The recognition of teams as complex dynamic systems was a hallmark and among the earliest considerations of research on team functioning. However, the popularization of conceptual heuristics such as the input-process-outcome (IPO) framework and the accessibility of methodological, analytical, and meta-theoretical principles from multilevel theory (MLT) have resulted in a disconnect between contemporary theory and empirical research on teams and this foundational perspective. Thus, the primary motivation for the present paper is to facilitate and stimulate future research on team phenomena that embraces systems thinking. To do so, we describe key concepts, terminology, and ideas from specific branches of the systems sciences—namely open systems theory, dynamical systems, and agent-based systems—that have direct relevance for researching team phenomena as complex systems. Additionally, a comparison between two example models of team performance that are rooted in an IPO+MLT versus a systems-oriented perspective is offered to highlight the difference in foci, applications, and inferences these approaches offer. The paper concludes with a summary of key advantages as well as potential obstacles for reintroducing systems-thinking back into team science.

Team cognition has been identified as a critical component of team performance and decision-making. However, theory and research in this domain continues to remain largely static; articulation and examination of the dynamic processes through which collectively held knowledge emerges from the individual- to the team-level is lacking. To address this gap, we advance and systematically evaluate a process-oriented theory of team knowledge emergence. First, we summarize the core concepts and dynamic mechanisms that underlie team knowledge-building and represent our theory of team knowledge emergence (Step 1). We then translate this narrative theory into a formal computational model that provides an explicit specification of how these core concepts and mechanisms interact to produce emergent team knowledge (Step 2). The computational model is next instantiated into an agent-based simulation to explore how the key generative process mechanisms described in our theory contribute to improved knowledge emergence in teams (Step 3). Results from the simulations demonstrate that agent teams generate collectively shared knowledge more effectively when members are capable of processing information more efficiently and when teams follow communication strategies that promote equal rates of information sharing across members. Lastly, we conduct an empirical experiment with real teams participating in a collective knowledge-building task to verify that promoting these processes in human teams also leads to improved team knowledge emergence (Step 4). Discussion focuses on implications of the theory for examining team cognition processes and dynamics as well as directions for future research.