| This dissertation concerns the fundamental organizational problem of integrating fragmented knowledge in order to accomplish important tasks. A perspective on how knowledge is integrated across subunits in a multiunit company is developed and tested. Organization structure is conceptualized as a network of lateral relations among organization subunits, and search for and transfer of knowledge are seen to take place through this network.; This model is applied to the task of developing products in the least amount of time. Two arguments are made. First, the more central an R&D team is in the network with respect to subunits possessing relevant expertise, the easier the network search, and hence the shorter the completion time. Second, weak (distant and infrequent) network relations lead to a search-transfer problem: weak ties facilitate search by providing non-redundant contacts but impede transfer of complex knowledge (i.e., noncodified and dependent knowledge), which typically requires strong ties between the two parties to a transfer. Weak ties therefore slow down R&D projects when the transferred knowledge is very complex.; This perspective is tested in a network study of 41 divisions and 120 R&D projects in a large electronics company. Archival and survey data are used to construct the interunit network and project measures. Statistical analysis of project completion time in 120 projects supports the two main arguments. |
