ABSTRACT
In the current study, we utilize the Yelp database as an arena to test how
population-level differences might relate to language use. While previous work
suggests online business reviews may provide insight into the psychological states
of its individual reviewers (Jurafsky, Chahuneau, Routledge, & Smith, 2014), we
expect that structural differences in one’s social community as a whole, where
language is crucial to conveying ideas, will affect language use. We focus on
how a language user’s social niche influences the amount and rate of information
transferred across a message. Agent-based simulations (Chater et al., 2006; Dale &
Luypan, 2012; Reali et al., 2014) and recent studies on the influences of interaction
in social networks (Bond et al., 2012; Choi, Blumen, Congleton, & Rajaram,
2014) indicate that the structure of language use may be influenced by structural
aspects of a language user’s social interactions. From an exploratory standpoint, we
aim to determine if one’s social-network structure predicts the probability structure
of language use.