Acquisition of segmental structure in first language acquisition is accomplished through the interaction of Universal Grammar and the learner's detection of phonemic contrasts in the input (Rice and Avery, 1995; Brown and Matthews, 1993,1997). This article investigates the acquisition of the English /l–r/,/b–v/ and /f–v/ contrasts by second language learners whose L1s do not contrast these segments. Based on L1 phonological acquisition and infant speech perception research,a model of phonological interference is developed which explains how the influence of the L1 phonology originates and identifies the level of phonological knowledge that impinges upon L2 acquisition. It is proposed that if a learner's L1 grammar lacks the phonological feature that differentiates a particular non-native contrast, he or she will be unable to perceive the contrast and therefore unable to acquire the novel segmental representations. In order to evaluate this hypothesis, two experimental studies were conducted. Experiment 1 investigates the acquisition of /l/ and /r/ by Chinese and Japanese speakers; the acquisition of the /l–r/, /b–v/ and /f–v/ contrasts by Japanese speakers is compared in experiment 2. The results from an AX discrimination task and a picture selection task indicate that successful acquisition of a non-native contrast is constrained by the learner's L1 grammar. Differences between Chinese and Japanese speakers (experiment 1) and differences in the acquisition of several different contrasts among Japanese speakers (experiment 2) are argued to reflect subtle phonological properties of the learners' respective L1s. These findings demonstrate that a speaker's L1 grammar may actually impede the operation of UG, preventing the L2 learner from acquiring a non-native phonemic contrast.

No abstract

No abstract

In this paper we argue that the prosodic well-formedness conditions of a learner's native language phonology interfere with accurate perception of non-native segmental sequences. This interference induces a perceptual illusion effect due to which learners perceive an illusory vowel that, in fact, does not exist in the acoustic signal. In this way, a learner's percept of the L2 input, or L2 intake, actually includes more segmental material than is physically present. As a product of the native language phonology, these perceptual illusions are shown to occur only in environments where L2 segmental sequences cannot be licensed by L1 prosodic constituents. In addition, the quality of the perceived illusory vowel is shown to coincide with the vowel category inserted by a learner's native language phonology in environments conditioning epenthesis. Finally, by varying the duration of interstimulus intervals in an AX discrimination task, we reveal latent capacities to overcome the perceptual illusion effect when finer grained acoustic representations are made accessible. We conclude that perception of illusory segments is a direct result of a learner's native language phonology generating a phonological representation that abides by the prosodic constraints of the learner's native language.

No abstract