Reviewed by
Srebrenka Letina
Institute for Analytical Sociology, Linköping University
Centola’s book is about a powerful and thought-provoking – yet simple – idea about the effectiveness of wide (strong) versus long (weak) ties for the diffusion of complex behaviors. Given the spread of corona virus (simple contagion) and the way people react to it at the moment (complex contagion), this already important topic could not be more well-timed. As the title correctly suggests, this book is relevant for researchers working in computational social science, social networks, behavior change, and all adjacent fields (e.g. sociology, psychology, epidemiology) – which covers quite a large and heterogeneous audience. The fact that the book is written in a language accessible to a wider audience (there are no formulas and no required pre-existing knowledge about computational methods) makes it of relevance beyond the academic ivory tower. This may have been a very ambitious attempt, but Centola managed to deliver the storyline of his research about complex contagions, which is rather slow-paced at some points for a more informed reader, but still able to keep the reader’s attention. The understanding is further facilitated by carefully placed elegant figures throughout the book that both help in understanding and highlight the most essential messages.
In the first part of the book (“Theory”), Centola builds the story of his research around two central points: a) complex behaviors spread differently than simple contagion; and b) Granovetter’s idea about the most efficient network structure for spreading information does not hold true for complex contagions. He starts by cleverly putting the reader in the right mindset by using the example of perceptual illusions that remain even after we become aware of them. Then he proceeds by slowly building the case from the intuitive (mis)understandings, early theories in social science to research about simple contagion, case studies, and new approaches that use web-based experiments and simulations. Clearly, the book is heavily based on methods of computational social science and large-scale studies, but it also shows the appreciation of theoretical work and early studies in social science.
The second part will be of interest to more practically oriented readers, as the author turns to implications of his research in health interventions, the diffusion of innovations, and diffusing change in organization. The third part about social design begins with very profound and even more reaching implications of research on complex contagions – due to technological advances we live at times when social world is changing drastically compared to social worlds of the previous generations (the world is becoming “smaller, the ratio of weak ties to strong ties in people’s social networks will increase”, p.137), resulting in fewer complex contagions that are essential for the cultural transmission. However, instead of dwelling on its possible negative repercussions, Centola takes a more optimistic and proactive approach by showcasing the research which focuses on the mechanisms that make complex contagion possible, and therefore can be used to facilitate it (e.g. empathy based on homophily in context-relevant traits). In the spirit of the book, the author uses the results of his cutting-edge research about inefficiency of health interventions based only on social support to make a strong cautionary statement about how our intuitive ideas can lead as astray. In Conclusion, the lessons of the book are reiterated, the main takeaway being that wide bridges will be more effective than brokers for diffusion of the behavior that requires an effort or has a cost. However, when the goal is to design and apply a successful diffusion, Centola emphasizes that specific social context needs to be considered, and offers some concrete directions for future research.
Finally, as an encore, the book ends with an epilogue about experimental sociology and three appendices (The Ethics of Social Design, Methods of Computational Science, and Technical Appendix for Models, respectively) that can easily be used as short stand-alone introductory texts for the beginners in the field of computational social science. As the author clearly illustrates, we are not yet at the stage when the diffusion of behavior is completely understood, and therefore this is not a cookbook about how to make a diffusion efficient. Nevertheless, it will be a valuable resource for practitioners, as well as researchers with diverse professional profile because it provides firm guidelines and not less importantly, many cautionary warnings. In fact, considering its powerful message and writing style, it is comparable to some bestsellers with somewhat related topic, such as Christakis (2009), or Barabási (2003), or the recent books by same authors (Christakis 2019; Barabási 2018), respectively. However, due to its very didactical and focused approach that also includes questions and paradigm for future research, I see it as most inspiring and especially useful for the beginners in the field (graduate students in network science, computational social science, etc.). In difference to textbook-like materials (e.g., Salganik 2019), this book sets a highly accessible example how to build carefully on the previous theories, and research done with traditional methods, and then proceeds to test new ideas with the use of the full potential of “new” methodologies (e.g. web-based experiments, simulations). As one learns from the book, given the clustered nature of most academic networks, I have no doubts that the message of the book will diffuse successfully.
References
BARABÁSI, A. L. (2018). The Formula: The Universal Laws of Success. Hachette UK.
CHRISTAKIS, N. A., & Fowler, J. H. (2009). Connected: The Surprising Power of Our Social Networks and How They Shape Our Lives. Little: Brown Spark.
CHRISTAKIS, N. A. (2019). Blueprint: The Evolutionary Origins of a Good Society. Hachette UK.
Return to Contents
of this issue