Foreign Accent Research Articles

Acquisition of a Foreign Accent by Native Speakers of English Living in the Czech Republic

First language attrition has been shown to affect many areas of linguistic performance in immigrants to other countries. In phonetics, there is often a shift towards the majority language phonetic features, and, in some cases, the speakers may cease to be perceived as native altogether. This article presents the results of a foreign accent rating study, showing that even Anglophone expatriates are not immune to L1 attrition despite the (relative) lack of pressure to linguistically assimilate due to the status of their mother tongue as a prestigious and desirable-to-master language. The quantitative results are augmented by personal narratives of the expatriate informants, showing that what is often dismissed as mere anecdotal evidence may in fact bear a strong correspondence to the quantitative data.

EFFECTS OF AN ARABIC ACCENT ON EFL LEARNERS' PRODUCTIVE INTELLIGIBILITY

This study aimed at investigating the effects of a foreign accent, namely the Iraqi Arabic accent, at the segmental level on the productive intelligibility of Iraqi EFL learners. Drawing on an intelligibility pronunciation principle, i.e. Gimsons (2001) Minimum General Intelligibility (MGI), the study applied a mixed-methods research approach to measure the extent to which features of this accent impede the productive intelligibility of these learners and to identify the communication strategies they use to overcome intelligibility failures. To achieve these aims, two data collection tools were used: a production intelligibility test and a speaking task. Although the overall quantitative findings revealed that Iraqi EFL learners foreign-accented English was intelligible at the segmental level, most intelligibility failures were ascribed to the mispronunciation of non-existent English phonemes. The qualitative aspect of the study aimed at identifying the communication strategies Iraqi EFL learners use to overcome these intelligibility failures. In this respect, several strategies were identified, namely the let-it-pass strategy, the replacement strategy, the repetition strategy and the time gaining strategy. The article concludes with the implications and applications of the findings.Corresponding author: Ahmad NazariDOI:: doi.org/10.24071/llt.2020.230201

Foreign accent classification using deep neural nets

Foreign accent classification using deep neural nets

A New Acoustic-Based Pronunciation Distance Measure.

We present an acoustic distance measure for comparing pronunciations, and apply the measure to assess foreign accent strength in American-English by comparing speech of non-native American-English speakers to a collection of native American-English speakers. An acoustic-only measure is valuable as it does not require the time-consuming and error-prone process of phonetically transcribing speech samples which is necessary for current edit distance-based approaches. We minimize speaker variability in the data set by employing speaker-based cepstral mean and variance normalization, and compute word-based acoustic distances using the dynamic time warping algorithm. Our results indicate a strong correlation of r = −0.71 (p < 0.0001) between the acoustic distances and human judgments of native-likeness provided by more than 1,100 native American-English raters. Therefore, the convenient acoustic measure performs only slightly lower than the state-of-the-art transcription-based performance of r = −0.77. We also report the results of several small experiments which show that the acoustic measure is not only sensitive to segmental differences, but also to intonational differences and durational differences. However, it is not immune to unwanted differences caused by using a different recording device.

Іноземний акцент персонажів художніх творів як проблема перекладу

Іноземний акцент персонажів художніх творів як проблема перекладу

Listener perceptions of foreignness, precision, and accent attribution in a case of foreign accent syndrome

This paper will examine how listeners perceive foreignness, precision, and speaker origin in the speech of a speaker with neurogenic foreign accent syndrome, as compared to an unimpaired native speaker, an unimpaired foreign speaker, and a native speaker with ataxic dysarthria. Listeners with and without experience in speech-language pathology rated degree of foreignness and precision of speech sounds of a speaker with neurogenic foreign accent syndrome as compared to (a) an unimpaired native speaker, (b) an unimpaired foreign speaker, and (c) a native speaker with ataxic dysarthria. Listeners also attributed national origin to each speaker and rated confidence in all their attributions and ratings. Listeners rated a native speaker with foreign accent syndrome as less foreign than a true foreign speaker. They agreed less and were less accurate when attributing national origin to a speaker with foreign accent syndrome and were less confident in their attributions and ratings. Unexpectedly, listeners were equally as unsure when attributing origin to a native speaker with ataxic dysarthria as they were when attributing origin to a native speaker with foreign accent syndrome. While previous literature has characterized foreign accent syndrome as perceptually unique in comparison to true foreign accents and other motor speech disorders, our results suggest that listeners do not universally discern a clear difference between foreign accent syndrome speech and other disordered speech. This may suggest that foreign accent syndrome is not a discrete diagnosis. If it is a discrete diagnosis, these results rebut the idea that foreign accent syndrome is the only motor speech disorder in which the speaker is perceived as foreign. • Foreign accent syndrome (FAS) and ataxic dysarthria have perceptual similarities. • FAS speech rated more foreign, less precise than other speech. • Accuracy and confidence decrease when placing national origin of speakers with FAS.

Hearing faces and seeing accents?

Abstract In this paper, the McGurk effect displayed by native Mandarin Speakers is examined in the light of reaction time (RT) and response types. Two within-subject factors, FACE and ACCENT, and one between-subject factor, English Proficiency, were incorporated in the experiment. The results showed that FACE and ACCENT, but not English Proficiency, had effects on the participants’ RT and response types. When a foreign ACCENT was dubbed onto a familiar FACE, the RT was the longest, and the McGurk effect was most likely to be found. Other kinds of McGurk stimuli composition did not receive different RT but induced different response types. When the FACE was foreign, regardless of the ACCENT, the participants tended to respond with perceptive illusion. The author concluded that the expectations of the perceiver influenced the use of multisensory integration and thus the longer RT and the appearance of the McGurk effect.

Non-native speaker pause patterns closely correspond to those of native speakers at different speech rates.

When speaking a foreign language, non-native speakers can typically be readily identified by their accents. But which aspects of the speech signal determine such accents? Speech pauses occur in all languages but may nonetheless vary in different languages with regard to their duration, number or positions in the speech stream, and therefore are one potential contributor to foreign speech production. The aim of this study was therefore to investigate whether non-native speakers pause ‘with a foreign accent’. We recorded native English speakers and non-native speakers of German or Serbo-Croatian with excellent English reading out an English text at three different speech rates, and analyzed their vocal output in terms of number, duration and location of pauses. Overall, all non-native speakers were identified by native raters as having non-native accents, but native and non-native speakers made pauses that were similarly long, and had similar ratios of pause time compared to total speaking time. Furthermore, all speakers changed their pausing behavior similarly at different speech rates. The only clear difference between native and non-native speakers was that the latter made more pauses than the native speakers. Thus, overall, pause patterns contributed little to the acoustic characteristics of speakers’ non-native accents, when reading aloud. Non-native pause patterns might be acquired more easily than other aspects of pronunciation because pauses are perceptually salient and producing pauses is easy. Alternatively, general cognitive processing mechanisms such as attention, planning or memory may constrain pausing behavior, allowing speakers to transfer their native pause patterns to a second language without significant deviation. We conclude that pauses make a relatively minor contribution to the acoustic characteristics of non-native accents.

Pragmatics and social meaning: Understanding under-informativeness in native and non-native speakers

Foreign-accented non-native speakers sometimes face negative biases compared to native speakers. Here we report an advantage in how comprehenders process the speech of non-native compared to native speakers. In a series of four experiments, we find that under-informative sentences are interpreted differently when attributed to non-native compared to native speakers. Specifically, under-informativeness is more likely to be attributed to inability (rather than unwillingness) to say more in non-native as compared to native speakers. This asymmetry has implications for learning: under-informative teachers are more likely to be given a second chance in case they are non-native speakers of the language (presumably because their prior under-informativeness is less likely to be intentional). Our results suggest strong effects of non-native speech on social-pragmatic inferences. Because these effects emerge for written stimuli, they support theories that stress the role of expectations on non-native comprehension, even in the absence of experience with foreign accents. Finally, our data bear on pragmatic theories of how speaker identity affects language comprehension and show how such theories offer an integrated framework for explaining how non-native language can lead to (sometimes unexpected) social meanings.

Rater Effects on L2 Oral Assessment: Focusing on Accent Familiarity of L2 Teachers

ABSTRACT In the present study, I examined the effects of rater characteristics, in particular, raters’ familiarity with a foreign accent, on the assessment of second language (L2) pronunciation. Forty-three native English-speaking teachers were divided into three groups according to their reported types of familiarity with Korean accents: heritage, familiar, and unfamiliar. The teachers assessed global proficiency and accentedness of speech samples provided by 40 Koreans speaking L2 English. The ratings made on 1–9 numerical scales were analyzed for interrater reliability and rater severity. While the three teacher groups overall exhibited high interrater reliability, the findings suggest that prior exposure to, or regular contact with, foreign-accented speech may affect teachers’ consistency in ratings. A comparison of the three groups’ severity measures revealed that the heritage teachers were more lenient than the non-heritage (i.e., familiar and unfamiliar) teachers in evaluating global proficiency and accentedness. The difference in the level of severity between the familiar and unfamiliar teacher groups was not significant.

Americans’ attitudes toward foreign accents: evaluative hierarchies and underlying processes

ABSTRACT This study examined Americans’ attitudes toward standard American English (SAE) and nine, non-Anglo foreign accents: Arabic, Farsi, French, German, Hindi, Hispanic, Mandarin, Russian, and Vietnamese. Compared to SAE speakers, all foreign-accented speakers were rated as harder to understand, more likely to be categorised as foreign (rather than American), and attributed less status and solidarity. However, not all foreign accents were equally denigrated on status and solidarity traits. Instead, an evaluative hierarchy emerged, with speakers of some varieties (e.g. French, German) consistently rated more favourably than speakers of others (e.g. Arabic, Farsi, Vietnamese). This variation in language attitudes was associated with variation in social categorisation – i.e. the higher the percentage of nonstigmatized foreign categorizations (i.e. Anglosphere, Western Europe) for a given foreign variety, the more favourably speakers of that variety were rated – and listeners’ processing fluency – i.e. the easier speakers of a given foreign variety were to understand, the more favourably they were rated.

Experiences of International Teaching Assistants in the US Classroom

Most research focusing on the challenges that international teaching assistants (ITAs) encounter in U.S. classrooms employs a linguistic perspective. The present study furthers that research by examining other challenges unique to ITAs, through the lens of an intercultural competence framework. Through individual interviews with 15 ITAs, the study highlights the challenges related to competencies in knowledge and skills faced by ITAs in U.S. classrooms. Findings reveal that knowledge about the U.S. education system, expectations of the classroom culture, and assumptions about student–instructor relationships pose the greatest difficulties. Additionally, the ability of an ITA to demonstrate communication skills remains a significant challenge, including the negative perception of speaking with a foreign accent and selecting effective word choices to accurately represent content. The study findings present practical implications for training ITAs for their pedagogical duties at U.S. colleges.

The tolerance of English instructors towards the Thai-accented English and grammatical errors

Although Thai English has emerged as one variety of World Englishes (Trakulkasemsuk 2012, Saraceni 2015), it has not been enthusiastically embraced by Thai educators, as evidenced in the frustration expressed by ELT practitioners over Thai learners’ difficulties with pronunciation (Noom-ura 2013; Sahatsathatsana, 2017) as well as grammar (Saengboon 2017a). In this study, we examine the perception English instructors have on the different degrees of grammar skills and Thai-oriented English accent. We investigated the acceptability and comprehensibility of both native-Thai and native-English instructors (ten of each), as these subjects listen to controlled passages produced by 4 Thai-English bilingual speakers and another 4 native-Thai speakers. There were 3 types of passage tokens: passages with correct grammar spoken in a near-native English accent, passages with several grammatical mistakes spoken in a near-native English accent, and the last being a Thai-influenced accent with correct grammar. We hypothesized that (1) native-Thai instructors would favor the near-native English accent over correct grammar, (2) native-English instructors would be more sensitive to grammar than a foreign accent, and (3) there is a correlation between acceptability and comprehensibility judgment. The findings conformed to the first hypothesis given that most Thai instructors were tolerant towards the near-native English accent, regardless of grammatical errors. The second hypothesis is rejected since native-English instructors were less tolerant towards both grammatical errors and the foreign accent. The third hypothesis was proved correct, that acceptability correlates with comprehensibility. Our study suggests that English instructors should devote proportionate attention to teaching both pronunciation and grammar. They should also be made aware of the negative attitude against Thai-accented English, so that learners would be treated fairly and without discrimination based on their Thai-influenced accent.

When processing costs impact predictive processing: The case of foreign-accented speech and accent experience.

Listeners use linguistic information and real-world knowledge to predict upcoming spoken words. However, studies of predictive processing have focused on prediction under optimal listening conditions. We examined the effect of foreign-accented speech on predictive processing. Furthermore, we investigated whether accent-specific experience facilitates predictive processing. Using the visual world paradigm, we demonstrated that although the presence of an accent impedes predictive processing, it does not preclude it. We further showed that as listener experience increases, predictive processing for accented speech increases and begins to approximate the pattern seen for native speech. These results speak to the limitation of the processing resources that must be allocated, leading to a trade-off when listeners are faced with increased uncertainty and more effortful recognition due to a foreign accent.

Is segmental foreign accent perceived categorically?

The second language learning process involves acquisition of sounds that differ to varying degrees from the sounds of a learner’s native language. Learners’ productions are strongly influenced by their native language particularly for sounds which are similar but non-identical in the two languages. However, foreign accent is typically investigated at the level of utterances and as a consequence the segmental basis of foreign accent and its role in communication remain unclear. The principal issues addressed by the current study are whether accentedness at the segmental level is judged categorically by native listeners, and whether consonantal and vocalic segments are treated similarly. British English listeners judged as native or non-native a series of words in which a single vowel or consonant had been replaced by its Spanish-accented counterpart. The degree of segment accent was varied in equal amounts along a 21-step continuum using a blending technique based on native and non-native segments excised from words spoken by a balanced bilingual talker. Listeners assessed 24 distinct consonant and vowel continua. Averaged across all vowel or consonant continua, listeners’ nativeness judgements varied with segment nativeness in a non-categorical fashion. However, most individual consonant continua, as well as those vowel continua that involved durational changes, were perceived categorically. These results suggest that while overall segment-level foreign accent might be considered to vary in degree, in reality at the level of individual segments –where second language learners’ pronunciation awareness and control has to be focussed– small acoustic changes can convey a foreign accent.

Intersectionality of Symbolic Violence against Highly-Qualified Female Migrants in Poland

The article describes the manifestations of non-physical hurt experienced by highly skilled female migrants from Belarus and Ukraine. Referring to intersectional approach, I show how three social categories: nationality, age and gender affect their everyday experience of living and working in the corporate sector in Poland with Polish colleagues. Using examples from in-depth interviews with migrants I analyse the occurrences of symbolic violence such as staring, ignoring, social rejection, double standards, being asked infuriating questions, hearing insulting tirades, conditional forms of acceptance, discourses of suspicion, pointing out foreign accent. Being young female foreign specialists but coming from the former USSR republics they go against the normative expectations of the host society representatives and distort the symbolic power structures, which results in them experiencing various forms and degrees of symbolic oppression. The paradox of their situation is that on the macro level, these migrants were encouraged to come to Poland and could benefit from certain institutional facilitations, but on the micro level these women experienced unequal treatment through interaction with the members of the host society.

Using deep learning to classify English native pronunciation level from acoustic information

The main purpose of this research is to test the use of deep learning for automatically classifying an English learner’s pronunciation proficiency, a step in the construction of a system that supports second language learners. Our deep learning dataset consists of 28 speakers – ranging in proficiency from native to beginner non-native – reading the same 216-word English story. In the supervised deep learning training process, we first label the English proficiency level of the data, but this is a complicated task because there are a number of different ways to determine someone’s speech proficiency. In this research, we focus on three elements: foreign accent, speech fluency (as measured by total number of pauses, total length of pauses, and speed of speech) and pronunciation (as measured by speech intelligibility). We use Long Short-Term Memory (LSTM) layers for deep learning, train a computer on differently labeled data, test a computer on separate data, and present the results. Features used from audio data are calculated by Mel-Frequency Cepstrum Coefficients (MFCCs) and pitch. We try several combinations of parameters for deep learning to find out what settings are best for our database. We also try changing the labeling method, changing the length of each audio sample, and changing the method of cross-validation. As a result, we conclude that labeling by speech fluency instead of by speech intelligibility tends to get better deep learning test accuracy.

How I Fell in Love with Spain

How I Fell in Love with Spain Laurence de Looze (bio) When I was in seventh grade in Chicago, a girl in my class who had moved to the US from Spain, Paloma Larramendi, stood up one day and gave her full name. Maybe we all had to do it—I don't really remember. I just remember her doing it. What she recited was: "My name is Paloma María Eugenia Gabriela González de Fernández de Mutis de Larramendi." In my memory I hear these names rolling out with rolled Rs and rounded vowels for something like several minutes. I'm sure it did not take Paloma nearly that long to say her name, but as an American kid I was used to people having just a first, middle, and last name, and that was all. I had never heard of anything more than three names, except in very rare cases when someone had a second middle name because of a dead grandparent. As Paloma recited this long name, which was more of a poem than a name, I felt as though I were being lifted and transported into some mysterious world in which the air was different and unknown forces moved people about in ways unknown to me. I was dumbfounded, and I remember that I looked around the room to see if any of my comrades were stunned the way I was. But I saw no sign of surprise on their faces. Mark Cleary, who did not even have a middle name, was sitting at his desk, blasé as always and probably wondering if any of this would be on a test later. And Harold Steinman was too busy passing notes to Edward Fuller to have even heard her. Not even Emily Gardiner, usually the most sensitive among us and given to writing poetry, seemed to have taken much notice. When Paloma finished, she seemed quite satisfied with herself, and before she sat down again at her desk she smiled as though she were in communication with something that none of us could quite understand. Paloma's act of rolling out her complete name was evidently a powerful moment for her. In the way of Americans, the kids in my class tended to shun anyone who was different or foreign, and Paloma was both of those things. She had a foreign accent, her clothes were old-fashioned, and she was clumsy at all sports. That was enough in any American junior high at that time to make her an outcast. Not every country is this way, incidentally. A Brazilian family I know did an exchange with a North American one: the son of the American family was treated as a wonderful, exotic apparition in Rio de Janeiro, and on his first day at school all of the girls came up and kissed him. When the Brazilian family sent their [End Page 167] daughter north the following year, they were surprised that she could not make friends because, as one student told her, she was "weird." I was no better than any of my fellow students. I would love to be able to say that I befriended Paloma, but I did not. In fact, it is probably just as well that I did not know that "paloma" meant "dove" in Spanish, because I might have teased her about it. After Paloma's one brief, glowing moment I never really noticed her again. She may have been in the same high school choir as I, but if she was, I have absolutely no memory of her at rehearsals or concerts. In fact, I thought of her again only when I sat down to write this essay. I Googled her name and discovered that there is indeed a Paloma Larramendi Wilson in the Chicago area, so I assumed that she must have found love and marriage in some fashion, most likely followed by children as well. Indeed, Spain and all things Spanish pretty much disappeared from my consciousness. I fell in love with France and things French. France was the land of grand culture and refinement (somehow I had this idea despite knowing almost nothing about...

The non-linguistic factors of the nature of the foreign accent in non-native speech

The non-linguistic factors of the nature of the foreign accent in non-native speech

English Language Accent Classification and Conversion using Machine Learning

Language is considered an integral part of human communication. Sometimes, a non-native speaker who is as fluent as the native speaker may not be comprehended suitably because of dissimilitude in accents. Also, detecting a person’s accent itself poses an arduous task due to subtle heterogeneity in pronunciation. In order to curtail these communication barriers arising because of accents distinctness, we have proposed a system to detect and convert speech which can conveniently differentiate one accent from another which has achieved an accuracy of 68.67%. The main motivation was solving the difficulty of Indians to understand the foreign accents and of foreigners to understand the Indian accent. The scope of this project would be limited to the English language. Detection of accent would involve identifying the speaker’s native language. We introduced a novel method of augmenting the traditionally exercised techniques by using Convolutional Neural Network (CNN) fused with Deep Neural Network (DNN) along with Recurrent Neural Networks (RNN) to subsequently enhance the accuracy. For accent conversion, audio features mainly consist of Mel-Frequency Cepstral Coefficients (MFCCs), Aperiodicity (AP) and Fundamental Frequency (F0). These features once extracted are processed by a Generative Adversarial Network (GAN) to convert source accent to target accent.