Development of Liaison Representation and its Top-Down Influence on Word Processing in Infants

ABSTRACT

We examined how toddlers process lexical ambiguity where different underlying forms are neutralized at the surface level. In a preferential-looking procedure, French-learning 30-month-olds were familiarized with either liaison-ambiguous phrases (i.e., sentences containing a determiner and a non-word, e.g., ces /z/onches, “these onches”, “these zonches”) (Experiment 1), or non-ambiguous (non-liaison) phrases (sentences containing un zonche, “a zonche”) (Experiment 2). Infants in both experiments showed a vowel-initial interpretation for the non-word, i.e., perceiving /z/ as an independent unit. In Experiment 3, 36-month-olds accepted both vowel- and consonant-initial forms (e.g., zonche, onche) after hearing the non-ambiguous cases (un zonche), suggesting an emerging but unstable understanding of the relationship between specific determiners and liaison consonants. Overall, infants represented the liaison consonant /z/ as an independent unit, consistent with the adult grammar. Furthermore, liaison knowledge biased infants’ interpretation of liaison-ambiguous cases (same as in adults) and even non-liaison cases.

Acknowledgment

We thank all the families who participated in the study.

Funding

This research was supported by SSHRC, NSERC, and CFI grants to the second author.

Notes

¹ The Test trials were not the same length in Experiment 3 as in Experiments 1 and 2. To examine if the same pattern of results would have been found in 30-month-olds had the trials lasted only 10.1 s, we analyzed the data focusing on the first 10.1 seconds of the trials. The results were similar to those obtained with 19.3 s of trial length. That is, there was a significant looking difference for Familiarized vs. Non-familiarized trials in Experiment 1a, t (15) = -2.741, p = .01, but no difference in Experiment 1b, t (15) = 1.173, p = .259. Similarly, no difference was found in Experiment 2b, t (15) = -.201, p = .844. However, the looking difference between Familiarized and Non-familiarized trials in Experiment 2a was not found with shorter Test trials, t (15) = 1.402, p = .181. This is because many infants looked throughout the whole 10.1 s part of the trial, suggesting that this initial shorter period was not sufficient for revealing younger infants’ abilities.

Additional information

Funding

This research was supported by SSHRC, NSERC, and CFI grants to the second author.