Welcome to AAI Reads!. This week, the book that we’re highlighting is Data Feminism. Written by Catherine D’Ignazio and Lauren F. Klein in 2020, it’s a great introduction to how we can incorporate feminist practices into data literacy.
Continuing our previous theme, explained in the last AAI Reads! review, we’ll have a TIFF and a JPEG perspective on the book. Since it’s been published so recently, none of us are quite at the Thumbnail level but maybe we’ll revisit that in a few years! This time Paulina will be the TIFF reviewer and Meghan will be the JPEG.
Paulina – TIFF review
I have been consuming a lot of books on data recently so the lossy compression is close to creeping creeping in. But before then, I get to be the TIFF review of Data Feminism and I really liked it. As we’ve been accumulating works for our Aggregative series, this book was an easy addition. Bringing a feminist perspective to data through critique and as a manual for how to change practice was a key part of this book and something we wanted in our data stories.
I liked that this work, like others in the vein of critiquing and understanding the intersections of data and social justice, takes the time to identify the positions of authors in relation to their subjects of study. In the book, D’Iganzio and Klein state that they are white women coming from particular backgrounds. These statements respond to critiques of white writers by writers of color by racializing the authors as white women. It acknowledges they have a racial identity, rather than ignoring that fact.
I also like that this illustrates one of my favorite points in the book. Drawn from a variety of scholars’ work, Data Feminism points out how “privilege hazards” prevent people from seeing the harm done to communities dissimilar to their own. This also extends to an inability to understand the world in a way contrary to their experience. By knowing the authors’ backgrounds, we can consider what might be missing in their discussion.
Another key point was that “more data” does not actually lead to change. Instead, the requirement to “prove discrimination,” or other similar issues, is part of systemic oppression by data because data does not speak for itself.
Data must be read and worked with, to form arguments (I’m manipulating the definitions for data literacy discussed by Bhargava et al.). However, there are people who will not read the data or do not care to change, regardless of how much work goes into disproving their points.
It’s a great reminder that data without people and human connection for translation is useless. Data cannot speak without people to give it voice. And we should always consider who is doing the talking.
Meghan – JPEG review
The first thing I want to mention is that this book is Open Access. That means, dear readers, that you can read it for free. While I’m 100 percent in support of authors being compensated equitably for their labor, because writing is labor, it’s important to acknowledge that knowledge under capitalism is gatekept, and that one of the key ways information is kept from the disenfranchised and marginalized is through paywalls.
The chapter that resonated the most with me was Chapter 5, “Unicorns, Janitors, Ninjas, Wizards, and Rock Stars”. This chapter looked at the notion of insider and outsider in data design and implementation, and how crucial intersectionality is to creating data sets and data visualizations that are for ‘good’.
“Embracing the value of multiple perspectives shouldn’t stop with transparency and reflexivity. It also means actively and deliberately inviting other perspectives into the data analysis and storytelling process—more specifically, those of the people most marginalized in any given context.”
Creating data with this level of transparency requires planning from the outset. It requires a different way of thinking about process and research. To me, it seems like a no-brainer to begin with perspectives outside of your own, and outside of the dominant context. The whole point of this chapter though, is that it’s not ‘normal’ practice. It’s still radical. Most data-focused job postings don’t stress intersectionality as a desirable. They focus on what the authors describe as unicorns, ninjas, wizards, and rockstars, data scientists who are unique, who have complex moves, who do magic, and who outperform everyone else. These are not skill-sets that are based around thinking about the perspectives of others. They’re those who are outside the mainstream in terms of skills, but who sit solely within it in terms of cultural and socio-political context.
I get worked up just thinking about it.
This is a great book for graduate students, which is the context in which I read it (and why I’m the slightly lossy JPEG, and not the TIFF). It explains feminist theory as related to data science, it gives concrete examples, and it makes arguments through clear, well-referenced case studies. It’s a great example of what a mass-market work about data can look like, and what a book written by academics for a public audience can look like.