On this week’s episode of the PolicyViz Podcast, I’m excited to welcome Catherine D’Ignazio and Lauren Klein, authors of the new book, Data Feminsim. The book is currently open for public review and comment, so you can head over to the website, take a read, and send your feedback and suggestions. The comment period closes on January 7th, so if you get a bit of a break during the next few weeks, I suggest you go on over and check it out.

The three of us talk about their book, how they define ‘data feminism’, and how data and data visualization can biased or otherwise incomplete because it lacks certain viewpoints. If you’re interested in learning more, I recommend you check out their book and writings, as well as reach out to them on Twitter.

If you’re interested in supporting the show, please consider becoming a Patreon supporter, writing a review of the show on iTunes, or simply sharing the show with your friends and networks.


Catherine D’Ignazio is a scholar, artist/designer and hacker mama who focuses on feminist technology, data literacy and civic engagement. She has run breast pump hackathons, designed global news recommendation systems, created talking and tweeting water quality sculptures, and led walking data visualizations to envision the future of sea level rise. Her research at the intersection of technology, design & the humanities has been published in the Journal of Peer Production, the Journal of Community Informatics, and the proceedings of Human Factors in Computing Systems (ACM SIGCHI). D’Ignazio is an Assistant Professor of Civic Media and Data Visualization in the Journalism Department at Emerson College, a Senior Fellow at the Emerson Engagement Lab and a research affiliate at the MIT Center for Civic Media & MIT Media Lab. Learn more: www.kanarinka.com.

Lauren F. Klein is a scholar and teacher whose work crosses the fields of data visualization, digital humanities, and media history, among others. She has designed platforms for exploring the contents of historical newspapers, recreated forgotten visualization schemes with fabric and addressable LEDs, and, with her students, cooked meals from early American recipes—and then visualized the results. Her writing has appeared in American Literature, Digital Scholarship in the Humanities, and Feminist Media Studies, among other venues. With Matthew K. Gold, she edits Debates in the Digital Humanities, a hybrid print-digital publication stream that explores debates in the field as they emerge. Klein is an Associate Professor in the School of Literature, Media, and Communication at Georgia Tech, where she also directs the Digital Humanities Lab. Learn more: www.lklein.com.

Episode Notes

Data Feminism Community Review Site

Feminist Data Visualization, paper by Catherine and Lauren

Information+ Conference

Tapestry Conference

Responsible Data Forum

2016 New York Times election needle

Jessica Hullman’s (Northwestern) website

ProPublica story on maternal mortality, Lost Mothers

Femicides in Mexico Map


Data, Designed with Stefanie Posavec 1/16/19 in Amsterdam

Data Visualization & Tableau with Brittany Fong 1/31/19 in Washington, DC

Support the Show!

PolicyVizPodcast Patreon Page

Write a review on iTunes, Stitcher, or Google Play

Write a review of Better Presentations on Amazon


Welcome back to the PolicyViz podcast. I’m your host, Jon Schwabish. On this week’s episode, I chat with Catherine D’Ignazio and Lauren Klein authors of the new book, Data Feminism. The book is actually out for review right now, community review, so you can go to the website, you can read the book and you can provide your thoughts and your suggestions and your comments. And I’ve put a link to the community review site on the show notes, so I encourage you to listen to the interview and then go over to the website and take a look at the book and see what you think and provide them with your thoughts and your reflections. Before we get into the interview, just a couple of notes; this show is completely supported by listeners, so if you would like to support the show, to help me buy better audio equipment, better editing services, more transcription services, please consider going over to my Patreon page and supporting the show. You can support as little as a dollar/month, $3/month or even more. I really would appreciate your support so I can keep the shoe going. I have one more episode coming up for the end of the year and then we’ll take a couple of weeks off to rest and relax.

A couple other announcements; I have a couple of workshops coming up in January. I’ll be teaming up again with Stefanie Posavec to put on our data design workshop. We’ll be in Amsterdam this time in mid-January. I’ll put the link up to that workshop as well so you might want to check that out. Stefanie and I have been doing a few of these data design workshops where I come from the data side of data visualization, she comes from the design side of data visualization and we try to take a balanced approach to teaching introductory data visualization.

And then at the end of January in DC, I’ll be teaming up with my friend, Brittany Fong to teach an intro data visualization and tableau workshop here in DC. I’ll be teaching intro data visualization and Brittany will then be teaching intro tableau. So if you are interested in those skills, I encourage you to check out the website and go over to the registration pages for both of those workshops.

So this week’s episode as I mentioned is an interview with Catherine D’Ignazio and Lauren Klein who are the authors of the new book, Data Feminism. The comment period for the draft is going to close on January 7th, so fortunately, I was able to get this interview in under the wires so that you can have your holiday break if you’re lucky to get one to go over to the website and check out the book and maybe provide your comments and your thoughts and your perspectives on this important topic, so over to the interview.

Great, so welcome to the show Lauren, Catherine, very nice to have you both on the show. Happy holidays to you.

Catherine D’Ignazio: Happy holidays.

Jon Schwabish: We’re getting towards the end of the year here. It’s finally starting to get cold in DC. So I think we’re all up and down the East Coast here, right? So Catherine, you’re up North.

CD: That’s right. I’m in Boston.

JS: Right. And Lauren you’re in Georgia?

Lauren Klein: I am. I’m in Atlanta.

JS: So it’s still warm down there.

CD: So we’re jealous.

LK: I went on a bike ride this morning although I did wear my warmer bike outfit.

CD: I mean I would not be going on a bike ride right now in Boston.

JS: I was at [inaudible 00:03:37] in Miami a couple of weeks ago and went out in the morning for a run and it was like low 50s which is cool for Florida and I was in just shorts and a t-shirt and feeling like, “Well, maybe it’s a little too cold,” and then I saw a guy with a winter hat and gloves and a big jacket and I was like, “You know what? I’m just going to play the Northern card right now and I’ll be okay.” Well, thanks for coming on the show. You have this really exciting project, Data Feminism that we’ll dive into and talk about and also I guess solicit people to come and give comments on the book which is great. Why don’t we start by having each of you introduce yourselves so for us to know where you’re coming from. Catherine you want to talk a little bit about your background?

CD: Sure, so my name is Catherine D’Ignazio and I’m currently an assistant professor of data visualization and also civic media at Emerson College which is in Boston. And my background is both an artist and designer as well as a software developer. So I come to data with those two hats as well as of course being an educator as well. I guess I should mention too one of the interesting things about the context that I’m in which actually has led to a lot of this work around data is that in teaching data visualization in the context of a journalism department at Emerson College here so I’m teaching students how to work with data who are non-technical, mostly arts and communications focus, so really looking at data journalism as a kind of interesting intersection of these things, so Lauren.

LK: Yeah hi, so yeah my name is Lauren Klein. I am an associate professor in the school of literature, media and communication at Georgia Tech. We colloquially [inaudible 00:05:33] humanities and so we teach most of the humanities courses to undergrads and grad students at Georgia Tech who primarily pursue engineering degrees. My background, I am trained actually of early American Literature and even before that, I was also a web developer. Catherine and I have actually realized that we might have known each other in the late ‘90s, early 2000s when we were both involved in the first dot com wave and during that crash I went back to grad school in English because I thought you could always rely on books. And then all around me this new field called digital humanities started to take shape and this is just a field that’s looking at how computational methods and questions about computation where generally it can be applied to humanistic research questions and then sort of the flip side to how humanities scholars can ask questions about computation and then technology more generally. We can talk a little bit more about that in relation to this book later.

JS: Yeah so how did you two end up teaming up together for this particular project?

CD: Yeah so it was really interesting, this is Catherine and I wrote this blog post and it was in the late 2015 and I was headed actually to the event where Jon you and I met which is the Responsible Data Forum event on data visualization and I wrote a blog post in event of that to coincidence with the launch and the post was called, What would Feminist Data Visualization Look Like? And the post ended up going all over the internet and one of my good friends, Patsy Baldwin, who at that time I think was a librarian still at MIT and she was like, “You have to meet this person, Lauren Klein. She came through Boston and did this amazing talk on feminist data visualization. I don’t know when that was Lauren, but I’d heard it wasn’t like that long ago in the past. And then did she connect us or did I use [inaudible 00:07:38].

LK: Well you know and then we realized that we were going to meet each other anyway because you were already scheduled to come to this workshops that I was organizing or co-organizing about what data visualization looks like for humanities, data and humanities research or something. I don’t remember the exact chronology, but that was basically it. It was like [inaudible 00:07:56] to this amazing blog post and then Patsy connected us because she had come to my talk, that’s right. That’s exactly how it happened.

CD: Yeah and then we met at that event that Lauren organized and realized that we had lots of really shared common interests and also complimentary backgrounds and skills and that’s what let us write our first paper together called Feminist Data Visualization and then realized there was probably enough to talk about that it warranted a book and so that’s where the book came from.

JS: Okay, so now you have this book and it’s also open which we’ll talk about in a little bit, but I’m not just going to ask, what is the book and we’ll just turn there and give people a sense of what you mean or you think of when you use that phrase, ‘data feminism’.

LK: Sure so the book I mean at its most basic level, it’s a book about feminism and data science. It’s what can data science learn from the past several decades of feminist activism and critical thought. We can to this idea because all of the language and use around data that has to do with power. We hear people saying data is the new oil. You see how corporations are really mining data from consumers and citizens for tremendous profit and you understand how data actually is this incredibly powerful tool, substrate, and vehicle for all sorts of things. It turns out that actually feminism has a lot to do with power too. You assume when you hear the word feminism that it might just only be about women or only for women, but really feminism it’s about challenging power and equalities and inequities wherever they exist. I don’t know Catherin do you want to talk a little bit more about that?

CD: Yeah so we’ve been doing like a let’s get on the same page about feminism whatever we talk about in this book. The word has a lot of baggage and people bring in like wildly different perceptions and history so like what their version of feminism is so we like to start with what our version of feminism is. Then we talk about it, thinking about what is feminism. Well, first that’s a belief, so it’s a belief in the equality of all of the sexes and genders. It’s also organized activities just like activism on behalf of working towards that equality because we don’t live in that equality right now. You just have to look at any data visualization about sex and gender differences in different sectors let’s say to see that that world is not equal, we haven’t achieved that yet. It’s belief, it’s activity on behalf of that belief and then feminism is also this long tradition of theory and thinking about how we can achieve that kind of equality. So here there’s so many rich and diverse contributions over the past really what we’re looking as like probably the past like 40 years or so, but just very different fields.

So we looked at feminist contributions in fields like geography, the humanities, science and technology studies to computer interaction design and so on. But I think one thing that’s been important to understand and particularly about contemporary feminism is this word intersectionality, so what we’re talking about is intersectional feminism and this is an insight that was formally fortified by Kimberle Crenshaw in the late ‘80s, but also came out of work by the Combahee River Collective even as far back as the ‘70s where their insight is that look if we’re talking about inequality in the world, we cant just talk about gender inequality because gender inequality is just really inextricably linked with all these other things like class inequality or like race inequality or like ability or like so many other of these different systems of oppression.

So that’s why when you read the book, you’ll see that not all the examples are about women, not all the examples are about gender, but like Lauren said at the beginning, the examples are all about power and so a lot of what we have tried to do with the book is try to bring a deeper and more nuance analysis of power to the conversation about data visualization.

JS: Can you give us an example of power imbalance in data? I think a lot of people who work with data or data science or data visualization, they have some data and then they go work with – they present a table or a graph and yeah there are people back there behind who answer the questions, but like I have the data and I’m just going to do something. Can you maybe give an example of a power imbalance in the data sets?

CD: Yeah totally. So one of the things that we talk about and we try to bring the conversation a little bit let’s say lift it out of the data set to start to consider well like what in the world produces a data set and looking at the ways that the production of a data set involves different kinds of power struggles. Let me be like really concrete about this. One of the interesting examples that we talk about in the book is this example of femicides in me Mexico. Femicides are gender-based killings. This is not a situation that’s new to Mexico. This is basically what you might call both intimate partner violence, domestic violence as well as other reasons why women are killed simply for the fact of being women.

In Mexico, there has been rising awareness in this problem as there’s been rising awareness in other places of this just systemic patterns of violence against women and yet there is no data even though this has been called out by Amnesty International, by the UN, even an international court has told the Mexican government that this is a problem in the context that Mexico. There have been folks working on this in the country of Mexico for the past 20 and more years, but there is still no comprehensive data. So it’s this environment which actually produces a lack of a data set if that makes sense. This is I mean pretty reflective of what happens a lot in relationship to women, but also to people of color, also to folks who are otherwise disenfranchised so like they’re underrepresented in data because right now we collect data about those things that we care about and there are other things that we care about comparatively less and we don’t collect data about.

For example ProPublica has been doing this recent really excellent reporting on maternal mortality and one of the things they uncovered is that we actually don’t have good, solid data on maternal mortality statistics on this country and this is something you would think we would – this is stuff we should know, right? We should be collecting, this is important information and yet the environment has not prioritized it and has prioritized other things.

The interesting thing we talk about in relation to femicides is that there is one woman activist who goes by a pseudo name of Princessa and she has determined so that she is going to collect every femicide that she can get her hands on and she actually logged them from media reports and from crowdsourcing every single day and she’s been doing this now for I think about three or four years. So she has this map so if you google Mexico femicides apps it will definitely turn it up and so she actually now has this one individual citizen has the most comprehensive data that is open at least on this phenomenon of femicides, but it shouldn’t have to be like that right? There shouldn’t be like this one person that is like opening herself up to basically safety issues just to collect a data set that folks should be taking responsibility for in the larger institutions, but maybe I’ll pass it to Lauren because she probably has an example.

LK: Sure yeah I mean there’s so many examples. I mean I’m trying to think about, Jon you’d also asked about not just data sets, but data visualizations that come under this umbrella of data feminism and some of the examples that I really like to give are ones that you might not think of necessarily as being feminist or able to be understood in terms of feminism because it don’t have to do with women, so issues of femicides and maternal mortality, right? Those involve women and their bodies, but ideas about feminism really carry over to everything that we do.

So an example, this isn’t a data set, but it’s a visualization, the election gauge from the 2016 election that actually Catherine and I were just talking about that they just brought it back for the midterms, the New York times featured that wobbled like a speedometer and I’m sure most people listening to this podcast are intimately familiar with this because it was like the most reviled visualization in 2016, right? People hated this visualization. They said it was manipulative, that it was ethically immoral. I mean people really responded hard and negatively to this visualization because it made them feel something and we tend not to think of visualizations as things that should make us feel things, but it turns out that people are really bad at interpreting more of the technologies that we use when conveying [inaudible 00:17:56] to you like error bars or any number of other things radiance.

There’s a research where Jessica Hallman actually he’s done a lot of user charts and writing on this, like people when they look at visualizations that are designed to convey uncertain outcomes, they just misinterpret them very badly. This one actually made people feel the way that the uncertainty was playing out and then it made people feel really uncomfortable about that. Catherine and I like to point to this example as an instance of – it’s a feminist visualization because it uses a different way of knowing. It doesn’t use necessarily like objective information or direct visual, sort of a direct correspondence between the data and then what is seen in terms of some sort of stratigraphic or something that is conceived to be a mutual method of conveying what the data really has to say, it actually evokes emotion right? But the only reason why we feel that emotional knowledge is somehow less good than other forms of knowledge like suggestive knowledge or factual evidence or something like that is because we have this really, really old entrenched hierarchy of forms of knowledge and this is a gendered hierarchy. You can see this in everything from – again I’m trying to think of an example of women like the way that medical knowledge if you’ve gone to medical school is valued more than a home remedy, something that is learned or experienced with the home. We tend to think like, “School is better.” The way you see they’re trained like a chef who had gone to culinary school is deemed to have more knowledge or different kinds of knowledge than an intuitive home cook, things like this. This also breaks down on gender lines as the more general sense that scientific or objective knowledge is somehow better than knowledge that you obtain for your senses.

It has a really, really long history and when you take a feminist approach to it, you can ask yourself, “Why shouldn’t visualization not make me feel something?” If it conveys uncertainty and it conveys it really well and does so in a way actually that made me feel uncertain to my core, isn’t that a really good visualization? Not a really bad one? And so these are the range of examples that we try to talk about in our book.

JS: So help me understand the link between the emotional characteristic, the draw of that piece and then the power and balance that Catherin was talking about earlier. How do those come together in this example or in any example where the concept of uncertainty relays that there’s this emotional reaction because many people at least are probably generally uncomfortable with this idea of uncertainty, wanting answer it’s 5% chance. How does that relate to this power and balance that you’re talking about in the book?

LK: That’s such a good question. I think that – so here is another example that actually is a little bit more relevant to some of the examples that Catherine was talking about before, but I’m going to bring that back to the election gauge at the end. People like answers, people like concrete information in the world, but the reality is that the world is complex and confusing and rarely resolves to certainty even these are especially the election data, it’s all probabilistic, right? It’s all running in simulations upon simulations and so even the data itself is data that is hypothetical. The election hasn’t happened right? We don’t really know what is happening and yet to display it in a way that it is certain is trying to either deliberately or unintentionally capitalize on the power that data has, that data has concrete evidence.

There are a couple of people who have tried to theorize like how data is different from evidence or is different from facts and what people say is, “People sort invoke data or employ data when they want to establish a stable basis on which future arguments can be made. They want something that itself can’t be challenged, but the example of the election data is really good, because it’s not like you actually collected it from people. You didn’t like take someone’s temperature and say, “Oh you have a fever because I put the thermometer in your mouth and the thermometer says 103. You have to stay home from school or whatever.” They run these simulations so already the data itself, it’s interesting and it can tell you something, but the data in itself is not objective or tied to a fact in the world in a direct way. And then when you have it presented in a way that conveys facts, that’s where problems arise. This is something that Donna Haraway, the feminist philosopher of science has been saying since the ‘80s.

She talks about something called the view from nowhere and this is actually a really foundational idea from our book which we apply. It comes from Donna Haraway and we apply it here broadly. There is always a perspective that is conveyed when you’re visualizing data or even as some of these theorists would say when you’re presenting information as data. It’s just that some of those perspectives are not made visible or not acknowledged and being aware of what those perspectives are, who is creating the visualization, where the data came from. If you’re more open about that, then people can actually work towards a better sense of the meaning of the data that they’re being presented with. I feel like Catherine you might have a better way to say this.

CD: Yeah totally and then what I’ll mention about the Donna Haraway, View from nowhere is that the kind of sociological updated version of that right now is this recent paper that Helen Kennedy and a team of researchers out of UK did called, ‘The Work that Visualization Conventions Do’ and it’s a great paper. It’s an empirical study of people’s perceptions of visualization. One of the things that they showed is how the sort of clean geometric forms, straight lines, clean lines, the facticity and then the concreteness with which data visualizations communicate these conventions of the form work to present this idea of completeness and totality and they make these [inaudible 00:24:50].

I think part of what Lauren is talking about with that gauge that was so interesting is that of course the visualization slipped on itself like it’s using some of those similar conventions, but then it was like wobbly and it destabilized our conventional assumptions about visualization, that is just true. You know what I mean? Visualization often is like, “This thing in the world is true and here is the observation of it or whatever,” and the gauge was like, “I’ve now destabilized and it’s giving me this emotional reaction to that.” So it was kind of almost like a hack of our own perception of visualization, so yeah that would be my addition.

JS: Yeah and it’s really interesting. I want to ask you about the book itself. The interesting thing about the book is that you’ve opened it up for public comment and I’m curious what that experience has been like so far and also why you decided to do that. I mean it seems like a lot of people say, “I’m writing a book. Stay tuned 24 months later and my book is out,” but you’re taking a different path with a topic I would guess that people probably have pretty strong feelings about.

LK: Yeah I mean we had a couple of reasons for wanting to open it up and definitely the slow pace of publishing is one of them and I do think it’s important to acknowledge for people who have never written a book, there’s a lot. It’s not like nothing is happening, it definitely takes two years to publish a book, but it’s not like nothing is happening during that time. It goes out for peer review. It gets reviewed by the editorial board of the press. It gets copy-edited. It gets typeset. It goes through legal teams to make sure that all the image permissions are okay. Actually there’s tons of work that happens and actually as I have done this a couple of times, I feel like it’s important to tell people about that because usually I mean I thought this too, you like set your book into the void and it comes out later as a book, book.

We put it online primarily because we do have a lot of audiences that we want to speak to. Catherine and I, our interests are overlapping, but they’re also as Catherine said earlier complimentary. I come at this from a humanities perspective. I wanted to speak especially to students and also to humanities scholars who were interested and looking for examples of how what they knew could apply to data science. Catherine, do you want to talk a little bit more about who you were hoping to speak to?

CD: Yeah. A lot of my other work has been about literacy, but data literacy for a very specific purpose not just because in general it’s good for people to know about data. I mean I do think it is, but one of the things that comes when we have – has been characterized like data has a lot of power, I suppose economic power and political power and social power, people are talking about even using this metaphor of the fourth industrial revolution to mean this changing methods of data and artificial intelligence.

One of the things that’s apparent to me is that not all voices are at the table for whatever this revolution is, we’re not all there and this is a revolution that’s being led primarily by folks in engineering and technology which is not bad. I mean I’m in one of those fields, but I really feel strongly we need other voices at the table, so that’s why I do a lot of work on data literacy specifically for what I would characterize as public information professionals, so folks like journalists, librarians, municipal government, non-profit sector, community-based organizations and all of that to think about stimulating more robust public dialogue about what the shared future of data and technology should look like. That’s what brings me to doing this work is thinking about how do we bring more folks to the table.

But then with the open review process, I definitely appreciate that we can get something out there in one year versus two years. This is my first book that I’ve ever written and it was like, “Oh my gosh I write something in one year and it doesn’t come out for two and a half other years.” But then also I feel like it’s very values aligned for us, so one of our core principals of data feminism is around pluralism.

So what that means is multiple voices in the process and running a more participatory process. And so this to me it seems to really enact those principals by saying, “Here. It’s open. We want your comments and then we are not just doing it because we want your eyeballs, we’re actually going to be reading all the comments and they’re going to be used as an input into our revision process in this way. I think for that reason as well. Already people have been extremely generous, like the folks that are strangers, who we don’t even know have given some very long and detailed and really generous comments there like, “Oh yeah we didn’t think about that or we should elaborate on that example more. This thing is not clear and so it’s really gratifying to see that kind of pluralism.

LK: I mean I think like we’re just two people and we have our individual perspectives, we’re both academics, we come from our environments that we live in and we cant know everything and this is true I mean of anything. No single person knows everything, but especially when working with data which is usually or can easily be martialed to make to make these sweeping claims. I think it’s a really important process for us to do and also for people to see to say like if the goal is actually better knowledge and more knowledge and more complete knowledge and we really think that we can use data in order to get there, but in order to do so, you need to acknowledge also the limits of what any single person or any single perspective can bring.

JS: Great, then try to acknowledge at least your own biases and then let people tell you where you’re wrong.

LK: Yeah.

JS: How do people provide comments to you? What’s the process like?

LK: Yeah they just go to the website which I’ll say now is –

JS: I’ll put it on the show note.

LK: Okay good yeah [inaudible 00:31:25]. Yeah so just you go to the website that Jon is going to post and when you read the draft of the text, you just highlight some texts. It’s very similar to medium.com. So you just highlight the text and post a comment that way. And you can also [inaudible 00:31:45].

CD: I think you need to register first and the link to register I think is in the top right, I’m not sure.

JS: Okay and how long are you keeping comments open?

LK: Till January 7th.

JS: Okay so people have their holiday break to just sit down for a couple of days –

LK: Now start with the conclusion.

JS: Start with the conclusion and work your way back. Okay so then January 7th and then you go back to editing and then we expect the book in our hands when?

LK: Spring 2020. It sounds forever.

JS: Spring 2020, so another year or so everybody has to hold their breath, but I’ve noticed that people are excited about the fact that there is a book on this topic and I know Catherine you’ve been around talking about this topic, did a great talk at InfoPlus back in the fall and I’ll link that talk as well. Well, it’s great. Thank you both for coming on the show. This is the last podcast episode of 2018. I think it’s a great topic to end on especially this year. So yeah thanks for coming on the show.

LK: Thank you. Thank you for having us.

CD: Thank you.


JS: Thanks everyone for tuning in to this week’s episode. I hope you will go check out the website for the Data Feminism book, put your comments in, take a look, read, maybe even read some of the later chapters as I’m guessing there’s probably a drop off, people get in, edit and it drops off. And so do take a look and provide your thoughts and your comments to get to this concept of pluralism. See I’m learning all sorts of new things today, so it’s great. Yeah so thanks everyone for tuning in. Have a very happy and safe holidays and a happy new year and this has been the PolicyViz podcast. Thanks so much for listening.