How uncertain are you? Disentangling expected and unexpected uncertainty in pupil-linked brain arousal during reversal learning

The cornerstone of adaptive behavior is to continuously update the model of the environment which predicts the outcomes of our actions. The links between actions and outcomes, however, are often uncertain, which can be attributed to two different sources. On the one hand, unexpected outcomes might be a consequence of the fact that our expectations are of probabilistic nature, and stimuli or actions provide the expected outcome only with a certain probability (expected uncertainty). On the other hand, negative feedback might also signalize that expectations are not valid due to changes in the environment (unexpected uncertainty). Because faster model updating is required for the latter case, disentangling expected and unexpected uncertainty is crucial for adaptive decision making. Not surprisingly, brain mechanisms coding such uncertainty have evolved. Several lines of research show that pupil-linked brain arousal is a sensitive indirect measure of these brain mechanisms, thus we investigated whether it is also involved in disentangling these two sources of uncertainties. To this aim, we measured pupil size during a probabilistic reversal learning task, where participants had to figure out which of two response options led to reward with higher probability, whereby at some points the identity of the more advantageous response option was switched. Expected uncertainty was manipulated by varying the reward probability of the advantageous choice option, whereas the level of unexpected uncertainty was assessed using a Bayesian computational model estimating the beliefs of the participants about the more advantageous location. We found that the level of expected uncertainty influenced behavioral performance but was unrelated to pupil-linked brain arousal. In contrast, irrespective the amount of expected uncertainty, unexpected uncertainty influenced pupil responses, confirming that pupil-linked brain arousal is involved in model updating after unexpected changes in the environment. These results emphasize the role of pupil-linked brain arousal and underlying neural structures in signaling and handling situations in which the previously established contingencies are no longer valid and the mental model of the world must be updated.