Reformulations in Multi-Party interactions in English as a Foreign Language in a Turkish Higher Education Context
Özet
Reformulations are “forms of talk which attribute to some prior speaker words and/or
ideas purported to have been authored or implied in some prior talk” (Gonzales
1996, p. 158) and are effective devices for describing, explaining, re-stating or
summarising the prior talk (Waring, 2002). Although they have been investigated in
a variety of contexts such as psychotherapy (e.g. Davis, 1986; Perakyla &
Vehvilainen, 2003; Kurri & Wahlstrom, 2007), radio call-in programmes, (e.g.
Hutchby, 1996; Drew, 2003), seminar discussions (e.g. Waring 2002), and other
educational contexts (e.g. Kapellidi, 2015; Hauser, 2006), the interactional function
of reformulation in L2 learner-learner interaction remains a gap in L2 literature.
Keeping this gap in mind, this study aims to document the characteristics of
reformulation and learners’ orientation to reformulation in an out-of-classroom group
discussion task in English as a Foreign Language (EFL). Track the phenomenon in
multi-party L2 interaction in a Turkish higher education context, this study will
investigate the phenomenon through the robust methodological underpinnings of
ethnomethodological Conversation Analysis. To provide a micro-analytic
investigation of sequential unfolding of reformulation in multi-party L2 interaction, 20
extracts based on a collection of 84 episodes, which were analysed in alignment
with the data-driven nature of conversation analytic research methodology, were
involved in the study. The audio data comes from discussion tasks designed as outof-classroom activities for students taking the Oral Communication Skills 1 and 2
classes at Hacettepe University Division of English Language Teaching. The data
was collected by Assist. Prof. Dr. Olcay Sert and labelled as L2 Discussion Tasks
Corpus (L2DISCO, Sert, 2016; 2017a; b). Participants were all non-native speakers
of English except one who was born in Australia, a bilingual Turkish-English
speaker. The dataset includes the participants’ audio recorded conversations at 6
different times over 1 year and consists of a total of 174 multi-party conversations
(average 20 minutes each), which amounts to 58 hours of audio recording. For this
study, a sub-corpus of 60 recordings (approximately 1200 minutes) were analysed.
The analysis shows that the sequential position of reformulation constitutes three
viii
categories: adjacency pairs (Sacks, 1967; Schegloff, 1968), independent Turn
Constructional Unit (TCU) in the second turn and triadic sequence. Additionally, the
analysis illustrates that learners employ reformulation to index a variety of actions
(i.e. agreement, confirmation, recipiency) as they primarily demonstrate their
understanding. In addition, reformulation is produced to provide other-repair and
other-correction when a party signals a trouble which can be a word search, a codemixed utterance, a grammatical mistake or difficulty in formulating an appropriate
turn. The findings also suggest that learners’ orientations to reformulation bring
evidence for their novice and expert roles displayed through reformulations.
Reformulations expose learners to the linguistic varieties in a target language and
they create their own learning by providing opportunity for a potential uptake of
reformulated utterances. Thus, the findings of this study also have some
implications for reformulation practices of learners as they can create their learning
opportunities through reformulations.
