people construct for themselves and each other as part of their lives in society is the enterprise that we ( Laboratory of Comparative Human Cognition, 1979) have called elsewhere an "ethnographic psychology of cognition".
Finally, our analysis of thinking in test, club, and classroom has, somewhat unexpectedly, highlighted one hallmark of modern cognitive psychology, the active organism. It has done so in an unusual manner: by contrasting behavior across settings and positing the within-setting factors that control between-setting variation. In attempting to understand Archie's behavior in the club setting we were forced to consider the many ways in which he actively modifies his problem environment. Once alerted to the pervasiveness of such behaviors in the club environment, we were encouraged to reexamine the experimental and classroom contexts. They were there to be seen, but only upon rather close observation.
This asymmetrical situation, in which the experiment highlights one form of interaction and the club another, has far reaching consequences for cognitive psychology in its attempt to characterize the active nature of human cognitive functioning. Experiments the contexts constructed to make behavior analyzable, are also contexts that discourage the expression of active, adaptive behaviors. This point was made many years ago by one of the few genuine interactive cognitive theorists, L. S. Vygotsky ( 1978) who pointed out that:
All stimulus-response methods [which in this context refer to all standard experiments in the tradition following Wundt] share the inadequacy that Engels ascribes to naturalistic approaches in history. Both see the relation between human behavior and nature as unidirectionally reactive. My collaborators and I, however, believe that human behavior comes to have that "transforming reaction on nature" which Engels attributed to tools. We must, then, seek methods adequate to our conception [p. 61].
If our analysis in this chapter has a future, that future will have to include the invention of new experimental and analytic techniques that permit a principled study of interaction in addition to reaction.
Early phases of this research were conducted collaboratively with L. Hood, R. P. McDermott and Carla Seal to whom we are indebted for their invaluable insights. Conduct of the research and preparation of the manuscript were made possible by a Grant from the Carnegie Corporation.