Language-Style Similarity and Social Networks

Area of Science:

• Social network analysis
• Computational linguistics
• Social psychology

Background:

• Social network formation is influenced by various factors, including shared interests and proximity.
• Linguistic style, a reflection of individual cognition and social identity, has been less explored as a predictor of social ties.
• Understanding the interplay between language and social networks can reveal mechanisms of social cohesion and influence.

Purpose of the Study:

• To investigate whether linguistic similarity predicts the formation and persistence of social network ties.
• To examine whether social ties lead to linguistic convergence over time.
• To explore the coevolution of linguistic styles and social networks.

Main Methods:

• Analysis of linguistic styles using the Linguistic Inquiry and Word Count (LIWC) framework.
• Examination of social network structures in two distinct datasets: a student cohort and online Yelp reviewers.
• Longitudinal data collection across two waves to track changes in both linguistic styles and network ties.

Main Results:

• Linguistic similarity was a significant predictor of friendship formation and persistence in both studies.
• Friendship ties were associated with a subsequent convergence in linguistic style between individuals.
• The findings support a dynamic, coevolutionary relationship between language use and social network structure.

Conclusions:

• Linguistic style is a key factor in the formation and maintenance of social relationships.
• Social networks and linguistic patterns mutually influence each other, leading to dynamic changes.
• The coevolution contributes to the creation of 'relational echo chambers,' influencing information flow and social dynamics.