Native English-speaking Listeners Research Articles

Listener perceptions of customer service agents’ performance

Abstract This study investigated listener-based assessment of the job performance of second language (L2) speakers employed as customer service agents in outsourced foreign-based call centers, focusing on agents’ job performance as a function of the comprehensibility, fluency, and accentedness of their speech. Using Amazon’s Mechanical Turk crowdsourcing platform, 116 native English-speaking listeners evaluated two-minute recordings of actual customer service conversations featuring 18 Filipino agents, assessing them for three global speech dimensions (comprehensibility, accentedness, and fluency) and three performance indicators, including agents’ confidence, competence, and listeners’ interest in future communication with agents (a measure capturing customer patronage). Comprehensibility and fluency consistently predicted how the listeners assessed the agents on all job performance scales, and accentedness was additionally associated with how strongly the listeners wished to communicate with the agents. Findings generally highlight the importance of fluent and comprehensible L2 speech in workplace settings.

Manipulating listener attitudes toward accented talkers influences speech perception

We previously demonstrated that listeners' attitudes toward talkers with foreign accents is related to listeners' ability to perceive foreign-accented speech. In this study, we sought to determine if listeners' attitudes toward talkers with foreign accents are malleable and, if so, whether changing listeners' attitudes toward talkers with foreign accents would change the strength of the relationship between listeners' attitudes and listeners' speech perception accuracy. Specifically, we hypothesized that supporting listeners’ biases regarding non-native speech would lead to better speech perception performance. To test this hypothesis, we created short vignettes that either conformed to or violated listeners’ biases regarding non-native talkers. Native English-speaking listeners rated and transcribed speech produced by native and non-native English speakers. Half of the listeners read the vignette prior to rating the speech, half received no information about the talkers; vignettes were randomly paired with talkers. Contrary to our hypothesis, there was no effect of vignette type. Listeners rated non-native listeners as more accented and less likeable when vignettes were present. Interestingly, listeners were also more accurate transcribing speech by both native and non-native talkers when the vignettes were present.

Transient speech processing costs depend on accent type and talker variability

In a dual-task paradigm, native English-speaking listeners (N=55) made judgments about visually presented digits while simultaneously listening to speech that varied by talkers-per-accent (between-subjects: 1 vs 3) and accent type (within-subjects: native[Canadian]/regional[Australian]/nonnative[Mandarin]). Adaptation occurred within only six exposures but was only observed for nonnative and native-accented speech (p < 0.05) and was also dependent on talkers-per-accent, indicating that transient speech processing demands are uniquely impacted by the interactions between accent type and talker variability.

The effects of lecturers’ non-native accent strength in English on intelligibility and attitudinal evaluations by native and non-native English students

The rapid spread of English medium instruction (EMI) across the globe has led to a growing number of non-native lecturers teaching in English to student populations that are increasingly international. The aim of the present study was to investigate to what extent lecturers with slight or moderate Dutch accents or native British English accents are evaluated differently in terms of intelligibility, comprehensibility and attitudinal impressions by non-native and native English-speaking listeners. In an experiment, 189 Dutch listeners, 175 international non-native listeners and 158 native English listeners evaluated fragments recorded by moderately accented Dutch, slightly accented Dutch and native English speakers. Findings showed that the moderately non-native accented lecturers were evaluated more negatively than lecturers with slight or native accents by both Dutch and international non-native listeners, but not by native English listeners. This suggests that non-native listeners evaluate the accents of non-native lecturers according to native speaker pronunciation norms.

SOCIAL ATTITUDES AND SPEECH RATINGS

AbstractThis study examined whether social bias manipulation can influence how naïve multiage listeners evaluate second language (L2) speech. Sixty native English-speaking listeners (Montreal residents) rated audio recordings of 40 Quebec French speakers of L2 English for five dimensions of oral performance (accentedness, comprehensibility, segmental accuracy, intonation, flow) using 1,000-point continuous scales. Immediately before rating, 20 listeners heard critical comments about Quebec French speakers’ English language skills, while 20 heard positive comments. Twenty listeners (baseline group) received no manipulation. Compared to baseline listeners, positively oriented listeners (younger and older) rated four of five dimensions more favorably. However, listeners’ behavior diverged under negative bias. Compared to age-matched baseline listeners, younger listeners upgraded speakers while older listeners downgraded speakers for all targeted measures. Findings cast doubt on the relative stability of L2 speech ratings and point to the importance of social context and generational differences in untrained rater assessments of L2 speaking performance.

Distinguishing Dick from Jane: Children's voices are more difficult to identify than adults' voices

Previous research on talker recognition has focused on the processing of talker identity in adult voices; however, little is known about how listeners identify child voices. This study examined adult listeners’ ability to differentiate and identify child voices (2.5-year-olds) and adult voices. In Experiment 1, native English-speaking listeners completed an AX voice discrimination task with both child and adult voices. Results revealed that listeners were significantly worse at discriminating between child voices relative to adult voices, regardless of the fact that child voices were mixed gender and adult voices were all female. In Experiment 2, we examined whether listeners could learn to identify child and adult voices. Adult listeners completed an adaptive training paradigm, where they learned to identify 4 child voices on one day and 4 adult voices on a separate day. Preliminary results indicate that with training, listeners can learn to identify voices from both age groups above chance. However, listeners were still faster at learning and more accurate at identifying adult relative to child voices. Taken together, these findings are an initial step towards understanding how talker recognition is influenced by the acoustic-phonetic characteristics and articulatory capabilities that vary with talker age.

English phoneme recognition of vocoded speech by adult Mandarin-speaking English-learners

The present study examined the joint influence of listeners’ language experience and the degree of spectral degradation of speech signals on English phoneme recognition of L2 listeners. The participants included 27 native English-speaking listeners and 43 native Mandarin-speaking listeners who used English as an L2. The L2 participants varied in chronological age, age of English learning, length of residency in the United States, and the amount of daily-based English usage. The speech stimuli included 12 English vowels embedded in a /hVd/ context produced by four speakers and 20 English consonants embedded in a /Ca/ context produced by two speakers. The speech stimuli were processed using 2-, 4-, 6-, 8-, and 12-channel noise vocoders. The processed and original stimuli were presented to the listeners for identification in a random order. The results showed that spectral degradation had more adverse effects on the L2 listeners but the L2 disadvantage became more evident as the number of frequency increased. The L2 listeners showed different confusion patterns from the L1 listeners, which was affected by the L2 listeners’ native language experience. Furthermore, the regression analysis revealed that the L2 listener’s length of residency in the United States was a significant predictor for their phoneme recognition outcomes.

The determining factor of L2 vowel perception: The establishment of L2 categories

The L2 speech learning theories argue that the perception of L2 phonemes is mainly influenced by the relationship between and among L1 and L2 phonemes. In the present study, we hypothesized that L2 vowel identification is primarily determined by the solidity of the L2 vowel category establishment and the perceptual distance between L2 vowel categories. A group of 31 native Mandarin Chinese-speaking listeners and a control group of 9 native English-speaking listeners participated in the study. All listeners were asked to identify natural English/Chinese vowel identification and categorize and rate synthetic English/Chinese vowels. Major findings include: 1) Chinese learners’ English vowel perception was associated with their capacity to establish English vowel category; 2) the perceptual distance between English vowel contrasts played a key role in Chinese learners’ English vowel perception, rather than the distance between the Chinese vowel and English counterpart. Overall, results of this study suggest that the individual variability of Chinese-native listeners’ English vowel perception is interpreted by the establishment of English vowel categories. [Work supported by China Natural Science Foundation 31628009.]

Segmental versus prosodic predictors of perceived accentedness

Previous work has identified both segmental and prosodic influences on the perceived foreign accentedness of non-native speakers, but their relative contributions remain poorly understood, in part because they are often investigated independently. In this study, both segmental and prosodic properties of word-length stimuli were assessed. Native English-speaking listeners rated the accentedness of trochaic English words produced by talkers from four language backgrounds: Hindi, Korean, Mandarin, and Spanish. Each token’s deviation from native English productions was quantified for two segmental properties (VOT and vowel quality) and three prosodic properties (ratios of duration, intensity, and f0 across the two syllables). For each language background, a linear mixed-effects regression model was created to predict accentedness ratings from the phonetic deviations, and the significance of each fixed effect was examined. In each model, the significant predictors included both segmental and prosodic propertie...

Clear speech benefits for perception of center-only and edge-only syllables

Clear speech is a speaking style that has been shown to enhance intelligibility in noise; however, the underlying reasons for this enhancement are less well understood. One hypothesis is that listeners require less acoustic information to identify syllables spoken in clear than in conversational speech, allowing them to make use of briefer dips in fluctuating noise. The present study tests this hypothesis by presenting six /bVd/ syllables (“bead, bid, bed, bayed, bad,” and “bod”), produced in clear and conversational speech styles, to 20 monolingual native English-speaking listeners in a six-alternative forced-choice task. These syllables were modified to present varying portions of the syllable to listeners (20, 40, 60, or 80 ms of the syllable preserved). First, center-only syllables were created, in which acoustic information around the vowel midpoint was preserved. Second, edge-only syllables were created, in which information from the vowel center was silenced and formant transitions preserved, with ...

English vowel identification and vowel formant discrimination by native Mandarin Chinese- and native English-speaking listeners: The effect of vowel duration dependence

The purpose of this study was to examine the relationship between English vowel identification and English vowel formant discrimination for native Mandarin Chinese- and native English-speaking listeners. The identification of 12 English vowels was measured with the duration cue preserved or removed. The thresholds of vowel formant discrimination on the F2 of two English vowels,/Λ/and/i/, were also estimated using an adaptive-tracking procedure. Native Mandarin Chinese-speaking listeners showed significantly higher thresholds of vowel formant discrimination and lower identification scores than native English-speaking listeners. The duration effect on English vowel identification was similar between native Mandarin Chinese- and native English-speaking listeners. Moreover, regardless of listeners' language background, vowel identification was significantly correlated with vowel formant discrimination for the listeners who were less dependent on duration cues, whereas the correlation between vowel identification and vowel formant discrimination was not significant for the listeners who were highly dependent on duration cues. This study revealed individual variability in using multiple acoustic cues to identify English vowels for both native and non-native listeners.

Continuing research endeavor in room acoustic effects on speech communication and psychoacoustics

I received the best student paper award in architectural acoustics at the San Francisco ASA meeting in 2013. I presented the first half of my dissertation project on designing acoustics for linguistically diverse classrooms by investigating the effect of room acoustics on speech comprehension by native and non-native English-speaking listeners. In the following year, I continued to complete my dissertation and received my PhD in architectural engineering from the University of Nebraska-Lincoln. Beginning in 2015, I am a postdoctoral fellow at RWTH Aachen University in Germany as part of the iCARE (icareitn.eu) research program funded by a Marie-Curie Initial Training Network (ITN) grant from the European Union. My role is to help improve an existing electronic system in extending its application of virtual acoustics testing on children with hearing impairment. Upon receiving the best student paper award, I continue to enjoy multidisciplinary research that combines architectural acoustics, speech communica...

How room acoustics impact speech comprehension by listeners with varying English proficiency levels

The authors previously reported the preliminary results from an investigation on effects of reverberation and noise on speech comprehension by native and non-native English-speaking listeners at ICA, Montreal. The results showed significant main effects of reverberation time (from 0.4 to 1.2 s) and background noise level (three settings of RC-30, 40, and 50). Non-native listeners performed significantly worse than natives on speech comprehension in general. Furthermore, the negative effect of reverberation was more detrimental for non-native than for native English-speaking listeners. However, the preliminary analyses have not yet accounted for effects of English proficiency on speech comprehension in addition to the room acoustic environments tested. All test participants were screened for three measures of English proficiency (listening span, oral comprehension, and verbal abilities). Non-native listeners as a group scored lower on all three proficiency measures than native English-speaking listeners. In this paper, English proficiency is investigated as confounding factor affecting speech comprehension performance alongside noise and reverberation. The results help to further the understanding of how room acoustics impacts speech comprehension by listeners, native and non-native English-speaking, with varying English proficiency levels. [Work supported by a UNL Durham School Seed Grant and the Paul S. Veneklasen Research Foundation.]

Room acoustic effects on speech comprehension by English-as-a-second-language versus native English-speaking listeners

English-as-a-second-language (ESL) listeners are generally more impaired than native English-speaking listeners on speech intelligibility tasks under reverberation and noise; however, more study is needed to ascertain these groups’ performances on speech comprehension tasks, rather than intelligibility. A recent study (Valente et al, 2012) showed that speech comprehension by both native English-speaking adults and children was more negatively affected by adverse acoustic environments than sentence recognition. The current project investigates the speech comprehension of both ESL and native English-speaking adult listeners under combinations of reverberation and noise. Sets of 15-minute long listening comprehension tests based on the format of the Test of English for International Communication (TOEIC) were developed, anechoically recorded, convolved with binaural room impulse responses (BRIRs) to produce five mid-frequency reverberation times (0.4 to 1.1 seconds), and presented over loudspeakers in the Nebraska acoustics chamber. Background noise was also varied at three levels (RC-30, 40 and 50) through an Armstrong i-Ceiling system. In total then, subjects were individually exposed to 15 acoustic combinations of reverberation and background noise. Preliminary results will be presented and compared between ESL and native English listeners. [Work supported by a UNL Durham School Seed Grant and the Paul S. Veneklasen Research Foundation.]

Lexical-neighborhood and competing-task effects on word recognition and word recall by native and non-native listeners

In spoken-word recognition, high-frequency words with few and less frequently occurring minimal-pair neighbors (lexically easy words) are recognized more accurately than low-frequency words with many and more frequently occurring neighbors (lexically hard words). This easy-hard word effect has been found to be larger for non-native listeners with a relatively late age of immersion in an English-speaking environment. Previous research found no effect of a competing digit-recall task on spoken-word recognition and no effect of listener group or listening environment on digit-recall accuracy. The present study compares word recognition by native English-speaking listeners and non-native listeners with either earlier (age 10 or earlier) or later (age 14 or later) ages of immersion in an English-speaking environment. Spoken word lists composed of equal numbers of lexically easy and hard target words were presented in an open-set word-identification task. Spoken words were presented in quiet and in moderate background noise and preceded by a list of zero, three, or six visually presented words, which listeners were asked to recall following the spoken word-recognition task (competing task). The size of the easy-hard word effect, spoken-word recognition accuracy competing-task word-recall accuracy, and word-entry response time will be compared across listener groups and listening conditions.

Conversational and clear speech intelligibility of /bVd/ syllables produced by native and non-native English speakers

The ability of native and non-native speakers to enhance intelligibility of target vowels by speaking clearly was compared across three talker groups: monolingual English speakers and native Spanish speakers with either an earlier or a later age of immersion in an English-speaking environment. Talkers produced the target syllables "bead, bid, bayed, bed, bad" and "bod" in 'conversational' and clear speech styles. The stimuli were presented to native English-speaking listeners in multi-talker babble with signal-to-noise ratios of -8 dB for the monolingual and early learners and -4 dB for the later learners. The monolinguals and early learners of English showed a similar average clear speech benefit, and the early learners showed equal or greater intelligibility than monolinguals for most target vowels. The 4-dB difference in signal-to-noise ratio yielded approximately equal average intelligibility for the monolinguals and later learners. The average clear speech benefit was smallest for the later learners, and a significant clear speech decrement was obtained for the target syllable "bid." These results suggest that later learners of English as a second language may be less able than monolinguals to accommodate listeners in noisy environments, due to a reduced ability to improve intelligibility by speaking more clearly.

Intelligibility, Comprehensibility, and Accentedness of L2 Speech: The Role of Listener Experience and Semantic Context

Abstract: This study investigated how listener experience (extent of previous exposure to non-native speech) and semantic context (degree and type of semantic information available) influence measures of intelligibility, comprehensibility, and accentedness of non-native (L2) speech. Participants were 24 native English-speaking listeners, half experienced and half inexperienced with L2 speech, who transcribed and rated 90 English utterances spoken by six English and six Mandarin speakers. The utterances varied along two dimensions: real-world expectations (true vs. false utterances) and semantic meaningfulness (meaningful vs. meaningless utterances). Listeners with more experience understood more speech from the L1 and L2 speakers than listeners with less experience but did not rate it differently in comprehensibility and accentedness. All listeners understood and rated the utterances from L2 speakers based on the semantic context available: true–false utterances were understood and rated best, meaningless utterances least. These findings have implications for evaluating learner pronunciation and for training learners in successful L2 communication strategies.

The Production of English Vowels by Fluent Early and Late Italian-English Bilinguals

The primary aim of this study was to determine if fluent early bilinguals who are highly experienced in their second language (L2) can produce L2 vowels in a way that is indistinguishable from native speakers’ vowels. The subjects were native speakers of Italian who began learning English when they immigrated to Canada as children or adults (‘early’ vs. ‘late’ bilinguals). The early bilinguals were subdivided into groups differing in amount of continued L1 use (early-low vs. early-high). In experiment 1, native English-speaking listeners rated 11 English vowels for goodness. As expected, the late bilinguals’ vowels received significantly lower ratings than the early bilinguals’ vowels did. Some of the early-high subjects’ vowels received lower ratings than vowels spoken by a group of native English (NE) speakers, whereas none of the early-low subjects’ vowels differed from the NE subjects’ vowels. Most of the observed differences between the NE and early-high groups were for vowels spoken in a nonword condition. The results of experiment 2 suggested that some of these errors were due to the influence of orthography.

CHANGES OVER TIME IN GLOBAL FOREIGN ACCENT AND LIQUID IDENTIFIABILITY AND ACCURACY

This study assessed global foreign accent in sentences and the production of two English consonants, /[upside-down r]/ and /l/, by 11 Japanese college students during their freshman and senior years (T1, T2). In Experiment 1, native English-speaking listeners rated five sentences spoken by the Japanese speakers and five native English control speakers. Experiments 2 and 3 examined 25 word onsets containing /[upside-down r]/ and /l/. Auditory evaluations by native English-speaking listeners were used to determine: (a) to what extent the consonants produced could be identified as intended at T1 and T2; and (b) whether /[upside-down r]/ and /l/ were produced more accurately at T2 than at T1. The results provided little support for a markedness hypothesis based on statistical frequencies and mixed support for a hypothesis based on perception studies. Some speakers made significant improvement, however, in both global foreign accent and liquid identifiability and accuracy.

Effects of experience on non-native speakers' production and perception of English vowels

This study assessed the effect of English-language experience on non-native speakers' production and perception of English vowels. Twenty speakers each of German, Spanish, Mandarin, and Korean, as well as a control group of 10 native English (NE) speakers, participated. The non-native subjects, who were first exposed intensively to English when they arrived in the United States (mean age=25years), were assigned to relatively experienced or inexperienced subgroups based on their length of residence in the US (M=7.3vs. 0.7years). The 90 subjects' accuracy in producing English /iIε æ/ was assessed by having native English-speaking listeners attempt to identify which vowels had been spoken, and through acoustic measurements. The same subjects also identified the vowels in syntheticbeat-bit(/i/-/I/) andbat-bet(/æ/-/ε/) continua. The experienced non-native subjects produced and perceived English vowels more accurately than did the relatively inexperienced non-native subjects. The non-native subjects' degrees of accuracy in producing and perceiving English vowels were related. Finally, both production and perception accuracy varied as a function of native language (L1) background in a way that appeared to depend on the perceived relation between English vowels and vowels in the L1 inventory.