1 Hakkarainen, K. (in press) Learning as interaction. In H. Hyötyniemi (Ed.), Feedback to Future: Pro- ceedings of the “Feedback” Symposium on Cybernetics of Finnish Society of Artificial Intelligence (Step 2000), 28 th of August, 2000, Helsinki University of Technology. Draft of LEARNING AS INTERACTION Kai Hakkarainen Department of Psychology, University of Helsinki Kai.hakkarainen@helsinki.fi Abstract. The purpose of the study was to examine the role of interaction in human learning and devel- opment at multiple levels. Cognitive growth is conceptualized in terms of interaction between cognitions (current and new information, informal and formal knowledge, crystallized and flu- id knowledge) within the human mind. Sociocultural theories, further, examine cognition in terms of interaction between mind and cultural knowledge objects and artifacts. Many re- searchers also argued that an interaction between several agents within social communities is the main mechanism of cognitive growth. Accordingly, learning is conceptualized as a process of participating in social communities rather than one of knowledge acquisition. It is concluded that cognition can neither be reduced to individual mental processes nor social practices, but takes place through dynamic interaction between these two. Keywords: dynamic cognition, interaction, distributed cognition, sociocultural theory, periph- eral participation 1. Introduction The purpose of the present paper is to examine the role of interaction in human learning and the development of expertise. The paper addresses interactive processes important for human development at multiple levels. In this article, the concept of interaction is loosely used to refer to processes which are dialectical or reciprocal in nature so that the development cannot be ex- plained by referring to a single source or driving force; interaction between several partially independent factors has to be taken into consideration. There are always several mutually de- pendent factors that jointly control human development and their specific configuration and the interaction between them appear dynamically to change their effects. The following discussion should, however, be taken only as a very tentative effort at conceptualizing the very complex issues involved. The interactive view of human cognition does not, however, have a very strong background in history of psychology; it has been much more common to construct mechanistic or reductionist explanations of human behavior that are in accordance of unexamined everyday accounts of cognition. The traditional psychometric approach to human intelligent activity, for instance, explained human intelligent activity by relying on individual characteristics. Accordingly, in- dividual mental abilities are assumed to be fixed and genetically determined and used to ex- plain how well an actor succeeds in one or another task (Olson, in press). The problem is that even if these kinds of explanation may help us to make sense of the everyday world, they do