Counterfactual Curiosity Activated Striatum Reward Circuits

TL;DR: A 2026 Social Cognitive and Affective Neuroscience fMRI study found people paid a waiting cost to learn counterfactual outcomes, and striatum activity tracked both the choice to know and the size of missed rewards.

Source: Social Cognitive and Affective Neuroscience (2026) | Lau et al.

Regret hurts but people still go looking for it.

After a choice is over, many of us want to know the alternate outcome — the house we missed, the stock we did not buy, the answer we almost gave.

This fMRI study turns that everyday impulse into a neural problem: why would the brain treat information about an outcome that can no longer be chosen as something worth pursuing?

A Balloon Task Made the Alternate Outcome Concrete

The standard Balloon Analogue Risk Task asks people to pump a virtual balloon for points, with a hidden pop limit. This version added a step. After banking points, participants could spend time to reveal the balloon’s hidden limit — learning how many more points they would have earned if they had kept going.

Crucially, that reveal was useless for the next trial. Each balloon had its own random limit.

Knowing the previous one did not help with the next one. The information answered a “what if?” comparison, not a strategic one. That is exactly what separates counterfactual curiosity from ordinary learning.

The waiting cost mattered too. Participants did not get the answer for free; they paid up to six seconds of delay. If people still chose the reveal under that friction, the information had motivational pull strong enough to compete with impatience.

Half the Time, People Paid for an Answer That Could Hurt

Across 38 analyzed participants, counterfactual reveals were chosen in 52% of bank trials. People were more likely to reveal when stakes were higher and waiting costs lower — the signature of motivated wanting rather than idle curiosity.

The behavioral ratings showed a real emotional cost. Participants reported more negative emotion after learning what they had missed.

The same answer that satisfied the urge to know also created regret. That tension is the whole point of the study: a choice made under friction, with foreseeable emotional cost, that people still made about half the time.

Reward Circuitry Activated Before the Answer Arrived

The caudate is part of the striatum — deep brain structures involved in reward, motivation, action selection, and learning. In this study, caudate activity rose when participants chose to seek counterfactual information, before they learned whether the answer would feel good or bad.

That timing is the central finding. The brain treated the act of pursuing information as rewarding in itself, not because the answer would be satisfying. Curiosity can have incentive value — the “wanting” component of motivation — before any content arrives. The result places counterfactual curiosity inside a reward-and-motivation system instead of treating it as a purely intellectual habit.

Missed Gains Also Pulled in the Nucleus Accumbens

After the reveal, the second result emerged. When participants discovered that they could have won more points, the caudate and nucleus accumbens responded more strongly to bigger missed gains. The nucleus accumbens is another striatal region tightly tied to reward and motivation.

That is biologically uncomfortable. A painful answer can still feel compelling because missed opportunity is being processed through reward circuitry, not only through regret or disappointment. The brain treats both the search and the answer as motivationally salient — even when the search is voluntary and the answer hurts.

Counterfactual Curiosity Differed From Adaptive Learning

Some curiosity is genuinely useful. A student checks an explanation; a clinician reviews a missed diagnosis; a trader studies a bad call to improve the next one. Those reveals carry information that improves future decisions.

The balloon task stripped that out by design. The hidden limit varied randomly trial to trial, so the reveal did not predict the next balloon. The motivational pull of “what could have happened” survived even when the answer had no strategic value. That makes the study more than a paper about regret — it is a paper about why the brain may treat some forms of information as rewards, regardless of whether they help.

Striatum Activity May Explain Real-Life Counterfactual Checking

The sample was young and university-based, and the stakes were game points. A lab balloon task is not the same as checking an ex-partner’s social media, replaying a financial miss, or rumination over a medical decision. The mechanism travels surprisingly well, though.

Counterfactual checks often promise closure while risking pain. The striatum result suggests that the drive to resolve the alternate outcome can be rewarding before the answer is known — which is why these checks can become compulsive even when they predictably worsen the person’s mood.

The result does not make curiosity bad. It shows curiosity has more than one flavor.

Some information helps us learn; some mainly rewards the urge to know; some does both while leaving regret behind. Once the alternate outcome is available, the decision is no longer only about usefulness — it is also about whether the brain wants uncertainty resolved.

The open issue for follow-up work is regulation. Can reminders about emotional cost, mindfulness strategies, or repeated exposure to useless counterfactuals change the choice to reveal? That would move this work closer to habits people recognize outside the scanner: rechecking a result that cannot be changed, then paying for the answer with frustration, rumination, or regret.