There are many different views of scientific theory, but perhaps one of the most popular is the realist view of scientific theory. Scientific realism is based on the belief that unobservable entities, such as gravity or electrons, are scarcely different from ordinary observable things, such as tables and chairs.
To go into more detail, the scientific realist makes three basic commitments. Scientific realism states that the claims that any particular theory makes about unobservable identities are either true or false, and they are true or false depending on whether the entities talked about by the theory exist and are correctly described by the theory. This is the semantic commitment of scientific realism. Secondly, there is the metaphysical commitment. A scientific realist believes that unobservable entities do actually exist in the intransitive, independently of our thoughts about them. The third main point is that we can know about these unobservable entities; theories can describe and explain them in the same way as they do observable things. This is the epistemological commitment.
There are many reasons for holding such a scientific view, two of which I will describe below.
One of the main reasons for holding a realist view of science, put crudely, is that it seems to work. That is, by using our knowledge of unobservable entities, we are able to predict phenomena correctly. For example, I may use my knowledge of gravity (an unobservable), in order to predict how far an object will travel when projected at a certain velocity. If this prediction is correct, it lends credence to the fact that gravity does exist independently of us. For example, a scientific realist would point out that science must derive some ontological support for protons and neutrons from the outstanding phenomenological success of all the theories using them. Therefore, the fact that when we use our knowledge of unobservable entities we tend to get reliable and accurate predictions lends a huge amount of ontological credence for our believing in them objectively and independently of ourselves.
Another reason for postulating the existence of these unobservable entities can be found by looking at the case were these entities not to exist i.e. what would the world be like if these entities, such as gravity, did not have independent existence. As Jack Smart correctly stated, If the phenomenalist about theoretical entities is correct we must believe in a cosmic coincidence’ (1963:39). That is to say that without the existence of these unobservable entities the universe would be in some kind of flux, with events happening by chance. Objects would fall to the floor, not because of an unobservable entity called gravity, but through random chance. Existence in this universe would be treacherous, as we would have no way of knowing whether objects would continue to fall to the floor, or mysteriously cling to my left shoe. Therefore, we either postulate the real objective existence of these unobservable entities, or we are reduced to an absurd universe with an equally absurd existence.
To conclude, we should hold a realist view of science for the reasons that it successfully explains scientific phenomena and that other views are reduced to absurdity as they lead to a view of the universe as some kind of cosmic coincidence.