APDSI 2000 Full Paper (July, 2000) An Empirical Research on a State of Knowledge and Knowledge Acquisition Behaviors in Organizations Kenji Hirata 1) 3) , Ryo Asaba 1) , Takashi Kusumi 2) , Hiromitsu Muta 3) 1) The Sanno Institute of Management, HRM Research Center (HIRATA_Kenji@hj.sanno.ac.jp) 2) Kyoto University, Graduate School of Education 3) Tokyo Institute of Technology, Graduate School of Decision Science & Technology Abstract Recently, both in academic and business domains, "Knowledge" or/and "Learning" in organizations have been focused on. A lot of psychologist or sociologist have suggested that researcher and practitioner should assume that knowledge is dynamic, is not static when they study about knowledge. However in Management researches, knowledge has been tended to consider it static, such as concepts of "Intellectual Asset" or "Intellectual Capital". And also there are few empirical studies about knowledge in organizations. We attempted to survey from the point of view that was knowledge as a serial utilized activity in real work context, not as an amount of storage. 308 Japanese companies attended. In this paper, we report about knowledge acquisition behaviors in problem solving, and a state of important knowledge in organizations. Former, it was found that workers accessed and used some information sources properly according to situation in order to solve problems. Workers were going to acquire knowledge optimally; case by case. Later, we found that a state of organizational knowledge was not fully specified, especially the knowledge related to innovation was imperfect. Furthermore, as an issue became more and more complex, its resolution was not so smooth and needed more various kind of knowledge. Knowledge such "Change" and "Optimizing" especially needed to be actualized and communicated in Japanese companies. Finally, we will discuss about embodied the scheme, Learning of Workers, for managing knowledge. 1. Introduction In late years there is a growing tendency for companies to shift from stable hierarchical structure to flat type or taskforce teams centered structure, or to introduce organization formation in which both characteristics are coexisting each other. Some systems of new organization utilizing information technology is studied or applied with a change of organizational formation. In particular, in the side of human resource management, the systems such as Knowledge Management（Spencer 1995[ 1 ], Shum 1998[ 2 ]）, EPSS（Gery 1991,1995[ 3 ] [ 4 ]）, e-Learning and Web Based Training have lately attracted considerable attention. Each of these systems is distinguished in terms of differences of its theoretical background or the information and behavior that it targets at, however, their fundamental purposes are same. In other words, all these systems are able to provide appropriate information, which can be taken as knowledges  , through utilizing information technology to process a large quantity of information related to job for the purpose of performance. Moreover, they have something in common, where organization system is considered from the viewpoint of knowledge utilized in a process because it has been considered from the viewpoint of a task and a task process conventionally. On the other hand, in sociology and cognitive psychology, it is difficult to treat knowledge with information technology. For example, Hutchins（1990[ 5 ]） pointed out the following from a study of navigation operations of a ship. In various instruments utilized in navigation operations, there are many technologies remained unchanged though they are possible to be replaced with the latest information technology of computers or automated by. For example, such technologies are a compass, † There is confusion because interpretation of knowledge is different in some fields such as Business Administration, AI, and Psychology. In this study, "Knowledges" is considered as Knowledge that has the wide meaning utilized in Business Administration and Mgt. Information. The information that can be formalized as declarative and procedural knowledge in AI is expressed as Knowledge.