Friday, February 15, 2019
Dark Matter – Or What?
Yesterday I gave a colloq about my work with Tobias Mistele on superfluid dark matter. Since several people asked for the slides, I have uploaded them to slideshare. You can also find the pdf here. I previously wrote about our research here and here. All my papers are openly available on the arXiv.
71 comments:
COMMENTS ON THIS BLOG ARE PERMANENTLY CLOSED. You can join the discussion on Patreon.
Note: Only a member of this blog may post a comment.
Subscribe to:
Post Comments (Atom)
Hi Sabine,
ReplyDeleteHave you come across this paper https://arxiv.org/abs/1812.08169 from Mariangela Lisanti's group at Princeton? From measurements of the accelerations of stars transverse to the galactic plane in the Milky Way, they find that quite generically from minimal assumptions it would be quite difficult for a modified theory of gravity to explain the relative magnitudes of the radial and transverse accelerations. I was wondering if you had any thoughts on this as I am a graduate student interested in modified gravity but this paper seems to be not rock solid but strongly suggestive evidence against modified gravity and I can't find a hole...
Olds,
DeleteWe know that MOND is wrong. Why are people still trying to rule it out? It's an approximation that captures some phenomenological relations which are confirmed by data, but it's not a theory that you can reasonably expect to make detailed predictions. I haven't read the paper (will look at it if I find the time) but just reading the abstract I get the impression that the authors miss the point.
Hi Sabine, Olds.
DeleteIt just so happen that I'm one of the authors of the discussed paper.
I find it a little funny that you, Sabine, decide to trash our paper without even reading it.
This paper is actually applicable to every theory/model that predicts amplification to the baryoinc acceleration. It was most easily demonstrated using the language of "pure" MOND vs. NFW, however the next easiest thing would be to do the same for Justin's superfluid. (If my word as a PhD student isn't enough, you can poke either Justin or Stacy, they are both on the same page with us.)
I don't think it is needed to repeat the paper's method here, but I'll just say that the main Idea, as Olds mentioned, was to ask: given the "amplification" needed to the acceleration in the radial direction (s.t. we get the right rotation curve), what is the implication on the vertical motion of gas/stars close to the solar position?. This is exactly where any model trying to mimic MOND should be in its "MOND regime", and thus the analysis should be robust.
Once again, as Olds mentioned, the statistics isn't grate. We always find that the NFW profile gives better results, however it is still complicated to say how significant this statement is. In the future it should get clearer.
I did not "trash" your paper. I explicitly said I didn't read it. I said I do not understand why people are still trying to rule out MOND. Please stop trying to fabricate opinions I do not hold and have never voiced.
Delete"Please stop trying to fabricate opinions I do not hold and have never voiced." Didn't mean to, have I done it before?.
DeleteI'll say it again, we didn't try ruling out MOND, we tried to ask whether MOND-like models can be consistent with MilkyWay dynamics.
Personally BTW, I don't think that, on a galactic scale, a striking evidence against MOND exists, correct me if I'm wrong.
This comment has been removed by the author.
DeleteNadav,
Delete"we didn't try ruling out MOND"
You write in your own paper:
"For the current study, we test MOND using the local profile of the rotation curve..."
You also use a particular equation, eq (2), that you have not actually derived from the approaches you mention in the introduction.
Tell me a model other than MOND that you have ruled out.
MOND is a Newtonian limit. It has been ruled out dozens of times by all experiment that requires post-Newtonian corrections.
I also would like to draw your attention to the question I have been responding to here, which illustrates how the claims in your paper are being received:
"quite generically from minimal assumptions it would be quite difficult for a modified theory of gravity..."
This is the impression that your paper raises, despite the fact that you probe a particular equation which you have not derived in any actual model of modified gravity that is not already known to be wrong. There is nothing "generic" about the equation you are testing. In contrast, it is highly specific.
You write in the introduction about Verlinde's Emergent Gravity and mention Khoury's superfluid, yet you did not actually do the calculation for any of those. Your paper still raises the impression you did.
Hello, Sabine.
ReplyDeleteA couple of questions about your colloq:
Could be the phase change in the dark matter (normal to superfluid) you comment in the slide 23 the result of a spontaneous symmetry breaking in an unknown field that produces something "similar" to a higgs mechanism over another unknown field (the dark matter one)?
And could that unknown higgs "similar" field be also responsible of the dark energy due to an hypothetical non-zero potential in its ground state?
Samu,
DeleteYes, it is the result of a spontaneous symmetry breaking, but not a gauge-symmetry. Using the Higgs doesn't work, it's just too heavy. For the Higgs the phase-transition is at very high energies. We know that from many observations. For the superfluid dark matter you need a phase transition that happens at very low energies/temperatures. The parameters just don't work out.
And what about a new (and unknown) hypothetical "dark" (scalar) field (less heavy) acting like the Higgs but only over the dark matter field (and leaving totally unaffected the ordinary particles)?
DeleteThat's very similar to the Khoury-model, except that for the superfluid you use a different potential than for the Higgs, so that you have a non-zero chemical potential.
DeleteBee, since the topic is MOND and dark matter, can you comment on this?
ReplyDeletearXiv:1901.01228
Conformal Gravity and the Radial Acceleration Relation Authors: James G. O'Brien
In this work, we summarize how conformal gravity can explain the Radial Acceleration Rule in a fashion consistent with the findings of the original authors without the need for dark matter.
arXiv:1812.03152
Radial Acceleration and Tully-Fisher Relations in Conformal Gravity
We discuss how the empirical Tully-Fisher relation follows as a consequence of conformal gravity.
Bee, could conformal gravity be the modified gravity that reproduces MOND and GR that astrophysicists have been looking for?
Dr Sabine H. -
ReplyDeleteLong time reader, first time commenter. Thank you for an insightful & accessible post, as always, & the slides of your talk. They really make the latest ideas on an important problem in fundamental physics accessible to laypersons like myself. just one question on the slides - are there any disadvantages to covariant emergent gravity? there is a slide on the advantages, but it would be nice to know the cons as well.
PS: please feel free to edit out anything unnecessary as you please
curious one,
DeleteI cannot edit comments. I can only entirely delete them.
Yes, you asking for disadvantages is a good point, though maybe let me be clear that the slide with the advantages is concretely advantages over Verlinde's approach. I really think there are only advantages.
If you are asking disadvantages over CDM, I said that during the talk, we don't have a cosmological solution. I just don't know how to solve the equations. We also don't have a UV-completion (second-to-last slide). I am also still not sure if you actually need the vector field. I have a vector field in CEG simply because Verlinde has one. Alas, if you look at the galaxies you really only need the scalar field that Khoury has. Now you can get a scalar field out of a vector, so that's no problem, but do you really need the rest of the vector field? I don't know.
Biggest disadvantage, I guess, no idea how structure formation works. Again it's not so much that I know it doesn't work, but that it's simply a calculation I can't do (read: don't have people/money, same story as always).
Sabine,
ReplyDeleteI am curious as to over just what distances the Newton/Einstein law of gravitation has been actually verified. I mean it seems to be very accurate within the solar system, but at galactic distances it presumably can't be verified because the effect of DM has to be subtracted out of the measurements. The nearest star is about 10^4 times as far away as the distance to the edge of our solar system.
If it can only be verified inside our solar system, is it a good assumption to extrapolate this law to vastly greater distances?
Testing for dark matter in the solar system seems like an obvious thing to do, but it is tough. To get an idea of what dark matter in the standard sense, not MOND, does can be seen with elementary Newtonian dynamics. We should all know Newton's second law of motion for a mass m with gravitation with a large mass M
DeleteF = ma = -GMm/r^2.
We can then reproduce Kepler's third law easily enough with ma = -mω^2 where ω = 2π/T for T the period of the orbit. This then gives ω^2 = GM/r^3, and it is not hard to see Kepler
T^2 = 4π^2r^3/GM.
Now suppose we have a region filled with dust or a gas with some mass density ρ. This would model the dark matter halo. The mass in a volume 4πr^3/3 is then M = 4πρr^3/3. This may also be seen with Newtonian mechanics and Gauss's law that the mass contained in a volume bounded by a shell of area A = 4πr^2 is in a simple setting
∫F·da = FA = (GMm/r^2)4πr^2 = 4πGMm
and with F = mg, here g is used for gravity acceleration, we have gA = 4πGM. Now let me substitute M = 4πρr^3/3 in for the mass M and with A = 4πr^2 is is not hard to see
g = -Gρr/3.
There is a sleight of hand I did here, where this elementary approach I use leads to the loss of a sign that I restored. Anyway this acceleration is Hooke's law for the force of a spring or the restoring force on a harmonic oscillator. If I put in the centripetal acceleration for g this gives ω^2 = Gρ/3. This means the orbital frequency and period is independent of the distance!
Of course to work this for a galaxy both the straight Newtonian gravity and this Gauss law result are put together and things get complicated, though not conceptually strange. We can estimate what the magnitude of that Hooke's law force is with the density of dark plus ordinary matter ρ ~ 10^{-26}kg/m^3. The extent of the solar system to Kuiper belt objects is about 5×10^{12} m. With the constant G ≈ 6.7×10^{-11}N-m^2/kg^2 the relative acceleration between the sun or earth and a body such as Pluto would then be 3.4×10^{-24}m/s^2. The magnitude of this acceleration is too small to consider measuring directly. It might have some measurable influence with chaotic dynamics, but that would require observations of these Kuiper belt objects over almost geological time periods. We can of course do the same across the distance of a galaxy, r ≈ 10^{21}m to get g ~ 10^{-16}m/s^2. This is small but its influence on the structure of a galaxy can be estimated. This was a motivator for the idea of halos of dark matter.
Hi Sabine,
ReplyDeleteThank you for making your scientific papers freely available.
It does not appear to me that super-fluid is falsifiable as you have described it. The testable aspects are more about contrasting with the few alternatives that you have indicated are on offer. If you have doubts about proving Dark Matter, just attack MoND. I fear they are both wrong, but MOND at least has the merit of a predictive rule for Galaxy Rotations, which Dark Matter is the other way around. ie. DM is inferred from Galaxy Rotations. One cannot measure the Dark Matter distribution independently of Galaxy Rotations, and then predict them from that.
The fact that fluid and superfluid Dark Matter can more be where it needs to be seems to be a roundabout way of promoting the beauty of the formulae that infer the DM. The formulae are too beautiful to be the cause of requiring unseeable, unconfirmable matter by their error in context. They always give the right answers, so this invisible but powerful stuff must be the right answer. Maybe the beautiful formulae are wrong in this instance. Maybe the ugly rule of thumb of MOND is our clue that something completely different has to be considered. One of those things you have failed, yet again, to mention is the simple but powerful theories based on Modified Inertia rather than Modified Gravity.
Regards
Marco Parigi
Why would you say it's not falsifiable after I told you how it's falsifiable?
DeleteThe Galaxy Lensing example appears to be hedged criteria, meaning that any results wouldn't necessarily falsify.
DeleteParticles, phase transitions and collision interference patterns are all difficult to interpret and are unlikely to have conclusive results on stated criteria.
It gives me the impression that superfluidity gives more arbitrariness than Cold Dark Matter without it and that any results are likely not to fail the idea on that basis. It reminds me of the SUSY style arguments that we need higher energies. I think superfluid DM is a dead end by the same arguments you use against SUSY.
The main reason the galaxy lensing is not currently conclusive is that the data isn't detailed enough. We have one average value for the velocity dispersion. Still, not all combinations of lensing and kinematic measurements are possible to accommodate in the superfluid model, simply because a superfluid does not form in all cases.
DeleteAs we explain in the paper, there is indeed one outlier in the sample. It's still within 2\sigma, so nothing to worry about. But hopefully demonstrates that not everything can be fitted. And certainly not so if we had better data.
Hi Sabine,
DeleteI really appreciate your time on this, and obviously you covered it in all the detail believed necessary. However, I really think you are, as a physicist, highlighting the failing of experiment with Null results for SUSY correctly outing it as a dead end waste of time, while on the other hand promoting research into Dark Matter, when it appears obvious to me that null results are similarly looking at a dead end waste of time for the same reasons. The particles of SUSY are hiding in realms that we can argue are just outside our reach ad infinitum, and the same with Dark Matter - the discovery of what it is seems arguably just outside our observable reach and has been for many decades also.
regards
Marco Parigi
Marco,
DeleteI do not know what null results you are referring to. Experimental evidence for dark matter has only become stronger since it was first discovered. Those are not null-results.
In case you are referring to direct detection experiments that search for specific particles that supposedly make up dark matter, that's a different thing entirely. As I clearly said elsewhere, I am not advocating more of this because there is no particular reason to think those specific particles are the right ones, hence it's basically poking in the dark, hoping to get lucky - not a promising path to progress.
Instead, find a way (hopefully largely model-independent) to identify the best explanation for the existing data, or collect more data until you have a good idea what it's made of, if it's made of anything. It's really the theory-development that's behind here.
I was referring to Null results in terms of what dark matter is. The experimental evidence strengthening for Dark Matter would only be believable if there was a narrowing down of what DM could possibly be.
DeleteThe kind of narrowing down for another science, say continental drift is that as we get more and more evidence, we narrow down what is actually happening and where the continents are going and where they have come from.
This kind of narrowing down, even after many decades is just not happening with Dark Matter. There is a diverging range of Dark Matter varieties under investigation, and patterns associated with measurement eg. Galactic Rotation rules are also diverging the candidates. Thus we have no evidence to point us where Dark Matter is coming from nor where it is going. Dark Matter has never predicted a Galactic Rotation even once. Ignoring the DM and just plugging in the maths of MOND does far better. I would consider that to be an absolute null result for DM of any description. Ie. it is completely unnecessary to consider its existence to estimate how it will effect movements around a Galaxy.
If one could estimate the position of a rock formation based on counting the number of earthquakes that have been in the region, and a theory of tectonic plates told you nothing useful, why would we bother with continental drift?
regards Marco
Marco,
DeleteWhat you say is not correct. Every time you get better measurements you are ruling out options.
The reason that the narrowing down isn't efficient is that experiments are looking in the wrong places.
No, ignoring dark matter and "just plugging in the maths of MOND" does not do far better. It does better for a selected set of observations.
A lot of the measurements we make are useful regardless of what they say about Dark Matter. I do agree that better measurements rule out options, but they appear to rule out different random options of contradictory Dark Matter candidates.
DeleteI'm not sure which kind of Dark Matter we have ruled out. Particle Dark Matter, Cold Dark Matter (that could be the same thing), Lamda CDM, Self Interacting Dark Matter, WIMP's, MACHO's, fluid, superfluid.
The selected set of observations of velocities within Galaxies are at least tangible and relatable to our experience as a star within a galaxy (conversely, DM saves mass equivalence). Evidence based on early universe considerations are actually a more extreme variant of the same phenomenon of a saving hypothesis. Dark Matter is *needed* for the theories of the early universe to be true. That is not the same as being evidence for DM. It is the precise structure that Popper talked about when describing "saving hypotheses". A good amount of Dark Matter in the early universe saves the theory regarding the CMB's relationship to the start of the universe. I would go further and say that the same could be said of a God hypothesis for the early universe. Creation theory just doesn't work without God. Early universe cosmology just doesn't work without Dark Matter. Is that really evidence for Dark Matter? Is "needing" a God, evidence for Theism? I know I have picked on the kinds of evidence that can be interpreted this way, but is there evidence for Dark Matter that cannot be interpreted as "DM saves the day"?
It just really looks and feels like we are barking up the wrong tree in a very similar way to SUSY. Dark Matter is just slippery enough that it evades both proof and disproof. As far as Popper would put it - Dark Matter is definitely not a bold hypothesis.
regards Marco
Marco,
DeleteYou are confusing specific particle physics models for dark matter with the CDM hypothesis in astrophysics/cosmology.
These ae looked at differently for sure. Particle Dark Matter is subject to ever increasing power levels in particle colliders, which you criticise greatly, and with good reason.
DeleteMeanwhile the CDM hypothesis in astrophysics cosmology appears immune to experimental falsification in the way a hypothesised particle is. What I am comparing is not the need for it to be a particle, it is the need for DM in haloes for Galaxies and the need for DM in cosmological narratives in *PRECISELY* the amount required to suit the formulae and models. No amount of observations will nullify the arbitrariness of DM and CDM respectively to fit the models. Generalised non-particle models or unseen particle models are still invoked in some way as the object of discussion when talking about DM or CDM - It has to fit *A SPECIFIC MODEL* for Dark Matter - I still hear "can't rule out particle, fluid, self-interacting etc." with any or all of these observations. Measurements based on cosmological narratives are not converging on one, they are diverging amongst the possibilities of what DM actually is.
regards
Marco,
DeleteIt seems that you do not really understand the relevant astronomical observations.
Yes, for galaxies, there certainly is a degree of arbitrariness (“wiggle room”), but for the CMB, and galaxy clusters?
Re galaxies: what’s often not discussed, or even mentioned, is “selection bias”. Galaxies are very complex systems, arguably more so than large mammals, and many/most also have complex histories. Good examples of the complexity is “mergers” ... how many models of these systems have you come across? Models which test any kind of CDM/alternative gravity? Models which are a good match to detailed observations?
I believe that I do understand the relevant astronomical observations.
DeleteScience is strictly about the observable and repeatable. Models can certainly be observed and made to fit what we believe are the results of Galaxy Mergers. However, we do not observe galaxy mergers. We do observe star and gas velocities. We do not observe the Big Bang in progress. We do observe the CMB. We do not observe the long term evolution of Galaxy Clusters. We do observe different snap-shots of galaxy clusters at different particular point in the past.
Therefore there is just as much arbitrariness in statements about DM and the CMB and Galaxy clusters as there is in Galaxies.
With regards to selection bias - Again, we do not observe the complex history of any particular galaxy. We do observe snapshots at various points in the past of different galaxies. We should not pretend that we know enough to model them with any significant certainty.
Maro,
DeleteThanks for the clarification.
It seems that , by your criteria, most of astrophysics, and much of astronomy, cannot be science ... no black holes, no neutron stars, no white dwarfs, no surface of last scattering, and perhaps even no gravitational wave radiation.
Hi JeanTate,
DeleteThat is the jist of it yes. A lot of those things have presumed formation mechanisms which cannot be “observed” in the way science intends repeatable observations.
We certainly can observe snapshots of things we can surmise the properties of, like snapshots over decades of the Galactic centre, and say that those observations match what we believe a black hole should look like, but to then say a black hole “formed by gravitational collapse” or any formation mechanisms is not science in my view at all.
Ideas such as these about cosmology are not “self correcting” because it might take centuries, or even millions of years to get clear long-term observations that would nail actual formation mechanisms.
Regards
Marco
If you consider a system undergo deformation overtime, the developed stress deposits stress energy into particle production. It satisfy the requirement of multiple ground state simultaneously to exist and for symmetry breaking.it also satisfy the requirement for gravity, and it's relation to spacetime curvature.
ReplyDeleteThe inflationary phase of cosmology is scale invariant. The spectrum is not dependent on frequency or energy. This scale invariance ends or is broken with the reheating phase when the extreme energy density of the false vacuum collapsed to the much smaller energy density of the current physical vacuum. I worked the power spectrum of this breaking and did some calculations with the fluctuation-dissipation theorem. The spectrum of this breaking has a set of frequencies spaced apart in k space with approximate scale invariance in these voided regions. In the fluctuation-dissipation theorem it is remarkably similar to the CMB power spectrum. This would suggest for quantum fields in a phase transition that this breaking would occur for certain energy states and not for others.
ReplyDeleteThis might then be the basis for this mixed phase. What ever dark matter is it was generated by this reheating process. In there mass-energy was generated out of the vacuum, or curvature of spacetime, by breaking this scale invariance with reheating. However, maybe some particles are in a mixed phase where they retain some properties of spacetime. This might then be the superfluid. I would have to think a lot to see how this could recover something similar to Verlinde's idea or your covariant extension thereof.
Hello, in your slide there is a point "Reshift-Dependence of Radial Acceleration Relation". I would call it "Redshift-Dependence of Radial Acceleration Relation".
ReplyDeleteRegards
Charlyms
Thanks for spotting this!
DeleteHi Sabine,
ReplyDeleteThat 2015/16 paper by Khoury & Berezhiani is one of my favorites,
https://arxiv.org/abs/1507.01019
https://arxiv.org/abs/1711.05748
It got me wondering... with all that angular momentum in their superfluid, and the tendency of angular momentum in a superfluid to quantize, at what point would such quantization stop?
If such quantization went all the way (to the quantum of angular momentum, h), then their insight into this primitive transparent mass might give back more than they thought?
PS: loving your book, and thanks for all your work!
Nigel
nnunn,
DeleteYes, very good question. Unfortunately, I don't have an answer.
Sabine, Do you have a few recommended videos to watch explaining these derivations or modeling.
ReplyDeleteMichael,
DeleteNo, sorry. The best I have to offer is this rather basic introduction to the idea of modified gravity.
So, without Dark Matter...
ReplyDelete* Galaxies in clusters are too fast.
* Stars in galaxies are too fast.
* Galaxies and galaxy structures bend light too strongly.
* The CMB doesn't come out right.
* Structure formation starts too slowly.
And so there must be dark matter, or if there is no DM (as the empirical evidence strongly suggests), then at minimum, the standard model and its assumptions are called into question. This second possibility is not widely admitted. Negative empirical results simply carry no weight relative to theoretical proclivities.
This second possibility, while not necessarily correct, is quite logical and reasonable. However, it is effectively, given no consideration in modern theoretical cosmology. Consequently we have a circular argument:
The standard model is considered correct because it can be said to agree with observed phenomena. In order for the model to agree with observed phenomena, there must exist dark matter. Since the standard model is correct, then dark matter must exist despite strong evidence to the contrary. And merrily we roll along...
This is incorrect. The evidence strongly suggests that there is dark matter. I do not know what negative empirical results you are referring to. Dark matter is a model that astrophysicists use to make predictions and for the bullet points on your list that works just fine.
DeleteThe standard model, at least on scales we can test, works very well—it explains all our observations, except a handful which are most likely statistical fluctuations. If it's incorrect, it must be incorrect only at scales which we haven't tested.
DeleteIs the superfluid a quantum field theory? The particles in superfluid seem incompatible with graviton. The latter is massless and move at light speed. The former has mass of 1 eV/c^2 and move less than light speed (like neutrinos). How do you reconcile the particles at galactic scale become Bose-Einstein condensate while gravitons are continually absorbed and emitted by gravitating massive particles?
ReplyDeleteWhat does superfluid predict about gravity at nano scale like near black hole singularity? It is not quantum gravity if it cannot explain gravity at nano scale. It is modified general relativity or MOGR
The underlying theory is a quantum field theory, yes, but it has a classical limit (as usual). The particles in the superfluid are not massless, they have masses of about an eV. I don't know what you think the problem is with gravitons. This has nothing to do with gravitons.
DeleteThe superfluid is an approximation that works only on large scales, it breaks down on short distances.
I didn't say it's quantum gravity.
In slide 3, you mention that Dark Matter rarely interacts with itself. How do we know that? Do we assume that DM interacts only through the 4 known fundamental forces?
ReplyDeleteWe know that because otherwise structure formation doesn't work properly. It can have a self-interaction. There are models of self-interacting dark matter, but it can't be a strong self-interaction. It's generally a different force than the standard model forces.
DeleteSabine,
ReplyDeleteDark matter is a model that astrophysicists use to make predictions and for the bullet points on your list that works just fine.
Dark matter is a free parameter of the standard model of cosmology that was specifically added to address the failures of that model to replicate observations. Dark matter's failure to actually appear in physical reality (via empirical observations) is cloaked in the pretense that, because it can be deployed, in ad hoc fashion, anywhere that the SMC makes erroneous predictions, it must, therefore, exist.
As stated previously this is just a form of circular argumentation: any ad hoc addition (whether entity or event) to the SMC is acceptable as long as it successfully corrects the existing errors. This is held to be true, whether or not there is any empirical evidence for the existence of the posited entity or event.
All the evidence that can be cited for the existence of dark matter, such as the five bullet points, consists of model dependent inferences. Take away the 'special glasses' of the SMC and the evidence for dark matter disappears - it is not a part of observed reality, it is only part of the model that requires it.
Peter Shor's assertion, that the SMC 'works' (as in agrees with observations) and is therefore largely correct, is logically falsified by the historical example of Ptolemaic cosmology.
Like the SMC, Ptolemaic cosmology could be said to agree with observations. However, the Ptolemaic physical description of the solar system was completely erroneous. It is most likely that the SMC is Ptolemaic in the sense that it clearly does not describe the cosmos we actually observe. It is incongruent with empirical reality. There is no big bang, expanding spacetime, dark matter, or dark energy in observed reality. They all exist beyond the realm of science in the imaginary spaces of mathematics.
When I first learned science, it was definitionally presented as the study of those things that can be observed and measured. This provided a sound distinction from, and pathway out of, the fuzzy supernaturalism that pervaded the worldview otherwise presented to me in my youth.
The modern 'scientific academy' has reconstituted supernaturalism under the guise of mathematics, and presents to us a picture of the cosmos that is immune to verification by observation and measurement. We are required to take the existence of things that cannot be otherwise observed on the word of mathematicists who insist, that the fact that their models can be made to agree with observations, constitutes sufficient evidence for the supernatural (unobservable) elements of their model.
In this mathematicists have become like the high priests of old. They hold forth unassailable on the existence of invisible things, wrapped in a cocoon of self-generated authority. That formulation and disposition is antithetical to science as I learned it. And so, I object.
bud rap,
DeleteDark Matter is a model. It fits the data, it makes predictions. That's how science works.
"Take away the 'special glasses' of the SMC and the evidence for dark matter disappears..."
Assuming that SMC means LCDM, if you take it away, you have no theory and cannot explain anything, so what's your point?
Assuming that SMC means LCDM, if you take it away, you have no theory and cannot explain anything, so what's your point?
DeleteThe point is that physical reality should carry more weight than a model and when your model diverges significantly from reality (cf. 95% of the model's 'universe' is invisible in physical reality), it is highly likely that it is the model that is at fault and not physical reality.
Such a conception is not even admissible in modern theoretical physics. There, the model is always right because it can always be mathematically massaged into agreement with actual observations via the liberal use of free parameters. The fact that the free parameters do not correspond to actual observables and therefore exist only in the model is simply treated as an irrelevancy.
When you say that's how science works, I have to agree with you. At least, that's how it appears to work in the realm of theoretical physics. But that is wrong. That is not the way science is supposed to work.
Science is supposed to be the study of physical reality. Theoretical cosmology has become, instead, an exercise in the care and feeding of a preferred mathematical model. The incongruence of that model with reality isn't even an afterthought to modern theorists. The fundamental assumptions of the model have become internalized, atavistic paradigms that are not subject to reevaluation.
Consequently, the standard model of cosmology, currently referred to as Lambda-CDM, is physically, both incoherent (the inexplicable original condition) and absurd (95% of the cosmos is composed of invisible matter and energy the only salient characteristic of both being that they reconcile the model with observations).
The current situation is exactly as it was at the end of the Ptolemaic era. You can't fix the scientific mess that is modern cosmology by tweaking its failed model. The model has to be discarded and its fundamental premises (now approaching 100 years old) need to be reconsidered in light of modern, observationally-based knowledge.
Do I see that happening anytime soon? Hardly. But that, too, is my point. Modern theoretical physics is sclerotic with mathematicism. Until math returns to being a useful and essential tool of science instead of the oracular "language of science" it is now widely touted to be, theoretical physics will remain at a dead end.
bud rap,
Delete"The point is that physical reality should carry more weight than a model and when your model diverges significantly from reality (cf. 95% of the model's 'universe' is invisible in physical reality), it is highly likely that it is the model that is at fault and not physical reality."
Whether it's "visible" or not is entirely irrelevant.
Sabine,
DeleteI understand full well that you and those who share your mathematical point of view consider the failure, of a model's free parameters to actually appear in physical reality, an irrelevancy. That is the prevailing paradigm throughout the theoretical sciences of the academy. My argument is: that approach has crippled theoretical physics and delivered it to a dead end.
The upshot of this capture of theoretical science, by a mathematical paradigm, is that the primary object of study for scientists is no longer physical reality - theorists now mainly study mathematical models. And the upshot of that is clearly seen here:
https://www.newyorker.com/science/elements/a-different-kind-of-theory-of-everything?
All the concluding metaphysical dross aside, the takeaway from this article is that mathematical models that "work" cannot be relied on as accurate descriptions of the underlying physics. In fact, most models seem to misrepresent the underlying physical system, leaving us with no understanding of, for instance, the causal chain of the gravitational effect.
You may consider the absence of empirical evidence for the existence of dark matter irrelevant but I consider that an unscientific point of view. Without empiricism, science founders on the shoals of supernaturalism. Without empiricism you can do perfect math but you will do no science. Without empiricism, science does not exist.
bud rap,
DeleteFirst, please stop blaming me for some article about Nima Arkani-Hamed.
Second, your statement
"You may consider the absence of empirical evidence for the existence of dark matter..."
Is wrong. I have listed empirical evidence for the existence of dark matter in my slides.
Sabine,
DeleteFirst, please stop blaming me for some article about Nima Arkani-Hamed.
I honestly don't know what you mean by that. I cited the article in support of my argument. Why would you think I am blaming you for an article somebody else wrote? Could the German word for blame have some connotation that does not translate in this context?
I have listed empirical evidence for the existence of dark matter in my slides.
What you list are empirical observations from which you infer via the standard model the existence of dark matter. Those empirical observations, since they do not entail the direct detection of dark matter, do not constitute empirical evidence for the dark matter claim. They provide model-based, inferential evidence only.
bud rap,
DeleteYou claim "theorists now mainly study mathematical models" quoting an article about Nima Arkani-Hamed. I am pointing out that you are blaming me for what someone else has said in an interview.
"What you list are empirical observations from which you infer via the standard model the existence of dark matter."
The observations document that there is something we cannot explain with the current theories. Dark matter is a parametrically explanation. That is standard procedure in science and it is good scientific methodology.
bud rap,
DeleteI have refrained from commenting on your comments, so far, but your last contains one misunderstanding/misrepresentation too far: “They provide model-based, inferential evidence only.”
What, may I be so bold as to ask, do you think astronomy is, other than exactly what you state?
This comment has been removed by the author.
DeleteSabine,
DeleteThe observations document that there is something we cannot explain with the current theories. Dark matter is a parametrically explanation. That is standard procedure in science and it is good scientific methodology.
I agree with everything you say there - except the last clause. The ability to freely parameterize a model essentially renders it unfalsifiable. That in turn renders the model impervious to reconsideration, and particularly to reconsideration of the model's axiomatic structure. That is a profoundly unscientific state of affairs.
The free parameterization of a model is not good scientific methodology, despite the fact that it is now common practice in the scientific establishment. While such an approach is perfectly acceptable in the mathematical realm (because axioms there can be true by definition), the consequences are disastrous for science.
When the axiomatic structure of a scientific model is, by consensus, not subject to reconsideration, the model has devolved into a belief system. LCDM is just such a belief system. If the incoherent and absurd claims of LCDM, about the nature of physical reality, were cloaked in theological rather than mathematical terms, its current proponents would sneer condescendingly at its crudity.
On the other matter, I remain mystified as to what it is you think I was blaming you for. Nonetheless, it is quite apparent that I offended you. That was not my intent, I sincerely apologize.
bud rap,
Delete"The free parameterization of a model is not good scientific methodology..."
That's just wrong. If you have a model with one free parameter, you measure that parameter and that makes the model predictive. That's standard scientific procedure.
"despite the fact that it is now common practice in the scientific establishment."
There's no such thing as "the scientific establishment" and clearly you have very little idea about what is and isn't common practice in science.
Look, I am not a fan of LCDM either, but your statements are are incoherent. LCDM can be criticised but not for the reasons you bring up.
You did not offend me, I pointed out that you make an inappropriate inference from a sample of one. Arkani-Hamed is not "theorists." He is one theorist.
Sabine,
DeleteIf you have a model with one free parameter, you measure that parameter and that makes the model predictive.
LCDM has six free and six fixed parameters. The fact that the model can be tweaked to agree with observations on the basis of those parameters is irrelevant to the question of whether those model specific parameters correspond to anything in physical (observable) reality. You appear to think that the lack of correspondence between such parameters and observed physical reality is irrelevant, I do not.
I also consider the free addition of both free and fixed parameters to a model to be inherently problematic, since it precludes reconsideration of the model when discrepancies arise. If empirically baseless parameters can be freely added whenever a model is discordant with observations, that model is essentially unfalsifiable. You seem to disagree, and there we appear to be at an impasse.
...your statements are are incoherent. LCDM can be criticised but not for the reasons you bring up.
Since you don't specify which statements you find incoherent, let me summarize my view of LCDM for you:
1. The original condition of the model at T=0, is logically, mathematically and physically incoherent.
2. The statement that the "universe" is 13.8 billion years old is a statement of universal simultaneity, which is a direct contradiction of relativity theory. (I am unimpressed with the invocation of co-moving reference frames as a mitigating argument.) If it is not a statement of universal simultaneity then it is a meaningless statement.
3. There is no direct empirical evidence for the existence of the dark matter and dark energy that are said to comprise 95% of the matter-energy content of the "universe".
4. It is a foundational assumption (early 1920s) of LCDM that the cosmos is a "universe" that can only be engaged in either a unified expansion or contraction. There has been no significant reconsideration of this assumption since its inception.
5. In light of the "universe" assumption, the redshift-distance relation discovered by Hubble (late 1920s) was causally attributed to a universal expansion. This assumption has also never been seriously reconsidered since its inception.
You may disagree with these points and no doubt you do. They are, however, not incoherent in the sense that all accounts of T=0 are.
Lastly, I did not quote Arkani-Hamed or the article. I dismissed Arkani-Hamed's comments as metaphysical dross and only cited the article as illustrative of a problem with mathematical models.
The statement "theorists now mainly study mathematical models" is simply my opinion. The qualifier "mainly" means that it is not a categorical statement - it does not apply to all theorists. The implication is, at worst, most theoretical researchers. I have always found your own willingness to step outside that framework refreshing.
bud rap,
Delete1) I don't know what 'T' is.
2) It's an approximation and everyone knows that. (Except you, possibly.)
3) So what?
4) I have no idea what a "unified expansion" is supposed to be.
5) Just wrong. There have been various attempts to explain the observations by other means, they just work very badly.
And in any case, as I said, these are not points of criticism that any scientist can take seriously.
Sabine,
Delete1) T=time, as always when referring to the inexplicable original condition of LCDM. Surely, you know that. I have to assume you are trying to deflect the point about the incoherence of the standard model's original condition as being moot, since the official position now seems to be that the model provides no account for its origin. If that is what you're getting at, it's a very weak argument.
2) An approximation of what exactly, the simultaneous age of the universe? And that approximation is approximately and simultaneously the same for all observers throughout the universe? Either the "universe" is, or is approximately, 13.8 billion years old or it is not.
If it is, then you are discussing the approximate timing of a universal simultaneity, otherwise (if it doesn't apply to the entire "universe") then it is an approximately meaningless statement.
3) "So what?", is the crux of the matter. That remark simply dismisses significant empirical evidence against the existence of dark matter as inconsequential, a mere triviality. To me, that represents an abandonment of fundamental scientific principles. Essentially the dark matter model has been declared correct by most modern scientists, despite its glaring inconsistency with observed reality.
It is my considered opinion, that science cannot proceed, and will not succeed, on such a trajectory. It could be argued that since the elevation of QCD in the late '70s to early '80s, theoretical physics has ceased to advance, entering a cul de sac of mathematical speculations.
At the time, QCD represented the formal acceptance of theory over observation since it posited the existence sub-nucleon particles (quarks) that were held to be undetectable as individual particles, due to a theoretical charge characteristic, asymptotic freedom, that resulted in "color confinement" (the quarks could only exist as multiples in the guise of ordinary observable particles).
This was essentially an inflection point in the history of science, and especially, in the nature of science education. Everyone receiving a science education subsequently, has been taught to accept this inversion of theory over evidence as standard scientific procedure.
And there we stand, on two sides of that great chasm. You consider negative empirical evidence to be irrelevant to the validity of a model. I consider it a falsification of the model.
You are correct to state that my position is not widely accepted among academic scientists. But, as is sometimes noted, science is not a popularity contest. Time will tell which view is scientifically justifiable.
4) Unified expansion, as in universal expansion. So why didn't I say universal? Because I wanted to allude to the simultaneous nature of the expansion that modern scientists prefer to elide.
The overall point, of course, was that a "universe" is a baked in assumption of the model. That assumption has never been seriously reconsidered - never.>. I could also make the point that applying GR to the universal FLRW metric inherently generates a self-contradictory model.
...
...(continued)
Delete5. Here you are correct, I overstated the case. Zwicky made a good faith effort to exam alternative explanations and a few other attempts have been made over the years. They have all failed.
However, what has been overlooked is the straightforward GR consideration of gravitational redshifting applied to the expanding spherical wave fronts of light emitted by galaxies.
Even a crude calculation, using a reasonable mass density assumption to estimate the mass at each iteration, yields a redshift-distance relationship for a wavefront without the usual tired light drawbacks. Amuse yourself, a crude calculation can be done on a spreadsheet.
_______________________________
Outside of your point #5, you made no serious attempt to address the issues raised but rather deflected the criticisms with feigned incomprehension of standard terminology and perfunctory dismissals, all covered with a transparently weak argument from authority. You're better than that Sabine. You know how to think - maybe you hadn't had coffee yet. I'm rather grumpy, myself, pre-coffee.
bud rap,
DeleteCapital T commonly stands for temperature, not time. Believe that or not.
No, I do not "feign" incomprehension, you indeed do not make sense.
I have neither the time nor the patience to teach you cosmology. If you cannot comprehend that no model based on a differential equation ever explains its initial condition, I cannot help you.
bud rap,
DeleteI do not know why you continue to write lengthy comments which do little more than display an appalling ignorance of not only cosmology, but astrophysics and even astronomy (I’m not as blunt as Sabine).
“significant empirical evidence against the existence of dark matter” Huh? Perhaps I missed it, but you have not even hinted at any, have you?
And I for one have no idea what your last point is supposed to mean (“Even a crude calculation ... ”).
Hi Sabine,
DeleteI find Bud Rap's comments to be coherent and important. I won't say that I completely agree with them, but theoretical physics has really veered away from repeatable observable and predicting new facts...
regards
Marco
Sabine,
DeleteIf you cannot comprehend that no model based on a differential equation ever explains its initial condition, I cannot help you.
You seem unable to grasp that it is just that inadequacy of the standard model, that I am criticizing. Explaining to me that the model doesn't explain its original condition is like explaining to the Pope that he's a Catholic.
You obviously don't want to have a conversation about the shortcomings of the orthodox model as a description of physical reality. This is your blog and I respect your choice. I will not trouble you with my opinions on the matter any further. Thanks for your efforts on behalf of science.
Regardless, I really like the equations. I think they are a real hint to what is actually happening.
ReplyDeleteYour two-phase theory is intriguing, especially in how it captures some of the strangely predictive aspects of MOND without resorting to point-blank modification of GR.
ReplyDeleteIf space and gravity are emergent from and thus dependent on matter, it would seem plausible that space might also subtlety share certain aspects of some of the particles of matter, e.g. of spin-1 vector bosons.
So my question is this: Is it possible, without violating known observational data, that the two-phase fluid you are proposing to be in some sense "part" of space itself, that is, a fluid without which space itself cannot emerge?
Terry,
DeleteIn a sense I believe that's what Verlinde is trying. Alas, I don't see how this would work, at least not at present.
Let me also say that as a pragmatist, it doesn't seem to me a useful idea to pursue. The reason is that the type of theory we can test with observations would be an effective theory regardless and you wouldn't be able to draw the distinction.
Concretely, look at the Lagrangian in my model. It's based on Verlinde's model. But that Lagrangian in an by itself looks like any other modified gravity Lagrangian, emergent or not. So really dark matter/dark energy I think will not help you tell what's going on fundamentally.
Sabine,
DeleteWow, this is a lot messier than I realized from looking only at your slides. While my poor quick reading of those slides left me with the impression that you advocate the straightforward 2016 Berezhiani and Khoury particle-based boson condensate interpretation of DM, I'm assuming from your comments above that you really still prefer your own 2017 covariant modification of Verlinde's emergent gravity using a baryon-coupling vector field.
I'm not at all wedded to the idea of "space as condensate" for explaining DM. I just don't mind brainstorming a bit, as long as everyone looking at some wild-hare idea keeps the presence of mind to recall that no idea works if it leads to contradictions with experimental reality. (On the other hand: Dirac's sea of electrons idea was about as nutty and patently unreal of a physics brainstorming idea as one can imagine... yet somehow it ended up predicting positrons! His sea idea also had amazingly accurate resonance with finite-sea solid state electron-hole theory, even though that did not seem to be Dirac's intent. Dirac was a person of deeply astonishing intuition, able to navigate physics in ways that remain baffling. What a truly amazing mind!)
To be honest, I like the "space as condensate" idea not so much for dark matter, but as a nice way to capture my own just-for-fun version of emergent space as deriving from the need for large ensembles of particles to share a common 3D orientation framework that guarantees absolute conservation of angular momentum. If you assume that Planck's constant is best represented by one unit of angular momentum, then you cannot avoid the implication that Planck's unit as usually stated is incomplete (!). Even though angular momentum is in Planck units, accurate mapping of its properties relative to other units requires adding a 3-vector spin axis. Planck-as-spin-1 thus unavoidably leads to the odd idea that "orientation" is a conjugate of the common Planck unit, and that such "orientations" might collectively behave like exceptionally ephemeral bosons. Emergent space in that case would be the literal Bose condensation of these ephemeral "orientation bosons" that are required by a spin-1 definition of the Planck unit.
All of which I mention here only because I can't help wondering if such ideas might link into your baryon-coupling vector field idea.
If one unit of angular momentum truly defines the full Planck's constant, then the usual Planck's constant is an incomplete approximation because it lacks a spin orientation 3-vector. Adding that orientation to the classical definition of Planck might create some interesting and nicely direct relations between small-scale quantum and emergent large-scale flat space, without any direct need for holography (at least at the first order of analysis).
Cheers,
Terry
Fascinating, thanks! I will read more deeply into this.
ReplyDelete