Abstract

In virtue of its limitations to explicate human behaviour in complex and dynamically changing situations, the traditional symbolic approach has been repeatedly subjected to different challenges and critiques, albeit a number of past supporting empirical evidence is available. A new perspective, known as Situated Action (SA), has emerged in recent years. With its antisymbolic stance, the central focus of the situated action approach hinges on the interactions between people and the historically and culturally constituted contexts in which they are embedded. The major point of departure between the two camps of thought lies on the importance attached to internal symbolic representations. Unfortunately, there has been no consensus on the definition of situated action. SA proponents are a group of interdisciplinary researchers, including psychologists, anthropologists, philosophers, neurobiologists and computer scientists. They tend to emphasize those aspects relating to their specific expertise. The main arguments of the four representative views put forward by Suchman, Greeno, Clancey, and Lave are depicted and commented. Of all the theoretical issues so identified, the most concerning problem is that there is a lack of operationally defined and testable set of mechanisms for understanding and evaluating a SA's system behaviour. In spite of its drawbacks, SA approach has significant implications to various areas, and three of them are particularly noteworthy, namely,human-computer interaction, the role of planning, and transfer of learning (including the emergence of cognitive apprenticeship). Each of these implications is thoroughly discussed. Nevertheless, the prominent issue is not simply the replacement of the old, or the rejection of the novelty, but rather should be a dialectical synthesis by which a comprehensive unified theory can be generated