Entropy

February 24, 2004 9:25 AM

AskMe physicists: Assuming there were any, where did the entropy from previous universes go? [more inside]

My understanding is that the universe at the time of the big bang was in a state of very low entropy, and that we've been using that inherent order to do work ever since the universe began. So, assuming that there was some previous universe, or string of universes, what happened to all then entropy that they generated when they expanded? When gravity collapses gases into stars, for instance, the outpouring of heat and light and whatnot compensates for the "lost entropy" of the highly ordered star. So where would all the entropy have gone in those collapsed universes?

posted by vraxoin to science & nature (13 comments total)

I've seen it suggested that the Second Law of Thermodynamics--entropy always increases in a closed system--applies only to an expanding universe, and in a contracting universe entropy might actually decrease.

One thing to remember about the Second Law is that, unlike many other physical laws, it is not actually an absolute law; it is only a probabilistic law. Instead of saying "entropy always increases" it would be more accurate to say "it is very, very, very, very, very unlikely that entropy would ever decrease." But it's not theoretically impossible.

As an analogy, imagine that you have 10 dice sitting on a table. Every so often, the table jerks up and down, causing the dice to bounce around and end up with possibly different numbers showing on top. If you started with the dice showing all 6's--a highly ordered state--it would be very likely that the dice would end up in a less ordered state after the table shakes. From a disordered state, it would be very unlikely--but not impossible--for the dice to end up in a highly ordered state.

Now imagine that instead of ten dice, there are billions of dice on the table. The likelihood of them all showing the same number at once is so small that for all practical purposes it's zero. But it's still theoretically possible.

To take this to a more realistic example, imagine you have a certain amount of gas trapped in a container with a plunger on one end. Suppose you want to get all of those gas molecules into a volume which is only one-half the volume of the current container. This involves a decrease in entropy of the gas (if the temperature of the gas stays the same, which we'll assume it does for the sake of this example). Well, you've got all these gas molecules flying around inside this container--you could just wait for them to all end up in one half of the container at once! A spontaneous decrease in entropy! Like billions of dice all showing the same number, this is not actually theoretically impossible, but is so very unlikely that it might as well be impossible. If you want to do this in a reasonable amount of time, you push the plunger in on the container, compressing the gas, which also involves a decrease in entropy of the gas, but is compensated for by an increase in entropy outside of the container, due to the effort you're putting into it. (In practice, and contrary to my earlier assumption, this would actually heat up the gas so that the entropy of the gas itself decreases somewhat less than if its temperature were constant, and only a small increase in entropy--various biomolecules in your muscles being converted to simpler molecules) outside of the container is needed to compensate.

Now, what if our container is the entire universe, and the "gas" is all the matter in the universe? Well, as long as the universe is expanding, entropy is increasing, and that's all well and good. What if it starts contracting? There's nothing outside the universe which could increase in entropy to compensate for the apparent decrease in entropy inside the universe. At least some physicists have said, well, that's not a problem since the second law isn't an absolute law anyway--entropy just decreases in a contracting universe, and what is highly unlikely in an expanding universe is actually to be expected in a contracting universe.
posted by DevilsAdvocate at 10:42 AM on February 24, 2004 [1 favorite]

In Jasper Fforde's charming "Lost in a Good Book," the sequel to "The Eyre Affair," the heroine believes someone is trying to kill her via unusual coincidences. (For instance, right before a sniper shoots her, she realizes the other seven women in her vehicle are all name Erma Cohen.) Her inventor uncle agrees that she is experiencing localized entropy drops and so gives her an entroposcope -- a jar or rice and lentils that she she can shake to see if entropy is decreasing. If it is, the rice and lentils should separate into improbable patterns.

Anyway, it's a very clever book.
posted by blueshammer at 11:09 AM on February 24, 2004

Isn't the Big Bang itself a big old increase in entropy?

But leaving that aside, since the laws of physics break down inside a singularity, I wouldn't expect any of them to operate in a smooth, trackable way through a universal contraction to a singularity and subsequent explosion.

To paraphrase DevilsAdvocte, the law is probablistic, meaning that it's extremely unlikely for entropy to decrease, or that entropy only decreases under extraordinary circumstances (like a contraction of the entire universe to a singularity).

/talking out of my ass
posted by scarabic at 11:10 AM on February 24, 2004

since there's no evidence at all for "previous" universes, how can we know anything about them? this included their entropy, size of underpants, whatever... previous universes are pure speculation with no observational support.

as we go back in time, to nearer and nearer the big bang, our knowledge of physics becomes less and less certain, because the physical conditions are so different to what we have today - so we don't have a clue what happened "before".
posted by andrew cooke at 11:19 AM on February 24, 2004

I am afraid this is too complicated for Ask.Me.

It was thought up until recently, even by Hawking, that a re-collapse would mean a decrease in entropy and a reversal of time. This seems to not be the case as was shown to him by Laflamme, one of his own students. The thermodynamic arrow of time continues in the same direction it seems even during a collapse.

All current work seems to indicate that, yes, entropy continues to increase and increase. Even current string theory models of the singularity (the so-called ekpyrotic models) don't really solve this problem of how a cyclic universe would discard this excess entropy.

I tend to favor basic inflation theory which essentially only takes hold when the sufficient entropic conditions exist, thus making a low-entropy state a contingent condition for the creation of a universe.

As andrew states, going beyond our universe to examine a multi-verse scenario is a little beyond us at this point and yet most of modern cosmology was metaphysics at some point. One proposal, similiar to what DevilsAdvocate says, is that since there is no definite stable state of maximum entropy, our entire universe may just be a random fluctutaion of low-entropy in a much larger metaverse of extremely high entropy. This is basically then the anthropic principle - we are here to witness this extremely unusual universe that started in a low-entropy state simply because only in this type of universe can life exist to make that observation.

In any case, for anyone that is interested in this I highly recommend Andreas Albrechts overview. It is highly readable, math-free and even reviews the basic concepts of entropy with some nice diagrams. He concludes:

So one important conclusion is that inflation, or any other attempt to dynamically
explain the initial conditions of the observed universe, will necessarily require some
special initial conditions itself, in order to have an arrow of time. These special initial
conditions are the vestiges of Boltzmann’s original “rare fluctuations” which can never
be completely excised from this sort of dynamical approach.

This conclusion is particularly directed at those who hold up the special initial
conditions of inflation as a serious flaw of the idea. However, I know of no fundamental
law that prevents one from hoping that some improved dynamical process could produce
the universe we observe using initial fluctuations that are even less rare than those which
initiate inflation, so perhaps it is just as well that the critics keep the pressure on.
But the discussion in this article also relates to another debate about initial conditions.

There are those who find the dynamical approached inherently flawed. Instead, they
wish to uncover broad principles or fundamental laws that will uniquely specify the state
of the universe (see for example Hollands and Wald, 2002). As I discussed in section 7,
several current ideas (such as the ekpyrotic model and eternal inflation) fall under this
category. The field seems to be divided among people who strongly favor a dynamical
approach, and those who strongly favor defining a unique state of the universe based on
principles.
posted by vacapinta at 12:47 PM on February 24, 2004

I'll talk out of my ass too: Is it possible to have a sort of "anti-entropy"? It seems that quantumists (I made that word up) are constantly discovering anti-thises and anti-thats. Could one universe's increase in entropy be another's decrease?
posted by jpoulos at 12:59 PM on February 24, 2004

Wouldn't a black hole basically be a giant anti-entropy machine? Sucking in matter and energy into a concentrated point...
posted by LionIndex at 2:15 PM on February 24, 2004

our entire universe may just be a random fluctutaion of low-entropy in a much larger metaverse of extremely high entropy.

Low entropy? You haven't seen my desk.
posted by adamrice at 2:16 PM on February 24, 2004

Black holes have entropy.
posted by vacapinta at 3:29 PM on February 24, 2004

I'll talk out of my ass too: Is it possible to have a sort of "anti-entropy"?

Probably not. Entropy isn't a "thing"--it's a measure of a system. To talk about anti-entropy seems about as meaningful as talking about anti-length.

That said, I don't want to say absolutely not, since physicists are always coming up with seemingly counterintuitive things.
posted by DevilsAdvocate at 5:16 PM on February 24, 2004

It seems that quantumists (I made that word up) are constantly discovering anti-thises and anti-thats.

I just read last night that quasars are speculated to be the opposite of black holes. Once matter/energy passes through the one-way membrane of a black hole's event horizon, it can never escape and becomes inaccessible to this universe. In a sense, you could say that it enters another universe.

An average quasar is supposed to be about the diamater of our solar system, and put out the equivalent energy of 150 billion stars. It's speculated that quasars bring matter into this universe from others, and black holes take it out of this one and deposit it elsewhere (to use crude terms).

Sort of makes you stop and decide what your definition of "universe" is.
posted by scarabic at 5:30 PM on February 24, 2004

Oh, given that vraoxin asked for physicists I should make it clear that my only qualifications are: an undergraduate degree in physics, physicist friends, and keeping up with major journal articles.
posted by vacapinta at 11:42 PM on February 24, 2004

Heck, mine are even less than vacapinta's--an undergraduate degree in chemistry. Chemists need to know a thing or two about entropy, but it's more entropy of chemicals in a flask or gases in a container. Entropy of the entire universe, not so much.

And Physical Chemistry--Thermodynamics was the hardest course I ever took. Yes, even harder than Physical Chemistry--Quantum Mechanics. That might partly be due to the fact that I had a very good professor for QM, though, and a rather poor one for thermo.
posted by DevilsAdvocate at 8:04 AM on February 25, 2004