Karmazyn Raz, H. & Smith, L. .
Indiana University
As novice word learners, 1-year-old children acquire language by linking heard words to visual scenes. However, the everyday visual world is highly cluttered, with many potential referents. By the time children are 2 years old, they are experienced learners, having many resources to resolve referential ambiguity, such as social cues, and category knowledge. The question of how visual statistics during learning moments reflect novel learning versus experienced learning, is essential for understanding learning across many domains.
Using head cameras, a 10-minute playtime was recorded from infants? egocentric point of view. The final sample included 32 children ages 12 and 30-months-old (N=18 for each age group). The preliminary results show that during moments where a parent named an object, the size of the referent object was larger in view for the 12-month-olds (mean =22, sd=9, n=401) than the 30-month-olds (mean=15, sd=11, n=470). The types (number of unique words) to tokens (total word count) ratio of the naming moment utterance is also larger for the 30-month-olds (mean= 0.85, sd = 0.05, n=470) than the 12-month-old (mean= 0.79, sd=0.1, n=401). These preliminary results suggest that naming moments during playtime may be a different practice for novice learners than experienced learners.