How Communities Mediate the Diffusion of Science: The Case of Granovetters Weak Ties Hypothesis

How do scientists adopt new ideas? This question was high on the agenda of science studies in the 1960s and 1970s. Using primarily survey methods to study the spread of scientific ideas, researchers discovered key dynamics of scientific diffusion. They found, amongst others, that the diffusion of a scientific idea bears similarities with the diffusion of other types of innovation, in that the growth fol- lows an S-shaped curve, just like the growth curve of the adoption of innovations [4, 13]. These studies further uncovered the central role of informal communities, sometimes called invisible colleges [4] or coherent groups [8], in the organization of scientific research. Such communities foster vocabularies and narratives through which researchers interpret scientific findings [5, 3]. Although it opened up a new field of research, science studies of the 1960s and 1970s faced limitations in terms of available data and methods. The explosive development in the availability of data and sophisticated analytical techniques have reinvigorated the field of science studies since the 2000s, allowing researchers to study the development of science at scale [6,12,14]. However, such computational analyses tend to focus on the structural properties of scientific networks while leaving the study of the interpretative work that scholarly communities do to ethnographers. As a consequence, our knowledge of how communities adopt and diffuse ideas is fragmented and limited. This study uses newly available digital datasets and computational techniques to revisit this important long-standing topic. We use citation network analysis, topic modelling and close reading, to study the role that communities play in interpreting scientific ideas as they spread within the academic field. Our case study is the diffusion of Granovetters Strength of Weak Ties hypothesis, first presented in 1973 in American Journal of Sociology. We map the community structure of the citation network through which the hypothesis diffuses and examine how different communities develop specific interpretations of Granovetters hypothesis. Our methodological contribution is to develop an approach that bridges the gap between the computational analysis of network properties and the interpretative analysis of meaning (see also [7]). Our substantive contribution is to show that communities reinterpret and adapt ideas as they adopt them