Organizational memory: Socio-technical framework and empirical research

The purpose of this research was to deepen our understanding of organizational memory, a concept related to issues in organizational learning, planned change, organizational communication, human expertise, and management information systems. The work addresses several questions: How is learning embedded socially and technically in organizations? What is the role of expert systems or culture? What function do social networks play in storing and making available organizational knowledge? How can expertise be mapped out in organizations? Which variables contribute to or thwart the development of organizational knowledge? Does past knowledge facilitate or impede organizational goals? The approach taken is socio-technical and constructive--the notion of organizational memory is viewed as a framework for understanding organizations as socio-technical information systems. Six parameters are proposed for describing organizational memory: (1) the contents of memory; (2) social and technical methods of storage; (3) processes for maintaining and developing organizational knowledge; (4) use of organizational knowledge; (5) the relationship between organizational structure and memory; and (6) the relationship between past patterns and current behaviors. These parameters are measured in two field settings by analyzing communication networks, personnel records, and data collected via interviews and questionnaires. The latter helped identify roles, training practices, routines, information characteristics of tasks, and changes in norms over time. Personnel records were analyzed to measure turnover. Network analysis was used to map information flows, retrieval rates, distributions of expertise, and to identify information-processing roles. This knowledge was used to characterize the socio-technical information-processing aspects of the field sites, and to flesh out the framework and methodology. One interesting finding was that members at both field sites retrieved information from peers more often than from information systems. Another finding was that retrieval rates and experience were positively correlated. Of methodological importance, network analysis was a valuable tool for mapping distributions of expertise. Overall, the work demonstrated the value of a memory information-processing perspective to organization theory. It concludes with reflection on the application of the framework to expert systems development and human resources planning, and with speculation on ideal memory design.