Complex gendered agency in Mexico: how women negotiate hierarchies of fear to search for the disappeared

Fear and agency are complex, interrelated and gendered phenomena for the madres buscadoras , the women searching for the disappeared in Mexico. These women operate in a context of unrelenting, multisided violence. At the same time, they choose to engage in activism that puts them at heightened risk of violence at the nexus of criminal organisations, state corruption and insecurity. This article investigates how the madres navigate contexts of gendered violence in Veracruz, Mexico, to engage in expressions of complex gendered agency. It makes the argument that we can understand why the fear of violence does not necessarily lead to demobilisation or inaction when we locate their activism within a hierarchy of fears . By recognising that the fear of never knowing about their missing loved ones outweighs the fears of violence that they are exposed to on a day-to-day basis, we gain insight into why they choose ‘fight’, rather than ‘flight’.


Introduction
'I am afraid, but I live with fear in all parts of my life. My biggest fear is not knowing whether my son is alive or dead.' (Interview, 2 March 2022) Fear and agency are complex, interrelated and gendered phenomena for the madres buscadoras, the women searching for the disappeared in Mexico. These women operate in a context of unrelenting, multisided violence. This term, conceptualised by Menjívar (2011: ch 2), is summarised as a 'potent combination of structural, symbolic, political, gender and gendered, and everyday forms of violence' (Walsh and Menjívar, 2016: 586). The majority of the madres buscadoras have no knowledge about the fate of their disappeared children and loved ones. At the same time, they choose to engage in activism that puts them at heightened risk of violence at the nexus of criminal organisations, state corruption and extreme insecurity (ICG, 2017). In this article, I investigate how the madres navigate contexts of gendered violence in Mexico to engage in expressions of complex gendered agency.
In recent years, literature on the phenomenon of the 'disappeared' in Mexico has begun to gain increased attention (see Ansolabehere and Martos, 2021). Other research has focused on those left behind, primarily the mothers and wives who sacrifice their safety and their ways of life not only to find their loved ones, but also to make demands for truth and justice (Schwartz-Marin and Cruz-Santiago, 2016;Cruz-Santiago, 2017;Wright, 2017;Orozco Mendoza, 2019;Gallagher, 2022). What is less explored, however, is how these women negotiate and understand the violence and fears associated with the high-risk activism they undertake. 1 In this article, I highlight how women's activism is complex, nuanced and must be understood in relation to the 'everyday scripts of violence' that colour the communities in which they live (Hume and Wilding, 2015). 2 The article makes the core argument that we can understand why the fear of violence does not necessarily lead to demobilisation or inaction for the madres buscadoras when we locate their activism within a hierarchy of fears. That is to say, by recognising that the fear of never knowing/finding out about their missing loved ones outweighs the fears of violence that they are exposed to on a day-to-day basis, we gain insight into why they choose 'fight', rather than 'flight'.
Fear is a complicated emotion; Ana Villarreal (2015: 136) writes that it 'may both tear the social fabric and bring people together, both destroy public space and create new forms of social life'. Beyond asking why people are afraid, though, she asks what people do when they are afraid (Villarreal, 2022). Joining a group of madres buscadoras can be a dangerous activity; mothers I spoke to have been stalked, threatened, stabbed and even shot at point blank range in the course of their public action. They are clear, though, that the love they have towards their missing children and towards each other is more important than shielding themselves from other sources of fear through inaction. Moreover, they are morally outraged by both the senseless violence their relatives have been subjected to and the lack of action by -and complicity of -the state to support and protect them. This is not to say that they do not feel fear; they do. However, the madres strategically negotiate the hierarchies of fear they feel. This negotiation is an expression of complex and gendered agency.
Drawing on qualitative fieldwork with mothers searching for the disappeared, human rights lawyers, security specialists and journalists, this article explores gendered explanations for how and why the madres buscadoras are able to overcome the barriers to high-risk mobilisation, namely, those related to fear of violence. The article uses the case study of a colectivo ('collective') of women searching for their loved ones in Veracruz, Mexico, to explore the ways in which fear and agency are conceptualised and negotiated by those who engage in high-risk activism. I begin by outlining how chronic violence and insecurity have gendered implications, as well as how women navigate these violent spaces, and tie these bodies of scholarship to the case of Mexico. Next, I outline the methods I used to engage in this sensitive research. Finally, I discuss the case of a colectivo of women who actively search for their missing family members in Veracruz. I outline the multiple ways that their lives are shaped by multisided violence, how this violence informs their expressions of gendered agency and how they are driven by different emotions that ultimately allow them to renegotiate their hierarchies of fear.

Gendered violence in Mexico
Different bodies of feminist scholarship provide useful tools for understanding how chronic violence and insecurity impact women's day-to-day lives in Mexico. For example, the gendered continuum of violence is a framework premised on the idea that gendered power differentials within patriarchal societies shape the ways in which men and women move through life; it facilitates understandings of violence as interrelated, overlapping and reinforcing (see Kelly, 1988). In 2001, Moser developed an operational framework that connects political, social and economic violence. This violence impacts men and women differently given their socially ascribed gender roles, as well as their differential access to resources and decision-making spaces (Moser, 2001: 37). Cockburn (2004: 43) further writes about the gendered continuum of violence, as it reflects the 'connectedness between kinds and occasions of violence'.
While acknowledging that women face multiple and intersecting violences across time and space, feminist scholars are increasingly illustrating how women express agency (see Berry, 2018;Zulver, 2022) and micro-resistances to violence (see Baines, 2016;Sutton, 2018;Zubillaga and Hanson, 2020;Stallone, 2021;Sarac, 2023) in conflict settings. Scholars have noted that contexts that feature extreme and chronic insecurity but that do not usually fall under the category of countries at war are sometimes excluded from these analyses. 3 Within urban violence studies, however, scholars are increasingly focusing on how enduring and everyday violence impacts women's lived experiences (Moser, 2001;Hume, 2009;Wilding, 2010;Menjívar, 2011;Hume and Wilding, 2015;2019;Walsh and Menjívar, 2016) and interferes with progress towards equality. For example, in Wilding's work in Rio de Janeiro, she argues that literature on violence and insecurity creates a false binary between violence that impacts women (private violence, namely, intimate partner violence) and violence that impacts men (public violence, such as gun crime and other forms of gang-related violence). Her research explores how gendered violence, in fact, permeates all realms of life in the insecure urban setting (Wilding, 2012). Hume's (2009) research in El Salvador further compellingly demonstrates how the violence women experience 'lives on the knife's edge between public and private violence'.
When it comes to women's agency in these violent settings, it is important to situate their actions within the complex conditions in which they live (Hume and Wilding, 2019). This means that a lack of action is not necessarily a reflection of passivity, but rather a considered decision about what other options (or lack thereof) for action exist; the authors ask us to consider agency as a process, rather than an end goal. This perspective of complex agency is evident in Zubillaga and Hanson's work on collectives organised by mothers in poor neighbourhoods in Caracas. Their research details how mothers 'perform everyday forms of resistance' in relation to the armed actors who operate in those spaces; depending on the setting, they assess whether to engage in 'dramatic, discursive actions' that are 'more openly oppositional, such as shouting, scolding and talking, to more hidden ones, such as both "circulating gossip" and "captive gossip" to more helpless ones, such as whispering' (Zubillaga and Hanson, 2020: v). Even silence can be a reflection of agency; Sarac's (2023) research about Yazidi women under ISIS captivity shows that they used silence both as a coping strategy and as a liberation-enabling strategy to facilitate their survival and escape.
Where does Mexico fit within these different categories and boundaries? How can we illustrate how violence impacts the mothers searching for their disappeared children in gendered ways? In this article, I overlay a gendered lens on the intersecting and 'multisided' (Walsh and Menjívar, 2016) ways in which mothers' experiences of searching for their children and other relatives are shaped by the violent contexts in which they continue to live. 4 In Mexico, the violence that women experience is often assessed using narratives that place it on a spectrum of violence against women, which is then framed as private violence (see Wright, 2017). Activists and scholars have pushed for recognitions of this violence as intrinsically connected to public expressions of insecurity, impunity, corruption and neoliberal economic policy (Lagarde, 2006;Staudt, 2008;Monárrez Fragoso, 2015;Staudt and Méndez, 2015;Wright, 2017;Orozco Mendoza, 2019;Frías, 2023). Indeed, Orozco Mendoza's (2019: 213) work on maternal activism recognises that 'when faced with institutions that disavow feminicide, the victims' mothers risked their own lives through bodily exposure and created communities of resistance that challenged dominant constructions of Mexican women as passive and docile'. Largely, however, the linkages between gender-based violence and the violence produced by gangs, organised criminal groups and state militarisation goes un-interrogated; as Kloppe-Santamaría (2021: 4) writes: 'violence against women has been, at the very best, addressed as an afterthought, as part of those "other" forms of violence impacting citizens in the "private realm"'. It is in this liminal spacebetween public and private -that the madres buscadoras operate.

Methods
The data in this article are drawn from 27 qualitative interviews collected during two trips to Veracruz and Mexico City. Interviewees included members of civil society organisations, human rights lawyers, academics, journalists, security specialists and leaders and members of a collective organised by mothers and other family members of the disappeared in Veracruz. Beyond interviews, I also engaged in participant observation when I attended a colectivo general meeting and a protest at the mayor's office, and accompanied the women in their annual Mother's Day march.
The madres (both leaders and members) face ongoing risks of violent targeting in their communities, and they are wary of speaking to strangers. I was able to contact them after a colleague put me in touch with the leader of the collective. When I expressed interest in travelling to Veracruz, she introduced me to leaders who arranged my schedule with the mothers themselves. We met in a hotel in the centre of town, and interviews were conducted in a quiet corner where they felt able to speak openly. Before beginning the interviews, I explained the broader goals of the project by reading the project's information sheet, gained informed consent and opened space for interviewees to ask any questions they had for me. 5 By the time of the Mother's Day march (my second trip), I had built rapport with certain members, which allowed me to meet and speak with other members. By marching alongside the women on one of their most emotional days, I demonstrated solidarity, which reinforced the connections I had made during my previous research trip.
My interview questions were semi-structured, which created room for interlocutors to modify the conversations according to what they wanted to share with me. All interviews were conducted in Spanish. The questions focused on how the members had joined the group, what they liked about it and the ways in which they actively expressed and felt solidarity. Although I purposely never asked about the details of their children's disappearances, without fail, every mother I interviewed wanted to begin the conversation by talking about the details of their loss. They showed me photos of their missing children, smiling at parties or sharing meals together; they also showed me pictures of the grandchildren who they are now responsible for raising. They showed up to the interviews with official colectivo bandanas, facemasks and banners that featured photos of their children; 'Until I find you, my son' read the bold letters superimposed over the face of a handsome young man, his tough expression visible in the photo.
Despite an ethical protocol that prioritised asking about participation and activities within the colectivo to avoid potentially distressing questions, the mothers wept while they told me about their lives. In the spirit of feminist, relational research, I let them guide the interviews in the way that felt meaningful to them; in a few circumstances, when a woman became too distressed, I stopped taking notes and sat with her until she felt calmer. One woman brought me a bag of drinks and snacks, "in case I got hungry" while I interviewed her. I explain these moments of intimacy to illustrate the care that the mothers showed me, a stranger. As I will detail in the following, the spaces where they feel safe and supported are few, and are generally tied to the colectivo. I felt that the interview spaces we shared were based on a certain level of trust.
By way of reflection, I want to note that becoming a depositary of the madres buscadoras' trust and care was both an intense privilege and also emotionally challenging, particularly when bearing witness to raw expressions of grief. Choosing to engage in relational interviewing allowed me to prioritise the respect and dignity of the interviewee (Fujii, 2018), while also acknowledging my own humanity as the researcher. Moreover, like Marie Berry, who writes about '[embracing] the persistent discomfort' (Berry, 2022: 953) when telling stories that are not her own, I too feel buoyed by the trust bestowed upon me by the madres. To borrow Berry's (2022: 8) words: 'their trust in me affirms my trust in the power and importance of sharing [their stories]'.

The case of a colectivo in Veracruz
"Bad things will happen" Veracruz lies nestled in the Gulf of Mexico, and at first glance, it appears to be a beautiful tourist destination. The palm-lined malecón ('boardwalk') is full of vendors selling balloons and souvenirs. The famous La Parroquia cafe serves coffee, while buskers play marimba music, and across the water is the port, where massive ships move in and out of the harbour. Yet, Veracruz is a city plagued with horrific violence; as the taxi driver who drove me from the airport said: "remember that you need to behave like Cinderella here; as soon as it gets dark, you need to be at home or bad things will happen" (field notes, 2 March 2022).
These "bad things" run the gamut from extortion, to kidnapping, to human trafficking, to forced disappearance, which take place at the nexus of drug-trafficking organisations (primarily, the Zetas cartel and their successors), state officials and corruption. The city entered the notoriously bloody rule of Governor Javier Duarte Ochoa from 2010 to 2016. As International Crisis Group (ICG, 2017: i) writes: 'In Veracruz, an alliance between criminal groups and the highest levels of local political power paved the way to an unbridled campaign of violence through the capture of local judicial and security institutions, guaranteeing impunity for both sides.' In fact, in 2017, one of the ICG's recommendations to break the cycle of violence, impunity and state-criminal collusion in Veracruz was for the state government to 'admit to and thoroughly investigate state involvement in crimes against humanity such as forced disappearances' (ICG, 2017: ii, emphasis added).
Disappearances are widespread across Mexico; the country has officially registered 100,000 cases (Lopez, 2022). It is only recently that official statistics are beginning to emerge on the numbers of disappeared in the country; civil society and academic organisations (see Observatorio sobre Desaparición e Impunidad en México, 2022) are clear that the official government numbers grossly under-represent the real figures. It was in 2010, when Governor Duarte came to power, that Veracruz's worst phase of violence and disappearances began; the government arrested several Zetas leaders, which led to the splintering of the cartel into smaller groups that competed violently for territorial control. The ICG identifies different strategies of violence at the time, which include inter-cartel violence (with state involvement), violence in the context of kidnappings for ransom, violence against accidental witnesses and 'violence against civilians to create a general climate of fear and thereby compliance', among others (ICG, 2017: 17). In this context, the report identifies techniques to cover up crimes committed by the Duarte administration, including state officials refusing to file missing person reports (or attempting to dissuade family members from doing so), shifting around cases between multiple officers to limit progress on investigation, prematurely closing cases, refusing to act on clues, leads or intelligence about clandestine graves, destroying evidence (including the illegal burial of unregistered corpses), and slander against disappeared people (ICG, 2017: 18-19). Importantly, another strategy involved 'threats against victims' organisations and family members carrying out their own investigations, as well as against witnesses, both by organised crime groups and the political officials', and 'the criminalisation of civil society-led search of clandestine graves under the argument of undue interference' (ICG, 2017: 18).
Violence before and after the disappearance of a loved one 'My two sons went to a football game and they never came home from the stadium.' (Interview, 2 March 2022) 'My son studied in Puebla. He came for the weekend and went to a party. People who were there said that a white truck with tinted windows showed up, armed men got out and then took my son and his friend away.' (Interview, 3 March 2022) The stories are painfully similar; young men and women 6 who went to a party, went to have a beer with a friend, or went to work and never returned. Their mothers told me, we are "muertas vivas" ("the living dead") (interview, 2 March 2022), unsure if their children are alive or dead, if they will one day come home, or if they will find their bones in the excavations they undertake in Veracruz's mass hidden graves. In this section, I detail the various forms of violence that the madres experience, which result in their activism being high risk: when their children go missing, when they begin to search and when they join the colectivo. I frame this violence using a definition of multisided violence as a 'potent combination of structural, symbolic, political, gender and gendered, and everyday forms of violence' (Walsh and Menjívar, 2016: 586).
Upon realising that their loved ones were missing, most of the madres went to the state Fiscalía ('Prosecutor's Office'). They reported their disappeared relative and opened a case. Of those who reported to the authorities, however, few have ever received any clues or information about their cases. One woman told me: "The Fiscalía takes your information, but then … nothing. You'd think that when you put in a claim, they'd go to look for him. But no" (interview, 3 March 2022). In the case of Ciudad Juárez, Orozco Mendoza (2019: 218) notes, the authorities ascribe blame to the victim: 'a woman's dead body was a testament, in the eyes of the authorities, to the victims' criminality', a guilt that is transferred to the 'families struggling for justice', thus justifying a perverse logic whereby the state both fails to search or investigate the case, and also deployed law enforcement against the families. As cited earlier in the ICG report, this dynamic holds true in Veracruz. Others did not report their cases immediately. They were afraid of violent repercussion from the same actors who had taken away their children; some expressed that they did not register the cases to protect their other children.
Even before the disappearance of a relative, however, women reported that extreme violence was common in their neighbourhoods. After the disappearance of their children, they felt the violence and social fragmentation more acutely. Indeed, one effect of chronic violence is a breakdown of trust between potential allies (see Moser and McIlwaine, 2001): the madres told me that there is a general consensus that if a child goes missing, "they must have been up to something [bad]" and thus deserved it (see also Centro Prodh, 2020). Being the mother of a disappeared person tarnishes a person; as one woman said: "I joined the colectivo because within my family, no one understood me" (interview, 2 March 2022).
The madres buscadoras report that after their family members go missing, they begin to experience fragmentation within the environment of the family along other lines of (structural) violence as well. Most of the women I spoke to come from poor families; becoming a buscadora is an expensive job, sometimes directly (in terms of transport, excavation supplies and legal fees) and always in terms of opportunity costs of not working in formal employment. As the mothers quit their jobs to search for their children, their remaining children and husbands often become angry and make demands that the women take care of "those of us who are still here" (interview, 2 March 2022). Indeed, becoming a madre buscadora ironically places women in a contentious position vis-a-vis traditional gender roles: as mothers, society expects that they fulfil maternal expectations and look for their children (be a 'good mother'); yet, at the same time, there are contrasting expectations that relegate them to the private sphere where they should take care of their husbands, other children and the grandchildren who are left behind by the disappeared. When this deeply personal site of security and community -the family -begins to crack, it exposes women to heightened anguish. It is perhaps not surprising that many women I interviewed spoke about the physical illnesses that they had developed after their children went missing; from diabetes, to blindness, to memory loss, they considered these ailments to be physical manifestations of their suffering. 7 Moreover, there are direct risks linked to being involved in the colectivo. As one woman told me: "[They] don't like it when you start to kick up dust" (interview, 3 March 2022). As a leader who no longer lives in Veracruz after receiving multiple threats from criminal organisations and directly from the former state governor told me: "I feel like a sitting duck" (interview, 15 February 2022). Other women told me about personal threats they had received; after joining the colectivo, one woman received a series of angry-face emojis that had been sent from her missing son's phone. She took this a sign that "the bad people" were watching her and were threatening her to stop her search. Another leader described that in 2014, she was driving on a highway when her car was cut off by a truck; a group of armed men got out of the car and pointed guns at her. They told her that she needed to stop searching for her son and that she had "been warned".

Searching for the disappeared
Despite the threats they face, women continue to search for the disappeared. Across Mexico -including in Veracruz -colectivos of family members have coalesced in the search for their loved ones (see Gallagher, 2022). Arely Cruz-Santiago (2017: 13) writes about how the relatives of the disappeared engage in 'a grassroots knowledgemaking process permeated by loss, care, and love'. She focuses on how they transform themselves into 'forensic citizens' who make rights claims for themselves and disappeared family members.
For the most part, these colectivos are made up of women, and primarily of mothers. A report by a Mexican human rights organisation identifies why it is usually women who search, primarily linking the phenomenon to traditional gender roles, where women are carers and men are economic providers (Centro Prodh, 2020: 31). While this report briefly mentions the risks and threats associated with searching for truth and justice -for example, when women discover cases of collusion between state officials and organised crime (Centro Prodh, 2020: 40) or patterns or plots of organised crime (Centro Prodh, 2020: 58) -the high-risk nature of the activism is largely unmentioned. When I asked a leader in Veracruz why it was mainly women in her colectivo, she told me: 'The [politically correct] answer is that the men support the women directly by working and making money for the family. But the real reason is not like this at all. The real reason is machismo; in Mexico, the men are the kings of the house … some men have even told their wives to stop searching so that they can take care of them instead.' (Interview, 15 February 2022) 8 Since founding the collective in 2013, the madres have found hundreds of human remains in the hills and swamps of Veracruz. In interviews, the women told me what this entails: they go to a particular plot of land and look for where the soil is uneven, then they take long metal spikes and push them deep into the ground. When they pull the rods out, they smell them for decaying matter. If they can detect the scent of decomposing bodies, they begin to dig, shovelling dirt onto wooden sieves so that they can sift out small bones. When they find a body, the women call the authorities. Speaking about this process, one leader said wryly: "The perpetrators can do whatever they want to the bodies, but us, they'll take us to jail" (interview, 15 February 2022). The women raise funds to pay for materials -including shovels and work boots -by holding raffles and bingos, and selling used clothes. Sometimes, they hire local labourers to help them with the heavy work of shovelling soil, but for the most part, the women are actively involved in the búsquedas ('searches'). 9 When I asked how they learned to search for their loved ones, multiple interviewees told me about an incident that took place during the colectivo's annual Mother's Day march in 2016. During the march, a piece of paper was passed between women until it made its way to the hands of the leaders. The paper included a map, which indicated a certain piece of terrain it suggested the mothers investigate. The land had supposedly already been dug up by state authorities, but when the women went and began to dig, they found the remains of 302 bodies. Allegedly, this piece of paper was given to the women by members of Los Zetas. The leader told me in an interview that when they began to work in this particular terrain, they hired a forensic specialist and also began to reach out to other colectivos around the country to learn how to properly engage in their search (see also Cruz-Santiago, 2017).
While searching for the bodies -what they refer to internally as their "tesoros" ("treasures") -of their missing and murdered relatives, they put themselves in grave danger. In this sense, the experience of the madres in the colectivo can be compared to the experience of the Madres de Plaza de Mayo in Argentina's Dirty War; in fact, the leader of the colectivo told me that in the aftermath of losing her son, she was inspired to action by reading about the Madres (interview, 23 February 2022). Widespread scholarship (see Navarro, 1989;Jaquette, 1994) has detailed how the Madres' activism -searching for their disappeared children during and in the aftermath of state violence -was deeply political. Yet, despite assessments that the mothers were able to protest the repressive state because of a unique political opportunity they created using their identities as mothers (Navarro, 1989), they were not necessarily safe and suffered attacks in the course of their activism. Similarly, the madres buscadoras also do political work. They exercise a particular type of complex gendered agency that both seeks justice in contexts of impunity and facilitates solidarity in contexts of social fragmentation. In the following section, I outline the ways in which their mobilisation is an embodiment of love and care, as well as injustice and outrage, and how these feelings facilitate a renegotiation of the hierarchies of fear that govern the violent contexts in which they live.

Solidarity in embodied love and care
Krystalli and Schulz ask us about how we could understand world making after violence if we take love and care seriously. The ability to love and care fall under direct attack when family members are disappeared. Thus, finding ways to resist this loss through the creation on new forms of solidarity can be considered a political act.
When I asked what solidarity meant to them, the madres said: "We are sisters who suffer the same pain" (interview, 2 March 2022); and "We all suffer, but we all have to be strong together" (interview, 3 March 2022). Such findings echo Marie Berry's work, which documents how when women learn of others' experiences of suffering, they find themselves 'compelled … to keep going and find ways to support those that were most vulnerable' (Berry, 2022: 12; see also Zulver, 2022). Goodwin and Pfaff's (2001: 285) work on the management of fear in high-risk social movements notes that activists have to 'manage, but not eliminate, their fears in more or less explicit and self-conscious ways'. Their research revealed encouragement mechanisms that mitigate fear, among them '"intimate social networks" that [underpin] these movements' (Goodwin and Pfaff, 2001: 287). These networks refer to extremely close personal ties between activists, which keep morale heightened, and make the cost of dropping out of the organisation high. In the case of the madres, maintaining these intimate, relational social ties is key to explaining why the women participate in high-risk social movements. Indeed, the main word they all used to describe why they joined the group, as well as why they continued their participation, was apoyo ('support'). This support ranged from the emotional to the economic. 10 Multiple interviewees told me that "when you fall, they pick you back up again" (interview, 2 March 2022), referring to how they feel able to speak openly about their moments of pain and trauma with the other women. This solidarity is prominently manifested through a WhatsApp group that the colectivo runs (and that had particular salience during the COVID-19 pandemic when in-person connections were limited). They use this digital space to share everything from photos of their children to special prayers on their loved one's birthday. Here, we see an example of what Berry (2022: 7) calls a 'radical' approach to resilience, whereby 'survival … becomes a collective project; the most mundane acts of … listening, sharing, and supporting become revolutionary acts of resisting domination, because they sustain those who have been harmed'.
To gain insight into the extent to which women felt committed to the colectivo in relation to solidarity, I finished every interview by asking if the interviewee would continue participating in group activities if she were to find her tesoro. Every single woman said yes. In their answers, they detailed that they could not give up on their "sisters" just because they found their "treasure". One said: "the colectivo is part of my family, and you don't abandon your family" (interview, 2 March 2022). Another said: "My struggle is for [all] of the disappeared, not just my son. It would be a selfish act to just search for one person" (interview, 23 February 2022). During the annual meeting, one woman who has found the remains of her daughter stood up to speak: "Unity makes us strong", she told the others (meeting, 9 May 2022). Later, in a private interview, she told me that she continues in her search for three reasons: "Empathy. Resilience. Solidarity" (interview, 9 May 2022). Such sentiments translate to other colectivos around the country; "their pain has become my pain" expressed a madre leader in Sinaloa when asked why she continued to search with other mothers after having found her son (Zatarain, 2020).

Expressions of injustice and outrage
Certain studies of high-risk activism have shown that people are more likely to participate -despite the potential risks of doing so -when they are able to frame their actions in terms of anger, rather than fear (Lerner and Keltner, 2001). Beyond the love and care that they show for their loved ones, and for each other, the madres are also outraged at what has happened to them. One key facet of their solidarity is related to "doing something" when the state "does nothing" (interview, 3 March 2022). Writing about the campesino insurgency in El Salvador, Wood (2003: 18) notes that for many, participation was an 'act of defiance of long-resented authorities and a repudiation of perceived injustices'. For those who have suffered, defiance is a reason to support mobilisation; it is both a 'refusal to acquiesce' and an expression of 'dignity and personhood' in the face of repression (Wood, 2003: 233). Beyond their grief, many of the madres are profoundly outraged not only that their children were disappeared, but also that the state is not taking steps to find them.
Such indignation and outrage were made particularly clear in the lead-up to the Mother's Day march. During their annual meeting, one of the leaders took the microphone to tell members that "the authorities are putting barriers in your way. They are steamrolling you. We need to apply pressure" (field notes, 2 May 2022). The meeting finished and I found myself bundled into the back of a taxi with three other women. We drove to the mayor's office, where dozens of colectivo members began to demand a meeting with the local official. They chanted and held up photos of their missing relatives. Eventually, they managed to unchain the heavy wooden gates and flooded into the building. Later, one of the leaders told me that the meeting itself "did not make much of a difference", but the symbolic act of storming the building was an important step, as it gave them visibility. Such sentiments are in keeping with studies that document how 'appeals to pain and a sense of injustice are a strategic resource for movements when they bring their social demands into the public arena, partly because of the moral force that these carry' (Ariza, 2021: 161).

Renegotiating hierarchies of fear
Studies on fear document that it is an isolating emotion: 'it makes us stay at home, diminish night-life activities, abandon parks, and suspect others' (Villarreal, 2015: 137). However, I argue that the madres' decision to engage in high-risk collective action -searching for their relatives -is a reflection of how they choose to re-rank their hierarchies of fear in a setting of chronic insecurity. Petersen and Liaras (2006) outline different tactics that can be used to overcome fear in war; in this article, I suggest that expressions of love and care, as well as injustice and outrage -when understood in relation to gendered violence -can be harnessed to overcome barriers to action. This process is collective; it is derived from the 'radical resilience' that is generated through solidarity (Berry, 2022). In the remainder of this section, I outline how, despite the risks they face, they find ways to negotiate hierarchies of fear to engage in acts of complex gendered agency.
The epigraph at the beginning of this article was spoken by one of the madres. In this statement, she characterises the fear of not finding her son as worse than the fear of violent repercussions for her activism. She was responding to a question I had asked her about whether or not being part of the colectivo was something that made her feel fear. Instead of answering directly, she described four different situations that had taken place in her neighbourhood; all of them involved witnessing acts of extreme violence. After a while, feeling fear becomes an everyday occurrence, one that she cannot fully mitigate as a poor woman living in a dangerous barrio. As she expressed, over time, "My biggest fear is not knowing whether my son is alive or dead." Others described their fears in terms of no longer having the ability to share their suffering. McIlwaine and Moser (2007: 117) write that 'little is known … about how the poor themselves actually perceive violence and its associated fear and insecurity'. At first glance, it may seem as though participating in highrisk activism means that the madres are not afraid of retributive violence. Yet, by listening to women's testimonies of participation, it becomes clear that they have developed different ways of prioritising and ranking their fear. The fear of violence for participating in the colectivo is not necessarily greater than the fear that they feel on a daily basis, living in hyper-violent neighbourhoods. Nor is it bigger than the fear of not being able to build new bonds of solidarity with other mothers. Nor does it outweigh the moral outrage they feel.
Therefore, it is not that the madres are engaging in foolhardy behaviour or that they are unaware of the risks they run; rather, they assess risk based on the realities and constraints of the 'chronic violence that surrounds them' (Hume and Wilding, 2019: 15). As outlined in the previous section, when their agency -their ability to mobilise -is driven by love and care, as well as moral outrage, women's capacity to negotiate and rank fear changes: they are more afraid of not participating than the participation itself. This dynamic reflects the explanations for why women participate in high-risk feminist movements in Colombia (Zulver, 2022): when the risks of simply being a woman in a violent context are high and inaction does not guarantee safety, participation (and the potential benefits that can be accrued through group membership) become comparatively justifiable.
To be sure, the women in the colectivo are at different stages of grief, and thus their ability to overcome or re-rank their understandings of fear are not homogeneous. Moreover, through observing meetings of the colectivo, it became clear that the discursive narrative of solidarity is a strategy imparted by the organisation's leadership to encourage ongoing participation. In reality, most of the women I engaged with continue to suffer depression and live in ongoing contexts of poverty and violence. Yet, even in private settings, many women spoke to me about no longer feeling fear in the same way since they joined the group. One woman told me that after losing her two sons, five years apart, she is not afraid anymore: "I've lost everything, the only person I fear now is God" (interview, 10 May 2022).
Understanding fear in this way -as a function of a context of fragmentation and where violence is effectively normalised -has the potential to deepen the ways in which we think about gendered complex agency. Mobilisation as madres is not necessarily solely about finding the missing family member; it is about showing love and care, and about searching for sources of solidarity in a high-risk setting where inaction does not guarantee safety. It is also a way of embodying outrage at the violence they have suffered. Women negotiate spaces in which they value solidarity and support more than they act on the personal risks of their activism.
One final example of how the madres negotiate hierarchies of fear is evident when they need to engage in uncomfortable interactions with purveyors of violence. For example, despite being given 'permission' to dig in one area (as outlined earlier), the leader told me that the búsquedas make the women visible and put them in direct danger at the hands of violent actors, who sometimes appear in armoured cars and holding weapons as a sign of intimidation. Even in the new space where they are searching for remains, in a swampy terrain outside the city, members of their team have been threatened with violence by armed men if they stray beyond the acceptable perimeters. Another example is evident in the experience of the leader mentioned earlier who was stopped in her car. The assailants asked her how much money she needed to stop her search: "I told them, I have my price: bring me my son and forget about me forever" (interview, 25 February 2022). This willingness to both follow the 'rules' imposed by violent actors and sometimes even stand up to them directlyeven when they use threats to inspire fear -is also an illustration of how the madres manage their fears in a violent context.

Conclusion
In this article, I have detailed a theory of complex gendered agency in the Mexican context. For the women who take part in colectivos, solidarity is about making worlds make sense when senseless violence ends the life of loved ones. Women participate in colectivos to express gendered agency and do so even in the face of intimidation. To illustrate this theory of complex agency, I included empirical examples to document how women actively re-rank their hierarchies of fear based on the realities of the violent communities in which they live. Eventually, the fear of not mobilising becomes worse than the fears associated with taking action.
One of the inherent goals of this article has been to avoid instrumentalising the dynamics of women's emotions -and how these facilitate action -by implying that they are automatic or take place without critical reflection. I have framed these actions as an expression of complex gendered agency, whereby emotions like fear are shaped by unrelentingly violent neighbourhoods, yet can be re-ranked when conceptualised in relation to the possibilities for solidarity amid suffering. To take action in the face of fear is a choice; as the madres buscadoras chant during their marches: "Por qué los buscamos? ¡Porque los amamos!" ("Why are we looking for them? Because we love them!").

Notes
1 While it does not focus on madres buscadoras, Villarreal's (2022) work on the logistics of fear has useful insights for how people manage fear in high-risk environments over time. 2 Not all of the women in the colectivo are mothers searching for disappeared children, though the vast majority are. Increasingly, wives, sisters and daughters of the disappeared are also participating in the search for missing loved ones. Much of the uniting narrative of the group, however, centres on motherhood. 3 Exceptions to this trend are found in Wright (2006), Orozco Mendoza (2019), Zubillaga and Hanson (2020) and González (2022), though these are not necessarily framed in terms of high-risk collective action. 4 Orozco Mendoza recognises this link in her work in Ciudad Juárez; she notes that to explain the emergence of mothers' ongoing struggle for justice, one needs to explain the 'context of extreme violence … and the activist response it engenders' (Orozco Mendoza, 2019: 215). 5 This project was approved by the Oxford Social Sciences and Humanities Interdivisional Research Ethics Committee (Ref: R60335). 6 In some interviews, women told me that there are different stigmas associated with having a son versus a daughter who is disappeared. For more on the disappearances of women in Veracruz, see (Ernst, 2017). For more on the stigmatisation of murdered women, see Orozco Mendoza (2019). 7 For more on the types of 'feminine emotional distress' women suffer in contexts of poverty, see Enríquez (2009). 8 I interviewed a few fathers who participate in colectivos, who are also dedicated to finding their missing children. A discussion of the discrimination they face because of this participation -due to what are seen as culturally acceptable behaviours for fatherswould make an interesting separate study. 9 For a detailed and thoughtful account of the process of searching for the remains of loved ones in Sinaloa, see Zatarain (2020). 10 There are also certain economic benefits that come from being part of the colectivo.
For example, the national Victims' Commission gives certain financial resources to the family members; these include money for food, scholarships and rent. As a leader said, however: "When you give resources to people in the struggle, it's pure poison; you lose the integrity of the struggle" (interview, 23 February 2022).

Funding
This work was supported by the European Union's Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie Grant Agreement No. 838513.