## Sunday, October 23, 2016 ... /////

### Dark energy and unsettled science

Anthony Watts wrote a text

‘Settled Science’ syndrome hits Astronomy and the Nobel Prize
which clearly explains a fresh Oxford paper by Sarkar et al.
Marginal evidence for cosmic acceleration from Type Ia supernovae
in Nature's Scientific Reports. My understanding is that Watts took the text from another source and invented the title. He tells us: Look, there's no consensus about Nobel-prize-winning cosmology. Well, yes, no, it depends.

As the bulk of Watts' text explains, Sarkar et al. looked at a bigger ensemble of supernovae in order to determine the acceleration rate of the cosmic expansion. And the equivalent cosmological constant that is determined by the data is roughly speaking $3\pm 1$ in some units. So we're only "three-sigma" or 99.7% certain than this acceleration-based cosmological constant is nonzero.

The article suggests that this certainty is even lower than the near-four-sigma "certainty" that we had about the existence of the $750\GeV$ diphoton particle at the LHC. Because the latter turned out not to exist, the cosmological constant perhaps doesn't exist, either, we're told.

Well, I think that this monologue is absolutely demagogic. We can't compare these two situations because they dramatically differ when it comes to the prior probability of the null hypothesis. In the diphoton case, the null hypothesis is the Standard Model – an extremely efficient, successful, renormalizable quantum field theory that's been enough to describe all of particle physics for 40 years.

The analogous null hypothesis – the simplest hypothesis saying that the new effect doesn't exist – is problematic or non-existent in the case of the cosmology. Why? You may say that the supernovae's indication that the cosmological constant is nonzero is just a statistical fluke. But this statement doesn't contain a sensible definition of your alternative hypothesis – one that should be simpler and null.

The fact that the three (yellow, green, blue) ellipses overlap is nontrivial. It means that three sources of data seems to agree on a model – there is one non-trivial check of the model – and this agreement is known as "concordance" and the standard model of cosmology with the parameters near the red circle was also known as the "concordance model".

You may see that the supernovae – the thick blue ellipse – aren't the only source of information indicating that the cosmological constant is nonzero. Clusters (green) and the data from the CMB (yellow), the cosmic microwave background, are relevant, too. With these three sources of the data, it's rather hard to argue that the "right point" is somewhere on the horizontal axis instead of the red circle. Clusters themselves mostly determine the matter density (they say it's around 30% of the critical density) but in combination with the CMB, they determine the dark energy or cosmological constant to be near 70%, too.

Sarkar et al. address this obvious complaint in approximately one sentence. They say that the CMB argument is "model-dependent" and the model could be qualitatively wrong which could make the CMB argument invalid. They don't mention clusters at all but they could make a similar comment about it.

It's plausible but this vague excuse proves that the situation is in no way analogous to the Standard Model in the case of the diphoton excess. You just don't have any good null hypothesis to offer, Dr Sarkar, and your ill-defined claims about the "dust's not being an ideal gas" aren't enough as a predictive hypothesis.

So while it's conceivable that the existing model is not right, it just wasn't overthrown by this paper. The standard model of cosmology – where the dark energy is some 70% of the energy density – is still the best and simplest model we have that may describe the cosmological observations. And this statement would remain true even if you allowed the alternative models to deviate from observations by 3 sigma! The paper basically says that if you only pick some part of the empirical evidence and spin the broader situation in a certain way, the evidence in favor of the standard model of cosmology may look rather weak. I am not terribly impressed by that line of thought. Creationists basically criticize the evolution in the same way. It's the kind of a criticism that is always possible – so its existence isn't evidence in favor of anything – and some prejudice may be seen in the very fact that someone tries to frame the facts in this way.

Let me mention that Dr Sarkar has been trying to attack the standard model of cosmology at least since 2007 and he's been also active in a highly bizarre research questioning special relativity.

So of course it's possible that the most widely believed model is wrong and a better one exists. Except that the paper by Sarkar et al. is in no way the paper that establishes such a statement. So even after Sarkar et al., most cosmologists will surely consider the standard model of cosmology to be their preferred default explanation of the cosmological observations.

I want to make a sociological comment on Anthony Watts. Just like the left-wing abusers of science often say that "science has to be settled and everyone who doubts is a lunatic", their opponents – like Anthony Watts – often love to write stories saying that "everything is uncertain, doubts about absolutely everything are totally OK". I find both extreme viewpoints equally ideological and ultimately equally idiotic.

Of course, it's not true that everything in science that is believed to be true by a majority of scientists is guaranteed to be true. It's not even true that everything that has been rewarded by a physics Nobel prize has to be true forever. For example, my uncle Sir Nevill Mott got a Nobel prize for condensed matter physics. At least his proposed minimum metallic conductivity was wrong. Mott has done things that are right and things that are wrong and the justification for his 1977 Nobel prize is a bit imprecise. But I think that many people would say that the "key" things that were rewarded were actually wrong.

Even if you adopt their viewpoint, this situation is rather rare.

But the Watts-style skeptical attitude "everything is uncertain" is lacking any scientific substance. The reason is that whether one doubts a scientific law or result is ultimately a subjective matter. It depends both on the character of the claim as well as the personality of the potential doubter. If you want to doubt everything, including the downward gravitational attraction exerted by the Earth, you may do so. But this attitude doesn't say anything about the gravitational attraction only. It only says something about the combination of the gravitational law and yourself. And, well, it mostly says something about yourself. If you seriously doubt it, you're probably an idiot.

Watts clearly has no attachment to one cosmological model or another – well, it's because he doesn't really have a clue what they say and why they're sensible (or not). But it's completely unjustified for him to use someone's skepticism about a particular theory that Watts has no idea about – the standard model of cosmology, in this case – as a confirmation of some "I told you so" paradigm by Watts. Watts' decision to copy-and-paste a story about cosmology and to add an ideological title above the story mainly shows that Anthony Watts himself is eager to embrace some highly controversial claims about cosmology. But this eagerness doesn't make these statements right. It mainly shows that the standard or quality requirements that Watts imposes on claims about cosmology are rather low. The Sarkar et al. claims may ultimately be ludicrous, indeed (along with Watts' support for them), and even if they're not, Watts has contributed neither to the standard model of cosmology nor to the Sarkar et al. ideas so it's just irrational for him to "align" himself with either side.

After all, while the climate alarmism is being presented by the cultists as a "mandatory belief" because it's believed by a certain majority today, we may also interpret the explosion of the climate alarmism in the past as a consequence of the scientific community's inability to "impose" well-established results or a "consensus" on the members. Just thirty years ago, pretty much every scientist would agree that it's downright silly to be worried about the future of the life on Earth because of the increasing CO2. No ground-breaking results affecting this question have emerged since that time but the fringe community of the climate alarmists was allowed to grow, anyway. It shouldn't have happened.

So I obviously agree with Watts that it's a denial of the basic character of science when someone demands everyone to agree with some statements – especially when they are self-evidently wrong and unhinged statements such as those worshiped by the climate alarmists. But he goes further, to the opposite extreme, and I disagree with that, too.

Just like the climate alarmists and other "consensus scientists" would love to demonize everyone who dares to disagree with what they call a "majority opinion", Watts and others sometimes seem enthusiastic about demonizing everyone who actually feels rather certain about anything, especially if this anything is believed by a majority of some researchers. I am sorry but this demonization is as crippling for science as the climate cultists' attitude.

Science couldn't work if revisions and doubts were outlawed. But it couldn't work if some "de facto certainty" about some assumptions were outlawed, either. A real-world researcher simply has to trust lots of previous scientific results – they are necessary assumptions that need to be made for the further research to be done at all. Sometimes the assumptions may be wrong – but one of the aspects of the scientific atmosphere is that it's not the end of the world. A scientist simply has to take the risk much of the time. When it comes to sufficiently well-established assumptions, the risk ultimately turns out to be very low.