Deciding with muscles

Abstract

Sequential sampling models have a rich tradition in psychology and play a central role in our understanding of human decision-making. According to these models, decisions arise from the gradual accumulation of noisy evidence over time until a decision threshold is reached. This framework accounts for both choice behavior and reaction times and is further supported by neurophysiological evidence: neural activity in various brain regions exhibits accumulation-to-bound dynamics consistent with model predictions. Recent empirical findings extend this view, revealing that similar accumulation-like signals can also be observed in the electrical activity of response muscles, suggesting that decision signals may propagate continuously from perceptual processing through motor execution. This raises a fundamental theoretical question: How are effortful actions initiated when sensory evidence is weak, particularly under time pressure? We hypothesized that evidence-independent urgency signals provide the additional drive needed for the translation of decision signals into muscle activation in such contexts. To formalize this hypothesis, we extended the gated cascade diffusion model, a computational framework that links decision formation, motor preparation, and motor execution. We tested this model extension against behavioral and neuromuscular data from two experiments manipulating sensory evidence quality and required response force, with speed pressure additionally manipulated in Experiment 2. Model fits and formal comparisons with alternative accounts provided strong support for our hypothesis. These findings enhance our understanding of the interface between decision-making and motor systems, particularly in contexts that require effortful actions, which are common in everyday behavior.