United States

This study explores the impact of multilingualism on early number concepts in preschool children. We compared the performance of mono- and multilingual preschoolers on four tasks that test early number concepts. The tasks were counting, Give-N, magnitude judgements, and number identification. The sample included 59 Australian children aged 3.5 to 5.5 years, with 25 multilingual children exposed to at least one language in addition to English. Our results indicated that age was a significant predictor across all tasks, and mono- versus multilingual experience alone did not have a substantial effect on these tasks. However, there was a significant age-by-language experience interaction in the counting task, where older multilingual children counted significantly higher than the older monolinguals. These findings lend new insights into the nuanced role that multilingualism plays in the development of number concepts.

CogSci 2025

How does multilingualism interact with early number concepts?

cognitive development

concepts and categories

psychology

poster

### Welcome to CogSci Conference 2025!

The 47th Annual Meeting of the Cognitive Science Society was a hybrid meeting held in San Francisco. 

<div style="position:relative;padding-top:0;width:900px;height:500px;"><iframe style="position:absolute;border:none;width:100%;height:100%;left:0;top:0;" src="https://online.fliphtml5.com/ebtyf/amvr/"  seamless="seamless" scrolling="no" frameborder="0" allowtransparency="true" allowfullscreen="true" ></iframe></div>

#### About

The Cognitive Science Society brings together researchers from around the world who hold a common goal: understanding the nature of the human mind. The mission of the Society is to promote Cognitive Science as a discipline, and to foster scientific interchange among researchers in various areas of study, including Artificial Intelligence, Linguistics, Anthropology, Psychology, Neuroscience, Philosophy, and Education.

The Society is a non-profit professional organization and its activities include sponsoring an annual conference and publishing the journals Cognitive Science and TopiCS.

#### Our History 

* **Society Creation**<br>
The Society was incorporated as a 501(c)(3) non-profit professional organization in Massachusetts in 1979. The organizing committee included Roger Schank, Allan Collins, Donald Norman, and a number of other scholars from psychology, linguistics, computer science, and philosophy. 
<br><br>
* **Conference Creation**<br>
The first conference on cognitive science was held at La Jolla, California in August, 1979, and has occurred annually since then. The proceedings of each conference are published, and those from most years are available through Lawrence Erlbaum Associates, Inc. The annual proceedings of the Cognitive Science Conference represent a major source of information on new work and new ideas in the scientific study of thinking. In 1990, the Society, with help from an anonymous donor, established the David Marr Prize for the best student paper at each annual meeting.
<br><br>
* **Journal Creation**<br>
The Journal, Cognitive Science, began publication in 1976, and is now published by Wiley-Blackwell. The Executive Editor is currently Richard P. Cooper of Birkbeck, University of London, and there are 18 Associate Editors and a 30-member editorial board. It serves as the premier outlet for research reports that intersect two or more disciplines. Copyrights for articles published in the journal are held by the Society. The Governing Board of the Cognitive Science Society voted in late 2006 to found a new journal, Topics in Cognitive Science (topiCS). The Editor in Chief is Wayne Gray, Cognitive Science Department, Rensselaer Polytechnic Institute. The journal seeks to fill a niche not occupied by Cognitive Science Journal or other cognitive science journals. Membership in the Society includes a subscription to Cognitive Science and TopiCS. Copyrights for articles published in the journal are held by the Society.
<br><br>

#### Code of Conduct

By attending the CogSci 2025 Conference, you are required to adhere to the society’s **[Code of Conduct](https://drive.google.com/file/d/1ChPuihLy6jE_BWqfO7J2KKgX35JW2zsM/view?usp=sharing)**.
<br><br>


You need to log in with the email address you registered with. 

Login credentials were sent to you from Underline -  subject line "Welcome to the CogSci 2025 Conference". Please be sure to check your spam/promotional inbox  if you do not see an email confirmation right away.





Please log in to join this event.

To access the site, please register [**here**](https://cognitivesciencesociety.org/registration/).

If you are registered and feel like you are seeing this message by mistake, please make sure you are logged in with the same email that you registered with. 

Please register!

The 47th Annual Meeting of the Cognitive Science Society presents the latest research across cognitive science and highlights the theme of Cognition in Context.

Self-regulated learning involves making decisions about how to study, but students often choose suboptimal strategies. Our experiment investigated how preferences for instructional methods—video, practice, or both combined—affect learning outcomes. We randomly assigned 130 participants to either receive their preferred method of instruction (honoring initial choice) or a different method (dishonoring initial choice). Contrary to previous research showing preferences for lectures, our participants initially selected practice-based approaches. However, when asked again after instruction, the majority of participants chose the combined approach. This shift in preferences suggests that students may overvalue comprehensive approaches, even when practice alone was equally effective and reduced instructional time by 66%. Whether preferences were honored or dishonored did not significantly affect performance or efficiency, thus control over instructional methods may be less important than the methods themselves. Based on our findings, future research should focus on guiding students to utilize practice to optimize learning efficiency.

To Honor or Dishonor Student Choices? The Impact of Self-Regulation on Instructional Methods and Learning Outcomes

It has long been known that human observers can identify actions based on how people move, even from very impoverished motion depictions such as Point Light Displays (PLDs). This study investigates how humans classify actions, and what types of motion information they use to do so. Using a newly available technique (OpenPose) for extracting human joint locations from natural video, we created three types of reduced displays: PLDs, stick figures, and motion flow videos. Participants identified actions in these videos through verbal responses, and these responses were analyzed for semantic similarity using a Natural Language Processing model. A Hierarchical Bayesian Model further compared semantic similarities across video conditions. Results showed the highest intersubjective agreement (a proxy for proportion correct) for stick figures, followed by PLDs, and the lowest for motion flow videos. These results suggest that dynamic pose representations are crucial for accurate action classification, with motion flow supporting only coarse classification. The same pattern held across different action categories, such as instrumental versus locomotion and upper versus lower limb actions.

Human Action Classification from Naturalistic Videos

Lexical networks may vary as a function of individual differences in vocabulary knowledge and word-level features. Analyses often rely on descriptive network statistics, which do not support robust inferences. This study introduces the latent space model as a method for assessing the degree to which network structure is accounted for by word-level features. We analyze lexical networks from adults with below-average vs. above-average receptive vocabulary knowledge (n = 22 per group). We used latent space models to assess effects of semantic, taxonomic, and phonological similarity between words on network structure as well as effects of part-of-speech, concreteness, age-of-acquisition, and word frequency. For both groups, we found significant effects of semantic and taxonomic similarity, with additional effects of phonological similarity and concreteness for the low vocabulary group. These findings suggest increased reliance on fewer cues in lexical networks of adults with larger vocabularies. Implications for inferential modeling of lexical networks are discussed.

Investigating the Impact of Vocabulary Size on Lexical Networks using Latent Space Modeling

Face prototype formation plays a pivotal role in translating face experience into robust internal representations. However, the early developmental trajectory and experiential influences on this cognitive capacity remain underexplored. Across five within-subject experiments conducted in Canada and China across different time periods, we investigated how face race and environmental context modulate infant prototype formation. Using a novel prototype formation paradigm, we discovered infants consistently exhibited stronger prototype formation for unfamiliar other-race faces compared to own-race faces, and this bias increased significantly with age. Furthermore, we demonstrated remarkable plasticity. Infant cohorts tested during and after COVID-19 lockdowns show opposite age-related trajectories in face prototyping, reflecting differential environmental exposure to diverse faces. These findings illuminate the experience-dependent nature of early face processing specialization, suggesting infants’ prototype formation dynamically adapts to optimize face processing within specific environments. We discuss implications for understanding the developmental origins of face processing biases and potential social consequences.

Enhanced Prototype Formation for Other-Race Faces in Infancy: Developmental Trajectories and Environmental Adaptations

The question of whether mental disorders are brain disorders has sparked curiosity in cognitive science for years. But does framing mental disorders as brain disorders actually help the public better understand and engage with mental health? What do people understand when we call something a brain disorder, and why does it matter if mental illnesses are described as brain-based? To explore these questions, we conducted three quantitative vignette studies with a UK-based general public sample, focusing on perceptions of seven mental disorders: ADHD, ASD, OCD, major depressive disorder, anorexia nervosa, schizophrenia, and bipolar disorder (as defined in the DSM-5-TR). Our findings show that seeing mental disorders as brain disorders is linked to beliefs about greater severity, earlier onset, longer duration, lower chances of recovery, and higher effectiveness of medication. These results highlight how public perceptions might impact reasoning and decision-making about mental health.

The (side) effects of medicalization: How viewing mental disorders as brain disorders shapes perceptions of onset, recovery, severity, and treatment efficacy in the general public

The typicality effect is the finding that some members of a category are more ``central'' and others more ``peripheral''. This effect is seminal for understanding the mental representation of concepts. Recently, researchers have looked for typicality effects in the representations learned by machine learning models as evidence of their cognitive alignment. Studies of the typicality effect in Large Language Models (LLMs) have focused on models trained on English corpora and category norms collected from English speakers. Here, we use existing norms to investigate the typicality effect across five languages: English, French, Portuguese, German, and Spanish. We focused on eight categories common across these norms, and asked whether a multilingual LLM, GPT-4o-mini, shows human-like typicality effects across these languages. The results show variation in typicality gradients across languages. Importantly, GPT-4o-mini's typicality judgments show strong alignment with human norms for some languages: English and French. The strong performance for French, in particular, cannot simply be attributed to the representation of that language in the training corpus. We discuss the implications of these findings for future studies exploring alternative model prompting approaches, different languages, and the modeling of new category norms collected using uniform methods.

Cross-Language Typicality Effects in a Multilingual Large Language Model

This study examined the effects of regularity, frequency, and phonological complexity on morphological production in Turkish-speaking children with developmental language disorder (DLD) and phonological disorder (PD) compared to typically developing (TD) peers. Thirty children (ages 4–6) completed elicited production tasks using real and nonce words with regular and irregular noun and verb suffixes. DLD children showed lower accuracy in tasks involving consonant voicing and epenthesis and performed significantly worse on irregular suffixation, often substituting irregular forms with familiar ones. Nonce word production confirmed these challenges. Random Forest analyses indicated that phonotactic probability best predicted TD performance, while lemma frequency and phonological neighborhood density were more influential for DLD and PD groups, respectively. These findings suggest that DLD children rely on familiar, regular forms to manage morphological complexity, reflecting distinct processing strategies compared to PD and TD peers.

Exploring the Impact of Regularity, Frequency and Phonological Complexity on Morphological Production in Children with DLD and Phonological Disorders

Semantic distinctions are encoded variably in kinship terminology, the set of words that denotes family members. Nonetheless, it has been suggested that kinship terminology, like other linguistic domains, is constrained by opposing pressures to be simple yet expressive. Here, we use this insight to explore how the meaning space for kinship is structured cross-linguistically. Under the assumption that kinship systems map forms to meanings in a compressible, structure-preserving manner, we designed a metric for identifying which semantic features are most important for distinguishing individuals in a kinship system. For 1229 kinship systems, we calculated the correlation between semantic similarity (the weighted sum of shared semantic features between individuals) and wordform similarity (the edit similarity between terms).  We then identified the optimal weight for each semantic feature in each language, confirming that kinship systems vary in which semantic features they encode, and that the features themselves vary in the extent to which they are encoded. Additionally, we identified that semantic features are encoded hierarchically; more simple and more informative features are weighted highest in general. By identifying this constraint on the distribution of forms and meanings in kinship terminology, our results provide new insights on how kin terms are structured for efficient communication.

Efficient communication drives the semantic structure of kinship terminology

Do large language models (LLMs) learn like people do? We investigate this question with a simple task that compares human learning and LLM finetuning on the same set of novel inputs. We find that while humans learn and generalize robustly, finetuned LLMs largely fail to generalize from what they learned and are more influenced by prior expectations than humans are. We then analyze human solutions of our task and find that stronger performance is characterized by the proactive formation of efficient representations that aid learning and generalization. Although LLMs can use in-context learning to match the performance of humans who do not form these representations, and can use similar representations provided in-context to match the performance of those who, they do not form these representations on their own. Given these findings, we then consider how future theories of human learning might be built in the age of LLMs

Humans learn proactively in ways that language models don't

Learning about complex, scientific topics often involves reading competing explanations posited by multiple disagreeing sources. This necessitates comprehending both explanations, understanding the extent of their disagreement, and determining which is more likely. In a series of three experiments, we investigated the role of features of explanations and their sources in readers’ evaluations of the explanations. Specifically, we presented participants with pairs of disagreeing explanations that varied in their simplicity, the expertise of their source, and the salience of each feature. We examined the extent to which these features individually and interactively affected readers’ evaluation of explanations, the causes they attributed to the disagreement, and curiosity about the topic of disagreement. We also examined the role of individual differences between readers, namely their prior topic knowledge and trust in science, in these outcomes. The findings inform theory about how people evaluate explanations and learn about science.

Downloads

Next from CogSci 2025

To Honor or Dishonor Student Choices? The Impact of Self-Regulation on Instructional Methods and Learning Outcomes

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from CogSci 2025

To Honor or Dishonor Student Choices? The Impact of Self-Regulation on Instructional Methods and Learning Outcomes

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES

Downloads