Workload as Coordinative Instability: A Dynamical Reframing for Human Factors

Workload remains one of the central constructs in Human Factors, yet its dominant theoretical treatments continue to frame it as the occupation of limited internal resources. This paper argues that the main difficulty is not simply how workload is measured, but what kind of phenomenon it is. Resource-based models such as Multiple Resource Theory and Cognitive Load Theory have been productive, but they rely on indirect constructs, unstable category boundaries, and a processing architecture that becomes less convincing in dynamic, interactive task settings. As an alternative, this paper develops a coordination-stability account in which workload is defined as instability in the coupled agent--environment system: the erosion and reorganization of perception--action coordination under shifting task constraints. This framework is first developed conceptually through ecological psychology, coordination dynamics, and representative design. It is then evaluated against a resource-based account using findings from the Multi-Attribute Task Battery (MATB), where the two frameworks generate different expectations about performance, individual differences, behavioral dynamics, subsystem coupling, and workload measurement. The MATB results are more consistent with the coordination-stability view: performance changes are non-uniform and partly non-monotonic, workload signatures are stable within individuals but transfer poorly across them, behavioral dynamics reorganize under increasing demand, and movement-based measures outperform traditional task metrics in classifying workload. The paper concludes that reframing workload as coordinative instability resolves persistent anomalies in the literature, and provides a richer methodological and practical toolkit for Human Factors research, including representative task design, dynamical measurement, and interventions aimed at improving performance by supporting stable coordination.