognize the capacities of village courts, church structures and hybridized community policing arrangements (p. 512). Similar arguments shape debate in the Pacific Islands region where the security benefits of hybridized arrangements that reflect ‘the grain of local beliefs and practices are recommended (Dinnen and Peake, 2013: 572).
Yet, Rousseau and Demian’s reflections should prompt more caution. They remind us that the effort to integrate customary security practices into the state security architecture can produce rigid and uncompromising articulations of ‘beliefs and practices’. Masculinized expressions of customary entitlement are frequently favoured in this process as authentic to Indigenous cultural protocols (Monson, 2023). In contrast, the institutionalization of codified customary forms can produce profoundly constraining regulation of ‘rightful’ gendered comportment that undermines rather than advances women’s security (George, 2017). These are important considerations relevant to research findings discussed below. These make clear the varied logics that inform participants’ reflections on security and insecurity. Before I go on to explain these responses, I next explain the PE research methodology that was used to generate data for this study.
Appropriate methods for sensitive contexts
Like Downing (2020) or Atakav et al. (2020), the PE approach I discuss here advances innovation and creativity in vernacular security research methods. PE is an interview-based methodology that asks research participants to respond to photographs (or any other visual material) and then records those responses. Proponents of this methodology contend that it enables comprehensive interactions with research participants (Harper, 2002: 15), potentially uncovering ‘knowledge that may have been previously unknown to the researcher’ who uses methods such as interviews or questionnaires (Packard, 2008: 65). Comparisons of participant responses to visual images provide the basis for analysis. As Stephen Harper observes, ‘two people standing side by side looking at identical objects, see different things’. The academically productive point arrives when ‘the differences in perception can be defined [and] compared’ with a view to understanding how they are ‘socially constructed by both parties’ (Harper, 2002: 22).
While not commonly featured in studies on gender violence (although see Frohmann, 2005), photo elicitation is particularly valuable as a research methodology that can minimize power hierarchies between study participants and researchers as well as potential communication hurdles. This was important given the linguistic diversity of the case study sites identified for the research and the need to make lines of enquiry both easily understandable and comparable within and across the case study sites.
The decision to develop a PE approach was encouraged by the idea that a spoken exchange ‘anchored in an image’ facilitates an easier, more conversational and less interrogatory form of cross-cultural communication than ‘in-depth’ interviewing. This is because an image is ‘understood, at least in part, by both parties’ (Harper, 2002: 20).
While a PE approach held promise as a productive research strategy for uncovering hidden ‘security speak’ (Jarvis et al., 2025), it was also important for managing another crucial ethical consideration: the avoidance of study participants’ exposure to trauma which might be caused by asking them directly about personal experiences of gender discrimination and violence (Jansen, 2010). PE was valuable in this respect because it allowed participants to reflect in an anonymized and depersonalized fashion on the scenarios depicted in photographs. There were, of course, numerous occasions during fieldwork where respondents ‘broke the frame’, as Harper (2002: 21) puts it, and, after some moments, began to reflect on the ways that the image they were looking at resonated with their own experience. In these instances, participants sometimes recounted experiences of violence that had never been shared with anyone else. When this happened, neither I, nor any of the fieldworkers who assisted me in data collection, encountered instances of participant distress. Indeed, these situations often made us feel that our respondents appreciated the opportunity to share experiences with an engaged listener. One participant confirmed this, stating that although women may endure insecurity, they may keep these experiences to themselves. ‘Our stories are just our own’, she stated (Fiji Interview 34). This reflection indicates how the use of innovative vernacular security research methods can also integrate ethical care considerations into the research design.
Constructing the images
The first stage in developing the PE images involved a broad-ranging desk review of relevant studies, including incidence studies of gender violence in the Pacific region, secondary academic sources, grey literature and relevant literary works by Pacific Island writers, including Gorodé (2011) and Gope (1997). My aim with this work was to identify scenarios where tensions might arise in women’s familial relationships and generate violence or insecurity. From this analysis, I identified three themes that that might guide the development of images: gendered familial expectations, gendered expectations on economic autonomy, and gendered expectations on participation in decision making.
From here, I collaborated with a community theatre group in Fiji called Women’s Action for Change and a photojournalism student from the University of Queensland, Ali Rae. The actors helped workshop initial story ideas into small vignettes that were then photographed in settings in and around Suva. Each image was constructed to capture the look and feel of family life in the Western Pacific so that the representations were legible to research participants across the case study sites. The scenarios depicted moments of tension in relationships between women and men but avoided depictions of actual violence so as not to provoke a traumatic response for the participants. Here I discuss responses to three of the 18 images that comprised the full study. All fall under the theme of gendered familial expectations.
Women’s Action for Change in Fiji, Bougainville Women’s Federation and Union Femmes Francophone d’Océanie – New Caledonia assisted me with the recruitment of study participants and the facilitation of interview sessions. 5 With the aim of generating responses to each image that were as open as possible, participants were asked to answer four questions while examining each image: If they thought the woman depicted in each was safe? Why? Who they identified as responsible for generating insecurity if that is what they saw? What actions the women could take to improve their security?
By the end of the fieldwork period in 2018, we had conducted PE interviews with just under 100 women study participants in Fiji and Bougainville and with 70 Kanak women in New Caledonia. Participants were all cis-gendered women. Their socio-economic standing was varied. Many lived in rural and remote villages, or the squatter settlements that are a feature of urban settings in each of the study sites. In New Caledonia, a more significant number resided in urban, middle-class contexts than in Fiji or Bougainville. Participants’ ages ranged from young adults to women who were elders and held some authority as leaders in their communities. Research collaborators eagerly embraced the work as did study participants, who were often keen to explain in detail what they saw in the photographs. This meant each interview often took up to an hour, confirming Harper’s views on the richness of the PE research methodology. As one participant in Fiji stated, ‘I see the photo, and plenty words come’ (Fiji interview 41). 6
Responses to the images were coded thematically to identify where and how principles of rightful gendered behaviour were expressed by participants responses and how they perceived this to relate to the security of the women featured in each image. These perceptions emerged most clearly in participants’ explanations of why women might be vulnerable to violence and what they might do to manage this situation. Within these testimonies, principles of gendered respect, gendered authority and gendered obedience were identified as key themes that might explain women’s insecurity. On the other hand, participants also nominated principles of gendered respect, reciprocity and sharing that might be invoked by women to contest constraint or resist insecurity. Difference in perspectives did not correlate with the demographic characteristics of respondent groups in ways that were significant. 7
As I show next, the vernaculars that women constructed in response to the images were varied, but provide unique insights on women’s repertoires of security knowledge and how this can be drawn upon to challenge the normalization of gender violence.
Photo elicitation responses
Image 1: Gender, household labour and security
The first photograph discussed depicted a man returning from his workplace to find his wife sleeping and his home in disarray. This photograph was taken inside a dwelling located in one of Suva’s squatter settlement communities. It showed a single low-income living arrangement. Pots and pans lay in the middle of the main living space and a woman was asleep on a mat in the background. The man was depicted in high-visibility clothing, suggestive of a demanding manual occupation, entering the doorway of the dwelling.
Many respondents across the three case study sites focused on the principle of women’s domestic servitude to the household, and to the man himself. In Fiji, this view was articulated by 44 of the 97 respondents. They were critical of the woman for ‘not doing her part’, ‘her job’ or ‘her work’ (Fiji interview 58; Fiji interview 16; Fiji interview 21). Other respondents contended that she was ‘not taking seriously her responsibility’ or contrasted the woman’s perceived inactivity with the husband who ‘does his share and he’s tired’ (Fiji interview 24; Fiji interview 21).
In Bougainville, these perspectives on gendered obligation were expressed less frequently (by 27 of the 88 participants) but with more detail than by respondents in Fiji. Participants were quick to observe the woman’s neglect of ‘her duty’ or assessed her to be ‘lazy’, for example (Bougainville interview 14; Bougainville interview 6; Bougainville interview 21; Bougainville interview 64). One respondent asserted that it was ‘custom in Bougainville that men never do housework, only women’ (Bougainville interview 3). Another stated ‘we all know that all house chores are a woman’s responsibility’ (Bougainville interview 56). Notably, one participant responded to the photograph with a more personal observation, explaining ‘we don’t have the strength to defend ourselves, women have to control the situation to avoid violence’ (Bougainville interview 13). These observations suggested that women ‘control’ the risk of insecurity by ensuring their household labour obligations are rightfully performed.
These perspectives were far less frequent in Kanak women’s responses (10 of 65). Many more references were made to the principle of gendered authority within the household and the need for women to be obedient to male authority. These responses included comments that the husband is the head of the house (chef du maison) (New Caledonia interview 23) and that the woman’s inattention to her roles would be interpreted as an affront to that authority. Some respondents from Fiji also shared this view. They observed the importance of meeting husbands’ expectations regarding household duties as a show of respect and that doing otherwise would cause male partners to become ‘short-tempered’, angry or ‘get wild’ (Fiji interview 73; Fiji interview 39; Fiji interview 53).
In Bougainville, references to the ways violence is used to enforce women’s submission to male familial authority were evident in just under a quarter of responses. Some observed that ‘men nowadays easily get mad or fed up’ in these sorts of situations (Bougainville interview 65), or that the ‘man can get angry and hit her‘ (Bougainville interview 83). One respondent went further, describing the woman’s behaviour as ‘displaying disloyalty to marriage and . . . abusive to a husband’ (Bougainville interview 35).
Although the view that women should adhere to norms of maternal servitude and patriarchal authority was strongly in evidence in participants’ responses, it was not universally expressed. One respondent from Fiji commented, ‘She’s safe because they are married and they need to help each other’ (Fiji interview 62). Another observed that the husband was not angry but helping the woman, explaining that she was sleeping because she ‘must have worked last night in a factory. The husband must be very helpful’ (Fiji interview 81). A third simply argued that ‘The husband’s expectation is wrong’ (Fiji interview 27).
Similar responses were also voiced by some study participants in Bougainville. One responded critically to the scenario, stating ‘women do most of the work and the man isn’t helping’ (Bougainville interview 11). Another emphasized the importance of partnership and reciprocity: ‘Sometimes men have to put themselves into their wives’ shoes. They have to be humble and committed to be happy to do women’s responsibilities’ (Bougainville interview 53). Others emphasized the importance of negotiation and shared planning: ‘Both of them should plan how things should go in the house’ (Bougainville interview 6; Bougainville interview 9). Respondents also articulated strong criticism of the man depicted in the image, stating that he had failed in his own gendered obligations as a reliable provider. As one participant stated, ‘The man is to blame [for the tension]. He should build a proper kitchen and a separate room to put the clothes’ (Bougainville interview 9).
Although still a minority view, Kanak women in New Caledonia were even more inclined to be critical of the man in the image. Some argued that men should share in household tasks (New Caledonia interview 10), that household labour requirements should not be the cause of anger, and the image simply depicted a certain ‘type of man’ (New Caledonia interview 13).
Overall, responses to this photograph indicated that while gendered norms of familial servitude were most commonly voiced by respondents from Fiji to explain the insecurity that the women faced, in the other two case study sites respondents were more concerned about the insecurity that women face if they challenge masculine familial authority and control. Nonetheless, in all cases, a small minority, rising to 15 out of 64 respondents in New Caledonia, discussed the photograph in ways that identified how more reciprocity between women and men, more shared decision making and more latitude for women to decide how to manage household obligations could be critical to the achievement of security in this environment.
Image 2: Gender, paid labour and security
The second photograph I discuss here was designed to reverse the representation of gendered roles. It showed a woman dressed in professional work attire, with car keys in hand, leaving an affluent-looking middle-class home while her husband reclined on a sofa reading a newspaper. The image was constructed to suggest that the woman is agitated as she departs for work and sees her husband idly passing time.
Contrary to the thesis that women in paid work are able to ‘purchase’ more gendered security in the home (Tauchen et al., 1991 cited Lenze and Klasen, 2017: 3), a majority of respondents in Fiji (58%) and Bougainville (77%) observed the woman to be unsafe in this image. In explaining their responses, 29% of participants from Fiji and 12% from Bougainville argued that the woman was ‘not supposed to go to work’ (Fiji interview 72) or ‘should stay home’ (Fiji interview 73) and ‘do the housework’ (Fiji interview 72), thereby avoiding distractions from duties to the household and the family.
In contrast, these perspectives were far less commonly held by Kanak respondents. Here, 64% of participants argued that the woman was safe and 15% made a direct connection between her employment and her autonomy. As one Kanak respondent put it, ‘a woman that works can’t be in danger because she brings money into the household’ (New Caledonia interview 16).
Logics of gendered morality strongly informed how respondents from Bougainville explained the insecurity that they detected in the image. For example, the idea that professional work puts women in close proximity with men and might see a woman’s reputation compromised in some way was mentioned by 32% of Bougainvillean participants. These responses frequently included examples of verbal slurs made against professional women, such as ‘working mothers’ are ‘dishonest’, ‘not respected’, or encourage ‘broken families’ (Bougainville interview 42; Bougainville interview 35) and engage in ‘love affairs’ (Bougainville interview 16; Bougainville interview 17). Another 32% mentioned that the woman would be leaving home without the company of her male partner, or that professional work challenged men’s ability to rightfully control or monitor women’s movement outside the home. This proximity to men in the workplace or capacity for financial autonomy were explained as factors that create intra-familial tension and expose wage-earning women to violence from male partners.
This was only a minority view amongst study participants in Fiji (17%). In this context, those who identified the man as responsible for tensions and any ensuing violence explained that this was not because the woman was working, but because the man was not, and therefore not fulfilling ‘bread winner’ expectations (just under one-third of respondents). Only two respondents from Fiji nominated logics of reciprocity or sharing and argued that the man might do alternative work within the household to support his partner. Instead, 25% of respondents in Fiji held exactly the opposite view and argued that it was inappropriate for the woman in the photograph to demand household work from the man in ways that challenged his ‘authority’ and that this would result in tension or conflict. This view was also shared by just under half of the Kanak respondents, with 44% assigning responsibility for any violence to the man but also observing that security problems are caused when women try to tell men what to do.
This suggests that while logics of reciprocity and sharing have a strong everyday resonance in each case study site, participants viewed women’s unpaid labour within the home to align more closely with principles of gendered rightfulness than their work outside the home. Across all sites, security was equated with the woman’s ability to balance her varied responsibilities rather than the woman and man working out together how to balance obligations. This contrasts starkly with the ways that principles about balance were discussed by participants in response to the previous image. Indeed, of the three images discussed in this article, responses to this particular image most strongly emphasized how restrictive gendered principles around women’s wage earning might generate insecurity for women.
Image 3: Gender and social autonomy
The final image focused on the theme of gendered sociality and autonomy and depicted a woman talking to a female friend at the front of her house with her male partner observing the conversation from the front door.
When discussing this image, many participants defined the conversation between the women as gossip and a potential source of insecurity in and of itself. Some participants from Fiji observed that the women were ‘wasting time telling stories’ or ‘gossiping’, an activity that they described as ‘not good for them’ and dangerous for their relationships with partners and their community (Fiji interview 10; Fiji interview 39; Fiji interview 24; Fiji interview 66). Reflections on masculine control were also a consistent feature in response to questions about how the women might manage the familial tensions that this kind of social situation might produce. Just under half of the respondents commented on the fact that the women would be more secure if they stayed home and focused on household duties and the family and doing ‘what she is supposed to do’ (Fiji interview 30).
Kanak participants also referred to gossip or ‘bavardage’ and observed that the women should make better use of their time. One questioned ‘doesn’t she have work to do?’ (New Caledonia interview 27), while others commented that if she doesnot attend to her ‘housework tasks, the man will become angry’, thereby increasing the threat of insecurity (New Caledonia interview 49; New Caledonia interview 8). On the question of how to manage that insecurity, respondents argued that the woman should ‘stay at or inside her home’ and focus on household tasks and chores (New Caledonia interview 13; New Caledonia interview 26).
Respondents from Bougainville also referenced principles of gendered virtue and the potential for these to be undermined by dangerous gossip in close communities. Some respondents commented that it is ‘not good for women to gossip’, that ‘gossipers mind each other’s business’ and that this could be a violation of familial privacy (Bougainville interview 11; Bougainville interview 3; Bougainville interview 56).
Some respondents also observed that the woman’s safety in the image was at risk because she was acting in ways that challenged the male partner’s authority. Just under 20% of respondents from Fiji observed that the women’s behaviour would cause male irritation because men become resentful or angry when they are not included and want to know what women are talking about. Others contended that the image depicted women being disloyal to their husbands. Participants who referenced these principles also mentioned the importance of women’s submission in this context as a strategy to achieve security. For example, they stated that the women should ‘call the friend into the house and talk while the husband is there’ (Fiji interview 62) or ‘explain what she is talking about’ to her husband (Fiji interview 61), or apologize to her husband, or ask forgiveness (Fiji interview 15; Fiji interview 54).
Bougainvillean women responded in similar ways to the image. One-third of respondents contended that the husband was unhappy that his wife was talking to her friend, with some also mentioning the husband’s jealousy or suspicion because the women were talking ‘secretly’ (Bougainville interview 83). When asked how the woman might manage this situation to avoid insecurity, respondents who were critical of the woman’s apparent disobedience argued that she should ‘tell [the] husband about the conversation’ (Bougainville interview 71). As one participant put it, ‘Men are the head of the household. Communication is best’ (Bougainville interview 19).
Kanak participants in the study were far less inclined to interpret the image in ways that reflected on the rightfulness of patriarchal control in the family environment. This was mentioned only by 10 respondents, who commented that the women had not met the husband’s expectations, with one respondent observing ‘They hit us for this’ (New Caledonia interview 35). In response to questions about how the women might achieve security in this kind of scenario this group of participants stated that the women should ‘do the work that the husband expects’ (New Caledonia interview 49) and that safety would be more likely when they chose to ‘shut up’ (New Caledonia interview 7).
In contrast to these responses, there were others in each site that referenced principles upholding women’s autonomy and freedom to safely socialize with others. In Fiji, these perspectives were defended at length, even though these also represented a minority view amongst Fiji-based respondents overall. Some argued that the woman was ‘looking happy’ (Fiji interview 3) and entitled to spend a ‘fun moment with friends’ (Fiji interview 11). Others challenged the tendency to interpret women speaking together as gossip and as dangerous. Instead, they stated that the women were ‘sharing ideas or problems’ in order to ‘solve’ them (Fiji interview 13). Criticisms were also made of the male figure in the image, who was described as someone who ‘interferes’ (Fiji interview 39), refusing to ‘give her the freedom’ and controlling the domestic environment like a ‘prison’ (Fiji interview 78; Fiji interview 80). Where respondents held views that challenged the rightfulness of masculine household control, they were more inclined to recommend that the woman in the image access external assistance to achieve safety if violence threatened. Notably only one respondent mentioned state sources of support, however, stating that if she was threatened or harmed ‘she could go to the police and report’ (Fiji interview 36). Others who nominated the importance of intervention to assist the woman argued that this would be helpful if it involved the friend she was talking to or a neighbour (Fiji interview 81), or ‘some other man’ who ‘could protect the woman’ (Fiji interview 82).
Far fewer respondents from Bougainville translated the image in this way. Only a minority observed that the man in the image was ‘to blame’ or was an ‘abusive husband’ (Bougainville interview 14; Bougainville interview 31). A small number of respondents observed that it was important for women to talk to each other because men did not ‘listen attentively’ to them, and that ‘women are not often recognised in their [men’s] talks’ (Bougainville interview 42; Bougainville interview 16). While few in number, these comments are significant because they challenge the idea that familial or community authority accrues naturally to men, and that women’s social interactions are focused only on ‘dangerous gossip’.
Bougainvillean respondents who held these views were also more likely to observe the usefulness of external authorities, nominating either state agencies or customary and religious figures as providing valuable sources of support. One participant used the language of ‘rights’ to explain how the woman might respond, stating ‘She has to know her rights as a woman in order to be aware of where she stands’ (Bougainville interview 22). Two others discussed the role that state police might play to ‘put this matter under law and order’ (Bougainville interview 72; Bougainville interview 13). Another participant suggested that the police should play a mediation role: ‘both of them should be brought to the Police station to sort out the matter’ (Bougainville interview 54). Others identified religious figures, chiefly authorities or other people in the village who might intervene to ensure the security of the women (Bougainville interview 18; Bougainville interview 65; Bougainville interview 31).
In contrast to the responses described above, more than half of the Kanak women participants argued that the woman depicted in the image was safe or at least should be safe. For example, one participant stated, ‘Men don’t like gossips. They are like chiefs; they have the authority [la hauteur]. Women are pushed to the side’ (New Caledonia interview 23). Another reflected that this was ‘a bad situation for the household’ (New Caledonia interview 46). Three respondents commented that the woman was too ‘dependent’ upon her husband and illegitimately constrained by his authority (New Caledonia interview 41; New Caledonia interview 29; New Caledonia interview 43). One explained further that ‘she is not free. For me I am free to do what I want’ (New Caledonia interview 42).
These attitudes also shaped how participants responded to further queries about what might be done to increase the woman’s security in this kind of situation. One respondent stated that she should ‘be less dependent on her husband and have more autonomy, [be] less submissive’ (New Caledonia interview 41). Another made a specific reference to rights (notably the only one amongst Kanak participants), stating that the women should ‘make the husband understand that she has the right to talk to whoever she wants’ (New Caledonia interview 29). Another made specific reference to principles of rightfulness that favoured the woman’s situation, stating they should ‘reconsider the rules of living together as a couple and the freedoms of each’ (New Caledonia interview 43).
These responses contrast in interesting ways with the evaluations of the situation offered by study participants from Fiji and Bougainville. Kanak participants responded in ways that suggested that principles of rightfulness in the familial or household environment were, or should be, relatively settled and would allow the woman to speak freely with her friend. Those that recognized danger explained it as an infraction of gendered rightfulness, something that wrongfully enforced the woman’s submission and restricted her autonomy. Notably, however, these respondents did not mention the role that might be played by outsiders – neither chiefly figures, religious leaders nor the police – as sources of support that might intervene to make the woman more secure. This is an interesting contrast to the views expressed by respondents from Bougainville and Fiji who felt the woman should be safe in this context but would also probably require the assistance of people outside the family to achieve that safety. Respondents from New Caledonia seemed far less confident in the benefits of external authorities and suggested the woman herself would have to alter the situation. The fact that no Kanak women mentioned agencies of state support – and particularly the police – was notable, and reflective of the ways in which women’s capacities to shape debates on gendered norms of rightfulness have diminished as efforts to codify Kanak customary law have intensified over the previous decade – a point I reflect on further in final section of this article.
Conclusion: ‘Custom – the roots of the tree that protects’
We might conclude from the discussion above that for many respondents, security from gendered violence depends upon women’s willingness or ability to submit to masculine authority, to limit autonomy and to demonstrate dutifulness to the family. These views repeatedly mention constraining gender norms that in many cases accord with parochial customary elites’ perspectives on the rightfulness of masculine privilege and power. Yet they were not universally endorsed in any case study site by the research participants. As I have shown, respondents also read the images in disruptive ways that challenged assertions that women’s security is dependent upon their subordination to masculine authority. These more resistant perspectives were most muted amongst Bougainvillean participants and also more muted across all three contexts in discussion on women’s participation in paid work. It is also noteworthy that for all three images, resistant perspectives were most pronounced amongst Kanak women in New Caledonia. In this context, integration of customary law into state judicial process is most advanced, yet based on rigidly conservative principles (Salomon, 2017) that emphasize the rightfulness of ‘patriarchal’ authority in Kanak society, the centring of ‘rights powers and responsibilities’ on men, and women’s servitude as critical to ‘social cohesion’ (Sénat Coutumier, 2014: 21, 22).
It is critically important to note here that while these ideas emphasize restrictive gender norms as a foundation of custom, this does not incline Kanak women to identify French state regulatory authorities as alternative or more effective guarantors of their security. Indeed, women’s reflections suggest that systems of customary knowledge and practice should provide a foundation for gendered security, but this can only be realized if ‘la hauteur’ or authority of Kanak men is balanced in ways that also recognize women’s customary standing. When considered alongside Rousseau’s earlier-cited differentiation, it can be argued that it is not kastom that is being questioned by these respondents but custom.
This idea is perhaps best summed up in the words of Kanak women leader Valentine Holle, who described custom as ‘the roots of the tree that protects me’. ‘I do not want to harm that tree’, she vowed when asked about the gender-restrictive provisions of custom favoured by the Kanak chiefly elite. To achieve security, Holle went on, there must simply be stronger recognition of the customary ‘prerogatives of women’ (Holle and Rosada, 2014).
The important point here is that these prerogatives are not erased by the words of men; they sit alongside them. The vernacular security approach developed in this article has uncovered their substance. They were manifest in PE responses that equated gendered security with principles of shared authority, shared responsibility and balanced obligation between women and men in familial contexts. The conclusion we might draw from this is that there is much to be gained from bottom-up study of the vernacular repertoires of cultural knowledge as they are articulated by women in their reflections on gendered security. Conversely, there is much that is obscured if custom is harnessed to the state-building enterprise in ways that are so rigid and tightly managed that these perspectives are unable to be heard.
Footnotes
Funding
ORCID iD
