Spurred by reading a recent obituary of Penzias, who with Wilson discovered the cosmic microwave background, I was led to read their letter in a 1965 issue Astrophysical Journal Letters, which was published side-by-side with a companion letter in that same issue by Dick, Peebles, Roll and Wilkinson.

The DPRW letter lays out a diagram of a proposed "thermal history of the universe", envisioning an early plasma universe dense with ions --- free nucleons and electrons --- in which photons are constrained to have a very short mean free path by constantly bouncing off all those ions. There is a transitional moment labelled "plasma recombines", at which the nucleons and electrons form neutral atoms and the photons are freed. The point of the letter is these photons, viewed at present with wavelengths shortened by eons of red-shift, form the cosmic microwave background observed by Penzias and Wilson, thus providing solid physical evidence for the "thermal history".

One interesting feature of the DPRW letter is that they do not take the cosmic microwave background observed by Penzias and Wilson as evidence for the big bang per se. The explicitly say that their theory of temperature of the early universe could apply in either of two scenarios: a "closed universe, oscillating for all time"; or a universe with "a singular origin".

The moment labelled by DPRW as "plasma recombines" seems to be the same as what I have seen currently labelled as the "recombination era" which occurred 378,000 years after the big-bang. In current web articles on big-bang theory I have seen this "recombination" terminology critiqued: nothing is being "recombined" because it was never "combined" in the first place.

This leads to my first question:

  • Is it true that origin of the phrase recombination era lies in the possibility of a closed universe, oscillating forever?

One way or the other, it looks like the discovery of the cosmic microwave background was not originally taken as evidence for the big bang per se, which is what I usually see it sold as nowadays. This leads to my more substantial followup question:

  • What further physical evidence is there that separates the big-bang theory from the oscillating universe theory?
  • 6
    $\begingroup$ I believe it simply comes from atomic physics, where the process of electron capture by an ion is called recombination. It reflects laboratory experimental conditions, not cosmological. $\endgroup$
    – John Doty
    Jan 24 at 15:51
  • $\begingroup$ The historians on hsm.stackexchange.com may have more insights into this question. $\endgroup$
    – PM 2Ring
    Jan 24 at 16:25
  • 2
    $\begingroup$ John Peacock writes in his textbook (1999) that "The historical origin of the name lies in HII regions: interstellar plasma is ionized by ultraviolet radiation from a hot star, and the region emits characteristic recombination radiation as the electrons and ions continuously re-form atoms before being photoionized once more." (emphasis from source) $\endgroup$
    – Sten
    Jan 24 at 16:27
  • $\begingroup$ What is going on with the title? $\endgroup$
    – Mauricio
    Jan 24 at 16:36
  • 1
    $\begingroup$ The title is meant to question the usage of the prefix "Re" in the word "Recombination", as explained in more detail in the body of the question. $\endgroup$
    – Lee Mosher
    Jan 24 at 20:26

2 Answers 2


Regarding the term "recombination", this is a misnomer inasmuch as before the so-called recombination era the plasma had not ever "combined", re-ionized, and then subsequently re-combined. Nonetheless, the term has stuck.


The CMB has a lot of information about the universe at 378000 years of age. See Secrets of the Cosmic Microwave Background for a summary. But it isn't the earliest information we have. As described below, the CMB provides data on conditions of the universe back to about $1$ second after the big bang.

The conditions back then are predicted from physics models. We have experimental data from particle accelerators that gives hard data to confirm our understanding of physics under condition that were present at $t = 10^{-13}$ seconds. The theories themselves provide predictions back to $10^{-43}$ seconds. Before that, the universe was so hot and dense that all theories break down. See What happened before the Big Bang? for a summary. It includes speculations about what happened before the big bang. None of those include an oscillating universe. But they don't say why an oscillating universe has been ruled out. So I don't have a real answer to your question.

The evidence from accelerators allows us to predict what would happen as the universe cooled to a state where nucleons could combine to form nuclei. The prediction for the ratios of numbers of H, He, and Li nuclei match what we observe. So this is confirmation that if the universe oscillates, it the oscillations go back much much earlier than t = 380000 years. This talks about it. The birth, life and death of the universe – Public lecture by Dr. Don Lincoln

Here is more detail from PBS Space Time on what can be learned from the CMB. The Cosmic Microwave Background Explained! Episodes 4 and 5 go It goes into Acoustic Baryon Oscillations. These allow us to predict the relative abundances of baryonic matter, dark matter, and dark energy.

Here is information on the Cosmic Neutrino Background. What Exactly Is The Cosmic Neutrino Background? Here's Why It Matters It talks about what it is and how difficult it is to directly measure these neutrinos.

The CNB left an imprint on polarization of the CMB. Here is a longer description. The Polarized CMB: From Neutrinos to Gravitational Waves. Here is information about the polarization of the CMB from the European Space Agency. Polarisation of the Cosmic Microwave Background: full sky and details )

  • 2
    $\begingroup$ Earliest hard data about conditions of the universe are at about 1 second (neutrino decoupling and its imprint on the CMB) and a few minutes (nucleosynthesis and its imprint on element abundances). Particle accelerators tell us what the universe would look like at earlier times under the simplest assumptions, but those assumptions don't have to hold. $\endgroup$
    – Sten
    Jan 24 at 16:40
  • $\begingroup$ @Sten - Thanks. I corrected some statements and added more references. $\endgroup$
    – mmesser314
    Jan 25 at 16:22

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.