ABSTRACT
However we resolve the debate about the nature of the scientific method, most parties seem to accept that science is usually the best guide we have to the future behaviour of things we can observe, for example comets, bridges, power plants and rainforests. Our scientific knowledge is fallible, partial and approximate, but usually it is the most reliable means we have for predicting phenomena in the world around us. However, science is often taken as telling us much more than this. The natural sciences seem to tell us about the ultimate nature of things and are often thought to have replaced metaphysics as the study of the fundamental structure of reality. Modern science presents us with what seems to be a detailed and unified picture of reality that describes the composition of things and the laws that they obey, from the internal structure of atoms to the life cycles of stars. Contemporary genetics and brain science even seem to offer the prospect of a physical science of human beings and their behaviour. Many of the entities postulated by modern science, such as genes, viruses, atoms, black holes, and most forms of electromagnetic radiation, are unobservable (at least with the unaided senses). So, whatever the scientific method is and however scientific knowledge is justified, we can ask whether we ought to believe what science tells us about reality beyond the appearances of things. Roughly speaking, scientific realism is the view that we should believe in the unobservable objects postulated by our best scientific theories.
