Explanations of Phenomena and Strict Empiricism

As everybody knows, science aims to explains things—scientists propose theories to help them understand or explain observable phenomena. Since an explanation is not adequate unless it is true, scientists seek true theories.

A. Musgrave (is)

University of Otago, Dunedin, New Zealand

e-mail: alan.musgrave@otago.ac.nz

© Springer International Publishing AG 2017 71

E. Agazzi (ed.), Varieties of Scientific Realism,

DOI 10.1007/978-3-319-51608-0_4 theory-choice. There is nothing to choose between theories that empirical evidence cannot choose between.

So suppose we have two theories that say the same about the observable phe­nomena, two theories that are equivalent as far as the observable phenomena go, two theories that are ‘empirically equivalent'. No observation or experiment can decide between these two theories. Observation or experiment may, of course, tell us that both of them are false—for they may both predict something that turns out to be wrong. But what if the two theories are ‘empirically adequate' as well as empirically equivalent? What if they tell us the same things about observable phenomena and everything they tell us is true? Strict empiricists will say not only that there is nothing to choose between them, but also that they equally well serve the aim of science. For according to strict empiricism, also known these days as ‘constructive empiricism', the name of the scientific game is ‘saving the phe­nomena' (Van Fraassen 1980, p.

This means that there is nothing to choose between any scientific theory T and what I call its surrealist transform T*: ‘The observable phenomena are as if T were true'. The surrealist transform T* of T is by design empirically or observationally equivalent with T. They ‘save the observable phenomena' equally well. So according to strict empiricism, there is nothing to choose between them.

You may think that surrealist transforms of scientific theories are a mere philosopher's plaything. Not so. They have a history, and served serious purposes. It was the ancient astronomers who discovered the first real example of empirically equivalent theories. They proved that two quite different theories about the sun's (apparent) motion made exactly the same predictions about the (apparent) positions of the sun. Both theories saved the phenomena equally well, yet both could not be true. However, astronomical observation could not tell the ancients which of them is true, and it is hard to see what else might have told them. Perhaps astronomers should forget about truth, perhaps the whole task of astronomical theory is just to save the phenomena.

2