Paul DiMaggio Amir Goldberg Princeton University
Systems of classification, such as those that distinguish between people of different ethnicities, lifestyles or occupations, are central to how social order is sustained. While sociologists have thoroughly investigated how divisions reified by social categories have concrete consequences for the unequal distribution of resources, a central question remains largely unanswered: how do social categories come about? This study attempts to address this question by comparatively examining how categories emerge through social interaction in two online communities: a music social network website and a financial investment community. The online data used in this study, which record the entire set of activities of tens of thousands of individuals over a period of almost one year, provide a vantage point that is rare in social scientific research: the ability to trace, with a microscopic lens, the dynamic co-evolution of social interaction and individual behavior. Using innovative data mining techniques, this study aims to investigate the extent to which and the ways in which, by distinguishing between different types of musical pieces or financial assets, people learn from one another to associate certain objects as making up various categories, which carry different social meanings.
This project's potential scholarly impact is twofold. Substantively, it has the potential of demonstrating that the two inherently different domains of finance and music, one seemingly motivated by calculative rationality, the other by emotion, are shaped by similar rudimentary social mechanisms. Methodologically, it introduces a set of novel techniques as tools for modeling and analyzing complex dynamic social processes that will be of use for future researchers in harnessing the vast repositories of data made available by the internet revolution for social scientific research.
Broader Impacts: Project results will be well-publicized within sociology and, more broadly, through publication in general-interest scientific journals and participation in interdisciplinary meetings and conferences devoted to web-based research and the study of complex systems. Concrete payoffs of two kinds are anticipated. First, the Internet has been touted as a way to overcome disadvantages resulting from network homophily for underrepresented groups to participate in economic and social activities. This study may cast light on the extent to which members of underrepresented groups are integrated into and appear to benefit from participation in the networks that the websites create; and may help us develop ways to enhance participation more effectively. Second, the data used in this study fully overlap with the first ten months of the 2008 financial crisis. Insights drawn from this study may be useful for understanding better the social dynamics that fueled the most severe economic recession since the Great Depression.
Perhaps the most vexing problem associated with the study of culture is the question of meaning: how different cultures assign different meanings to the same realities. Meaning is highly complex. Things do not have meanings in and of themselves; rather, they become meaningful only insofar as they are associated with one another. Studying cultural emergence and evolution therefore necessitates studying the complex processes through which ideas, behaviors and symbols become associated with, or dissociated from one another. This study examines culture through the prism of classification: how different objects become collectively classified as belonging to the same category of meaning. It examines two domains of social activity that are conventionally seen as antithetical to one another – music and financial investment – and demonstrates that in both arenas categories emerge and change through interpersonal interaction. Though this has been suggested by social scientists in the past, it has never been demonstrated in a systematic way before. The study’s major findings are threefold: The more people interact in closed cliques, the more they perceive the world through clearly delineated categories. This happens because people coordinate their categorical perceptions with those they interact with most frequently. Social status has an effect on whether people are influential in changing the categorical perception of others. It is particularly consequential in domains – like music – where there is no consensus on how to ascertain quality and performance. Though highly theoretical, this study is grounded in concrete empirical evidence and extremely large volumes of data gathered from two online social networks. It demonstrates that processes of culture diffusion are far more complex than the simple spread of fads, and, perhaps more importantly, that irrespective of the information being exchanged, who we exchange it with and how these interaction partners behave has an impact on how we perceive the world and consequently act.
