- Physics - World. Hello and welcome to the Physics World Stories Podcast. I'm Andrew Lester. And in this episode, we're gonna be exploring the debate raging between dark matter theorists and modified Newtonian dynamics or mon theorists. The debate centers around two competing theories, which are proposed to explain the discrepancies between the observed gravitational effects in the universe and the predictions made by Newton's laws of gravity and general relativity.
There are particular issues in the way that galaxies behave. For example, the rotation curves of galaxies are faster than you would expect. Dark matter theory suggests that the missing mass causing these gravitational anomalies is due to this unseen form of matter. Dark matter doesn't emit absorb or reflect light, meaning it's invisible and detectable only through its gravitational effects. Dark matter is thought to make up around 27% of the universe's mass energy content.
Mon, on the other hand, suggests that these discrepancies can be explained without invoking unseen matter. Proponents of mon propose modifications to Newton's laws of motion and gravity at very low accelerations, which are typical of the conditions found at the outer regions of galaxies. According to mon, the gravitational force behaves differently at these low accelerations of gal rotation curves without requiring the unseen dark matter in the astronomical community.
There is something of a tribal debate about it on the physics world website. There's the first of three features by Keith Cooper exploring this topic, cosmic combat, delving into the battle between dark matter and modified gravity. Having read that article, I wanted to find out more. So in this podcast, we'll hear from two researchers, one who's moved from the mon camp to the dark matter camp, and first Professor Stacey Magor, who moved from dark matter to be a supporter of mon.
- I am, uh, on the faculty at Case Western Reserve University in Cleveland, Ohio and the United States. Uh, I work on galaxies and cosmology, and, and those things have sort of brought me to be interested in the dark matter problem and, and the modified Newtonian dynamics of Milgram. So in, um, both of those, uh, astronomical objects, I mentioned galaxies and cosmology. You find that things don't add up.
What you see is not what you get in the sense that if you use the law of gravity as taught to us by Newton and Einstein, um, applied to the visible matter, uh, things don't work out in galaxies. Stars and gas are orbiting faster than one would predict. Um, the cosmology as a whole, um, doesn't really add up unless you have both some extra mass and, and for that matter, some extra, um, dark energy.
Uh, and in particular in the case of cosmology, we know, uh, from the early universe that the abundances of the light elements, um, the isotopes of hydrogen, uh, helium and lithium, uh, we can understand that as, uh, the universe having passed through, uh, a, a, a phase in which it was one giant nuclear reactor. And we get the, uh, observed elemental abundance, right, for a particular density of, uh, bariums, uh, neutrons and protons, normal matter that make up most of the mass.
Uh, and that is maybe 5% of the critical density, which is the over under, uh, between expanding forever or relapsing. And yet, if we go out as astronomers and move, measure how things move, and large scales and try to estimate the total gravitating mass density, uh, then it's much bigger than that, uh, density of normal matter. Uh, so we infer that there is extra mass in both kinds of systems and that mass visible to us, and it have participated in this, uh, early big bang nucle synthesis.
So we infer that it is some kind of new particle outside of the standard model of particle physics. - So where does the modified new dynamics come into - This? So, um, so that, that background about the dark matter is sort of where I came to the problem from and where I, where I think most, uh, scientists come to it. Um, but I was mostly interested in those astronomical objects.
In particular, a class of things called low surface brightness galaxies, which, uh, when I was a young scientist, were, were new and not much was known about them. Uh, and I, and so a low surface brightness galaxy is just a, a, if you think of a pretty spiral galaxy, it's, it's just take that and stretch it out. It's, uh, has a much lower average surface density of stars. Um, and so it's not as pretty, but still a galaxy.
And so, um, I had ideas about how those, uh, would form in terms of dark matter other people did too. And, um, none of those ideas proved correct. And so in studying those things, I, I really got concerned about whether we understood dark matter or if it could even explain these galaxies.
And, and that's a long story because one doesn't consider, uh, something as radical as the, uh, a modification of gravity, the, the so-called modified Newtonian dynamics until you're really convinced that there's a problem with the dark matter picture.
Um, and I became aware of that, or, or I become concerned for the, the dark matter, um, before I became aware that, that anybody had really even hypothesized that, um, at the time I was a postdoc in Cambridge and, um, uh, ma Mai Milgram, who was the person who hypothesized the modified dynamics happened to come by. Uh, so it was a sort of a chance meeting in middle Earth, as it were. And, and at a critical time when I was, um, already concerned about whether the dark matter picture was viable.
Um, and like I say, there's a huge backstory just to that statement, but I'd gotten worried about it. Um, nevertheless, I was still convinced that that had to be the case. And so when I saw his talk advertised on a poster, uh, something about modified gravity, I was like, oh, who wants to hear that nonsense? It can't possibly be right. Um, and so I did go anyway.
And, uh, he basically not knowing who I was or what I did, derived in a few lines on the board, um, a prediction for what low surface bytes galaxies should do in his modified dynamics. Uh, that was exactly what I was observing and, uh, was part of what had caused me deep concern for whether or not, uh, we could explain it with dark matter. And it was basically a fine tuning problem.
So, um, what the modified dynamics hypothesize is that rather than there being a whole lot of dark matter in the universe, that all the discrepancies that we observe are because Newton and Einstein didn't teach us everything. And so there's a change the force law. Um, and then his hypothesis is change happens at a particular acceleration scale.
Um, that's already kind of hard to wrap our heads around because of course, people, uh, early on when, when the dark matter problem became obvious, um, people did think, well, maybe it's a change in gravity. Uh, and the first thing our brains think of is size, right? Cosmic scales are huge. The galaxy is huge. The solar system is huge, but it's tiny compared to those other things.
So it's easy, uh, to imagine that, okay, Newton's inverse square law law and, and Einstein's additions to it work perfectly in the solar system. But maybe when you get to the much, much bigger scales of galaxies, some intermediate scale there, the, the force law changes, um, that does not work. Um, you, you cannot make a, a length scale dependent modification, uh, to the law of gravity and, and fit it to all the data. Um, so people pretty much gave up on that pretty quickly.
But, but Milgram like, I guess he suggested, well, there are other scales, and the one that seems to work is this very low acceleration scale. Um, and so he had hypothesized, um, basically he wanted to cook up a theory that made rotation curves flat, which was the, the hot topic at the time.
Uh, but once you do that, once you write down a, a modified force law, it's, it's an equation and, and you're stuck with it, and it makes lots of predictions and, um, acceleration follows from surface density. So I had inadvertently done the perfect, um, experiment to test this theory. , I didn't know about the theory when I, I was , I was interested in the low surface brightness galaxies, um, but they were low surface density, and so that meant low acceleration.
So they were strong tests of, uh, his theory. And in fact, after, uh, I hearing his talk, I went back and read his original papers, and I found a, a quote in there that said, low surface brightness galaxies will provide particularly strong tests. And I was like, wow, I have now the data to falsify this stupid theory, , that that was my first thought, because I was coming at it from this dark matter picture where that had to be true.
And even though I was already worried about whether it, whether it could work or not, I still assumed that this, this strength theory had to be wrong. Um, now the concerns I had were because of the fine tuning, and again, a long story short, that fine tuning comes about because galaxies, uh, turn out to obey this modified force law that Milgrim wrote down.
And not just in terms of having flat rotation curves, but in being able to map the details of the observed mass distribution to the kinematics. That is how the entire shape of the rotation curves was, how, how stars and gas moved in galaxies. Um, and so what I said before is that, you know, we have the dark matter problem because what we see is not what you get. Um, but it is in this modified force law. Um, and what since then is, you know, 20 years of hard work.
But, um, we've been able to show that that's true empirically in galaxies, um, without reference to, to milgram's particular theory, that if you measure the mass distribution, uh, of the stuff you can see, then it maps, uh, directly to the total kinematics. Whether that's dark matter or this modified gravity or something else, we have not yet dreamt up. Uh, and that's a critical part of understanding what's going on. That is unanticipated in dark matter theory.
And as yet, not satisfactorily explained by it, - We'll hear much more from Stacey later in the podcast. But while Professor Magor has moved from studying dark matter towards mon, Dr. Adrian Nik has moved in the other direction. - So, um, I am a postdoc in the, uh, department of, uh, physics and astronomy at St. Andrews. Um, I work on, uh, right now I work on testing the idea that we are living in a local under density or void, and this could, um, perhaps solve the Hubble tension in cosmology.
Um, but originally this, this postdoc was about, uh, implementing the, um, wide binary test of gravity at low accelerations. Um, and that took, uh, first couple of years. Um, and that's something I've been working on for much longer is the idea of grooving dynamics Ormond, um, which, uh, is about, um, galaxy rotation cuffs, perhaps being understandable without any dark matter at all. - You've changed your mind, right, as a result of the evidence.
- So if, if I just work o outwards on from the smaller scales to the larger ones, and that, that I think will make most sense. So, uh, the first thing, um, well, uh, the, the smaller scale thing, uh, failure was with regard to the solar system FM fides. So, um, because modern acceleration dependent sort of theory, um, the deviation from Newtonian mechanics sets in at only about a 10th of a light year for an individual star.
Uh, it, it's, yeah, uh, which is still very far compared to all the planets. But the thing is, um, it's not that much further out than, uh, the gas giants. It's like 7,000 astronom units. Um, whereas, uh, in particular Saturns are 10 astronom units. Um, so you could imagine that if you're able to make very precise measurements, you might be able to notice some, some mon effects.
Um, and it turns out that, uh, a certain anomaly is predicted in mon, um, which, uh, is inconsistent with the observations that nine Sigma confidence. Um, this is based on a precise, uh, radar pulse is being sent to the Cassini spacecraft with, or which is Saturn for many years, and finding exactly the distance between Earth and Saturn, uh, which basically I is, uh, constraining the sunset and distance.
Um, so yeah, we, uh, that, um, the, there was a very detailed study on this recently, looking at different possible adjustments to the mon, uh, theoretical parameters and getting the consistency between rotation curves and solar system IFM rights. And yeah, this, this can't be done.
Um, this, uh, um, uh, the, the other thing, um, that came up, uh, so I'll talk about the white binaries last because that's more complicated, but the other thing that came up was the, uh, galaxy clusters, um, galaxy clusters, um, of course, uh, one, uh, already problematic in the sense that they needed extra mass. Uh, but, uh, this argument didn't really convince anybody, because generativity also requires galaxy clusters of extra mass or nutrient gravity can't work either in.
But, uh, the thing is, um, galaxy clusters have less gravity than predicted by mon in their outskirts. Uh, and that's only become clear recently because we've sort of got observation extent out further. But, uh, uh, there've been a few studies now which show that essentially galaxy clusters seem to follow the inverse square law, even in the low acceleration regime.
So until recently, we didn't have data that went sufficiently far out from a cluster to get into the low acceleration regime, but now we do, and it's clear that they're following the inverse square law. Um, not only that, but they're following the inverse square law with, uh, with, uh, but the wrong sort of normalization, if you will. So they're following it, uh, as if the Galaxy cluster has six times as much mass as the visible.
Um, however, in the standard cosmological paradigm that makes total sense because it's supposed to be five times as much dark matter as a visible. Um, so, uh, in the outskirts of galaxy clusters, which are, by the way, it's so far out that, um, there a mass enclosed interior to that radius. Uh, it, I mean, the proportions of variance to, to dark matter should be the same as the cosmic one.
Um, it's not realistic for bar to have been blown out of a whole cluster, uh, even though for individual galaxies that could happen. Um, but for whole class it's unrealistic. So, so it is always known that galaxy clusters, uh, should get sort of around this, uh, bar feedback issue if you had measurements far out, which recently can be available.
Um, the other thing to realize of course, is that because in qualo, even if the normalization is higher, uh, ultimately the in qualo will end up like giving less gravity than the sort of inverse distance operation in wand. Um, so because of this, um, ultimately, uh, it recently has become clear that there's less gravity in the outskirts than than one predicts just on the, the basis of the visible mass.
But because you also need to add a lot of dark matter e even in m to explain the central regions, um, it's clear that mon substantially over predicts the gravity in the outskirts, and that cannot be, uh, reconciled because you'd have to add like negative, uh, dark matter for that to work. H historically, actually the cluster result came first and then the wide binary result, and then the solar system one, uh, in case people are wondering.
Um, but, um, of course they're all in a very short, uh, period. Um, but, uh, regarding the wide binary results, um, so this is look at an intermediate scale of, uh, about, um, 0.1, uh, light years. Uh, the idea here is that, um, binary stars, uh, with reasonably wide separations, um, in the solar neighborhood, uh, should orbit around each other 20% faster than the neuturing expectation, uh, if mon is correct.
Um, so, uh, that's, um, quite a lot compared to the planetary separations, but it's much smaller than the typical separation between stars. Uh, so there's, uh, plenty of, um, wide binaries. Uh, in fact, the nearest start to the sound is actually itself part of a white binary. Um, so these are not that, um, rare.
- The European space agency's Gaia mission has been surveying the stars and giving much more precise data than has ever been available before Ira and his colleagues have turned to that data.
- So, uh, what we did is we built like a detailed sort of statistical model for wide binaries, which, uh, I've been working on and off towards since like, uh, late 2017, um, calculating the, in detail, the mon predictions for how wide binary orbits should behave in mon, also in the Newtonian case, uh, then we had to add, um, the possibility of a third star orbiting one of the two stars in the binary, and actually a fourth star, because there could be a closed binary around both
of the stars in the wide binary. Uh, but these would be on, on very short orbits, and you could just focus on this motion of the sort of closed binary as a whole, uh, around the more distant, like wide binary companion. Um, and this only in the cases where you're not actually detecting the closed binary, because if you are, then that would be removed from the sample anyway, um, but still, um, by doing a very detailed analysis.
So we, we fixed the, uh, analysis protocol as much as possible in advance of getting the actual data. Um, we posted like a detailed like, uh, blueprint for how the test should be done in autumn 2021. Um, and we were able to more or less prepare for the actual data because there were previous guide data releases, which weren't quite accurate enough. Uh, and these more or less showed us like how to get, how to do the analysis.
Like, um, the actual analysis, uh, was sort of something we did as quickly as possible after the guide h release three came out in June, 2022. Uh, and I explained to the other authors that we would most likely have clear results by the end of the year, um, that that actually turned out to be the case.
So on fifth November, we found out that, uh, there was an approximately 20 Sigma failure of mon, um, but the observations were consistent with the Newtonian, uh, prediction within one or two standard deviations. Um, so that was then written up, uh, subsequently and, uh, that's been published, uh, recently.
Um, so there were, uh, there was also another study by, um, a couple of researchers in Queen Mary that, uh, were authors on, on my paper, but before I did that they managed to publish their own, like much faster, uh, sorry, much simpler and, and much faster analysis, um, which should have, um, got to the, uh, it should have captured the physics, but, uh, effectively it's like calling someone on a, on a break phone instead of like a smartphone.
It'll probably still work, um, but it's, it's a bit less, uh, ideal. Um, so essentially what I did is we explored the parameter space much more thoroughly, whereas they just found like what was like the best model. Um, but, but still this was sufficient to see that there was a very strong preference for, uh, Newtonian gravity of a wand. Um, that's why my results were not, uh, completely surprising.
Um, but nonetheless, it wasn't clear until we did the analysis, uh, uh, that we would actually get the same results. It's always a possibility that it would somehow be different. Um, but no, it, it is all quite consistent. Um, we also like, um, found out, uh, the main problem with some other groups that claimed there was actually a signal in the white binary test that looks like Bond. Um, we borrowed the dataset from them, and, uh, we identified like what the problem was.
And that's explained in, um, one of the sections in a paper, which is, um, uh, titled what comparison with other, uh, you know, implementations of the White binary Test. So, yeah, um, the, uh, it's, it's very clear that wide binaries are Newtonian, uh, to quite high precision about like a percent or so. Uh, and that's why the mon prediction is completely incompatible with the observations. Um, it's also totally in line with the solar system result.
If you were to as believe mon is correct, uh, and ignore all the rotation curve results, but just, um, believe the casini radio tracking result, that would automatically show that wide binaries need to be Newtonian, um, to like within a percent or so, um, that wouldn't by itself disprove bond in that case, that would simply prove that wide binaries need to be close to Newtonian.
Um, but then if you combine it with rotation curve results, um, then you would see, uh, the failure of, of mon um, yeah. So it's these three failures, like essentially, like in in context, what that means is you can't extend mon to even slightly smaller or larger scales in the Galaxy
regimes for which it was designed. Okay, - But so you were previously a proponent of, I mean, you were looking at it, uh, uh, tell me how it feels to you when you start looking at it at something like that and you go, oh, hang - On. Well, this is quite a shock when we found out like, uh, that, uh, it, it didn't fit the wide binary test.
The, the thing is like, uh, because I, I knew sort of the risk of moral hazards was very high, that's why we, uh, tried to take mitigating measures, you know, so we'd agreed, like, um, so basically there's a Gravity law parameter in the code, which is like zero for Newton and one for one, but it's not like an inte, it's a floating thing. So it, it, it could have values between minus two and plus, uh, four, uh, plus 3.6.
So, uh, what we decided in advance is basically I either would give zero, it would give one, and the other value would suggest this something not quite right with analysis. Also, we agree on what some kind of variations to the analysis that should be done. And obviously if the result, um, more or less holds up when we change these assumptions, um, that would be good.
Uh, then some other authors suggested some other, like less complicated, uh, simple sort of ways of looking at the data, which should also clarify if Terrine mechanics is preferred or, or of Mons preferred. Um, and all these things gave very consistent results.
So, um, yeah, for, in terms of how it felt like, uh, yeah, this, this was quite bad , um, I mean, there's no way to describe it any other, this was, um, yeah, especially in this scenario where, uh, it wasn't really clear what, uh, I would work on next. And, uh, um, but more generally like, um, because, uh, it seemed likely that other problems would come and, uh, yeah, that, that this was very bad situation. - Is that kind of like an overnight thing?
I mean, how long was it in that process before you were sort of going, okay, I've really gotta stop? - No, no, it's, it's basically an overnight thing on the 5th of November, 2022. - Really? Okay. Did you sleep well that night? - Um, , well, uh, I, I I, I can't remember like, uh, but ba basically, um, because we'd agreed like the, all the, um, analysis procedures in advance and there was a very clear result, um, then, um, there was, um, essentially no way around it.
Um, it was just very precisely on the Newtonian value, uh, rather than the mod one in principle, any other value could have risen theoretically, um, like if there were problems of some sort of the test or, or if something else was going on.
Um, but, um, in that scenario, uh, why the code just sort of prefers the, the Newtonian result, um, especially given there was some previous hints that the outcome would be Newtonian, um, because of this, um, yeah, there, there wasn't really any, uh, way to like, it, it was clear at that point that, uh, this, this couldn't have been, the mon couldn't have been correct - Clearly. I wanted to know Stacey's response to this study of the binary stars from the Gaia data.
- Okay, so let me provide some broader context, because like I said, my first thought was this, this stupid theory I will falsify. And then when I actually went through the predictions, it was like, check, check, check, check. And, and, and I had corroborated all the predictions that Milgram had made while falsifying my own. I had my own dark matter based theory, and it was wrong. It did not do the right thing. And so, you know, what was I supposed to say that, that he was wrong?
Um, but there are lots and lots of other things besides individual galaxies. So I spent several years reviewing all the evidence, and I mean all the evidence I did, uh, a tremendous amount of, of work about this. Um, and for the most part, not all of it, but for the most part, all the systems in which we inferred dark matter ma worked perfectly well, sometimes better. Um, and I would say in fact, well, I had been working really hard to explain my way out of the hole.
I had dug with the data in terms of dark matter. And as soon as I allowed myself to think of all these problems in terms of mon, I found I was working less hard. And so what I found at the time, and this was the nineties basically, um, and I've reviewed it many times since then. So the statement still holds that you can explain maybe 80% of the data with ma depending on how you want to define, uh, percentages. And it really depends on how you choose the way the various lines of evidence.
And so the other 20% just didn't work. And the big thing that didn't work at the time was clusters of galaxies. Um, and that's, that's still true. Um, and that, you know, that gives me concern about whether or not it can be true. But what I don't accept is that we have to automatically default to dark matter. That just was the first thing we thought of. Um, and we do have to understand why mon gets so many predictions, right?
Even if it's wrong as a theory, and we're not really engaging with that as a community. Um, most people seem to retain the attitude that I had, that we know it's dark matter. This theory has to be wrong, and we just grasp at the first straw to say it's wrong, um, and then never think about it again. Um, so Inverne is not one of those people. Um, he took it very seriously on very seriously from the beginning.
And he constructed a, a test, which should be, um, decisive, and that's to do with wide binaries. Basically hunt through the Gaia database, identify pairs of stars that are, are wide binary candidates, measure their motions also with Gaia, and see if you can see a statistical signal. Basically you should see an enhancement of orbital velocities relative to the Newtonian case. Um, and Indra Mill saw nothing.
So he, you know, and, and I think he went into it expecting to see it, um, and this would prove mon and in fact, he, he found the opposite. So he, he correctly said, no, that doesn't work. Whether or not that amounts to a falsification, uh, depends on what I said before. It depends on how you weigh all the different lines of evidence. If you put a hundred percent of the weight onto ev, uh, in Nils experiment, then yeah, it's a falsification.
If you put a hundred percent of the weight on the experiment I did long ago on low, low surface brightness galaxies, it's as close to a confirmation as you'll ever get - Under re reputation of dark matter. - So Malin, I said is binary. It either works or it doesn't, and it works way more often than it should if it's wrong. Um, dark matter isn't like that. We often do not have a well defined prediction that comes out of it. It's more of an inference. So we see things going wrong there.
So there's some dark matter, okay, well how much, how much should there be? What do you actually expect? Um, and if you're only talking about it at the level of, oh, there is a discrepancy from what we see, so there's, we infer there's some dark matter, it always works. Um, if you actually try to predict something, then it goes badly wrong in a lot of cases. And one thing that you really do not predict is that galaxy should look like mon.
Um, basically if you tried to build dark matter models, then you have a Newtonian, um, collection of stars and a Newtonian dark matter halo, and you don't know what the mass distribution of that halo is. So you have to make some assumptions in order to try to predict something and build a model. And you can make a whole bunch of assumptions. And what you find is that there is this enormous parameter space of what galaxies and, and other systems might plausibly do.
When you have both luminous stuff and dark matter mixed together in mon, there's only one thing it can do. It has to follow this force law that was written down. And for some reason, nature always seems to be plucking this mandian, uh, appearance out of, uh, dark matter universe if that's how the universe works. And that's, uh, you know, in a sort of a basian statistical sense, that's just completely improbable. It should never happen. 'cause there are all these other possibilities.
Um, and the only way it happens in that conventional dark matter context is if you make it so right. And so there are lots of models out there where after the fact I've told them, okay, the data do this, and then the modelers are like, oh, yeah, I can do that. But they never did it in advance, and they can't actually use that to predict anything.
Whereas mon, I can take and look at a galaxy and say, okay, I measured the mass distribution, I can predict what the motions will be, and you know, nine times outta 10 I'm correct. Um, and I'm more interested in that nine times outta 10 than I am out of the one. You know, they're astronomical data are goofy, they're always some weird cases. Now, Indra no founding opposite with wide binaries, and I take that very seriously.
Um, so maybe that is informing us about the deeper theory, not necessarily that mon is wrong, but we know it's not complete as a theory, and so maybe there's some theory development to be done there. Um, but it's, it's a little worse than that in that there is as yet no consensus on the wide binaries. So, um, both Xavier Hernandez and, and independently, uh, Kahu Kai have done the same experiments with Gaia, basically the same data looking for wide bearing areas.
And they both claim to see a mon like effect. So where Indra nil doesn't see it and rules it out, they say they do see it and must have it. And I don't know who to believe. - If you don't, can't sort it, how can anybody else know what the answer is here? How, how can anybody look at this and go, oh, well, that's what's going on? - Well, I think it's like I say I, let me step back. There have been controversies like this in astronomy since its inception.
Um, in modern times we now have the Hubble tension, right? Is it 67 or 73? Um, when I was young, that was, is it 50 or is it a hundred? And, um, it's really hard to sort through that sort of thing took decades to solve the first one. It's may take as long to solve the new one.
Um, and the ideal way to do this is for all the principles to get together, set and agreed rules that they're all going to play by and then compare data, take notes, and agree that they've all come to the same answer, uh, in the proper way. And, and then somebody has to admit they're wrong. Um, that is the sticking point. Very few people are willing to do that. Um, and, and you know, I would say I would do that. I had a prediction.
I tried hard to fudge it, and I had to admit that it was just wrong. I what I predicted for low surface brightness galaxies specifically, that they would shift off of the Tulley Fisher relation that didn't happen. Um, and so, you know, I would prefer that they sort it out among themselves. Um, now, - But, but hang on. Isn't that what peer review does?
I mean, isn't that essentially - Oh, oh my, um, in principle, yes, in practice, um, we are all people doing these things and we bring our own baggage to the reviewing process. And, um, I've been on both sides of that many, many times, and I've seen all sorts of strange going. And so it, it's a, it's a bit like, I think it was Churchill, right? The peer review is, is the worst system, uh, except for every other, right?
I mean, you need some method, um, to, to, to review what goes into a journals and what counts and what qualifies. Uh, and that is the best method that we have. But it's certainly not perfect. Um, and you have these controversies because people, um, you know, are willing to accept publication. In fact, you know, for astronomy, the, the acceptance rates of most astronomy journals, the good ones are pretty high.
Um, and part of that is because we have a long history that is informed by exactly these kinds of controversies. So, you know, I have many times gotten a manuscript, a review where I'm like, I don't believe this, but, but I could be wrong. And so if, um, you know, the, the authors make a good case and they maybe have a kind of data that I haven't seen before, then I will usually recommend that for publication, even if I seriously doubt it.
And of course, I will explain that in my report, and there is some back and forth, and sometimes you can get people to moderate what they say and sometimes not. Um, and it's pretty rare though. It does happen when you can, when you get a paper that is, oh, this is just wrong. Um, and, and even then sometimes they get published because of this long history of getting things wrong. Um, I can think of cases where papers I think should not have been published, did get published.
I cannot think of any case. Um, and I've done hundreds of these at this point. I can't think of any case where a paper that was really good did not eventually get published. I can think of cases where, uh, it had to be shopped around to different journals, um, before it got accepted. - Uh, that's interesting. Has it always been like that? - You know, 20 years ago, 30 years ago, the only people submitting papers to astronomy journals were astronomers. It was a modest sized community.
If you didn't know people, you knew of them. And so there was a basic presumption that papers were being submitted in good faith. Um, they at least tried right , even if they got it wrong. Um, that's not really true today. There's a, a whole bunch of predatory journals.
Um, there are people usually physicists, uh, people trained mostly as particle physicists who suddenly think they're qualified to do astronomy writing papers that make some of the basic mistakes that the astronomical community made 50 years ago. And you point these out as a reviewer and they're like, oh, I never heard of that, so it can't possibly be right. It's like, no, no, really, we've been through this before. You're, you're making an obvious mistake.
Um, but the editors are also physicists who never heard of whatever that issue is. And so it gets published anyway. Um, and, uh, some of these papers are not even offered in good faith. Um, and there's a range of ways in which that happens. One is just ignorance of the history of the field. And so it's like, oh, you, you really missed a huge section of the literature. And if you're trained as a physicist, you don't even know what journals to look those up in.
So it can be an honest mistake. Um, but it is a mistake. And, um, there are also people, especially on this subject of, of the modified dynamics, um, you know, if you're a particle physicist who think you're gonna get a Nobel Prize for discovering a dark matter particle, you know, you do not want to hear that there's some other solution.
Uh, you just don't. Uh, - But isn't that true of if you think you're gonna get a Nobel Prize for proving mon that you wouldn't wanna - Hear, I don't think anybody's gonna get a Nobel Prize for that. Um, not anytime soon anyway. Um, and and I, I, I laugh at that because I very consciously chose to report what I thought was true in the data, knowing full well that I would suffer a social penalty for doing so. Uh, and had I realized how steep that social cost would be, I might not have done it.
And a lot of people choose not to. Um, so it's, it's funny, I have conversations with people and they will say different things sometimes in private than they do publicly, um, because it's a controversial issue, and they don't want to be seen to be on the wrong side of that. Um, so I may be wrong, but I'm, I'm certainly being honest about it, and that's why I say I don't know what the Y binaries are teaching us.
I know that Inverne is convinced that he's disproved bond, and I know he has made an honest attempt at doing that. I just don't know if it's right - , I asked Dr. Adrian Banick if he was surprised that people were still looking at mon as a possibility. Well, - Uh, the thing is, um, you, you can try to circumvent the falsification by changing the how, how the theory works. I'm not surprised that there'll be some people working on it.
First of all, like when I start understand people who've worked right for many decades would, in any case, not easily give it up. Um, but more importantly, um, I would say it does work well in the rotation curve, uh, was to to predict galaxy rotation curves.
Um, and, uh, in our case, for example, uh, when we do cosmology, um, we often rely on sort of calculations that use more, but as a, a proxy for what could happen if gravity was enhanced in some way, uh, just to see what that kind, how that kind of simulation might differ. Even though by now it's clear that the extent of any possible modification to gravity is much smaller than Bond gives. Uh, but it, it's, uh, helpful as a sort of an extreme sort of case.
Um, but in terms of people working on bond, uh, there's actually much fewer people working on one now, in, in terms of doing like numerical bond simulations, uh, a lot of the people working on this in v uh, have left. Uh, that's just because of the PhD's finishing. But the thing is, like, it's not, they, they hire new PhD students to work on bond. Um, so, uh, that, that's, uh, sort of thing. And, uh, certainly there aren't very many postdocs working on bond.
Um, and the postdoc stage especially, you'd have to write like specific grant proposals. And this can't really, it's not really investible, uh, at the moment. Uh, obviously I'm not entirely sure if it, if it will be in the future, but, um, that seems very unlikely. But certainly right now it's, it's not. Um, so yeah, there's also much fewer in terms of early care researchers working on one.
There's very few, there's, there's people who already have tenure, a lot of free time and can afford to work on one regardless of what other people think in terms of people who actually have to write a grant to work on wand. Uh, you had that in the past, but, uh, you don't really have that now. A apart from, um, obviously, you know, a handful of people who got grants before it was falsified and then they're still working. That's, uh, one of the exceptions - More from midrail soon.
But let's go back to my conversation with Professor Stacey McGill. But if it is, and he is disproven bond for the wide binaries, does that, does that mean you abandon it and you look at something else? Or, or what does that mean for you? - That is a great question and I wish I had a great answer. Um, because I have spent a lot of time trying to find that intermediate case, right? So, um, and there are hybrid models out there and they all sort of feel wrong to me.
And, and I don't wanna go into a great length about each one, but I've tried to do that and I have failed. Um, but ultimately, like I say, one needs to be able to explain the phenomena that ma does predict correctly. And there is a lot of that. And it does not sit well with the normal dark matter picture. I mean, uh, you can find people who say, oh, yeah, I, I can do that, but no, I've been through that. If I thought I could explain it easily with dark matter, I would do so.
Um, I am still personally way more comfortable with that. 'cause basically that's what I grew up with. Um, and I would also love to stop having this argument with people. It gets old. Um, but it's, it's just not true. Um, it is very hard to explain, uh, the phenomenology of galaxies and a lot of other things with our conventional notions of non-BAR cold, dark matter, just some new particle that's not good enough.
'cause what we see in the data is a clear connection between the distribution of the stuff we see and the dynamics that we get. And so the dark matter knows, uh, what the normal matter is doing. There needs to be some direct link between them that is way stronger and more compelling than just gravity, which is a really weak, um, force. It's, it doesn't organize things.
And so, so to give you an example, a lot of times I will make this point, oh, the, the barian tail is lagging the dark matter dog. That doesn't make sense. Uh, 'cause the, basically you can have a dark matter distribution that does what you need it to do, but you can only do that 'cause you don't know. You can't see it. So you just say, oh, it is whatever I need it to be. You can't predict it. And so the, basically the luminous matter is telling the dark matter what to do.
That doesn't make sense. Um, 'cause there's a lot more dark matter. So people say, oh, well dark matter tells the luminous matter what to do. It's like, okay, that sounds wise, but actually it's stupid because you cannot use that statement to construct a model that predicts anything successfully. I know I've tried, I've tried really hard. Um, and so it's really the difference between, um, taking some data and fitting it, that's easy.
I've also done that you can always find a dark matter model that fits once you're told what the data is. The challenging thing is to make a model that predicts what we see in advance, uh, and that nobody has demonstrated so far - Mm-Hmm. Does the fact that we can't use something to predict it mean it's not true? Or does it just mean we just can't predict - It? Well, so no, and that's one of the frustrating things.
So at one point, oh geez, almost a decade now ago now, I wrote a review in which I said that the two things were in commiserate and they really are in commiserate. Um, so yeah, just 'cause something can't predict it doesn't mean it's wrong. Um, on the other hand that something does predict it, we usually take to be an indication it might be right. Um, and that is basically the answer to the question you posed is what would I do next? Well, that's what I've been trying to do.
And what I've been trying to get other people to do is like, okay, I am open to an a dark matter interpretation, but you actually have to demonstrate how this mon like force law follows from that. I have not been able to do that in a satisfactory model. Um, in fact, I thought at several points in my career I had done it and I realized I had assumed something along the way that made it so it was basically a tautology. Um, and I see lots of models being published that say, oh yeah, we do this.
And I look at them, it's like, oh yeah, I made that tautology 20 years ago. You're just making the same mistake. Um, and you know, so one has to actually demonstrate that with a satisfactory model, and that's a much higher standard than its sounds. Um, and people are not doing that in large part because they're already convinced it has to be dark matter. And so there's a very low bar, um, to trip over to say, oh yeah, I got that right. Can you give me an example of that?
Uh, so the standard dark matter halo is this thing called an NFW halo that specifies a particular density profile. Um, and it has its problems, but if you take that and the stellar mass halo mass relation that people think work and put some bunch of exponential discs in there, which is the standard sort of approximation we use for the luminous parts of galaxies, then it almost adds up to something that I call the radio acceleration relation.
Um, which is the sort of manifestation of mon in a conventional sense. If I didn't know about mon then I could still say this relation is there in the data, and it happens to look exactly like mon, but let's do it some other way. And that, that modeling procedure I just described almost works. Um, and so if you don't look too closely, then it's like, oh, that's fine, and you go on your merry way.
Um, but it has some assumptions built into it like that still a mass halo, mass relation, a a relation between how big the dark matter is, halo is for a galaxy of a given, still our mass. Um, that's not native. We already had to fit that to data. And in fact, that was one of the things that gave me problems with the dark matter paradigm in the first place, is that the obvious thing to, uh, assume was that there's a one-to-one relation.
You know, basically there's a fixed ratio between the amount of dark matter and the amount of luminous matter. Uh, and that is such an obvious thing, uh, to assume that we assumed it all through the 1990s and we fought really hard to not do that, right? We did not want to abandon that. Uh, and by the turn of the century we realized that we had to do so. And one of the reasons is the kind of data that we've just been talking about, that it simply doesn't work.
And so if you now 20 years later make a model that assumes, um, that there is some magic stellar mass halo, mass relation, well, you're already assuming a large part of the answer that was established as something that didn't work a long time ago. Um, it's just been long enough we've forgotten that, that we've granted ourself this role in fudge factor. Um, even after that, it still doesn't work, right?
But I, I mentioned this particular case because, um, Julio Navarro, who's the inn in NFW Halos, built exactly a model like this, it's like seven years ago now, and it's like, and, and said it worked. And I'm like, didn't I try that? And sure enough, I went back and looked and, and the last time I had touched the files was in 2000, but it was exactly the same kind of model, and I had dismissed it as not working.
And the reason was what I just described, because I assumed something different about the ratio between luminous and dark matter. Well, I could update that and then I could reproduce his result, but I was just, it's another tology, right? I was choosing, uh, relation between the masses, the dark matter halo, and the luminous mass that made it. So, um, and that kind of thing is now so deeply buried in our psyches. It is very hard, right? We've gotten way, way deep in this rabbit hole.
You know, once you're down a rabbit hole , you don't wanna change, right? It's, - But I've spoken to two people today who have done exactly that, who have changed, right? But in different directions. So, so people do change. They do look at the evidence and change, but you've gone different ways. - , fair, fair, you know, some of us do, but it's actually in minority in my experience. Um, and it's, it's even more complicated than that.
'cause for the most part in, in the community, I think, like for me, dark matter was something, oh yeah, we need that. Uh, when we make galaxies, I wasn't really concerned about that. I knew it was an element, but I was specifically trying to understand these low surface brightness galaxies. Um, and I did not care one way or another about there being dark matter. It was just one of the ingredients that goes into these models.
Um, and so the data made me think carefully about that in a way that I simply hadn't been obliged to do before that. Uh, and I think that's where most people are. It's not that they, uh, there's simply not challenge to change their minds. It's not that important to what they're doing, uh, day by day. I mean, same thing for cosmology.
Um, you know, if mon is correct, and we do not understand a lot of the things that we think we understand in terms of, uh, our modern Lambda CDM cosmology and, you know, putting on a cosmologist hat, I, I get why we arrived at the model we are at. And there are a lot of compelling reasons for that.
Um, you know, putting on a mon hat, I'm like, well, yeah, you got this invisible mass plus now this dark energy and, you know, , these are the sort of ancillary hypotheses or, or tooth fairies if you want, that you have to invoke when a theory is on its last legs. And it's conceivable that we're doing that in order to use the theory to approximate some deeper unknown theory that hopefully would encompass both general relativity and mind.
But we don't really have that theory and there aren't really any people working on it. And so who knows, you know, maybe it's impossible, but, um, we haven't really tried that hard. - I wondered if IL thought that we do still need something else, something that isn't dark matter to explain the way that things are.
- Well, if you look at scales smaller than a, than a MegaPath and, uh, leave beside the cosmological tensions for a moment, then, um, for the galaxy scales and, um, and, and larger, uh, cluster scales, then, uh, I think you probably do need something else, as in you need some adjustments to the properties for dark matter, so it doesn't cluster so efficiently. And in terms of the, uh, tightness of the radial acceleration relation, um, yeah, I'm, I'm still not entirely sure how that would arise.
There may be, for example, a hidden or an unknown, maybe an unknown interaction between the dark matter and the barons. Uh, maybe, uh, that helps to keep the two in sort of line and following the RAR maybe. Um, so yeah, there, there is some chance that you, you need some additional ingredients.
In fact, it's very likely because, um, for example, if dark matter reached the very high central densities that are predicted, um, many problems would arise observationally, uh, including the rotation speeds of galaxy bars in the centrals of galaxies becoming slowed down by sort of gravitational friction against the dark matter. Um, which doesn't really happen. Um, but, uh, I I, I do think that, uh, yeah, there, there's a decent chance, um, you almost actually do need some additional thing.
It can't be purely the Lambda CDM model, even on the galaxy scales. Um, yeah, but the basic picture of the dark matter halos plus the an inverse square gravity law, uh, being the, the main sort of things that main, main thing that's going on that I think is probably correct. Um, it just, it might need to be, uh, tweaked in some way. - What could possibly bring mon back into your kind of like research area. You thought, oh, yeah, no, I will look at it again.
- So, uh, the original mon field equations have been falsified in high significance on, on the galaxy scales. Uh, if, if additional tests of the so-called, uh, dynamical friction may help to clarify if, if there is dark matter. So there have been some studies recently which show that, um, essentially as the large magnetic cloud is mo orbiting the Milky Way, it, it seems to have experienced dynamical friction.
Um, but, uh, studies to try and clarify like this issue further, uh, and also sort of how much the milk curate disc has recoiled due to the large mag cloud, which places a limit like it, it measures the mass ratio between the two, right? Um, and that's actually very different to the ratio visible mass. Uh, it's, it's more in line with, uh, the ratio it's expect in the dark matter picture.
Um, but, uh, measurements like this, uh, in, in theory, if someone were to, uh, if, if these measurements would somehow to turn out to be not correct, and actually the mass ratio between the milky and the LMC is about one to 20 ish, which is the ratio of the visible mass instead of the one to five ratio that's observed and also more or less expected in the standard model, um, then, uh, that would obviously provide compelling evidence.
The galaxies are purely bar, uh, in theory, uh, that that could happen. Um, and, uh, things like if you managed to prove that even such a massive galaxy as LMC or something that should be so massive in the standard model, given it's stellar mass and typical sort of stellar to dark halo mappings, um, if that experience like no dynamical friction, uh, or new place, really tight upper limit on it.
So there's some ways in, in theory that code, um, maybe, uh, show that you don't have much dark matter around galaxies, um, which in turn would, um, show that, well, if you have to make do with the visible mass, you, you'd have to modify the, the law of, of gravity or invent additional forces in in some way. There is a very high likelihood of additional problems with the Lambda CDM on even larger scales with, with regards to the Hubble tension.
Uh, and, uh, this is what I'm working on now, the idea that structure formation, uh, and perhaps the law of gravity most likely needs to be modified in much larger scales of, uh, several dozen, like up hundreds of MegaPath. Uh, and, uh, so hundreds of millions of light years.
And what, uh, might happen there is if you form a significant local under density, uh, well, you'd form densities as well, but if we happen to be living in an under density, then locally the universe would appear to be expanding faster than it kinda actually is. 'cause there's additional outward motion due to the void. Um, and that may solve the Hubble tension. Um, we're actually able to recently use a model along those lines that we published in 2020.
Um, it, it sort of had, uh, parameters for like how the voy should work. Um, but using that model, we are able to correctly predict the, uh, a, a key feature of the velocity field in the local universe that was subsequently measured last year. And that's a so-called bulk flow curve. Um, so there, there's, uh, definitely, and, and I should say the bulk flows are also much faster than expected in the standard cosmological model.
So, uh, there's definitely a high chance of some modifications to Lambda CDM. Um, I, I don't want people to think that it's sort of going to, to work, uh, completely without any further modifications. Um, that's, that's not really possible given the current evidence. Um, but um, that doesn't mean uh, that the, um, modern field equations provide a viable alternative. - Let's have one final thought from Stacy. - We do not understand how the universe works.
We think we do, but neither we, we don't know what the dark matter is. We just know we need it. Um, and we don't really understand how mon can be true as a theory, and I think that's a wonderful thing, right? There's still fundamental physics to be learned about how the universe works, and so, um, it's not like science is over. It's not like we're filling in the last few decimal points of, of everything. There's fundamental issues here still to be addressed,
and so that's a good thing. We're, we're not done - Here. I'd like to thank Stacy and Idra nil for talking to me for this episode of the Physics World Stories Podcast. And if you'd like to know more, I can highly recommend Keith Cooper's feature on the physics world website. It is, of course, the first of three, this one entitled Cosmic Combat, delving into the Battle Between Dark Matter and Modified Gravity. The Second Coming.
Soon we'll explore some of Dark Matter's, recent successes and the serious challenges that it's also facing. Thank you very much for listening, and we'll be back next month with something else from this wonderful world of physics, physics world.