The document discusses the causal interpretation of statistical models in social research. It outlines different perspectives from staunch causalists to moderate skeptics. Interpreting a statistical model causally is described as an epistemic activity to decide if a model is valid, rather than determining a physical causal relation. The causal interpretation depends on the statistical information and machinery used to make inferences from the model. Keeping statistical and causal inferences distinct is important, while acknowledging the role of background knowledge in interpretation.