Paper short abstract:

This talk investigates how technicians in a molecular genetics laboratory manage to constitute a unified sample over the course of radical material transformations – from liquid blood to digital code – with a particular focus on the role of informatic ‘noise’ and its recognition and removal.

Paper long abstract:

When patients come into the clinic for genomic testing, they carry with themselves an enormous variety of potentially meaningful materials. As their cases move through the testing process - from sample collection to DNA extraction to analysis - the types of objects that serve as evidence for the clinicians and scientists working on their cases change radically. The object of analysis may be - for example - a physical body, peripheral blood, a DNA strand, or a printed report of Gs, Cs, Ts, and As. Because of the diversity of qualities thrown up by these divergent avatars of the patients' cases, significant social work has to be practiced in order to maintain the ontological coherence of the case across sociotechnological time. At moments of transformation, the materials act symbolically as boundary objects, and I lay out the modalities by which different conceptualizations are enacted as ways of referring to "the same thing." I discuss the 'transduction' of patients' cases in terms of both signifiers (naming practices) and signifieds (physical objects). I look at how the patient's body is represented as singular while erasing the multiplicity inherent in these acts of representation. In particular, I focus on the types of qualities that get sloughed off with the 'noise' during the process. (For instance, when a tissue sample is transformed into DNA, it is no longer intrinsically obviously from which organ it was derived.) Such sieving is itself standardized, and the types of 'noise' that remain inseparable from desirable data complicate this purification process.