Research Article The neurophysiology of phonemic contrasts perception in L two/L three learners: The role of acquisition setting
Research Article The neurophysiology of phonemic contrasts perception in L two/L three learners: The role of acquisition setting
Abstract
Phoneme discrimination is believed to be less accurate in non-native languages compared to native ones. What remains unclear is whether differences in pre-attentive phonological processing emerge between the first foreign language (L two) and additional ones (L three/L n), and whether they might be influenced by the acquisition setting (formal vs. naturalistic). We conducted an event-related brain potential oddball study with native Polish learners of English (L two) and Norwegian (L three/L n). The results revealed a graded amplitude of the mismatch negativity effect, which was largest in L one, smaller in L two, and smallest in L three/L n. Considering the previously obtained results for naturalistic/mixed learners with the same language combination, we believe that the acquisition setting is an important factor influencing the perception of phonemic contrasts. In the naturalistic group, no difference was observed between L one and L two, while the instructed group exhibited more fine-grained distinctions between all tested languages.
One. Introduction
One. Introduction
The focus of this paper is on non-native phoneme discrimination in multilingual learners, i.e., those who, in addition to their second language (L two), have acquired at least one additional foreign language (L three/L n). The field of L three/L n phonology has been extensively explored in recent years, with the last decade witnessing manifold studies on the phonological system developed by multilinguals. Notably, the majority of this research has focused on speech production rather than perception, although this trend seems to be partially reversing in recent years. Still, a modest number of researchers have directly explored the processing of non-native phonemes in L two when compared with L three/L n, and the experimental techniques used to investigate the non-native phoneme perception in multilinguals have typically involved forced-choice goodness tasks, perceptual assimilation tasks, foreign-accentedness ratings, or discrimination tasks. What has been scarcely studied so far is the neural markers of phonemic contrast perception by multilingual listeners, with most of the work prioritizing bilingual research. In order to bridge this gap, we report the results of a phonemic perception event-related potential study, which focused on trilingual L one Polish - L two English - L three Norwegian speakers who acquired their L three/L n in a formal (i.e., classroom-instruction) setting. The selection of trilingual instead of bilingual listeners enabled us to elucidate the issue of language status (operationalized as L two or L three/L n) as a factor affecting non-native phoneme processing.
Categorical perception is an empirical phenomenon, when the ability to distinguish two sounds is affected by the categories they belong to. In a native language, we can discriminate sounds that belong to two different phonemes, but not when they belong to a single category. Categorical perception helps us process the incoming speech signal fast in the L one, but in a non-native language, it tends to hinder differentiation between phones that belong to two different phonemes in L two, but are assimilated to a single L one phoneme. Major theories of L two acquisition of speech, such as the Speech Learning Model, revised Speech Learning Model, Perceptual Assimilation Model - L two and Second Language Perception Model associate category formation in a non-native language with the ability to detect differences between L one and target language sounds. Whereas a large body of research dealt with studies of conscious non-native phone discrimination using behavioral techniques, the present study concentrates on neural, pre-attentional discrimination of native and non-native sounds. This is particularly novel with respect to investigating online phonemic perception in multilinguals, comparing two non-native languages.
As a next step, we juxtapose the results of the current study with the data obtained earlier in the same task by a group of trilingual speakers who acquired the foreign languages (i.e., English and Norwegian) in a naturalistic (i.e., immersion) setting. In this way, we examine whether the learning context (i.e., the acquisition setting) may potentially affect the neural discrimination of phonemic contrasts. Even more crucially, the current research also contributes to the ongoing debate concerning the effectiveness of foreign language learning, with non-native phoneme perception and consequently, their production as its central components.
One of the most pertinent questions both within the field of L two and L three/L n acquisition concerns the effectiveness of the learning process shaped by the learning context (or setting), with the nature of this context typically instantiated as either naturalistic (i.e., immersive) or formal (i.e., instructed). The former one typically progresses in natural communicative contexts, holistically, without interruptions and in a country or region where the target language is used on an everyday basis. The latter one is assumed to be "notably less than ideal with respect to the natural ecology of language learning", as it takes place in a restricted classroom setting, does not extend much outside classroom time, it happens in smaller portions, often with the learner's L one used for instruction, and usually centered around grammatical and lexical information. Moreover, the exposure quality in the formal classroom setting is specific, because it primarily involves interactions with L one-accented teacher and fellow classmates, entail- ing many mistargeted phonetic details. Obviously, there are also numerous other acquisition contexts situated somewhere in between these two archetypal learning scenarios. One could mention here, for instance, a mixed context where formal education proceeds by means of immersion or self-learning attempts facilitated by such resources as podcasts, videos, or language learning apps. However, since most studies to date have focused on non-native language acquisition in either immersive or formal settings, an apparent need arises for more systematic comparisons of patterns observed among two learner types with shared language repertoires. Hence, in the current paper, we will focus on two markedly different learning scenarios, i.e., one involving formal classroom instructions of L two English and L three/L n Norwegian (the study reported in this paper) as compared to the one involving naturalistic acquisition of L three/L n Norwegian and naturalistic/mixed acquisition of L two English by Polish migrants in Norway.