LAUSR

This article examines the creation and stabilization of early-stage boundary objects by voluntary teams spanning divergent professional and scientific fields. Cross-disciplinary collaborators can share similar goals, yet nonetheless face frictions from differences in professional expertise, practices, and technical systems. Yet if boundary objects help to span disciplinary divides, the same challenges are likely to hinder initial boundary object development. Comparative ethnography of three projects adapting Grid computing technology to fields of science highlights challenges for boundary object creation, including a “mind-set shift” before the technology could stabilize. Enriching our knowledge of boundary object beginnings, we find successful stabilization requires both appropriate localization and further resources, which enable the simultaneously global–local nature of boundary objects. This essential feature is understudied in management research. Developing the boundary object concept on its own terms enhances empirical and theoretical application, particularly when researchers prefer one main theory of objects, rather than a “pluralist” approach.