Olivia: "The data doesn't lie." We hear that all the time, Jackson. It’s become this mantra for objectivity. Jackson: Right. Numbers are pure. Numbers are truth. If you can’t measure it, it doesn’t exist. Olivia: Exactly. But what if the biggest lies aren't in the data itself, but in the data that was never collected in the first place? The stories that were deliberately erased before a spreadsheet was ever even opened. Jackson: Hold on, erased? That sounds like a conspiracy theory. Are you saying someone is out there hitting 'delete' on reality? Olivia: Not exactly hitting delete, but systematically choosing not to hit 'record.' And that's the revolutionary idea at the heart of the book we’re diving into today: Data Feminism by Catherine D’Ignazio and Lauren F. Klein. Jackson: Data Feminism. Okay, that's a combination of words I did not expect. It sounds... intense. Olivia: It is, but in the most illuminating way. What’s fascinating is that the authors come from different fields—one in data visualization and justice, the other in digital humanities—and they came together because they saw the same problem from different angles. They even made the book completely open-access online, which tells you their goal isn't just academic. It's a mission. Jackson: An open-access mission. I like that. So, this idea of 'erased data'… give me an example. What are we not recording?

Olivia: Let's start with something incredibly visceral: pregnancy and childbirth complications. For decades, we’ve had statistics. But for a very long time, the horrifyingly disproportionate rates at which Black women in the U.S. die from these complications were not a mainstream topic of conversation. The data points might have existed in disparate places, but they weren't being collected, aggregated, and centered as a crisis. Jackson: Wow. So the problem was there, people were dying, but because no one with power was actively looking for that specific pattern, it was effectively invisible in the 'official' data? Olivia: Precisely. The authors call these "missing datasets." They argue that these gaps are not random accidents. They are the result of power. The book opens with a foundational quote: "All data have biographies." They are collected, cleaned, and compiled by people, and every one of those steps is a human decision shaped by biases. Jackson: A biography. I love that. It’s not a raw material like oil that you just pull from the ground. It’s more like a finished product that’s been shaped and polished by many hands, and some parts might have been sanded off. Olivia: Perfect analogy. And to understand who does the sanding, the authors introduce a concept from sociologist Patricia Hill Collins: the "matrix of domination." Jackson: That sounds like a villain's weapon from a sci-fi movie. What does it mean in simple terms? Olivia: Think of it like a net. Each knot in the net is a system of power—racism, sexism, classism, colonialism. They all intersect and reinforce each other. This 'matrix' determines whose life is seen as valuable, whose problems are worth solving, and therefore, whose data is worth collecting. The lack of data on maternal mortality for Black women wasn't a simple oversight; it was a symptom of a system that has historically devalued Black women's lives. Jackson: Okay, that’s heavy. And it’s a bit overwhelming. If these power structures are so huge and invisible, how can you even begin to fight back? It feels like you’re fighting a ghost. Olivia: You fight back by making the invisible visible. And this is where the 'feminism' part of Data Feminism becomes a powerful tool for action. It’s not just about critique; it’s about building something better. It gives you a new set of questions to ask. Jackson: What kind of questions? Olivia: The book says data feminism asks a series of "relentless 'who' questions." Who benefits from this data? Whose goals are prioritized? Who is harmed? And, crucially, on whom is the burden of proof placed? Jackson: On whom is the burden of proof placed… That’s a big one. It’s like, you have to prove your suffering is 'statistically significant' before anyone will believe you. Olivia: Exactly. And that’s why some people decide to stop waiting for permission and just start collecting the data themselves.

Jackson: So they create their own datasets? How does that even work? Olivia: It works through what the book calls "counterdata" and "data activism." There's this powerful story they analyze about a community mapping project. In a city, Black children were dying on the streets due to various forms of structural violence—accidents, crime, neglect. But the official city data didn't capture the systemic nature of it. It was just a collection of isolated incidents. Jackson: Just tragic, disconnected dots on a map. Olivia: Right. So the community started their own project. They began documenting and mapping every single death, creating a visual archive of the loss. By doing this, they created a new dataset from the ground up, one that told a story of structural oppression, not just individual tragedies. They were challenging the official narrative. Jackson: They were building their own proof. They weren't waiting for the city to decide their children's lives were a pattern worth investigating. Olivia: Exactly. They were fighting back against what the book, citing scholar Ruha Benjamin, calls "The New Jim Code." Jackson: The New Jim Code? Olivia: It's this idea that software and a false sense of objectivity can come together to control the lives of people of color. An algorithm that predicts crime hotspots might just be reflecting historical over-policing in Black neighborhoods, creating a feedback loop of more policing. The code looks neutral, but the outcome is discriminatory. This community mapping project was a direct challenge to that kind of coded bias. Jackson: That’s incredible. It’s like data as a form of protest. But it also sounds like it requires a ton of work and emotional labor from people who are already suffering. Olivia: It does. And that brings us to another core principle of data feminism: elevate emotion and embodiment. The authors argue that traditional data science has this obsession with being cold, detached, and "objective." It strips out all the human feeling. Jackson: The whole "just the facts, ma'am" approach. Olivia: Yes, and they argue that this is not only dishonest—because, as we've said, choices are always being made—but it’s also less effective. They give this brilliant example from data visualization. Imagine you see a bar chart of shooting deaths. It shows numbers. Maybe one bar is higher than another. You register it as information. Jackson: Okay, I can picture that. It’s a standard news graphic. Olivia: Now, what if instead of a bar chart showing the number of deaths, you saw a visualization that showed the number of years stolen from each person who died? It calculates the years they likely would have lived and presents that. Jackson: Whoa. Olivia: It’s not just a number anymore, is it? Jackson: No, that hits completely differently. It’s not a statistic; it’s a stolen future. It's the loss of potential, of experiences, of life itself. That’s not just information, that’s… grief. Olivia: That’s what they call "data visceralization." Making you feel the data in your body. It’s not about manipulating the audience; it’s about conveying a deeper, more complete truth. The truth is that these deaths are not just data points; they are profound human losses. Valuing that emotional knowledge is a feminist act.

Jackson: Okay, so data should be more human, more emotional, more contextual. I’m on board. But in the real world, data scientists are obsessed with one thing: 'clean data.' How does that fit with this idea of embracing the human messiness? Olivia: It doesn't. And that's the next sacred cow the book goes after. The authors argue that the very idea of "clean data" is deeply problematic. The obsession with it often leads us to erase the most important parts of the story. Jackson: What do you mean? Isn't cleaning data just about fixing typos and removing duplicates? Olivia: On the surface, yes. But it can go much further. Imagine a survey about gender. A "clean" dataset might force everyone into a male/female binary, deleting or re-categorizing anyone who identifies as non-binary. In that act of 'cleaning,' you've just erased an entire group of people's existence from the dataset. The authors say, "clean data hides diversity." The 'messiness'—the outliers, the non-standard answers—is often where the richest information is. Jackson: Huh. So the mess is the message. Olivia: In many ways, yes! This leads them to advocate for a shift in thinking, from "datasets" to "data settings." A dataset is a static file. A data setting, a term from scholar Yanni Loukissas, includes the whole context: the people who collected it, the tools they used, the political climate they were in, the choices they made. It's the data's whole biography. Jackson: It’s like the difference between reading a single text message versus knowing the entire history of the relationship between the two people texting. Context is everything. Olivia: Exactly. And when analysts don't have that context, the authors say they become "strangers in the dataset." They are working blind, and they risk doing real harm, what they call "epistemic violence"—misrepresenting a group's reality because you don't understand their world. Jackson: I can see this happening all the time. It reminds me of the Open Data movement, where a city government will just dump a massive, 10-gigabyte spreadsheet of, say, 311 calls online and say, "Look how transparent we are!" Olivia: You’ve hit on their next major point. They critique the Open Data movement for this very reason. It succeeds in "opening up" data, but it fails in contextualizing it. That spreadsheet is technically public, but it's practically useless to almost everyone. Who can even open a file that big? What do the column headers mean? What are the known biases in who reports issues to 311? Jackson: Without that context, it's not data; it's just digital noise. Olivia: And the solution they propose is so simple and brilliant: create "data user guides." A short, narrative document that accompanies the dataset, telling its story. Its history, its limitations, its ethical considerations. It’s about taking responsibility for the data you put out into the world.

Jackson: So, if we pull this all together, it feels like the book is a complete reframing of what data is. It’s not this cold, hard substance. It’s a living, breathing thing, shaped by power, full of stories, and carrying real emotional weight. Olivia: That’s a perfect summary. It dismantles the myth of neutrality, first by showing us how power creates 'missing data.' Then, it hands us a toolkit—data feminism—to fight back, by challenging power with 'counter-data' and by elevating emotion to tell truer stories. And finally, it grounds us in the real world, reminding us to embrace the messiness and always, always prioritize context. Jackson: It all comes back to those "relentless 'who' questions" you mentioned earlier. Who, who, who. Olivia: Yes. Who benefits? Who is harmed? Whose labor is visible? The book is a call to stop pretending we are objective observers floating above the world—what Donna Haraway called "the god trick of seeing everything from nowhere"—and to instead acknowledge our position and use data with care and a commitment to justice. Jackson: It’s a huge shift from data science as a purely technical skill to data science as a moral and political practice. It makes you wonder, what 'missing data' exists in your own life or your own community? What important stories aren't being told because no one with power has decided to count them? Olivia: That’s the question, isn't it? It’s something we can all ask. Think about your workplace, your neighborhood, even your own family. What truths are being overlooked? We encourage everyone listening to sit with that question. The answers might be the first step toward your own act of data feminism. Jackson: A powerful thought to end on. Olivia: This is Aibrary, signing off.