The structure and quality of classroom interaction and the ways in which children learn how to interact appropriately was the initial focus of the research and field work described in this paper. The site was a kindergarten/first grade class in a suburban Boston Title I elementary school with many students from second and third generation Italian-American families. At the outset the primary research tools used were collection and analysis of videotaped classroom activities and behavior and discussion sessions involving the teacher and the researcher. A participant observation component was later added to the project, and the researcher then visited the classroom on a weekly basis. This change in research technique indicated a new direction for the project--that of making the teacher a co-researcher. Since the research was being done in a context of discovery rather than proof, the researcher and the teacher together were able to focus on isolating, describing, and discovering the dynamics of the classroom environment (how it functions, what is predictable, what kinds of breakdowns occur). It was suggested that teachers could share in this discovery process in the absence of a trained researcher and become sources of idea exchange and dialogue for one another. This method might be applied on a larger scale in schools where there is a need for inservice training and continuing education courses designed to help teachers share experiences and raise individual self-esteem and awareness. (MM)