This paper contributes to the current debate about the “knowledge sharing-protecting” tension in R&D collaborations by introducing a new angle for analyzing this paradox. Considering the role of internal knowledge sources of the firm, the paper introduces and empirically analyses the interplay between three knowledge governance mechanisms (selective, contingent, and orchestrated openness) and the different technological fields of the firm (core, related non-core, and distant non-core technologies). Based on a unique first-hand large-scale project-level dataset from a Global 100 firm across a 10-year span, this research finds that R&D collaborations selectively conducted in firm’s related non-core technological fields promise the best innovation performance. Further, by orchestrating its projects across core and non-core technological fields in the same knowledge portfolio, the firm can leverage a network of inter-connected projects when collaborating with external partners. As such, the firm can strategically “distribute” collaboration risks across multiple projects, while optimize its access to external knowledge that it intends to get. Finally, contrary to the common assumption, this study does not find any discriminating factors against market-based partners in their contribution to innovation in both core and non-core technologies.

本文通过引入一个新的角度来分析这种悖论，为当前有关研发合作中“知识共享-保护”张力的辩论做出了贡献。考虑到企业内部知识源的作用，本文介绍并实证分析了三种知识治理机制（选择性，偶然性和有组织的开放性）与企业不同技术领域（核心，相关非核心和开放性）之间的相互作用。遥远的非核心技术）。根据来自全球100强公司10年跨度的独特的第一手大规模项目级数据集，这项研究发现，研发合作是在公司的有选择地进行的。相关的非核心技术领域有望实现最佳创新绩效。此外，通过与同一知识组合中的核心和非核心技术领域协调项目，该公司在与外部合作伙伴协作时可以利用相互关联的项目网络。这样，该公司可以在多个项目中策略性地“分布”协作风险，同时优化对计划获得的外部知识的访问。最后，与通常的假设相反，本研究没有发现基于市场的合作伙伴对核心和非核心技术创新做出贡献的任何歧视性因素。