Davies' Dubious Defense of a Double Standard

In my blog post several days ago I complained about what I called a double standard followed by many a modern physicist. The double standard is that theoretical physicists spend huge amounts of time speculating about strange, far-out concepts such as spacetime wormholes, time travel, parallel universes, string theory and multiverses (things for which there is no observational support), but the same individuals dismiss as nonsense many possibilities such as ESP for which there is a great deal of evidence (much of it accumulated by scientists and documented in scientific papers that have been published for decades). That evidence includes compelling ganzfeld studies that show a success rate of about 31% greatly in excess of the expected success rate of 25%. For example, this study says, “if one considered all the available published articles up to and including 1997 (i.e., Milton and Wiseman’s 30 studies, plus 10 new studies), 29 studies that used standard ganzfeld protocols yielded a cumulative hit rate that was significantly above chance (31%).”   An ESP study involving only artistically gifted people reported a success rate of 50%, twice the success rate expected by chance.

Schematic depiction of ESP

The very next day I read a book The Eerie Silence by physicist Paul Davies that contained a passage that almost seemed to have been written in response to my blog post. He discussed exactly the same discrepancy I discussed, using the same term that I used (“double standard”) to describe it; however Davies tried to defend this double standard.

Before trying to rebut this reasoning by Davies, let me make clear that I am a longtime Paul Davies fan. I have enjoyed many books he has written, and in 19 cases out of 20 I find his reasoning to be convincing. But in the example I will now discuss, I think Davies fails to come up with a convincing argument.

Davies starts out like this:

The point about modern physics is that weird entities like dark matter or neutrinos are not proposed as isolated speculations, but as part of a large body of detailed theory that predicts them. They are linked to familiar and well-tested physics through a coherent mathematical scheme. In other words, they have a place in well-understood theory. As a result, their prior probability is high.

Davies is on extremely dubious ground here. Neutrinos are predicted by the Standard Model of Physics, but dark matter is not at all predicted by that theory. There are no particles of dark matter mentioned in the Standard Model of Physics. Physicists believe in the likelihood of dark matter not for theoretical reasons but because of observational reasons, because they need dark matter to help explain certain observations. Exactly the same thing can be said about ESP.

Dark matter is not at all “linked to familiar and well-tested physics through a coherent mathematical scheme.” It is instead a completely mysterious alleged thing that we basically know nothing about. We have zero tested equations that describe dark matter, and also zero equations that describe ESP. That really leaves dark matter and ESP in the same ballpark.

In the case of a similar and equally important mysterious phenomenon – dark energy – we can say that there is a strong theoretical basis for believing that there is something like dark energy. However, the problem is that the theory (quantum field theory) tells us dark energy should be at least a trillion trillion trillion trillion trillion times more powerful than it is. This is the widely discussed “vacuum catastrophe,” often referred to as the worst prediction in the history of science. Scientists continue to believe in dark energy, even though their theories are almost infinitely out of whack with our observations about how much dark energy exists. So is dark energy something that is “linked to familiar and well-tested physics through a coherent mathematical scheme”? No it isn't. You can't use such language when the theory gives an answer that is wrong by a factor of 1,000,000,000,000,000,000,000, 000,000,000,000,000,000,000. So any claim of theoretical rectitude can't be used for dark energy – but scientists continue to believe in it.

In truth, neither dark matter nor dark energy is part of “well understood theory.” They are mysterious things that we do not have any well-established theory for (at best we have half-baked, super-speculative or “way off the mark” theories). The same thing can be said about multiverses, parallel universes, time travel and wormholes.

I may also note that Davies errs when he suggests that the “prior probability” of something is “high” when it can be “linked to familiar and well-tested physics through a coherent mathematical scheme.” With sufficient ingenuity, a physicist can create all kinds of bizarre, improbable theories that are linked in some way to existing physics. Inventive physicists have done that, creating a thousand and one conflicting theories of time, space, particles, and forces, including countless different varieties of string theory, inflation theory and quantum gravity. The fact that you can somehow link your theory in with existing physics does not show it has even one chance in 100 of being correct. As Davies himself says in another book, “It is easy to construct artificial universe models, albeit impoverished ones bearing only a superficial resemblance to the real thing, which are nevertheless mathematically and logically self-consistent” (Information and the Nature of Reality, page 68).

Regarding ESP, Davies says the following:

Telepathy is not obviously an absurd notion, but it would take a lot of evidence for me to believe in it because there is no properly worked out theory, and certainly no mathematical model to predict how it works or how strong it will be in different combinations. So I assign it a very low (but non-zero) prior probability. If someone came up with a plausible mechanism for telepathy backed up a proper mathematical model which linked it to the rest of physics, and if the theory predicted specific results – for example, that the 'telepathic power' would fall off in a well-defined way as the distance increases, and would be twice as strong between same-sex subjects as mixed-sex subjects – I would sit up and take notice. I would then be fairly easily convinced if the experimental evidence confirmed the predictions. Alas, no such theory is on the horizon, and I remain extremely skeptical about telepathy in spite of the many amazing stories I have read. 
 

Here Davies professes quite a demanding set of criteria before we can believe something: (1) that we should only believe something if it is predicted by some mathematical model; (2) that this model should be “linked to the rest of physics”; (3) that the model should make specific numerical predictions that can be confirmed. Does it make sense to advance such a set of criteria as a prerequisite for believing in something? It certainly does not.

One reason is that a very large fraction of all things that we do believe in (inside and outside of the sciences) do not satisfy such a set of criteria. But we accept such things nonetheless because we have observations that compel us to believe in them. I refer to things such as love, hate, psychological discomfort, newly discovered species, gamma ray bursts, and earthquakes. Consider, for example, when a biologist discovers a new species of animal in Brazil. It is not at all something predicted by some mathematical model, but it is accepted as a new scientific finding nonetheless. Even in the hard physical sciences we have things that are accepted even though they are not at all predicted by existing models. A recent example is the discovery of the acceleration of the universe's expansion, which left scientists stunned, having been predicted by no popular theory.

What Davies describes here is a model of verification that is sometimes followed (but often not followed) in the world and physics and astronomy, but such a model of verification is actually uncommon in many other sciences. Sciences such as psychology, geology, and biology make relatively little use of mathematical models and prediction. If you pick up a textbook on zoology you will see almost no equations or mathematical models anywhere. In such sciences there is no standard at all that some new scientific conclusion must be supported by some mathematical model.

As for Davies' idea that ESP would need to be supported by some theoretical model “linked... to the rest of physics,” such a prerequisite does not make any sense. It amounts to saying, “I refuse to believe in something unless it is similar to things I have already learned.” I may note that some of the greatest advances in physics and astronomy occurred when scientists started to postulate things that did not fit in with the previous framework of ideas. When quantum mechanics was introduced, it did not fit in at all with physics as it had been previously understood. When the Big Bang theory was introduced, it did not fit in at all with the previous cosmological ideas of scientists such as Einstein, who favored the idea of an eternal, static universe.

Davies suggests the idea that we should not take an observed phenomena seriously unless we have a theory of how it works, one that makes good predictions. That is a misbegotten principle that scientists themselves do not follow. When scientists observe a new type of phenomenon, they accumulate observations for the phenomenon, and at the same time start working on theories to explain the phenomenon. It may be decades or centuries later that a theory arrives that finally explains the phenomenon. An example is pandemics. The phenomenon was studied for centuries before scientists such as Pasteur finally came up with a decent theory to explain it.

It would have made no sense around 1600 to have said, “Don't believe in pandemics – we don't have a good theory to explain it.” Another example is lightning. Scientists observed it for a long time, but did not develop a decent model to explain it until the 18^th century. It would have made no sense around 1500 to have said, “Don't believe in lightning – we don't have a good model to explain it.”

What Davies' criteria amounts to is a kind of plea that he should be excused from taking something seriously whenever that thing does not have characteristics that allow him to study it in the way that he is most familiar and comfortable with studying things. That's a lame type of reasoning. We can imagine the same type of reasoning being used by a biologist: “I refuse to believe in galaxies or other reputed deep space-objects, because I cannot view them through my microscope, or place them in my test tubes, or study them in a cage or an aquarium.”

A wiser approach is that we should take a phenomenon seriously whenever we have repeated compelling evidence for its existence, regardless of whether we have familiar off-the-shelf methods for studying the phenomenon, and regardless of whether the phenomenon comfortably meshes with our preconceptions.