Hacker News new | past | comments | ask | show | jobs | submit login
Cosmologist claims Universe may not be expanding (2013) (nature.com)
192 points by hairytrog on Apr 6, 2019 | hide | past | web | favorite | 94 comments

I know Christoff Wetterich! From the time of my Ph.D. when we were working the Renormalization Group. I met him a few times, we were working on similar things but in different groups, we were "competitors" if you wish. Funny to see him cited in HackerNews. He is certainly a respectable physicists, but that can be said for many others. I have not read the cited article, it is in my field, so I could understand it if my memory worked well, but after 20 years from my Ph.D. I am too rusted now.

The central concept behind Christof Wetterich's proposal is scale invariance: Imagine you try to measure the size of the universe with a yard stick. If you measure that the universe is expanding, you can't be sure that it really is the universe expanding, or maybe just your yard stick shrinking.

This is true in so-called scale invariant theories. These theories do not contain any intrinsic length scale. The moment your theory contains a fundamental length l you break scale invariance. In this case you are able to compare your yard stick to the length l to find out if your yard stick was shrinking or the universe really is expanding.

Our present model of the world still contains a few fundamental length scales (such as the Planck length, or a length scale associated to the nuclear force), but people such as Christof Wetterich are building models that try to get rid of (some of) these fundamental length scales. A major approach in this direction is asymptotically safe quantum gravity, that also led Christof Wetterich and Mikhail Shaposhnikov to the most precise prediction of the Higgs mass prior to the Higgs' discovery in 2009, see https://arxiv.org/abs/0912.0208.

But the yardstick argument assumes uniform expansion. Small scale structure like our yardstick is held together by electromagnetic forces which counteracts the expansive force. That's my understanding anyway.

If we create a light year long line of yardsticks then we would expect the line to expand faster than the yardsticks that make it up. Locally this would look like growing gaps between the yardsticks. we never used the fact that the line was a light year long. So let's just put two yardsticks next to each other in a vacuum. They drift apart and eventually so far they are not part of each others' observable universes.

That’s accurate. Metric expansion only dominates loosely bound structures on a scale beyond individual galalxies. The yardstick, solar system, and Milky Way are all on scales too small, and are too strongly bound, for metric expansion to have an impact. Galaxy clusters and the space between galaxies however, are impacted.

This is (2013). Christof Wetterich continued research, his current theory is called emergent scale symmetry: https://arxiv.org/abs/1705.00552. It got peer-reviewed and published at Physical Reviews D, the leading journal on cosmology.

Interesting take on things. There's value in thinking about equivalent models that allow potentially entrenched assumptions to be set aside. Worth mentioning that the kilogram has since been defined more robustly than the way mentioned in the article. https://www.sciencealert.com/it-s-official-the-definition-of...

> The idea may be plausible, but it comes with a big problem: it can't be tested.

Doesn't that imply that the accepted model can't be tested? I mean, in the sense of distinguishing it from this one.

Yes, depending on your favorite definition of "different model" there are models that can not be distinguished even in principle. Most famous example is the different interpretations of quantum mechanics.

In my opinion the reasonable way to deal with that, is to assume that they are the same model since they describe the same physics and leave the semantics to the philosophers.

Yeah, I was thinking along similar lines, but it seemed too outrageous. The various interpretations of QM are so-far indistinguishable because they're based on the same (or equivalent) mathematics, and so make the same predictions.

This one, though. I mean, the fundamental observation is that red shift increases with distance. The accepted model has red shift caused by motion away from us, and so we conclude that recessional velocity increases with distance. That is, the Universe is expanding.

This new model has blue shift caused by increasing mass. And so we are blue-shifted relative to distant stuff, which we see only in the past. So it looks red-shifted to us. And distance is proportional to time delay, so stuff that's farther away is more red-shifted.

Bu I don't see how the temporal rate of mass increase could be related to -- or rather, another reflection of -- the expansion of the Universe. But then, IANAP, so hey.

Not a physicist, but I recall reading that there is enough empirical evidence to support the current model of expanding universe.

OK. But if that evidence were adequate to exclude this new hypothesis, then one could reject it. And TFA says that one can't. That implies that existing empirical evidence is just as consistent with the new hypothesis as it is with the accepted one.

I always wonder about that. I remember very specifically learning 40(?) years ago that the expansion was slowing, supposedly proven by some sort of red shift in light. That is now out of favor, but a few more new facts just might swing things the other way again.

> That is now out of favor

It's not "out of favor" just as a matter of fashion: a lot more data was collected, extending much further away and therefore much further into the past.

> a few more new facts just might swing things the other way again.

No, it would take a lot more than that, because our current model of accelerating expansion for the past few billion years (note that before that, the expansion was decelerating) is based on a lot more data than we had 40 or so years ago.

Accelerating universe is supported by supernova data. We have incomparably more supernova data compared to the past. Discovery of accelerating universe was mainly due to radically scaling up supernova discovery process.

Yes, but that’s still dependent on a differential velocity model for the origin of redshift. This paper changes that core assumption, meaning the same data can be interpreted to have a radically different meaning.

Afaik, we have no direct observations of galaxies moving away from us - over such enormous distances we have not had enough time since the start of observations to actually see a galaxy recede.

No, the origin of the cosmological redshift is not proposed to be due to differences in velocity, but metric expansion.

... which causes a change in relative velocities

That makes it sound like cosmological redshift is a form of Doppler shift. In reality, the target object emits light towards us with little Doppler shift as it will have experienced little acceleration in its reference frame. The light becomes redshifted as it moves through the universe while the universe is expanding, effectively stretching out the wavelength of the light.

The change in wavelength is exactly proportional to the increase in size of the universe during the time of flight, and it is not directly related to speed of recession except in common and inaccurate discussion. The link to relative velocity (if that concept makes sense for objects in widely separated and unrelated reference frames) is model dependent, while the metric expansion measurement is directly measured.

Ok. I was under the impression that expansion had no effect on local phenomena because local phenomena always counteract expansion. But if you view light more as a wave instead of a particle then could see the argument that the light is locally stretched if there is no force keeping it from being stretched e.g. gravity or energy.

Within galaxies, however, gravity dominates dark energy expansion, right? If so, with enough measurement accuracy shouldn't we be able to look for standard candles within our galaxy that either show or don't show said mass differences while compensating for any shift due to relative motion? I'd think that given enough accuracy we'd have something that's testable eventually.

Reminds me of my funny notes in a childhood diary. I had a personal pet theory that Dinosaurs were so huge because they were not as heavy as we assumed them to be. Imagining what it felt like to be a dino was a steady staple for boring afternoons.

There is a fringe hypothesis around called the "expanding Earth theory." The idea is that the Earth and presumably other planets gain mass over time by some unknown mechanism like particle capture in their core. Dinosaurs didn't weigh as much because the Earth had less mass.

It actually proposes a plate tectonics model where the continents all fit back together as you shrink the Earth and it kind of works.

AFIAK the hypothesis contradicts too much other data and is likely wrong, but I count it as an example of a good creative out-there hypothesis. This kind of thinking is good in general.

Maybe they're big because they're a lot closer than we realize.

Wouldn't the gravitational pull increase? For example you would see less gravitational lensing in older galaxy clusters.

I think it's worse than that: it looks to me like this model will either not predict what it's claimed to predict, or will violate local stress-energy conservation.

Basically, if stress-energy is locally conserved, particle masses can't just increase out of nowhere: the energy that goes into the increasing masses has to come from somewhere, and so the overall energy, which is what drives the dynamics of the universe, doesn't change, so the model can't predict anything different from standard cosmology.

OTOH, if stress-energy is not locally conserved, then we should be able to observe such violations of conservation. But no such observation has ever been made: we have extremely strong evidence of local stress-energy conservation.

>we have extremely strong evidence of local stress-energy conservation.

Can you observe tiny scale invariant change in just few hundred years?

> model can't predict anything different from standard cosmology.

I think that's the goal of emergent scale symmetry at this point. Alternative to standard cosmology that explains the same things. To differentiate between the two you need to figure out what is different.

I can see your point, but is the introduction of dark energy in the current favored model better? Instead of having an energy coming from nowhere that increases particle mass, we have something coming from nowhere that expands space.

> Instead of having an energy coming from nowhere that increases particle mass, we have something coming from nowhere that expands space.

Dark energy does not "come from nowhere". It's locally conserved everywhere. It has to be: the covariant derivative of the metric is identically zero, and the "stress-energy tensor" of dark energy is just a constant times the metric.

Yes, but it wouldn’t be locally measurable, as as the article notes mass is dimensional, meaning from your reference frame you’d notice no change.

As to looking at older galaxies - maybe - but if the relativistic mass of photons is also changing due to the same phenomenon, there would be no difference in observations.

The universe isn't expanding. It's everything in it that's shrinking!

Strange that your comment is being downvoted, because it's a good summary of the article!

For instance size of hydrogen atom is inversely proportional to the mass of electron [1], so with heavier electrons, sizes of atoms shrink.

And in general in units where planks constant and speed of light are equal to one, mass is measured in 1/length.

[1] https://en.m.wikipedia.org/wiki/Bohr_radius

This seems testable enough... The speed of light isn't in question here, so the frequency is the yardstick. Can't we just record the frequencies given off by the elements every hundred years or so and see that they are the same?

That’s true, as with this cosmology light evidently isn’t affected by the increase in rest mass, as photons are massless. If this cosmology affected relativistic mass, then light would also be affected and would not be seemingly redshifted by the mass gain mechanism.

So, yeah, we should be able to measure a gradual emission spectrum shift here on Earth with current methods - we can measure energies of photons to very high precision, and you’d likely only need a few years of data to prove or disprove the hypothetical cosmology.

We should do it then. Maybe nobody ever looked because we assumed these things to be invariant.

Not necessarily, because the currently accepted theory is that the universe is expanding at an accelerating rate. The frequencies should change in both models.

What are the theories on what we think the edge of the universe looks like? What does it even mean? Will we just hit an invisible wall like in a video game? Or do we think it’s curved around the edge? Or what?

If it’s expanding, what is it expanding into?

> edge of the universe looks like? What does it even mean?

I think you should start with:


And only then with:


In short, what can ever influence "our" part of the universe is limited by the speed of light being constant and being the physical "speed limit", and by the fact that the age of the universe is less than 14 billion years. (Earth alone is 4.5 billion years old).

Fascinatingly enough, we can "see" in each direction we look at up to the distance of around 46 billion light years, even if the age of the universe is just 13.8 billion years (that 46bly in one direction is the result of the accumulated expansion of the universe since the beginning). And the farthest things we see are also the oldest, as we effectively "look back in the past" when we look at the distance (due to the speed of light being constant). So when we look far enough we see almost up to the beginning of the universe (up to the point when the universe was "clear enough" for us to see after the beginning). The CMB (Cosmic Microwave Background) is the "leftover glow" of the big bang that reaches us now, from that earliest past:


Then one can try to guess what's there behind the physical limits that are fixed for us, but we will never have the direct data about that:


I believe that’s what this Wikipedia article discusses:


This is something that has always intrigued me. What could possibly contain the universe? Another universe perhaps?

I think such a thing is difficult to comprehend except maybe mathematically. We are our universe and our universe is us.

I think when you are trying to envision the universe being contained in anything, you are already on the wrong path.

Maybe akin so asking: "What came before time?". The question itself is a contradiction.

Why is that the wrong path? I think these are valid questions. Please, I’m curious to hear your thoughts.

It’s a meaningless question to ask because the universe (according to the typical definition) encompasses all of space and time. It does not make sense to ask what came before/after it, and it does not make sense to ask what is outside the universe.

Your question seems to assume there’s an edge but there doesn’t have to be. Space could be curving in on itself (if you had FTL travel maybe you could make a circuit) or it could be infinite. What I’m not sure about is the distribution of matter in an infinite universe. Is there an edge to the matter or is it infinitely distributed with the universe itself.

We observe from a far distant star the red shifted spectrum. This is explained by the expansion of the galaxy and the expanding room between us and this star. So the photons coming from there have a red-shifted and hence lesser engery. What has happend to the energy of the photon? Where's it gone to?

I think the theory is suggesting that the photon was emitted with less energy.

I think the question is that if the current accepted theory for the redshift is correct, then where has the energy gone?

Would the star then appear to be a different color to "someone" its own solar system than it does us? Does OUR star thus appear to be moving TOWARD them since it's "our" color vs what "they" would consider to be the correct color for a star our size?

Doesn’t this require the photon to be emitted differently depending on where it ends up, as it would experience different amounts of redshift depending on the length of flight?

That’s some serious retro-causality if a photon from 13B ly hits us instead of Andromeda.

Forget about the expansion of the universe; you could ask the exact same question for an observer moving away from a source and detecting a different frequency (hence energy) of the light emitted due to "normal" Doppler effect in a non-expanding universe. Energy (on its own) is not an invariant in (special/general) relativity. You have the same situation for measurements of lengths (i.e. Lorentz contraction). Energy is one coordinate of a four-vector whose "length" is the true invariant.

This is similar to a situation where you draw a set of perpendicular axis on a plane and use them to give coordinates (x, y) to a point. You then rotate the set of axis, keeping the point fix. Your question about the energy amounts to asking "what happens to the x-coordinate of that point?".

The energy of a photon doesn't go anywhere, it just seems to us to be lower. The observed energy of a photon depends of the relative motion. If you move at the same velocity as the observed object, the redshift disappears and the energy of photon is there.


That’s a nonsense reply.

Those calories are known to dissipate via motion, light, and heat.

By contrast, the energy lost to expansion is simply lost, not emitted by the traveling photon through some mechanism. There is no conservation of energy in such a system, so it’s completely unlike the case of burning calories.

Should I add a /s?

It doesn't seem to make any sense as sarcasm either, so while you're free to do that you'd be better off explaining.

Why does it have to make any more sense than the original question? I didn't mean anything by it, it was flippant. And all the same, I have a hard time understanding what could otherwise be meant by the GP question. I heard that question in high school physics. It doesn't have an answer today (nor did it in 1998 when I took high school physics) that I'm aware of (which admittedly doesn't mean much). But, it especially doesn't have an answer today (or, in 2013, since that's when the OP article was written) if our notion of what happens to red-shifted photons and their energy depends on the expansion of the universe.

Maybe you'd like to postulate a theory to the original question with your own energy expenditure instead of deriding me for mine?

If you think there's something wrong with his question, this is the kind of place where you formulate that in a clear way, instead of dismissing it flippantly.

The question is why the red shift happens, and the proposed answer is that the photons were actually emitted with less energy.

It's not a nonsense question. You shouldn't treat it like one. And your weird sarcastic question is just off-topic, not nonsense, so it's very confusing as a counter to begin with.

Well, in my defense, I never intended it as a counter. Like I explained, I was being flippant. I responded with nonsense to something that I originally perceived as being nonsense. I can accept that I was wrong. And, absolutely, in retrospect, at the very least, it was all in bad taste.

> The question is why the red shift happens, and the proposed answer is that the photons were actually emitted with less energy.

I didn't get that from TFA or the GP question at all. What are you basing that on? What am I missing? A more careful reading?

isn't /s a reddit thing, and a really annoying one at that?

Sure, yes, this was all in bad taste. I accept that. My apologies.

I can't see why the article would say the effect is not measurable. We are trying to contrast two explanations: In one, galaxies are receding and the expansion of space in the intervening distance red shifts the spectra. In the other masses are changing (globally?) as some rate, changing the spectra.

The best modern atomic clocks depend on atomic spectra, a shift in which would be evidence for or against this model.

Given the precision of modern atomic clocks (<10^-17)[1] and the fact that the galactic spectral shifts are significant at large distances this seems plausible. I don't have precise #'s to do the calculations at hand so maybe I've made an error.

[1] https://en.wikipedia.org/wiki/Quantum_clock

But if our atomic clocks' frequency is changing, how would you tell? You can't use a more precise clock; we don't have one. You'd have to do something like measure how far light travels in the same time. But if our length measurements are changing too... do they cancel? Or could we still tell?

What you would need to determine is whether a dimensionless quantity is changing.


If the fine structure constant α is found to change, then would only be able to say that the relationship among the speed of light c, the elementary charge e, and the Planck constant ħ are changing, but IMO if we discovered that (say) α changed from .007300… to its present value of .007297…, we would not be able to then say "the speed of light changed (and the elementary charge and Planck constant remained the same)"

…but as a matter of practicality, scientists could feel free to use a different value for c in the past (and the present value for the elementary charge and Planck constant), if that is the most convenient way to accommodate a change in α. It wouldn't surprise me in the least if holding c and ħ constant is better for astronomers and holding e and ħ constant are better for particle physicists.

(Saying that ħ changes would be similar to saying that mass changes, or at least, the units of ħ include mass, and the Planck constant "is the basis for the definition of the kilogram." in the SI system [well, technically, this is not the case until next month. Right now the kilogram is still defined by the physical "international prototype kilogram" object, but the above quote is from wikipedia])

However, while the best experiments are consistent with changing α, the change in α is constrained to be WAY less than it would need to be to explain cosmological red shift observations.

There is also the opposite scenario, the Big Rip: https://www.youtube.com/watch?v=gEyXTQ9do-c

Kind of related to something that I've been chewing on:

I've read things (outside of the OP) that say the universe is expanding and also speeding up it's expansion, as neighbors exert less of a pull on each other the further spread out things get. However, that gravitational pull is still there -- will it at any point slow down the expansion and retract back inwards?

> that gravitational pull is still there -- will it at any point slow down the expansion and retract back inwards?

No. If we are talking about the current best model we have of the universe, the dark energy density (which is what drives the acceleration of the expansion) is constant in time, while the density of matter (which is what causes the expansion to decelerate) decreases with time. Up until a few billion years ago, the expansion was decelerating, because the density of matter was high enough to dominate. But since a few billion years ago, the density of matter has been small enough that the dark energy dominates, and that is never expected to change.

Though to be fair, our expectations on what dark energy will do aren’t very well grounded.

We already have inflation and dark energy but no good reason to expect either of them to be there at all. Like, the book on dark energy I’ve seen go: ”here are a bunch of pretty good guesses”, thats not a field full of confidence in its expectations (though admittedly the book is 9 years old by now).

> We already have inflation and dark energy but no good reason to expect either of them to be there at all.

I have seen plenty of heuristic arguments for why we should expect inflation to be there, but I'm not really qualified to judge. None of them seem to have convinced enough physicists to be considered standard.

For the cosmological constant, I think it depends on what you consider the "natural" condition to be. The simplest way to derive the Einstein Field Equation is to start from the Einstein-Hilbert Lagrangian and vary it with respect to the metric. The justification for using the Einstein-Hilbert Lagrangian is that it contains the only scalar which contains up to second derivatives of the metric and is quadratic in the metric and its derivative, namely, the Ricci scalar. But that statement is not actually true: there's another obvious scalar that meets this requirement, namely a constant. So the most "natural" version of the Einstein Field Equation contains the cosmological constant: it should be there on theoretical grounds, and that means "dark energy" should be there as well.

Of course that also raises another problem, which is indeed one we don't have a good answer to at present: why is the cosmological constant so small?

The Big Crunch will probably not come about - https://en.wikipedia.org/wiki/Big_Crunch

> However, that gravitational pull is still there -- will it at any point slow down the expansion and retract back inwards?

No. Even if dark energy stopped right now, almost ever pair of galaxies has long reached escape velocity. Their relative speed will slow over time, making them drift apart slower, but it will never reach 0, or even become negative (= contraction).

Wouldn't this theory rule out dark energy since I believe dark energy is used to explain an accelerating, expanding universe?

Yes, I think that's the point.

> Conversely, if the particles were to become lighter, the frequencies would become redshifted.

> If all masses were once lower, and had been constantly increasing, the colours of old galaxies would look redshifted

Can anyone explain this apparent contradiction?

If all masses were once lower, and had been constantly increasing, then those galaxies are now blue-shifted, and blue shifted is what we have come to think of as normal. But billions of years ago, when they were lighter, they were emitting photons that, to us now, in the present, seem red-shifted. They seem redshifted because they come from the past, when things were lighter.

Or in very straightforward terms:

lightweight is red, heavy is blue

If everything constantly grows heavier then the past is red and the future is blue.

Thank you. My head hurts. Your explanation helped.

>Instead, the Big Bang stretches out in the past over an essentially infinite period of time.

So the timeline according to this theory is [Big Bang] - >[Inflation] - >[Increase in mass] & we're seeing the redshift from that?

Head hurts, I love it.

Alvie would be pleased to know that Brooklyn is not expanding.

Does this mean that G is not contant? And would it throw away all our calculations of distances of galaxies in the known Univers?

We know there is an equivalence between mass an energy.

Does this view imply that new energy is appearing in the universe?

The traditional view is that the universe is expanding due to "dark energy", which does presume that yes, new energy is appearing in the universe all the time (at extremely low amounts, but over galactic distances it adds up to cause expansion)

I suppose you are right, I didn't though about that. I find the implications surprising.

Does that means that the universe can't be considered a closed system and entropy increasing is not unavoidable?

As my imagination, the Big Bang is the collision of two universes in the past.

"The ekpyrotic theory hypothesizes that the origin of the observable universe occurred when two parallel branes collided." - https://en.wikipedia.org/wiki/Brane_cosmology#Models_of_bran...

Only a collision between super massive materials could produce such huge larger energy as we see in Big Bang.

Like a cyclical interaction with a neighbor brane which is essentially another universe? Because that's the ekpyrotic universe model. And it fits with your imagination.

What is "ekpyrotic" ? Google Translate can't translate it, sorry.

The first hit from DuckDuckGo for 'ekpyrotic' is the Wikipedia page https://en.wikipedia.org/wiki/Ekpyrotic_universe . It describes the reason Steinhardt created that word as the name for the model.

The Steinhardt and Turok preprint is at https://arxiv.org/abs/hep-th/0111030 , published in the journal Science at http://science.sciencemag.org/content/296/5572/1436.full .

> We propose a cosmological model in which the universe undergoes an endless sequence of cosmic epochs each beginning with a ‘bang’ and ending in a ‘crunch.’ The temperature and density are finite at each transition from crunch to bang. Instead of having an inflationary epoch, each cycle includes a period of slow accelerated expansion (as recently observed) followed by slow contraction. The combination produces the homogeneity, flatness, density fluctuations and energy needed to begin the next cycle.

See also https://en.wikipedia.org/wiki/Paul_Steinhardt#Cyclic/ekpyrot... .

What does the Universe expand into?

Nothing. The concept seems foreign because we're always in something, but there is definitionally nothing the Universe can expand into. If it's easier to conceptualize, just think of 'expansion' as a short way of saying 'distances are getting longer'.

So does it mean that everything in the Universe is getting bigger but since everything is getting bigger we don't know it because relative sizes are the same?

could this help explain the mass of the several copies of the kilogram drifting apart?

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact