Confederated developments in computer science, psychology, and other fields have created a real revolution in human understanding of causal relationships that have taken place throughout history. Causal modeling provided us with a precise, formal framework to represent causal relationships. This framework is undeniably influential and powerful due to its utility in raising concrete scientific questions.
During the talk, Professor Ismael reviewed some important aspects of formalistic reasoning and stated that the type causality is fundamental. It is the process of determining the different types of causation in events associated with a network. Using directed acyclic graphs (DAGs), it is possible to visualize the causal network, and the arrows indicate the causation connections and the nodes represent the events. In this framework, judgements are made using loose and tacit assumptions about which variable to consider and what to hold fixed during the process. In reality, variables cannot be separated from their underlying causes. Hence, to be able to study causal structure in practice, there is the possibility of randomizing the values of variables, which effectively separates it from its own past causes so that changes in its values can be treated as interventions.
According to traditional analyses, causal relations can be viewed as chains of interconnected events, while causal interactions are those involving local exchanges of an observed quantity among events. A causal process can also be defined in terms of micro dynamical loss, so this causal concept is closer to the dynamic laws found in Newtonian physics. As for interpretations of interventionists, causal relationships, and causal processes, they are all found in classical physics. Our thinking about causation changed significantly as a result of structural causal modelling and now we can understand why causation doesn’t run backwards and why time seems to flow from the past into the future.
Causation establishes a connection to agency and connects with people’s pre theoretic idea that causation, like interest and action, can be used as a means to an end. As a result of the thermodynamics gradient, if we hold fixed all of the information embedded in the observed macrocosmic state of the universe, the state is somehow fixed. Therefore, we can manipulate some features of our local environment and apply the probability postulate to obtain the probability distribution over the possible microstate of the world and then project that forward or backward in time based on classical mechanical laws.
Professor Jenann Ismael concluded the talk by explaining the existence of causal pathways in nature and how those pathways are neither asymmetric nor compulsive. The complete understanding of causal concepts depends on facts about the world and our place in it, which support our learning and understanding.