I Before the Big Bang

[Jaime Rudas]

TL;DR Summary

Did time and space exist before the Big Bang?

The article Where did the Big Bang happen? states:

According to standard cosmological models, which are based on general relativity and are found to agree well with observations, time and space did not exist before the Big Bang

Is this really the case?
Wouldn't it be more correct to say that the current consensus cosmological model is unable to describe anything that happened before, for example, the Planck time?

 

[DaveC426913]

TL;DR Summary: Did time and space exist before the Big Bang?

The article Where did the Big Bang happen? states:

Is this really the case?
Wouldn't it be more correct to say that the current consensus cosmological model is unable to describe anything that happened before, for example, the Planck time?

I believe the idea is that space-time was created by the Big Bang.

No BB = no space and no time.

 

[phyzguy]

One possibility, known as the Hartle-Hawking no boundary proposal, is that if you go backward in time towards the big bang, you reach a point where space and time cease to exist. In this proposal, asking, "What was there before the big bang?" is similar to asking "What is north of the north pole?"

Here's a Wikipedia link: https://en.wikipedia.org/wiki/Hartle–Hawking_state

 

[phinds]

TL;DR Summary: Did time and space exist before the Big Bang?

Wouldn't it be more correct to say that the current consensus cosmological model is unable to describe anything that happened before, for example, the Planck time?

I agree that that's the case but I'm not an expert.

 

[Jaime Rudas]

I believe the idea is that space-time was created by the Big Bang.

No BB = no space and no time.

One possibility, known as the Hartle-Hawking no boundary proposal, is that if you go backward in time towards the big bang, you reach a point where space and time cease to exist. In this proposal, asking, "What was there before the big bang?" is similar to asking "What is north of the north pole?"

Here's a Wikipedia link: https://en.wikipedia.org/wiki/Hartle–Hawking_state

It seems to me that, according to these two hypotheses, time was created in the Big Bang. Which makes me realize that saying that time did not exist before the Big Bang is, in fact, an oxymoron because if there was no time, there could not have been a before.

 

[Jaime Rudas]

"What is north of the north pole?"

Polaris?

 

[DaveC426913]

saying that time did not exist before the Big Bang is, in fact, an oxymoron because if there was no time, there could not have been a before.

I disagree. It certainly wouldn't have been anything like we understand it, but that doesn't mean it couldn't exist.

 

[PeterDonis]

if there was no time, there could not have been a before.

In Hawking's "no boundary" model, the portion of the model "before" the initial time (which is shaped, heuristically, like a hemisphere, so there's no singularity anywhere, and it joins to the initial time spacelike hypersurface smoothly) is purely spacelike; that's why there's no "time" there. But it is "before" the initial time in the sense that it's not at that time or any later time. "Before" might not be the best ordinary language word to capture the math, but the math itself is consistent, although counterintuitive.

 

[phyzguy]

Polaris?

Polaris is "up" from the north pole. North is a direction along the surface of the Earth.

 

[javisot]

TL;DR Summary: Did time and space exist before the Big Bang?

The article Where did the Big Bang happen? states:

Is this really the case?
Wouldn't it be more correct to say that the current consensus cosmological model is unable to describe anything that happened before, for example, the Planck time?

Space and time, spacetime, is GR. If there was a point in the past where we failed to construct a reasonable notion of space and time, it means that GR fails at some point. Perhaps a theory of quantum gravity would resolve this.

 

[PeterDonis]

If there was a point in the past where we failed to construct a reasonable notion of space and time, it means that GR fails at some point.

In the Hawking no boundary proposal, as I noted before, the manifold describing the universe "before the Big Bang" is purely spacelike. This could be interpreted as GR "failing" since GR is based on the manifold having a Lorentzian signature.

Perhaps a theory of quantum gravity would resolve this.

As I understand it, the Hawking "no boundary" proposal is essentially a quantum gravity proposal.

 

[javisot]

In the Hawking no boundary proposal, as I noted before, the manifold describing the universe "before the Big Bang" is purely spacelike. This could be interpreted as GR "failing" since GR is based on the manifold having a Lorentzian signature.


As I understand it, the Hawking "no boundary" proposal is essentially a quantum gravity proposal.

One question Peter, the term "big bang" predates the inflation proposal. Yet, we call the beginning of reheating the "big bang", a reheating that begins when inflation ends.

Would you say the article's question "where the big bang happens?" refers to the start of reheating or the start of inflation?
(I'm not sure..)

 

[PeterDonis]

One question Peter, the term "big bang" predates the inflation proposal. Yet, we call the beginning of reheating the "big bang", a reheating that begins when inflation ends.

The terminology in the literature is unfortunately not consistent on this point. To me, it makes the most sense to use the term "Big Bang" to refer to the hot, dense, rapidly expanding state that, according to our best current models, happened at the end of inflation, when reheating took place. (In papers that are the most careful about using proper technical terminology, that is how the term is used.)

Would you say the article's question "where the big bang happens?" refers to the start of reheating or the start of inflation?

Neither. Since the article explicitly says that "time and space did not exist" before the big bang, it is clearly using the term to mean the initial singularity. In models which have both inflation and an initial singularity, the initial singularity is before even the start of inflation. But there are also "eternal inflation" models in which there is no initial singularity--inflation extends infinitely far back into the past. And there are models like Hawking's no boundary proposal in which, while there is a start of inflation, there is no initial singularity.

The article says it refers to "standard cosmological models", but I don't know that there is a "standard" model for what happened prior to reheating. So for the article to assume that "standard cosmological models" include an initial singularity strikes me as a pop science oversimplification, not a reliable description of the current state of theory in this area.

 

[Jaime Rudas]

But there are also "eternal inflation" models in which there is no initial singularity--inflation extends infinitely far back into the past.

Remember that, under reasonable assumptions, the Borde-Guth-Vilenkin theorem demonstrates that inflationary cosmological models doesn't extend infinitely into the past.

 

[PeterDonis]

under reasonable assumptions, the Borde-Guth-Vilenkin theorem demonstrates that inflationary cosmological models doesn't extend infinitely into the past.

Yes, that's true. But one of the assumptions is that there has to be an everywhere expanding congruence in the spacetime. de Sitter spacetime extends infinitely into the past by violating that assumption--basically it's one version of a "bounce" model where the expanding phase is preceded by a contracting phase. There are also other more complicated "bounce" models in the literature. So the BGV theorem can't be asserted as entirely ruling out models that extend infinitely into the past.

It does rule out at least most "eternal inflation" models, which I believe was the intended point of those who proved the theorem. However, I think there are still other "eternal inflation" models (though highly speculative ones) that it doesn't rule out. I think we had a thread on this some time ago.

 

[Jaime Rudas]

I think we had a thread on this some time ago.

Yes, that's why I said "remember."

 

[PeterDonis]

Yes, that's why I said "remember."

Ah, thanks for the link. Note that in the further discussion in that thread, a model due to Linde was mentioned which he still describes as "eternal inflation" even though it satisfies the conditions of the BGV theorem.

 

[javisot]

Yes, that's true. But one of the assumptions is that there has to be an everywhere expanding congruence in the spacetime. de Sitter spacetime extends infinitely into the past by violating that assumption--basically it's one version of a "bounce" model where the expanding phase is preceded by a contracting phase. There are also other more complicated "bounce" models in the literature. So the BGV theorem can't be asserted as entirely ruling out models that extend infinitely into the past.

It does rule out at least most "eternal inflation" models, which I believe was the intended point of those who proved the theorem. However, I think there are still other "eternal inflation" models (though highly speculative ones) that it doesn't rule out. I think we had a thread on this some time ago.

If I understand you correctly, Peter, the theorem you mentioned rules out inflationary models that extend infinitely into the past and also contain singularities, since they would imply that the universe had infinite temperature and density at some point in the past. (Is that correct?)

Perhaps an analogy would be: we have an infinitely hot cup of coffee. If we try to cool it by finite, even infinite amounts, it will remain infinitely hot. There is no reasonable reason for the coffee cup to change state if it was infinitely hot at some point.

Inflationary models that extend infinitely into the past are only valid if they do not allow for an infinitely dense and hot state of the universe at some point. Right?

 

[PeterDonis]

If I understand you correctly, Peter, the theorem you mentioned rules out inflationary models that extend infinitely into the past and also contain singularities

No. There are no such models; "extend infinitely into the past" and "contain singularities" are contradictory. They can't both be true of the same model.

Inflationary models that extend infinitely into the past are only valid if they do not allow for an infinitely dense and hot state of the universe at some point. Right?

There is no such thing as "an infinitely dense and hot state of the universe". That would violate the laws of physics.

A model with a singularity, such as FRW models with an initial singularity, does not have such a state anywhere; that's not the correct definition of a singularity. The correct definition of a singularity is that worldlines only extend for a finite proper time. So in FRW models with an initial singularity, worldlines only extend for a finite proper time into the past.

It is true that in most known spacetimes with a singularity, there is some physical quantity that increases without bound as the singularity is approached. In the case of FRW spacetime, that's the density. (Temperature is usually taken to increase without bound also, but technically that's not part of the FRW spacetime model, it's part of whatever separate model of the matter and radiation is being used.) But the value of such a quantity is still finite at every point in the actual spacetime. The singularity itself is not part of the spacetime.

 

[javisot]

No. There are no such models; "extend infinitely into the past" and "contain singularities" are contradictory. They can't both be true of the same model.

But what's the correction Peter? You said that the BGV theorem can't be asserted as entirely ruling out models that extend infinitely into the past.

So, according to you, the BGV theorem allows models that extend infinitely into the past but meet another condition. Which one?

 

[PeterDonis]

what's the correction Peter?

Correction to what?

according to you, the BGV theorem allows models that extend infinitely into the past

In the sense that, while every worldline has a finite extent into the past, there is no upper bound to those finite extents. That's the sort of model that Linde was referring to in the paper referenced in the other thread that @Jaime Rudas linked to. Note that this kind of model would be very different from a standard FRW model, where there is an upper bound to the finite extent into the past of any worldline (in our best current model of our actual universe, this is about 13.8 billion years.)

but meet another condition.

I'm not sure what you're referring to here. Please quote specific statements I made that you are asking about, and say what you think they mean or what you think is unclear about them.

 

[javisot]

I'm not sure what you're referring to here. Please quote specific statements I made that you are asking about, and say what you think they mean or what you think is unclear about them.

Ok, I'll try to be more direct, what models exactly does the BGV theorem ruled out?, what properties should they have?

At one point you say:

It does rule out at least most "eternal inflation" models, which I believe was the intended point of those who proved the theorem. However, I think there are still other "eternal inflation" models (though highly speculative ones) that it doesn't rule out. I think we had a thread on this some time ago.

But it's not very specific, can you elaborate on this?

 

[PeterDonis]

what models exactly does the BGV theorem ruled out?

It rules out models that (a) are expanding everywhere, and (b) do not have a past spacetime boundary.

As I noted earlier, de Sitter spacetime meets condition (b), but does not meet condition (a); it's not expanding everywhere (in the sense that's relevant for the theorem).

Most "eternal inflation" models meet condition (a), but do not meet condition (b); they have no past spacetime boundary.

The unusual "eternal inflation" models referred to in the Linde paper meet both conditions, but the way they meet condition (b) is counterintuitive; there is no way to give a maximum proper time along a worldline to the past spacetime boundary in such models. By contrast, in a model like our best current model of the universe, we can give such a time (about 13.8 billion years, as I said before).

 

[Jaime Rudas]

It rules out models that (a) are expanding everywhere, and (b) do not have a past spacetime boundary.
[...]
Most "eternal inflation" models meet condition (a), but do not meet condition (b); they have no past spacetime boundary.

Wouldn't it be better to say that they do not meet condition (b) because they have a past spacetime boundary?

 

[PeterDonis]

Wouldn't it be better to say that they do not meet condition (b) because they have a past spacetime boundary?

Yes, the way I phrased it was odd. I should have said that the eternal inflation models that are ruled out meet condition (b) (because models that meet both conditions are the ones that are ruled out), while the Linde eternal inflation model is not ruled out because it doesn't meet condition (b).