Cross‐Situational Statistics Present in an Early Language Learning Context
Evidence From Naturalistic Parent–Child Interactions
Bibliographic Data
| ID | 7150330 |
|---|---|
| Authors | Ellis S Cain (0009-0005-6619-1983, Department of Cognitive and Information Sciences University of California, Merced, corresponding author), Rachel Ryskin (0000-0002-9516-4467, Department of Cognitive and Information Sciences University of California, Merced), Chen Yu (0000-0002-4310-1923, Department of Psychology, Center for Perceptual Systems University of Texas, Austin) |
| Year | 2025 |
| Volume | 49 |
| Issue | 6 |
| Pages | e70078-e70078 |
| Publication date | 2025-06-01 |
| Peer Reviewed | Yes |
| Open Access | Yes |
| Type | ARTICLE |
| Venue | Cognitive Science (JOURNAL) |
| Journal identifiers | ISSN: 0364-0213 • E-ISSN: 1551-6709 |
| Publisher | Wiley (PUBLISHER • GB) |
| DOI | 10.1111/cogs.70078 |
| PMID | 40535997 |
| OpenAlex | W4411458454 |
| Language | EN |
| Citations received | 1 |
| References cited | 35 |
According to the cross‐situational learning account, infants aggregate statistical information from multiple parent naming events to resolve ambiguous word‐referent mappings within individual naming events. While previous experimental studies have shown that infant and adult learners can build correct mappings based on statistical regularities encoded in multiple learning situations in an experiment, other studies that use more naturalistic stimuli (e.g., real‐world video) reveal poor performance in adults' ability to infer the correct referent. Based on those results derived from more naturalistic stimuli, the cross‐situational learning solution cannot be useful to solve the mapping problem in the real world because cross‐situational statistics from the real world are much more ambiguous than those created in experimental studies. To examine the feasibility of cross‐situational learning in everyday contexts, the present study aims to quantify visual‐audio statistics from one of everyday activities—parent–child toy play. We analyze parent naming events in a video corpus of infant‐perspective scenes during parent–child toy play in a naturalistic lab setting, where we found three distinct properties that characterize statistical regularities perceived by young learners: (1) there are a limited number of visual scene compositions perceived by young learners at the moments when they hear object names; (2) the frequencies of parent naming events are distributed in a skewed, Zipfian fashion; and (3) cross‐situational statistics in naturalistic toy play are comparable to those used in laboratory experiments. Our results underscore the importance of quantifying the statistical regularities in the input from the learner's perspective in order to shed light on the mechanisms supporting early word learning
Affordance · Cognitive psychology · Linguistics · Naturalism · Referent · Situational ethics · Child and Animal Learning Development · Computer Science · Language Development and Disorders · Psychology · Social Psychology · Speech and dialogue systems
Object name Learning Provides On-the-Job Training for Attention
Embodied attention and word learning by toddlers
Zipf’s word frequency law in natural language
Mask R-CNN
Rapid Word Learning Under Uncertainty via Cross-Situational Statistics
Infants rapidly learn word-referent mappings via cross-situational statistics
Most people are not Weird
Zipfian Distributions in Child-Directed Speech
The learnability consequences of Zipfian distributions in language
Zipfian distributions facilitate children's learning of novel word-referent mappings
Learning Through Processing
Five factors that guide attention in visual search
From Play to Language
| Unique citing works | 1 |
|---|---|
| Citations per year | 1 |
| Citation span | 2025 - 2025 (1) |
| Citation velocity | recent |
| Highly cited | No |
| Citation types | Neutral: 1 |