Poster #121 - The Development of Children’s Judgments and Perception of English Consonant Clusters

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

Integrative Statement

To achieve adult-like language perception, children must learn to be both accurate and flexible when listening to speech. On the one hand they must accurately attend to phonetic details to correctly interpret words (ball vs. doll, sport vs. support), but they must also remain flexible enough to recognize words with unexpected phonetic details, as when listening to an unfamiliar accent. By 14-18 months, children perceive mispronunciations in familiar words (e.g. Swingley, 2009) and can store phonetic detail in novel word representations (e.g. Yoshida et al, 2009), while their interpretation of accented speech begins around 4-7 years (Nathan et al, 1998; Bent, 2014). How does this later flexibility develop? Do children use their knowledge of words to accept or reject unexpected phonetic detail? And are some unexpected details easier to interpret than others?

This study reports children’s perceptual judgments of English words that begin with two-consonant clusters (like ‘spoon’), when compared with mispronounced variants with either a deleted consonant (‘poon’,’soon’) or an inserted vowel (‘uhs-poon’, ‘suh-poon’). In one task, children heard two cartoon aliens labeling unfamiliar objects in their ‘alien’ (invented) language and judged whether the aliens “said the same word or not”. In half of the trials, one word had an initial cluster and one had a variant pronunciation. In the second task, the same aliens attempted to label familiar English objects and actions, and children were asked which alien “said it best”. On each of these trials, one alien produced an accurate cluster and the other produced an accented variant (see table 1).

The results come from 11 typically-developing monolingual 4-6 year olds (mean age 5;7), who in a separate task all successfully produced the target consonant clusters. (15 monolingual adult controls all performed near ceiling). When judging real words, children easily rejected consonant deletions like ‘poon’ (96% correct), but were significantly more likely to accept inserted vowels like ‘uhs-poon’ and ‘suh-poon’ (70% correct). When judging novel words children were less adult-like: they were significantly better at identifying words as same (86%) vs. different (56%) – and a significant interaction between error type and cluster position showed that word-initial deletions were identified more accurately, but word-initial epenthesis was less accurate (see figure 1). The clusters’ consonants was not a significant predictor, nor was age.

Our results suggest multiple asymmetries in children’s word-level perceptual development. One finding is that children are better at recognizing consonant deletion, and less willing to accommodate deletion in accented speech, as compared to vowel insertion; this aligns with evidence elsewhere in both child and adult phonologies (Fleischacker 2005; Farris-Trimble, 2008). Differences between the tasks also suggests that children do not accept extra vowels in accented speech simply because the words are interpretable – using the context and available consonants to figure out what “uhf-reezer” is – since children are more likely to judge nonsense-word pairs with and without inserted vowels as the same. We will discuss these results in the context of ongoing perceptual tuning and the mental representation of words in later childhood.

Authors

• Anne Michelle Tessier, Simon Fraser University
  Presenting Author

• Claire Moore-Cantwell, Simon Fraser University
  Non-Presenting Author

• Kevin Kong, Simon Fraser University
  Non-Presenting Author

• Ashley Farris-Trimble, Simon Fraser University
  Non-Presenting Author