What (Exactly) Is Novelty in Networks? Unpacking the Vision Advantages of Brokers, Bridges, and Weak Ties
区分了多样性、非冗余性和独特性三种新颖性，利用企业邮件面板数据检验它们如何与网络结构结合创造视野优势，发现强弱关系在不同新颖性维度上各有贡献。
The strength of weak ties and brokerage theory both rely on the argument that weak bridging ties deliver novel information to create “vision advantages” for actors in brokerage positions. However, our conceptualization of novelty is itself fundamentally underdeveloped. We, therefore, develop a theory of how three distinct types of novelty—diversity, non-redundancy, and uniqueness—combine with network structure to create vision advantages in social networks. We test this theory using panel data on an evolving corporate email network. Three main results emerge from our analysis. First, we confirm the diversity-bandwidth tradeoff (DBT) at the heart of the vision advantage. As brokers’ networks become more diverse, their channel bandwidth contracts, creating countervailing effects on access to novel information. Second, we uncover a mechanism driving the DBT, which helps explain differences in vision advantages across strong and weak ties. Strong, cohesive ties deliver greater information diversity and non-redundancy, whereas weak bridging ties contribute the most unique information (the information that is most different from what other contacts deliver). Third, we find network diversity (in contrast to network constraint) to be positively associated with longitudinal entropy, a measure of the accumulation of novel information over time. This suggests that weak bridging ties, which provide the most unique information through low bandwidth, structurally diverse channels, contribute the most to one’s aggregation of novel information over time. Collectively, these results take a step toward resolving a long-standing debate in network theory about whether strong, cohesive networks or weak bridging networks contribute more to vision advantages. This work firmly establishes that it depends. This paper was accepted by Kartik Hosanagar, information systems. Supplemental Material: The code and data files are available at https://doi.org/10.1287/mnsc.2022.4377 .