Arizona teachers' speech: Phonological features and listener perceptions

While it has been shown that students' attitudes regarding teachers' speech (particularly accent) can affect comprehension of that teacher (Ahn & Moore, 2011; Rubin, 1992), less research has investigated the specific linguistic features that contribute to favorable or unfavorable perceptions of teaching suitability (Kang, 2012), especially in a high school context (Boyd, 2003). Likewise, there is another gap in the related research: While undergraduate students are often the evaluators of instructors' speech, other educational stakeholders have been generally overlooked. To address these weaknesses, this study explored three main areas. First, brief (1.5-2 minute) speech samples from ten currently certified Arizona teachers underwent an extensive phonological analysis in order to investigate linguistic features that differentiate native English speaking (NES) teachers from non-native English speaking (NNES) teachers with Spanish as a first language. These 10 speech samples were then evaluated impressionistically by 141 listeners (32 current teachers, 46 teacher candidates, 30 parents of middle or high school students, and 33 students in grades 8-11) on comprehensibility, accentedness, and perceived teaching suitability. The listener groups' evaluations of the 10 teachers' speech samples were examined as a whole and compared across listener groups. Last, the way in which certain phonological features predicted listeners' ratings on comprehensibility, accentedness, and perceived teaching suitability was examined. Results from parametric and non-parametric measures showed that NNES and NES teachers' speech samples were indistinguishable in many suprasegmental features, only being significantly different in percent rising tone choices. Despite this, listeners rated NNES teachers as significantly less comprehensible, more accented, and less suitable to teach than NES teachers. Further, there were significant differences among the listener groups in terms of ratings. Finally, many of the phonological features analyzed (on both NES and NNES teachers) were found to predict comprehensibility, accentedness, and perceived teaching suitability, but each of the three latter constructs was predicted by a different set of features. Because NES teachers' and NNES teachers' speech samples did not differ significantly on most phonological features analyzed, the findings suggest that listener evaluations of NNES teachers can be explained, in part, by biased perceptions and other speaker-irrelevant factors rather than objective phonological characteristics, a phenomenon that should be accounted for before any policies are enacted that would restrict the responsibilities of non-native English speakers.