I've seen several reports that the entropy of the Universe at the time of the Big Bang was lower than it is now.

This seems impossible.

If we view entropy as a function of the number of micro states consistent with a given macro state, I cannot imagine how the entropy of what was basically a giant fire (at the time of the Big Bang) is not higher than the entropy of the current state of the Universe, which has stable planets, and life.

This is not inconsistent with the Second Law as a local phenomenon, it just says what observation plainly suggests -

Gravity caused the universe to become less complex over time, eventually supporting life.

On a smaller scale, the entropy of a fire must be higher than the entropy of a stable object, since the fire is obviously rapidly changing state, and therefore, presumably, is capable of being in a larger of number of states, while still being a fire. In contrast, a cup probably does change states somewhat, but obviously not as often, and obviously not on the same scale, in that the macro state of a cup is roughly constant.

As a result, common sense suggests that the entropy of a fire is higher than the entropy of a stable object.

Also, the temperature of the universe has consistently decreased over time:



So, why do physicists say that the entropy of the Universe around the time of the Big Bang was lower than it is today?

It doesn't make any sense, yet I've heard two reputable physicists make this claim, which is on its face almost certainly incorrect.

I'm wondering if someone can put forward a simple calculation that explains where this claim comes from, rather than simply quoting the Second Law, and saying it must be so, when available evidence strongly suggests otherwise.

This would involve taking the temperature at the time of the Big Bang, and expressing the entropy as a function of that temperature, and showing that entropy decreases as you move away from the Big Bang in time - and I don't see that happening.

The bottom line is that the universe is both colder and more structured than it once was. This means you can describe its structure with less information. So any statement that the universe has increased in entropy, or complexity, since the Big Bang, borders on nonsense, in my opinion.

It's also perfectly reasonable to assume the universe became more structured with time, since it started out so hot, that its kinetic energy overpowered any of the stabilizing forces like gravity, charge, etc.

You can pretend otherwise, but the reality is that gravity decreases local complexity over time, and other small scale forces do the same, forming atoms, eventually molecules, and then apparently, life itself.

  • 4
    $\begingroup$ Why do you think the universe was a giant fire? Why not cold at absolute zero? $\endgroup$
    – safesphere
    Nov 3, 2019 at 15:28
  • 4
    $\begingroup$ The temperature quoted in the link is Fermi temperature. An object can have a very high Fermi temperature and be at absolute zero of thermodynamic temperature. For example, it would feel cold to touch - en.m.wikipedia.org/wiki/Fermi_energy $\endgroup$
    – safesphere
    Nov 3, 2019 at 15:34
  • 19
    $\begingroup$ @Feynmanfan85 are you sure that expressions like "You can pretend otherwise..." and "That's total rubbish" encourage constructive exchange of ideas? $\endgroup$
    – IMil
    Nov 4, 2019 at 6:22
  • 2
    $\begingroup$ What happens to the entropy of a proton-electron system when they combine to form a hydrogen atom? Hint: don't forget about the photons. The same is true of gravity - for a cloud of matter to collapse under gravity, it needs to shed its (non-negative) gravitational potential energy. The planet has lower entropy than the cloud of matter; the universe as a whole doesn't. $\endgroup$
    – Luaan
    Nov 4, 2019 at 13:56
  • 1
    $\begingroup$ @IMil This question could do with having far less rhetoric / ranting. It could probably be fairly easily shortened to 2-3 concise paragraphs whilst still conveying the point of the question. I actually removed my initial +1 as I progressed through reading it. $\endgroup$
    – JBentley
    Nov 4, 2019 at 20:44

4 Answers 4


Two issues:

1) Your objection would seem to apply to any explosion. The initial state is hot, the final state is cold. What happens is that the initial state has high entropy density, but small volume. During the explosion volume increases, and entropy density decreases. Overall, the total entropy goes up somewhat, because of non-equilibrium processes (as required by the second law).

2) The main issue with the big bang is a different one. The entropy in non-gravitational degrees of freedom is large to begin with and does not increase by very much. However, for a system that interacts gravitationally the state of maximum entropy is a black hole. The initial state of the universe is believed to have had a fairly smooth geometry, so its entropy was much smaller than it could have been.

  • $\begingroup$ No offense, but what you're saying is that the basic definition of entropy, which is the logarithm of the number of micro states consistent with a given macro state, doesn't apply. Because if you apply that definition, entropy has obviously decreased since the big bang, since the universe has cooled. You're taking a useful concept, entropy, and turning into something useless and counterintuitive. The bottom line is, the universe is colder, and more stable than it once was. Call that whatever you want, it's the truth. $\endgroup$ Nov 3, 2019 at 16:34
  • 20
    $\begingroup$ @Feynmanfan85 it cooled down but it also got bigger. Bigger things have higher entropy (logarithm of number of states) than smaller things. An expanding gas can increase its entropy while decreasing its temperature. This is the basis of any heat engine and historically is one of the fundamental insights on which thermodynamics was built. $\endgroup$
    – N. Virgo
    Nov 4, 2019 at 1:33
  • 4
    $\begingroup$ @Feynmanfan85 : From reading your Quetsion and some of your comments, you seem to be conflating entropy with heat. Why are you taking a useful concept, entropy, and turning it into a redundant label for heat? Additionally, you seem to completely ignore positional degrees of freedom, of which there are currently around $10^{90}$ more than during the Big Bang. You also seem to ignore low temperature distinguishability of particles that unify at higher temperatures (increasing microstates during cooling). $\endgroup$ Nov 5, 2019 at 0:41

Judging by the various comments and answers I have seen posted so far, it seems to me that the main mistake made by the questioner is to assume that cold things always have lower entropy than hot things. This is wrong because the states counted in entropy measurements are states in position-momentum space (called phase space), not just momentum space. During an adiabatic expansion, for example, the entropy is constant but the temperature goes down.

Having said that, assessments of any property of "the whole universe" are always going to be metaphysically tricky. It is perhaps better to restrict our statements to identifiable parts of the universe, such as for example the part of it within the particle horizon. Assuming that the heat flow across the boundary of this region is small (and there is no reason for there to be much heat flow because the temperature is close to uniform), the entropy of that part has grown, for the same reason the entropy of any other isolated system grows under spontaneous processes.

The other interesting part of the picture is that large gravitating clouds are unstable against gravitational collapse, and in such a collapse some parts get hotter while other parts get colder. That can seem surprising when you first meet it, but in such processes the overall entropy is again increasing. After such collapse you get things like stars, and planets in orbit---and sometimes (well, at least once), life!

  • $\begingroup$ My definition of entropy is not based upon temperature, it is the logarithm of the number of micro states consistent with the macro state, which is the correct definition of entropy. Temperature informs the calculation of the number of micro states. That's not a mistake, it's how reality works. What you're arguing is that a BBQ has lower entropy than a house plant. I disagree. $\endgroup$ Nov 3, 2019 at 17:10
  • $\begingroup$ @Feynmanfan85 Do you agree that in an adiabatic expansion of an ideal gas the temperature falls? $\endgroup$ Nov 3, 2019 at 17:24
  • $\begingroup$ If you keep kinetic energy fixed, and expand volume, then yes, temperature should fall. But that's not what's happening here: the universe started out with no structure whatsoever, and was insanely hot. Then, it cooled, and structure formed. When a gas expands, structure is either constant, or decreasing. After the big bang, the universe became simpler in structure, and there's no way around this fact. Specifically, life exists, i.e., most structured matter we're aware of; solar systems exist. All this decreases the amount of information necessary to describe the universe, decreasing entropy. $\endgroup$ Nov 3, 2019 at 17:29
  • 13
    $\begingroup$ @Feynmanfan85 You contradict yourself: "the universe started out with no structure whatsoever" and "After the big bang, the universe became simpler in structure". What is simpler than none? You also insist on confusing Fermi temperature and thermodynamic temperature: ".. was insanely hot" - is a misconception. Things with a high Fermi temperature can be cold. In addition, things like Bose condensates don't even have a defined temperature. For example, if the Big Bang was a burst of coherent photons, they all could be in the same state with zero entropy. $\endgroup$
    – safesphere
    Nov 3, 2019 at 17:57

The statistical mechanics definition of entropy for a system at thermodynamic equilibrium is

"a function of the number of micro states consistent with a given macro state"

In order to apply it to the Universe (does not matter if in its early stages or later) one should reply positively to a few preliminary questions

  • is the universe a system in thermal equilibrium?
  • do we know how a macro-state of the universe is characterized?
  • is it conceivable to know what are the available micro-states of the universe at present time?

As far as I can see, answers to all these questions are negative.

As a consequence, I cannot assign a verifiable meaning to any statement about the entropy of the Universe. I just take this kind of sentences as a loosely speaking by analogy. Nothing more than that.

  • $\begingroup$ The cosmic microwave background is the most perfect black body spectrum ever observed. $\endgroup$
    – Thomas
    Nov 3, 2019 at 22:40
  • 1
    $\begingroup$ @Thomas I know. But this is just saying that at some time radiation got equilibrated. Can you say that before that time and nowadays the whole Universe is a system at equilibrium? $\endgroup$ Nov 3, 2019 at 22:48
  • 1
    $\begingroup$ To whom doesn't like my answer: please, feel free to downvote if you do not like it. However, you would make a great service to the community if you could justify a positive answer to at least one of my questions. I am always glad to learn something more. $\endgroup$ Nov 3, 2019 at 23:10
  • 1
    $\begingroup$ We know that at 300k yrs the universe was in almost perfect thermal equilibrium. We can take the known laws of physics and run them backwards to conclude that we had near perfect equilibrium up until the electroweak epoch. Of course, before that we don't know. $\endgroup$
    – Thomas
    Nov 4, 2019 at 16:08
  • 1
    $\begingroup$ After 300k yrs various pieces of the universe fell out of global equilibrium. The CMB is at a different temperature than the intergalactic gas, which is at a different temperature than stars, which are at different temperature then planets, etc. However, all these bits and pieces are lin local equilibrium, have a perfectly well defined entropy, which we can try to add up. $\endgroup$
    – Thomas
    Nov 4, 2019 at 16:10

I also don't understand why people sometimes say that the universe after the big bang was in a "low entropy" macrostate. (But, personally, I have never heard a physicist say this, so I don't think you should give the statement much weight.) According to the inflation model, the universe was all basically in thermal equilibrium right before inflation, which would suggest it was at maximum entropy.

The confounding factor here is indeed gravity. To my knowledge, there's no good way to ascribe entropy to a gravitation field that is not asymptotically flat, as is the case in an expanding universe. (How do you "count states," exactly?)

The reason that the universe cooled down since the big bang is that, due to general relativity, the universe has been expanding, separating all the hot stuff. Perhaps the entropy has not decreased, but the entropy density, entropy per unit volume, has decreased. This has allowed for life, etc. Obviously the universe is not in thermal equilibrium anymore.

(Also note that high entropy $\neq$ high temperature in general, these are different concepts.)

While I'm here, I'd like to address one misconception in your question:

On a smaller scale, the entropy of a fire must be higher than the entropy of a stable object, since the fire is obviously rapidly changing state, and therefore, presumably, is capable of being in a larger of number of states, while still being a fire. In contrast, a cup probably does change states somewhat, but obviously not as often, and obviously not on the same scale, in that the macro state of a cup is roughly constant.

The reason that a hot gas will have more entropy than a cold gas is that there are a larger range of velocities (actually, momenta) than the hot gas particles could be moving with: hence, more microstates. When there is more energy around, more states are "unlocked." In order to rigorously count microstates, which is a subject in an introductory statistical mechanics course, you must find the "phase space volume" of all the microstates, i.e. if you have $N$ particles, you must integrate $$ \int_{\rm Vol} d^3 x_1 \ldots d^3 x_N d^3 p_1 \ldots d^3 p_N $$ where $\rm Vol$ is the volume in phase space of your microstates, $x_i$ is the position of the $i^{\rm th}$ particle and $p_i$ is the momentum of the $i^{\rm th}$ particle. That will give you your "number" of states, up to a dimensionful factor. The logarithm of the number of microstates is the entropy.

  • 1
    $\begingroup$ And of course there is the fact that the universe seems to be made up of more matter and energy than what we see and know. Whether or not that matter and energy is in some entropic state now that is compareable to "then" is a moot point insofar as our equations of state don't add up to all the mass and all the energy... $\endgroup$
    – Stian
    Nov 5, 2019 at 11:06

Your Answer

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

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