Reflective teaching and classroom research: a case study

The second form of research drawn on here is observation and reflection on the author's own classroom practices and the data her courses produce. Its aim is to improve teaching and learning with a more detailed needs analysis, reflecting students' specific academic purposes, and detailed feedback on oral performance.

The research draws on two main sources of data and results. The first of these is course-related. As a teacher, the author always takes personal notes in class when participants introduce each other, writes down observations after class, and jots down suggestions or concerns while negotiating with individuals, pairs or small groups about cooperation and presentation topics. Extensive notes support later adaptation of a new course to what learner needs and profiles appear, as was the case for two C1 Academic English courses drawn on here. Students' oral presentations in class are also occasions for teacher's written notes, on the basis of which detailed face-to-face feedback is given. Written peer feedback is requested from the audience, but goes to the presenters themselves and is shared only on a voluntary basis. Audio recordings of the presentations and subsequent discussions are also provided so that recordings of co-presentations can be listened to together or individually and passages that require attention be focused on. Students can also request a copy of their audio recording for their own purposes.

Data in terms of notes from oral presentations derive from critical moments for the teacher, who is also an ELF listener. They influence feedback to students, and they also result in teacher learning about the dynamic nature of intelligibility in ELF and the role of familiarity. Students can, however, exercise their learner agency through access to their own recordings, becoming aware of what strengths to play on and what weaknesses to attend to.

The second source is data from an ongoing process of reflective pronunciation teaching over many years involving individual work in one-to-one situations coupled with collaborative annotation and note taking, in which students are helped to realise their agency as oral users of ELF.

From the first source, two oral co-presentations are discussed below, each one involving an individual, one Chinese and one Russian, whose oral contributions in class were extremely difficult to understand for the teacher, an ELF listener. This judgment was formed through instances in which not enough words were recognisable from a stretch of speech for the listener to be able to understand more than the general topic before the speaker's intention became clear(er) through (dis)confirmation of the next turn, a form of implicit other-initiated, coconstructed repair (cf. Smit 2010: 211ff.). Its purpose is comparable to Mauranen's (2012: 225) "echoing in search of form", which "is done by somewhat fuzzy matching, as if imperceptibly doing corrective or clarificatory language work". The teacher, as an ELF participant, uses identified or uncertain words from the learner's utterance to trial a partial 'candidate' confirming and developing the topic, thus beginning what might reflect the learner's intention. Then a pause and a quizzical face from the teacher signal that her hypothesis is incomplete and invite other-continuation/help from the speaker. Class participants also cooperate by supplying something they recognised. However, class cooperation happened less frequently in relation to oral contributions by the Chinese and the Russian participants and the teacher sensed a lack of joint achievement.

The Chinese law student was paired off with a Francophone Swiss, while the newly arrived Russian computer science student joined a 'similect' Slavonic group with a Polish and a Czech student. The Russian participant generally spoke at such a speed that it was virtually impossible to develop a hypothesis about what she was saying quickly enough to be able to process what was coming next in order to confirm or disconfirm the original hypothesis. Judging from the help the Polish student was able to offer occasionally, it appeared that the Slavonic-language speaker found her more intelligible. Interestingly, the Russian student took offence when asked to slow down and refused to accept that she was difficult to understand. The analysis below reflects the author's attempt as an ELF listener and as an EAP teacher to come to terms with not understanding an assertively fluent and rhotic course participant.

The respective joint presentations in which the Chinese and the Russian participant were involved reflected the efforts they had made and the way the two speakers had used preparation advice. The Chinese student still had difficult pronunciation, but he and his Swiss partner had produced excellent slides with picture illustrations and textual explanations that included many of the words they felt were difficult to pronounce/understand. Their talk was clearly structured and sign-posting was frequent. They had also rehearsed their co-presentation and they added interactivity by taking turns repeatedly during the presentation, each time recapitulating some of what the other had just said. The peer feedback confirmed the overall fluency, interactivity and topic- relevance of the slides, but almost half the respondents still found the speakers hard to understand, one commenting that pronunciation was not clear. Half of the remaining respondents mentioned two other factors as weaknesses: "watched too much on the screen and not enough to the audience", "standing sideways to the audience ... it would be better to stand face-to-face to the audience".

The peer feedback received by the three Slavonic co-presenters is not on record. They spoke on a topic that clearly interested the audience (lycanthropy) but they sounded under-rehearsed to the author or as if performing at the limit of their capacity, using many words, including proper names, that they had perhaps never pronounced in English, e.g. 'psike for psyche (the Czech student) or Zeus as 'servus with an apical r instead of zju:s (the Polish student) or 'gari for the first name of Harry Potter (the Russian student). The Slavonic threesome's colourful slides were audibly appreciated and they had lifted expressions and whole sentences from a written source. The Russian student, who spoke second, referred to a previous presentation to establish a connection. She was reading from notes but frequently looked up at the audience. She had also slowed down enough for the teacher to take some notes, here complemented by notes from the recording, which revealed that pronunciation was not the only factor in making her difficult to understand. Table 5.1 shows pronunciation variations co-occurring in the same presentation.

While invariable id for -ed endings often causes no intelligibility problem, occasional id became confusing in combination with nonstandard features in

Table 5.1 Pronunciation variation during oral presentation

stress placement and/or use of prepositions, for example, when the intended word opposed was pronounced hposit and so was heard as opposite, and only the recording revealed that she had said opposed of this theory. This holds more generally: unexpected realisations of words may have been recognised had the listener not been distracted by nonstandard variations in article use, some interchangeable uses of verbs and nouns and intermittent absence of copular be. The most unfamiliar feature for this author was the student's hyper- rhoticity (Hickey 2014: 150): using rhotic post-vocalic r where there is none (etymologically/in the spelling); for example, wDrs was heard as wars instead of was, bi'kors was heard as becourse instead of because, 'orBor as orthor instead of author. The teacher had not encountered this before and had not recognised either the words or the phenomenon, a feature of L1 English varieties under pressure from General American (cf. Kramer 2012), or, in terms of stylising, indexing white American speech (cf. Cutler 2014). The teacher was confused as an ELF listener by something unexpected, non-transparent and unilaterally persisting (cf. Seidlhofer 2011: 134-135). For the presenter, however, indexing American pronunciation may have been a safe positioning. Whether it actually was a 'BTA' or a reflection of her ambition to earn one cannot be ascertained, but it was likely her agentive way of coping with the high-stakes situation. Preparing her presentation in a Slavonic-language 'similect' threesome, and co-presenting in it, may have activated more Slavonic features than had she been able to converge with speakers with a different multilingual repertoire communicatively requiring her to adjust her pronunciation interlingually (cf. Jenkins 2000: 169ff., 214-215). It was high-stakes because the student's mid-semester written progress test had been the weakest of the group and the assessment of the presentation was important. Speaking speed was a signal of fluency, which, while not compensating for lack of accuracy, projected competence. However, she appeared reluctant to accommodate to the multilingual ELF context even in an EAP class, which she may have found as taxing as the trilingual context in which she was now studying computer science.