ABSTRACT
For the realist, this problem results from analysing scientific explanation as consisting of a form of logical argument. Even if it is a necessary condition of explanation that such arguments can be constructed, it is not a sufficient one. What is additionally required is the delineation of the causal processes involved in the generation of the phenomena we are trying to explain. Instead of establishing, as in the D-N model, relations of logical necessity between the explanans and explanandum statements, we should discover the relations of natural necessity that exist in the physical world. To show that the truth of the conclusion follows necessarily from that of the premisses is no substitute for describing the necessary, causal connections between things in nature.
