When people act on objects, their goals can depend on the objects’ intrinsic properties and conventional uses (e.g., using forks, not knives, to eat spaghetti), locations (e.g., clearing the table, regardless of what is on it), or both (eating with the fork next to your plate, not your dining partner’s). For adults, objects’ intrinsic properties matter more than their locations in most action contexts. Whereas 5-month-old infants privilege objects’ intrinsic properties in attributing goals to people reaching for objects, 3-month-old infants do not. Do younger infants fail to view reaching as goal-directed, or are they uncertain which properties of objects are relevant in different contexts? Here we show that 3-month-old infants attribute goals to others’ reaching actions when given information that their actions depend on what, not where, an object is. Our findings suggest that 3-month-old infants can learn about others’ object goals, before they reach for objects themselves.

Our everyday decisions are driven by costs, risk, and reward.How do people take these factors into account when they pre-dict and explain the decisions of others? In a two-part exper-iment, we assessed people’s perceptions of other people’s riskpreferences, relative to their own. In Part 1, participants re-ported their relative preference between a guaranteed payoutand lotteries with various probabilities and payouts, and madepredictions about other people’s preferences. In Part 2, partic-ipants estimated the lottery payout that generated a given rela-tive preference between a guaranteed payout and a lottery, bothfor themselves and others. We found considerable individualvariability in how people perceive the risk preferences of oth-ers relative to their own, and consistency in people’s percep-tions across our two measures. Future directions include for-mal computational models and developmental studies of howwe think about our own and each other’s decision-making.

Not every prosocial act is equally praiseworthy. As adults, we tend to evaluate helpers depending on others’ needs; helping someone with a hard task may be more burdensome but also more prosocial than helping someone with an easy task. Despite growing evidence that infants are sensitive to the costs of others’ actions, whether such representations inform their evaluation of prosocial acts remains an open question. Here, we ask whether 16-month-olds’ preference between two helpers is sensitive to task difficulty. Infants preferentially reached for (Exp. 1), and looked at (Exp. 2), an agent who helped someone facing a high-cost task over an agent who helped someone facing a low-cost task. Such preference disappeared when the agents completed the same tasks in a self-serving context. These results suggest that infants use action cost not only to predict and explain others’ behaviors but also to evaluate others in prosocial contexts.

In studies of infant cognition, why do babies look longer when objects pass through each other, or someone behaves inefficiently? We test 3 candidate explanations (domain-specific prediction error, domain-general endogenous curiosity, and perceptual novelty), each with a distinct, non-mutually exclusive, predicted pattern of neural activity. We scanned 17 adults using fMRI while they watched videos of agents and objects, adapted from infant behavioral research [1–4]. Cortical regions preferring social vs physical information [5,6] showed similar preferences for these stimuli. These regions, in the first run of the experiment, responded to physical and social violations (unexpected > expected outcomes), with a greater response to violations from the corresponding domain. Regions that respond to general perceptual novelty also responded to social and physical violations, regardless of domain. Thus both domain-specific, and general perceptual, cortical regions encode violations of expectation involving agents and objects. References 1. Baillargeon R, Spelke ES, Wasserman S. Object permanence in five-month-old infants. Cognition. 1985;20: 191–208. 2. Needham A, Baillargeon R. Intuitions about support in 4.5-month-old infants. Cognition. 1993;47: 121–148. 3. Gergely G, Nádasdy Z, Csibra G, Bíró S. Taking the intentional stance at 12 months of age. Cognition. 1995;56: 165–193. 4. Woodward AL. Infants selectively encode the goal object of an actor’s reach. Cognition. 1998;69: 1–34. 5. Koster-Hale J, Saxe R. Functional neuroimaging of theory of mind. Understanding Other Minds. 2013. pp. 132–163. doi:10.1093/acprof:oso/9780199692972.003.0009 6. Fischer J, Mikhael JG, Tenenbaum JB, Kanwisher N. Functional neuroanatomy of intuitive physical inference. Proc Natl Acad Sci U S A. 2016;113: E5072–81.

When adults see one solid object pass through another, or see a person take the long route to a destination when a shortcut was available, we classify those events as surprising. Infants look infants look longer at the same unexpected outcomes, compared with visually similar but expected outcomes, in violation-of-expectation (VOE) experiments. What domain-specific and domain-general cognitive processes support these judgments? In a pre-registered experiment, we scanned 32 adults using functional magnetic resonance imaging (fMRI) while they watched videos designed for infant research. One region implicated in physical reasoning responded selectively to unexpected physical events, providing evidence for domain-specific physical prediction error. Multiple demand regions responded more to unexpected events regardless of domain, providing evidence for domain-general goal-directed attention. Early visual regions responded equally to unexpected and expected events, providing evidence against stimulus-driven prediction error. Thus, in adults, VOE involves domain-specific, and high-level, domain-general computations.

When predicting or explaining another person’s actions, weoften appeal to the physical effort they require; a person whoworks hard for something, for instance, must really like it (Liu,Ullman, Tenenbaum, & Spelke, 2017). But people are notonly motivated to avoid physical effort; they also seek to avoidmental effort (Shenhav et al., 2017; Kool & Botvinick, 2018).Here, we ask whether mental effort enters into preschoolers’understanding of other people’s actions. Across 4 experiments(N=112), we presented 4- and 5-year-old children with anagent (naive in Exp 1, 2 and 4, and knowledgeable in Exp 3)who can either move through a simple or complex maze envi-ronment with a specific goal (in Exp 1-3, to reach a play struc-ture beyond the mazes, and in Exp 4, to practice solving themazes). We found that children were sensitive to the physicaland mental effort associated with more complex mazes, and tothe trade-offs between effort and gain in skill. The intuitionthat choices impose costs on our bodies and minds appears toguide children’s understanding of other people.

When we reason about the goals of others, how do we balance the positive outcomes that actions led to, with the potentiallybad ways those actions could have ended? In a four-part experiment, we tested whether and how adults (full study) and6- to 8-year-old children (ongoing study) expect other agents to take account of the ways their goal-directed action couldhave failed. Across 4 different tasks, we found that adults expected others to negatively appraise perilous situations (deeptrenches), to minimize the danger of their actions, and to trade off danger and reward in their action plans. Our preliminarychildrens study shows similar trends. These results suggest that people appeal to peril-how badly things could go if onesactions fail-when explaining and predicting other peoples actions, and also make quantitative inferences that are finelytuned to the degree of peril and reward that others face.

Infants understand that people act in order to achieve their goals, but how can they tell what goals people findworthwhile? Here, we explore the thesis that human infants solve this problem by building a mental model of action planning,taking into account the costs of acting and the rewards actions bring. Consistent with this thesis, we found that 10-month-oldinfants, after viewing an agent approach two objects equally often, inferred that the agent preferred the object whose attainmentrequired a costlier action. Infants’ responses generalized across changes in perceptual variables that distinguished one actionfrom another (e.g. path length, angle of incline), suggesting that an abstract cost metric based on force or effort supported theirjudgments. These findings suggest that infants’ knowledge about agents may be expressed as a generative model for actionplanning, which can then be inverted to identify the probable hidden causes for observed actions.