CONCLUSION
' w e began this project with a particular purpose--one that we saw as especially apt for a time when references to social contexts are increasing in frequency and diversity. We wanted to introduce a concept, to demonstrate several of the ways in which it is being thought about and pursued in research, and to point to some propositions that cut across the variety
We are well aware that the coverage has gaps: inevitable in a short volume with a strict page limit. (Every contributor concluded with a sense of important points foregone.) The temptation, then, in any final comment is to try to mention everything that has not yet been covered.
In the face of that impossibility> we shall end by noting a single new direction-one that flows from Michael Cole's commentary. He has pointed to the need for some integration of practice theories and activity theories. We see the need as well to bring together approaches that emphasize actions and approaches that emphasize meanings. These, too, lack integration. "Meanings" appear under the labels of belief systems, cultural models, folk theories, consensus models, social representations, explanatory styles, and the interpretation of practices (see, for example, DAndrade and Strauss, 1992; Duveen and Lloyd, 1990; Harkness and Super, in press; Modell, 1994).
Lave (1993) has suggested that these two broad approaches to contextualizing development differ both in their emphasis and in their history. The first concentrates on the nature of engagement with an activity; its tradition is likely to be activity theory. The second "focuses on the construction of the world in social interaction" (Lave, 1993, p. 17); its tradition is likely to be phenomenologcal social theory. The contrast is provocative, and we join Lave in urging attention to both these approaches and to the continuing analysis of their interconnections and of what each contributes to the overarching problem of contextualizing development.