Communication Modelling (LAP-2003), 1-2 July, 2003, Tilburg, The Netherlands
Conversational Analysis as a Theoretical Foundation for Language Action Approaches?
Göran Goldkuhl
VITS Research Network
Department of Computer and Information Science,
Linköping University and
Jönköping International Business School
Sweden
ggo@ida.liu.se
Abstract
Communication modelling approaches within the Language Action Perspective (LAP) are based on two important theoretical cornerstones: 1) considering language use as action and 2) organisation of communication in accordance with pre-defined patterns.
The basis for the first cornerstone can be found in speech act theory, but the basis for other must be searched for elsewhere. The paper investigates the origin for many LAP approaches, the Conversation-for-action schema of Winograd & Flores. It tries to detect an influence from the sociological approach of conversational analysis. Could conversational analysis be a basis for the study of communicative patterns within LAP?
A discussion is pursued concerning differences between and possibilities to combine speech act theory and conversational analysis within LAP. This discussion, which is supported by the use of a simple example of business interaction, concludes with a contestation - in a spirit of conversational analysis - of a too heavy use of pre-defined patterns in communication modelling.
1
Introduction
There are several approaches to business modelling and information systems modelling that build on the language action perspective (LAP). This perspective emphasises that communication is one kind of action. It objects to a restricted referential view on communication, i.e. to limit communication to a mere transfer of information. Speech act theory, as formulated by Austin (1962), Searle (1969) and Habermas (1984) has been a major source of inspiration for LAP approaches to business and information systems modelling.
Prominent approaches within this LAP tradition are Action Workflow (Medina-Mora et al,
1992) and DEMO (Dietz, 1999). There are other approaches as SAMPO (Auramäki et al,
1988) and BAT (Goldkuhl, 1996). The speech act basis is obvious in these different approaches. When using this LAP approaches different communicative actions are classified in accordance with established speech act typology, which is mainly the classification scheme by Searle (1979). This gives however not the whole picture. In the LAP approaches,
1
mentioned above, the interest goes beyond single speech acts. There is a great interest for speech act patterns, i.e. how different speech acts are related to each other. In Action
Workflow as well as DEMO there is a pattern of four sequentially organised speech act types.
In words of DEMO this pattern is: 1) request, 2) promise, 3) statement 4) acceptance (Dietz,
1999). Consequentially, one can say that the LAP approaches are built upon these two theoretical cornerstones:
1. Communication is action in accordance to generic speech act types
2. Communicative acts are organised and framed in accordance with pre-defined communicative patterns
For both cornerstones there is a use of pre-defined constructs. The first one is directly derived from speech cat theory (SAT). The other can however not be seen to be a part of
SAT. This theory has been criticized for being too monological (e.g. Linell, 1998). The communication pattern orientation must be searched for elsewhere.
Most LAP approaches has the origin from the seminal work “Computers and cognition: a new foundation for design” by Winograd & Flores (1986). In this book they introduced their
Conversation-for-action (Cfa) schema. The basis in speech act theory was here acknowledged, but besides this, they introduced this scheme as a communicative pattern of speech acts. This idea of communicative patterns has later being brought further in succedent LAP approaches.
In order to assess