A variety of alternative orientations to psychological inquiry have arisen in response to attacks on mechanistic and reductionistic assumptions about human activity. Despite their differences, these alternative viewpoints seem to be converging on a set of common themes that can be subsumed within a world view known as contextualism. This paper examines the basic assumptions and principles of a contextualist perspective, highlighting its emphasis on the unity, plurality, spontaneity and ecological dependency of human activity. We advance an essentially pluralistic and relativistic thesis in exploring the theoretical and methodological implications of a contextualist socio-psychological science.' " Knowing" ', claims William McGuire (1983), 'involves the inherent tragedy that[the] necessary representations of the known are necessarily misrepresentations : one cannot survive without doing it [i.e. using representations] but one cannot do it very well'. The inadequacy of knowledge, he suggests, stems from the nature of the human cognitive system. As the environment's richness of information often exceeds one's capacity to process it directly, one's representation of what is 'out there' is reduced and distorted to fit in with one's available cognitive categories. One needs to know, that is to represent a thing-in-itself in a reduced and simplified form, just so that one can cope with that richness and complexity.The abstract representation of knowledge in theory, McGuire (1983) argues, offers even less cause for optimism. By its very generalized and abstract nature, theory runs the risk of dangerously oversimplifying and distorting the events and experiences it is said to represent. Theory represents each situation only vaguely and often becomes overgeneralized to situations beyond its scope. In this perspective on theory and knowledge, the knower, particularly the theorist, is likened to 'a ropewalker who, on arriving at a precipice of ignorance, ties one end of a chain of inferences to a stake at its brink, and, flinging the free end as far as possible out over the abyss, runs quickly along the thrown chain to get the maximum distance before plunging to disaster ' (McGuire, 1983, p. 3).This image of knowledge as inherently tragic can, we think, be understood, not