Function word facilitate word segmentation.

Abstract

Infants’ word segmentation is facilitated by bottom-up cues, such as prosodic cues (stress-patterns, position in a sentence, etc.). Research shows that infants also use top-down cues, i.e., known words, to segment new words [Bortfeld et al. (2005); “mommy” and the child’s name]. In two experiments, infants were tested on their ability to use a known function word “the” to segment vowel-initial words (the ability to segment them does not appear until 13.5–16 months), without any additional bottom-up cues. In experiment 1, using the headturn preference procedure, monolingual English-learning 11-month-olds were tested on their ability to segment sentence-medial vowel-initial words that followed the function word “the.” First, infants were familiarized with passages containing the target words “ash” and “eff.” Next, infants were presented with two familiar and two novel isolated words. Mean listening time to the familiar and unfamiliar words was significantly different [t(12) = 4.220, p < 0.001]. Out of 13 infants, 12 listened longer to the familiar words compared to the novel words. In experiment 2, monolingual English-learning 8-month-olds are currently being tested using the same task. Preliminary results show three out of five 8-month-olds listen longer to the familiar words compared to the novel words.