Gravitational Anomalies

[Rev. Green]

Is there an archive of answerbank questions? I'd like to see a question I asked some years ago about the possibility that dark energy and dark matter could be explained away by modifying the way that the gravitational force is described.

[douglas9401]

https://www.theanswerbank.co.uk/Science/Question1656683.html

This one?

[ladybirder]

Crikey, what took you so long Dougie.

[Rev. Green]

Many thanks Douglas, yes, that's the one. It might have received a different response today.

[douglas9401]

And thank you, Rev.

I was between bites of cake, ladybirder. :-)

[Old_Geezer]

From what I've heard the gravity modifying conjecture tends to produce other issues to answer, and doesn't explain all that dark matter does. But all avenues are worth trying.

Meanwhile you need no archive as such. It'll be on your account past questions, but you'd need to look back. Failing that, there's a search function.

[ClareTG0ld]

// It might have received a different response today.//

Quite possibly -- science is often in flux, and answers can well carry an expiry date. Why do you wonder this, though? If, for example, it's because of the news about "muon g minus 2", then the BA in that earlier thread already hinted at this anomaly. For what it's worth, I don't think the new results change the picture from the same answer four years ago.

[Rev. Green]

No, Clare, it wasn't the muon, it was the 5 sigma result of gravity deviating from its expected value for binary star systems with large separation of the stars, which appears to make dark matter redundant.

[ClareTG0ld]

Oh.

Well I haven't come across that story, and while in itself that doesn't mean much, I'm surprised that a result that would effectively rule out Dark Matter, something that is extremely well-established at this point to the extent that we even know its distribution in the Universe and in our local galaxy, would go essentially unreported.

Where'd you see this?

[lankeela]

CTG as Sheldon would say do you want some aloe vera? Cos you just got burned!

[ClareTG0ld]

I'll be fine.

I found it -- it is, as I suspected, not what Rev. Green said.

https://bigthink.com/starts-with-a-bang/binary-stars-prove-modified-gravity/

Put simply: no, it doesn't disprove the existence of Dark Matter.

[Rev. Green]

Thanks Clare, I hadn't seen that well-reasoned report. News of the original paper merely said that MOND had been verified at a 5 sigma level, without mentioning that the data were selected suitably. Perhaps I can show that all dogs are Collies by studying only sheep farms?

[Zacs-Master]

‘ Is there an archive of answerbank questions?’

Erm, yep, that’s why the name has ‘bank’ in it. You sure your up to discussing gravitational anomalies, rev? ;-)

[ClareTG0ld]

// News of the original paper merely said that MOND had been verified at a 5 sigma level ... //

To a certain extent, anybody should have stopped reading at least the *report* at that point. I won't comment on the merits of the original paper, which have been discussed in depth in the link I provided; but your ***-detector should be in high alert at the point that it's claimed that such a paper is in any sense definitive.

Take the "at 5 sigma level" part, for example. I wouldn't here wish to undermine at all the idea that "5 sigma" is a useful threshold, a gold-standard even, for confirming some given observation. But that comes with a few caveats. First, and perhaps most important, is that in a paper like this it's being attached to a secondary analysis. The author of the study claiming to have verified MOND is not part of the "Gaia EDR3/DR3" team, which is to say that he's analysing somebody else's data set. Analyses like that are evidently not necessarily wrong, but it's a factor to take into consideration: whenever you are using somebody else's data set, you might be less familiar with how it was generated, what its limitations may be, what assumptions went into evaluating the given uncertainties, etc. You'd hope that all this would be made available alongside the data set, and indeed often it is, but even still there's always that barrier between the "author" and the "reader" that can play a role.

In a case like this, it's even more relevant. Without wishing to accuse the author of definitely publishing results that happen to favour their pet theory, what is at least true is that somebody who hadn't come across the "AQUAL" variation of MOND is rather less likely to have written such a paper with such a bold conclusion. As an example I'm more familiar with, I mentioned muon anomalies. The one that appeared in the media recently is about the "g-2" anomaly, but there have been a few others all linked to muons from the last decade or so. If you want to look them up, search for "RK anomaly" and "P5 prime anomaly", or have a look at the (now slightly out-of-date) review in https://arxiv.org/pdf/2110.13270 . The technical details don't particularly matter, but what *does* matter is this: firstly, the anomalies seem to be disappearing one-by-one as more data are gathered and as theoretical analysis improves; secondly, at least individually the anomalies were never at 5 sigma level, and it's only secondary analysis (eg Alguero et al http://arxiv.org/abs/2104.08921 , claiming up to a 7.4 sigma deviation(!)) that suggested otherwise; thirdly, even if after all they remain, some of the wilder interpretations touched on in that review aren't necessarily responsible, and a more plausible interpretation is that we're messing up the predictions somehow, by missing or miscalculating something that we are already "able" to account for within the current theory. London/Matias seemed to downplay this explanation, and it's interesting to me to see them argue that even a recent experimental result ("note added" at the end of the review), leading one of the anomalies to almost disappear, end up "reinforc[ing]" the case for the new physics that Matias has spent much of his recent career looking for. (NB I know Matias, he's a good physicist, and in his defence once told me that he spent his entire career "hoping to be wrong", because that's the more interesting outcome. A good attitude to have, in other words; but "reinforc[ing]" is a subjective term here that I cannot agree with).

TBC

[ClareTG0ld]

A third common feature, actually, is how "local" the anomaly is. Take the claim about binary stars throwing shade on dark matter. Well, maybe. But there are maybe a dozen other separate pieces of evidence for the existence of Dark matter, so even if one of those fell
down a rather more plausible explanation is that we don't fully understand how to compute binary star distributions in the presence of dark matter, rather than that it doesn't exist at all (here I overlook the other criticisms of the MOND study in the review I posted above). Going back to the muon anomalies, the P5 prime one I mention was first seen in an LHCb paper back in 2013 ( https://arxiv.org/pdf/1308.1707.pdf ), and amounts to a minor deviation from the then-prediction in one out of 24 separate measurements. Again, that it's only 1/24 of the measurement doesn't automatically mean that nothing "new" is happening, but it's definitely worth being aware of this point: other factors could be at play that affect just how reliable a measurement and prediction are in that one region. Conversely, any new model that would be needed which *does* manage to explain that deviation also needs to be carefully "tuned" in order to not mess everything else up, which at least is tricky. There's historical precedent for minor "tuning" leading to major breakthroughs, one such example being how a small correction to the precession of Mercury's orbit was a major clue that led to General Relativity over Newtonian Gravity. But, at a risk of being proved wrong in time, I think there are other conceptual reasons why you should be more sceptical about *this* minor correction (in the muon/ P5 prime anomaly) leading to the New Physics breakthrough we're all waiting for.

Going back to the Dark matter story, then: MOND has been found to fail to explain many other observations that Dark Matter does, and to the extent that it's been partially successful it's often been with some fusion of the two (ie, some variations on Newtonian gravity but also with a Dark matter candidate added anyway). Put quite simply: one paper, by one author, analysing one data set with that author's own obvious bias in favour of MOND, is not evidence that MOND is correct. The paper itself is, I think, still worthy of attention, but not because it's definitive. Papers that end up being "wrong" are still worth publishing, because the very act of trying to show that they are wrong advances the field of science: and, for that matter, trying to show that this one is "wrong" and ultimately failing would be an even more exciting development.

For the people outside the field, any paper that purports to overturn an entire discipline should be treated with the utmost scepticism. Some of this is also because journalism: generally, the headline-grabbing result is the "omg we didn't expect this at all!" rather than the "oh wait never mind, things are just as we anticipated". In that sense, there will almost certainly never be a follow-up to this report; no doubt at some point in the next year, yet another challenge to Dark Matter will be picked up, and then quietly dropped when the challenge fails...

[Rev. Green]

Thanks Clare, I'll remember that the next time I read an American journal.