How do we know laws of physics are the same throughout the universe? Intuitively I would say they would vary in two natural ways: the constants in the equations may vary or the math in the equations may vary. As a guess they could change over a long time. What is the furthest radius we can prove from earth, with absolute certainty, that the laws of physics do not vary? I am aware this may not be a radius but a more complex shape that cannot be simply described by a radius.

The nearest answer I can think of for a radius is a guess. And that guess is based on the furthest physics experiment we have done from earth. Which I think is an experiment with mirrors on the moon. Therefore if we assume (I don't know if this assumption is totally 100% reasonable) all physics laws hold because this experiment works. Then the radius is to the moon. This doesn't give an concrete answer for the radius, merely an educated guess.

  • 4
    $\begingroup$ Well, the experiments people did on Earth in the 18th century is far, far further in space (and time) than the Moon is from Earth today - and that's peanuts to how long life has existed here. Are you supposing that we "drag" our physics with us through time and space? This is starting to look awfully complicated, for something that has no problem to explain and no predictions to make. I'd hazard to call the idea unscientific - you're trying to replace something that's simple and seems to work with something that's very complex, and doesn't have any observations (or even reasoning) to go with. $\endgroup$ – Luaan Aug 31 '16 at 11:54
  • $\begingroup$ The second paragraph you are referring to about the moon, as I wrote before in the question-it is merely an educated guess. Which of course means the second paragraph may not be correct about the radius because as I wrote before (in that paragraph) its a guess. But its the best answer I could guess for myself at the time I wrote the question. $\endgroup$ – qwerty10 Aug 31 '16 at 18:17
  • $\begingroup$ So you dont think its a good guess then $\endgroup$ – qwerty10 Aug 31 '16 at 18:25
  • 2
    $\begingroup$ I think it's not a good question, once you understand the core of the scientific method. It's a bad idea to think about things that can't be disproven - that way lies madness and dragons :) Look at all those crackpot theories out there - most of them either actively avoid a scientific test, or have no way to be disproven in the first place. And anything that is Earth-centric is suspicious - you think your approach is conservative, since we didn't observe something yet, but it actually introduces more complexity to the model, which inherently gives it more of the burden of proof. $\endgroup$ – Luaan Aug 31 '16 at 19:14
  • 1
    $\begingroup$ This is assumed by the cosmological principle. Scientists generally trust whatever hypothesis requires the fewest assumptions — and assuming the Universe is homogenous is more supported than assuming it is not. $\endgroup$ – Sir Cumference Oct 20 '16 at 17:44

Nothing can be proved "with absolute certainty"; that is not how science works.

We adopt a working hypothesis that the constants of nature are exactly that; both constant in time and space. Then we conduct experiments that attempt to falsify that hypothesis or at least place limits on by how much things might vary.

For reasons that are explained in answers to this Physics SE question (see also this question), only the dimensionless parameters like the fine structure constant can be assessed for their variation - other constants like $G$, $c$ and $h$ are tied up in our system of (measuring) units so we are unable to say whether they are changing or not.

Taking the example of the fine structure constant, observations of absorption lines towards distant quasars put strong limits on by how much this can have varied in space and time (the two are inseparable, since it takes finite time for information to travel to us). So you can find lots of different attempts to do this in the literature - I dug out a few. Albareti et al. (2015) say the variation is less than a couple of parts in a 100,000 out to a redshift of 1 (a lookback time of about 8 billion years or so. Similar constraints exist for experiments carried out in different parts of the solar system. On the other hand, some authors do claim variations of a few parts per million on similar lookback times or in different directions (Murphy et al 2008; King et al. 2012), but these claims are disputed by many, if not most workers in the field.

There is a massive review of this topic by Uzan (2011), which you could read - this really is a broad question. My summary would be - at the moment there is no convincing evidence for any variation in space and time.

  • 1
    $\begingroup$ Closer to home, the natural Oklo nuclear fission reactors yield very strong evidence that the fine structure constant has been constant (to within a very tight bound) for the last 2 billion years. $\endgroup$ – David Hammen Aug 30 '16 at 4:42
  • 1
    $\begingroup$ @DavidHammen Absolutely correct. If someone wished to write an answer based on "lab-based" experiments, I would upvote it. $\endgroup$ – ProfRob Aug 30 '16 at 7:01
  • $\begingroup$ Perhaps a deeper question would be "why are the laws of physics what they are"? (or "what they appear to be"). If a fundamental constant varied wildly we may not recognize the associated law for what it is. $\endgroup$ – JonesTheAstronomer Dec 7 '16 at 1:28

Let's start in the middle:

What is the furthest radius we can prove from earth, with absolute certaintity, that the laws of physics do not vary?

Zero. Proofs are found in mathematics and court rooms, and are impossible in natural science. The best we can do is have falsifiable theories. This holds for every description of reality - there's no "proof" even for the Laws of Gravity.

So, what could we observe that would tell us that physical constants or relationships between physical quantities are different in other parts of the universe, or at other times during its existence?

  • Gravity: For galaxy clusters, we have independent mass measurements from several different sources that agree within their (admittedly large) error bars. Gravitational lensing, velocity dispersion of the member galaxies and X-ray temperatures are all in agreement. So the laws of gravity seem to work even at redshifts up to 0.5 or even higher.
  • Atomic physics: We observe highly redshifted objects. The wavelength of the light emitted by these objects is made longer by the expansion of the universe. Observing redshifted spectral lines of different chemical elements (or molecules) tells us that atomic physics worked the same when and where this light was emitted. If the transition levels between electron orbits had changed over time, we would get different redshifts for the same objects depending on what element's spectral line we observe.
  • Nucleosythesis: Shortly after the big bang, the temperature lowered such that protons and neutrons were no longer created and destroyed constantly. A free neutron has a half live of about 8.5 minutes before it decays into a proton and an electron. Our theories predict that we'd get a helium (2x proton, 2x neutron) content in the universe of about 25%. (The rest of the "normal" matter being essentially all hydrogen), and that is indeed what we observe. Now, the helium content is dependent both on the matter density at the time this took place and the half live of the neutron. From other observations (BAO come to mind) we are fairly certain that we got the matter density about right. Which leaves only a small wiggle room for the half live of the neutron, and hence for changes in the weak force.
  • We've covered gravity, electromagnetism, and the weak force. I don't know any good test for the strong force.

For a change of natural laws over time, we can look at the isotope distribution in rocks here on earth. We should be able to tell whether the decay rate of various elements was different at earlier times by looking at how many of each of their decay products are around.

To summarize, we cannot say with "absolute certainty", but what we observe seems to indicate that natural laws are the same throughout the universe.

  • 3
    $\begingroup$ "court rooms" I would argue that court room "proof" has much, much more in common with science's way of "proving" things (that is, via an inductive reasoning process) than it does with mathematical proofs. Also, "A free neutron has a half live of about 8.5 minutes before it decays into a proton and a neutron." ...Is that sentence correct? I'm not completely sure, but that sounds like it's nearly doubling its mass when it decays. $\endgroup$ – jpmc26 Aug 29 '16 at 21:11
  • 1
    $\begingroup$ Yes, the courtroom bit wasn't too serious. Maybe I should take it out. And yes, it's of course an electron, sorry. Thanks to @sds for the edit! $\endgroup$ – Alex Aug 29 '16 at 22:11
  • 6
    $\begingroup$ It never ceases to amaze me how our education system can force people to take over ten years of science classes to get a diploma and still fail to really teach people what science is. Not long ago, I blew the minds of an entire meeting room full of college educated adults by saying "How about we make a couple hypotheses, draw up some tests that can confirm or invalidate those hypotheses, and regroup to go over the results" - while I'm always happy to look like a wizard, this really is something that any third grader should know how to do. $\endgroup$ – corsiKa Aug 29 '16 at 23:15
  • 1
    $\begingroup$ @MartinArgerami As Temyr wrote, redshift is a constant stretching of the light waves. Example: The sodium D lines at ca. 600nm are 1nm apart (numbers rounded). At redshift 1 we'd see them at 1200nm and 2nm apart. If their separation were different than their redshift would suggest, we'd know that electromagnetic force was different where the light came from. Spectral lines from other elements and other orbitals would be affected differently, too. I don't think it's possible to change electromagnetism in a way that would keep all the energy levels of the orbitals consistent. $\endgroup$ – Alex Aug 30 '16 at 16:00
  • 1
    $\begingroup$ A free neutron decays into a proton, electron, and an electron antineutrino. $\endgroup$ – Howard Miller Aug 30 '16 at 18:55

We can't know for certain. However, we can state with confidence what would be broken were it not to be true, provided a certain mathematical formulation is valid. This is Noether's theorem https://en.wikipedia.org/wiki/Noether%27s_theorem

TL;DR what breaks is the conservation of linear momentum. If you consider that the laws of physics may vary with time rather than place, what breaks is the conservation of energy. Both subject to the constraint that a Lagrangian formulation is valid.

I've encountered serious physicists discussing the possibility that that time-invariance might not hold for the early stages of the universe. The consequence would be non-conservation of energy on the largest cosmological scales, which is where the evidence for this conservation law is least strong. (We have to posit the existence of dark matter and dark energy, and also not all the universe is observable).

  • 2
    $\begingroup$ You mean symmetry under positional or time translation surely? Well, the universe is not symmetric under time translation and the "law" of conservation of energy is not a fundamental part of the General Relativistic description of the universe. $\endgroup$ – ProfRob Aug 30 '16 at 15:33
  • $\begingroup$ @RobJeffries It is not symmetrical (entropy comes to mind), but does it violate conservations? (My guess is not; if conservation laws hold in one time direction, it seems trivial that they hold in the reverse direction as well.) It still gets complicated on the microscopic level (T symmetry alone is violated, but CPT is not), but I'm out of my depth here. $\endgroup$ – Peter - Reinstate Monica Sep 1 '16 at 1:52
  • 1
    $\begingroup$ @RobJeffries Re: "The "law" of conservation of energy is not a fundamental part of the General Relativistic description of the universe": That surprised me; bummer ;-). Red shift is a nice simple example for energy "going away". After reading a bit (preposterousuniverse.com/blog/2010/02/22/… helped, including some of the discussion) it seems that more complex forms of energy-momentum conservation still hold... but maybe that is a misunderstanding (cf. physics.stackexchange.com/a/35438/72043.) $\endgroup$ – Peter - Reinstate Monica Sep 1 '16 at 2:22
  • $\begingroup$ The conservation of energy already is broken on the cosmological scale. Yes, think of your Noether’s theorem argument. $\endgroup$ – Incnis Mrsi Sep 11 '16 at 12:29

One problem with your question is that it is a bit of a paradox. If a Law of Physics appears to vary depending on time/place being observed then, because of what it means to be a physical law, we've simply misunderstood the law itself or aren't observing all forces at work.

Here's a super simple example.

enter image description here

These people haven't found a place in the universe where gravity acts differently, they're simply being pushed harder by a fan than the gravity is pulling down on them. Of course, if the only information you had about them was this picture you wouldn't know that and might think that gravity acts differently where they are.

If scientists observe variances is how a law behaves and simply waves their hand saying "oh the law works differently there" then that isn't science anymore. We'd want to know why the law appears to work differently in one place vs another.


One example that is maybe more to OP's point is dark energy. We observe that the Universe is expanding at an increasing rate even though our Laws of Physics, specifically gravity predict that its expansion would cause it to decelerate. Rather than shrugging their shoulders and saying "well the Laws of Physics just work differently at the edge of the Universe" scientists theorized something called dark matter to explain why the Universe's expansion is accelerating despite gravity.

  • $\begingroup$ I dont think this is a good answer. The other three answers (discussing math e.g. papers) that were posted earlier are more what I had in mind for a potential answer. And my question is written from the point of view of a physicist thinking about the problem. $\endgroup$ – qwerty10 Aug 30 '16 at 16:06
  • $\begingroup$ So in my opinion your solution is an invalid answer to my question. Its not what I was looking for in an answer. $\endgroup$ – qwerty10 Aug 30 '16 at 16:13
  • 2
    $\begingroup$ I agree that the other answers are wonderful and even better than mine. However I also think that there is a flaw in your question, specifically the implication that the science community would accept that a physical law simply works differently in one area of space than it does in another. If an observation in distant space ever contradicts our known laws, the explanation will never be "The laws of physics simply work differently out there". See my edit for another example. $\endgroup$ – Dean MacGregor Aug 30 '16 at 22:16
  • $\begingroup$ My question does not mean that the discovery of new mathematical physics (or new theories) is not possible to explain experimental information such as your dark energy example. Its considering the actual laws such as the mathematical physics equation for example- for dark energy or another math physics equation for another phenomenon .So your edit/paragraph about dark energy ,in my opinion, is not an answer. $\endgroup$ – qwerty10 Aug 31 '16 at 3:21

"They (the laws of physics) would vary in two natural ways:"

  1. the constants in the equations may vary or

    Possible. We are fairly certain about the values of constants up to smaller astronomical scales (sub-galaxy). On the galactic scale and beyond we have strange deviations from what we would expect. On the galactic scale we currently attribute the deviations to "dark matter" which to me seems little more than a placeholder for the unknown.

    On a universal scale the apparently accelerating expansion of the universe is usually attributed to a different placeholder for the unknown, "dark energy"; or it may be that general relativity as we understand it does not hold on large astronomical scales, so that for example the gravitational constant is not in fact a constant, or whatever. This is fairly strong evidence that what we think to know is wrong or incomplete, so the answer is "at the universal scale we know that we are wrong".

  2. the math in the equations may vary.

    That's the one thing which we are fairly sure about: The math will not vary. It may be incomplete, or wrongly applied, or whatever; but the math is the one thing which does not vary.

  3. Let's also not forget that there is famously "plenty of room at the bottom". We don't even know the number of dimensions at very small (sub-nuclear) scales, we don't know how the single threads of the fabric of space time are knit together, etc.

  4. On a more speculative level this may not be the only universe but for example just one shard of a multiverse; Lee Smolin wrote about the idea of an evolution of universes. The other ones would have most likely different constants, or differ in some other funny way.

  5. On an even more speculative level: If you ask Elon Musk and others, we live in the Matrix anyway, and all laws of nature are subject to change at the whim of a keystroke equivalent by the sys admin. Something like /gamemode 1 qwerty10, and your credit card runs never empty.

  • 1
    $\begingroup$ What are "single threads of the fabric of space" that "are knit together" ? $\endgroup$ – Roland Pihlakas Aug 31 '16 at 19:59
  • $\begingroup$ @RolandPihlakas Interacting Higgs Bosons? Entanglement? Whatever. We don't really know, but a metaphorical fabric it is, and thus has metaphorical threads in it ;-). Robert Laughlin seems to imagine some kind of sophisticated ether; space (-time) just behaves so much like a vibrating medium that the modern classic's dismissal of a medium may be denying the elephant in the room. $\endgroup$ – Peter - Reinstate Monica Sep 1 '16 at 1:41

Science is based on guesses, paraphrasing Feynman. We guess that something works in a certain way. A good guess explains the existing data and makes predictions which can be tested The best guess is the good guess which is simplest i.e. minimises the number of additional assumptions. So the assumption by Newton that gravity works for planets in the same way as it does for throwing stones while walking along a beach was, in essence, just a guess.


i am by no means a scientist, thus nor an astrophysicist. i have a background in electrical engineering and a curiosity for cosmology. i ended up here essentially because i'm looking for answers to the question asked above.

it seems to me that the following information is revelant to the question : a fairly recent (2017.09.20) article published on the NASA website mentions a study which reveals that the two methods used to compute Hubble's constant (one is based on observations of type 1a supernovae, the other on the CMB) disagree (although the Standard Model of Cosmology predicts their agreement) :

« A recent study using the first method yielded an 8% greater expansion rate than the second method’s result. » - https://science.nasa.gov/science-news/news-articles/hubbles-contentious-constant-news

the article doesn't mention a clear explanation for this discrepency. for instance, perhaps there are holes in one or both computation methods.

if i understand correctly : since it is believed that the CMB informs us about the early universe, but this is not so for type 1a supernovae, then another possible explanation is that both measurements are valid, and the discrepency means that something has changed over time. for instance, the article asks the question « Or are dark energy’s or dark matter’s properties changing over time? ». given the importance of Hubble's constant, perhaps this points to the fact that physics have changed over time.

  • $\begingroup$ Wendy Freedman, Sullivan professor of astronomy and astrophysics at the University of Chicago says, “It could be that we don’t understand the uncertainties well enough to know why these two methods differ.” $\endgroup$ – ProfRob Feb 18 '18 at 10:03
  • $\begingroup$ thanks for that comment. possibly you can enlighten me on what Freedman actually means by that statement ? i.e. uncertainties about what ? and what is not understood about those uncertainties ? $\endgroup$ – poligraf Feb 18 '18 at 13:48
  • $\begingroup$ is this a third method, or another instance of the type 1a supernovae method ? nasa.gov/feature/goddard/2018/… « The difference between the two values is about 9 percent. The new Hubble measurements help reduce the chance that the discrepancy in the values is a coincidence to 1 in 5,000. » $\endgroup$ – poligraf Feb 24 '18 at 2:36

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.