Some questions on the Cosmological Principle

[rede96]

Firstly, I am just a very interested layman so please forgive my ignorance and non mathematical approach.

As I understand it, the cosmological principle states that on larger scales the universe is homogeneous and isotropic. So could someone help me to understand the following:

1) Does this imply that the universe is infinite? As any observer must be able to see the same from any point in space?

2) Wouldn't this also suggest that the gravitational forces acting upon galaxies (or groups of galaxies) across large distances cancel out, so we should have a static universe on larger scales not an expanding one? (EDIT: If we were to ignore dark energy)

3) Assuming a big bang, then doesn't that imply there must be an edge to our universe?

Thanks for any help.

 

[DaveC426913]

'Finite yet boundless' is the phrase often used.

The surface of a sphere is homogeneous and isotropic, yet it is not infinite in extent, nor does it have a boundary.

The universe is thought to be finite in a higher dimension. i.e. it wraps around

 

[PeterDonis]

Does this imply that the universe is infinite?

No. As DaveC points out, a spatially finite but unbounded universe, such as a 3-sphere, also satisfies the principle.

Wouldn't this also suggest that the gravitational forces acting upon galaxies (or groups of galaxies) across large distances cancel out, so we should have a static universe on larger scales not an expanding one?

No. Look up the Friedmann equations.

Assuming a big bang, then doesn't that imply there must be an edge to our universe?

No. Either the universe is spatially infinite, or finite but unbounded. In either case there is no edge.

 

[rede96]

The surface of a sphere is homogeneous and isotropic, yet it is not infinite in extent, nor does it have a boundary.

Thanks for the reply. I sort of get that but find it very difficult to understand in 3 dimensions. I understood observations showed that the universe was pretty flat. Also, if there is a slight positive curvature, which would make the universe round on larger scales, then the universe would be much bigger and possibly older than we have calculated?

 

[PeterDonis]

I sort of get that but find it very difficult to understand in 3 dimensions.

It's the same as in 2 dimensions, just with one more dimension added. If you can see how a 2-sphere has a finite area but no boundary, then a 3-sphere having a finite volume but no boundary works the same way.

I understood observations showed that the universe was pretty flat.

"Pretty flat" is not the same as "absolutely flat". The best fit to our observations is a spatially flat and infinite universe, but the "error bars" are large enough that a spatially finite universe with a very, very large radius of curvature is still possible.

if there is a slight positive curvature, which would make the universe round on larger scales, then the universe would be much bigger and possibly older than we have calculated?

No. Our estimates of the age of the universe do not depend on our estimates of its spatial size/finiteness.

 

[DaveC426913]

No. Our estimates of the age of the universe do not depend on our estimates of its spatial size/finiteness.

Especially since we believe the observable universe is about 92 billion light years across, yet only 13.7 billion years old.

 

[rede96]

No. Look up the Friedmann equations.

I was just watching some of Leonard Susskind's lectures on cosmology, but haven't quite understood the Friedmann equations yet. I was more thinking of Newton's (shell) theorem and how that is applied to a homogeneous and isotropic universe and it just seemed that everything would cancel out.

 

[rede96]

No. Look up the Friedmann equations.

I suppose the way I am looking at it is, if the universe is slightly curved then taking the 2d sphere analogy, if I place a number of galaxies equidistant around the equator of the sphere and assume there is no other matter, then the gravitational forces acting on the galaxies cancel out and they don't move wrt each other. Of course the sphere can 'expand' and the distance between the galaxies still remains equidistant, but in that analogy I can't see how gravity has anything to do with the sphere expanding as the Friedmann equations would suggest.

 

[rede96]

If you can see how a 2-sphere has a finite area but no boundary, then a 3-sphere having a finite volume but no boundary works the same way.

Sorry, my mind just doesn't grasp that! :) I can see how that works on the surface of a sphere, which is 2 dimensional, but we live in 3d space. So I can't see how there is another dimension 'expanding' which in turn causes our 3d 'space' to expand and the distance between galaxies grow.

 

[marcus]

I suppose the way I am looking at it is, if the universe is slightly curved then taking the 2d sphere analogy, if I place a number of galaxies equidistant around the equator of the sphere and assume there is no other matter, then the gravitational forces acting on the galaxies cancel out and they don't move wrt each other. Of course the sphere can 'expand' and the distance between the galaxies still remains equidistant, but in that analogy I can't see how gravity has anything to do with the sphere expanding as the Friedmann equations would suggest.

I like the question, but I can't think of a succinct response right at the moment. Hopefully someone else will.

...Sorry, my mind just doesn't grasp that! :) I can see how that works on the surface of a sphere, which is 2 dimensional, but we live in 3d space. So I can't see how there is another dimension 'expanding' which in turn causes our 3d 'space' to expand and the distance between galaxies grow.

Around 1850 Georg Riemann showed how you could describe a 2-sphere without referring to any surrounding 3d space. You could describe the curvature and distances entirely from the POV of a 2d amoeboid creature living in the 2d sphere. All existence was concentrated on that infinitesimally thin 2d spherical surface. There was no "inside" or "outside" or "3rd dimension". Respect and gratitude to Riemann. He contributed a lot else besides, to mathematics. Oh, it was not limited to the 2-sphere. All kinds of manifolds with different shapes could exist and be described and studied without assuming any higher dimension space for them to live in. they did not need to be "embedded" in order to exist.

So the analog, a 3-sphere, does not have to have an inside or an outside or a "4th dimension". Just think about the experience of the 3d creature roaming around in it, measuring distances and angles of triangles etc , comparing radius and volume of 3d balls. circumnavigating. It doesn't have to be "embedded" in order for him to report his experiences exploring the geometry of his 3-sphere world. At least I think it doesn't.

It can expand just by all the internal distances getting larger---it doesn't have to be embedded, or have a center point.

 

[PeterDonis]

I was more thinking of Newton's (shell) theorem and how that is applied to a homogeneous and isotropic universe and it just seemed that everything would cancel out.

The shell theorem says that, if the configuration of matter outside some region is spherically symmetric, you can ignore its gravitational effect on the matter inside the region. But that doesn't mean everything cancels out. See below.

if the universe is slightly curved then taking the 2d sphere analogy, if I place a number of galaxies equidistant around the equator of the sphere and assume there is no other matter, then the gravitational forces acting on the galaxies cancel out

No, that won't work, because the configuration is not symmetric in a circle (since we're in 2d instead of 3d, the shell theorem applies to a circle instead of a sphere) around a given point. You would have to have a homogeneous distribution of galaxies everywhere on the 2d sphere. Then, if you pick any point on the 2d sphere and consider a small circular region around it ("circular region" meaning a region of the surface of the 2d sphere bounded by a circle drawn on the 2d sphere at some radius along the 2d sphere around a given point), the distribution of matter outside that small circular region is symmetric ("circularly symmetric", by analogy with "spherically symmetric" in the 3d space case), so that distribution has no net gravitational effect on the matter inside the small circular region. However, different pieces of matter inside the small circular region can still have effects on each other. See further comments below.

Of course the sphere can 'expand'

Exactly; all of the above was talking about the sphere at some instant of time. But it doesn't say anything about how the size of the sphere itself changes with time.

in that analogy I can't see how gravity has anything to do with the sphere expanding as the Friedmann equations would suggest.

The Friedmann equations don't say that gravity is what's causing the expansion. The expansion itself is due to initial conditions and inertia; the model assumes that the universe started out expanding. The shell theorem doesn't rule that out; in the 2d sphere analogy, the shell theorem doesn't prevent the 2d sphere from having some given expansion rate at a given moment of time. In Newtonian terms, gravity affects acceleration, not velocity; we can give the 2d sphere any velocity we like at some instant of time as far as gravity is concerned.

What the Friedmann equations tell you is how the rate of expansion changes with time, depending on the matter and energy present. In the 2d sphere analogy, it tells you how the rate of expansion of the sphere changes with time. If ordinary matter and radiation are present, that rate of expansion will decrease with time, and that makes intuitive sense because the matter and radiation causes attractive gravity. Your question basically is, how is that consistent with the shell theorem?

Let's go back to the 2d sphere analogy, and let's assume that the 2d sphere is expanding. From the viewpoint of an observer at some point on the 2-sphere (i.e., living in some particular galaxy), this expansion appears as all the other galaxies moving away from him. If we choose a small circular region around his galaxy, so small that no other galaxies are in it, then we see that the gravitational effects of all those other galaxies on him cancel out; he experiences no net force, and stays where he is, which makes sense. But this analysis can't tell us anything about whether, or how, the motion of other galaxies relative to him will change with time.

In order to investigate that, we have to pick a circular region that is large enough to contain at least one other galaxy besides his. Let's suppose the galaxies around him are symmetric, so that if we enlarge the circle just enough, it will contain some number of galaxies evenly distributed around the edge of the circular region. The gravitational force of all these galaxies on our observer will still cancel, so he will still stay where he is. But the force his galaxy exerts on all the other galaxies does not cancel out; each of those galaxies will be pulled towards him by the gravity of his galaxy. Since all those other galaxies are moving away from him, the effect of the pull of his galaxy will be to decelerate them. And since the same analysis applies no matter which galaxy we pick to be at the center of a circular region, we conclude that the expansion as a whole will decelerate. And that is what the Friedmann equation describes (with one spatial dimension added), for the case of a universe filled with ordinary matter and energy.

I can't see how there is another dimension 'expanding'

There isn't. The 3d space we live in does not have to be embedded in any higher-dimensional space in order to expand. Neither does a 2d sphere need to be embedded in a 3-space in order to expand; the fact that we visualize a 2-sphere embedded in a 3-space, or the fact that the 2-sphere we happen to live on is embedded in a 3-space, does not mean any 2-sphere must be embedded in a 3-space. Logically, mathematically, and physically, that simply is not a requirement.

 

[rede96]

Around 1850 Georg Riemann showed how you could describe a 2-sphere without referring to any surrounding 3d space. You could describe the curvature and distances entirely from the POV of a 2d amoeboid creature living in the 2d sphere. All existence was concentrated on that infinitesimally thin 2d spherical surface. There was no "inside" or "outside" or "3rd dimension". Respect and gratitude to Riemann. He contributed a lot else besides, to mathematics. Oh, it was not limited to the 2-sphere. All kinds of manifolds with different shapes could exist and be described and studied without assuming any higher dimension space for them to live in. they did not need to be "embedded" in order to exist.

Ok, thanks for that. I can't say I fully understand that but will have a read on related material.

So the analog, a 3-sphere, does not have to have an inside or an outside or a "4th dimension". Just think about the experience of the 3d creature roaming around in it, measuring distances and angles of triangles etc , comparing radius and volume of 3d balls. circumnavigating. It doesn't have to be "embedded" in order for him to report his experiences exploring the geometry of his 3-sphere world. At least I think it doesn't.

I guess where my mind struggles is that I can see the natural properties of a 2d surface on a sphere lend itself to having a finite area and no boundaries, but I don't know of any 3d shape / geometry that is the same. All 3d shapes have boundaries, so that implies that our 3d universe must have a boundary or be infinite.

 

[rede96]

The shell theorem says that, if the configuration of matter outside some region is spherically symmetric, you can ignore its gravitational effect on the matter inside the region. But that doesn't mean everything cancels out

I suppose what I was thinking was if we have 3 galaxies all of the same mass and all equidistant along some x axis, then the two outside galaxies would move towards the middle galaxy. I could place another 2 galaxies at either side of the end galaxies which would cancel out their effect, but the result would still be that the outside 2 galaxies would move towards the centre. I could keep adding galaxies in this way but the net effect would always be movement towards the centre unless there were an infinite amount of galaxies.

However taking the 2d analogy, I can place a finite amount of galaxies on the surface of there sphere so they are isotropic and hence all the effects of gravity cancel out. (I've put more detail below)

So in that case, the only way we could have a big crunch, would be if the galaxies were not all isotropic or some galaxies had a lot more mass than others.

The Friedmann equations don't say that gravity is what's causing the expansion. The expansion itself is due to initial conditions and inertia; the model assumes that the universe started out expanding. The shell theorem doesn't rule that out; in the 2d sphere analogy, the shell theorem doesn't prevent the 2d sphere from having some given expansion rate at a given moment of time. In Newtonian terms, gravity affects acceleration, not velocity; we can give the 2d sphere any velocity we like at some instant of time as far as gravity is concerned.

What the Friedmann equations tell you is how the rate of expansion changes with time, depending on the matter and energy present. In the 2d sphere analogy, it tells you how the rate of expansion of the sphere changes with time. If ordinary matter and radiation are present, that rate of expansion will decrease with time, and that makes intuitive sense because the matter and radiation causes attractive gravity. Your question basically is, how is that consistent with the shell theorem?

Let's go back to the 2d sphere analogy, and let's assume that the 2d sphere is expanding. From the viewpoint of an observer at some point on the 2-sphere (i.e., living in some particular galaxy), this expansion appears as all the other galaxies moving away from him. If we choose a small circular region around his galaxy, so small that no other galaxies are in it, then we see that the gravitational effects of all those other galaxies on him cancel out; he experiences no net force, and stays where he is, which makes sense. But this analysis can't tell us anything about whether, or how, the motion of other galaxies relative to him will change with time.

In order to investigate that, we have to pick a circular region that is large enough to contain at least one other galaxy besides his. Let's suppose the galaxies around him are symmetric, so that if we enlarge the circle just enough, it will contain some number of galaxies evenly distributed around the edge of the circular region. The gravitational force of all these galaxies on our observer will still cancel, so he will still stay where he is. But the force his galaxy exerts on all the other galaxies does not cancel out; each of those galaxies will be pulled towards him by the gravity of his galaxy. Since all those other galaxies are moving away from him, the effect of the pull of his galaxy will be to decelerate them. And since the same analysis applies no matter which galaxy we pick to be at the center of a circular region, we conclude that the expansion as a whole will decelerate. And that is what the Friedmann equation describes (with one spatial dimension added), for the case of a universe filled with ordinary matter and energy.

Thanks for the great explanation. But here is where I am still struggling to understand:

Taking the 2d sphere analogy, if I place a finite amount of galaxies on the surface of the sphere in such a way that all galaxies are isotropic, i.e. all equidistant and all with the same mass, and I assume no expansion or other forces at work, then the net effect must be that all gravitational forces cancel out. No galaxies move towards each other. It can't be any other way.

So if the sphere does expand and the distance between galaxies grow, as the surface still remains isotropic, then there is still no net effect of gravity that could exert a force on the expansion. Does that make sense?

Now I could imagine that if we could assume another dimension (inside the sphere) and that gravity could permeate through that other dimension, then all the mass from all the galaxies would also be centred in the middle of the sphere. And that force would of course have an effect on the expansion. So if there was some initial force that 'inflated' the sphere in such a way that the force of expansion was just slightly greater than the net effect of gravity from all the galaxies accelerating towards the centre of the sphere (or in other words the outward force was just slightly greater than the escape velocity), then I could see how the rate of expansion would be effected and slow over time but never reaching zero. Which in some way I say analogous to our universe.

Of course that is just my simple way of looking at it and in any case I'd have no idea how that would transform to a 3d universe.

 

[Pizza]

Reading this thread was useful for me too.

 

[PeterDonis]

Taking the 2d sphere analogy, if I place a finite amount of galaxies on the surface of the sphere in such a way that all galaxies are isotropic, i.e. all equidistant and all with the same mass, and I assume no expansion or other forces at work, then the net effect must be that all gravitational forces cancel out.

No, it isn't. If you take a universe like that, which is not expanding or contracting at some instant, then in the next instant, it will start contracting. That's what the Friedmann equations say.

Deriving that result from the simplified model I gave, that you quoted, is easy: just consider the same circular region around a chosen galaxy, just large enough that there are some number of other galaxies at the edge of the region. In what you quoted, I assumed those galaxies were all moving away from the chosen galaxy, because that's the case in our actual universe. In that case, the other galaxies will decelerate, relative to the chosen galaxy, because of the chosen galaxy's gravity.

But if we assume, instead, that all the other galaxies are at rest relative to the chosen galaxy, then the chosen galaxy's gravity will cause them to accelerate inward, towards the chosen galaxy. And since that is true regardless of which galaxy we choose, the universe as a whole must begin contracting.

It can't be any other way.

Yes, it can. You have not responded to the argument I actually made; you're just ignoring it. Don't. Go back and re-read it.

 

[rede96]

Yes, it can. You have not responded to the argument I actually made; you're just ignoring it. Don't. Go back and re-read it.

Sorry, didn't mean to ignore it, just didn't understand it fully.

Your question basically is, how is that consistent with the shell theorem?

I'm not sure it was but I am probably not making myself clear, sorry. The issue I am struggling to understand is how can an isotropic and homogeneous universe with matter it 'contract' unless there is a finite amount of matter in it. (see my rationale below) And a finite amount of matter suggests that there are boundaries, at least as far as how matter can be distributed in that universe.

If you take a universe like that, which is not expanding or contracting at some instant, then in the next instant, it will start contracting. That's what the Friedmann equations say.

Would you mind helping me understand this better as I think it is where a lot of my confusion comes from. Accepting what the Friedmann equations say, when I think about our universe 'contracting', I think of it as if we place all the galaxies on some imaginary coordinate system. The ratio of the distances between the galaxies always stays the same but the actual distances between them reduce. But for this to happen then doesn't it mean that there must be a finite amount of matter / energy in the universe? Otherwise it would be impossible to trace the universe back to a singularity, as the universe would never stop contracting if there were an infinite amount of matter.

Also, taking the 2d analogy again, when we talk about contraction I assumed it meant that the galaxies are not moving around on the surface of the sphere, like above it meant that the distances between them that are getting smaller. But for that to happen then the contraction is moving the galaxies towards the centre of the sphere, so the sphere is 'shrinking', which in the analogy would mean the galaxies are moving into a space that was previously in a different dimension. So are we saying that as our 3d universe contracts that it is moving into a space that was previously in a different dimension too?

Maybe if you could help clear these two points up then I think the rest would fall into place. Again sorry for the my confusion.

 

[PeterDonis]

a finite amount of matter suggests that there are boundaries

No, it doesn't. This has already been addressed. The fact that you find it difficult to imagine a universe with spatial sections that are 3-spheres, containing a finite amount of matter but without any boundary, does not mean that model is not perfectly valid and self-consistent. It is.

Accepting what the Friedmann equations say, when I think about our universe 'contracting', I think of it as if we place all the galaxies on some imaginary coordinate system. The ratio of the distances between the galaxies always stays the same but the actual distances between them reduce.

This is a reasonable description, yes.

doesn't it mean that there must be a finite amount of matter / energy in the universe?

No. There are self-consistent contracting models with finite spatial slices (3-spheres without boundary, as I noted above), and also with infinite spatial slices. Both are perfectly valid models.

Otherwise it would be impossible to trace the universe back to a singularity

This is not correct. Contracting models with infinite spatial slices still have a singularity. The singularity theorems of Hawking & Penrose, proved in the late 1960's and early 1970's, cover this case.

What may be confusing you is the claim often made in pop science treatments that the singularity has "zero size". That is not correct. The singularity is not actually part of spacetime; when we talk about a singularity in such models, what we are really talking about is a mathematical limit that can be taken, and the fact that certain quantities increase without bound in that limit, while others remain finite. In the case of a collapsing universe (with either finite or infinite spatial slices), invariants related to the spacetime curvature increase without bound when the limit is taken, but the proper time along the worldline of any "comoving" observer does not; it converges to a finite value in the limit. The latter property (which is called "geodesic incompleteness"--there are geodesics, in this case the freely falling worldlines of comoving observers, which end at a finite value of their affine parameter) is what makes the singularity a "singularity". But since the singularity is not actually part of spacetime, there is nowhere, in the model with infinite spatial slices, where an infinite slice suddenly "turns into" a zero-size slice. Every spatial slice that is actually part of the spacetime is infinite in this model.

taking the 2d analogy again, when we talk about contraction I assumed it meant that the galaxies are not moving around on the surface of the sphere, like above it meant that the distances between them that are getting smaller

Correct.

But for that to happen then the contraction is moving the galaxies towards the centre of the sphere, so the sphere is 'shrinking'

Yes.

which in the analogy would mean the galaxies are moving into a space that was previously in a different dimension.

No. In the 2d analogy, the "space" inside and outside of the 2-sphere is not physically real. It's just a crutch that we need in order to visualize what is going on. The 2-sphere is a perfectly valid mathematical space in its own right and does not need to be embedded in a higher-dimensional space in order to exist. In the 2-d version of the model, the 2-sphere is the only "space" that exists, and it is perfectly possible for it to "contract" or "expand" without needing to be embedded in any other space.

So are we saying that as our 3d universe contracts that it is moving into a space that was previously in a different dimension too?

No. See above.

 

[rede96]

No. In the 2d analogy, the "space" inside and outside of the 2-sphere is not physically real. It's just a crutch that we need in order to visualize what is going on. The 2-sphere is a perfectly valid mathematical space in its own right and does not need to be embedded in a higher-dimensional space in order to exist. In the 2-d version of the model, the 2-sphere is the only "space" that exists, and it is perfectly possible for it to "contract" or "expand" without needing to be embedded in any other space.

ok, got that now. thanks.

No. There are self-consistent contracting models with finite spatial slices (3-spheres without boundary, as I noted above), and also with infinite spatial slices. Both are perfectly valid models.

Ok, thanks. Not really understanding this (see below).

No, it doesn't. This has already been addressed. The fact that you find it difficult to imagine a universe with spatial sections that are 3-spheres, containing a finite amount of matter but without any boundary, does not mean that model is not perfectly valid and self-consistent. It is.

Correct, I have no idea what you mean by spatial sections that are 3-sphere. So will need to do some research. However my point was not based on my lack of understanding per se, but on observations and logic. If the following is acceptable:

Accepting what the Friedmann equations say, when I think about our universe 'contracting', I think of it as if we place all the galaxies on some imaginary coordinate system. The ratio of the distances between the galaxies always stays the same but the actual distances between them reduce.

then I imagine we place our current universe against some imaginary coordinate system as above. But as the universe contracts, and the distances between galaxies reduce, then for each time interval the I compare new positions of the galaxies to original starting points what would I see? I would see all the galaxies moving towards me I assume. And just and it is valid for any observer to see all galaxies moving away from them, then it must follow that is just as valid for any observer to see all galaxies moving towards them. But how can all the galaxies reduce to an infinite set of different coordinate points? As I understand it, there is nothing in 3d geometry that would allow this. If something is shrinking from all directions that it must end up at a single point... no?

Just to be clear I am not stating any of the current models are wrong, I'm just trying to understand them.

 

[PeterDonis]

I have no idea what you mean by spatial sections that are 3-sphere.

Go back to the 2d sphere analogy again. If we include time, we don't have just one 2d sphere; we have a whole succession of them, one at each instant of time. Saying that the 2d sphere universe is "expanding" or "contracting" just means that the 2d spheres at different instants have different sizes. Each 2d sphere is a "spatial section" of the universe; the whole universe, from the GR point of view, is the "spacetime" comprised of all the 2d spheres, "stacked up" in the order in which they occur in time.

The case of the 3-sphere universe is the same, just with one spatial dimension added. The whole spacetime is a succession of 3-spheres, one at each instant of time, whose sizes can vary. Each individual 3-sphere is a spatial section of this spacetime.

as the universe contracts, and the distances between galaxies reduce, then for each time interval the I compare new positions of the galaxies to original starting points what would I see?

How do you propose to do this comparison? You can't compare new positions to old positions, because there is nothing marking out the old positions; the only "markers" are the galaxies themselves, and by assumption the distances between them are decreasing. Those distances are the only "markers" you have, and saying that the universe is "contracting" is just another way of saying that those distances are decreasing with time.

how can all the galaxies reduce to an infinite set of different coordinate points?

I'm not sure what you mean by this. Where in the model do you think this is happening?

If something is shrinking from all directions that it must end up at a single point... no?

No. The 3-sphere never contracts to a point; the "singularity" that pop science presentations refer to as a point is not actually part of spacetime. It's just an abstract limit that can be taken, as I said before.

Also, you have to stop thinking of the 3-sphere as embedded in some higher dimensional space. It isn't. It's a space in its own right. It isn't "shrinking from all directions" when it's contracting, like a balloon with the air escaping. It's just that 3-spheres at later instants of time are smaller (have less total volume) than 3-spheres at earlier times. That's all there is to it.

 

[rede96]

The case of the 3-sphere universe is the same, just with one spatial dimension added. The whole spacetime is a succession of 3-spheres, one at each instant of time, whose sizes can vary. Each individual 3-sphere is a spatial section of this spacetime.

I think I've got this now, thanks. Well at least I understand it is mathematical model equivalent to the surface of an n+1 dimensional ball. (Looked that up) The only thing that still confuses me is how we can be on the surface of a 3-sphere. Something can be on the surface of a 2-sphere, as it is a natural property of any 3 dimensional object to have a surface. So struggling to see how we can just assume an additional dimension exists?

Also, I assume you don't mean that each 'slice' exist in time simultaneously?

How do you propose to do this comparison?

By measuring blue shift (Assuming a contracting universe) in the same way cosmologists have measured red shift and deduced the universe is expanding and was smaller in size in the past. So if I measure the blue shifts in a receding universe I can map out the positions of galaxies over time. The data would (as I understand it) tell me that all galaxies are receding 'towards' me. In much the same way we say now that all galaxies (on larger scales) are moving away from us.

I'm not sure what you mean by this. Where in the model do you think this is happening?

Currently anyone in any galaxy can say that all the galaxies are moving away equally from their point of view. So if Alice is in her galaxy and Bob in a different distant galaxy, then in a contracting universe Alice would observe all the galaxies moving towards her and Bob would observe all the galaxies moving towards him.

However it is not possible for the all the galaxies to eventually converge at both Alice's and Bob's coordinate positions simultaneously. So at some point in time Alice's coordinate position and Bob's coordinate position must merge as the distance between them reduced to zero. (My rationale for this is below)

No. The 3-sphere never contracts to a point; the "singularity" that pop science presentations refer to as a point is not actually part of spacetime. It's just an abstract limit that can be taken

I wasn't actually thinking of pop science to be honest. I was thinking of the 2-sphere analogy. In the 2-sphere analogy as the distance reduced between the galaxies, the galaxies are moving closer and closer to the centre of the 3 dimensional sphere. We agreed that in an earlier post. So at some point in time all the galaxies must converge into a single point, or at least try to. So if the analogy between 2-sphere and 3-sphere holds then it must be the same for our universe. Sort of like if I ask everyone on Earth to simultaneously throw a ball into the air, then when the balls fall back to earth, the are all heading for the centre of mass.

So I assumed all that mass must collapse into a singularity.

Also, you have to stop thinking of the 3-sphere as embedded in some higher dimensional space.

I didn't realize I was? Can you give me an example?

It's a space in its own right. It isn't "shrinking from all directions" when it's contracting, like a balloon with the air escaping. It's just that 3-spheres at later instants of time are smaller (have less total volume) than 3-spheres at earlier times. That's all there is to it.

Ah, ok. That's confused me. We use the balloon analogy to explain the expanding universe with us being on the surface of the balloon. So why doesn't the analogy work in reverse?

 

[Torbjorn_L]

I was just watching some of Leonard Susskind's lectures on cosmology, but haven't quite understood the Friedmann equations yet. I was more thinking of Newton's (shell) theorem and how that is applied to a homogeneous and isotropic universe and it just seemed that everything would cancel out.

Odd, it was precisely Susskind's free Stanford University lecture series that made me superficially grok this. He starts out by covering the scale factor description and Newton's shell theorem over cosmological scales, then show how you can make a Newtonian expanding universe out of that. The physics will be that of a throw 'from the shell', with a parabola as the bounding case of just so escape at infinity analogous to escape velocity from Earth.

All the general relativity and the curvature comes later, and the global topology even later. I cheated and haven't seen his GR lectures before or after. (As he himself notes, you don't have to in order to follow his cosmology lecture series.) Nor does he introduce the full Friedman's equations but derives a successive useful set of heuristic solutions as he covers the cosmological eras. Maybe you should plow through the series, a rewarding experience, before attempting the FE. If so. note that he has at least two series, and the material is not identical. [Embarrassing disclaimer: His lectures enabled me to cheat on Friedman's derivations too! ]

TL;DR: In Susskind's cosmology lectures, expansion is an observation. It solves a lot of problems of a static universe such as that a static universe is incompatible with the cool (non-star surface temperature) space. It matches the observations of CMB et cetera, and he makes a really simple to understand walk through how that works out in numbers (CMB temperature, flat space, matter/antimatter, expansion rates in different eras, ...).

I can't remember if he also refers to GR's problems of a static solution - which possibly one can infer from the similarly just so solution for the Newtonian universe - I guess not since there was little of quantifiable GR in it.

 

[PeterDonis]

I assume you don't mean that each 'slice' exist in time simultaneously?

This question doesn't make sense. "Exist in time simultaneously" isn't a concept that even applies to spatial slices of a spacetime. Spatial slices of a spacetime are not the same as individual events in the spacetime, which are the kinds of things that the concept of "simultaneous" applies to.

By measuring blue shift (Assuming a contracting universe)

This will tell you that distant galaxies are moving towards you (instead of away from you as the redshifts we currently observe tell us), yes. If that is what you meant by "compare the new positions of galaxies to their original starting points", then that's fine. It just wasn't clear from that phraseology that that was what you meant.

if Alice is in her galaxy and Bob in a different distant galaxy, then in a contracting universe Alice would observe all the galaxies moving towards her and Bob would observe all the galaxies moving towards him.

Yes.

it is not possible for the all the galaxies to eventually converge at both Alice's and Bob's coordinate positions simultaneously. So at some point in time Alice's coordinate position and Bob's coordinate position must merge as the distance between them reduced to zero.

This is incorrect. The model of the contracting universe does not include a "point" at which all the galaxies converge; this "final singularity" is not actually part of the model. It is only a sloppy way of referring to mathematical limits that can be taken along the worldlines of galaxies.

In the 2-sphere analogy as the distance reduced between the galaxies, the galaxies are moving closer and closer to the centre of the 3 dimensional sphere.

NO THEY ARE NOT. The emphasis is because I have repeatedly told you that the 2-sphere in that analogy is not embedded in any higher-dimensional space, yet you continue to talk as if it is. If you can't stop doing that, there's no point to this discussion, because you will be using an incorrect model.

We agreed that in an earlier post.

I agreed to no such thing. I did precisely the opposite. See above.

I didn't realize I was? Can you give me an example?

See above for an example where you did it with the 2-sphere in the 2-d analogy. Also please see previous posts in this thread where I have responded to specific statements of yours with this same issue. Please read more carefully; you are repeatedly missing important points that I am trying to make.

We use the balloon analogy to explain the expanding universe with us being on the surface of the balloon.

And in that analogy we emphasize that only the surface of the balloon is "real"; the 3-dimensional space that we can't help visualizing the balloon as being embedded in is not "real" and is not part of the analogy.

 

[rede96]

I agreed to no such thing. I did precisely the opposite.

My Post:

taking the 2d analogy again, when we talk about contraction I assumed it meant that the galaxies are not moving around on the surface of the sphere, like above it meant that the distances between them that are getting smaller.

Your reply:

Correct.

My post:

But for that to happen then the contraction is moving the galaxies towards the centre of the sphere, so the sphere is 'shrinking',

Your Reply:

Yes.

What am I supposed to read into that? If the sphere is shrinking, and the space between galaxies getting smaller and smaller, where do the galaxies physically end up?

This seems to be what is causing all the confusion for me, so perhaps you can explain just where all the galaxies would finish in relationship to each other in a contracting universe?

 

[PeterDonis]

What am I supposed to read into that?

Read the very next quote and response in the same post. You said:

which in the analogy would mean the galaxies are moving into a space that was previously in a different dimension

I responded:

No. In the 2d analogy, the "space" inside and outside of the 2-sphere is not physically real.

You have to look at everything I'm saying, and look at it in context. Yes, I didn't catch the fact that you said "towards the center of the sphere" once, so I gave a "yes" response to a statement that actually wasn't completely correct. (Part of it was--the part that said the sphere is shrinking. I just didn't catch the other part.) But I said explicitly "the space the sphere is embedded in is not physically real" or words to that effect multiple times. I'm not going to catch everything and you can't look at one thing I say and ignore everything else I say.

If the sphere is shrinking, and the space between galaxies getting smaller and smaller, where do the galaxies physically end up?

On the sphere. See below.

perhaps you can explain just where all the galaxies would finish in relationship to each other in a contracting universe?

They don't "finish" anywhere. As I said before, the "final singularity" of "zero size" is not actually part of the model. In the model, the galaxies just get closer and closer and closer together, but never actually "finish" at one point.

Remember also that this is just a model, and most physicists would say that, because of the "singularity" (or more precisely the fact that the galaxies don't "finish" anywhere even though only a finite time elapses on their clocks), the model, or at least the part of it at the very end, is not physically reasonable. Some other physics must take over when the spacetime curvature becomes large enough (or, in the analogy, when the 2-d sphere becomes small enough), and cause something different to happen. But we don't know what that other physics is; the best we can say is that it probably has something to do with quantum gravity.

 

[rede96]

You have to look at everything I'm saying, and look at it in context. Yes, I didn't catch the fact that you said "towards the center of the sphere" once, so I gave a "yes" response to a statement that actually wasn't completely correct. (Part of it was--the part that said the sphere is shrinking. I just didn't catch the other part.)

Ah, ok. And sure I know and wouldn't expect you to proof read everything I write, but unfortunately that particular point was quite a key lever to my previous way of thinking so I just didn't understand your other points due to my mind being locked on the above. Anyway, its all clear now. But I can assure you I wasn't intentionally ignoring what you wrote and I really appreciate your time in responding.

Remember also that this is just a model, and most physicists would say that, because of the "singularity" (or more precisely the fact that the galaxies don't "finish" anywhere even though only a finite time elapses on their clocks), the model, or at least the part of it at the very end, is not physically reasonable.

Are there any other theories / speculation about what happens after the math breaks down?

Some other physics must take over when the spacetime curvature becomes large enough (or, in the analogy, when the 2-d sphere becomes small enough), and cause something different to happen

Although the physics breaks down, are there any parallels drawn between this and how black holes are formed?

But we don't know what that other physics is; the best we can say is that it probably has something to do with quantum gravity.

Is that related to trying to find a theory of everything?

They don't "finish" anywhere. As I said before, the "final singularity" of "zero size" is not actually part of the model. In the model, the galaxies just get closer and closer and closer together, but never actually "finish" at one point.

Do you mind if I ask at what point or distance does the math breakdown? And I take it this is for an open universe?

 

[Pizza]

But since the singularity is not actually part of spacetime, there is nowhere, in the model with infinite spatial slices, where an infinite slice suddenly "turns into" a zero-size slice.

And how should a minimalistic slice turn into an infinite?

 

[PeterDonis]

its all clear now. But I can assure you I wasn't intentionally ignoring what you wrote and I really appreciate your time in responding.

No worries. I'm glad it's clear now.

Are there any other theories / speculation about what happens after the math breaks down?

Yes; they all come under the broad heading of quantum gravity. But there are quite a few different proposed quantum gravity theories, and quite a bit of dispute over which one has the better chance of being right.

Although the physics breaks down, are there any parallels drawn between this and how black holes are formed?

Only in the sense that the very early universe and the region close to the singularity of a black hole are both regions where classical GR is expected to break down. But the two spacetimes overall are very different, and there aren't really any fruitful analogies between them (though some pop science sources try to make it seem like there is).

Is that related to trying to find a theory of everything?

A lot of quantum gravity theorists appear to think so. Personally, I'm not so sure; the pursuit of a quantum gravity theory will, I think, give us a more comprehensive theory than we have now, but for it to be a theory of everything, we would have to be sure that it covered all possible phenomena, not just the phenomena we know about, and I don't see how we could ever be sure of that.

at what point or distance does the math breakdown?

That's one of the things we'll only know for sure when we have a more comprehensive theory. The common belief is that classical GR breaks down near the Planck scale--roughly, when the "radius of curvature" of spacetime approaches the Planck length, or when the density of matter/energy approaches one Planck mass per Planck volume. But that's really just an educated guess; we have no evidence of where the breakdown actually occurs. Our most accurate experiments are still twenty or so orders of magnitude away from the Planck scale, so the fact that we have seen no breakdown of classical GR in the regime we have explored doesn't tell us much; classical GR could break down ten orders of magnitude before the Planck scale is reached, and we would still see no evidence of it in our current experiments. So there is a very wide range of uncertainty still.

I take it this is for an open universe?

Not just an open universe; a collapsing closed universe has the same problem.

 

[PeterDonis]

how should a minimalistic slice turn into an infinite?

It doesn't. The expanding universe is just the time reverse of the collapsing one; for the spatially infinite case, every spatial slice is infinite. The different slices have different densities of matter and energy, but they're all spatially infinite.

 

[rede96]

note that he has at least two series, and the material is not identical.

Thanks for the heads up, I was wondering why it was getting confusing. Found both now.

 

[rede96]

I'm glad it's clear now

I need to do some more reading to catch up / better understand some parts, but thanks again for the help.

 

[Pizza]

Am sorry, but the slices have to have a shape near the big bang too. And from there it should describe slices - and I wonder how they should grow from finite to infinite!?

 

[PeterDonis]

but the slices have to have a shape near the big bang too.

What "shape" are you talking about?

 

[Pizza]

What "shape" are you talking about?

the slices shape, form, appearance - from the outside view. A slice is like a cut from a bigger thing, like a 2d from a 3d.
A slice here, like you described, is like a screenshot of a moment of time in space. Isn't it?

So what's about the big bang, what's about the slices there? and how can a finite slice turn to infinite?

Let me ask otherwise. How can grow finite to infinte? Thats all I want to know, to be honest.

 

[PeterDonis]

A slice here, like you described, is like a screenshot of a moment of time in space. Isn't it?

More or less.

what's about the big bang, what's about the slices there? and how can a finite slice turn to infinite?

As I've already said, it doesn't. Every slice in the spatially infinite model is infinite. If by "the big bang" you mean "a slice with t = 0 that is just a single point", as I've said before, there is no such slice. The "initial singularity" is not part of the model.

 

[Pizza]

Okay, let's talk then about a finite thing growing to infinite. The thing was never zero big at any time. And then, wow, got infinite, Is it that, what I have to believe?

How could you really make sure (and prove) that something grows to infinite?

 

[PeterDonis]

let's talk then about a finite thing growing to infinite.

That does not happen. Please, read more carefully. I explicitly said, several times now, that every spatial slice is infinite. Nothing grows from finite to infinite. It's always infinite.

 

[Buzz Bloom]

Sorry, my mind just doesn't grasp that! :) I can see how that works on the surface of a sphere, which is 2 dimensional, but we live in 3d space. So I can't see how there is another dimension 'expanding' which in turn causes our 3d 'space' to expand and the distance between galaxies grow.

Hi rede96:

I will try a different approach to explaining what I think you are not getting.

First consider the entire 3D universe at a fixed time, so its expansion is not relevant. Then think of a plane, say P, though the point where you are extending in all directions. In a flat infinite universe, P would be an infinite flat plane. If our 3D universe is finite, P really isn't a plane, but locally it almost is. As P extends from where you are as far as it can go, it will be a 2D curved surface of a sphere. Now this may be confusing because the center of this 2D sphere is not in the 3D universe at all. The math does not require that the center be discussed, because the equations related to the properties of the 2D P can be expressed without discussing where the center is. However, it may be conceptually helpful to think of the center of P as being in a direction from you that is outside of the 3D universe, and P is curved as the surface of a sphere in a 4th dimension. Of course, the particular plane P is not special. Any plane in any orientation at any point of the 3D universe would result in the same sized 2D spherical surface, with the same center point in the 4th dimension.

Using the analogy of the 3D curved universe to a 2D surface of a sphere like the (approximate) Earth's surface, if you are at the North pole, the analogy of plane P becomes a line through the North pole which when extended becomes a circle, that is, a circle of longitude. The center of all of these circles is the center of the Earth, which is not a point on the 2D surface.

I hope this is helpful.

Regards,
Buzz

 

[Torbjorn_L]

Thanks for the heads up, I was wondering why it was getting confusing. Found both now.

Been there, (unfortunately) done that. Wasted 30 minutes of productive study time.

Though I recouped it later by studying the differences too. There are some good material that is unique in both. (You don't need to watch both though, either one will do. The later is more up to date on the CMB results if I remember correctly, but that isn't too important.)

 

[Pizza]

That does not happen. Please, read more carefully. I explicitly said, several times now, that every spatial slice is infinite. Nothing grows from finite to infinite. It's always infinite.

Probably I didn't read it careful enough. But I do my best and read it again.

Lets say then there is no singularity (do not like it). Never was etc. That is, what I am talking about. I could imagine, that if something is zero and grows, would end in infinite. Just Imagination. Why should it end in finite? Why should zero should do anything else... (not mathematically of course, in a different manner, more physically)

So - and now - please be carefully - there was never any zero nor singularity, so the bing Bang happens from what? Not out of a singularity, not from zero. From Something. Not from nothing, right? I have some ideas about coming something out of nothing (without Religion), not kidding (discussed in Germany). But the case of the Big Bang, the standard, I've problems to understand? Why?

There is no evidence, that "nothing" "was" before. There is only something, coming out of something else, as I have to assume. According to the readings.Am sure.

 

[PeterDonis]

Lets say then there is no singularity (do not like it).

More precisely: in the idealized model we are talking about, there is no spatial slice labeled $t = 0$ that is of "zero size". All the spatial slices have infinite size. But there are spatial slices labeled by values of $t$ that are arbitrarily close to $0$.

I could imagine, that if something is zero and grows, would end in infinite. Just Imagination. Why should it end in finite?

I don't know what you are trying to say here, but once more: the spatial slices are always infinite. Nothing "grows" from zero to anything; nothing "ends in finite". Those things do not happen in the model.

there was never any zero nor singularity, so the bing Bang happens from what?

The idealized model we are talking about is not believed to apply to the very beginning of the universe; it only applies from the end of inflation forward. The end of inflation, where the universe was filled with matter and energy in a very hot, very dense, rapidly expanding state, is what is properly referred to as the "Big Bang", not the "initial singularity". Before that hot, dense, rapidly expanding state, we don't know for sure what existed; it depends on which version of inflation (or possibly some other model--the inflation models are currently thought to be the most likely, but they are not the only ones under consideration) ends up being confirmed by experiments.

 

[Pizza]

All the spatial slices have infinite size. But there are spatial slices labeled by values of t that are arbitrarily close to 0.

Thank you for your explanations. [Sorry for replying a little late - I had been overwhelmed by the latest events here in Europe eventually. My latest post here more or less a try to come back to everyday life - trying it again now]
I have probably too many difficulties to grasp the concept of the model, which - to make it even worse - starts just after the lynchpin, excluding it.
An infinite slice is no problem for my imagination, because every slice of the Euklidean space is infinite by default. This does not imply that "content" is infinite as well. Like a white paper without any border, hard to draw something "inflation-or-not" literally everywhere...

I have less difficulties with t the time. Time is when something is existent, consequently it can only start with the existence of "content", no matter what before. But how can something (content) grow to infinite isn't answered, in my eyes.

 

[PeterDonis]

This does not imply that "content" is infinite as well.

It does in the case of the universe--there must be, on average, the same density of matter everywhere in the infinite space. A finite region of matter surrounded by emptiness would not look the way our universe looks.

Time is when something is existent, consequently it can only start with the existence of "content", no matter what before.

In terms of models in the context of GR, this is not true; it is perfectly possible, theoretically, to have a universe with a meaningful "time" but empty of "content". Such a model does not match our actual universe, but it is still a consistent model.

how can something (content) grow to infinte

You still don't understand: it was always infinite. Nothing ever "grew" from finite to infinite. That includes "content" as well as "space"; in every infinite spatial slice, there is, on average, the same density of matter everywhere. (The density varies from one spatial slice to another, but it is the same everywhere in each slice.)

 

[Pizza]

Hmm, ok. But still difficult, emptiness with time, by the GR?

A finite region of matter surrounded by emptiness would not look the way our universe looks.

And why not? The principles would maybe allow that observation, or the observations, leading to these principles are maybe still consistent, if the border is, let's say, 100 times farther than the observable universe!?

 

[PeterDonis]

The principles would maybe allow that observation, or the observations, leading to these principles are maybe still consistent, if the border is, let's say, 100 times farther than the observable universe!?

No, they wouldn't. It's not enough to say "maybe"; you have to actually look at the math. In this case, the math is the Einstein Field Equation. A solution of the Einstein Field Equation describing a finite region of matter surrounded by emptiness would not look the way our universe looks, even if the finite region were much larger than our observable universe. That's why we don't use such solutions in cosmology; we use the FRW solutions, because they match observations.

 

[Pizza]

Allright then, will have to have a look on the math. I was not aware of that, that it would look differnt. Have to believe you. Are you sure? Even if it is 10000 times larger, we not sitting in the middle but somewhere far enough from any border?

 

[PeterDonis]

Even if it is 10000 times larger, we not sitting in the middle but somewhere far enough from any border?

Yes. Remember that we can see back 13.7 billion years or so, so we have an entire history of development that any model has to match. It's not just a matter of "well, we can't see an edge".

 

[Pizza]

Remember that we can see back 13.7 billion years or so, so we have an entire history of development that any model has to match.

But this is exactly what did not imply - in my thoughts - that edges are impossible, so may exist. Looking back in time would have the same appearance I imagined.

 

[PeterDonis]

But this is exactly what did not imply - in my thoughts - that edges are impossible, so may exist. Looking back in time would have the same appearance I imagined.

You may have "imagined" it that way, but that is not what the math actually says. "Imagining" is not a good strategy when dealing with this subject area; even cosmologists can have trouble "imagining" the things that the models tell us must be the case in order to match observations.

 

[Pizza]

Lets have a look then, if there are no loopholes in this math in question. For me actually "just" a matter of some (many) years or never. I like math, but am far beyond the necessary skills.
Ahm, I mean for the math itself, to find loopholes probably likely never of course.

Thanks for your patience with a layman.

 

[PeterDonis]

Lets have a look then

Are you familiar with the math of the FRW models? As discussed, for example, in these two Wikipedia articles?

https://en.wikipedia.org/wiki/Friedmann–Lemaître–Robertson–Walker_metric

https://en.wikipedia.org/wiki/Friedmann_equations

And are you familiar with the general class of asymptotically flat solutions to the Einstein Field Equation? ("Asymptotically flat" is GR-speak for "a finite region of matter surrounded by empty space".) See, for example, here:

https://en.wikipedia.org/wiki/Asymptotically_flat_spacetime