Psychological Inhibition

Abstract

Psychological inhibition, which MeSH classifies under learning, is the suppression of a response, thought, or memory that is prepotent or already under way but no longer wanted, the braking function that lets a goal override habit and reflex. This article sets out how it is measured: the stop-signal race model that makes the latency of an invisible stopping process quantifiable, the interference-control tasks that pit an automatic response against an instructed one, and the cognitive forms that suppress unwanted thoughts and memories. It traces the construct to the individual-differences work asking whether inhibition is one ability or many, and to the right frontal circuitry that may implement it. Three interactive demonstrations let the reader run a stop-signal race and watch its inhibition function build, drive the congruency effect in a colour-word conflict task, and commit the errors of restraint a go/no-go task exposes.

Keywords: inhibition, response inhibition, stop-signal reaction time, interference control, cognitive control

Psychological inhibition is the capacity to suppress a response, thought, or memory that is automatic, habitual, or already initiated, so that behaviour can be governed by a current goal rather than by whatever the situation most strongly invites (Diamond, 2013). In the Medical Subject Headings vocabulary it is catalogued as descriptor D007266, defined as the interference with or prevention of a behavioural or verbal response even though the stimulus for that response is present. That definition points to the paradox at the centre of the construct: inhibition is an act that leaves no trace in behaviour, because its success is the absence of a response, and this is what has made it both hard to measure and conceptually slippery. The sections below build the idea up in the order the field found tractable: what inhibition is and why it resists direct observation, a working taxonomy of its kinds, the stop-signal race model that solved the measurement problem for response inhibition, the interference-control tasks that quantify the resolution of conflict, the cognitive and mnemonic forms that reach beyond action, the individual-differences programme that asks whether inhibition is unitary, the prepotency that every inhibition paradigm must first create, and finally the neural circuitry that carries the stop.

What Inhibition Is

Inhibition earns its place among the core executive functions by doing negative work. Where working memory holds information available and cognitive flexibility switches between tasks, inhibition suppresses: it overrides a response the situation would otherwise compel, whether that response is a reflex, a well-learned habit, or an action already launched (Diamond, 2013). The everyday examples are unremarkable precisely because inhibition usually succeeds without notice. A driver lifts off the accelerator as a light turns amber rather than pressing on; a reader ignores a flashing advertisement beside the text; a speaker catches a tactless remark before it is uttered. In each case the measurable outcome of inhibition is that something expected did not happen, and this is the construct's defining difficulty. A response that is successfully withheld leaves no reaction time, no error, no visible act to time or count, so inhibition cannot be observed the way a response can. It has to be inferred, and the history of the field is largely the history of the ingenuity required to make an absence quantifiable (Logan & Cowan, 1984). The problem is compounded by the term's breadth. Inhibition names a neuronal event, in which one cell reduces the firing of another; a behavioural operation, in which a prepared action is cancelled; and, in the older clinical literature, a personality disposition toward restraint. These senses are related only by metaphor, and treating them as one thing has caused lasting confusion (Nigg, 2000). The cognitive sense that concerns this article is the deliberate or automatic suppression of a prepotent mental or motor response in the service of a goal, and even within that sense, as the taxonomy below shows, several distinct operations travel under the single name.

A Taxonomy of Inhibition

The word inhibition is applied to a range of operations that share a family resemblance but differ in what is suppressed and how it is measured, and progress in the field has depended on keeping them apart. Joel Nigg, reviewing the disorder-linked literature, argued that the term had been stretched to the point of vacuity and proposed a taxonomy that separated effortful executive inhibition from automatic and motivational forms, a corrective that later cognitive work adopted in spirit (Nigg, 2000). For present purposes the useful cut is threefold, set out in Table 1. Response inhibition is the outright cancellation or withholding of a motor action, the stopping of a hand already moving or the restraint of one poised to move. Interference control, sometimes called resistance to distractor interference, is the narrower suppression of an irrelevant stimulus dimension or competing response so that a task-relevant one can govern behaviour, as when the meaning of a colour word must be held off so its ink colour can be named. Cognitive inhibition reaches beyond action to the contents of thought and memory, suppressing irrelevant representations, intrusive thoughts, and no-longer-relevant information. These are not merely three labels for one process. They dissociate in the tasks that measure them, in their developmental trajectories, and, as the individual-differences section will show, in how weakly they correlate across people, which is the strongest evidence that inhibition is a family rather than a faculty (Friedman & Miyake, 2004). What unites them is a common computational situation: in every case a strong or prepotent tendency must be overridden by a weaker, goal-relevant one, and one influential account holds that this is achieved not by a dedicated stop mechanism at all but by the ordinary competition among representations, the goal-relevant one being boosted until it wins (Munakata et al., 2011).

Table 1

Three Kinds of Psychological Inhibition

Note. The three kinds share the logic of overriding a prepotent tendency with a weaker goal-relevant one, but they dissociate across tasks, development, and individuals, which is why treating inhibition as a single faculty misleads (Nigg, 2000; Friedman & Miyake, 2004).

Response Inhibition and the Race Model

The measurement problem was solved most elegantly for response inhibition, the outright stopping of an action. Gordon Logan and William Cowan proposed that stopping be understood as a race between two independent processes: a go process, set in motion by the primary stimulus and running toward the response, and a stop process, set in motion by a later stop signal and running toward cancellation (Logan & Cowan, 1984). Whichever finishes first wins. If the go process reaches the finish before the stop process, the response escapes; if the stop process wins, the action is cancelled and nothing is seen. This deceptively simple idea does real work, because it turns the unobservable into the calculable. The stop process has a finishing time, the stop-signal reaction time, that can never be observed directly, since a successful stop produces no response to measure. But the race logic lets it be recovered. The experimenter varies the stop-signal delay, the interval between the go stimulus and the stop signal, and records how often a response nonetheless escapes. A short delay gives the stop process a large head start and almost always succeeds; a long delay lets the go process get so far ahead that the response escapes almost every time. The resulting curve, the probability of responding rising with delay, is the inhibition function, and its position on the time axis fixes the latency of the stop process. Because a response escapes exactly when the go process finishes before the stop process does, the go reaction times that beat the stop are the fast ones, and treating the observed escape probability as the fraction of the go distribution that outran the stop yields the stop-signal reaction time by subtraction (Verbruggen & Logan, 2008). A typical adult value is roughly 200 milliseconds, and it lengthens in attention-deficit disorder, in ageing, and after damage to the right frontal lobe. The race is set out in Figure 1. The demonstration below runs it, letting the reader set the stop-signal delay and watch the inhibition function build point by point as the escape probability climbs from near zero to near certainty.

Interference Control

Not every act of inhibition cancels a whole response; often the task is subtler, to hold off one dimension of a stimulus so that another can govern the answer. The paradigm case is the task devised by John Ridley Stroop, who found that naming the ink colour of a word takes far longer, and produces far more errors, when the word itself spells a conflicting colour than when it spells a neutral string or a matching one (Stroop, 1935). Reading is so overlearned that the word meaning is processed automatically and primes its own naming response, which must be suppressed for the ink colour to be reported; the extra time on incongruent trials, the congruency effect, is a direct index of the interference and of the control brought to bear against it (MacLeod, 1991). The same logic drives the flanker task of Barbara and Charles Eriksen, in which a central target arrow or letter is flanked by distractors that point the same way or the opposite way, and the incongruent flankers, priming the wrong response, slow and disrupt the answer to the target (Eriksen & Eriksen, 1974). What makes these tasks theoretically central is that they yielded the first mechanistic model of the control that resolves the conflict. Jonathan Cohen and colleagues showed that a parallel-distributed-processing network with a strong, well-trained word-reading pathway and a weaker colour-naming pathway reproduces the Stroop effect exactly, provided that an attentional signal biases processing toward the weaker, task-relevant pathway; interference control, on this account, is not a separate stop but the top-down boosting of the pathway that would otherwise lose (Cohen, Dunbar, & McClelland, 1990). A later addition specified where the bias comes from: a conflict-monitoring system, centred on the anterior cingulate, detects the co-activation of incompatible responses on incongruent trials and recruits control in proportion to the conflict, so that inhibition is summoned only when it is needed (Botvinick et al., 2001). The demonstration below sets the congruency effect out directly, comparing the mean time to name an incongruent colour word with the mean time to name a congruent one, and letting the reader widen the interference by making the reading response more automatic.

Cognitive and Memory Inhibition

Inhibition is not confined to the motor system. The same braking logic appears when the object to be suppressed is a thought, a percept, or a memory, and some of the strongest evidence that inhibition is an active process rather than mere decay comes from these cognitive forms. Negative priming is the clearest case: when a stimulus that was ignored on one trial becomes the target on the next, responses to it are slower than to a wholly new stimulus, as though the act of ignoring it left a lingering suppression that must be undone (Tipper, 1985). The cost is small but reliable, and it is difficult to explain without positing that the ignored item was actively inhibited rather than simply not selected. A broader proposal casts inhibition as the gatekeeper of attention itself. Lynn Hasher and Rose Zacks argued that efficient thought depends less on the capacity to hold information than on the capacity to keep irrelevant information out, and that much of the cognitive decline of ageing, and the intrusive, unfocused quality of some older adults' thought, reflects a weakening of this inhibitory gate rather than a loss of storage (Hasher & Zacks, 1988). On this view the functional capacity of working memory, so often treated as a fixed store of a few items (Cowan, 2001), depends heavily on how well irrelevant contents are inhibited from occupying it. Inhibition can also be turned deliberately on memory. Michael Anderson and Collin Green had people learn word pairs and then, given the cue, practise not retrieving the associate, and found that this repeated suppression made the avoided memories harder to recall later than baseline items that had simply not been practised, a suppression-induced forgetting that implicates the same executive machinery used to stop an action (Anderson & Green, 2001). Across action, attention, and memory, then, a common signature recurs: a prepotent or intrusive representation is actively held down, at a measurable cost, in the service of a current goal.

Unity and Diversity

If response inhibition, interference control, and cognitive inhibition all deserve the name, a natural question follows: are they expressions of one underlying ability, so that a person good at one is good at all, or a loose collection of task-specific skills? The question is answered by measuring many people on many inhibition tasks and asking how the scores covary, and the answer has been sobering for anyone expecting a unitary faculty. Akira Miyake, Naomi Friedman, and colleagues gave a large sample a battery of executive tasks and used latent-variable modelling to separate what the tasks share from what is specific to each, finding that inhibition, updating, and shifting are separable but correlated functions rather than one general control (Miyake et al., 2000). When they looked within inhibition, the news was starker still: tasks meant to tap the same inhibitory ability, prepotent-response inhibition and resistance to distractor interference among them, correlated only weakly with one another, so that a common inhibition factor was hard to extract at all (Friedman & Miyake, 2004). The individual differences that do exist are substantially heritable, and much of what the tasks share is shared genetically (Friedman et al., 2008). These findings forced a striking revision. In a later synthesis the same group concluded that the variance once attributed to a distinct inhibition factor was better explained by the general factor common to all executive tasks, so that inhibition, uniquely among the executive functions, had no reliable variance of its own once the common factor was removed (Miyake & Friedman, 2012). Whether this means that inhibition is not a separate function, or that the tasks are simply too impure to isolate it, remains debated. But it fits a wider theoretical current, discussed next, that questions whether a dedicated inhibitory mechanism exists at all, or whether the appearance of inhibition emerges from more general processes of activation and competition.

Prepotency and the Go/No-Go Task

Every inhibition task must first manufacture the very thing it asks the participant to suppress. A response cannot be inhibited unless it is prepotent, strong enough that it would occur by default, and much of the craft of designing these tasks lies in building that prepotency so that inhibition has real work to do. The go/no-go task makes the logic explicit. The participant is shown a stream of stimuli and told to respond as fast as possible to frequent go signals and to withhold the response to rare no-go signals. The high frequency of go trials, and the premium on speed, together forge a strong tendency to respond, so that the occasional no-go signal arrives with a prepared action already building, and the participant must restrain it (Diamond, 2013). Failures, presses to a no-go signal, are commission errors, and their rate is the task's index of poor response restraint. The go/no-go and the stop-signal task are often bracketed together as measures of response inhibition, but they differ in a way that matters. In the go/no-go task the no-go signal appears at the same time as the imperative stimulus, so the response is restrained before it is truly launched; in the stop-signal task the stop signal follows the go stimulus, so an action already under way must be cancelled in flight (Verbruggen & Logan, 2008). Restraint and cancellation are not the same feat, and the two tasks load on partly different processes, a further instance of the diversity that troubles any attempt to treat inhibition as one thing. The demonstration below runs a go/no-go stream in which the proportion of go trials sets the strength of the prepotent response, letting the reader see how a higher go rate builds a stronger habit and drives up the commission errors on the rare no-go trials.

Neural Basis

The search for where inhibition happens has centred on the right frontal lobe. Adam Aron, Trevor Robbins, and Russell Poldrack, drawing together lesion, imaging, and stimulation evidence, argued that the right inferior frontal cortex is critical for stopping: damage there lengthens the stop-signal reaction time in proportion to the extent of the injury, the region activates when a response is cancelled, and disrupting it impairs stopping (Aron, Robbins, & Poldrack, 2014). Their account traces a fast pathway from the right inferior frontal cortex through the subthalamic nucleus of the basal ganglia, which can apply a global brake to the motor system quickly enough to catch an action in flight. The basal ganglia are central to the fuller circuit. A broader review of the neural basis of response control locates stopping in the interplay between frontal cortex and the basal ganglia pathways that gate movement, and ties individual differences in inhibitory control to the same circuitry whose dysfunction marks impulsivity, attention-deficit disorder, and addiction (Bari & Robbins, 2013). Yet the localisation is not without challenge, and the challenge returns to the conceptual question. Yuko Munakata and colleagues proposed that much of what looks like dedicated inhibition is better understood as the outcome of competition: a unified framework in which goal representations in prefrontal cortex bias the competition among possible actions, so that an unwanted response is defeated not by a special stop signal but by the strengthening of its rival (Munakata et al., 2011). On this reading the right inferior frontal cortex may support the representation and application of goals that resolve competition, rather than housing an inhibition module as such. The debate between a dedicated braking mechanism and an emergent, competition-based account remains live, and it mirrors at the neural level the psychological uncertainty over whether inhibition is a thing or a name for what several other things accomplish.

Worked Example

The stop-signal reaction time is worth deriving by hand, because its logic is the whole of the race model and it is exactly what the stop-signal demonstration computes. Suppose a participant's go responses have a mean reaction time of 500 milliseconds, distributed around that mean with a standard deviation of 80 milliseconds, and that the true latency of the stop process is 200 milliseconds, the quantity to be recovered. On stop trials the experimenter sets the stop-signal delay to 260 milliseconds. The stop process therefore finishes, on average, at the delay plus its own latency, 260 plus 200, which is 460 milliseconds after the go stimulus. By the race logic a response escapes exactly when the go process finishes before that, so the probability of responding on a stop trial equals the proportion of the go reaction-time distribution that falls below 460 milliseconds. A value of 460 lies half a standard deviation below the mean of 500, since 460 minus 500 is negative 40, and 40 is half of 80, so about 0.31 of the go distribution is faster than 460 milliseconds. The model therefore predicts that roughly 31 percent of responses will escape at this delay, and that is what the experimenter observes. Now run the inference in reverse, which is what a real experiment must do, because the stop latency is unknown. Given an observed escape probability of 0.31, the integration method finds the reaction time below which that proportion of go responses fall, the 31st percentile of the go distribution, which is 460 milliseconds. This is taken as the moment the stop process finished. Subtracting the stop-signal delay recovers the latency of the stop process itself: 460 minus 260 is 200 milliseconds, exactly the value built in. A staircase procedure that adjusts the delay until responses escape on half of stop trials reaches the same answer by a shorter route: at a 50 percent escape rate the stop finishes at the median go reaction time, 500 milliseconds, so with the delay converged to 300 milliseconds the stop-signal reaction time is 500 minus 300, again 200 milliseconds. Setting the demonstration's stop-signal delay to 260 reproduces the 31 percent escape rate and the 200-millisecond estimate; moving the delay traces out the inhibition function while the recovered stop latency stays fixed, as a genuine property of the participant should.

Discussion

Psychological inhibition is a construct built to name a non-event, the response that did not occur, and its intellectual history is the story of making that absence measurable and then asking whether the thing measured is one ability or many. The first achievement was methodological. The stop-signal race model turned the unobservable latency of a cancelled action into a number, the stop-signal reaction time, by reading it off the probability that a response escaped, and the interference paradigms of Stroop and the Eriksens turned the resolution of conflict into the reliable arithmetic of the congruency effect (Logan & Cowan, 1984; Stroop, 1935; Eriksen & Eriksen, 1974). The second was mechanistic: connectionist and conflict-monitoring models showed that the interference effects need not imply a dedicated stop at all, but fall out of a strong pathway biased against by a top-down control signal that is itself summoned by conflict (Cohen et al., 1990; Botvinick et al., 2001). The third was to extend the same braking logic beyond action, to the negative priming of ignored percepts, the inhibitory gating of attention, and the deliberate suppression of memories (Tipper, 1985; Hasher & Zacks, 1988; Anderson & Green, 2001). Running through all of this is an unresolved tension. The individual-differences evidence finds that inhibition tasks correlate weakly and that a distinct inhibition factor may have no variance of its own once a general control factor is removed (Friedman & Miyake, 2004; Miyake & Friedman, 2012), and a strand of theory holds that inhibition is emergent, the visible result of goal-driven competition among representations rather than a mechanism in its own right (Munakata et al., 2011). The neural evidence is caught in the same tension, torn between a fast right-frontal braking pathway and a more distributed, competition-based picture (Aron et al., 2014; Bari & Robbins, 2013). What is not in doubt is the phenomenon. Prepotent responses are overridden, conflicting information is held off, unwanted memories are pushed down, and the failure of these functions is among the most reliable markers of developmental and clinical disorder (Nigg, 2000). Whether inhibition is best explained as a faculty or as the shadow cast by more general processes, the work it does is real, and the paradigms that measure it are among the most productive in the study of cognitive control.

Common Misconceptions

A successful inhibition can be timed like a response.

A successful stop produces no response, so it has no directly observable reaction time. The latency of the stop process must be inferred from a model of the race between going and stopping, by reading it off the probability that a response escapes at each stop-signal delay (Logan & Cowan, 1984; Verbruggen & Logan, 2008).

Inhibition is a single ability that a person either has or lacks.

Tasks meant to measure the same inhibitory ability correlate only weakly, and latent-variable work finds that a distinct inhibition factor may carry no reliable variance of its own once a general executive factor is removed. Inhibition behaves more like a family of loosely related skills than one faculty (Friedman & Miyake, 2004; Miyake & Friedman, 2012).

Inhibition only concerns motor actions.

The same braking logic operates over thought and memory. Ignored percepts leave a lingering suppression that slows later responses to them, and deliberately not retrieving a memory makes it harder to recall afterwards, so inhibition reaches well beyond the control of movement (Tipper, 1985; Anderson & Green, 2001).

The Stroop effect shows that inhibition failed.

The extra time on incongruent trials measures the interference and the control mounted against it, not a failure of inhibition; on most trials the correct colour is still named. The effect can be modelled without a dedicated stop at all, as a strong reading pathway biased against by a top-down control signal (MacLeod, 1991; Cohen, Dunbar, & McClelland, 1990).

Glossary

Cognitive inhibition.

The suppression of an irrelevant thought, representation, or memory, as distinct from the inhibition of a motor action; measured by negative priming and suppression-induced forgetting.

Commission error.

A response made to a no-go signal in a go/no-go task; the failure of response restraint, and the task's index of poor inhibition.

Conflict monitoring.

The detection, attributed to the anterior cingulate cortex, of simultaneous activation of incompatible responses, which recruits cognitive control in proportion to the conflict present.

Congruency effect.

The increase in reaction time and errors on incongruent trials relative to congruent ones in a conflict task; the index of interference and of the control brought against it.

Flanker task.

A task in which a central target is surrounded by distractors that prime the same or the opposite response; incongruent flankers slow the response, indexing interference control.

Go/no-go task.

A task requiring fast responses to frequent go signals and withholding of the response to rare no-go signals; the prepotency built by the go rate makes restraint the object of measurement.

Inhibition function.

The curve relating the probability of responding on a stop trial to the stop-signal delay; its position on the time axis fixes the latency of the stop process.

Interference control.

The suppression of an irrelevant stimulus dimension or competing response so that a task-relevant one can govern behaviour; measured by the Stroop and flanker congruency effects.

Negative priming.

The slowing of a response to a stimulus that was ignored on the preceding trial, taken as evidence that ignoring an item actively suppresses it rather than merely failing to select it.

Prepotent response.

A response that is strong enough to occur by default, whether through reflex, overlearning, or task structure; inhibition has work to do only against a prepotent tendency.

Response inhibition.

The cancellation or withholding of a motor action, whether restraining one not yet launched (go/no-go) or cancelling one already under way (stop-signal).

Stop-signal delay.

The interval between the go stimulus and the stop signal in a stop-signal task; short delays favour successful stopping, long delays let the response escape.

Stop-signal reaction time.

The estimated latency of the stop process, recovered from the inhibition function; the standard measure of response-inhibition efficiency, roughly 200 milliseconds in healthy adults.

Stroop task.

The task of naming the ink colour of a colour word; incongruent words slow naming because the automatic reading response must be suppressed, indexing interference control.

Unity and diversity.

The finding that executive functions, including inhibition, are separable but correlated, so that inhibition is neither wholly unitary nor wholly independent of the other functions.

Key Researchers

Adam R. Aron. Professor of Psychology at the University of California, San Diego; his work localised response inhibition to the right inferior frontal cortex and a fast fronto-basal-ganglia pathway, and refined the account across a decade of lesion, imaging, and stimulation evidence. Faculty Page - Google Scholar - Wikidata

Matthew M. Botvinick. Senior Director of Neuroscience Research at Google DeepMind and formerly of Princeton University; he formalised conflict monitoring, the account in which the anterior cingulate detects response conflict and recruits the control that resolves it. ORCID - Google Scholar - Wikidata

Adele Diamond. Canada Research Chair in Developmental Cognitive Neuroscience at the University of British Columbia; she positioned inhibitory control as one of the three core executive functions and charted its development from early childhood. Faculty Page - ORCID - Google Scholar - Wikipedia

Naomi P. Friedman. Professor at the Institute for Behavioral Genetics, University of Colorado Boulder; her latent-variable studies showed that inhibition tasks share little specific variance and that individual differences in executive function are strongly heritable. Faculty Page - ORCID - Google Scholar

Lynn Hasher. Professor Emerita of Psychology at the University of Toronto; she reframed inhibition as the gatekeeper of attention, arguing that suppressing irrelevant information, and the age-related weakening of that function, shapes the effective capacity of thought. Faculty Page - Google Scholar - Wikipedia

Gordon D. Logan. Centennial Professor of Psychology at Vanderbilt University; with William Cowan he formulated the horse-race model of stopping, making the latency of the unobservable stop process measurable as the stop-signal reaction time. Faculty Page - Google Scholar - Wikipedia

Akira Miyake. Professor of Psychology and Neuroscience at the University of Colorado Boulder; his unity-and-diversity framework established inhibition as one of three separable but correlated executive functions, and later work questioned whether it has variance of its own. Faculty Page - ORCID - Google Scholar

Frequently Asked Questions

What is psychological inhibition?
Psychological inhibition is the suppression of a response, thought, or memory that is prepotent or already under way but no longer wanted, so that behaviour can follow a current goal rather than habit or reflex (Diamond, 2013). Because a successful inhibition produces no response, its strength is inferred from a model rather than observed directly (Logan & Cowan, 1984).

What is response inhibition?
Response inhibition is the cancellation or withholding of a motor action. It covers both restraining a response not yet launched, as in the go/no-go task, and cancelling one already in flight, as in the stop-signal task, and the two feats load on partly different processes (Verbruggen & Logan, 2008).

What is the stop-signal reaction time?
The stop-signal reaction time is the estimated latency of the stop process in the race model. It cannot be timed directly, since a successful stop leaves no response, but it is recovered from the probability that a response escapes across stop-signal delays, and it is roughly 200 milliseconds in healthy adults (Logan & Cowan, 1984).

How is inhibition measured?
Response inhibition is measured by the stop-signal reaction time and by commission errors in a go/no-go task; interference control is measured by the congruency effect, the extra time on incongruent trials, in the Stroop and flanker tasks; cognitive inhibition is measured by negative priming and by suppression-induced forgetting (Stroop, 1935; Tipper, 1985).

Is the Stroop effect a measure of inhibition?
The Stroop effect indexes interference control, the suppression of the automatic reading response so the ink colour can be named. It can be modelled without a dedicated stop, as a strong reading pathway biased against by a top-down control signal recruited by conflict (Cohen, Dunbar, & McClelland, 1990; Botvinick et al., 2001).

Is inhibition a single ability?
Probably not. Tasks meant to tap the same inhibitory ability correlate only weakly, and one influential analysis found that a distinct inhibition factor had no reliable variance of its own once a general executive factor was removed (Friedman & Miyake, 2004; Miyake & Friedman, 2012).

Where in the brain does inhibition happen?
Response inhibition has been tied to the right inferior frontal cortex and a fast pathway through the subthalamic nucleus of the basal ganglia (Aron, Robbins, & Poldrack, 2014). A competing view holds that inhibition emerges from goal-driven competition among representations rather than from a dedicated brain module (Munakata et al., 2011).

Can memories be inhibited on purpose?
Yes. Practising not retrieving a learned association makes it harder to recall later than an unpractised item, a suppression-induced forgetting that appears to draw on the same executive machinery used to stop an action (Anderson & Green, 2001).

References