Edmondson, AmyBransby, Derrick2026-07-0720262026-06-052026Bransby, Derrick. 2026. Dreams to Machines: How Teams Navigate the Messy Middle of Innovation. Doctoral Dissertation, Harvard University Graduate School of Arts and Sciences.32696883https://dash.harvard.edu/handle/1/42744037On their journey to innovate, teams confront daunting gaps between their vision for the future and the constraints of reality. This distance inspires creativity but complicates execution. Novel ideas often fail to materialize not for a lack of potential, but because teams struggle to execute and learn in tandem. They strive to keep shared understanding stable enough to coordinate, yet flexible enough to adapt. Prior research shows that constraints help create common ground for coordination. However, these studies usually treat constraints as features of a task, not as boundaries that teams shape over time. In this dissertation, I develop a process theory of constraining: the ongoing practice of articulating, negotiating, and revising boundaries of emerging problems and solutions. Constraining explains how innovation teams harness creative potential and make progress without prematurely resolving the tension that drives creative work. Using longitudinal qualitative data from field study of project teams at a live event production firm (Kinetiq), I examine how teams translate clients’ ambitious aspirations (“dreams”) into custom technology (“machines”) under immovable deadlines. I show how teams create tentative problem representations that make clients’ aspirations legible by distilling their essence into evaluative criteria and tethering them to shared references to establish working assumptions. These criteria and assumptions define the contours of emerging problems and establish a minimal basis for coordination. Over time, teams refine this basis through cycles of proposing, probing, and pivoting—a process my informants called the “design spiral”—clarifying and reshaping the problem space as solutions emerge. Within this larger process, I identify provisional commitments—open-ended, mutually affirmed decisions that enable action despite uncertainty—as the mechanism for progress. Such commitments establish a basis for coordination that is stable enough to support coordination but flexible enough to adapt, allowing teams to execute without hardening their creative trajectory. Provisional commitments make ambiguous problems actionable, converting tentative problem representations into task structures (scaffolds) that allow teams to learn and execute in tandem. This dissertation makes several contributions. First, it advances research on coordination and team cognition by specifying the practices through which teams generate the stability required to execute while preserving the flexibility to learn and adapt. In doing so, this work explains how common understanding emerges over time rather than being assumed at the outset. Second, it connects creativity and innovation research by shifting attention from ideas to the collective work that carries them toward realization, showing how teams actively shape the space where ideas develop. Third, it extends existing research that views uncertainty as a condition to navigate rather than mitigate, showing how teams enact disciplined flexibility by making provisional commitments that allow them to execute and learn in tandem. Together, these findings offer practical guidance for leaders navigating the persistent tension between creative ambition and the demands of implementation.application/pdfenConstraintsCoordinationCreative workTeam cognitionTeam innovationUncertaintyManagementOrganizational behaviorThesis or Dissertation2026-07-070000-0002-0402-4187