The September 19, 2023 podcast with transcript Why the earliest galaxies are sparking drama and controversy among astronomers includes the following:

An article published earlier this year in the journal Monthly Notices of the Royal Astronomical Society came to this conclusion after combining two models of the universe. One is the commonly accepted model for the expansion of the universe. This model explains that as the universe expands, the light from galaxies must travel further and therefore shifts from a bluer to a redder spectrum of light. The other model it is combined with has been debunked. It's called the tired light model, and it alleges that as light travels across the universe, it gets redder because it gets "tired," or loses energy.

and addressing that is astronomer Jorge Moreno:

Moreno says that while he thinks that combining the models is clever, it is not supported by scientific evidence.

"I think in science, if you already have a model that's simpler than that, you should stick to it—unless you have extraordinary evidence to do otherwise."

Moreno also cautions people against quickly jumping on this supposition that the universe is twice as old as previously thought. If it were true, scientists would be able to prove it through the direct observation of stars and galaxies that are older than 13.8 billion years old—the current accepted age of the universe.

No such evidence has been found.

I'm not asking about Gupta's inclusion of "tired light", though I'm also not overly enthused about Moreno's advice that we should "stick to" simple models unless there is "extraordinary evidence to do otherwise" either.

Instead I'd like to better understand:

Question: What would evidence of stars and galaxies that are significantly older than 13.8 billion years old look like? In what part of space has such evidence been looked for and not found?

  • 9
    $\begingroup$ "the commonly accepted model ... explains that as the universe expands, the light from galaxies must travel further and therefore shifts from a bluer to a redder spectrum of light." That doesn't sound right. It sounds more like the "tired light" model. Rather, in the commonly accepted model, Hubble's law observes that further galaxies are receding from the earth faster than nearer ones (implying the expansion of the universe), and this greater recession speed, not the distance itself, is what causes redshift. This doesn't affect your question, just seems like poor journalism. $\endgroup$
    – LarsH
    Commented Sep 19, 2023 at 12:56
  • 2
    $\begingroup$ FYI, the article linked to with your Monthly Notices of the Royal Astronomical Society was also linked to, and discussed (actually, mostly debunked), in an answer on the Physics SE site. $\endgroup$ Commented Sep 19, 2023 at 16:58
  • 2
    $\begingroup$ @LarsH I don't think it's not poor journalism; rather it's a matter of interpretation: The emitting and the observing galaxy do not move through space, and light in an expanding universe is redshifted along its path. The longer a photon travels in an expanding space, the more it is redshifted, so I think it is a perfectly fair way to describe it. $\endgroup$
    – pela
    Commented Sep 20, 2023 at 6:00

1 Answer 1


Assuming the cosmological principle still holds, then we might expect that your own galaxy and those around us should contain evidence of stars/objects older than 13.8 billion years old.

These might include:

Very cold white dwarfs. White dwarfs colder than around 2000 K with relatively low mass ($\sim 0.5M_\odot$) would probably be older than 13.8 billion years old. This a combination of the likely progenitor lifetime and the length of time it would take a low mass white dwarf to cool to these temperatures.

Low mass $(<0.8M_\odot)$, very low metallicity $(z< 10^{-4})$, single, evolved (giant) stars. The main sequence lifetime of a $0.8M_\odot$, low metallicity star is longer than 13.8 billion years, so finding evolved examples, where any binary mass transfer can be excluded would be key evidence. Estimating the mass of an isolated, low metallicity giant would probably require asteroseismology.

In our own and local group galaxies we could look for globular clusters (or actually any cluster of stars) that has a main sequence turn-off below about $0.8M_\odot$. Again, this is because the main sequence lifetime of such stars would be older than 13.8 billion years.

The latter two examples don't really involve any exotic or poorly understood physics. We pretty much know how low mass stars evolve and it isn't much of an extrapolation from known elderly low mass examples. Thus, providing mass estimates of sufficient precision could be obtained, there is unlikely to be much "wriggle room".

The white dwarf evidence might be more equivocal and model-dependent. There are aspects of the physics of cold white dwarf cooling that are less settled (a pun!). For example, the extent to which quantum effects like Debye cooling become important and whether diffusion of oxygen through carbon can release further gravitational potential energy at late times. Also, cooling times are quite mass-dependent and it's difficult to get an accurate mass for a white dwarf.

As to where has the evidence NOT been found - isolated very cool white dwarfs could only be found in infrared surveys (which have been done - e.g. with WISE) and probably only in the Solar neighborhood. Evolved, low mass stars would be much easier to find over a large fraction of the galactic halo and have, as far as I know, not been found among any metal-poor giants. Having said that, it may be the case that asteroseismology, and hence good mass estimates, are lacking for many candidates. Very old clusters would likely be spotted almost anywhere in the galaxy (apart from the galactic plane) and there would be a good chance of finding them if they existed in the Magellanic clouds or the closest local group galaxies.

  • 2
    $\begingroup$ I love these wonderful, insightful, definitive and yet easily understood answers - thanks! $\endgroup$
    – uhoh
    Commented Sep 19, 2023 at 12:29

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .