Hubble measured high redshifted galaxies to discover the cosmic expansion. In a hypothetical universe where only one galaxy exists, would there still be observational evidence for the Big Bang theory? In a very distant future when everything except our Milky Way fade away from the horizon, what would be the cosmology developed by civilizations in the Milky Way?

  • $\begingroup$ there are other ways of measuring expansion of a universe, see en.wikipedia.org/wiki/… $\endgroup$
    – anna v
    Jan 14 at 12:48
  • $\begingroup$ I think the question raises the experimental point how much information we can extract about the past if all the physical markers of the past are red-shifting to zero... CMB gone, gravitational background gone, neutrinos gone. I am not a relativist, so I can't answer the question conclusively, but intuitively I thought that in an expanding universe that's where we are going to find ourselves, eventually: a lonely island of a few galaxies or even less surrounded by blackness. A frightening thought, indeed. $\endgroup$ Jan 14 at 13:16
  • $\begingroup$ The question strikes soo many thoughts. Would the big bang theory apply to such a hypothetical universe? Maybe in a trillion years, the galaxies could move away but should still emit light towards us and would look like dots. Would not like to answer the question because I don't have complete knowledge on the subject. $\endgroup$ Jan 14 at 13:47
  • 1
    $\begingroup$ Leonard susskind mentioned this in one his lectures, when talking about a future when the cosmic horizon does not extend too much beyond the milky way. Basically, they would likely never figure it out. $\endgroup$ Jan 14 at 14:35
  • $\begingroup$ By what I've read about spatial expansion (mostly papers by the well-known team of Davis & Lineweaver, written at various levels of sophistication but all having titles that include the phrase "Expanding Confusion"), the galaxies remain stationary amid such expansion, which Davis has specified (in a remark to Chowderowski) "is not a force or drag": Consequently, I understand the question, but not the only answer posted so far. $\endgroup$
    – Edouard
    Jan 14 at 17:28

3 Answers 3


If there were just one galaxy, there would still be a microwave background (MWB) and we would still have general relativity. The abundances of hydrogen, helium, lithium, and beryllium in space are also clues. The Friedman equations that solve GR for the dynamics of the universe imply that the universe either expands or contracts. By studying the MWB, we would come to much the same conclusions that we have today. Model fits would be missing a few constraints, so values of cosmological parameters would have greater uncertainty, but not by a large factor.

Of course, it does depend on what laws of physics or cosmological parameters are different to result in only one galaxy forming.

  • 3
    $\begingroup$ I thought that if he can't see the redshifted galaxies any longer, then he also can't see the much colder microwave background? Am I making a mistake here? Is it a given that the total angular momentum of a very old galaxy is close to zero, which we would need for the cosmological isotropy assumption or can we end up with an elliptical galaxy with a well defined rotation axis that would define a highly anisotropic "visible universe"? $\endgroup$ Jan 14 at 14:11
  • 7
    $\begingroup$ @FlatterMann That's the problem with non-physical hypothetical questions: at some point, you got to decide where to break physics. $\endgroup$
    – JEB
    Jan 14 at 14:58
  • $\begingroup$ I interpreted the question as implying that only one galaxy formed, ours, and the rest of the universe is just gas that did not collapse. Our galaxy would be an Elliptical galaxy because rotation requires interactions with neighboring galaxies to get spun up. But, that would not change cosmology by much. $\endgroup$
    – eshaya
    Jan 16 at 20:08
  • $\begingroup$ Just a comment about the comment: "rotation requires interactions" is backwards... disks form because dissipative collisions allow gas to shed its velocity dispersion. That doesn't work for stars. So if you form a galaxy, it makes a disk out of gas, which turns into a disk of stars. After the galaxy converts its gas into stars, if you heat it with interactions, the stars disperse into a spheroid and remain in that form. $\endgroup$
    – Sten
    Jan 18 at 16:06
  • $\begingroup$ Pancaking does not get very far before star formation occurs. Dissipation plus rotation results in a disk. If a collapsing cloud has little or no rotation then the collapse continues to much higher 3d densities resulting in early star formation. That is how galaxy bulges, globular cluster, and the old giant ellipticals. All have low rotation. Specific angular momentum is a key difference between Hubble types. $\endgroup$
    – eshaya
    Jan 20 at 17:31

It would be pretty damn tough, to the point where how to do it is a research question astronomers might think about. But note that the hypothetical universe with only one galaxy would be very similar to our universe in the very far future, when the cosmic horizon is so close that we can no longer see other galaxies and maybe even the cosmic microwave background (because the wavelength of the CMB is larger than the observable universe).

Here's a relatively recent source, where the author considers the universe 1 trillion years in the future. With absolute darkness beyond Milkomeda (the result of the projected Milky Way + Andromeda merger), galaxy dynamics could still eject stars from the galaxy, and those stars would show Hubble recession. These hypervelocity stars would be rare and the effect small, but it can be done. The paper on arXiv: https://arxiv.org/abs/1102.0007

  • $\begingroup$ I think by that time expansion in our vicinity would be so high that even though we wouldn't any longer see other galaxies and CMBR, there could then happen other bizarre effects due to so large expansion in so close distance which could enable to make prediction of expanding universe. $\endgroup$ Jan 15 at 6:30
  • $\begingroup$ So if the expansion accelerates and most of the currently visible matter is gone from sight, then an astrophysicist will likely arrive at a false estimate for the original matter density and the original expansion rate, which also gives a wrong age. Yes, they can tell that their universe is falling apart, but they lost all information about the state it came from, similarly to how we lack observational evidence about the era "before" the big bang. There will, of course, be giant stone monuments on very old planets telling them the original parameters... in languages they can't read. $\endgroup$ Jan 15 at 10:55
  • $\begingroup$ Acceleration of ejected stars really just measures the energy density of the dark energy, which can also be measured by other means, e.g. the structure of the galaxy itself at large radii. $\endgroup$
    – Sten
    Jan 18 at 21:25
  • $\begingroup$ @Sten I imagine that would depend on where dark energy starts to dominate. If that horizon is far-enough away, one would still need hypervelocity stars. $\endgroup$
    – Allure
    Jan 19 at 2:39
  • $\begingroup$ That horizon would be about 1 Mpc, which is about the size of the Local Group currently. It's already been suggested to use Local Group dynamics to measure dark energy (arxiv.org/abs/2306.14963) and the final Local Group shouldn't be much smaller than its present size (maybe half its present size, just due to the virialization arguments; dark energy starves it of accretion to further contract it). $\endgroup$
    – Sten
    Jan 19 at 3:06

For your hypothetical scenario, the cosmologist would prove that the universe is static, and it's always been like this, without any bound on time that is it's an eternally static universe. The later scenario of our universe where everything falls beyond the observable universe would also imply the same thing as your hypothetical scenario: Lonely galaxy scenario. Let me know if you wanna know anything else.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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