Poster #122 - Val vs Vaz: Children Do Not Always Disambiguate Contrasting Words

• In Event: 2-055 - Poster Session 06
  In Poster Session: PS 06 Section - Language, Communication

Integrative Statement

BACKGROUND: When children hear a novel word, they prefer to assign it to a novel, unlabeled object over a previously labeled object (Merriman, 1989). One explanation for this preference is children honor the “principle of contrast,” the assumption that different words refer to different things (Clark 1987; 1990). The current study tests whether this preference varies by the amount of contrast between words. To test this, children heard pairs of novel words, some similar (e.g., FIM and VIM, differing on one phoneme) and some dissimilar (e.g., LEK and VOOM, differing on all phonemes).

METHOD: 3- to 6-year-olds (N=50) completed a disambiguation task on a touchscreen. Each child included 28 trials, 14 “same” trials and 14 “different” trials. In each trial, an object appeared on screen and was labeled with a word (“This is a fim.”) Words always consisted of three phonemes (e.g., fim). Then, two objects appeared on screen, the original object and a novel object, and children were asked to point to the referent of a word. In the “same” trials, children pointed to the referent of the original word (“Point to the fim.”) In the “different” trials, children pointed to the referent of a novel word (“Point to the vim.”) On the different trials, words pairs differed by the number of contrasting phonemes and by the position of the contrasting phonemes (see Table 1). Object choice (selection of the correct referent) and reaction time were the primary outcome variables.

RESULTS: For object choice, a 4 (Age: 3, 4, 5, and 6-yr-olds) X 4 (Number of Contrasting Phonemes: 0, 1, 2, or 3) a Mixed Factorial ANOVA with age as the between-subjects factor and contrast as the within-subjects factor showed a significant effect for contrast, p<.001, but no effect for age and no interaction (see Figure 2). Follow-up analyses showed children were less likely to select the correct referent when words contrasted by only 1 phoneme (M=.65) compared to 0, 2, or 3 phonemes (Ms>.82), ps<.001). Additionally, reaction time data showed that children responded most slowly to one-phoneme contrasts (M=2233 ms), including when compared to 0-phoneme contrasts (i.e., “same” trials) (M=1496 ms), p<.01. Finally, position data showed that children were least accurate when responding to one-phoneme contrasts located on the 3rd phoneme (e.g., “val” vs “vaz”) compared to those on the 1st and 2nd phonemes, ps<.05. On these trials, children performed at chance, p<.01.

CONCLUSIONS: Overall findings suggest that not all contrasts are created equal. Children made more errors when judging words that contrasted by a single phoneme compared to words that did not contrast (i.e., same) and words that contrasted by two or more phonemes. Additionally, children were slowest to judge one-phoneme contrasts, spending nearly twice as long on these as compared to same trials. Finally, for one-phoneme contrasts, children were incorrect more when the words differed on the 3rd phoneme, as in CAP and CAT. For these trials, children essentially guessed. These findings have implications for when children do (or do not) honor the principle of contrast.

Authors

• Catanya Stager, University of Alabama
  Presenting Author

• Joshua Mickler, University of Alabama
  Non-Presenting Author

• Grace Parente, University of Alabama
  Non-Presenting Author

• Jason Scofield, The University of Alabama
  Non-Presenting Author