Ever wondered what theorists do in their spare time? Why… philosophize of course! I received this in my inbox today from one of our fellow physicists with the subject line of “For your amusement…” .
If you have any comments for Byron, post here and he will respond. Enjoy!
– Jennifer Gagné, Web Publishing Coordinator
Sterile Neutrinos And The Problem Of Evil
B.K. Jennings
The thesis of this essay is not that sterile neutrinos are evil; though they may well be. Nor is it that physicists are evil though their significant others may, at times, think they are. Rather it is on the physicists’ self image and the idea of falsification. Karl Popper proposed in his work on the philosophy of science that what distinguishes science is that its hypotheses can be falsified. This is the known as the demarcation criteria and has had a large impact on physicists’ idea of what they are doing. In a lab, one hears the question “Is that falsifiable?” in hallway conversations. However, the philosophy community has largely dismissed Popper’s ideas. The main criticism is known as the Duhem-Quine Thesis: it is impossible to test a scientific hypothesis in isolation, because an empirical test of the hypothesis requires one or more background assumptions. Lets see how this works or doesn’t work by considering two examples: one from classical theology and one from modern physics.
Below, we have a statement of the problem of evil:
Is God willing to prevent evil, but not able?
Then he is not omnipotent.
Is he able, but not willing?
Then he is malevolent.
Is he both able and willing?
Then whence cometh evil?
Is he neither able nor willing?
Then why call him God?
The statement of the problem is first attributed to Epicurus, predating Christianity. By the Popper demarcation criteria this would be a scientific hypothesis. If you accept the assumption that there is an omnipotent, omnibenevolent God, then there is the falsifiable prediction that evil does not exist. At first blush, it looks like the assumption has been falsified and God cannot be both omnipotent and omnibenevolent. But now the Duhem-Quine Thesis kicks in. One of the many refutations of the problem of evil comes from Gottfried Leibniz (1646 – 1716). That’s right—the opponent of Newton on the priority of the discovery of calculus. There was no separation science and religion in those days.
Leibniz’ argument was this is the best of all possible worlds: a position known as optimism and ridiculed in Voltaire’s Candide. Putting that ridicule aside, let’s look at the argument. Leibniz said that humans have only seen a very small part of the entirety of existence. In his theology, souls exist forever and so it would be premature to judge if this is the best possible world since we have seen such a small fraction of it. He has attacked the “whence cometh evil?” question, one of the background assumptions and not the main assumptions of omnipotence and omnibenevolence. According to Leibniz, the reason we do not understand where evil comes from is due to our limited knowledge and understanding. Once we have a better understanding, presumably in the afterlife, we will see where evil, or the appearance of evil, comes from. This approach used the Duhem-Quine Thesis to get around the falsification. For the sake of argument let us accept his argument as valid. It avoids falsification, but at a price. While the initial statement meets Popper’s demarcation requirement for science, Leibniz has removed any possibility of falsification. Any outcome is consistent with our limited understanding and the problem of evil has been removed from science to the realm of theology or philosophy. This is perhaps where it should be.
Now before physicists get too smug and think this is only a problem for theology, let us consider another problem: sterile neutrinos and neutrino oscillations. Here again we will see that falsification can be avoided but only at the price of moving from science to metaphysics.
If we take the three known generations of particles, we have three neutrinos. Various experiments, at Super Kamiokande, SNO, and elsewhere, have shown that the neutrinos oscillate. Most of these experiments are consistent with three neutrinos oscillating among themselves. But there is a striking exception—the LSND experiment at the Los Alamos National Laboratory. There are two principle ways to explain the LSND results, 1) the experiment is wrong, or 2) there is another type of neutrino, a so-called sterile neutrino.
Both of these approaches avoid having LSND falsify the main idea that what is being seen in the previous experiments is due to neutrino oscillations. Both reduce the ability to falsify predictions. If we simply assume that all experiments that disagree with the prevailing wisdom are wrong, we have nothing left and falsification becomes impossible. So we need a good reason to throw out experimental results we do not like. Way 2) also makes falsification more difficult—not impossible—but more difficult. Adding an extra neutrino means we need more parameters which then can be adjusted, meaning we have more wriggle room to explain inconsistent results.
To help decide between these two ideas a new experiment was carried out: MiniBooNE. The results of this experiment were inconsistent with the LSND results – if you assume only one sterile neutrino. But Duhem-Quine strikes again. Why not assume three sterile neutrinos, one for each generation of particles? We avoid falsifying the hypothesis that the LSND is due to sterile neutrinos, but at the price of introducing more parameters and making the new model less falsifiable. If we add additional sterile neutrinos at will, we can completely avoid any possibility of falsification. In doing so we fail Popper’s demarcation criteria, and the model becomes unfalsifiable. In general, as we add more parameters we move down the slippery slope to unfalsifiable, from science to metaphysics. (I avoid saying from science to theology.)
But, what happens when, in avoiding falsification, we make the model more falsifiable? Well that’s another paradigm.
When Pauli first theorized the neutrino, he said “I have done a terrible thing, I have postulated a particle that cannot be detected.” As it turns out, he was wrong. Not only have they been detected, but remarkable properties have been measured. They (neutrinos) are now capable of helping us understand the solar interior, rather than vice-versa. All these things in only 80 years – a lifetime, yes, but only a moment in the context of human history.
I mention this only because I think that what seem like untestable things can become testable in (relatively) short timescales. The people who have written papers on 2+ neutrinos to explain this I think consider them fully falsifiable with adequate further experiment, and those that I know at least only pursue studying them because they want to consider falsifiable models. (I have not myself written any.) Can one come up with a model which is completely impossible to test? Sure, but equating these efforts to those is unfair. If we are scientists, we should at least be aware of what kind of physics explains the existing data, and we should pursue it experimentally if we feasibly can.
While it may not be this year or next year that it’s tested, it’s unreasonable to equate results than will not occur in this funding cycle or the next to eternity. Well, perhaps in the current fiscal climate…
Just a feeling, a hunch. I am not a physicist. I think sterile neutrinos are a cop-out.
Leibniz was and still is totally right. Our senses and what the brain makes of it is very limited. The scope of human perception is a fraction of reality. It is constructed to find food in the wild and not to figure out the foundation of existence. Furthermore, evil is the bigger part of benevolence. Without it we hadn’t been evolved. Its the drive of evolution.
Of course why even bother with evolution if you want to make some humans and are truly omnipotent? You can’t escape the problem this way. Personally I think the omnipotent assumption is an unreasonable one even for a god
.
In response to the first post, I just note that the Duhem-Quine Thesis implies that no model is falsifiable. The point of post is to point out the avoiding falsification comes at a cost.
I must be missing the point. The way you describe the process of dealing with neutrino oscillations is pretty much exactly the way Popper meant it to work: You have a theory, you find a counterexample, boom, theory dead, make up a new theory and repeat. Nobody avoided falsification. It happened at every step.
Falsification does not mean finding a fault and dropping the subject. It means finding a fault and then trying to get it right(er).
Hi, just wanted to mention, I liked this article. It was helpful. Keep on posting!