Social Connection Induces Cultural Contraction: Evidence from Hyperbolic Embeddings of Social and Semantic Networks

Abstract

Research has repeatedly demonstrated the influence of social connection and communication on convergence in cultural tastes, opinions and ideas. Here we review recent studies and consider the implications of social connection on cultural, epistemological and ideological contraction, then formalize these intuitions within the language of information theory. To systematically examine connectivity and cultural diversity, we introduce new methods of manifold learning to map both social networks and topic combinations into comparable, two-dimensional hyperbolic spaces or Poincaré disks, which represent both hierarchy and diversity within a system. On a Poincaré disk, radius from center traces the position of an actor in a social hierarchy or an idea in a cultural hierarchy. The angle of the disk required to inscribe connected actors or ideas captures their diversity. Using this method in the epistemic culture of 21st Century physics, we empirically demonstrate that denser university collaborations systematically contract the space of topics discussed and ideas investigated more than shared topics drive collaboration, despite the extreme commitments academic physicists make to research programs over the course of their careers. Dense connections unleash flows of communication that contract otherwise fragmented semantic spaces into convergent hubs or polarized clusters. We theorize the dynamic interplay between structural expansion and cultural contraction and explore how this introduces an essential tension between the enjoyment and protection of difference.

References in corpus (14)

• Emergence of scaling in random networks
• Efficient Estimation of Word Representations in Vector Space
• Distributed Representations of Words and Phrases and their Compositionality
• Semantics derived automatically from language corpora contain human-like biases
• Quantifying social group evolution
• Self-similarity of complex networks
• Man is to Computer Programmer as Woman is to Homemaker? Debiasing Word Embeddings
• Popularity versus Similarity in Growing Networks
• Publication Bias (The "File-Drawer Problem") in Scientific Inference
• The Spontaneous Emergence of Conventions: An Experimental Study of Cultural Evolution
• Network Mapping by Replaying Hyperbolic Growth
• Superlinear Scaling for Innovation in Cities
• Neural Embeddings of Graphs in Hyperbolic Space
• Centralized "big science" communities more likely generate non-replicable results