ABSTRACT. A functional analysis depends on causal explanations and stimulates causal explanations. The anti-positivist naturalism questions the empiricist concept of causality. The covering law theory of causality has no been successful when it comes to causal explanations of social actions. Manipulation accounts contend that a causal connection exists only where we can manipulate the cause to bring about the effect. Causal models allow us to predict the effect of changes and interventions, capitalizing on the locality of such changes. Empirical assessment of causal mechanisms does not have to go down to observations at the level of individuals.