Light from celestial objects is old. In the case of galaxies, it's millions of years old. It seems plausible to me that light might show signs of its age.

I was surprised that a Google search only turned up one study in this area: Measurement of the speed of light from extraterrestrial sources. It looked at the speed of light from several bright stars: Aldebaran, Capella, and Vega. The results showed that the speeds were different!

My question is, have there been other studies by physicists that looked at old light versus new? It would be so interesting to view in an interferometer light that is a million years old. I can think of many other tests, and I'm sure physicists could think of more. Why hasn't this been done or has it? Could we probably find an age marker by looking closely at the old light?

  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$
    – ACuriousMind
    Nov 6 '17 at 16:50
  • 2
    $\begingroup$ IMO, an ideal answer to this question would be (pointing out the obvious, to which standing answers at this instant, have done justice to) + (addressing how the linked article is compatible with this). I am really looking forward to the second part. :) $\endgroup$
    – 299792458
    Nov 6 '17 at 17:11
  • 1
    $\begingroup$ A lot of celestial light is not very old: xkcd.com/1342 $\endgroup$
    – Henry
    Nov 7 '17 at 12:45
  • $\begingroup$ Relevant, but not a duplicate: physics.stackexchange.com/q/69448/26076 $\endgroup$ Nov 8 '17 at 6:14

Light does not "experience" time, the concept "age" does not apply to light in a meaningful way (with respect to human experience). [As background; recall clocks slow for objects as they near the speed of "light" reaching a theoretical 0 if full light speed were attainable.] A thought experiment clock on a photon would therefore stand still. A photon's source does have an "age" in the traditional (human experience) sense, and it is standard that we say the light is as old as it's source. That "age" does not then carry with it the traditional effect of aging.

While the light source ages in a traditional fashion and may in fact be completely burned out though we can observe it today from our distant position in space, any photon from an object no matter how old the source is in no way different than a newly created photon presuming it is the same wavelength. As I view it you could not discern the "age" of light without knowledge of its source, because light is in reality timeless.

  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$
    – ACuriousMind
    Nov 6 '17 at 16:50

The oldest light in the universe is the Cosmic Microwave Background, which is only about 380,000 years younger than the universe. It shows signs of its age in the redshift that has occurred as the universe expanded- the CMB would have been visible when it was created, and is now in the microwave spectrum. Besides that, however, the results are consistent with it acting like normal "young" light.

No experiment can prove that light does not evolve in some way over some timescale- we can only really test specific ways we expect light might evolve, and set limits on the timescales they must occur over. So the idea that light might evolve over time in some way is not really falsifiable. If a grad student were to work on this, I'd definitely say it should be in the context of specific, falsifiable theories regarding how light could evolve over time.

As far as I know, all models that have the photon evolving at all require the photon to have a non-zero rest mass. If photons have a non-zero rest mass, they could decay into other particles over some long timescale, they could oscillate like neutrinos do to another particle, and they would slow down over time (very slightly) as they were redshifted. The Particle Data Group lists current limits on the photon mass as $m_\gamma < 10^{-18}~\rm eV$. Note that this implies that the speed of a CMB photon is within $2\cdot10^{-4}\rm\frac{meters}{universe\ age}$ of the nominal speed of light, so if it exists the slowdown would be far too small to measure directly.

  • 11
    $\begingroup$ "You can't prove a negative" is simply wrong. Every statement can be written in a negative form. Is it impossible to prove anything? "You can't prove a negative" is also a negative. How did you prove it? :) See departments.bloomu.edu/philosophy/pages/content/hales/… $\endgroup$ Nov 5 '17 at 10:14
  • 8
    $\begingroup$ @tfb: Please note the difference between "Here's a negative we cannot prove yet" and "You can't prove a negative". Also, I'd argue that we cannot prove anything in physics. We merely have models that are considered "good enough" for the time being. $\endgroup$ Nov 5 '17 at 18:20
  • 15
    $\begingroup$ If you can't prove a negative, I'm gonna have to hand back my math degree. $\endgroup$ Nov 5 '17 at 18:40
  • 11
    $\begingroup$ @EricDuminil I was referring not to a law of logic, but the general principle in science that it is impossible to show the nonexistence of something sufficiently open ended, and the general asymmetry in difficulty between proving the existence of something that exists and proving the nonexistence of something that does not. Still, I see your point and have removed the offending phrase ;) $\endgroup$
    – Chris
    Nov 5 '17 at 21:17
  • 5
    $\begingroup$ @EricDuminil We have no evidence that light does not decay on the timescale of $t\sim 10^{100}~\rm yr$, for instance. Or $10^{10^{10^{10}}}\rm yr$. Or someone could make a theory that light oscillates to a non-SM gauge boson that is identical to the photon except it couples to the Higgs boson slightly differently. Or any number of crazy theories that always have "some place to run" whenever an experiment excludes part of their parameter space. $\endgroup$
    – Chris
    Nov 5 '17 at 21:53

One can ask the question "how old is this piece of rock" and get good answers from studying the changes in the lattice structures with time, the dispersion of other atoms in the lattice which also depends on time, etc. We are able to do this because the basic lattices of the rock are constant in time. They might break down at a certain rate, interact with extra atoms at a certain rate etc., but the base is constant in time.

Light is made up of zillions of photons in superposition (not interaction), the photons' quantum mechanical wavefunctions adding in such a way as to produce the observed light. In contrast to atoms in a lattice, the structure is instantaneous, i.e. a slice in this classical light wave solution perpendicular to its direction of motion


will contain zillions of photons running along, keeping no history except the history of the frequency of the wave which is their energy $E=h\nu$. This can be Doppler shifted and spectral lines can give a history of the travel time and the origin. So it is possible to find a history in the wavetrains coming from afar for special situations, (like the Doppler shift) and where there are intermittent signals as also with pulsars, one might get extra information from the variations of the amplitude of the light arriving, as with neutron star mergers which give the history of the merge in electromagnetic radiation (recently also with gravitational radiation, but that is another story).

Actually this completely new research on time crystals might open a window in the future of something similar to the contaminations time induces in a normal crystal where a history can be unraveled, but I am not holding my breath. No natural time crystals have been proposed which might have sent wavetrains that could be used to extract a history from the light arriving.


Light emitted by cosmic sources usually has spectral lines of known elements, such as hydrogen. If the light source moves from us or light lost its energy en route, the spectral lines get shifted towards red end. Given the universe expands, so that all light loses its energy when travelling, we can roughly entimate what distance light has travelled by measuring the shift of the spectral lines.

  • $\begingroup$ yeah, but how to tell if it's an ancient photon from a distant star, or a young one from a light source in the same room. you can't ask it for a birth certificate. $\endgroup$
    – Jasen
    Nov 6 '17 at 8:33
  • $\begingroup$ @Jasen it cannot be a young one from the same room because it has hydrogen spectral lines shifted. $\endgroup$
    – Anixx
    Nov 6 '17 at 14:39
  • 1
    $\begingroup$ @Jasen The key is that you're not measuring one photon, you're measuring a population of photons that you know came from a star (but may not be sure how far away the star is) $\endgroup$
    – Random832
    Nov 6 '17 at 16:20

In my opinion, light slowly loses a small amount of energy as it travels very long distances. This manifests itself as a red shift in light from distant galaxies. This energy is not lost, but a result of the light particle causing a push of the light source and the light destination away from each other. This could be considered an example of the electromagnetic force pushing two galaxies away from each other as light bounces back and forth between them. Although the current scientific consensus is that this red shift is caused by space itself expanding, I believe that such an expansion of space would have caused the large scale structures of the universe such as clusters of galaxies to dissipate long ago. However, light travelling between galaxies would cause a force pushing the galaxies away from each other. This would have the same effect as space expanding, although without the need for "dark energy" or a violation of the law of conservation of mass and energy. Whether you share my opinion or strictly follow the scientific community, you can get a general idea of how old light is by measuring its red shift relative to the expected frequency.

  • $\begingroup$ That’s an interesting theory. Maybe you could conduct some experiments to see if nature supports your hypothesis. $\endgroup$
    – Lambda
    Nov 12 '17 at 14:45
  • $\begingroup$ At close range, when a photon bounces back and forth between two electrons, it causes the electrons to accelerate away from each other. This acceleration of the electron causes a red shift of the photon due to the difference in the relative speed of the electrons before and after the photon strikes the electron. $\endgroup$ Nov 12 '17 at 20:41

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