In practice, we develop new computational tools to study the media we produce and consume, the beliefs we hold and stories we tell, our myths and traditions, as well as the opinions, ideas, and narratives we propagate – how they develop, and how they evolve.
In doing so, we combine tools and approaches from data science, mathematical modelling, complex systems, algorithmic information theory, evolutionary biology, cultural evolution, cultural sociology, political communication, social psychology, and cognitive psychology.
For examples of ongoing projects, see below.
Over the last decade, the increasing influence of social media in public opinion has highlighted the interplay between the spread of (mis)information, social influence, and opinion formation. Buzzwords such as "virality", "fake news", and "infodemic" have been quickly adopted by scholars, journalists and policymakers. It's useful to have names for these phenomena, but the truth is: we still don't know how it all really works. Are we really vulnerable to misleading information like some special kind of disease? Or are we rather wary learners, carefully weighing the consequences of following our peers? With the unprecedented scale and impact of misinformation, there is now an urgent demand for rigorous research into how our beliefs evolve when faced with new information.
In the CC Lab, we are developing new methods to measure how people change their minds, from the individual to the crowd. We are applying these new methods to data on how we interact online drawing from mathematical models of opinion dynamics, data mining and natural language processing, to be able to develop more accurate theories of how beliefs evolve. Is belief change basically a rational process? Or are people simply trying to minimise some form of cognitive dissonance between their opinions and their identity? Or rather, do ideas and beliefs have their own ecology, interacting with each other in ways that resemble cooperation and competition? It's hard to tell – but that's what we're trying to do. Understanding how beliefs and opinions evolve will allow us to develop better, evidence-based policies regarding the spread of misinformation and the polarisation of public opinion.
Narratives are the stories we tell about the world. They are the way we connect new information to our preexisting beliefs and opinions. For instance, two people might read the same news article about how the government handled an important issue, and one might say the article shows how their government is incompetent, while the other person might say the article is biased, and their government is actually doing a fantastic job. Both might interpret the article differently, each reinforcing different combinations of beliefs and opinions they might hold – and ultimately, different sets of narratives – about the world.
In the CC Lab, we apply tools from natural language processing, machine learning and social network analysis, and draw from work on agenda setting, framing, and political communication, to look at the networks of narratives that form around particular political topics. Examples include how different political actors frame others as friend or foe, or how particular stances (pro-this, against-that) are often expressed together, as well as which metaphors are used, what they imply, what they highlight or play down. By developing methods to understand how these narratives unfold, how they connect and evolve, we hope to shine some light on problems about conspiracy theories, radicalisation, and modern forms of political propaganda.