No abstract

The current study investigated linguistic influences on comprehensibility (ease of understanding) and accentedness (linguistic nativelikeness) in second language (L2) learners' extemporaneous speech. Target materials included picture narratives from 40 native French speakers of English from different proficiency levels. The narratives were subsequently rated by 20 native speakers with or without linguistic and pedagogical experience for comprehensibility, accentedness, and 11 linguistic variables spanning the domains of phonology, lexis, grammar, and discourse structure. Results showed that comprehensibility was associated with several linguistic variables (vowel/consonant errors, word stress, fluency, lexis, grammar), whereas accentedness was chiefly linked to pronunciation (vowel/consonant errors, word stress). Native speaking listeners thus appear to pay particular attention to pronunciation, rather than lexis and grammar, to evaluate nativelikeness but tend to consider various sources of linguistic information in L2 speech in judging comprehensibility. The use of listener ratings (perceptual measures) in evaluating linguistic aspects of learner speech and their implications for language assessment and pedagogy are discussed.

This mixed-methods study examines the effects of rating scale length and rater experience on listeners' judgments of second language (L2) speech. Twenty experienced and 20 novice raters, who were randomly assigned to 5-point or 9-point rating scale conditions, judged speech samples of 38 newcomers to Canada on numerical rating scales for comprehensibility, accentedness, and fluency. Results yielded high Cronbach's alpha coefficients and no group differences for rating scale length or rater experience. However, Rasch category probability plots revealed that raters had difficulty differentiating between scale steps, particularly in mid-scale range-a challenge which was exacerbated in the 9-point scale condition. Evidence from verbal protocols and post-task interviews suggested that experienced and novice raters adopted strategies to either draw on or offset their perceived experience with L2 speech in conducting their ratings. Implications for L2 pronunciation research are discussed, including the need to draw on raters' perceptions to develop a greater understanding of major listener perceptual constructs in L2 pronunciation and to move beyond adherence to the native speaking standard in data collection procedures. L2 pronunciation conventions 3 Rater experience, rating scale length, and judgments of L2 pronunciation: Revisiting research conventions Rating scales provide the framework within which human raters score second language (L2) performance, which is taken to be an indicator of L2 learners' ability on the construct being measured. Such assessment schemes constrain, structure, or filter raters' responses, often through rigid scale descriptors associated with a fixed number of scale bands. Lumley describes the tension between the "simplified orderliness of the rating scale," which underrepresents the complexity involved in L2 performance, and raters' unconstrained reactions to the performance, which may be disordered and complex (2005, p. 248). The challenge for raters is to reconcile their possibly idiosyncratic, intuitive, or non-linear impressions of an L2 performance with rating scale specifications, including discretizing what might be a "grey area" performance. In L2 pronunciation research, 9-point numerical rating scales are increasingly becoming entrenched as a research convention to measure qualitative constructs of comprehensibility, accentedness, and fluency. 1 Nine levels are used due to the conclusion that, for accentedness at least, smaller scales might cause a ceiling effect (Southwood & Flege, 1999). Generally, only scalar endpoints are defined (e.g., very easy/difficult to understand for comprehensibility), with little guidance for raters on how to assess these dimensions other than to use the full range of the scale (Derwing, Munro, & Wiebe, 1998). A major appeal of these Likert-type scales to L2 pronunciation researchers is that their generic nature offers versatility, in that they can be used with learners from virtually any first language (L1) background or proficiency level on any task....

The current project aimed to investigate the potentially different linguistic correlates of comprehensibility (i.e., ease of understanding) and accentedness (i.e., linguistic nativelikeness) in adult second language (L2) learners' extemporaneous speech production. Timed picture descriptions from 120 beginner, intermediate, and advanced Japanese learners of English were analyzed using native speaker global judgments based on learners' comprehensibility and accentedness, and then submitted to segmental, prosodic, temporal, lexical, and grammatical analyses. Results showed that comprehensibility was related to all linguistic domains, and accentedness was strongly tied with pronunciation (specifically segmentals) rather than lexical and grammatical domains. In particular, linguistic correlates of L2 comprehensibility and accentedness were found to vary by learners' proficiency levels. In terms of comprehensibility, optimal rate of speech, appropriate and rich vocabulary use, and adequate and varied prosody were important for beginner to intermediate levels, whereas segmental accuracy, good prosody, and correct grammar featured strongly for intermediate to advanced levels. For accentedness, grammatical complexity was a feature of intermediate to high-level performance, whereas segmental and prosodic variables were essential to accentedness across all levels. These findings suggest that syllabi tailored to learners' proficiency level (beginner, intermediate, or advanced) and learning goal (comprehensibility or nativelike accent) would be advantageous for the teaching of L2 speaking.As many second language (L2) researchers have pointed out, it is crucial to set realistic goals for adult L2 learners, prioritizing understanding over © Cambridge University Press 2015 0142-7164/15

Link to this article: http://journals.cambridge.org/abstract_S1366728914000832How to cite this article: PAVEL TROFIMOVICH, TALIA ISAACS, SARA KENNEDY, KAZUYA SAITO and DUSTIN CROWTHER (2016 People are famously poor at judging their own ability, engaging in such behaviours as "errors of omission", "flawed self-assessment", and "faulty self-awareness"