Fugitive Data Portraits

self emancipation in Virginia with Tev’n Powers

Emily Chao: Thanks for taking the time! If you want to, introduce yourself, where you’re located, and what brought you to Collective Action School? 

Tev’n Powers: I’m Tev’n. I’m calling from Norfolk, Virginia. I came to Collective Action School because I’ve been following Logic(s) for a while and I saw that the school had had a previous cohort and was gearing up for another year. I’m often looking for spaces to build community and as far as finding community in tech, or tech adjacent spaces, that can be challenging if you’re also looking for folks who are value aligned and politically aligned in certain ways. The Collective Action School seemed like a great space to meet some folks who are also looking to think more critically about the ways we engage in tech, the way that our career and personal choices intersect with our values and what we produce and put out into the world. So that was sort of the fire that lit the interest in joining Collective Action School. 

Emily: Nice. Was there a community in mind that you were wanted to bring into Collective Action School? Is there someone that you were learning for, or you wanted to bring your learnings back to? 

Tev’n: Yeah, I have close friends, some in tech, some not in tech. I’m part of an organization called Black Praxis Project, and we do a lot of virtual political education, book clubs amongst ourselves, and then we open that up to the public as well from time to time. I felt like Collective Action School had a similar spirit of bringing people together to build community virtually. It was (1) a continuation of some of the work I’ve been doing in the spaces with my friends and community, and (2) an opportunity to see how another collective learning and collective growth experience can happen and to bring some of that back to my friends in our organization. 

Emily: Amazing! Let’s switch gears just a little bit. Tell me a bit about your project and how it began? 

Tev’n: So the working title, or maybe at this point it’s the actual title, of the project that I’ve been working on is called “Fugitive Data Portrait: Self Emancipation in Virginia.” This project started as a bookmark in my Firefox browser. I’m very much the type of person that, having worked in data and having a data science background, if I find a data set or archive or a set of information on the internet that looks like it someday might be interesting to dive into or that I would like to set some time aside to dive into, I’ll bookmark it. So I have a whole backlog of archives and projects that hopefully someday can see the light of day. This project was one that I found thanks to the Library of Virginia digitizing a lot of state records from the 1800s. I had no idea really where that it lead to, if it was going to be a project, or if it was just something that I would get a chance to read more closely. As I started looking through those records and opening up spreadsheets of thousands of rows, I started looking through them one by one. But it dawned on me that I also might be able to get an aggregate view or a bird’s eye view of some of the information as well. So from there I started poking around the web and looking into archival research. I’ve stumbled upon that word in my homework—seeing people dig into the archives and try to extract useful information and insights from them. I came across W. E. B. Du Bois’ Data Portraits, and in that work saw that there was sort of a history and lineage that I could possibly build on or be in conversation with. The Collective Action School provided the space for me to feel like I had the time to sit and dive deeper into that. But yes, it started from finding a dataset that I thought was interesting, wanting to use my tech skills and interests for something other than the type of projects that I’ve typically been involved in, and having the space finally to pursue that. 

Emily: Nice. I know the dataset is mostly focused on records of slavery in Virginia but are there any interesting things you’d never known before this or tidbits that you discovered when looking through it? 

Tev’n: Yes, there are things in there that I never knew, and I think what I’ve realized in digging through the dataset is that there are things in there that I’ve never really thought about, so to speak. For example, I’m for the first time looking at the distribution of the records dates—discovering that it does make sense that during some months it looks like there are peaks in the recorded runaways happening. Or I’m looking at the distribution of gender and age across the reported escapes, and speculating what that might mean. I’ve also been looking at some dates and trying to reference historical records of wars or rebellions that were happening at the same time and trying to speculate about how these records might be a part of some of the more noted historical events from Virginia. 

There was one other thing that I hadn’t really known about, or that’s in the same category of discovery, which was seeing certain records of folks escaping on the same day from the same plantation. I have friends who do archival work, specifically around marronage and runaway slaves, so have been in conversation with them, thinking about whether that might have been some sort of group or collective rebellion or escape or action that happened. And discussing other things that they know were happening during those time periods. So it’s opened up a lot of conversations, a lot more questions, and I’m still actively learning as I continue to poke around in the archive. 

Emily: Absolutely. And just curious, how many records did the library digitize? 

Tev’n: I want to say over a dozen actual collections.The collection that I’ve sent a mock up of, which I’m actively working on, is just the runaway slave records. That collection or sub-collection has just over 5,000 individual records in it. It ranged from the 1720s to 1863, but 95% of them are from 1861 to 1863, during the Civil War. There are also court records and sales and deeds and those sorts of records in the other collections, which also range from hundreds to multiple thousands of rows per collection. It seems like a lot of work went into making these accessible. 

Emily: That’s awesome. And, also, it is very much heartbreaking data as well. But it is very important that they’ve provided access to keep the history. 

Tev’n: Absolutely. Absolutely. 

Emily: Were there any challenges that you’ve encountered while doing this project? And what does it feel like to work through them? 

Tev’n: I think the challenges that I’ve encountered so far are mostly trying to stay focused on honoring the stories in the data. Like you said, these are real stories and heavy information that we’re poring through and looking at here. There are parts of the data that are very interesting to me from a tech perspective and for engaging with in that way, but, at the same time, I don’t want to come across as not treating these stories with the genuine respect that they deserve. And so throughout this process, I’m thinking about how to make sure that as I’m designing and building things that it’s not with a deliverable product focus mindset. It’s not like, “I’ve got to build this thing.” I find myself, at certain times, allowing deadlines to push me, to rush through something, and I’m trying to resist that urge when it comes to work like this. I want to remain cognizant of taking my time to approach this data in a thoughtful way and not in a rushed way. It speaks to both trying to honor the stories and the people who are represented in these records, and also to giving myself some time and space to sit with the type of records that they are and not let myself be emotionally detached as I’m thinking with these stories. 

Emily: For sure. I know you mentioned W. E. B. Du Bois’ Data Portraits as inspiration. Are there other sources of inspiration you’ve been drawing from? 

Tev’n: I’d be remiss if I didn’t first mention writer and scholar Saidiya Hartman as a major spark of inspiration. In an interview she says, “I work a lot with scraps of the archive. I work a lot with unknown persons, nameless figures, ensembles, collectives, multitudes, the chorus. That’s where my imagination of practice resides. That’s where my heart resides.” While reading her book, Wayward Lives, Beautiful Experiments, in Black Praxis Project’s book club, I found a tremendous appreciation for her ability to balance storytelling with the documented experiences of people she encountered through her archival research. It was her engagement with the archive that sparked my curiosity in wanting to know what these historical records might contain and what we might be able to learn about the people documented within them. This curiosity is a big part of what led me to wanting to build something that could provide a new way of accessing these records. 

So I’ve got that book that I’ve read and continue to reference. And someone who I haven’t read as closely, but who I’ve learned via video or podcast a lot about, is Ida B. Wells. I’ve heard folks reference her as one of the earliest examples of data journalism that we have due to the work that she did around lynchings in the 1900s. So, I have that in mind. And then I did mention a friend of mine, someone I went to school with. Her name is Dr. Celeste Winston. She’s a scholar and professor at Temple University right now and she does a lot of research on runaway slaves. I actually have her book somewhere over here called How to Lose the Hounds: Maroon Geographies and a World beyond Policing. She does a lot of work on marronage and slave rebellions and runaways in Montgomery County, Maryland, I believe. 

Having someone like that to speak to—both from the archival perspective of hearing the type of work that she does when she engages the archive and how she approaches that work—as well as these historical figures for inspiration, I’d like to think that I’m continuing or contributing to the type of conversations that they were also trying to evoke. Those are the people who I look to and try to reference as I figure out what I’m doing here. 

Emily: Nice. Where is your project now, and where do you want it to go? What’s your wildest dream for it? 

Tev’n: That’s a good question. I don’t know where I want it to go. That’s for darn sure. 

I’m actively working on a web page. Over the last couple of years, I’ve been learning web development and data visualization work. So, the front end part of that is what I’ve been learning over the last few years after having worked for nearly a decade in data science and backend work. Part of where I want it to go is into something that I think is visually appealing, to the point that people at least want to engage with it. When I work on projects like this, what I really would like to happen is for someone who has a little bit more expertise in either the research field or archival field would want to collaborate and add some words and writing to this. I’m able to do some data things and will put my own writing on this page and contextualize it and draw the insights that I think I found, but I also know that there are people who can add to and probably push these thoughts further. I just haven’t flexed those skills as much as I know other people have. It would be great to find someone who would actually want to add some words to it, whether it’s this archive or combined with another archive. I think that could make it a much more engaging site than just the visualization. So I’m going to do what I can as far as putting my words on it, but, you know, hopefully someone I already know, or someone else who might encounter it in the future, sees it and would like to collaborate in some type of way. 

Emily: Yeah. Of course. Do you mind showing me what you have so far? 

Tev’n: Yeah! This is the web page. There’s two parts. There’s the front end web page that I sent you some mock ups of and there’s this Jupyter notebook I’ve been working out of. 

So this is the two-step process. On this backend here, this is my Python notebook where I’m reading the CSV file and just prototyping and doing exploratory data analysis to look within the dataset. This chart is the histogram of months of escape. You can see me prototyping it here. And then if we go back over to the frontend piece, this is now the visualization that I’ve been working on in Svelte (which is the front end framework that I work with). 

Emily: It looks like the vast majority of events are during 1862 to 1863. 

Tev’n: Yeah, exactly. I’ve even trimmed off the early 1700s data that’s in here, but as you can see, the majority of the records shown are from those two years. Which then goes back to the question of what’s not in the data. It’s a question of whether this an accurate reflection of the correlation of the escapes during this time or if this more so an accurate reflection of what was recorded. Do these the years prior and after truly not have as many escapes there? 

Emily: Like, were people motivated to record more because it was happening more? Kind of that? 

Tev’n: Yeah, motivated to record more, and honestly, knowing records from this time, how well were records kept? I can poke through the actual spreadsheets and see that there are some counties that are way more represented in the data than others. Is that a reflection of their court process and record keeping versus the actual events that took place in that county? That’s sort of an open question that I don’t know that I can find the answer to from the data alone. 

So this is one of the visualizations (below). As you see in the mock up, there’s some other things still coming, but this here is a map where I’m just playing around with making sure that I could hover over and highlight the county in the map and that the corresponding county shows up in the text below. I only have it on one of the counties here, but that’s because I didn’t feel like going through and copy-pasting fifty times just before our call, but that is what is in flight there. This has been the process of prototyping visualizations with the easy Python library before I delve too much into doing it in the front-end. 

With the age distribution, I bucketed records so that you can see, somewhat expectedly, people in their twenties to thirties are the bulk of the recorded escapes. Remarkably though, you do see recorded examples of people even up to 80, 70+ years-old self-emancipate. What I found interesting here is that within the ages of 0 to 10, you see almost an even distribution in recorded gender. I believe it’s about 40 to 50% split between male and female, amongst the children, and then very quickly, from teenagers onward, it’s dominated by mostly males being recorded. 

Like I said, a lot of these things just raise more questions for me about what this means. What might that say about what was happening? Or, again, what was recorded and what was not? 

I think the last interesting piece that I’ve gotten to so far is from looking at people that were recorded to have escaped on the same exact day from the same plantation. Most of the records just record a single person, and the distribution tails off. Well, okay: these two people were reported from this plantation, two plus three, you know, four or five, six, however many people were recorded. 

Emily: Oh, okay. There’s some groups that go up to 11 or 12. 

Tev’n: Yeah. It begs the question of whether these are actual escapes that have taken place on the same day, or if someone over the course of a month submitted all these records at one time to the court. So there are some questions here. I plan to call out on the website that some of the conclusions that we might draw can’t be known for certain because so much is left unsaid in the dataset. 

Emily: Yeah. For sure. 

Tev’n: So, yeah, that’s been the process: Figma mockups, prototyping with some Python quick library just to see what the shape of data looks like, and then reading it a little more closely. I’ve been working on the frontend trying to build a web page, which I think could be a nice blog post and some visualizations that folks would like to engage with. That’s kind of where things are. 

Emily: That’s really great work so far! 

Tev’n: Thank you. Thank you. And I hope that by the end of February the landing page link that I sent you is actually a polished version, or at least a useful version of the site—so that when the yearbook’s ready a version of the blog post and site will be ready. 

Emily: But it’s also okay if it’s not. Right? 

Tev’n: And it’s okay if it’s not. 

Emily: Exactly. No deadlines to rush us. 

Tev’n: We’re gonna be perfectly fine if it’s not ready. 

Emily: Yeah! Well, anything else you want to add to finish off? 

Tev’n: It’s been nice to have had the space with the Collective Action School to begin something like this and to use it as a reminder and practice even after the school to approach tech work in different ways. 

It’s been a labor of love, so to speak, working on this project, and I’m thankful for y’all for providing that spark and space to begin with. I’m excited to see what the rest of my cohort has been up to since we last met. 

Emily: Awesome. Well, thank you. It’s been great having you in the community with us, and thank you for taking the time! 

Tev’n: Likewise. Thank you. Thank you again. 

Tev’n is continuing work on Fugitive Data Portraits after receiving a 2024 Public Humanities Fellowship from Virginia Humanities.