Domain-Adversarial Based Model with Phonological Knowledge for Cross-Lingual Speech Recognition

Phonological-based features (articulatory features, AFs) describe the movements of the vocal organ which are shared across languages. This paper investigates a domain-adversarial neural network (DANN) to extract reliable AFs, and different multi-stream techniques are used for cross-lingual speech recognition. First, a novel universal phonological attributes definition is proposed for Mandarin, English, German and French. Then a DANN-based AFs detector is trained using source languages (English, German and French). When doing the cross-lingual speech recognition, the AFs detectors are used to transfer the phonological knowledge from source languages (English, German and French) to the target language (Mandarin). Two multi-stream approaches are introduced to fuse the acoustic features and cross-lingual AFs. In addition, the monolingual AFs system (i.e., the AFs are directly extracted from the target language) is also investigated. Experiments show that the performance of the AFs detector can be improved by using convolutional neural networks (CNN) with a domain-adversarial learning method. The multi-head attention (MHA) based multi-stream can reach the best performance compared to the baseline, cross-lingual adaptation approach, and other approaches. More specifically, the MHA-mode with cross-lingual AFs yields significant improvements over monolingual AFs with the restriction of training data size and, which can be easily extended to other low-resource languages.

A Self-Supervised Model for Language Identification Integrating Phonological Knowledge

In this paper, a self-supervised learning pre-trained model is proposed and successfully applied in language identification task (LID). A Transformer encoder is employed and multi-task strategy is used to train the self-supervised model: the first task is to reconstruct the masking spans of input frames and the second task is a supervision task where the phoneme and phonological labels are used with Connectionist Temporal Classification (CTC) loss. By using this multi-task learning loss, the model is expected to capture high-level speech representation in phonological space. Meanwhile, an adaptive loss is also applied for multi-task learning to balance the weight between different tasks. After the pretraining stage, the self-supervised model is used for xvector systems. Our LID experiments are carried out on the oriental language recognition (OLR) challenge data corpus and 1 s, 3 s, Full-length test sets are selected. Experimental results show that on 1 s test set, feature extraction model approach can get best performance and in 3 s, Full-length test, the fine-tuning approach can reach the best performance. Furthermore, our results prove that the multi-task training strategy is effective and the proposed model can get the best performance.

Multiple Sources of Surprisal Affect Illusory Vowel Epenthesis

Illusory epenthesis is a phenomenon in which listeners report hearing a vowel between a phonotactically illegal consonant cluster, even in the complete absence of vocalic cues. The present study uses Japanese as a test case and investigates the respective roles of three mechanisms that have been claimed to drive the choice of epenthetic vowel—phonetic minimality, phonotactic predictability, and phonological alternations—and propose that they share the same rational goal of searching for the vowel that minimally alters the original speech signal. Additionally, crucial assumptions regarding phonological knowledge held by previous studies are tested in a series of corpus analyses using the Corpus of Spontaneous Japanese. Results show that all three mechanisms can only partially account for epenthesis patterns observed in language users, and the study concludes by discussing possible ways in which the mechanisms might be integrated.

Echoes of L1 Syllable Structure in L2 Phoneme Recognition

Learning to move from auditory signals to phonemic categories is a crucial component of first, second, and multilingual language acquisition. In L1 and simultaneous multilingual acquisition, learners build up phonological knowledge to structure their perception within a language. For sequential multilinguals, this knowledge may support or interfere with acquiring language-specific representations for a new phonemic categorization system. Syllable structure is a part of this phonological knowledge, and language-specific syllabification preferences influence language acquisition, including early word segmentation. As a result, we expect to see language-specific syllable structure influencing speech perception as well. Initial evidence of an effect appears in Ali et al. (2011), who argued that cross-linguistic differences in McGurk fusion within a syllable reflected listeners’ language-specific syllabification preferences. Building on a framework from Cho and McQueen (2006), we argue that this could reflect the Phonological-Superiority Hypothesis (differences in L1 syllabification preferences make some syllabic positions harder to classify than others) or the Phonetic-Superiority Hypothesis (the acoustic qualities of speech sounds in some positions make it difficult to perceive unfamiliar sounds). However, their design does not distinguish between these two hypotheses. The current research study extends the work of Ali et al. (2011) by testing Japanese, and adding audio-only and congruent audio-visual stimuli to test the effects of syllabification preferences beyond just McGurk fusion. Eighteen native English speakers and 18 native Japanese speakers were asked to transcribe nonsense words in an artificial language. English allows stop consonants in syllable codas while Japanese heavily restricts them, but both groups showed similar patterns of McGurk fusion in stop codas. This is inconsistent with the Phonological-Superiority Hypothesis. However, when visual information was added, the phonetic influences on transcription accuracy largely disappeared. This is inconsistent with the Phonetic-Superiority Hypothesis. We argue from these results that neither acoustic informativity nor interference of a listener’s phonological knowledge is superior, and sketch a cognitively inspired rational cue integration framework as a third hypothesis to explain how L1 phonological knowledge affects L2 perception.

The Correlation Between Phonological Knowledge and Pronunciation Ability

The purpose of this research is to know whether there is a significant correlation between phonological knowledge and pronunciation ability. The method used in this research is correlation design. Variable X is the students’ phonological knowledge and variable Y is students’ pronunciation ability. Population of this research is all of the fourth semester students. There are 80 students as population and the writer took 20 students as the sample. The data are analyzed using Pearson Product Moment formula to know whether there is a correlation between phonological knowledge and pronunciation ability of the students. The result showed there is a correlation between phonological knowledge and pronunciation ability of the students. Based on the calculation that is obtained, rxy value from variable X and Y is 0,223. It is categorized as a low correlation. Whereas the result of the calculation use tcount formula. The tcount value is 1,002. It means that there is no a significant correlation between X and Y variable.

Developing Reading Skills in Prelingually Deaf Preschool Children: Fingerspelling as a Strategy to Promote Orthographic Learning

The aim of this study was to clarify whether fingerspelling provides a sophisticated mechanism that promotes the development of detailed orthographic knowledge for deaf individuals even in the absence of paralleling phonological knowledge. An intervention program comprised of various procedures chaining between fingerspelled sequences; their written correlates and meaning were administered in a multiple probe single-subject research design across semantic categories to four children with severe to profound prelingual hearing loss (age 4.2–6 years). Results demonstrate the occurrence of rapid orthographic learning during intervention and reliable retention of it in maintenance checks, despite the participants’ insufficiently developed phonological skills. Observations of the participants’ behavior further suggest fingerspelling to function as an effective mediator in the initial development of robust detailed orthographic lexicon. Analyzes also indicate that “learning through action” and “relevance to the task” provide two key factors in relation to the promotion of orthographic learning, with their absence creating an obvious vacuum in this regard. Insights from the study are discussed with reference to their implication for the development of learning materials and learning environments for prelingually deaf and other novice learners.

EIL Pronunciation Research and Practice: Issues, Challenges, and Future Directions

Research has seen a growing recognition of the crucial role pronunciation plays in teaching English as an International Language (EIL), in achieving effective communication and its close link to the mastery of other aspects of language teaching and learning. In spite of its recognised importance, pronunciation is still a marginalised skill in many EIL programmes due to EIL teachers’ lack of required phonetic and phonological knowledge, and the paucity of comprehensive coverage targeted at EIL learners. While there has been much research on phonetic and phonological features of world varieties of English, there has not been a strong research–practice nexus in the field. There is therefore a need to comprehensively review EIL pronunciation teaching, to document what previous research tells us, and to discuss how research can be translated into practice. This article aims to cover a range of current issues concerning EIL pronunciation modelling and theorising, and provides a brief articulation of the current issues surrounding the global spread of English and its theoretical development. Specifically, it considers current EIL research issues, challenges, and their implications for pronunciation practice. It also considers the implications of Gardner’s (2008) Five Minds for the Future for EIL pronunciation teaching and postulates the necessity of a sixth mind to navigate the field in the post-pandemic era.

Effects of recasts, clarification requests on suprasegment development of English intonation

A total of 102 English as a Foreign Language (EFL) learners participated in the present study, which aimed to test how different types of corrective feedback-recasts and clarification requests-can differentially affect the suprasegment development of English intonation. All participants received 5 treatment sessions designed to encourage them to notice and practice the target feature in meaningful discourse; recasts or clarification requests were provided to the participants’ untargetlike production, except those in the control group (n=34), who received comparable instruction but without corrective feedback. Acoustic analyses were conducted on 7 intonation features including words/IP, pause, anacrusis, lengthing, pitch reset, improper tonicity and tone selection elicited via pretest and posttest measures targeting trained instances and untrained instances. The results showed that 1) recasts are more effective than clarification requests on EFL learners’ suprasegment development of English intonation; 2) recasts may not only lead learners to establish, reinforce and generalize their new phonological knowledge of English intonation that they had practiced during the treatments, but also help them transmit their attention from trained to untrained learning of foreign language input at a suprasegmental level.