**2.4 Effects of prior language teaching and tutoring experience**

Listeners' teaching and tutoring experience can be associated with the ratings of NNSs' speech. In Kang's [1] study, the undergraduate listeners who had the

experience of teaching and tutoring in languages tended to be less severe listeners. Such results are consistent with those of Hsieh [23]. In her study, the effects of listeners' teaching experience status on the ratings of ITAs' oral proficiency, comprehensibility, and accentedness are investigated. Hsieh reported that English as a second language (ESL) teachers were more lenient in their ratings of ITAs' oral proficiency, comprehensibility, and accentedness than American undergraduates with no teaching experience. However, in a more recent study, Kang et al. [3] did not find a statistically significant relationship between listeners' amount of teaching experience and their holistic ratings as well as their rating severity of L2 proficiency. Thus, although listeners who possess language teaching experience are assumed to be more lenient than their counterparts with no experience in teaching, this leniency is not consistent across the board.

#### **2.5 Non-native speaker linguistic stereotyping**

Prior research on US undergraduates' perceptions of NNSs' oral proficiency has established the validity of linguistic stereotyping [2]. As set forth by Lambert et al. [24], linguistic stereotyping is a phenomenon through which a short sample of L2 variety attributable to low-prestige groups arouses a cascade of demeaning evaluative judgments of speakers. They elicited NSs' perceptual judgments in regard to NNSs' productions according to a speech evaluation instrument developed by their research team. This pioneering study showed that NSs tended to pass negative judgment in relation to intellect, superiority, and friendliness when they heard accented speech. Rubin and Smith [25] postulated that one of the factors that moderated the proclivity of undergraduates' engagement in linguistic stereotyping was the amount of exposure they had with international instructors. This means that those undergraduates who have more cross-cultural experiences tend to be less critical and more tolerant of L2 accented varieties. In contrast, the undergraduates who harbor negative expectations toward speakers belonging to a particular social group would not be able to provide an impartial judgment of an L2 speaker's oral performance.

Another source of bias associated with listeners is reverse linguistic stereotyping (RLS), which is the converse of linguistic stereotyping [2]. In RLS, speakers' nonlinguistic cues can affect NS listeners' perceptions of the speaker. Such cues could include pictures of the speaker and knowledge about their ethnicity, race, gender, or social class. Thus, when listeners are misinformed that they are listening to an NNS, they tend to rate the speaker highly accented and less comprehensible [2, 26]. This problem is compounded if speakers represent members of stigmatized social groups and speak with a stigmatized accent, as it is the case with those NNSs who speak Spanish-accented English. The members of this community tend to be rated less favorably on the indices of status and solidarity by NSs [2]. In Kang et al. [3] study, RLS emerged as one of the listener-related variables which significantly predicted naïve listeners' deviation from expert scores, implying that those who had the propensity to become engaged in RLS were more likely to be influenced by subjective impressions and provide less accurate ratings. Overall, listeners' perception can be influenced by a variety of nonlinguistic or paralinguistic aspects of an utterance.

#### **2.6 Listener attitude**

The effects of listeners' expectations on the perception of L2 speech have been researched through listeners' attitudes in perceptual encoding of information. Using matched guise technique [25] and verbal guise technique [27], studies have revealed that listeners' attitude toward speakers' perceived social groups will not

**7**

*Listener Background in L2 Speech Evaluation DOI: http://dx.doi.org/10.5772/intechopen.89414*

language competence.

duced by speakers of the language.

instead of native for the given language" (p. 103).

only impact the listeners' ratings of the speakers' degree of accentedness but also their perceptions of the speakers "nonlinguistic characteristics," such as intelligence and language competence ([28], p. 570). Rubin [29] investigated students' perception toward non-native speakers. The guises (a picture of an Asian ITA and that of an American TA) were used to form participants' perceptions of the speech files and to mislead them in terms of the speaker's ethnicity. His finding reported that L1 English listeners had the proclivity to rate the speaker of the lecture more accented when the lecture was accompanied by the picture of an Asian woman than the same speaker of the lecture when it was accompanied by the picture of a Caucasian woman. Also, Brown's [30] study on American college students' attitude toward non-native instructors confirmed that listeners' preexisting knowledge of the speakers' country of origin significantly influenced their judgments of the speakers'

In addition to encoding a distorted perception of accent and language competence, the listeners who hold attitudinal biases toward a group of L2 speakers tend to have a less accurate perception of individual sounds in the speech signal transmitted by the L2 speakers. An illustrative example is the case of Cantonese L2 speakers of English who are stereotyped to have an unreleased word-final stop production. In this regard, Hu and Lindemann [31] investigated whether this existing stereotype would impair Cantonese perceptions of English word-final stops. To test their hypothesis, they selected a group of Cantonese participants and presented them with recordings of some sentences each one of which included an underlined keyword featuring a final stop. After each recording, the participants listened to three versions of the keyword that varied in the pronunciation of the final stop, and they were required to decide which one of the versions they listened to in the sentence. In one version, the final stop was totally unreleased, in another one the final stop had a released burst, and in the third version the released burst of the final stop was followed by an aspiration. The listeners who were told that the sentence was recorded by an American L1 speaker had the propensity to choose the aspirated version, and those who were told that the sentence was recorded by a Cantonese L1 speaker had the tendency to choose the unreleased version. The results of this study confirmed the hypothesis that the preconception notions held toward the speakers of a language have a direct bearing on how listeners perceive the phonemes pro-

An area which is comparably under-researched in the literature is the role of listeners' bias in perceived comprehensibility and intelligibility. Smith and Nelson [32] theoretically defined comprehensibility as understanding the overall message of an utterance and intelligibility as the accuracy with which the individual words are understood in an utterance. In a similar vein, Munro and Derwing [33] conceptualized comprehensibility as the relative ease with which an utterance is understood by the listeners, measured subjectively through scalar ratings, and intelligibility as word recognition, measured objectively through transcription tasks and cloze tests. Integral to the conceptualizations of the two speech constructs is the fact that they are regarded as the inherent characteristics of speech and the speaker is blamed for any breakdown in communication due to an incomprehensible speech. This misconception is reflected in van Wijngaarden's [34] argument stating that "the intelligibility of speech is known to be lower if the speaker is non-native

However, among the few scholars who pioneered appreciating the role of listener bias in the perception of speech comprehensibility and intelligibility is Munro [35] who stressed that the measurement of L2 speech is a function of three interrelated factors, namely, the characteristics of the speech per se, listener-related factors, and contextual factors. An active line of research has hitherto supported

#### *Listener Background in L2 Speech Evaluation DOI: http://dx.doi.org/10.5772/intechopen.89414*

*Metacognition in Learning*

ency is not consistent across the board.

**2.5 Non-native speaker linguistic stereotyping**

experience of teaching and tutoring in languages tended to be less severe listeners. Such results are consistent with those of Hsieh [23]. In her study, the effects of listeners' teaching experience status on the ratings of ITAs' oral proficiency, comprehensibility, and accentedness are investigated. Hsieh reported that English as a second language (ESL) teachers were more lenient in their ratings of ITAs' oral proficiency, comprehensibility, and accentedness than American undergraduates with no teaching experience. However, in a more recent study, Kang et al. [3] did not find a statistically significant relationship between listeners' amount of teaching experience and their holistic ratings as well as their rating severity of L2 proficiency. Thus, although listeners who possess language teaching experience are assumed to be more lenient than their counterparts with no experience in teaching, this leni-

Prior research on US undergraduates' perceptions of NNSs' oral proficiency has established the validity of linguistic stereotyping [2]. As set forth by Lambert et al. [24], linguistic stereotyping is a phenomenon through which a short sample of L2 variety attributable to low-prestige groups arouses a cascade of demeaning evaluative judgments of speakers. They elicited NSs' perceptual judgments in regard to NNSs' productions according to a speech evaluation instrument developed by their research team. This pioneering study showed that NSs tended to pass negative judgment in relation to intellect, superiority, and friendliness when they heard accented speech. Rubin and Smith [25] postulated that one of the factors that moderated the proclivity of undergraduates' engagement in linguistic stereotyping was the amount of exposure they had with international instructors. This means that those undergraduates who have more cross-cultural experiences tend to be less critical and more tolerant of L2 accented varieties. In contrast, the undergraduates who harbor negative expectations toward speakers belonging to a particular social group would not be able to provide an impartial judgment of an L2 speaker's oral performance. Another source of bias associated with listeners is reverse linguistic stereotyping (RLS), which is the converse of linguistic stereotyping [2]. In RLS, speakers' nonlinguistic cues can affect NS listeners' perceptions of the speaker. Such cues could include pictures of the speaker and knowledge about their ethnicity, race, gender, or social class. Thus, when listeners are misinformed that they are listening to an NNS, they tend to rate the speaker highly accented and less comprehensible [2, 26]. This problem is compounded if speakers represent members of stigmatized social groups and speak with a stigmatized accent, as it is the case with those NNSs who speak Spanish-accented English. The members of this community tend to be rated less favorably on the indices of status and solidarity by NSs [2]. In Kang et al. [3] study, RLS emerged as one of the listener-related variables which significantly predicted naïve listeners' deviation from expert scores, implying that those who had the propensity to become engaged in RLS were more likely to be influenced by subjective impressions and provide less accurate ratings. Overall, listeners' perception can be influenced by a variety of nonlinguistic or paralinguistic aspects of an

The effects of listeners' expectations on the perception of L2 speech have been researched through listeners' attitudes in perceptual encoding of information. Using matched guise technique [25] and verbal guise technique [27], studies have revealed that listeners' attitude toward speakers' perceived social groups will not

**6**

utterance.

**2.6 Listener attitude**

only impact the listeners' ratings of the speakers' degree of accentedness but also their perceptions of the speakers "nonlinguistic characteristics," such as intelligence and language competence ([28], p. 570). Rubin [29] investigated students' perception toward non-native speakers. The guises (a picture of an Asian ITA and that of an American TA) were used to form participants' perceptions of the speech files and to mislead them in terms of the speaker's ethnicity. His finding reported that L1 English listeners had the proclivity to rate the speaker of the lecture more accented when the lecture was accompanied by the picture of an Asian woman than the same speaker of the lecture when it was accompanied by the picture of a Caucasian woman. Also, Brown's [30] study on American college students' attitude toward non-native instructors confirmed that listeners' preexisting knowledge of the speakers' country of origin significantly influenced their judgments of the speakers' language competence.

In addition to encoding a distorted perception of accent and language competence, the listeners who hold attitudinal biases toward a group of L2 speakers tend to have a less accurate perception of individual sounds in the speech signal transmitted by the L2 speakers. An illustrative example is the case of Cantonese L2 speakers of English who are stereotyped to have an unreleased word-final stop production. In this regard, Hu and Lindemann [31] investigated whether this existing stereotype would impair Cantonese perceptions of English word-final stops. To test their hypothesis, they selected a group of Cantonese participants and presented them with recordings of some sentences each one of which included an underlined keyword featuring a final stop. After each recording, the participants listened to three versions of the keyword that varied in the pronunciation of the final stop, and they were required to decide which one of the versions they listened to in the sentence. In one version, the final stop was totally unreleased, in another one the final stop had a released burst, and in the third version the released burst of the final stop was followed by an aspiration. The listeners who were told that the sentence was recorded by an American L1 speaker had the propensity to choose the aspirated version, and those who were told that the sentence was recorded by a Cantonese L1 speaker had the tendency to choose the unreleased version. The results of this study confirmed the hypothesis that the preconception notions held toward the speakers of a language have a direct bearing on how listeners perceive the phonemes produced by speakers of the language.

An area which is comparably under-researched in the literature is the role of listeners' bias in perceived comprehensibility and intelligibility. Smith and Nelson [32] theoretically defined comprehensibility as understanding the overall message of an utterance and intelligibility as the accuracy with which the individual words are understood in an utterance. In a similar vein, Munro and Derwing [33] conceptualized comprehensibility as the relative ease with which an utterance is understood by the listeners, measured subjectively through scalar ratings, and intelligibility as word recognition, measured objectively through transcription tasks and cloze tests. Integral to the conceptualizations of the two speech constructs is the fact that they are regarded as the inherent characteristics of speech and the speaker is blamed for any breakdown in communication due to an incomprehensible speech. This misconception is reflected in van Wijngaarden's [34] argument stating that "the intelligibility of speech is known to be lower if the speaker is non-native instead of native for the given language" (p. 103).

However, among the few scholars who pioneered appreciating the role of listener bias in the perception of speech comprehensibility and intelligibility is Munro [35] who stressed that the measurement of L2 speech is a function of three interrelated factors, namely, the characteristics of the speech per se, listener-related factors, and contextual factors. An active line of research has hitherto supported

Munro's argument and lent credence to the role of listeners' attitudinal biases. For example, Kang and Rubin's [2] RLS study revealed that the listeners presented with the picture of an Asian scored lower in the listening comprehension test than those who were presented with the picture of a Caucasian. They attributed this phenomenon to the listeners' biased perception stemming from their prior experience such as their negative experience with international teaching assistants. In another study, Wolff [36] reported that in his experiment involving Nembe and Kalabari speakers, two languages of Eastern Niger delta, Nembe speakers reported to have found more linguistic similarity between their language and Kalabari and had a more complete perception of Kalabari speakers, whereas the opposite was reported by Kalabari speakers. Lindemann and Subtirelu [28] associated this discrepancy in perception with the "asymmetrical attitude" held by the listeners toward different language groups (p. 577). Therefore, it can be concluded that listeners' predispositions make significant contributions to what listeners perceive and the attitudinal biases held by listeners toward the speakers of a particular social group, especially a low prestigious one, would impair their perceptual encoding of the speech signal produced by the speakers from that stigmatized social group.
