Is an unfalsifiable theory automatically a bad theory?

Answer

Dear Giovanni,

Let us say that an unfalsifiable theory is scientifically irrelevant. By non-falsifiable we mean that it is impossible in advance to devise an experiment that can test the theory.

An example: I could say that a dragon lives in my pond that is invisible, makes no sound, does not stir the water, etc. It is impossible to check whether such a dragon actually lives in my pond. It is therefore not a scientific theory. The answer to this question is also totally irrelevant, because since the dragon can interact with us neither directly nor indirectly, it will never make any difference whether the dragon exists or not.

The concept of non-falsifiability arises from the fact that scientific theories are generally unprovable. Take Einstein’s theory of relativity as an example. That theory has been confirmed by an untold number of experiments in the lab and in everyday life (e.g. every time you use a GPS). But that is not absolute proof that the theory is correct. One experiment showing a particle traveling faster than light (hence the importance of the neutrino experiment) shows that the theory is wrong (or at least is only an approximation).

In summary: A scientific theory cannot generally be proven, but it can explain all existing and hopefully all future experiments. One negative experiment is enough to undermine the theory. But if it is a priori impossible to test the theory, ie if the theory is non-falsifiable, then the theory cannot teach us anything and is therefore totally irrelevant.

Kind regards,

Philippe Tassin
Physicist Vrije Universiteit Brussel & Iowa State University