 This study provides a longitudinal empirical examination of the basic elements of Nonaka's (1994) dynamic theory of organizational knowledge creation. First, the data illustrate the notion that knowledge creation in organizations proceeds through an intertwined four-phase process: (1) socialization (tacit knowledge amplification); (2) externalization (tacit knowledge is transformed into explicit knowledge); (3) combination (explicit knowledge amplification); and (4) internalization (explicit knowledge is transformed into tacit knowledge). Second, the study extends Nonaka's theory by comparing the relative amount of intra-organizational knowledge transfer occurring during periods of product redesign with the amount of knowledge transfer occurring during steady-state periods. The questionnaire data suggest that the overall level of knowledge transfer is higher during periods of product redesign than it is during the steady state, whereas the interview data indicate that there were more mentions of knowledge transfer during the steady state. Third, the data suggest that there may be benefit in adding tacit error correction as a fifth phase in the learning cycle. This phase is characterized by a dual emphasis on externalization and internalization. Implications of these findings are discussed.

Many organizations are made up of other organizations that have decided to act collectively as with research and development consortia, industrial alliances, trade associations, and formal political coalitions. These collective organizations can be characterized by their differing strategies: some are general in scope, while others specialize on a more narrow purpose. What explains the prevalence of generalism and specialism among collective organizations? We develop an ecological model in which collective organizations compete over member organizations. Assuming that an organization joins a collective when its objectives match that of the collective, our model predicts a generalism bias in the ecology of founding and growth among collective organizations. This outcome is predicted to be path dependent, however, emerging over time according to relatively minor differences in initial conditions. These predictions are supported in an analysis of founding and growth rates among US R&D consortia, and the model helps to account for the numbers, sizes, and strategic diversity of these consortia.

No abstract