Locking-in “Foreign DNA”: How Socio-Semantic Networks Select a GE–GM Classification Standard

• In Event: Science, Knowledge & Technology Section Roundtables
  In Refereed Roundtable (60 minute): Table 2: Distinguishing difference: classification, standardization, and racialization

Abstract

While prior research shows that standardization lets classifications travel without requiring full agreement, it rarely explains the upstream selection problem. When multiple portable criteria are available, why do some become the field’s standard reference point? I argue that socio-semantic networks make this standard reference selection possible through two coupled mechanisms. First, collaboration ties and durable carriers such as templates, repositories, and evaluative infrastructures, unequally amplify the reach and visibility of candidate formulations, making some far more likely to be encountered, reused, and reproduced as defaults. Second, a formulation succeeds when interpretive differences are clustered: within-group agreement and between-group divergence let communities absorb the same formulation into their routines with minimal disruption, lowering adoption costs and failure risks and making it easier to reuse and embed as a default. Empirically, I analyze 111,504 Genome Editing and Genetic Modification research articles and construct institution-level collaboration networks to trace how the “foreign DNA” criterion—a portable classificatory interface for sorting interventions and products across the GE/GM boundary—consolidates among alternative classification scheme, and how its implementation clusters by collaboration communities. The result shows that “foreign DNA” becomes the default interface because collaboration networks both amplify its replication through structured exposure and organize its evolving interpretations into community-aligned clusters that contain disagreement, lowering reuse risk and locking the idea of “foreign genes” in over time.

Author

• Xin Zhao, University of North Carolina-Chapel Hill