Christopher X J. Jensen
Associate Professor, Pratt Institute

E&E in A&D: Genetic profiling as art?

Posted 13 Feb 2015 / 0

Smithsonian Magazine Creepy or Cool? Portraits Derived From the DNA in Hair and Gum Found in Public Places

I find a lot of art to be gimmicky. I know as a professor at an art and design school, that could get me into some trouble, so let me explain what I mean. “Gimmicky” art to me is work that rests on the foundation of clever tactics, perspectives, or tricks… and nothing more. The artist comes up with a slightly off-kilter way of looking at something — often ironic — and then builds the work around that perspective. To me this kind of work exploits the same kind of human cognitive biases as humor; in fact, one could argue that this kind of art is really a kind of humor. Clever and funny and really about discordance, unexpected-but-logical combinations, and the unexpected. Most art that gets a showing has these characteristics, but I am interested in whether such work actually provides the viewer with the chance to reach any new insights. If the art is pure gimmick, it does not, as there is nothing there but the gimmick.

At first when I read this article about Heather Dewey-Hagborg’s work, I thought “man, there is a gimmick”. The work does have all the hallmarks of gimmicky art, as it rests on a single, novel, clever idea: to collect DNA from public places and try to use it to create portraits of unknown and unwitting DNA donors. But as I thought a bit more about the work, I began to realize that it does rise well above pure gimmick, as it has amazing potential to make people think a lot differently about the emerging genetic technologies that are transforming much of our society.

Dewey-Hagborg collects personal objects left behind by people that are likely to contain their DNA. Cigarette butts seem to be a favorite, apparently because the act of smoking leaves a lot of genetic material behind in the filter (one could also point out that cigarette butts are among our most ubiquitous throw-away items that end up being tossed everywhere, but that is a different conversation). She then extracts and amplifies this DNA and sequences particular regions with the potential to tell her about the facial features of that person. This genetic data gets plugged into a computer model that then spits out a 3D image of the DNA donor’s potential face, which she has rendered by a 3D printer (and I assume paints to add color). Lots of cool technology here, but why?

What I hope that people get out of this work is a new perspective on the promises and perils of emerging genetic technologies. An immediate reaction that this work ought to engender is “wow, I am leaving information about myself everywhere”. The fact that Dewey-Hagborg was able to make guesses about the general appearance of the people who left behind their gum or cigarette butt should provoke people to wonder how this kind of ubiquitous information might be used to profile us in the future. Whatever DNA sequencing does or will allow us to know, that information will be potentially knowable about pretty much everyone because we are leaving genetic traces of ourselves everywhere. This is one of the valuable features of Dewey-Hagborg’s choice to use anonymous donors, as the work is kind of ‘modern archaeological’ in its approach to reconstructing a particular event: we get to imagine something about an unknown person based solely on their genetic code. This choice also comes with some liabilities, which I discuss below.

I think this work also brings up some really important questions about the difference between genetic identity versus cultural identity. Looking at her work alone, I found myself struggling to categorize the faces. They are a bit creepy, and definitely suffer from living in the uncanny valley. But when I read about the information found in the genetic sequences of these DNA donors — including their probably regional origin — I immediately began to ascribe cultural qualities to them. I do not love that I do this, as it is a kind of ‘precursor to racism’: show me the face without context and I don’t make assumptions, but give me a region that I connect with a culture and I make all sorts of assumptions. I suspect that most people do this, and Dewey-Hagborg’s work should compel us to see that genes and culture are not the same things. For all we know her brown-eyed man of Eastern European descent was a yogi master who exclusively eats South African food: the DNA tells us very little about what this person actually is like. We tend to ascribe way too much significance to genetic ancestry, and most people falsely believe that what we call “race” or “ethnicity” has a genetic rather than cultural basis. Looking at these portraits, their origin, and our reaction to them will lead thinking people to examine their own understanding of the meaning of genetics versus culture.

Finally, I think that Dewey-Hagborg’s work raises important ethical questions. She raises these questions by crossing an ethical line: she is displaying information about people obtained by sequencing their DNA without their consent. How much information one can reveal about people based on the DNA they leave behind is an emerging bioethical question, one that is raised by other recent studies that use public sampling of DNA to answer scientific questions. I do not feel that Dewey-Hagborg has committed a major ethical violation here. Instead, her work sits close enough to that ethical boundary line that it compels us to ask where the line ought to actually exist. One might say “if you leave your DNA in public, you can expect that someone might sequence it”. But given how impossible it is to not leave your DNA behind, we cannot compare public DNA to say leaving your diary out for others to read. Dewey-Hagborg’s work is smart because it demonstrates proof of concept that very soon, anyone can genetically profile anyone else. We need to figure out how to ethically deal with that new technological reality.

I have a suggestion for Dewey-Hagborg on what should be her next work. What I would like to see her do is implied by her own self-portrait, which was included in her display of portraits of her anonymous DNA donors. Her self-portrait is not all that impressive, as it only very vaguely looks like her. Part of this has to do with age: as the Smithsonian article nicely points out, you cannot tell from the DNA collected whether the person smoking the cigarette is 15 or 95, so the model is of a 25-year-old version of the DNA donor. I am not aware of Dewey-Hagborg’s age, but her self-portrait seems to be younger than her. But the self-portrait also looks kind of generic, and it certainly does not capture the uniqueness of Dewey-Hagborg’s actual face. These unique characteristics in people are the very things that we use to recognize others, so in a way her portraits are very far off from representing the “real” people whose DNA they profiled.

What I am suggesting is that Dewey-Hagborg do a second exhibition using the same methodology but with known DNA donors (or better yet, she should not know who the donors are, but a collaborator would, making this a “single blind” project). These known DNA donors should be exhaustively photographed, or have an actual cast of their face made, or should actually be a part of the exhibit: whatever it takes to contrast the outputs of this DNA-driven computer model with the real people. What I suspect would be the outcome of this exhibition would be some level of disgust, distrust, and disappointment, as the DNA-derived portraits would likely be poor facsimiles of the real people. I think that it is important that people realize the limits of our genetic knowledge, and such an exhibit would focus on these limits. It is critical in the midst of our rising enthusiasm for genetic technologies that we do not hand over more trust than they deserve. Can they tell you coarse things about a person’s skin color or facial features? Perhaps. Can they provide an accurate profile worthy of a crime poster? I would suspect not, but such an exhibit could ask that question.

Or, alternatively, someone could subject her model to scientific analysis, asking the question “how accurately do her reconstructions represent the actual DNA donors?”. I am curious what such an analysis would tell us about the “profile value” of DNA based on our current understanding of genetics. I suspect that our understanding of how genotype translates into phenotype is still way too basic to generate meaningful predictions of appearance from genetic sequences, and I am skeptical about how soon we will be able to untangle the diverse genetic (and environmental) interactions that lead to actual phenotypes.

A Major Post, Articles, Computer Science, Ethics, Genetics, Human Evolution, Risk & Uncertainty, Science in Art & Design, Sociology

Leave a Reply