Matriculating through the STSS certificate program certainly influenced the way I think about my dissertation research. I came to understand my object of study as a complex ecosystem of direct-to-consumer (DTC) genetic testing and third-party interpretation, with the “raw” genotype data file as the technical artifact at the center. Focusing on that genotype file, I can trace it from DTC company to users/downloader to third-party interpretation tool, and the movements of that file circumscribe my scope of inquiry. I am of course still interested in people — DTC consumers and third-party tool developers (indeed, these are the people I’m surveying and interviewing!) — but viewing this as a technical system united by the existence of the genotype data file was informative in drawing and describing the boundaries of my project.
The material I was exposed to in COM 539, “Theories of Technology and Society,” with Dr. Neff was the most useful in terms of layering theory onto my project. The ideas I found most relevant were (1) the social construction of “raw” data, (2) dialectic of database and narrative, and (3) power of algorithms. I will briefly summarize their contributions here.
Social construction of “raw” data
First, reading ‘Raw’ Data is an Oxymoron (Gitelman 2013) helped me understand why I do (and should) put “raw” data in scare quotes, though not for the reasons I had initially thought. Before this class, my main reason for the quotes was because there are several upstream laboratory processes require to get a set of called genotypes — i.e. there are “raw-er” forms of the genetic information. However, Gitelman and Jackson’s edited volume helped me understand how data are never raw because they are not given pre-factually. Instead, data must first be imagined as data by individuals — the individuals who collect or generate it, and then the ones that use it. The genetic data generated by DTC companies is imagined in several ways — firstly by the DTC companies deciding which genetic variants to test for, and then subsequently by the customer who decides to download their data file and port it to third-party tools for further analyses.
Database and narrative
In Brown’s “Ethical Programs: Hospitality and the Rhetorics of Software,” he outlines how people’s experience of the world is a dialectic between database and narrative (Brown 2015). The two are interdependent, each requiring the other to make sense and be of use. This resonates strongly for me with DTC testing — why people do it, why they get their data for further self-interpretation, and how they “take in” (or not) all of the resulting information. For the full scope of information people may get, from ancestry to health to identifying family members, they filter and contextualize the results based on their existing understanding of themselves: their family, their health risks, their heritage. Empirical work has shown how users of DTC genetic ancestry tests end up weaving in the results with their existing familial, cultural, and racial narratives (Nelson 2008). The implication is that even if data were objective and predictable, the interaction of data with personal narrative is going to be far more unwieldy. Because of this complexity, it is hard to foresee whether people will be helped or harmed by pursuing self-directed analysis of their own genetic data. In practice, for my survey and interviews with DTC customers, this made me realize that my lines of questioning need to interrogate people’s databases and narratives.
Power of algorithms
Algorithms are central to interpretation of genetic information. Arguably, this is where most of the industry competition is in genetics, given that generating DNA sequence has become fairly fast, inexpensive, and straightforward. Algorithms are the site of responsibility and power, and indeed are often inscrutable “black boxes” to those affected by them (and indeed those whose data fuels them) (Pasquale 2015). Genetic interpretation algorithms are key to my project as well. Lay users of consumer genetic testing likely lack the expertise to analyze their data on their own. Instead, they turn to third-party tools (in addition to the DTC companies) to extract further meaning from their genomes. The algorithms of these third-party tools are what furnish people with information, with varying levels of transparency. I undertook this project in part because, through a decade working in human genetics research, I understand many of the analytic and bioinformatic approaches used by third-party interpreters. Despite my working knowledge, however, there are still many tools I find inscrutable. The algorithmic power, and responsibility, lies with the creators of these tools.
Conclusion
While I am not ultimately writing an “STSS dissertation,” the courses, readings, and writing I undertook through the certificate program have shaped how I approach my dissertation work. As an interdisciplinary scholar, the goal of bridging across disciplines and modes of inquiry is one my passions. Therefore, my exposure to STSS and my experiments layering these ways of thinking onto my dissertation work, only strengthen my capacity to do that bridging work.
References
Brown, James J. 2015. “Rhetorical Devices: Database, Narrative, and Machinic Thinking.” in Ethical programs: Hospitality and the rhetorics of software.
Gitelman, Lisa. 2013. “Introduction.” in Raw Data Is an Oxymoron, edited by L. Gitelman. Cambridge, MA: MIT Press
Nelson, Alondra. 2008. “Bio Science: Genetic Genealogy Testing and the Pursuit of African Ancestry.” Social Studies of Science 38(5):759–83
Pasquale, Frank. 2015. “Introduction – The Need to Know.” in The Black Box Society.