Video-based Interaction Analysis is a powerful tool in the investigation of human activity that is particularly effective in complex, multi-actor, technology-mediated work settings and learning environments. It is currently undergoing a period of rapid development, driven, in part, by researchers' dissatisfaction with conventional methods, and in part by the ubiquity of video equipment.

In this paper, we have tried to take stock of the accumulating wisdom of one particular community of practice, a community that comes together over particular research problems encountered by its members. Our purpose has been to describe what Interaction Analysis looks like at a particular point in time, from a particular vantage point, i.e. the practices of a group of industrial and academic researchers who are struggling with the requirements of research in complex work and learning environments.

The stock of wisdom that has accumulated in this community is difficult to communicate, in part because it is evolving, so that what constitutes accepted practice today may be outmoded three months from now. It is also difficult to communicate because we have not yet developed a mutually agreed upon language for talking about what we do. Our attempt in this paper to delineate a set of "foci for analysis" within video-based Interaction Analysis is a first step in that direction.

In spite of these difficulties, we do believe that Interaction Analysis has come far enough at this point to warrant such a stock taking. Yet we also want to caution the reader that practitioners of Interaction Analysis vary widely in their approaches. Others may report practices that differ substantially from ours. Our hope, and indeed expectation, is that this account will provide the basis for others working in the field to modify, to add, to elaborate, and to question.

Twenty years ago, when Harvey Sacks started the first Interaction Analysis Laboratory in a smoke-filled, windowless room at the University of California at Irvine, it was absolutely unclear to what extent new ways of looking (and listening) had to be forged, or to what extent we could rely on the findings and analytic categories of conversation analysis to guide us in looking at video tapes. Since then we have found, as does every developing research tradition, that certain kinds of principles have emerged, certain kinds of problems have crystallized as important, and, as time moves on, it has become increasingly clearer which parts of the world are illuminated by the new approach and which are not.

As more and more technology emerges to facilitate this way of working, we expect this practice to flourish and take off. We also expect new issues and opportunities to arise. For example, one of the questions currently emerging asks to what extent success and productivity are directly dependent on the fact that it is a collaborative methodology, and specifically one that, in most of our experience, has capitalized on analysts' face-to-face interaction. The prolific generation of observations and hypotheses, the control of analyst bias, and the ability to draw on comparative materials from other tapes, all depend on the fact that we meet regularly and work together in physical co-presence. Recently, however, members have begun to contemplate the idea of doing this kind of analysis with colleagues who are temporally or spatially removed.

As in any lively field, issues of this sort are rampant and constitute the daily struggles of doing this kind of work. There is also, however, a stable base now, a sound foundation that has been accumulating over the years, that, hopefully, will inform subsequent work by ourselves and a widening circle of practitioners of Interaction Analysis.