The importance of falsifiability

Many years ago, while I was still a graduate student studying particle physics, my supervisor Bob was very worried about supersymmetry. He was particularly worried that it will become the accepted theory without the need to be properly tested.

In those days, it was almost taken for granted that supersymmetry is the correct theory. Since he came from the technicolour camp, Bob did not particularly like supersymmetry. Unfortunately, at that point, the predictions of the technicolour models did not agree with experimental observations. So it was not a seriously considered as a viable theory. Supersymmetry, on the other hand, had enough free parameters that it could sidestep any detrimental experimental results. This ability to dodge these results and constantly hiding itself made supersymmetry look like a theory that can never be ruled out. Hence my supervisor’s concern.

Today the situation is much different. As the Large Hadron Collider accumulated data, it could systematically rule out progressively larger energy ranges where the supersymmetric particles could hide. Eventually, there was simply no place to hide anymore. At least those versions of supersymmetry that rely on a stable superpartner that must exist at the electroweak scale have been ruled out. For most particle physicists this seems to indicate the supersymmetry as a whole has been ruled out. But of course, there are still those that cling to the idea.

So, in hindsight, supersymmetry was falsifiable after all. For me this whole process exemplify the importance of falsifiability. Imagine that supersymmetry could keep on hiding. How would we know if it is right? The reason why so many physicists believed it must be right is because it is “so beautiful.” Does beauty in this context imply that a theory must be correct? Evidently not. There is now alternative to experimental testing to know if a scientific theory is correct.

This bring me to another theory that is believed to be true simply because it is considered so beautiful that it must be correct. I’m talking of string theory. In this case there is a very serious issue about the falsifiability of the theory. String theory addresses physics at the hypothetical Planck scale. However, there does not exist any conceivable way to test physics at this scale.

Just to avoid any confusion about what I mean by falsifiable: There are those people that claim that string theory is falsifiable. It is just not practically possible to test it. Well, that is missing the point now, isn’t it? The reason for falsifiability is to know if the theory is right. It does not help if it is “in principle” falsifiable, because then we won’t be able to know if it is right. The only useful form of falsifiability is when one can physically test it. Otherwise it is not interesting from a scientific point of view.

Having said that, I do not think one should dictate to people what they are allowed to research. We may agree about whether it is science or not, but if somebody wants to investigate something that we do not currently consider as scientific, then so be it. Who knows, one day that research may somehow lead to research that is falsifiable.

There is of course the whole matter of whether such non-falsifiable research should be allowed to receive research funding. However, the matter of how research should be funded is a whole topic on its own. Perhaps for another day.