Response inhibition, the intentional stopping of actions, is often considered a dispositional ability central to adaptive everyday functioning, normal development, and good mental health. This ability is proposed to rely on an inhibition process implemented by a specific brain circuit that causes the arrest of motor activity and cessation of muscular output. Consequently, the integrity and functioning of this circuit should allow us to differentiate between the response inhibition ability of individuals and offer opportunities for intervention if this ability is impaired.
However, this understanding is largely based upon experimental investigations using the stop signal task and its associated model-based estimate of response inhibition latency (stop signal reaction time or the SSRT). Using cognitive tasks to elucidate the neural basis of behavior and cognitive processes requires that we link what we can observe with what we want to understand. This calls for assumptions in addition to those that we can derive from theory. For instance, measures of inhibitory circuit functioning are often identified by contrasting successful action stopping with instances where stopping fails or is unnecessary. In doing this, one assumes that such contrasts allow isolating differences in an inhibition process. Similarly, measures of the inhibition process are often validated through associations with the SSRT. In doing this, one assumes that SSRT variation is caused solely by variation in an inhibition process. And when using the SSRT or neural inhibition measures as endophenotypes, either for specific psychopathologies or shared vulnerabilities, one assumes that they capture stable and dispositional traits. If these assumptions are flawed, findings based on them become uninterpretable and inferences invalid.
Through a series of studies using stop signal task performance measures together with measures of central and peripheral nervous system activity, we have investigated these assumptions and found that several of them do not hold – different measures of response inhibition do not appear to capture the dispositional workings of the outlined inhibition process supposedly implemented by a specific brain circuit. Not only do we find that inhibition measures are unstable over time, but also that we do not know why they vary, neither between individuals nor over time. Consequently, we do not know what these measures reflect. In sum, this suggests that flawed assumptions have led us astray, leaving our current understanding of the neural substrates of response inhibition unanchored and invalid.