The rationalities behind the adoption of cyberinfrastructure for e-science in the early 21st century U.S.A.

Based on grounded theory and thematic analysis of 70 in-depth interviews conducted over 32 months (from November 2007 to June 2010) with domain scientists, computational technologists, supercomputer center administrators, program officers at the National Science Foundation, social scientists, policy analysts, and industry experts, this dissertation explores the rationalities behind initial adoption of cyberinfrastructure for e-science in the early 21st century U.S. This dissertation begins with Research Question 1 (i.e., how does cyberinfrastructure's nature influence its adoption process in early 21st century U.S.?) and identifying four areas of challenging conditions to reveal a lack of trialability/observability (due to the participatory/bespoke nature), a lack of simplicity (due to the meta/complex characteristic), a lack of perceived compatibility (due to the disruptive/revolutionary quality), and a lack of full control (due to the community/network property). Then analysis for Research Question 2 (i.e., what are the rationalities that drive cyberinfrastructure adoption in early 21st century U.S.?) suggests that there are three primary driving rationalities behind adoption. First, the adoption of cyberinfrastructure as a meta-platform of interrelated technologies is driven by the perceived need for computational power, massive storage, multi-scale integration, and distributed collaboration. Second, the adoption of cyberinfrastructure as an organizational/behavioral practice is driven by its relative advantages to produce quantitative and/or qualitative benefits that increase the possibility of major publications and scientific reputations. Third, the adoption of cyberinfrastructure as a new approach to science is driven and maintained by shared visions held by scientists, technologists, professional networks, and scientific communities. Findings suggests that initial adoption by pioneering users was driven by the logic of quantitative and qualitative benefits derived from optimizing cyberinfrastructure resources to enable breakthrough science and the vision of what is possible for the entire scientific community. The logic was sufficient to drive initial adoption despite the challenging conditions that reveal the socio-technical barriers and risky time-investment. Findings also suggest that rationalization is a structuration process, which is sustained by micro individual actions and governed by macro community norms simultaneously. Based on Browning’s (1992) framework of organizational communication, I argue that cyberinfrastructure adoption in the early 21st century lies at the intersection of technical rationalities (i.e., perceived needs, relative advantages, and shared visions) and narrative rationalities (i.e., trialability, observability/communicability, simplicity, perceived compatibility, and full control).