Pure Genius

Q&A: Heather Dewey-Hagborg, information artist, on the intersection of art and science

Q&A: Heather Dewey-Hagborg, information artist, on the intersection of art and science

Posting in Cities

Heather Dewey-Hagborg creates forensic portraits based on the DNA found in stray items, such as chewing gum and cigarette butts.

At the intersection of art and science, Heather Dewey-Hagborg creates forensic portraits based on the DNA found in stray items, such as chewing gum and cigarette butts. Her work, Stranger Visions, will be on display at the Cyber In Securities exhibition in Washington, D.C., from August 30 to September 27.

I spoke recently with artist Dewey-Hagborg, who also has a background in programming, about the science behind forensic portraiture, reactions to her work and the intersection of art and science. Below are excerpts from our interview.

You got the idea for this work when you noticed a hair stuck in a crack in the glass covering a piece of art. Talk about the path from that spark of inspiration to the first completed piece.

That was a long path. The hair got stuck in my head. That led to the process of ideation, which led to an exploratory process. The first portrait I completed was my self portrait. I completed that using a 23andMe [personalized genetic] profile. I did that to get a sense of how good a portrait I could get using the best-case-scenario data. Just that took almost six months. To get the lab work down so I could do my own experiments -- collect my own samples, extract the DNA myself and have some success amplifying those traits -- probably took another few months. It took about a year to get to the first forensic portrait.

During the first leg of the project -- creating my self portrait -- the experiments included algorithms, data processing, looking at scientific research papers and going through the 23andMe website of traits. 23andMe links to the scientific research their conclusions are drawn from. That got me started in terms of looking at what traits are related to appearance. That created the groundwork for the trait information. From there, I started looking at SNPedia, a website resource of lots of traits.

From all of this information, I started building code that could go through a data file and pull out the traits and give a sense of your likelihood of having a specific trait. The next leg was working with 3D models. I have a background in machine learning. It was natural for me to work on facial-recognition algorithms based on 3D scans of people's faces. Then it was a matter of figuring out how to take those flat 3D models and integrate them into a rapid prototyping workflow, so I could get a 3D print. The final leg was the lab work. I went to a community laboratory in downtown Brooklyn and signed up for a biotech course. I learned to amplify and analyze DNA. I started working through my experiments with the help of the biologists there. Eventually, I started having some success. I also started working on a Ph.D. at Rensselaer Polytechnic Institute in upstate New York.

What are the samples you're working from?

This went through a number of iterations. When I first started working on the project, I collected a ton of different samples from all over the city. For most of the original samples, something about them didn't work out. I was still in my learning process. Then I started having more success when I started my Ph.D. program. I came down to New York City on the weekends and collected samples. I collected five samples -- cigarette butts and chewing gum -- from two neighborhoods in Brooklyn. Genetically speaking, they're interesting because they're diverse neighborhoods. That's my initial set of samples.

Did you have a science background before you took on this undertaking? Or were you learning as you went?

My science background was minimal. I had a computer science background. I'd taken the standard biology courses in high school and a little in college. I was coming into it from a pretty ignorant perspective. I knew about evolution, of course. But I had no idea about molecular biology or the lab techniques involved. A lot of the learning process for me on that end was simply mechanical. I was learning to do things like pipetting properly.

Without a strong science background, how did you know your idea to create these forensic portraits would actually work?

I wasn't sure at all it would work. This is part of the experimental art tradition. The way I often do art is by starting with a question. I don't know what the answer is ahead of time. It's something that gets stuck in my head that I want to experiment with. That was the case with this project. I wanted to see what I could do. As is often the case with this work, you start with a question and you experiment and see where it leads you. It's not necessarily a perfect answer.

Were you doing similar work before this or has this exhibit been a departure for you?

It fits into a trajectory I've been working on related to information and reinterpreting a surveillance system as something that can be creative. Back in 2008, I was experimenting with eavesdropping systems that would listen to people on the street. I've worked with facial recognition algorithms along the lines of what a computer that contained a facial recognition system would dream about. What would those images look like? This exhibit is part of that body of work.

What's the overall goal of Stranger Visions? What's the statement you're trying to make?

This is an art work and, by its very nature, meant to be subjective. I'm more interested in what it provokes others to think about than any specific intention of my own. I'm coming from a background in surveillance, so for me the project was an exploration of this new territory in surveillance culture -- an emerging set of issues. The idea is to call attention to privacy issues, but also explore and experiment and see how much I can know about a person from an artifact they leave behind. It is a kind of critical experiment.

Of your current exhibit, you've said that none of your subjects have recognized themselves yet and that the pieces are more likely to have a 'family resemblance' than an exact one. Why is that and do you hope to improve on the technique?

There isn't an exact resemblance because we don't know yet if there are genes for specific facial traits. There are a handful of things we know and a huge amount we don't. I've been taking the scientific research and interpreting it in a speculative direction. These studies would have to be replicated before they would be accepted in the scientific community. Some of the research I'm using is well established -- eye color, for example. Some if it is more speculative. I've looked at ancestry information markers and mitochondrial DNA. There's all kinds of interesting research.

I do plan to continue to improve this. As I find new research, I continue collecting new traits and adding them to my spreadsheet.

Have there been any unexpected reactions to the work? Are people concerned about the genetic and forensic implications?

This work, for me, is two-fold. On one hand, I have a genuine scientific curiosity and a love of these emerging technologies. I enjoy experimenting with these systems. On the other hand, it's also a conceptual work. It's meant to question things about privacy and our biotechnological future. When people are concerned about the privacy issue, that's entirely justified. It's the point of the piece to call attention to these questions. For me, that's the best reaction. I'm OK with it if people are upset with me. I'm hoping to draw attention to these broader questions and create a discussion.

The predominant reactions I get from people are fascination and curiosity and interest. Most of the emails I get are from people who want to send me samples and have their own portrait done.

Are you planning to do these personal portraits people are requesting?

Even if I wanted to do these commissions, it's difficult to find the time to do that work. I don't even have time to process the samples I have to the level of detail that I'd like. I started with five samples. I keep trying to improve on these portraits. It's an iterative process. At the end, I'll have maybe 10 or 20 versions of a person's portrait depending on the information I acquire. At the moment, I'm still trying to find the time to focus on the samples I have. But I can imagine, in the future when I've got the process pinned down that might be a possibility.

Your work is a prime example of the intersection between art and science. Do you expect we'll see more of these sorts of intersections?

It's something I've been working on for 10 years. In the past couple of decades, there has been a lot of this work coming together and taking off. It's exciting. I come from an interdisciplinary background. I studied art, as well as programming, in college. I do see the work coming together more and more.

The problem is there aren't that many venues that exhibit this work, including in New York. We have a lot of artists making fascinating, experimental work and not a lot of spaces willing to show it. That's a problem that could use addressing.

Photo, top: Heather Dewey-Hagborg. (Kari Mulholland/TED)

Photo, bottom: A portrait from the Stranger Visions collection.

Share this

Christina Hernandez Sherwood

Contributing Writer

Contributing Writer Christina Hernandez Sherwood has written for the Los Angeles Times, Newsday, the Philadelphia Inquirer, Diverse: Issues in Higher Education and Columbia Journalism Review. She holds degrees from the University of Delaware and Columbia University's Graduate School of Journalism. She is based in New Jersey. Follow her on Twitter. Disclosure