Skip to main content
SearchLoginLogin or Signup

Introduction

Published onNov 02, 2022
Introduction

In 2017, a colleague sent me an article from the citizen media outlet Global Voices. The headline read: "One Woman Is Behind the Most Up-to-Date Interactive Map of Femicides in Mexico"1. This was my first encounter with the work of María Salguero and I was both stunned with sadness and painfully curious at the same time. Since 2015, this Mexican activist has spent hours a day scanning news reports, digging into government websites and investigating crowdsourced tips. Her sole focus is logging feminicides – gender-related killings of women and girls. For each feminicide, Salguero plots a point on the map and logs up to 300 fields in her database, everything from name, age, whether the victim was transgender, and relationship with the perpetrator; to mode of death, the case status in the judicial system, and whether organized crime was involved. She also includes the full content of the news report where she sourced the information. María's map is called "Yo te nombro: El Mapa de los Feminicidios en México" ("I name you: The Map of Feminicides in Mexico") because one of her many goals is to show that each murdered woman or girl had a name and a life and a place and a community (figure 1a and 1b)2.

A map of Mexico with colored markers to represent locations where femicides have occurred. The color of the marker corresponds to the year in which the femicide occurred: red for 2016, purple for 2017, and light blue for 2018. There is an immense concentration of femicides near southern Mexico, and they become less concentrated further away.
A zoomed in version of the femicide map over Ciudad Juarez, a Mexican city just south of El Paso. A purple marker (representing a femicide of a trans woman from 2017) is selected and a description box to the right of the map contains information about the attack, including its date & time, its location, and a brief description. The description box reads the following: 

Nombre (Incident Title)
#Transfeminicidio Identidad Reservada

Fecha (Date)
15/08/2017

Lugar (Place)
Pedro Meneses Hoyos, Ciudad Juárez, Chihuahua, 32730 México

Hechos (Description)
MARTES 15 DE AGOSTO DE 2017 | POR EDITOR 12
Juárez, Chih.- Un individuo que aparentemente pertenecía a la comunidad LGBT fue localizado sin vida por la noche en un fraccionamiento ubicado al sur oriente de la ciudad, reportaron las corporaciones policicas. 
El cuerpo del hombre vestido de mujer y en avanzado estado de descomposición fue encontrado en el fondo de un pozo de contención de aguas pluviales. 
El occiso tení una bolsa de plástico en la cabeza, aunque personal de la Fiscalí General del estado asegura no le pudieron encontrar huellas externas de violencia. 
Al lugar de los hechos llegaron sus familiares y lo identificaron como Hilario Lopez Ruiz, de quien no se proporcionó más información. 
El cuerpo fue enviado al Servicio Médico Forense donde se le practicara la autopsia de ley y determinar de esa manera las causas reales de su fallecimiento.
Figure 0.1 María Salguero’s map of feminicides in Mexico (2016–present). (a) Map extent showing the whole country. (b) A detailed view of Ciudad Juárez with a focus on a single report of an anonymous transfeminicide. Courtesy of María Salguero.

María estimates that it takes her three hours a day, every day, to scan her daily sources, verify information and organize it into structured database fields, "It's just that it’s a mess of information from all the tweets and all the things I get sent."3 She takes breaks for mental health, but the work is still excruciating and massive in scale. In 2019, she logged 2,900 feminicides, an average of eight killings per day. Meanwhile, the Mexican government logged 1,0064. What accounts for this significant discrepancy? It's hard to know because despite the fact that Mexico has a law defining and criminalizing feminicide, which additionally guarantees women's rights to live a life free of violence, the official data are not publicly available in any disaggregated form5. As a result, María's unofficial database has become the largest publicly accessible dataset of feminicide in Mexico. Her map has been used to help find missing women, and María herself has testified multiple times before Mexico’s Congress about the findings of her work. She has made her data available to journalists, activists and nonprofits to support their efforts. Parents of victims have called her to give their thanks for making their daughters visible. María is honest about the broad aims of her work: “This map seeks to make visible the sites where they are killing us, to find patterns, to bolster arguments about the problem, to georeference aid, to promote prevention and try to avoid feminicides.”

But the question remains: why is this work left to one person?

In Data Feminism, Lauren Klein and I frame data about feminicide as missing data – the state and its institutions systematically ignore the phenomenon, neglect to count and register cases, and often neglect to conceptualize gender-related violence in such a way that it could even be counted precisely6. And thus it does not count. When the state and its institutions fail to collect important information, activists and civil society organizations are increasingly stepping into those data gaps and producing their own counterdata. As Alice Driver notes in the case of Mexico, “the most accurate records of feminicide are still kept by individuals, researchers, and journalists, rather than by the police or a state or federal institution.”7

This is a book about counterdata in relation to feminicide and gender-related killing. It is about the potential but also the limits of counterdata science – data production, analysis, visualization and circulation of data that happens outside of – and often in opposition to – mainstream counting institutions like governments and corporations. It is about feminist activists who use counting to make things count, but then must deal with the reductionist consequences that follow from reducing complex social phenomena into convenient, sortable, aggregable forms like numbers and database fields. In the famous words of Audre Lorde, "the master's tools will never dismantle the master's house."8 What are the limits of using the "master's tools" of counting and quantification when the goal is nothing short of the elimination of gender-related violence? This book will return to this question, for it is a central tension of using data science for social justice.

María Salguero's work has received an immense amount of media coverage. She has even become something of a folk hero in Latin American feminist movements which have been mobilizing fiercely against feminicide in the past decade. On March 8, 2021, International Women's Day, activists took to the streets in Mexico City to cover existing street names with names of notable women, victims of feminicide, women's groups and feminist events. At least two separate streets were named for Salguero and the results were showcased on social media with the hashtags #TomaLasCallesNoCalles and #LasCallesTambienSonNuestras9 (figure 0.2).

Figure 0.2 On International Women's Day, March 8, 2021, activists temporarily renamed several streets to honor the work of María Salguero. Courtesy of Fátima Araujo.

María's work deserves to be venerated. But her efforts are not unique. Over the last several years, my colleagues and I have cataloged more than 150 grassroots projects around the world working to challenge feminicide and gender-related killings by painstakingly scanning news reports and social media feeds, triangulating with official data, talking with families and friends of loved ones, and logging details of people’s murders into digital maps, spreadsheets and databases. This book is an exploration of their efforts, and how their data practices challenge conventional ideas about data science in the world today. It makes the case that feminicide data activists provide a powerful model for how data may be used in the service of justice.

What is Feminicide?

But first, what are we talking about when we use this word? Feminicide (or femicide) is the misogynous and gender-related killing of women. It is used to denote domestic violence or intimate partner violence that is fatal, and also murders perpetrated where a woman's gender, and her gender subordination, are part of the motivation for the crime. Both feminicide and femicide are evolving concepts whose exact definition and scope shifts across contexts, and you will see this surface in the various definitions of what "counts" as feminicide used by activists in this book. Yet despite this variation, feminicide and femicide are increasingly mobilized in legislation, national statistics and activism10. The term “femicide” emerged from the feminist work of Radford and Russell11, where they define femicide as a form of sexual violence that includes verbal, physical, visual, and sexual forms of abuse. Building on this work, Latin American activists and scholars introduced the term feminicidio as a way to capture the role of the state in enabling violence against women through either omission, negligence or complicity12. In recent years, feminicidio has traveled back into English as feminicide, as a way of capturing the public nature of such violence – basically implicating states for widespread failure to ensure women's basic human right to life. Feminicide, in other words, asserts that this violence is not a "domestic" or "private" problem, but a public problem in which the state is complicit. In their 2010 work Terrorizing Women, Fregoso and Bejarano define feminicide as (1) the murders of women and girls founded on a gendered power structure; (2) both public (implicating the state) and private (implicating individual perpetrators); and (3) intersecting with other inequalities. They write, "the focus of our analysis is not just on gender but also on the intersection of gender dynamics with the cruelties of racism and economic injustices in local as well as global contexts."13 This is the formulation of feminicide that I will use throughout this book.

Activism on the issue of feminicide from feminist and women’s movements, particularly across Latin America, has played a fundamental role in raising awareness and promoting policy change globally. Most notably, mobilization around the disappearances and murders of women and girls in Ciudad Juárez, Mexico since the 1990s and the powerful demonstrations held by the “Ni una menos” (“Not one [woman] less”) movement in Argentina and across the region since 2015 have brought worldwide attention to the issue of feminicide and contributed to the formulation of specific legislation and national plans for the eradication of violence against women14. The vast majority of Latin American countries now have a law defining feminicide or femicide and dozens of grassroots efforts to monitor the phenomenon have proliferated in the past decade15.

Because Latin American feminists and women's groups are leading this work, the geographic center of gravity in this book is focused on Latin American data activism on the topic of feminicide. As you will see, the book also intentionally weaves in feminicide data activism from groups based in the U.S. and Canada for three reasons. First, there is a equally long legacy of activism in North America organizing against fatal violence against Black women as well as missing and murdered Indigenous women and girls and two spirit people (MMIWG2S). Among many other efforts, we can see this in the Women's Memorial March which started protesting and memorializing the killings of Indigenous women in Downtown Eastside Vancouver in 1993 and continues to this day. Second, as I began to speak publicly about this work, I realized that there is a risk of people from Anglo-America dismissing feminicide as a problem "down there." This reaction is both ignorant of and incorrect about the current realities of feminicide, MMIWG2S and gender-related violence in North America. So one reason this book focuses on the Americas is as a corrective to the gringo exceptionalism that tries to invisibilize feminicide in the Global North.

Third, the book focuses on data activism in the Americas broadly because feminicide data activists themselves are increasingly in dialogue with each other across national and regional borders. They participate in networks like the Interamerican Anti-Femicide Network (RIAF); they set up one-on-one meetings to share tips and strategies; and they draw inspiration from the categories, concepts or maps used by each other. This very book emerges from a South-North participatory action research project called Data Against Feminicide. This way of working is what the editors of the anthology Translocalities/Translocalidades frame as "translocal": an effort to create feminist and anti-racist bridges, crossings, connections and movements across situated and local subject positions in the Americas16. Instead of transnational governance or formal movement building at the scale of the state, this approach emphasizes the construction of relationships at the scale of local and grassroots perspectives. These relations are born from shared struggle but respect local differences and variation in culture, context and history. This is how feminicide data activists are in relation to each other.

Counting Feminicide

This leads us to the book's title and its central questions: Why count such fatal violence? What may be gained by systematically tabulating, monitoring, documenting, aggregating, analyzing and visualizing cases of feminicide and gender-related killing? How is it possible to seek justice through databases and spreadsheets, charts and statistics? What does counting do and who do we count for?

According to estimates by the United Nations, in 2017, 87,000 women were intentionally killed across the world. Nearly 60% of them were murdered by intimate partners or other family members17 . In the United States, reports state that around three women are killed every day by their current or former partners, though we know that these numbers are under-reported18. Every day in Mexico, ten women die violent deaths19. In Latin America and the Caribbean (LAC), every two hours a woman is killed in incidents related to her gender20. The vast majority of killings of women are perpetrated by intimate partners and family members. But other feminicides are perpetrated by organized crime and narcotraffickers, sex traffickers, and the state itself in the form of overpolicing, systemic economic violence, the social exclusion of LGBTQ+ people, and racialized violence visited on low-income communities. A lower proportion are random assaults by strangers. Beyond the loss of lives, these murders have deep ripple effects on the mental health and livelihoods of relatives and communities. Not only do they cause and perpetuate different forms of intergenerational trauma, but because women are so often providers of care or income or both, these murders leave their loved ones more vulnerable to food, health and housing insecurities.

Yet, as terrible as these existing statistics are, we also know they are terribly incomplete. Official government data on gender violence and feminicide are often absent, difficult to access, infrequently updated, contested, and underreported. The reasons for these gaps range from stigma and victim-blaming to matters of legal interpretation and plain old pervasive patriarchy 21. This is a phenomenon called missing data that we will explore at length in Chapter 2. While numerous reform efforts exist to address these gaps from a national and international perspective – through the creation of standardized definitions, data standards, indicators and indices – the central focus of this book is on grassroots and community-based efforts which count from within specific places and communities. Citizen data monitoring of feminicide has stepped into the data gaps that exist in many countries and regions around the world. Such projects have proliferated particularly since a 2015 report from the UN Special Rapporteur on Violence Against Women called for the creation of femicide observatories in every country in the world 22. An observatory is an on-going monitoring effort – versus a one-time research project – that works continuously to document human rights violations. Individuals and groups compile cases of feminicide from open government data sources and media sources; they verify those data; they analyze those data; they circulate visualizations, reports and statistics; they accompany families through the justice system; and they memorialize and humanize victims.

While many counterdata efforts have taken shape in the past decade, they follow in a pre-digital activist tradition started by individuals like Esther Chávez Cano in Ciudad Juárez23 and groups like Comuna Mujer of Montevideo’s Centro Comunal Zonal 9 in Uruguay24, Women We Honour Action Committee in Ontario25, the Combahee River Collective in Boston26, the Women's Memorial March in Vancouver27, and the Movimiento de Mujeres del Oro and the Movimiento de Mujeres y Justicia in Ecuador28. These projects used fax machines, pamphlets, newspaper clippings and hand-written accounting ledgers to document feminicides and gender-related killings, to protest and memorialize, and to fight for policy change and justice. Scholars Fregoso and Bejarano place these data practices in a long history of Observatorios Comunitarios (Citizens Observatories) in Latin America, which have served to engage communities in monitoring state agencies and using collective action strategies to hold them accountable29.

You will meet contemporary groups working in this tradition, and explore their motivations and their methods for producing counterdata. These efforts are mainly initiated and sustained by women and Indigenous people and queer people. Most projects are explicitly aligned with intersectional feminist values, feminist and anti-racist movements and/or movements for Indigenous rights. There is variation across the projects: they count and classify different things, their reasons for counting are diverse, and the uses to which the data are put are diverse. While the majority of activists use feminicide as their organizing concept, those that focus on gender-related violence as it intersects with white supremacy and/or settler colonialism and/or cisheteropatriarchy name that violence with more intersectional specificity. This includes Missing and Murdered Indigenous Women, Girls and Two Spirit People (MMIWG2S) or Black women killed in police violence, or fatal violence against transgender people. Indigenous women and Black women are at the forefront of grassroots efforts that monitor racialized feminicide. Despite this variation, all of these groups are joined in their determination to count and document one of the most egregious violations of human rights in the world today: the violation of the right to life, liberty and security of person30. As the Combahee River Collective stated in their march to protest Black women's murders in Boston in 1979: "3rd World Women: We Cannot Live Without Our Lives."

One thread common to all feminicide data activism discussed in this book is that counting is never neutral; it is a deeply political act. For that reason, I will mainly refer to counterdata production in this book versus counterdata collection, to keep at top of mind that the work being undertaken is not about a neutral observation of what is already present in the world, but about the deliberate crafting of political visibility and advancement of concepts like feminicide that can work to name and challenge structural political problems31.

Data Feminism in Action

If counting is not neutral, which other assumptions about data might we need to challenge in order to bend it towards justice? Which other standard operating procedures of the Western Anglo White Rich Cis Man data science might need to be disrupted and decolonized? In Data Feminism, Lauren Klein and I outlined what a feminist approach to data science might look like. We drew from intersectional feminist theory, activism and writing to outline seven principles for working with data in a feminist way: Examine power, Challenge power, Elevate emotion and embodiment, Rethink binaries and hierarchies, Embrace pluralism, Consider context, and Make labor visible (see the full list in the toolkit in Chapter 8).

These principles were designed to challenge what one might call hegemonic data science – mainstream data science which works to concentrate wealth and power; to accelerate racial capitalism and perpetuate colonialism; and to exacerbate environmental excesses and social inequality32. This is the data science described in pathbreaking works of public scholarship like Weapons of Math Destruction, Race After Technology, Automating Inequality, Surveillance Capitalism, Algorithms of Oppression, The Costs of Connection, and Artificial Unintelligence33. Hegemonic data science can be directly implicated in feminicide and a number of other public emergencies if we draw from formulations of technical "glitches" by Ruha Benjamin and Meredith Broussard34. While glitches tend to be considered as temporary interruptions in a system, what if what they actually do is expose standard operating procedure? What if glitches actually "illuminate underlying flaws in a corrupted system"?35 Emergencies, glitches, crises – including feminicide as a crisis – these are not one-off, temporary accidents. They are entirely predictable outcomes of the way that our present political-economic system is organized to favor the lives, livelihoods and thriving of those at the top of the economic order and wage slow violence on the rest of us as the disposable middle and bottom. Hegemonic data science is the technocratic and ideological infrastructure of this extractive system. Theorists of feminicide and racialized violence call our attention to who is disposable, ignorable and disappear-able under neoliberal and extractive modes of governance and it turns out to be the majority of us36.

Yet, as media scholar Paola Ricaurte asserts in her call for decolonial approaches to data science, there are “possible alternative data frameworks and epistemologies that are respectful of populations, cultural diversity, and environments.”37. Some alternative epistemological approaches are being created and advanced in academia. I would situate data feminism here along with emerging and exciting work on feminist data refusal, radical data science, decolonial AI, Indigenous data sovereignty, queer data, Black data, and emancipatory data science38. But academia cannot and should not do this work alone. New media scholar Stefania Milan has spent the last decade theorizing and studying data activism. She and co-author Lonneke van der Velden call attention to the important role of people and communities outside the academy: "data activists function as producers of counter-expertise and alternative epistemologies, making sense of data as a way of knowing the world and turning it into a point of intervention. They challenge and change the mainstream politics of knowledge."39 Data activists are inventing and enacting alternative data epistemologies, new ethical visions, and expanded notions of citizenship in the digital age. We see this at work in the Indigenous data sovereignty movement, Data 4 Black Lives, the anti-eviction data activism taking root in the United States, in new communities that have emerged to connect data study and data practice such as Tierra Común, and, of course, in the central subject of this book: anti-feminicide data activism40.

As I learned more about María Salguero's work, and that of other data activists, I was struck by how their ways of working resonate with data feminism's principles. The role of emotion, care and lived experience is omnipresent (elevate emotion and embodiment). The approach to sourcing and verifying information is almost always pluralistic (embrace pluralism). Feminist data activists work in coalition and collaboration – though not always without conflict. The labor of counting is not always publicly visible but it is at the center of the internal work of activists (make labor visible). For me, different data feminism principles appeared to be more and less salient at different stages of a feminicide counterdata project. Data activists also navigate ethical issues not addressed by data feminism, such as consent to name someone in a database, which points to feminist concerns that the data feminism principles may have overlooked.

Throughout the book, I will highlight resonances and tensions between activist data practices and the principles of data feminism. As all practitioners know, practice is messy and rarely adheres cleanly to pleasing principles. As such, this book attempts to reflect on data feminism's principles as I, Catherine, see them surfacing in data activism about feminicide, and it can be thought of as an extended action-reflection on the frictions that arise when moving from speculative ethics to real-world relations. Here I wish to be careful about imposing data feminism on groups from the outside; from a Global North positionality; or from a white academic perspective. It would be tautological and misguided to think that I went into "the field" and Lo! I discovered data feminism! Instead, I will focus throughout this book on the idea of resonances and tensions in practices and principles, and on drawing out distinctions between hegemonic data science and counterdata science.

The stakes are high. There is currently enormous investment in the AI ethics and "data for good" space. Stanford and MIT are aiming to raise on the order of billions of dollars, foundations are creating funds in the millions, and Big Tech corporations are assembling large AI Ethics research teams (and then firing their Black and white women leaders who raise questions about potential harms)41. But perhaps these powerful institutions are looking to and investing in the wrong places for such ethical guidance. At best, these monies are funding work by well-intentioned people who need a lot more training in feminism, critical race theory, Indigenous and ethnic studies, queer theory, community organizing, post-colonial thought, social work, and other frameworks and fields of practice that rigorously tackle inequality and oppression. At worst, these are cynical efforts at ethics-washing42. Grassroots data activists at the margins – real-world people who are using data science in the service of real-world struggles for justice – have much more expertise to offer those of us trying to craft data frameworks and practices in the service of justice. Areas like health, housing, urban planning, policing, transportation and education are characterized by durable and extreme structural inequalities based on patriarchy, white supremacy, classism, colonialism and more. To work with data in these domains is, inevitably, to confront inequality and oppression. Grassroots organizations and movements know this and live this. They are far better equipped to deal with how power infiltrates data sets in these domains than today's hegemonic data scientists who are trained at elite institutions or in the elite institutions' money-making MOOCs. Activists and movements are increasingly challenging these inequities using data science as one tactic in a larger struggle. This book makes the case that feminicide data activists are at the forefront of a data ethics that rigorously and consistently takes power and people into account.

Hello Reader

Before we43 go further together, I would like to take this moment to introduce myself to you and explain how I got to the point of writing a book about feminicide and data activism. I often describe myself as a "hacker mama". I'm a hacker because I spent a good part of the first part of my career as a freelance software developer and database programmer. I am also a hacker in sensibility and method, where hacking means the "clever or playful appropriation of existing technologies or infrastructures or bending the logic of a particular system beyond its intended purposes or restrictions to serve one’s personal, communal or activism goals."44 Fundamentally, I have never believed that tech is deterministic. That is to say that just because technology is birthed from war and militarism and colonial violence doesn't mean that it cannot be employed towards serving as resistance to and liberation from the same. Over the course of my careers as a software developer, artist, and professor, I have sought to point out the power imbalances in tech; but also to find and use the cracks; to bend and borrow the tools; and to support the flourishing of alternative, feminist, humble visions for the role of data, information and technology in working towards social justice.

The "mama" part of hacker mama is because I am mother to three children and three cats and that takes up a whole lot of my time. "Mama" is also a relational term, existing in a relation of care with people and pets and places. These interconnections manifest as collaborations and I rarely work alone. This is my quiet but persistent refusal of the relentless individualism of academia, of the art world, and of Western society generally. I have worked under the banner of various collectives in the past, including iKatun, the Institute for Infinitely Small Things, and Platform2: Art & Activism. And I currently work behind the scenes to support various collaborative efforts such as Colab@MIT and Indigenous Women Rising (see the Statement on Sharing Power at the end of this book). Finally, I am also a "mama" in the sense of its gendered political identity. I care deeply about those experiences of mamas which are systematically overlooked and undervalued: birth, breastfeeding, and reproductive justice; care work; political work towards healthy, safe, thriving communities. In 2014 and 2018, I co-organized the Make the Breast Pump Not Suck Hackathon with a collective of awesomely wonderful women who I continue to admire. And while that work was about the beginning of life, and the present book is about lives cut short by violence, there are many resonances. Indeed, in much of the activism against feminicide and gender-related violence, mothers are out front and they have been organizing collectively for decades to demand systemic change.

Following a long period of freelancing and adjuncting, I have been full-time in academia since 2014 and I am currently a tenure-track professor at MIT in the Department of Urban Studies and Planning. Entering academia was not a given for me, since my training was mainly as an artist, software developer and design researcher. But once I started, I found that I love teaching and working with students. I love connecting with brilliant and committed colleagues. I love the intellectual and creative freedom. And I strategically sidestep and actively challenge some of the other parts of academia that make it a fundamentally conservative, hierarchical, individualistic, extractive enterprise.

I sit at the intersection of a number of dominant group identities: I am a white settler, cisgender, and mostly straight (or at least straight-passing) woman. I grew up in a middle-class family that moved around a lot, but mostly lived in the US South. I'm descended from working class Italian immigrants and some of the Italian side of our family emigrated to Argentina and, later, to Brazil. These family ties led me to live in Argentina for several years, both as a young person and a middle-aged person, and to develop my Spanish, which I speak (imperfectly) to my children and cats. This is part of the story of how I came to the work of data activism about feminicide.

While I researched María Salguero for Data Feminism, I became deeply inspired by both her work and the many-pronged efforts of Latin American feminist movements. My partner and I moved our family to Argentina in 2019 where I was able to attend gatherings, interview activists and data journalists, and make connections to feminist groups. I enter the topic of feminicide with a distinctly North American perspective, and you will see that surface in this book. I enter with appreciation for everything that the framing of feminicide as a feminist concept has been able to accomplish in legal, activist, policy and intellectual circles. And I enter with questions about the boundaries of that frame and whose killings may be sidelined or made invisible within our current formulation of feminicide, to name a few, Indigenous women or sex workers or trans women or Black women killed in police violence. Patricia Hill Collins teaches us that violence is a "saturated site" – a place where intersectional forms of domination are rendered most visible45. Obviously violence against women involves patriarchy and sexism, but how do these intersect with settler colonialism? With anti-Blackness? With cissexism? With the extractive logics of neoliberal capitalism? I hold this appreciation and curiosity about feminicide as a concept together with a commitment to intersectional thought and action. It turns out that many of the data activists working on feminicide also ask these questions.

Data Against Feminicide

In May 2019, I took a bus across Buenos Aires to meet Dr. Silvana Fumega, Research and Policy Director of the Latin American Initiative for Open Data (ILDA)46. I didn't know quite what to expect, just that three people in my network had told me that if I was interested in learning about feminicide and data, I must meet Silvana. From this very first meeting, which actually turned into a brainstorming session, Silvana and I drank coffee, shared interests, and discussed ideas around using machine learning to detect news articles about feminicide. Since 2017, Silvana had been leading the development of a Latin American data standard for feminicide data collection47. At the end of our meeting, she suggested we invite Helena Suárez Val into our emerging collaboration. Helena is a feminist activist, researcher and writer who has been producing data about feminicide in Uruguay since 2015 and was researching how feminicide data circulate on social media through her doctoral project. She had recently written her masters thesis on feminicide data activism and told us about the joy and connection she found by convening a group of feminist data activists to discuss their work48. Remarkably, Helena, Silvana and I have formed a tight and thriving collaboration (and friendship) despite the fact that, mainly due to COVID, all three of us did not meet together in person until the spring of 2022.

Three women standing together in the middle of a walkway smiling for a photo. There is a green canopy of trees lining the walkway behind them.
Figure 0.3 Silvana Fumega, Helena Suárez Val and I worked together for three years prior on the Data Against Feminicide project before we were able to meet in person in Montevideo in Spring 2022. Courtesy of the author.

Our collaboration evolved to be named Data Against Feminicide49. It is a bit of a sprawling project, as are our conversations and brainstorms, but perhaps best characterized as a South-North action-research collaboration across activism, academia, and civil society. We outlined three goals for the project, each stemming from an area that one of us had been working on:

  1. To foster an international community of practice around feminicide data

  2. To develop digital tools to support activists' production of feminicide data from media sources

  3. To support efforts to standardize the production of feminicide data where appropriate

The Data Against Feminicide project does not collect or aggregate activist data50. Our three goals are focused on supporting and sustaining the already existing practices of activists who care for femicide data in their own contexts. Since 2020, we have organized annual convenings around the International Day for the Elimination of Violence Against Women. This day of awareness falls on November 25 and was designated by the UN as a remembrance for the three Mirabal sisters – "las Mariposas" – from the Dominican Republic who were assassinated by order of dictator Rafael Trujillo in 196051. Our Data Against Feminicide virtual events have featured panels with data activists, workshops and hands-on learning sessions, brainstorming and network building opportunities and short talks by academics and graduate students52. Hundreds of participants have attended these events, from 38 countries and five continents. The primary language of the events is Spanish, with live interpretation in English and Portuguese to support participation from across the Americas. Between annual events, we host a network directory, email list, and Slack channel so that participants may connect with each other.

In Spring 2022, we experimented with creating a more formal learning structure and, through ILDA, offered a nine-week online course in Spanish about feminicide and data53. Around 120 participants from across Latin America participated in the asynchronous learning experience, with the first four-week module focused on theory and the second on practice. We read theoretical and empirical texts about feminicide together, ran tutorials about pivot tables, analyzed news articles, and engaged in lively discussions in the course forum.

In collaboration with my lab and students at MIT, we have also been co-designing and piloting digital tools and technologies to support the counterdata production work of activists and civil society organizations. Data Against Feminicide has released two free tools for activist use, and the process of creating these is the subject of a case study in participatory design in Chapter 7. The way I think about our events and tools is that our community-building work is about participating in a translocal social infrastructure around this work and the tool development is about co-creating translocal technical infrastructure. In both cases – social and technical – the infrastructure is deliberately small-scale, non-centralizing, and consists of tools and relations that support and sustain (not outsource, not automate) the difficult labor of activist data production.

To make the past three years of community-building and tool-building possible, we needed to better understand activist data practices: How widespread are practices of counterdata production about feminicide and gender-related violence? Why do groups and individuals begin their monitoring projects? What sources do they use for information? How do they classify and categorize cases? How do they publish and circulate their data and with what impacts and effects? What challenges do they face in collecting, analyzing and using the data they produce?

Helena had answers to these questions for herself. Since 2015, she has been logging cases of feminicide in Uruguay into a Google spreadsheet and publishing them on her website Feminicidio Uruguay54. The project started during a period of intense feminist organizing in Uruguay – there were mobilizations around the country to prevent the overturning the country's law legalizing abortion as well as increasing attention to feminicide. Helena, at the time working with a group called Cotidiano Mujer (Everyday Woman), remembers that they were collectively outraged by a series of murders of young women in late 2014 as well as by the sensationalist, victim-blaming media coverage of their killings. The group started going out into public space to protest each killing and Helena began the spreadsheet primarily as a record of protests. Eventually, the collaboration dissolved, but Helena kept recording each new case as it was reported in the press. For each violent death, she logs details in her spreadsheet and on her map, and then also publishes those to Twitter and Facebook. The work – of reading misogynist news reports about women who were beloved to their families or communities – is exhausting and she has often contemplated ending the monitoring, but has not55.

In order to understand Helena's experiences in relation to others in our emerging community, we began first to keep a list of all of the feminicide data activist groups we could find. We started with Helena's prior research into other feminicide mapping projects56. As we read research papers or hosted community events, each time we learned about a group who is monitoring or has monitored gender-related killings we would do research about them online and add them to our list. This informal spreadsheet of counterdata producers has now been formalized and expanded to include more than 150 efforts from around the world. It includes groups who document and produce data about cases of feminicide, femicide, MMIWG2, LGBTQ+ killings, Black women killed in police violence and other forms of gender-related killings. We also started, in 2020, to conduct in-depth interviews with data activists in the Americas. At first we interviewed ten groups. Then we began appreciating how truly widespread across the Americas this form of data activism was. Ten interviews became twenty, twenty became thirty, and to date, we have interviewed 35 individuals or groups.

We have published about Data Against Feminicide individually and collectively, with an eye towards supporting publications led by each other and supporting our students and project partners to take the lead on publications as well57 [CITE]. With the blessing of my collaborators, I proposed to document our learnings from the interviews and larger list of data activist projects, and this book is the form that has taken. Throughout the book, you will see work by Silvana, Helena and Data Against Feminicide surface many times as an object of reflection and discussion. Their prior work and our collaborative work together has been formative in shaping this book and indeed the book could not exist without the learnings that I have gained from our conversations and friendship, and without its grounding within the Data Against Feminicide project and community as a whole.

In fact, this book draws significantly from two collaborative projects – Data Against Feminicide and Data Feminism. In order to do right by my collaborators, as well as by the feminicide data activists who we interviewed and who we continue to be in community with, I set up a Community Peer Review Board for this book. Members of this board include Silvana, Helena, Lauren F. Klein, Annita Lucchesi from Sovereign Bodies Institute, Debora Upegui Hernandez from the Observatorio de Equidad de Género Puerto Rico, and Paola Maldonado and Geraldina Guerra from the Alianza para el Monitoreo y Mapeo de los Feminicidios en Ecuador are part of this board and participated in reviewing and providing detailed feedback on the manuscript. Everyone was either paid for their time or else we did a labor trade. While traditional peer review is designed to provide expert scholarly feedback, the purpose of the Community Peer Review Board, for me, was to invite feedback from the people that I am in community with in the process of doing this work. They are the ones who I feel most accountable to, as well as the ones who are best positioned to correct, question or challenge my interpretations or usage of our shared work. Appendix II describes this board and our process in more detail.

Overview of the book

Counting Feminicide is divided into three major sections. The first section provides background on some of the major concepts in the book. In A Short Genealogy of Feminicide and Data Activism (Chapter 1) we will explore how the concepts of femicide and feminicide evolved through multiple South-North border crossings. Counting feminicide is not a new activist practice, and we will visit several historical examples of documenting cases and situate them in relation to literature on data activism. Official Data / Missing Data / Counterdata (Chapter 2), describes a case study in Puerto Rico as a way to examine missing data – why official information about feminicide is often absent, sparse, incomplete, unreliable and contested. This chapter introduces the idea of counterdata and characterizes the data activists who produce counterdata, with a focus on data activists in the Americas.

The second section of the book is called Anatomy of a Feminicide Counterdata Science Project. The four chapters in this section are empirical, meaning, they come out of our research team's in-depth interviews with 35 data activist groups and individuals. From our qualitative analysis, the themes of resolving, researching, recording, and refusing and using data surfaced as common workflow stages for counterdata science projects about feminicide and gender-related killing. Each of these stages of work is the focus of a chapter in this section and the collaged diagram in figure 2.4 links these stages together. Resolving (Chapter 3) addresses activists' motivations for starting a database of cases about gender-related killings. This includes their framing of the problem, their theory of change, and why they believe counting may help to challenge power in their context, as well as the limitations they see for what can be accomplished with data. Researching (Chapter 4) is the process of seeking information and discovering relevant cases of feminicide or gender-related killing for groups' databases. This theme also includes on-going research to follow cases as more information develops or as criminal charges are pressed and a case proceeds through the legal system. The stage of recording (Chapter 5) relates to the process that data activists use for information extraction – typically moving from the unstructured text of press reports, social media or personal exchanges – into structured datasets, the verification and management of those data, and the classification of cases according to diverse typologies and the management of data. The recording stage is where activist conceptions of gender-related violence interact with legal, regional and/or international standards. While some groups use the laws in their country to determine if a case is feminicide, others deliberately count feminicide with a more expansive conception of the phenomenon.

Refusing and using data (Chapter 6) is the final stage of a counterdata science project about feminicide and refers to where counterdata go and who uses them and towards what ends. This chapter surfaces five major ways that counterdata are used to contest feminicide. Activists that engage in repair work use their data to provide direct support and services to relatives and communities who have lost beloved members. There are a variety of ways that counterdata production serves activist efforts to remember and memorialize killed people. Often these take creative forms, like videos, music, plays, and art installations that tell stories that seek to bring life and humanity to individuals, and try not to reduce a woman to her death. Many activists are focused on narrative change efforts to reframe feminicide. They use data to communicate about lives lost on social media, or through data journalism. These accounts challenge the cultural stereotypes around the killings of women as isolated events perpetrated by pathological individuals and reframe them as a structural phenomenon and matters of public concern. Counterdata are also used to push to reform existing institutional practices around feminicide, and there are examples of collaboration and communication between government and activist groups around the production of feminicide data. In these cases, counterdata and official data begin to mix in fascinating and sometimes uneasy ways. Finally, there are efforts to revolt – to use counterdata to support large-scale mobilizations, usually in conjunction with social movements tied to specific political demands.

Each of the chapters in Section II: Anatomy of a Feminicide Counterdata Science Project focuses on the grassroots data practices of feminist, women, Indigenous and queer activists to produce and circulate data about feminicide and gender-related killing. These practices, and the goals and values and relations that motivate them, are often radically different from those of hegemonic data science. So, at the end of each of these chapters, I try to draw out key differences between counterdata science and hegemonic data science as a way of understanding alternative epistemological approaches to data science that are grounded in care, memory and justice.

In the third and final section of the book, Section III: Action-Reflection, I model some ways to carry forward and apply what our research team learned from our interviews. In Co-designing For Counterdata Science (Chapter 7), I present a case study about our how our Data Against Feminicide team has used these findings along with the principles of data feminism to undertake participatory design of digital tools and machine learning systems together with data activists. We have developed and piloted two tools, in three languages, with the goal of supporting and sustaining counterdata production across the Americas. This chapter describes our design process and how we used the principles of data feminism as guideposts, putting our work in dialogue with other justice-oriented work in human computer interaction (HCI). It also surfaces key unresolved tensions that principles alone are insufficient to address, and details how we navigated some of the intersectional shortcomings of our first designs.

The final chapter in this section and in the book is A Counterdata Science Toolkit (Chapter 8). While this book is focused on data activism about feminicide, there are many other domains in which data activism is happening. These include monitoring other forms of human rights violations, tracking and contesting evictions, monitoring government spending, logging voting rights violations and tabulating extractive environmental harms, to mention just a few. Chapter 8 is geared towards activists, citizen data scientists, data journalists, community-based organizations, researchers, urban planners and practitioners who aspire to produce their own counterdata science projects. It is a first pragmatic foray into drawing out some lessons from the epistemological approach and the data practices of anti-feminicide data activists and speculating how these may be useful to data activists working towards social justice in other domains.

My hope is that this toolkit, like the book itself, can work as a first step towards the study and practice of counterhegemonic uses of data science and technology design, and specifically to those uses that support the creation of intersectional, anti-racist, anti-ablist, decolonial, abolitionist, Black, Indigenous, queer, and feminist futures. The essential tension of counting and producing counterdata about feminicide is that we – and here "we" means myself and my collaborators, you the reader, and the activist groups and individuals that I profile in this book – we seek nothing less than the complete eradication of the phenomenon we are counting. We want the world where no one counts gender-related deaths because our political economy ceases to produce gender-related violence as a logical and completely predictable outcome of its white supremacist, extractivist, patriarchal, settler colonial economic logic. This is the world to fight for: the world where counting is not necessary and serves no purpose. We will hold this tension throughout.

Comments
47
?
Susana Galan:

At the EU level, the concept of femicide is no longer geographically limited to the Global South but has been incorporated as one of the main metrics to measure gender-based violence (see https://eige.europa.eu/gender-based-violence/femicide). Defined as “the killing of a woman or girl because of her gender,” it is stated that it can take different forms, “such as the murder of women as a result of intimate partner violence; the torture and misogynist slaying of women; killing of women and girls in the name of “honour”; etc.” This last part (honour killing), however, does connect to the idea of misogynistic (non-EU, immigrant) traditions and ideas, thus persisting on conceptually tying some forms of femicide to the Global South.

?
Susana Galan:

A relevant reference in relation to state complicity and impunity is Melissa W. Wright’s “National Security versus Public Safety: Femicide, Drug Wars, and the Mexican State,” in Accumulating Insecurity: Violence and Dispossession in the Making of Everyday Life, Shelley Feldman (Georgia University Press, 2011).

?
Susana Galan:

These definitions don’t seem to capture the fact that this violence ends in the death of the woman (and thus is at the end of the continuum of gender-based violence, to borrow Liz Kelly’s model)

?
Susana Galan:

Perhaps use parentheses to avoid the double em dash

Débora de Castro Leal:

maybe “alone” would be nicer than unique? it would highlight the potential for collaboration and solidarity in these efforts while leaving space for the uniqueness of María’s work and those of others, even if their efforts are similar?

?
Susana Galan:

Add comma

?
Susana Galan:

It may be worth explaining why you have opted for feminicide (the term more commonly used in Spanish) over femicide (more commonly used in English). Is it just because the book is focused on Latin America or are there broader underlying conceptual questions (i.e., in line with Marcela Lagarde’s analysis?

?
Sophie Toupin:

A great article by a south african scholar to add to your already great list in footnote 38: Adams, Rachel. “Can Artificial Intelligence Be Decolonized?” Interdisciplinary Science Reviews 46, no. 1-2 (2021): 176–97. https://doi.org/10.1080/03080188.2020.1840225.

?
Susana Galan:

many? several?

?
Shamina Vastani:

suggest using “That” instead of “This” demonstrative pronoun. This points to an idea or event occurring in the current or immediate future whereas “that” refers as in this case a future that we aspire to create. Use “a” instead of “the” after the colon the reason is the same as stated in previous comment by me.

?
Shamina Vastani:

suggest using indefinite article ‘a’ as it introduces an idea not mentioned before this sentence.

?
Shamina Vastani:

remove comma

?
Shamina Vastani:

remove commas

?
Shamina Vastani:

remove “our” before “how”.

?
Shamina Vastani:

“…. there is an…”

?
Shamina Vastani:

suggestion to rewrite as below as ‘whose’ is used to refer to ideas when it is supposed to be used to show relationship between two or more persons or a person and a thing. Ex; Whose car is? (relationship between person & thing). Whose is that? (relationship between person & person)

suggestion 1.

Feminicide and femicide are relatively newer concepts, i.e the exact definition and scope of each is still evolving and shifting across contexts. This will become increasingly apparent in the various definitions presented in this book that are used by activists of what “counts” as feminicide.

suggestion 2.

Both feminicide and femicide are evolving concepts the exact definition and scope of which shifts across contexts, and you will see this surface in the various definitions used in the book of what “counts’ as feminicide according to activists.

?
Shamina Vastani:

should be written as one sentence: “Maria’s work deserves to be venerated but her efforts are not unique.”

?
Alessandra Jungs de Almeida:

I really enjoyed this paragraph. I also liked the invisibilize word here. Maybe another dimension to analyze here, and I think this is the other side of “gringo exceptionalism,” is the idea of Latinxs as violent. Because it is not only about invisibilizing the feminicides in North America but also about putting LAC in a stereotypical place of violence and danger. I have experienced these stereotypical and usually not in-depth affirmations many times here in the US. This perception is also based on a colonial way of thinking (the white as reasonable, controlled and the Latinxs and women as emotional, dangerous, non-linear…).

?
Alessandra Jungs de Almeida:

maybe to cite the legal analysis table here?

Hana Gabrielle Bidon:

I’m not sure what the phrase down there means, but I’m guessing based on the context from the intro that it’s in Latin America and other countries not in the Global North.

Alan Colín-Arce:

Is there a way to join any of these channels?

Catherine D'Ignazio:

Yes! in fact we have two panels and two workshops coming up in late November - you can learn more about them and register here: https://docs.google.com/forms/d/e/1FAIpQLSdpcpUKnEgBGxddHVZMATLZ5vUJb-gJiyZspbh0nxL7s6sjTw/viewform

?
James Scott-Brown:

“an inveitable”?

”logical … outcome of its … logic” is redundant

?
James Scott-Brown:

It’s unnecessary to have a “we” both immediately before and after the parenthetical clause.

?
James Scott-Brown:

I’m not sure about the positioning of this paragraph break.

The previous paragraph outlines the first 3 chapters in Section II; I’m not sure why there’s a paragraph break before outlining the fourth.

Rather than combining the two paragraphs into one very long paragraph, you could have a separate paragraph for each of chapters 3/4/5/6.

?
James Scott-Brown:

“count femicides using“?

?
James Scott-Brown:

“definition“?

?
James Scott-Brown:

“is released”/”becomes available”?

The situation develops, but I’m not sure that “information” does.

?
James Scott-Brown:

“to add to” ?

?
James Scott-Brown:

overview?

?
James Scott-Brown:

“meaning, they” -> “meaning that they” (or “are empirical: they come out of…“)?

?
James Scott-Brown:

The current TOC numbers the appendices with Arabic, rather than Roman, numerals: it lists “Appendix 2“.

?
James Scott-Brown:

This “[CITE]” placeholder may be redundant now this footnote exists?

?
James Scott-Brown:

“Cite interview with Helena“ is presumably a placeholder that needs to be replaced.

?
James Scott-Brown:

italicise these book titles?

?
James Scott-Brown:

“Chapter 8 of this book“ (it’s currently ambiguous whether you are referring to chapter 8 of Data Feminism or Counting Femicide)

?
James Scott-Brown:

Consider rephrasing to something like: “As written on a banner carried during the Combahee River Collective’s march…“

You could also cite the LOC reference of a photo of this: https://www.loc.gov/item/2016648553/

(unless there is a specific statement from the Collective that you could cite)

?
James Scott-Brown:

reference as “Article 3” of the declaration specifically

?
James Scott-Brown:

“These more specific labels include…“ ?

?
James Scott-Brown:

by “organized crime *groups*“

?
James Scott-Brown:

ignore?

?
James Scott-Brown:

what distinction are you trying to make between being ignorant of something and incorrect about it?

?
James Scott-Brown:

I assume the use of “misogynous” rather than “misogynistic“ is a deliberate choice?

?
James Scott-Brown:

It would be more convenient for the reader to provide the translation in parentheses rather than a footnote (as was done for “Yo te nombro: El Mapa de los Feminicidios en México” above)

?
James Scott-Brown:

this “data” is redundant, as “data” appears after the list of activities.

?
derya akbaba:

I think it’s helpful for the definition but I would remove “of data” that comes after circulation

?
James Scott-Brown:

To me, “dataset about“ or “database of” sound more natural, but this might just be a personal preference, as search engines find uses of “dataset of“.

?
James Scott-Brown:

The figure is numbered 0.1, rather than 1.

?
James Scott-Brown:

Italicise?