About

Moira Dillon is an Assistant Professor of Psychology at New York University, where she has been serving since 2017. Her research focuses on infants’ and children’s precocious knowledge about the world, including objects, places, people, and animals, and how this knowledge derives from our evolutionary inheritance. She investigates how this foundational knowledge forms the basis of human cultural and intellectual achievements. Dillon employs cognitive, developmental, and computational approaches in her research, aiming to gain insight into the origin and development of human cognition. Her work involves active collaborations with economists, mathematicians, neuroscientists, educators, and humanists, and she maintains research partnerships with organizations such as The National Museum of Mathematics (MoMath), Lookit (an online infant and child lab), and the Abdul Latif Jameel Poverty Action Lab (J-PAL). Dillon’s expertise in human common sense is also used to improve machine common sense, with the goal of creating AI systems that better understand humans and that humans can better understand. Her academic background includes a Ph.D. from Harvard University obtained in 2017 and a B.A. from Yale University earned in 2008.

Research topics

• Computer Science
• Artificial Intelligence
• Psychology
• Cognitive psychology
• Political Science
• Cognitive science
• Developmental psychology
• Neuroscience
• Epistemology
• Database
• Medical education
• Business
• Knowledge management
• Public relations
• Medicine
• Data science

Selected publications

• Core misperceptions

  2026-04-07

  articleOpen access1st authorCorresponding

  The article: (1) misstates the core-knowledge hypothesis and consequently argues against a straw person; (2) misses replies to prior core-perception accounts, which become undefeated prolepses; and (3) focuses on properties that are common to, not distinctive of, perception and core knowledge. Core knowledge’s “territory in between” still provides guidelines for how we conceive of the mind, its states and processes.

  PublisherDOI

• Scaling cognitive science with classroom games for learning mathematics

  2026-05-22

  articleOpen access

  We report two randomized evaluations of educational interventions targeting children in low-income neighborhoods of Delhi, India. Four math games, played in social groups, associated early emerging and universally intuitive concepts of number and geometry with the mathematical language, symbols, and operations taught in Delhi government schools. The first experiment, conducted in 231 NGO-run preschool classrooms (1,986 children) showed that the games enhanced children’s mastery of mathematics both immediately after the intervention and a year later, after children’s first year of primary school. The second experiment, conducted in 141 kindergarten and first-grade classrooms in government schools (2,828 children), with no extra personnel, showed that the games remained effective when played cooperatively by larger groups of children, led by their regular teachers, during time allotted for math instruction. These experiments show how insights from basic research on children’s mathematical intuitions can be leveraged to design ready-to-scale interventions for the foundational years.

  PublisherDOI

• Descriptions of phenomenal versus intentional mental states are differentially associated with gaze direction

  2026-03-20

  articleOpen accessSenior author

  Across cultures, humans reason about others’ mental states as being either phenomenal, reflecting shareable experiences, or intentional, reflecting agentive goals. Might our dissociated judgments of others’ mental states be linked to dissociated judgments of others’ gaze direction, one of our earliest social cues from infancy? We presented adults (N = 600) and 6- to 8-year-old children (N = 200) with direct, definitional descriptions as well as descriptions of example phenomenal or intentional mental states and asked them to choose whether a face with averted or straight-ahead gaze better matched the character in each description. Adults and older children chose the averted-gaze face for the direct phenomenality description but the straight-ahead-gaze face for the direct intentionality description. Adults showed the same pattern of responses for the indirect, example descriptions, but children chose the straight-ahead-gaze face for both types of indirect descriptions. Descriptions of either phenomenal or intentional mental states might be intuitively associated with straight-ahead gaze, perhaps originating in infants’ understanding of straight-ahead gaze as communicating either mutual social engagement or general visible accessibility for goal-directed action. The association between phenomenality descriptions and averted gaze may, by contrast, arise only after learning that certain phenomenal contexts involve inner experiences with averted gaze (e.g., someone is looking at a particular emotion-evoking stimulus or looking “into their own mind”). If, in adulthood, our social concepts have indeed been built on prelinguistic intuitions that can be evoked by language, then these intuitions may continue to affect our mature, linguistically driven social reasoning.

  PublisherDOI

• Geometry

  MIT Press eBooks · 2026-03-16

  book-chapterOpen access1st authorCorresponding
  PublisherOA PDFDOI

• How does noun-label extension reflect another’s actions and goals?

  OSF Preprints (OSF Preprints) · 2026-01-04

  otherOpen access1st authorCorresponding
  Publisher

• Is there a “shape bias” for places?

  2025-06-19

  preprintOpen accessSenior author

  Early emerging biases support our learning of new words, like a tendency to extend novel nouns to objects of the same shape. The present work reveals that, in its focus on the domain of objects, the existing literature on this shape bias has been critically limited in its generalizability. Adults (N=72) and 6-year-old children (N=72), but not 3-year-old children (N=72), extended novel nouns to places of the same shape over places composed of the same-colored walls, regardless of whether places were described by novel nouns in labeling phrases or place-relevant prepositional phrases. All age groups, by contrast, showed a shape bias for objects that matched many features of the place stimuli. This relatively delayed shape bias for places suggests that at least some word-learning biases may emerge separately for different domains, thereby challenging existing theories of language acquisition to account for such domain specificity in word learning.

  PublisherDOI

• Descriptions of phenomenal versus intentional mental states are differentially associated with gaze direction

  2025-09-09

  preprintOpen accessSenior author

  Across cultures, humans reason about others’ mental states as being either phenomenal, reflecting shareable experiences, or intentional, reflecting agentive goals. Might our dissociated judgments of others’ mental states be linked to dissociated judgments of others’ gaze direction, one of our earliest social cues from infancy? We presented adults (N = 600) and 6- to 8-year-old children (N = 200) with direct, definitional descriptions as well as descriptions of example phenomenal or intentional mental states and asked them to choose whether a face with averted or straight-ahead gaze better matched the character in each description. Adults and older children chose the averted-gaze face for the direct phenomenality description but the straight-ahead-gaze face for the direct intentionality description. Adults showed the same pattern of responses for the indirect descriptions, but children chose the straight-ahead-gaze face for both types of indirect descriptions. A link between straight-ahead gaze and phenomenality present from human infancy may thus impact children’s judgments of others’ gaze direction. But by adulthood, language may come to evoke particular contexts in which a character’s inner experience may be more strongly linked to averted gaze. If, in adulthood, our social concepts have indeed been built on such prelinguistic intuitions and language, then these intuitions may continue to affect our mature, linguistically driven social reasoning.

  PublisherOA PDFDOI

• Scene and Heard: Spatial layout and language support young infants’ categorization of places

  2025-03-15

  preprintOpen accessSenior author

  Human infants are limited in their self-guided navigation. Can they nevertheless learn about the places they’ll go? Across three experiments, we tested the role of spatial layout and language on pre-crawling 6-month-old infants’ (N = 96) ability to categorize places in a novelty-preference looking-time paradigm. We found that, when places are labeled, young infants are sensitive to their distinct spatial layouts and can learn place categories. When places’ spatial layouts are disrupted or when places are not labeled, by contrast, young infants cannot learn such place categories. Our results shed new light on young infants’ sensitivities to foundational domains of everyday life prior to their capacity for self-guided exploration.

  PublisherOA PDFDOI

• Taking the C-nic Route: Object-Directedness and Path, Not Efficiency, Shape Adults' Word Extension

  2025-05-12

  preprintOpen accessSenior author

  What intuitions guide adults in extending the meanings of new words? Are these intuitions consistent with prelinguistic sensitivities operative in infancy? Across two preregistered experiments, adult participants saw simple grid-world environments in which characters moved in a “C” path efficiently (or not) to an object (or not). These events were labeled with a novel verb or noun. Participants were asked whether that word applied to new events varying in object-directedness, path, and efficiency. By contrast to infants’ focus on efficiency, adults instead focused on object-directedness and path, and they did so similarly for both verbs and nouns. Language might thus build on universal, prelinguistic assumptions of goal-directedness and efficiency to specify what goal an agent might have (e.g., object- vs. location-directed) as well as how an agent might achieve that goal (e.g., this vs. that kind of path), ultimately restricting the hypothesis space for action understanding and supporting learning.

  PublisherOA PDFDOI

• Scene and Heard: Spatial layout and language support young infants’ categorization of places

  2025-06-18

  preprintOpen accessSenior author

  Human infants are limited in their self-guided navigation. Can they nevertheless learn about the places they’ll go? Across three experiments, we tested the role of spatial layout and language on pre-crawling 6-month-old infants’ (N = 96) ability to categorize places in a novelty-preference looking-time paradigm. We found that, when places are labeled, young infants are sensitive to their distinct spatial layouts and can learn place categories. When places’ spatial layouts are disrupted or when places are not labeled, by contrast, young infants cannot learn such place categories. Our results shed new light on young infants’ sensitivities to foundational domains of everyday life prior to their capacity for self-guided exploration.

  PublisherDOI

Recent grants

• CAREER: Becoming Euclid: Characterizing the geometric intuitions that support formal learning in mathematics

  NSF · $1.8M · 2019–2024

Frequent coauthors

• Elizabeth S. Spelke

  Harvard University

  42 shared

• Esther Duflo

  Massachusetts Institute of Technology

  24 shared

• Keesler Welch

  17 shared

• Joshua Dean

  University of Bristol

  17 shared

• Véronique Izard

  Université Paris Cité

  17 shared

• Harini Kannan

  16 shared

• Yuval Hart

  Hebrew University of Jerusalem

  10 shared

• Brenden M. Lake

  8 shared