On the Aim of Scientific Theories in Relating to the World: A Defence of the Semantic Account