Anne O'Keeffe

Introducing pragmatics in use

1.Introduction 2.Researching Pragmatics 3.Deixis 4.Gricean Maxims in Context 5.Speech Acts in Context 6.Politeness in Context 7. Cross-cultural Pragmatics 8.Pragmatics in Specific Discourse Domains 9.Pragmatics and English Language Teaching 10.Coda: Looking to the Future

2. RESEARCH IN THE TEACHING OF SPEAKING

This chapter reviews research and practice in six main areas relevant to the teaching of speaking: (1) the growing influence of spoken corpora, (2) the debates concerning native speaker (NS) and nonnative speaker (NNS) models for spoken pedagogy,(3) the issue of authenticity in spoken materials, (4) approaches to understanding speaking in the classroom, (5) the selection of texts and aspects of spoken language for the teaching of speaking, and (6) developments in materials and methods for the teaching of speaking. Spoken corpora, whether NS corpora collected in “old” or “new” variety locations or NNS corpora based on learner data or expert/successful user data, have generated vigorous debate as to how spoken language should be modeled for teaching, and their influence is being seen in shifts in methodology toward language-awareness-based approaches as well as new materials based on lexicogrammatical and discoursal corpus evidence. Various approaches to understanding classroom speaking are also reviewed, including discourse analysis, conversation analysis, cognitive approaches, and the Vygotskian perspective. Applications of insights from these approaches are reviewed, especially how the approaches affect the selection of texts and language features to be taught. Finally, practical discussion on the teaching of specific spoken genres is reviewed and probable future directions are discussed.

What's in a name? - vocatives in casual conversations and radio phone-in calls

This paper looks at the use of vocatives across two corpora: the 5-million word Cambridge and Nottingham Corpus of Discourse in English (CANCODE) and a 55,000-word corpus of radio phone-in calls. 100 vocatives are sampled from the CANCODE corpus, using only informal, casual conversations among intimates, friends and close associates. All vocatives (n=232) were extracted from the radio data. The vocatives in both datasets were classified according to the contexts in which they occurred. The contexts were categorised under headings connected with topic and turn management, face concerns, general relational concerns, humour/badinage and summons. The distribution over the two datasets was compared, as well as the position of the vocative in the speaker turn. Overall, the CANCODE data revealed a preference for vocatives in relational, topic management, badinage and face-concerns, while the radio data revealed a tendency for vocatives to be used more in the management of phone calls, turn-taking, topic management and face concerns. The radio data showed a greater frequency for initial position, then final, while the casual conversation data was the reverse. Medial position was seen to be problematic in both datasets and an alternative analysis is proposed. In neither dataset did vocatives seem to be necessary except in a small number of cases. Overwhelmingly, the vocative serves pragmatic functions. Comparing linguistic features such as the vocative across datasets enhances the descriptive framework for spoken genres.

Second Language Speaking

Approaches to spoken language description have contributed to the understanding of second language speaking. Three theoretical frameworks have also provided insight. Language Identity looks at the impact an additional language has on an individuals identity. Language Socialization sees language as the symbolic means by which humans appropriate norms of verbal and nonverbal behavior. Sociocultural Theory draws on Vygotskys view of language acquisition as a sociocultural process linking the social/interactional with the cognitive. Speech acts research has also been important, but has generally used elicited data. Spoken corpora provide real data but raise issues concerning native and nonnative speaker status as models.