Raudaskoski1990
Raudaskoski1990 | |
---|---|
BibType | INCOLLECTION |
Key | Raudaskoski1990 |
Author(s) | Pirkko Raudaskoski |
Title | Repair Work in Human-Computer Interaction: A Conversation Analytic Perspective |
Editor(s) | Paul Luff, Nigel G. Gilbert, David Frohlich |
Tag(s) | EMCA, Artificial Intelligence, Repair, HCI, Conversation Analysis, AI Reference List |
Publisher | Academic Press |
Year | 1990 |
Language | English |
City | London |
Month | |
Journal | |
Volume | |
Number | |
Pages | 151-171 |
URL | Link |
DOI | 10.1016/B978-0-08-050264-9.50012-9 |
ISBN | |
Organization | |
Institution | |
School | |
Type | |
Edition | |
Series | |
Howpublished | |
Book title | Computers and Conversation |
Chapter |
Abstract
This chapter discusses repair in different areas of the artificial intelligence (AI) approach to human–computer interaction (HCI) and the approach taken in conversation analysis (CA). The repair is seen as unwanted as the smooth progress of the interaction has been disturbed and communication has broken down. This is the view that is taken by most AI and HCI researchers for repair work. CA is an interpretative approach to studying conversational phenomena with no predefined model for conversation. In CA, repair is essential for successful human communication and is considered as one of the resources for managing meaning in conversation. Within this perspective, repair becomes a condition, not a hindrance, for a successful conversation. Thus, CA researchers concentrate on the patterns of the conversation sequences which handle misunderstanding while HCI researchers are more concerned about the cause of miscommunication, looking at natural conversations to discover types of trouble sources. If conversation involves achieving understanding then repair work is an essential part of the apparatus, enabling the participants to check their interpretations or correct the other person's interpretations. The importance of evaluation to the design of human–computer interfaces is acknowledged by HCI researchers. It is necessary to study not only how people communicate between themselves, but also the specific nature of human–computer communication and the pressures of communication. Systems have to be designed in cycles where evaluation and re-implementation are as important to the final artifact as the original design.
Notes