Cognition

"Wink or blush? Pupil-linked phasic arousal signals both change and uncertainty during assessment of changing environmental regularities"

Check it out:
doi.org/10.1016/j.co...

A reversal learning paradigm. Four manipulations. Cluster based permutation test. Two pupil responses. Five linear mixed models. Blinks. Response times... and a double dissociation.

Environmental change... Belief uncertainty... Arousal... Pupil dilation... Two separable effects...

People can infer the magnitude of other people’s knowledge even when they cannot infer its contents Inferences about other people’s knowledge and beliefs are central to social interaction. However, people’s behavior is often consistent with a range o…

Full paper and materials: www.sciencedirect.com/science/arti...

Authors: Rosie Aboody, Isaac Davis, Yarrow Dunham (yarrowdunham.bsky.social‬), Julian Jara-Ettinger (julianje.bsky.social���)

These results suggest that epistemic inference goes beyond specific belief states: we can form precise representations of how much someone knows, even when we cannot determine the specific contents of their knowledge.

Their judgments closely matched predictions from a computational model based on utility maximization and Bayesian reasoning, suggesting that people infer knowledge magnitude by reasoning about the costs and efficiency of others’ actions under different quantities of knowledge.

Across two studies, we found that people can make precise, quantitative judgments about how much other people know from watching them make simple choices (like where to search for a hidden prize), even when they have no way to determine the contents of that knowledge.

Can we infer how much someone knows, even when we can't tell what they know?

A counterfactual explanation for recency effects in double prevention scenarios: Commentary on Thanawala and Erb (2024) Many cognitive scientists and philosophers take cases of double prevention to be one of the primary motivations for accepting causal pluralism, the vi…

"A counterfactual explanation for recency effects in double prevention scenarios: Commentary on Thanawala and Erb (2024)"

📢 New paper from: Tadeg Quillien, Kevin O’Neill, & Paul Henne

www.sciencedirect.com/science/arti...

These results suggest that time affects causal reasoning by influencing the events people mentally modify when imagining how things could have happened.

Parameters from the model fit confirm that recent events are less 'stable': people are more likely to modify them when imagining counterfactuals:

Our computational model (orange) is able to closely fit the original data (black) from Thanawala & Erb (2024):

Our explanation is that people make causal judgments by imagining how things could have happened differently. Since people prefer to imagine alternatives to recent events, we expect that the order of events in a scenario will affect causal judgments.

In a double prevention scenario, a double preventer prevents a preventer from preventing an outcome. What do people think caused the outcome? A recent study finds it depends on the order in which events happen. We give an explanation for this result:

Does culture shape how we look, gesture, and interact? A new study uses wearable eye-tracking to explore how Japanese and Dutch participants coordinate gaze and action during collaboration.

www.sciencedirect.com/science/arti...

"Action chunking as conditional policy compression"

📢 New paper from: Lucy Lai, Ann Huang, & Samuel Gershman

sciencedirect.com/science/arti...

9/ Blog post here: www.aesthetics.mpg.de/en/newsroom/...

Read the article: www.sciencedirect.com/science/arti...

8/ New paper: "Aesthetic appeal of dance actions depends on expressivity, liveness and audience characteristics"

Authors: Julia Christensen, Eva-Madeleine Schmidt, Klaus Frieler, Rebecca Smith-Chase, Luisa Sancho-Escanero, Georgios Michalareas, Fredrik Ullén, & @escross.bsky.social

7/ The findings suggest that individual differences play a decisive role in the audience’s aesthetic experience—alongside performance and setting.
Dance education and exposure may help cultivate future audiences.

6/ The format of the performance mattered too:
- Live > Pre-recorded video
- Human > Avatar

But: No difference between in-person vs livestream
Online audiences enjoyed the live-streamed dance just as much as in-person viewers.

5/ Subjective factors also played a role.
👩 Women liked the performances more than men
📈Interest increased with age
🧑‍🎓 Higher education = lower interest
🔁People high in openness reported more interest in attending again

4/ The dancer’s movements were measured via speed and acceleration, making it possible—for the first time—to distinguish expressive from non-expressive movements in a genuine live performance setting.

3/🎯 Key finding: Audience enjoyment depended largely on how expressively the dancer moved.
Expressive sequences were significantly more enjoyable than less expressive ones.

2/ Each audience saw:
-2 live performances by a dancer in a motion capture suit
-2 avatar performances on a screen
-2 pre-recorded video performances

About 40 people attended in person, and 40 more watched via livestream. After each, they filled out a questionnaire.

1/ Unlike ballet, contemporary dance is not necessarily intuitively “understood” by the general public. So which factors influence whether audiences enjoy a performance?
An interdisciplinary team explored this question using six versions of the same choreography.

🩰What makes contemporary dance enjoyable?

A new study by researchers at the Max Planck Institute for Empirical Aesthetics @ae.mpg.de‬ investigates what people like about contemporary dance—using live performances, motion capture, and audience questionnaires.
👇

We argue that these findings challenge dual-system models and instead support the idea that categorization might emerge from a single flexible system, or from tight interaction between systems, capable of adapting even in uncertain environments.

Importantly, this holds for both simple and complex categorization rules.

We find that participants are not only able to learn under probabilistic conditions—but that they also apply this knowledge in a new, explicit task. The more reliable the feedback, the more robust the transfer.

Across two experiments, we use a Probabilistic Categorization Task (PCT) to explore how participants form and transfer category knowledge under varying levels of uncertainty (70%, 80%, and 90% reliable feedback).