Cognitive Science Program Colloquium Series: Dr. Catherine S. Tamis-LeMonda from New York University
Date/Time: Friday, 10/3/25 from 4:00pm – 5:30pm EST
Location: McHugh Hall 302
Meetings: If you are interested in meeting with Dr. Tamis-LeMonda during the day or attending dinner in the evening on Friday, please email crystal.mills@uconn.edu.
Talk Title: Word learning in context: Disambiguating the ambiguous
Abstract: The pace and breadth of early vocabulary development is impressive to say the least. Infants grow from producing their first words around 12 months to using over 500 words by 2 ½ years. How do infants crack the code to acquire so many words in a relatively short period of time? Our theoretical framework emphasizes the embodied and embedded nature of learning: Infants actively engage with their environments in the presence of socially responsive partners who provide semantically relevant input within a tight time window during highly specific activity contexts. The tight temporal connection between infant action, caregiver speech, and activity context cuts across word classes—nouns, verbs, adjectives, prepositions—thereby functioning to ‘disambiguate the ambiguous’.
To illustrate the embodied and embedded nature of infant learning, I present several studies from our lab on infants’ exposure to different word classes in the ecologically-valid home environment. We videorecorded infants (12-24 months) and mothers during natural home activities (1 to 2 hours per visit, Ns=30-100). We transcribed interactions and identified all concrete nouns in mothers’ speech. We also marked mothers’ use of verbs and ‘math words’— adjectives and prepositions that refer to numbers, quantities (more, less), spatial relations (under, on top of), shapes, and magnitudes (big, long).
Annotations of the timing, behaviors, and contexts of infants’ speech exposure revealed several impactful characteristics: (1) Words in mothers’ speech contained high regularity in temporal structure. For example, mothers used the same word in a bout of repetition (e.g., dog, dog, dog; up, up, up) and they referenced objects from the same taxonomic category (e.g., animals) within a tight time window (e.g., dog, cat, horse); (2) Mothers provided multimodal input (speech with gesture/touch) that functioned to highlight word meaning; (3) Infants’ own actions were a reliable impetus for word exposure (e.g., verbs and spatial terms like walk, jump, down referred to infants’ movements through space), and (4) High regularity characterized the activity and location contexts of infants’ exposure to particular words (e.g., food nouns and words for magnitudes were frequent during snack time). Notably, characteristics of infants’ home experiences and language interactions predicted individual differences in skills across domains—including sustained attention, vocabulary growth, the production of words during interactions, school readiness and academic achievement years later. We discuss implications for theories of word learning and language interventions.
