Individual Submission Summary
Share...

Direct link:

Coding in a Lab: Toward a Micro-Sociology of Computer Programming

Wed, September 4, 1:00 to 2:30pm, Sheraton New Orleans Hotel, Floor: Four, Southdown

Abstract

Interest in machine learning (ML) has recently increased significantly among the STS community. However, few STS-inspired ethnographers have tried to account for computer programming, an activity that is yet central to the actual development of ML algorithmic systems. Among ML-related agitation, a surprising silence seems then to subsist: Even though writing numbered lists of machine-readable symbols is constitutive to the shaping of computerized methods of statistical learning, these scriptural practices and the sociotechnical issues to which they relate still resist ethnographic analysis.
Taking stock of this situation, we make methodological and theoretical suggestions aiming to support and promote the micro-sociological analysis of computer programming practices. First, based on an ethnographic inquiry conducted in a computer science laboratory for digital signal processing, we present an investigation technique that allowed us to make appear the material thickness of computer programming and document some of its constituent relationships. Second, drawing upon collected materials, we make the analytical proposition that a significant aspect of computer programming practices consists in aligning inscriptions in order to gain access to remote entities and, in turn, index a location within a numbered document. Finally, we present some of the macro-social phenomena these micro-sociological insights help to localize and make visible.

Author