The massive adoption of social network technologies makes it urgent to develop a deeper understanding of how they affect the formation of opinions and ideologies. Ideas are formed through the interaction with social neighbours, and it is known that the social network topology plays an important role in this process. Moreover, opinions are not formed in isolation but are affected by the evolution of related ideas, and the complexity and feedback introduced by this co-evolution introduces effects that have not been studied in detail.
In this study we analyse the joint evolution of an opinion vector, called a “worldview”, by developing a model that extends the well-known Tangled Nature model of evolutionary ecology . In this model, agents update specific opinions by estimating their affinity to their neighbours by contrasting their remaining opinions. Agents are, therefore, more influenced by neighbours with similar worldviews, which induces correlations and complex dynamics in evolution of each opinion. Simulations show that the worldview evolution exhibits intermittent polarization events when the social network is scale-free. This, in turn, trigger extreme crashes/surges in popularity of various opinions. Interestingly, the opinion dynamics generated by our model shows statistical similarities to the dynamics observed in financial markets. In the case of small-world networks we observe local intermittencies within the communities, which is unsynchronized with the rest of the networks.
Our formalism reproduces macroscopic social phenomena, like polarization, as an emergent consequence enabled by simple micro-dynamics. Moreover, the social network topology proves to have a critical effect, playing a key role in determining characteristics of the emergent dynamics of individual opinions.