An exploration of complex action stopping across multiple datasets: Insights into the mechanisms of action cancellation and re-programming.

A long history of psychological experiments has used stop signal paradigms to assess action inhibition. Recent studies have investigated complex stopping behaviours, such as response-selective stopping where only one component of a bimanual action requires cancellation. A current emphasis has been to use electromyographical (EMG) recordings to assess the temporal dynamics of action inhibition at the level of the muscle, beyond those based solely on observable behavioural events. Here, we combine EMG and behavioural data from 17 cohorts of healthy younger and older adults yielding over 42,000 response-selective stopping trials, providing unique insights into this emerging field. Expanding from past research in this area, our robust single-trial EMG analyses permit detection of cancelled (partial) and response-generating EMG bursts in both hands, revealing substantial overlaps in the distributions of timing of action cancellation and re-programming. These findings are consistent with recent experimental and modelling evidence, suggesting that response-selective stopping is best modelled as two independent processes: a discrete bimanual stop and initiation of a new unimanual response. This overlap may be incompatible with the recent pause-then-cancel model, and more consistent with a broader "pause-then-retune" account, where a slower process mediates any action updating, not just cancellation. Moreover, this independence means that cancellation can happen at any time during motor planning and execution, against the notion of an observable "point of no return" in terms of EMG and behavioural measures. We also discuss best practices for the analysis of EMG data and indicate how methodological aspects, such as choosing appropriate reference time points, can influence the outcomes and their interpretation.