Final Report Summary - L2 COMMUNICATION (Fostering cross-cultural communication: Identifying the linguistic factors that promote comprehensibility in academic and workplace settings) The overarching goal of the funded project is to identify the aspects of second language (L2) speech that are most important for achieving effective cross-cultural communication in workplace and academic settings. This is broadly accomplished by investigating both the systematic effects of listener background characteristics on listeners’ judgments of L2 speech, and the extent to which their perceptions of the speech align with the linguistic properties of learners’ L2 productions. The rationale for pursuing this line of research is that enhancing cross-cultural communication entails a greater understanding of the components of L2 speech that are most salient to listeners, and the extent to which the factors that they attend to contribute to communication breakdowns. In light of this overarching research program, the more specific overlapping research goals over the four-year term of the grant are to: (1) disentangle the overlapping dimensions of comprehensibility and accentedness, so that the degree to which a listener is comprehensible (or understandable) to the interlocutor can be modelled in L2 assessment instruments while leaving the notion of accent (or the degree of adherence to a native speaker standard) aside, since the consensus among applied linguists is that comprehensibility should be prioritized over accent reduction in L2 pedagogy and assessment.A major finding of this research priority is that accentedness is linked to pronunciation regardless of an L2 learner's first language (L1) background. Conversely, comprehensibility cuts across a much wider range of linguistic domains than accent, with pronunciation and lexical-grammar dimensions differentially contributing to L2 comprehensibility ratings depending on the speaker’s L2 background. (2) validate an empirically-derived L2 comprehensibility scale, originally developed in the Canadian context, for use in different English-medium instructional settings (e.g. the UK) by examining the generalizability of the criteria in the original scale for use with learners from different first language (L1) backgrounds. A major finding of this research priority is that raters reportedly attend to pronunciation and fluency to a greater extent than they do to accent and vocabulary when rating L2 English comprehensibility. A data-driven analytic scale was constructed to guide teachers’ ratings and help diagnose the linguistic factors that are most important to address for comprehensibility at different ability levels for international students on English-medium campuses. (3) examine linguistic and nonlinguistic (i.e. construct-relevant and construct-irrelevant) influences on listener (rater) judgments of L2 speech by investigating listener perceptions of speech quality in relation to more objective measures of learner speaking ability. A major finding of this research priority relates to individual differences in raters’ perceptions and the role of rater background characteristics (e.g. experience, familiarity with and attitudes toward the learner’s L1 accent), which are generally construct irrelevant sources of variance for the speaking ability being measured. Although raters may arrive at the same scoring decision when judging an extemporaneous speech sample using the same scoring rubric, they might do so for vastly different reasons. (4) examine different facets of learner performance on collaborative L2 speaking tasks, extending research on L2 speaking performance on monologic tasks (i.e. speaking into a recording device without the presence of a human interlocutor), which represents the majority of existing research, to different interactional patterns, including with an interviewer or examiner, and in learner-learner interactions. A major finding of this research priority is that parallel and collaborative interactional types appear to be the most prevalent on paired L2 speaking test tasks involving problem solving, with collaborative pairs reflecting most positively on their interactional experience. Conversely, asymmetric interactions tend to occur when there is a mismatch in paired test-takers’ L2 proficiency levels, with the dominant peer likely to be the least satisfied with the quality of the interaction. These research priorities have been achieved through both independent and collaborative empirical studies and state-of-the-art reviews designed to bridge the gap between research and practitioner communities that are unaware of substantive and methodological developments in each other’s areas. Taken together, these research priorities elucidate how communicative efficiency is mediated by listener perceptual and attitudinal variables in settings where the stakes for achieving successful oral communication are high. Ultimately, identifying the factors that are most important for achieving effective cross-cultural communication elucidates ways of reducing communication barriers and fostering the social integration of newcomers into society. Funding has enabled the development and launch of the Second Language Speech Lab (http://www.bristol.ac.uk/speech-lab/) at the University of Bristol, a state-of-the-art research facility designed to optimize the collection and analysis of high-quality L2 speech samples. The lab is a locus of research activities conducted in conjunction with Bristol’s Centre for Assessment and Evaluation Research in Education, which facilitates community building and assessment literacy training through seminars, outreach, and capacity-building activities. Together with spearheading dedicated graduate-level courses on teaching and assessing fluency and pronunciation in Bristol, the university is at the forefront of research and training in this area.
