Year: 2000
Type: article
Abstract: Of Words, Birds, Worms, and Weeds: Infant Word Learning and Lexical Neighborhoods. George J. Hollich ([email protected]) Department of Psychology Ames Hall; Johns Hopkins University Baltimore, MD 21218 USA Peter W. Jusczyk ([email protected]) Department of Psychology Ames Hall; Johns Hopkins University Baltimore, MD 21218 USA Paul A. Luce ([email protected]) Department of Psychology University at Buffalo Buffalo, NY 14260 USA How specific are infants’ representations of words? Do words that sound similar to each other present any special difficulties, or benefits, in early lexical acquisition? That is, experience in encoding certain kinds of phonotactic se- quences and metrical patterns could facilitate the acquisition of new word-to-world mappings (Jusczyk, 1997). Alterna- tively, competition from existing lexical items that share similar phonotactic and phonetic properties could also in- hibit children’s ability to encode a new item (see Luce & Pisoni, 1998; Marslen-Wilson, 1989; McClelland & Elman 1986; Norris, 1994). Thus, for example, children who know the word, hat, could conceivably learn the word, had, more quickly than a phonetically unrelated word because their experience with the ha- sound structure makes form- ing an acoustic package easier. On the other hand, competi- tion from the “hat” representation, could make had very difficult to learn and inherently confusable with “hat.” Two studies are reported that examine infants’ abilities both to detect the similarity among such “lexical neighbors,” words that differ by a single phoneme, and to learn a refer- ent for a novel neighbor after an exposure to a high number of these similar sounding words. In all studies, the lexical neighbors were constructed of CVC non-words that differed in the initial consonant, the vowel, or the final consonant of a prototype. All lists were controlled for word phonotactics, frequency, and their relation to English lexical neighbor- hoods. In study 1, 15-month-old infants exhibited a novelty pref- erence for a neighborhood prototype, after being familiar- ized in the head turn preference procedure with twelve lists of twelve neighbors. The mean looking time in seconds, with the standard error in parentheses, to the novel and pro- totypical words was 7.95 (0.52) and 6.70 (0.68), respec- tively. This suggests that, even by 15 months, infants are capable of detecting the neighborhood similarity among words. In study 2, 17-month-olds were tested on their ability to learn the referent of two novel prototypes after being ex- posed to their respective lexical neighbors. In one condition, the high-density condition, six lists of twelve neighbors were used. The low-density condition utilized six lists of three neighbors plus nine filler items. Results obtained with the intermodal preferential looking procedure indicated that word learning was significantly better in the low density condition, both in overall looking times and in infant reac- tion times to the targeted word. The mean difference in looking times between the target and non-target in the high- and low-density conditions was -0.14 (0.14), and 0.59 (0.21) seconds, respectively. Taken together, these results fit well with current models of spoken language recognition, many of which suggest a competitive effect for words arising from dense lexical neighborhoods. However, preliminary results from a con- trol study seem to indicate that some exposure to a neigh- borhood may be better than no exposure at all. Thus, 17- month-old infants that were tested on their word learning ability after being exposed to twelve lists of only filler items performed worse than those from the low-density condition reported above did. This suggests that some exposure to lexical neighborhoods might facilitate and strengthen in- fants’ ability to form a representation of the new word, while too much exposure might fatigue the system and/or intro- duce strong competitive effects. References Jusczyk, P. (1997). The discovery of spoken language. Cambridge, MA: The MIT Press. Luce, P. A., & Pisoni, D. B. (1998). Recognizing spoken words: The neighborhood activation model. Ear & Hearing, 19, 1-36. Marslen-Wilson, W. D. (1989). Access and integration: Projecting sound onto meaning. In W. D. Marslen-Wilson (Ed.), Lexical access and representation. Cambridge, MA: Bradford. McClelland, J. L., & Elman, J. L. (1986). The TRACE model of speech perception. Cognitive Psychology, 18, 1- Norris, D. (1994). Shortlist: A connectionist model of con- tinuous speech recognition. Cognition, 24, 1495-1520.
Authors: George Hollich, Peter W. Jusczyk, Paul A. Luce
Language: English
Subfield: Developmental and Educational Psychology
Field: Psychology
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W2561681568