The purpose of this work is (1) to investigate the discourse structure of email discussions, and (2) to investigate the suitability of Conversation Analysis (CA) and Discourse Analysis (DA) for the analysis of email discussion data. The study uses naturally occurring email discussion data from publicly available archives. Radio and newspaper discussion data are used for companson. The discourse map is developed to represent mUlti-party multi-topic asynchronous discussions. Tum-taking is investigated using CA, and a set of rules proposed for tum-taking in email discussions: the main difference from spoken conversation is the absence of priority for speaker selection, such that any speaker can self-select at any time. Utterance pairs are also investigated using CA, focusing on openings and closings and on question/answer pairs. Openings and closings are varied and optional, and do not occur in pairs. Questions and answers often occur in paired structures. Strategies for linking utterances in the absence of adjacency are identified. To investigate longer sequences, selected whole discussion threads are analysed using the exchange rank taken from the DA framework. Email-specific features include multiple responses to one initiate, and single messages responding to more than one earlier message. Additionally, multiple initiates and/or responses can be found in a single email message. Repair in email discussions does not need to deal with troubles arising from real-time spoken interaction, but all four types of self- and other-initiated self- and otherrepair are found, as well as refusal to repair. Finally, the elements of a model for the investigation of the discourse of email discussions are presented and the methods of analysis evaluated. CA provides a flexible approach which leads to the illumination of major aspects of the data. The use ofDA presents problems, some of which are inherent in the framework, and some of which spring from the complexity of turns in email discussions. A major influence on the structure of email discussions is that they are managed locally and interactionally by the participants themselves.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:272143 |
| Date | January 2002 |
| Creators | Harrison, Sandra M. |
| Publisher | Birmingham City University |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
Page generated in 0.0016 seconds