The discovery of new knowledge can shed light on how our universe works, uncover cures for complex and rare diseases, provide insights for using available knowledge resources better, and identify new products to enhance the productivity and safety of society. The generation of the latest knowledge, whether in scientific research or new product development firms, often occurs in teams due to the combined expertise that emerges when specialized experts share and exchange knowledge (Teodoridis, 2018). Scholars create new knowledge in the form of a jointly published paper, and contributors create intellectual output that they could not have done without their collaborators' diverse yet complementary inputs.

Such scholarly collaboration has even been called a ‘‘springboard for economic prosperity and sustainable development’’ (US Office of Science & Technology Policy 2000). nderstanding how these collaborations commence and evolve remains critical. Typically, interactions between contributors occur within a shared social context and emerge from and are perpetuated through social networks. These social networks can span disciplinary, departmental, or national boundaries. However, particular configurations of the networks in which collaborators are embedded can affect how much they can utilize one another’s expertise (Vestal & Danneels, 2023).

The social networks of collaborators can be shaped by various forces, which Leak and Chalkley (1997) argue can be attributed to endogenous or exogenous network factors. For example, investigators can change across time regarding their interests or skills, and the characteristics of who they want to work with can follow developmental curves associated with these shifts. On the other hand, external changes can come from macro forces like institutional changes, including deliberate efforts to draw scholars across fields, shaping network change.

The research design was partially planned and opportunistic. Leadership within this organization sought to promote opportunities for innovation by reducing the structural and social boundaries that often hindered interaction between disciplines and decided to encourage the establishment of internal, interdisciplinary teams. Following the official announcement of the intervention in October 2007, which encouraged faculty and staff to organize into problem or disease-focused teams, 64 interdisciplinary research (IDR) teams self-organized, producing proposals in the hopes of securing funding for their work. Teams drew together individuals from across departments and specializations, thus destabilizing organizational silos.

Our empirical research focuses on the networks of participants who participated in an organizational effort to foster new problem-focused teams. Joining one of these teams had the potential to disrupt an employee’s traditional networks because it shifted the location of collaborative activity from the department to the cross-disciplinary team.

To examine the evolution of collaborative relationships over time, we utilize a network approach and study newly formed scientific teams over fifteen years. We conceive of ties as joint scholarly publications produced within fifteen years. We propose that the initial pattern of intrateam networks will influence the co-production of knowledge and the evolution of collaborative networks for years to come. We focus most on patterns of interaction that we believe will affect members’ knowledge integration and creation, including (a) shared department, (b) shared team, and c) shared triangles; we will elaborate on how each of these patterns shape collaboration across time.

Our study of a bounded network within one organization focuses on the co-authorship ties that formed simultaneously due to an organizational initiative to draw people from across disciplines and fields into problem-focused teams. From this single point in time, we explore whether being on the same team results in greater strength of co-authorship ties compared to those outside of teams; we define a tie as co-authoring a published scholarly work, and tie strength is defined as the count of these co-authored works within a given period. We also investigate other factors influencing co-authorship tie strength, including prior collaboration, the number of mutual contacts, and being a member of the same primary department.

Analyzing data spanning 15 years, three years before intervention, and 12 years following the onset of these problem-focused teams, our findings shed light on how the social network of collaborators influences co-author productivity, giving us a better understanding of how particular exogenous shocks influence network dynamics in knowledge organizations.