Why do certain domains of human knowledge advance faster than others? This thesis sets out to answer this question by elaborating upon the notion that new ideas spring from hybridizations of existing ideas. Over time, such hybridizations form a path-dependent network structure of interconnected and interconnecting knowledge domains. This thesis develops a network-analytic approach to model this structure and to investigate its effects. Based on all technological knowledge patented in the United States between 1975 and 1999, this thesis demonstrates that the network position a knowledge domain is embedded in at a given point in time affects its subsequent growth prospects in important and predictable ways.