Behavioral Science Dictionary

Anterior cingulate cortex

Cognition & Dual-Process

The brain's monitor for conflict, error, and the need for control.

What it means

The anterior cingulate cortex is a medial frontal brain region implicated in detecting conflict between competing responses, monitoring errors, registering pain and negative feedback, and signaling when more cognitive control is needed. By flagging conflict and mistakes, it is thought to recruit the lateral prefrontal cortex to adjust behavior, making it a linchpin of the brain's performance-monitoring and control loop. Its activity generates the error-related negativity, an EEG signature appearing within a tenth of a second of a mistake, and it contributes to the cost–benefit weighting of whether exerting effort is worthwhile — relevant to debates over self-control and ego depletion. ACC function relates to individual differences in self-regulation, anxiety, and sensitivity to errors. It matters because it is where the brain notices that something is going wrong and decides to engage deliberate control, a precondition for catching and correcting biased or impulsive responses.

How the monitoring loop works

Conflict here has a precise meaning: two incompatible responses are active at once, and the degree of co-activation can be computed as a single number. Botvinick and colleagues modelled that quantity directly, letting a simple network estimate conflict from the overlap between competing response units. The proposal is that the region reads out this number and passes it forward, while lateral prefrontal cortex does the actual work of raising attention to the task-relevant input. The loop is reactive rather than clairvoyant: control tightens after conflict is felt, not before. That predicts two well-documented effects. People pay a smaller cost on a difficult trial that follows another difficult trial, and they slow down on the trial after a mistake. The region does not decide what to do; it registers that something needs deciding.

What the evidence shows

The founding imaging result cuts against the region's own reputation. Carter and colleagues (1998) found activity on errors, as expected, but also on correct responses when response competition was high, suggesting it registers the conditions under which errors become likely rather than errors themselves. Electrophysiology adds a generator and a rival reading of it: source models place the error-related negativity in medial frontal cortex, and Holroyd and Coles (2002) reinterpreted that signal as a reinforcement-learning message, a dopaminergic report that things went worse than predicted, rather than a conflict readout. Shenhav, Botvinick and Cohen (2013) later folded conflict, error, effort and reward into one account: the region estimates the expected value of control and allocates it accordingly. Rival models fit much of the same data, which is the honest state of play.

Limits and caveats

Activation is not necessity. Fellows and Farah tested four patients with dorsal ACC damage and found normal Stroop and go/no-go performance, intact post-error slowing, and normal adjustment to speed-versus-accuracy instructions, a hard result for any account that makes the region indispensable. The imaging signal carries a confound too: incongruent trials take longer than congruent ones, and Grinband and colleagues showed that dorsal medial frontal activity tracks time on task, so part of the apparent conflict effect may be trials simply lasting longer. Scale is a third problem, because the anterior cingulate is a long strip with distinct subdivisions that get pooled under one name. Reverse inference is weaker still: Lieberman and Eisenberger argued from meta-analytic data that pain best describes dorsal ACC function, and ten researchers replied that it responds robustly in non-painful tasks.

Reading ACC claims in practice

For practitioners the region is a mechanism to reason with, not evidence to cite. An imaging finding cannot make a behavioural effect real; if the behaviour does not replicate, a picture of a lit-up cingulate will not rescue it. Ego depletion is the case in point: large-scale registered replications failed to find the effect, so ACC-based stories about the cost of effort should not be read as ratifying it. What the research does license is a design principle. Control is recruited only when a mismatch is registered, so an interface in which a wrong action feels exactly like a right one gives the monitor nothing to catch. The errors that survive are the ones committed inside habitual, fluent, low-conflict flows. Making a consequential step feel different from the routine it resembles, through a changed layout or an enforced pause, manufactures the conflict the monitor needs in order to fire.

Examples

On the Stroop task, naming the ink color of the word 'RED' printed in blue triggers anterior cingulate activity that signals the conflict and summons extra control.

You mistype your password and know it half a keystroke before the screen does; that jolt is the error signal the region produces within a tenth of a second.

Reaching for the indicator in a hire car and hitting the wipers: the clash between the practised habit and the new layout is what the region flags so you correct it.

Weighing whether to proofread a contract line by line or skim it and accept the risk is the effort cost-benefit calculation the expected-value-of-control account attributes to the region.

A radiologist calls a shadow on a chest film benign and is right, but the read was contested: two plausible interpretations were live at once. On Carter and colleagues' error-likelihood finding the monitoring signal rises on exactly that trial, on a correct call made under high competition, not only when the call turns out wrong.

First described in Performance-monitoring research; Botvinick, Carter, Cohen (2001).

Key references

  1. Shenhav, A., Botvinick, M. M., & Cohen, J. D. (2013). The expected value of control: An integrative theory of anterior cingulate cortex function. Neuron, 79(2), 217–240. doi.org/10.1016/j.neuron.2013.07.007
  2. Shenhav, A., Cohen, J. D., & Botvinick, M. M. (2016). Dorsal anterior cingulate cortex and the value of control. Nature Neuroscience, 19(10), 1286–1291. doi.org/10.1038/nn.4384
  3. Carter, C. S., Braver, T. S., Barch, D. M., Botvinick, M. M., Noll, D., & Cohen, J. D. (1998). Anterior cingulate cortex, error detection, and the online monitoring of performance. Science, 280(5364), 747–749. doi.org/10.1126/science.280.5364.747
  4. Holroyd, C. B., & Coles, M. G. H. (2002). The neural basis of human error processing: Reinforcement learning, dopamine, and the error-related negativity. Psychological Review, 109(4), 679–709. doi.org/10.1037/0033-295X.109.4.679
  5. Wager, T. D., Atlas, L. Y., Botvinick, M. M., Chang, L. J., Coghill, R. C., Davis, K. D., Iannetti, G. D., Poldrack, R. A., Shackman, A. J., & Yarkoni, T. (2016). Pain in the ACC? Proceedings of the National Academy of Sciences, 113(18), E2474–E2475. doi.org/10.1073/pnas.1600282113
  6. Lieberman, M. D., & Eisenberger, N. I. (2015). The dorsal anterior cingulate cortex is selective for pain: Results from large-scale reverse inference. Proceedings of the National Academy of Sciences, 112(49), 15250–15255. doi.org/10.1073/pnas.1515083112
  7. Grinband, J., Savitskaya, J., Wager, T. D., Teichert, T., Ferrera, V. P., & Hirsch, J. (2011). The dorsal medial frontal cortex is sensitive to time on task, not response conflict or error likelihood. NeuroImage, 57(2), 303–311. doi.org/10.1016/j.neuroimage.2010.12.027
  8. Fellows, L. K., & Farah, M. J. (2005). Is anterior cingulate cortex necessary for cognitive control? Brain, 128(4), 788–796. doi.org/10.1093/brain/awh405
  9. Botvinick, M. M., Braver, T. S., Barch, D. M., Carter, C. S., & Cohen, J. D. (2001). Conflict monitoring and cognitive control. Psychological Review, 108(3), 624–652. doi.org/10.1037/0033-295X.108.3.624

Related terms

← All 1001 terms