The 2nd law of thermodynamics is not evidence of God

[Magical Wand]

Apologist William Lane Craig argues the 2nd law of thermodynamics is evidence the universe had an absolute beginning. And whatever begins to exist has a cause. The universe began to exist. Therefore, the universe has a cause. This cause must be non-spatio-temporal and non-material since all that is material had a beginning. Ergo, this cause is how God is described, or as Aquinas liked to say, "This is what people call God."

So, let me explain how it is supposed to be evidence of an absolute beginning.

This is because of the tendency of particles to approach equilibrium in closed systems. As a consequence the universe would evolve to a state of no thermodynamic free energy and would therefore be unable to sustain processes that increase entropy and may no longer be exploited to perform work (since work is obtained from ordered molecular motion). As a result, an infinitely-old universe should therefore have reached equilibrium (i.e., maximum entropy) long ago, but it did not, so the universe began at some finite time in the past.

In the book "Atheism: The Case Against God", George H. Smith replied as follows:

"Reconciling the Second Law of Thermodynamics with the present state of the universe is not as hopeless as theists like to pretend. To begin with, the Second Law is a statement of statistical probability, and there is nothing inherently contradictory in supposing that a closed system can decrease in entropy or fluctuate between increasing and decreasing entropy states. But this probability, while metaphysically possible, is extremely unlikely, so it is usually ignored in practical applications.

More importantly, however, the Second Law pertains only to closed systems, which, according to many physicists, renders it inapplicable to the universe as a whole. Professor Grunbaum, a physicist, writes:

'An inherent limitation on the applicability of the ... entropy concept to the entire universe lies in the fact that it has no applicability at all to a spatially infinite universe.'

Professor E. A. Milne, commenting on another physicist’s acceptance of the heat-death thesis, writes: 'Jeans’s own studies in the realm of the second law of thermodynamics were all concerned with the kinetic theory of gases, in which the specimen under discussion is supposed walled around in a finite vessel; and to such systems the notion of a heat-death is applicable. But by no means is the same result to be predicted of the whole universe.'

Finally, according to Landau and Evgeny, authors of Statistical Physics: '... in the general theory of relativity the universe as a whole must be regarded not as a closed system, but as one which is in a variable gravitational field. In this case the application of the law of increase of entropy does not imply the necessity of statistical equilibrium.'

Since the concept of entropy can be defined only with reference to closed systems, it cannot legitimately be applied to the universe as a whole. The theist takes a scientific principle derived from a specific context, and attempts to shift this context in order to manufacture a need for god. In the name of science, the theist posits a “god of the gaps,” a god who allegedly fills in the gaps of human knowledge. But gaps of knowledge eventually close, leaving god without a home. The entropy argument is a cosmological argument draped in scientific jargon—but an invalid argument, even when presented in scientific terms, is still invalid."

Smith's rebuttal seems robust to me (although I don't like the way he speaks of theists, i.e., "theists like to pretend". Just because some theists -- viz., apologists -- do it doesn't mean all or most theists do it).

Anyway, I'm eager to hear your thoughts on this.

[Edit: Many viewers commented the Big Bang theory also proves -- or indicates -- the past is finite. However, (1) this isn't correct as I explained in the comments below and (2) this isn't relevant to this post since the subject isn't whether the universe had a beginning; rather, it is whether the 2nd law is evidence of such a beginning. The difference is subtle, but relevant.]

 

[Polymath257]

One aspect of entropy increase being a statistical thing rather than a necessary thing is that there is a time period, called the Poincare recurrence time, in which it is *guaranteed* that entropy will cycle for a closed system.

Now, the Poincare recurrence time, even for a room full of gas, is incredibly long. But it is NOT infinitely long. Which means that any use of the second law to show the universe *must* be finitely old is doomed to failure.

There are also issues surrounding the fact that a treatment of gases in closed vessels on Earth does not deal with self-gravitation of gases in space. And the gravitational aspect is crucial for things like formation of stars (which decreases entropy for a core of gas while increasing it for regions further out) to universal expansion (which is close to being adiabatic: no increase or decrease of entropy).

 

[ChristineM]

And whatever begins to exist has a cause. The universe began to exist. Therefore, the universe has a cause. This cause must be non-spatio-temporal and non-material

Not necessarily so. The law of causation, like the second law of thermodynamics, did not begin to resolve until some short time after the BB so it is possible there was no cause.

As a result, an infinitely-old universe should therefore have reached equilibrium (i.e., maximum entropy) long ago,

It will take considerably longer than the universe has already existed to reach heat death/maximum entropy, around 15 trillion years.

so the universe began at some finite time in the past.

Around 13.8 billion years ago

And welcome to RF. Enjoy the cake

 

[exchemist]

Apologist William Lane Craig argues the 2nd law of thermodynamics is evidence the universe had an absolute beginning. And whatever begins to exist has a cause. The universe began to exist. Therefore, the universe has a cause. This cause must be non-spatio-temporal and non-material since all that is material had a beginning. Ergo, this cause is how God is described, or as Aquinas liked to say, "This is what people call God."

So, let me explain how it is supposed to be evidence of an absolute beginning.

This is because of the tendency of particles to approach equilibrium in closed systems. As a consequence the universe would evolve to a state of no thermodynamic free energy and would therefore be unable to sustain processes that increase entropy and may no longer be exploited to perform work (since work is obtained from ordered molecular motion). As a result, an infinitely-old universe should therefore have reached equilibrium (i.e., maximum entropy) long ago, but it did not, so the universe began at some finite time in the past.

View attachment 52492

In the book "Atheism: The Case Against God", George H. Smith replied as follows:

"Reconciling the Second Law of Thermodynamics with the present state of the universe is not as hopeless as theists like to pretend. To begin with, the Second Law is a statement of statistical probability, and there is nothing inherently contradictory in supposing that a closed system can decrease in entropy or fluctuate between increasing and decreasing entropy states. But this probability, while metaphysically possible, is extremely unlikely, so it is usually ignored in practical applications.

More importantly, however, the Second Law pertains only to closed systems, which, according to many physicists, renders it inapplicable to the universe as a whole. Professor Grunbaum, a physicist, writes:

'An inherent limitation on the applicability of the ... entropy concept to the entire universe lies in the fact that it has no applicability at all to a spatially infinite universe.'

Professor E. A. Milne, commenting on another physicist’s acceptance of the heat-death thesis, writes: 'Jeans’s own studies in the realm of the second law of thermodynamics were all concerned with the kinetic theory of gases, in which the specimen under discussion is supposed walled around in a finite vessel; and to such systems the notion of a heat-death is applicable. But by no means is the same result to be predicted of the whole universe.'

Finally, according to Landau and Lif****z, authors of Statistical Physics: '... in the general theory of relativity the universe as a whole must be regarded not as a closed system, but as one which is in a variable gravitational field. In this case the application of the law of increase of entropy does not imply the necessity of statistical equilibrium.'

Since the concept of entropy can be defined only with reference to closed systems, it cannot legitimately be applied to the universe as a whole. The theist takes a scientific principle derived from a specific context, and attempts to shift this context in order to manufacture a need for god. In the name of science, the theist posits a “god of the gaps,” a god who allegedly fills in the gaps of human knowledge. But gaps of knowledge eventually close, leaving god without a home. The entropy argument is a cosmological argument draped in scientific jargon—but an invalid argument, even when presented in scientific terms, is still invalid."

Smith's rebuttal seem robust to me (although I don't like the way he speaks of theists, i.e., "theists like to pretend". Just because some theists (i.e., apologists) do it doesn't mean all or most theists do it.

Anyway, I'm eager to hear your thoughts on this.

I'm a bit baffled by this. Surely the Big Bang hypothesis also points to the universe having a finite age, doesn't it? So who is disagreeing with this, and why does it allegedly tell us anything about a hypothetical creator?

Or are we just back to the good old First Cause cosmological argument for God, again?

 

[Polymath257]

I'm a bit baffled by this. Surely the Big Bang hypothesis also points to the universe having a finite age, doesn't it? So who is disagreeing with this, and why does it allegedly tell us anything about a hypothetical creator?

Or are we just back to the good old First Cause cosmological argument for God, again?

There are versions of quantum gravity in which time can be extended to previous to the BB. In those, it is often the case that time (and the resulting multiverse) is infinitely old. Craig seems to be arguing against such extensions.

 

[epronovost]

I personally can't engage much on the debate of thermodynamic laws, but I can tell the following thing.

1) Everything that exist has a cause is false or at least an overgeneralization.

We don't know the cause of many phenomenon and we can't say that they have one (or several since some phenomenon have no singular cause) until we actually find them. An example of that would be radioactive decay for example seems to be an uncaused phenomenon.

2) causality is a feature of the universe just like spacetime.

Since causality is a feature of the universe just like spacetime how could the universe be caused. That would imply that something can cause causality and that something could be prior to time which is of course contradictory.

3) this is what people call God

That's also not the case. The beginning of the universe or its cause, even if it could have one, isn't what people call God. What people call God is very wide, but its always something that's worship and in the overwhelming majority of cases God is also a personable creature with interest and interactions toward humans and the world they live in. Saying "that's what people call God" is basically talking through people and a bare assertion.

 

[exchemist]

There are versions of quantum gravity in which time can be extended to previous to the BB. In those, it is often the case that time (and the resulting multiverse) is infinitely old. Craig seems to be arguing against such extensions.

But as there is no accepted theory of quantum gravity, that merely means he is arguing against one speculative hypothesis which is not generally adopted, then, doesn't it?

 

[Polymath257]

But as there is no accepted theory of quantum gravity, that merely means he is arguing against one speculative hypothesis which is not generally adopted, then, doesn't it?

yes, although he is arguing against *any* extension that would include infinite time.

 

[Revoltingest]

Apologist William Lane Craig argues the 2nd law of thermodynamics is evidence the universe had an absolute beginning. And whatever begins to exist has a cause. The universe began to exist. Therefore, the universe has a cause. This cause must be non-spatio-temporal and non-material since all that is material had a beginning. Ergo, this cause is how God is described, or as Aquinas liked to say, "This is what people call God."

So, let me explain how it is supposed to be evidence of an absolute beginning.

This is because of the tendency of particles to approach equilibrium in closed systems. As a consequence the universe would evolve to a state of no thermodynamic free energy and would therefore be unable to sustain processes that increase entropy and may no longer be exploited to perform work (since work is obtained from ordered molecular motion). As a result, an infinitely-old universe should therefore have reached equilibrium (i.e., maximum entropy) long ago, but it did not, so the universe began at some finite time in the past.

View attachment 52492

In the book "Atheism: The Case Against God", George H. Smith replied as follows:

"Reconciling the Second Law of Thermodynamics with the present state of the universe is not as hopeless as theists like to pretend. To begin with, the Second Law is a statement of statistical probability, and there is nothing inherently contradictory in supposing that a closed system can decrease in entropy or fluctuate between increasing and decreasing entropy states. But this probability, while metaphysically possible, is extremely unlikely, so it is usually ignored in practical applications.

More importantly, however, the Second Law pertains only to closed systems, which, according to many physicists, renders it inapplicable to the universe as a whole. Professor Grunbaum, a physicist, writes:

'An inherent limitation on the applicability of the ... entropy concept to the entire universe lies in the fact that it has no applicability at all to a spatially infinite universe.'

Professor E. A. Milne, commenting on another physicist’s acceptance of the heat-death thesis, writes: 'Jeans’s own studies in the realm of the second law of thermodynamics were all concerned with the kinetic theory of gases, in which the specimen under discussion is supposed walled around in a finite vessel; and to such systems the notion of a heat-death is applicable. But by no means is the same result to be predicted of the whole universe.'

Finally, according to Landau and Lif****z, authors of Statistical Physics: '... in the general theory of relativity the universe as a whole must be regarded not as a closed system, but as one which is in a variable gravitational field. In this case the application of the law of increase of entropy does not imply the necessity of statistical equilibrium.'

Since the concept of entropy can be defined only with reference to closed systems, it cannot legitimately be applied to the universe as a whole. The theist takes a scientific principle derived from a specific context, and attempts to shift this context in order to manufacture a need for god. In the name of science, the theist posits a “god of the gaps,” a god who allegedly fills in the gaps of human knowledge. But gaps of knowledge eventually close, leaving god without a home. The entropy argument is a cosmological argument draped in scientific jargon—but an invalid argument, even when presented in scientific terms, is still invalid."

Smith's rebuttal seem robust to me (although I don't like the way he speaks of theists, i.e., "theists like to pretend". Just because some theists (i.e., apologists) do it doesn't mean all or most theists do it.

Anyway, I'm eager to hear your thoughts on this.

The simple view (one that even engineers can understand)...
We cannot assume the universe is a closed system.
So entropy might even decrease.

 

[SigurdReginson]

I don't have much to say on the subject, but welcome to RF.

 

[Magical Wand]

One aspect of entropy increase being a statistical thing rather than a necessary thing is that there is a time period, called the Poincare recurrence time, in which it is *guaranteed* that entropy will cycle for a closed system.

Now, the Poincare recurrence time, even for a room full of gas, is incredibly long. But it is NOT infinitely long. Which means that any use of the second law to show the universe *must* be finitely old is doomed to failure.

Exactly! The Poincaré recurrence theorem proves that entropy can decrease (or fluctuate) back to its initial state as long as sufficient time is included in the equation. And what can be more sufficient than infinite time?

Nevertheless, there are objections to this view. Every theorem makes certain assumptions and the universe must be in accordance with these assumptions. Some argue the universe is not in accordance with the Poincaré theorem. They claim that (1) the theorem is not valid for nonlinear dynamic systems or for any system that is capable of exhibiting deterministic chaos. Chaos is a common occurrence in our universe. (2) A physicist named P. D. Gujrati published a paper ( https://arxiv.org/pdf/0803.0983.pdf ) arguing Poincaré recurrences do not imply entropy will decrease even if these conditions are met.

I sent an email to cosmologist Anthony Aguirre (an expert on this subject) asking his opinion and he responded the following regarding (1): "The theorem applies pretty generally to any system with a fixed state-space and reversible evolution operator, so I don’t think chaos is a factor.” Moreover, in the paper “The Poincaré recurrence time for the de Sitter space with dynamical chaos”, physicist K. Ropotenko showed such reversions can occur without problems in these kind of states.

Regarding the paper by Gujrati, Dr. Aguirre replied: "The author uses Gibbs entropy, which is time-independent in a closed system undergoing Hamiltonian (or Liouville) evolution. So, indeed entropy does not go down, but it does not go up either. If you use Boltzmann entropy, it will generally go up, but also down again since it is just tied to the microstate. My favorite way of thinking about entropy is that developed by myself and collaborators. The most readable (but still fairly technical) version of it is here, if you’re interested. In this account, coarse-grained (or “observational”) entropy would, I think, fluctuate all the way back down to near its initial value."

Another possible objection to this view is that an entropy reversion is unlikely to produce a whole cosmology filled with low entropy galaxies. It would be much more likely for the recurrence to produce the minimum amount of matter necessary to support life (like a single solar system). But this is not the case (i.e., we see a huge universe), therefore, it is more probable the universe is not the product of a spontaneous reversion of entropy.

The problem with this argument is that it applies the rules of finite statistics (which are only logically valid for finite quantities) to transfinite quantities. If time is infinite, then no matter how unlikely it is for an event to happen, it will necessarily happen; this follows from statistical mechanics. Therefore, there will exist an infinite number of large universes like ours (with multiple galaxies) and an infinite number of small universes (with a single solar system).

The objector is incorrectly arguing that an infinite number of events is less likely than another infinite number of events, or that an infinity is less likely than another. This is clearly absurd. The probability of our large universe existing is infinitely big, so we don't have to worry about our existence being improbable.

This subject is very interesting! I love to talk about it.

 

[Magical Wand]

And welcome to RF. Enjoy the cake

View attachment 52494

Thank you very much for your hospitality! And I'll enjoy the cake, yes. haha

 

[Magical Wand]

I'm a bit baffled by this. Surely the Big Bang hypothesis also points to the universe having a finite age, doesn't it? So who is disagreeing with this, and why does it allegedly tell us anything about a hypothetical creator?

Or are we just back to the good old First Cause cosmological argument for God, again?

Well, Big Bang cosmology opens up a whole new can of worms. It is a very complex subject. But I can quote more than 20 recent books and articles by prominent cosmologists where it was argued the Big Bang doesn't provide evidence of an absolute beginning. As Polymath257 pointed out, classical theory (e.g., general relativity) doesn't apply at the Big Bang since extremely high energies are outside of its domain. However, an absolute beginning is only obtained if you apply classical theory. Thus, this result has no value to physicists.

But let me just say that some cosmologists disagree with that as well. I've talked to a cosmologist who wrote a very long and extensive paper arguing that even General Relativity can get rid of the initial curvature singularity by replacing it with a coordinate singularity (a coordinate singularity is not a real hole in the space-time manifold; it merely represents a bad choice of coordinates). And he told me this result shows that space-time can be extended prior to the expansion of the universe.

Much much much more can be said about this, but I'll stop here for the sake of space and time.

 

[Magical Wand]

I don't have much to say on the subject, but welcome to RF.

Thanks!

 

[Brickjectivity]

Apologist William Lane Craig argues the 2nd law of thermodynamics is evidence the universe had an absolute beginning. And whatever begins to exist has a cause. The universe began to exist. Therefore, the universe has a cause. This cause must be non-spatio-temporal and non-material since all that is material had a beginning. Ergo, this cause is how God is described, or as Aquinas liked to say, "This is what people call God."

Good point in the rebuttal.

A very short answer could be: 2nd Law is evidence that the universe will continue to move towards thermal equilibrium but not proof that more mass and more energy is not being added. If more energy is being added, then this supports a model with no beginning.

 

[Magical Wand]

I personally can't engage much on the debate of thermodynamic laws, but I can tell the following thing.

1) Everything that exist has a cause is false or at least an overgeneralization.

We don't know the cause of many phenomenon and we can't say that they have one (or several since some phenomenon have no singular cause) until we actually find them. An example of that would be radioactive decay for example seems to be an uncaused phenomenon.

2) causality is a feature of the universe just like spacetime.

Since causality is a feature of the universe just like spacetime how could the universe be caused. That would imply that something can cause causality and that something could be prior to time which is of course contradictory.

3) this is what people call God

That's also not the case. The beginning of the universe or its cause, even if it could have one, isn't what people call God. What people call God is very wide, but its always something that's worship and in the overwhelming majority of cases God is also a personable creature with interest and interactions toward humans and the world they live in. Saying "that's what people call God" is basically talking through people and a bare assertion.

Well, I agree with numbers (1) and (2), but (3) is incomplete.

The Christian God is generally defined (by theologians) as a being who exists outside of space. And that definition would fit the results obtained by this cosmological argument. Non-material is also standard. Now, "non-temporal" is more controversial. There are many books out there in which it was argued God is temporal. For example, Ryan Mullins' recent book "The end of the timeless God." But it is still fairly standard.

Finally, being a "personable creature with interest and interactions toward humans" also defines God, but the Kalam argument only tries to get some features of God. It doesn't have to give all features in order to provide empirical evidence of his existence. I know you didn't explicitly say that, but that's what it seems to me.

Correct me if I didn't accurately represent your comment, though.

 

[Polymath257]

Well, Big Bang cosmology opens up a whole new can of worms. It is a very complex subject. But I can quote more than 20 recent books and articles by prominent cosmologists where it was argued the Big Bang doesn't provide evidence of an absolute beginning. As Polymath257 pointed out, classical theory (e.g., general relativity) cannot be applied in the Big Bang since extremely high energies are outside of its domain. However, an absolute beginning is only obtained if you apply classical theory. Thus, this result has no value to physicists.

But let me just say that some cosmologists disagree with that as well. I've talked to a cosmologist who wrote a very long and extensive paper arguing that even General Relativity can get rid of the initial curvature singularity by replacing it with a coordinate singularity (a coordinate singularity is not a real hole in the space-time manifold; it merely represents a bad choice of coordinates). And he told me this result shows that space-time can be extended prior to the expansion of the universe.

Much much much more can be said about this, but I'll stop here for the sake of space and time.

I think the Hawking-Ellis theorem shows there has to be an actual curvature singularity and not just a coordinate singularity *if* GR is the correct description. So I am a bit skeptical of this claim.

As you said, though, GR is a classical theory and will certainly need to be modified to take into account quantum effects at high energies. Whether that allows for time to be extended depends, as far as I know, on the specifics of quantum gravity.

 

[Polymath257]

The objector is incorrectly arguing that an infinite number of events is less likely than another infinite number of events, or that an infinity is less likely than another. This is clearly absurd. The probability of our large universe existing is infinitely big, so we don't have to worry about our existence being improbable.

It *can* make sense to say one collection of infinitely many events is more likely than another collection of infinitely many events. The problem is that there does not appear to be a natural probability measure to use to make this claim.

 

[AlexanderG]

Scientific laws only apply in very specific circumstances, which others have already pointed out above with the discussion of closed systems. So, arguing from the Second Law has a few problems.

1. Scientific laws are descriptive, not prescriptive. We could discover an exception tomorrow and have to revise the law. Observed reality defines scientific laws; scientific laws do not dictate ultimate reality. Currently, we can't seem to observe ultimate reality.
2. We can't confirm that the circumstances in which the Second Law applies is a match with the circumstances of our universe, let alone with the circumstances prior to the Planck time.
3. The Second Law applies specifically to closed systems within our universe. We don't know if it applies to natural objects or forces outside our universe, like quantum fields, multiverses, etc.
4. If you wish to argue from the Second Law of Thermodynamics, it can be useful to also examine the First Law, which states that energy cannot be created or destroyed. This arguably implies that energy is eternal and thus there is no need for a god. Alternately, this First Law also implies that god had no mechanism to create the energy in our universe.

 

[Magical Wand]

I think the Hawking-Ellis theorem shows there has to be an actual curvature singularity and not just a coordinate singularity *if* GR is the correct description. So I am a bit skeptical of this claim.

Yes, so this cosmologist I mentioned argued against this result. His name is Ovidiu-Cristinel Stoica.

I sent an email to him asking the following:

"In your very interesting article "The Friedmann-Lemaitre-Robertson-Walker Big Bang Singularities are Well Behaved", you've proven a theorem showing that GR can be written in a non-singular form. But my question to you is about the genericness of your results. Do your results apply to different singularities? Black holes, naked singularities, weak singularities and etc? Because some authors argue that any resolution to the singularity problem, must be generic.

A following question would be: what about the extension before the Big Bang? Would you agree it is possible that it could have been contracting from the infinite past ... or are you proposing a cyclic scenario?"

He replied:

"I agree that the solution should be general. In a series of papers, and then in my PhD Thesis, I did the following:

- find a way to extend semi-Riemannian geometry to work with a class of singular metrics
- apply it to FLRW and to the three main types of black holes.
Here is my Thesis, which describes this approach, and also includes references to the articles I wrote about this up to that point:
Singular General Relativity

There are more ways in which a metric can be singular. The simplest one is if the determinant of g_{ab} becomes zero. Even so, the covariant derivative and the curvature requires g^{ab}, which is singular when the determinant of g_{ab}->0. So it was nontrivial even for the simplest case. I call the first case "bening singularity". The more complicated case is when g_{ab} has components that go to infinity, I call it "malign singularity". This is the case of black holes singularities. In this case, my theory doesn't work directly. But I found out that a coordinate transformation can sometimes make the metric that is a malign singularity to become a benign singularity. Remember how the Eddingtion-Finkelstein coordinate transformation shows that the r=r_m singularity, which is the event horizon, is not actually a singularity, but the Schwarzschild coordinates themselves are the singular ones, and not the metric. This works for r=r_m, but not for r=0, because for r=0 there are scalar invariants that are singular and remain singular even if you make the transformation (K=R^{abcd}R_{abcd}). But I found out that I can make coordinate transformations that remove the part of the singularity that is due to the coordinates, and makes the metric smooth at r=0 too. The metric is still singular, but the singularity is benign.

So my approach also works for black hole singularities. It doesn't work for all conceivable singularities, but it works for the known BB and BH singularities, and for some general classes. I don't know if in the real world there are singularities for which my approach wouldn't work. I conjecture that there are only benign and malign that can be turned into benign, but I didn't prove this.

About extension before the BB, yes, it is possible, I guess page 7 of Beyond the Friedmann-Lemaitre-Robertson-Walker Big Bang singularity contains just what you asked. Note that although in this paper the generalized Einstein equation is different from the one in the first one, the singular geometry is the same, and the solutions in the paper you found extend as well beyond the singularity.

If you are interested, I think my Thesis is the easiest way, because I describe more systematically than in the papers what I did."


I think this is certainly interesting, but I'm also a bit skeptical of his results.