Content area
Full text
Narrative is especially relevant to the analysis of organizational processes because people do not simply tell stories-they enact them. Narrative data have surface features that are useful for description, but explanatory process theories must be based on deeper structures that are not directly observable. To address this problem and to facilitate better process theory, in this article I use concepts from narrative theory to create a framework for analyzing structural features in narrative data.
Explanation is essential to theory and practice. If we see an organization doing well, we want to reproduce the success; if we see one doing poorly, we want to prevent failure. Either way, we need a theory-an explanation of what is causing the observed outcomes. As Sutton and Staw argue, explanation is the core of good theory; it is "the answer to queries of why" (1995: 378). In this article I adopt what DiMaggio (1995) calls the "theory as narrative" view: an explanation is a story that describes the process, or sequence of events, that connects cause and effect (see also Einhorn & Hogarth, 1986). In this view, good stories are central to building better theory.
The quest for "better stories" touches on a theme raised in an earlier AMR special forum on theory building, where Eisenhardt (1989, 1991) and Dyer and Wilkins (1991) articulated competing views on the apparent tradeoff between "better constructs" and "better stories." Dyer and Wilkins (1991) criticized Eisenhardt's (1989) approach to building theory with multiple case studies as sacrificing context and deep structure in a sense, ruining the story-in favor of building constructs that could be used in traditional variance models. Eisenhardt (1991) countered that better stories and better constructs are not mutually exclusive; if journal page limitations were less strict, one could have both.
In this article I argue that one can indeed have better constructs and better stories, because in the domain of process theory stories are constructs. More precisely, stories help explain the relationships between events in a process or a narrative (Bal, 1985; Barthes, 1977; Chatman, 1978; Rimmon-Kenan, 1983). In the same way that a survey contains indicators for the underlying constructs in a variance theory, narrative text contains indicators for an underlying process theory.
In process theory typical patterns of...