Individual Submission Summary
Share...

Direct link:

Cultural Bridges and Broken Chains: Cross-Cultural Connectivity among Knowledge Producers in 1000-1899CE

Sat, August 8, 2:00 to 3:00pm, TBA

Abstract

Human innovation relies on transmitting knowledge through social networks that span space and time. Yet how cross-cultural connectivity among knowledge producers evolved historically, and what disrupted it, remains poorly understood. We focus on cultural bridges, defined as network ties between individuals from different cultural groups, and examine how they varied across domains, relationship types, and centuries.

We construct a multiplex social network of 120,828 globally notable knowledge producers active between 1000 and 1899 CE, drawing on a cross-verified biographical database linked through Wikipedia references. Using an LLM-assisted pipeline, we classify 871,623 ties into relationship types such as influence, collaboration, mentorship, and rivalry based on biographical textual evidence. We further integrate publication records from the Universal Short Title Catalogue to examine whether social localization extended to knowledge content.

We find that cultural bridges vary systematically across domains in ways that reflect the social organization of knowledge production. Music, organized around court patronage, exhibits the highest bridging rates, while visual arts, organized through guild apprenticeships, shows surprisingly low rates despite being non-linguistic. Intellectual ties such as influence and rivalry are the most cross-cultural, while personal ties such as family and friendship remain the most local. Academic cultural bridges declined over centuries, counter to standard narratives of increasing intellectual globalization, while bridges in other cultural domains remained stable or grew. This divergence is linked to the decline of Latin as a scholarly lingua franca. Latin speakers consistently maintained higher cross-cultural connectivity, and their disappearance from academia accounts for a substantial share of the overall decline. Book-pair similarity analysis confirms that intellectual content nationalized in parallel, and that Latin attenuated this localization.

This sheds light on how social interactions among knowledge producers are regulated by the institutional and linguistic infrastructures in which they are embedded, and how shifts in those infrastructures can selectively reshape some fields while leaving others intact. By combining large-scale historical network analysis with computational text methods, we demonstrate how new data sources and analytical tools can bring theoretical claims about the social structure of intellectual life under systematic empirical test.

Authors