{
  "id": 12071428,
  "title": "Is confidence a second thought—or part of the first one?",
  "url": "https://urgent.news/2026/10/05/is-confidence-a-second-thought-or-part-of-the-first-one",
  "topic": "science",
  "section": "Science",
  "published": "2026-10-05T04:00:33.000Z",
  "source": {
    "name": "The Transmitter",
    "slug": "the-transmitter",
    "url": "https://www.thetransmitter.org/decision-making/is-confidence-a-second-thought-or-part-of-the-first-one/"
  },
  "original_language": "en",
  "account": "In certain scenarios, neural circuits may compute confidence alongside the act of decision-making. This revelation alters the inquiries researchers ought to pose. Every choice inherently prompts an implicit inquiry: How probable is this choice to be accurate? Confidence—the response to that inquiry—indicates to a decision-maker the degree of reliance to place on their own selection. Confidence affects how we articulate our convictions, assimilate feedback, and determine whether to proceed or gather supplementary information. For an extended duration, confidence held a role akin to Cinderella in cognitive psychology—valued for its utility yet frequently disregarded as a distinct variable. However, this perspective is evolving. Researchers are now probing the whereabouts of confidence within the brain, its computational mechanisms, its manifestation in psychiatric conditions, and even the potential to imbue artificial agents with rudimentary metacognitive competencies. A central question pervades this pursuit: What precisely is the brain ascertaining when it conveys confidence?\n\nThe prevailing theory posits that choice and confidence follow a hierarchical organization: The brain initially formulates a decision, regardless of its provisional nature, and subsequently estimates the likelihood of that decision being accurate. This perspective perceives confidence as an outcome derived from the decision-making process. The neural underpinnings of decision-making, however, propose an alternative, albeit less intuitive, proposition. When each option is pre-associated with a specific action, the decision transpires within the neural circuits primed for those actions. Envision a monkey contemplating the net motion of a cloud of moving dots—either leftward or rightward. Should the monkey articulate its choice via an eye movement, neurons engaged in eye movement planning, including a subset within the lateral intraparietal area, represent the evolving decision variable. Their activity does not predominantly encode an abstract judgment such as “the motion is rightward.” Instead, it mirrors the intensifying intention to gaze at the rightward target. This framework, termed “intentional,” posits that deciding is inextricable from preparing to act. It implies a distinct approach to comprehending confidence. If a decision can evolve as a contest among actions rather than abstract propositions, confidence can be integrated into this competition. Envision a decision-maker reporting both perceptual choice and confidence through a single eye movement, selecting among four targets: left-high confidence, left-low confidence, right-low confidence, and right-high confidence. According to the hierarchical account, this is construed as a sequence of computations: Determine the direction, evaluate confidence, and subsequently select the appropriate target. Conversely, the intentional framework offers an alternative interpretation: The four actions may serve as the alternatives over which the decision unfolds. Diverse states of sensory evidence favor distinct joint choice-confidence responses: Robust rightward evidence, for instance, favors the right-high-confidence action, while attenuated rightward evidence may favor the right-low-confidence action (as per the RTCON model). In this perspective, sensory evidence operates through four concurrent accumulators within eye-movement-planning circuits, each corresponding to a potential response. The earliest accumulator to attain a threshold determines both the choice and the confidence. The hierarchical and intentional accounts yield divergent predictions regarding the role of action-preparation circuits. Within a hierarchical framework, choice and confidence can be delineated upstream and then translated into the action utilized to report them. Even if confidence is continuously derived from an evolving decision, neural activity in action-preparation circuits might mirror the evolving decision rather than actively participate in the decision process. In contrast, the intentional account posits that the decision materializes within these circuits: Choice and confidence-related signals should emerge concurrently as the brain selects among the joint choice-confidence actions. While no neurophysiological study directly juxtaposes hierarchical and intentional accounts, a 2025 study offers some insights. It conducted a version of the four-target task described above: Macaques reported both motion direction and confidence by selecting one of four eye-movement targets while researchers recorded from lateral intraparietal neurons corresponding to individual targets. Activity distinguishing high- from low-confidence targets commenced to diverge approximately concurrently with activity differentiating leftward from rightward choices, roughly 200 milliseconds after motion onset. Confidence-related activity, accordingly, did not appear to trail behind choice-related activity. These findings challenge the strict hierarchical account, which posits that confidence evaluation commences only after commitment to a choice, and align more smoothly with an intentional account, wherein both dimensions of the report emerge within an eye-movement-planning circuit.\n\nHowever, this does not entail that confidence must invariably be embedded within the decision itself. In numerous tasks, confidence diverges from choice in manners that seem to necessitate a distinct evaluative process. My recent research suggests that confidence may be instantiated intentionally or hierarchically contingent on the context. The intentional account appears to elucidate behavior more effectively when confidence reports are explicitly incentivized—for instance, through a pecuniary reward.",
  "summary": "In some settings, neural circuits may calculate confidence in a decision simultaneously with the choice itself. That changes the types of questions researchers should ask.",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}