The article discusses different theories of causality and how using a variety of methods can help us better understand cause and effect relationships. It explains that no single theory is better than another and that we need to consider different perspectives. The researchers highlight the importance of making untestable assumptions when making causal claims in both experiments and observations. They also suggest that using computational tools like agent-based modeling can help us understand causality better. Overall, the article argues that to make convincing claims about cause and effect, we need to use different methods and viewpoints.