Activation and Perception of Native-Language Phonotactics in Bilinguals

Public Deposited

As monolinguals and bilinguals hear words unfold over time, they experience competition from words that share sounds within the same language (e.g., st- activates strict and stamp). Unique to bilinguals is that they are also prone to competition from similar sounding-words between their two languages. In the present dissertation, we examine how experience with multiple languages transforms the way in which speech sounds are activated and perceived. Specifically, we investigated how and if bilinguals activated and perceived native-language (L1) structures when processing words in their second language (L2) across three experiments. We used phonotactic constraints, which are rules for combining speech sounds, to examine activation and perceptual processes. A Spanish phonotactic constraint is that s+consonant clusters are not permitted at word onsets, and a vowel must precede the cluster (prothesis; e.g., English: strict, Spanish: estricto). In Experiment 1, we investigated whether bilinguals activated L1 phonotactic constraints when processing L2 words. We found that L1 Spanish-L2 English bilinguals accessed the L1 Spanish constraint. Bilinguals were faster to make lexical decisions on English-like words containing the activated onset rule (e.g., estomb) when primed with Spanish-conflicting words (e.g., strict), relative to non-conflicting stimuli (e.g., kneeling, prime: workers). In Experiment 2, the extent to which bilinguals perceived, or perceptually repaired, L2 words and L2-like non-words to sound more like L1 words was investigated. Response time results across vowel detection and AX discrimination tasks revealed mixed effects for L1 perception during L2 processing. Bilinguals perceived the onset in English s+consonant words (e.g., strict) when the beginning sound of the word was the focus of the task (vowel detection), but not when making low-level perceptual judgments on whether conflicting (e.g., strict and egg) versus non-conflicting (e.g., work and egg) word pairs were the same (AX discrimination). Experiment 2s results suggest that when the beginning sound of the word is the explicit focus of the task, perceptual repair to the onset occurs. In Experiment 3, eye-tracking methodology was employed across a word recognition task and a combined word recognition and AX discrimination task. The goal was to examine the relation between activation and perception of L1 sounds during L2 processing. On the visual display, when a Spanish-conflicting English target word was present (e.g., strict), along with the potentially activated onset competitor word (e.g., egg) and two unrelated filler items (e.g., work, can), Spanish-English bilinguals with lower L2 (English) proficiency looked more at the onset competitor than the filler items. This pattern was observed on the two measures of word recognition. Moreover, in the combined word recognition and AX discrimination task, bilinguals did not perceive L1 phonotactics during L2 processing. Together, in Experiment 3, lower English proficiency bilinguals activated, but did not perceive, Spanish phonotactic constraints during English processing. Overall, the findings in this dissertation provide insight into the structure of acoustic space within the bilingual mind. L1 representations for sounds and words influence how L2 input is perceived. Speech perception and language activation are susceptible to interference from the L1. These results highlight the importance of accurately identifying differences versus disorders in bilingual populations. An L1 Spanish speaker learning an L2 (English) might not be able to identify and differentiate between strict and estrict due to the L1 Spanish phonotactic rule. However, the L1 Spanish speakers behavior is due to a rule difference across languages and is not indicative of a disorder. Implications and future directions for the current research are discussed.

Last modified
  • 11/20/2019
Date created
Resource type
Rights statement