NB: can you try to make your next post a little more succinct please.
I apologize to everyone for the length of this post. It looks to me as though
Farsight has been sloppy in his terminology, speaking of "spacetime" when he really meant to say "space", or of FLRW solutions in general when he apparently meant to speak only of FLRW solutions in which 3-dimensional space is Euclidean. Fixing those misstatements wouldn't remove all of his errors, but it would help quite a bit, so I'm going to start by going through a severely abridged history of his recent posts to highlight uses of "spacetime", "FLRW", and other key terms he may wish to walk back.
[size=+1]
Review[/size]
"The Friedmann–Lemaître–Robertson–Walker (FLRW) metric is an exact solution of Einstein's field equations of general relativity; it describes a homogeneous, isotropic expanding or contracting universe..."
The first is from Einstein’s 1920 Leyden Address, where he describes a gravitational field as inhomogeneous space. The second is from the wiki FLRW page, which describes a homogeneous isotropic universe. There’s no gravity in that universe. Gravitational fields have been thrown out with the bathwater.
When I read that, I had no way of knowing that
Farsight was using the word "gravity" to mean "inhomogeneous space" or a Newtonian 3-dimensional gravitational field instead of the more modern notion (curved spacetime) that was introduced by Einstein, developed by Einstein's contemporaries such as Eddington and Weyl, and expressed nowadays by modern coordinate-free tensor notations instead of the coordinate-specific notations (such as affine connections) that were employed by Einstein and his contemporaries.
Furthermore, the FLRW solutions to which
Farsight explicitly referred include entire families of solutions in which space is definitely not Euclidean, and cannot be regarded as flat by any sane definition. So I wrote:
False. Although FLRW solutions (note the plural) assume the universe is homogeneous and isotropic, they allow for matter and gravity.
Farsight reiterated his error:
I reiterate: if you assume that the universe is homogeneous and isotropic, there's absolutely no gravity in it, spacetime is flat.
That's flat-out wrong, even if we substitute "space" for "spacetime". There are entire families of FLRW (hence homogeneous and isotropic) solutions in which neither space nor spacetime is flat. Einstein invented one of those solutions himself: the Einstein static universe.
A concentration of energy results in curved spacetime. This energy has a mass equivalence. Matter exhibits the property of mass but mass doesn't cause spacetime curvature, so much as the energy content of matter. Now go back to that Einstein quote where he referred to a gravitational field as inhomogeneous space, understand that if the energy distribution is absolutely uniform, space is homogeneous, there's no gravitational field, spacetime is flat, and light travels in straight lines.
Farsight is repeating his claim that homogeneous space implies flat spacetime. Even if we substitute "space" for "spacetime",
Farsight's claim is plainly incorrect, with Einstein's static universe serving as one of many FLRW counterexamples.
Of course I've seen them. The issue is in the interpretation, what the individual terms actually represent, what curved spacetime really is, and seeing the underlying simplicity.
Although we'd like to rescue
Farsight here by substituting "space" for "spacetime", he was responding to
Reality Check's observation that the curvature of spacetime appears explicitly in Einstein's field equations. In that context, we can't substitute "space" for "spacetime" without rejecting Einstein's equations, which
Farsight has said he does not wish to reject.
As if to prove he does not understand Einstein's field equations,
Farsight went on to write:
Let me try to get this across with a trivial example. You have a massive body in space. The energy present in this body "conditions" the surrounding space, the effect diminishing with distance, such that a light beam curves as it transits this space. Now add another massive body near to the first one, and shoot your light beam through the gap between them. It doesn't curve. However you can detect a Shapiro delay, so you can assert that spacetime is still curved. Now chop up your massive bodies into infinitesimal parts and distribute them evenly throughout the universe. Shoot a light beam through it, and that light beam doesn't curve. And you can't detect a Shapiro delay either. Forget about the expansion of the universe for a minute, and think about it. Light moves uniformly, in straight lines: your spacetime curvature has gone.
That example would make sense if we change "spacetime" to "space", but
Farsight gave this example in the context of Einstein's field equations, which involve the curvature of spacetime rather than space. Furthermore, as
sol invictus eventually pointed out, the example's Shapiro delay proves that curvature (of spacetime) can be detected even when light travels in Newtonian straight lines. Finally, the "forget about the expansion of the universe for a minute" is critical, because curvature would remain detectible if we were allowed to look at anything beyond a particular spacelike slice through spacetime.
I'm not misrepresenting them, I'm pointing out a clear issue. Here's the Einstein quote again followed by the quote from wiki:
”According to this theory the metrical qualities of the continuum of space-time differ in the environment of different points of space-time, and are partly conditioned by the matter existing outside of the territory under consideration. This space-time variability of the reciprocal relations of the standards of space and time, or, perhaps, the recognition of the fact that ‘empty space’ in its physical relation is neither homogeneous nor isotropic, compelling us to describe its state by ten functions (the gravitation potentials gμν)...
"The Friedmann–Lemaître–Robertson–Walker (FLRW) metric is an exact solution of Einstein's field equations of general relativity; it describes a homogeneous, isotropic expanding or contracting universe..."
...snip...
So you've got light moving uniformly in straight lines. Where's your curved spacetime gone? Space is homogeneous. There's no gravitational field. There is no gravity to make this universe collapse, even when it was much smaller, with a much higher energy density. And let's face it, if it had done, we wouldn't be here.
Although we can change
Farsight's use of "spacetime" to say "space" instead, neither he nor we are allowed to make that change when we're quoting others. In the example above,
Farsight quotes Einstein saying "space-time" and mentioning the 10 components of a 4-dimensional (spacetime, not space) tensor, but turns right around and scolds me for considering the curvature of spacetime instead of the curvature of a certain 3-dimensional slice of it. To top it off,
Farsight himself wrote "spacetime".
Don't throw Humpty Dumpty logic at me, sol. If light travels in straight lines spacetime isn't curved, and that's it.
Once again: It's a fact that spacetime can be curved even if all light travels in straight lines.
Farsight evidently meant to say "space" instead of "spacetime".
Unfortunately, he went on (in that same post, just one paragraph later) to scold
sol invictus for speaking of something happening in "spacetime" (which happens to be an unambiguous and standard usage, unlike
Farsight's):
Nothing
happens in spacetime sol, because spacetime is an "all-time-view", a block universe where we draw a world line to represent the motion of a body through space. Things happen in space. It's space that's expanding, not
spacetime. Again we see an issue over the distinction between curved space and curved
spacetime. Have a read of the
FLRW article on wiki and look out for it.
This curvature relies on the expansion of space, which at any one moment is flat and homogeneous and free of any overall gravity.
In that last sentence,
Farsight appears to admit that spacetime can be curved even when a particular 3-dimensional slice through spacetime isn't. If
Farsight had been willing to admit that most of his own statements about curved spacetime were rendered incorrect by his own sloppiness, there would be less to discuss and this would have been a shorter post.
Your definition of curvature seems to be elevating a manifold into something real rather than mathematical artefact. I think you’re confusing a spacetime manifold with physical space.
Is
Farsight accusing me of confusing a 4-dimensional spacetime manifold with 3-dimensional physical space? Or is he accusing me of confusing a 3-dimensional slice of a 4-dimensional manifold with 3-dimensional physical space? Or is he accusing me of confusing a 4-dimensional manifold with a 4-dimensional physical spacetime?
Farsight's accusation was wrong in any case, because I have not confused any of those things, but I answered him as though he were accusing me of the last of those confusions.
[size=+1]
Digression[/size]
Thanks. I think the issue with boards like this tends to be intellectual arrogance and dishonesty. Some people who think they understand relativity and cosmology cannot bear a hard evidential argument backed by cool clear logic. When their belief and expertise are challenged, they sometimes start getting nasty. I don't. My name (John Duffield) is public knowledge, and I have my reputation to think about.
Taking
Farsight at his word, I expect he will acknowledge
- his many uses of the word "spacetime" when he apparently meant to say "space",
- his misrepresentations of the FLRW solutions and their diversity,
- the fact that Einstein's field equations are expressed using spacetime curvature, and do not mention the curvature of any 3-dimensional space, and
- the fact that the spacetime curvature that appears in Einstein's field equations is mathematically well-defined.
If
Farsight will acknowledge those facts, along with the fact that his personal definition of gravity differs from the notion of Einsteinian gravity in common usage today, then we may be able to move on to a discussion of whether
Farsight's personal definition of gravity coincides with Einstein's. In connection with that, I want to quote
Vorpal:
Einstein said that the gravitational field is the connection, since he explicitly treated the connection coefficients as the components of the gravitational field. There's a good conceptual reason for this, since it explicitly dissolves any difference between gravity and inertia. But it doesn't affect Clinger's point here, since we're talking about curvature.
I'm really not seeing the difficulty here. Spacetime being curved is by definition having non-zero curvature. Since curvature is given by a tensor, you can't make it go away. If it's nonzero in some frame, it's nonzero in all frames. (Unlike the connection coefficients, which do not form a tensor, and can be transformed away [edit: locally]... which goes right to Einstein's point about things in freefall locally acting just like STR.)
Because the connection coefficients are not tensorial, identifying the connection with gravity leads to a coordinate-dependent notion of gravity, which goes against the spirit of general relativity. As Einstein wrote for a non-technical audience in the book that
Farsight has been quoting (
Relativity: The Special and the General Theory, italics as in the English translation by Robert W Lawson):
Einstein said:
We refer the four-dimensional space-time continuum in an arbitrary manner to Gauss coordinates. We assign to every point in the continuum (event) four numbers, x1, x2, x3, x4 (coordinates), which have not the least direct physical significance, but only serve the purpose of numbering the points of the continuum in a definite but arbitrary manner. This arrangement does not even need to be of such a kind that we must regard x1, x2, x3, as "space" coordinates and x4 as a "time" coordinate.
...snip...
The following statement corresponds to the fundamental idea of the general principle of relativity: "All Gaussian coordinate systems are essentially equivalent for the formulation of the general laws of nature."
From Einstein's own words, we can infer that identifying gravity with the coordinate-dependent connection is less than ideal. It seems to me that Einstein probably would have preferred to use the more modern, coordinate-independent tensorial notion had the modern tensor notation been available at the time. Einstein used explicit connections because that was the notation of his time. This may be a case of notion following notation.
[size=+1]
Summary of Farsight's errors[/size]
- You were flat-out wrong to deny that curvature is defined by the Riemann curvature tensor and its contractions (the Ricci tensor and scalar curvature).
- You were flat-out wrong to deny the curvature of spacetime in most FLRW solutions, including all "dusty" solutions.
- You were flat-out wrong to deny the existence of gravity in FLRW solutions.
...snip...
The following statements are mathematical:
- A non-zero Ricci tensor implies curvature.
- A non-zero scalar curvature implies a non-zero Ricci tensor, hence curvature.
- Most FLRW solutions, including all "dusty" solutions, imply curvature of the spacetime manifold.
When you deny those statements, you are flat-out wrong.
[size=+1]
Farsight's response to my criticisms[/size]
I will now respond to
Farsight's response to my criticisms.
When did mathematics, backed by citations to standard references, turn into "bible thumping"?
The bible thumping was you saying that I must be wrong because what I was telling you wasn't in agreement with Hawking and Ellis. That is bible thumping Clinger. Recognise it. Don't do it.
I was saying you were wrong because Einstein's field equations and their mathematical consequences (such as the FLRW solutions) don't agree with you. Rather than write out the derivation of those consequences (which would have made my post even longer!), I referred to standard references in which those consequences are worked out in painstaking detail.
As I
mentioned in post #4914, I have found two minor errors in the equations presented by Hawking & Ellis. I suspect there are more, because I still haven't read the entire book and I could have missed some errors even within the equations I studied with some care. If you believe you have found one or more technical errors in Hawking & Ellis that could eliminate the obvious contradictions between your claims and the mathematical consequences of Einstein's field equations, then I urge you to report those errors here.
Until you do so, I will interpret your references to "bible thumping" as a quaint way of admitting you have no substantive criticisms of the mathematics that contradicts you.
The universe isn't a 2-dimensional surface. It's perhaps a "ball", which the common man would call a sphere. Mathematically speaking the sphere is the surface rather than the ball.
Let's speak mathematically. In the FLRW solutions, neither spacetime nor space can be a closed ball. For the solutions in which space is Euclidean, space is homeomorphic but not diffeomorphic to an open ball. For the solutions in which space is a 3-sphere, space is a 3-sphere, and a 3-sphere is not even homeomorphic to a ball.
Lots of people have trouble visualizing a 3-sphere, so I used a 2-sphere for my example. You can criticize my example for being 2-dimensional instead of 3-dimensional, but your criticism of its sphericality is incoherent and suggests (once again) that you are unfamiliar with the FLRW solutions you
introduced into this discussion.
You have a bad habit of arguing as though there's some privileged coordinate system in which your beloved Euclidean space has been carved out of non-Euclidean spacetime, and that your argument automatically generalizes to all other coordinate systems.
There is, that of the rest frame of the CMBR. See
CMBR dipole anisotropy. From this you can gauge your motion through the universe. Whilst this isn't technically an "absolute" reference frame, it doesn't get much more absolute than that.
Okay, I'll accept that. From now on, we can discuss the FLRW solutions using the standard decomposition into space and time that makes space homogeneous and isotropic. In some of the FLRW solutions, the canonical decomposition yields a non-Euclidean space. You have been arguing as though the canonical decomposition always yields a Euclidean space, which is flat-out wrong.
That unpublished arXiv paper supports what I wrote: Curvature is measured by the Riemann curvature tensor and its contracted forms (the Ricci tensor and the scalar curvature)...
Don't be evasive. Look at the last line of the abstract:
The interpretation of gravity as a curvature in space-time is an interpretation Einstein did not agree with.
Yet Einstein chose to express his field equations in terms of the Ricci tensor and scalar curvature. Einstein must have thought the mathematical definition of curvature was more useful than the notion of gravity that
Farsight would attribute to Einstein almost a century later.
According to
Vorpal, who (like several other participants in this thread) knows more about this stuff than you and me combined, Einstein identified the gravitational field with the affine connection. As I explained above, Einstein should not have been comfortable with that identification, and several of his contemporaries had already adopted a more modern point of view by 1920. See for example
- Sir Arthur Eddington. Space, Time, and Gravitation. Cambridge University Press, 1920. (Chapter 5 uses the Riemann tensor, Ricci tensor, and scalar curvature to classify the curvature of spacetime.)
- Hermann Weyl. Raum, Zeit, Materie. 1921. (The title of section 15 is "Curvature", and Weyl uses affine connections and the Riemann tensor throughout the rest of the book.)
All of that's just a historical curiosity, however. The central fact is that the only curvature that appears in Einstein's field equations is the mathematical notion of curvature you have been rejecting. It is also a fact that the mathematical notion of curvature is used to classify the spatial component of an FLRW solution, and you have been rejecting that fact also by pretending that all FLRW solutions have flat space.
The technical error is in FLRW which starts with the assumption of homogeneity and isotropy of space and then allows Σ to range over a 3-dimensional space of uniform curvature, that is, elliptical space, Euclidean space, or hyperbolic space. It has to be Euclidean. Space is flat when it's homogeneous.
Unsupported. Semi-wrong. Flat-out wrong. Flat-out wrong.
Let me unpack that for you:
- "The technical error is in FLRW" part is unsupported because you have been unable to point to any technical errors in the FLRW solutions. Aside from your own bare assertion, we have no evidence of any technical error. When some random guy on the Internet is railing against a mathematical theory that was developed, reviewed, studied, and taught by many experts, and has been understood by generations of mathematicians and physicists, it would be a mistake to bet on the random Internet guy.
- "elliptical space" is semi-wrong because the only elliptical space that's homogeneous and isotropic is a 3-sphere. (Other 3-dimensional elliptical spaces are homeomorphic but not diffeomorphic to a 3-sphere.) You should have said "spherical space".
- "It has to be Euclidean" is flat-out wrong. Spherical space is not Euclidean, yet it yields a family of FLRW solutions. Einstein's steady state universe is an FLRW solution with spherical space, and it's just the most famous example of its type.
- "Space is flat when it's homogeneous" is flat-out wrong on several counts. Spherical space is homogeneous and isotropic but not flat. Elliptical space isn't isotropic, but it's homogeneous without being flat. There are many other examples, including some isotropic examples you forgot to include in your enumeration of FLRW spaces.
Spacetime is only "curved" in an arcane mathematical sense when space is expanding, because the homogeneity of space changes over time even though space in the universe remains homgeneous. Light doesn't bend round full circle and end up where it started.
As demonstrated by FLRW solutions with spherical space, including Einstein's steady state cosmology, space can be curved in the familiar spherical sense (albeit 3-dimensional instead of 2-dimensional) even when it is homogeneous and isotropic. In FLRW solutions with spherical space, including Einstein's steady state cosmology, a photon of light (travelling along a geodesic) does indeed end up where it started (speaking spatially, of course).
No. You, however, are confusing mathematics with subjective opinion.
You are definitely confusing a spacetime manifold with physical space. Space exists, that's what's out there. It's expanding, we have hard scientific evidence that tells us this. A spacetime manifold is an abstraction.
Yes, spacetime manifolds are an abstraction. I said so myself in a
previous post.
I am not, however, confusing a spacetime manifold with physical space. As I explained in that
previous post, the FLRW solutions are models (abstractions) of spacetime. The facts I have stated concerning those models are theorems of mathematics. You have been denying those mathematical facts, apparently because you are elevating your subjective opinions above the mathematical facts.
I'm making a distinction between mathemmatical abstraction and reality.
As am I, except I'm getting that distinction right and you aren't.
Go look at the Ricci curvature tensor, it relates to volume change. The universe is expanding. And in that universe space is homogeneous, at all times. And since a gravitational field is homogeneous space, there is no overall gravity in the universe.
The Ricci tensor is defined on a mathematical abstraction. When you say the universe is expanding, I don't know whether you mean the mathematical abstraction of space is expanding (as in some but not all FLRW models) or that the physical universe is expanding. When you say space is homogeneous in that universe, at all times, I have to assume you're talking about a mathematical abstraction, because the physical universe is definitely not completely homogeneous.
You then insert your subjective opinion ("since a gravitational field is homogeneous space", which probably wasn't even the opinion you intended to express), and conclude that there is no overall gravity in the universe. Since the physical universe in which I live has something that even you might accept as gravity, I have to conclude that your conclusion there was supposed to be about some mathematical model of the universe, not the physical universe. Since you are using the word "gravity" in some subjective sense that you attribute to Einstein, you appear to be making a subjective statement about a mathematical model, which seems pointless.
Let me put it this way: If you want to define "gravity" as a synonym for "homogeneous space" (which is what you wrote above) or "inhomogeneous space" (which is more likely to resemble what you intended to write), then you're free to do so. (Having done so, however, you shouldn't blame anyone but yourself for the confusion that's certain to follow from your redefinition of a technical word that most people (even non-experts!) interpret as something rather different from what you mean by it.) Once you've redefined "gravity" to mean "homogeneous" (or "inhomogeneous") space, then your argument basically boils down to this: All FLRW models have a homogeneous space, so they have "gravity" (in your personal sense above) or don't have "gravity" (in the personal sense you may have meant).
You could have saved yourself and everyone else a bunch of grief by saying that clearly and unambiguously at the outset.
The following statement is your subjective opinion:
- Curvature of the spacetime manifold does not imply gravity.
The universe didn't contract when the matter-energy density was high, now did it.
That's not my subjective opinion.

Are you suggesting that a baseball thrown upward becomes subject to gravity only after its vertical velocity changes sign?
Einstein's static universe, in which space is spherical, is one of the FLRW solutions. Extending the lane indefinitely brings it back to the bowler's feet.
And WMAP has demonstrated that space is flat. So you can discard that solution.
WMAP has not demonstrated that all of the FLRW solutions have flat space. Experimental results cannot disprove mathematical theorems; they can, however, show that some of the mathematical models are not realistic.
You, however, have been denying and continue to deny the mathematical fact that some of the FLRW solutions have spherical (or some other non-flat) space. That's yet another example of your failure to distinguish mathematical models from physical reality.
Spinning a yarn is no substitute for doing mathematics.
You've been spun a yarn, Clinger. Now come on, think it through: start with a high-energy-density universe. It expands, reducing the energy density if you adhere to conservation of energy. What happens next? It expands, reducing the energy density. What happens next? It expands, reducing the energy density.
Spinning a yarn is no substitute for doing mathematics.
The change in energy density over time is given by Hawking & Ellis equation (5.10). As I have explained, the rate of expansion decreases with time. Whether that rate of expansion will ever drop below zero depends upon the initial conditions; the equations that determine whether that will happen are well known, and can be found within standard references.
Your informal argument above is similar to an informal argument that a moon rock thrown upward will never fall back to the moon. In reality, whether the rock falls back to the moon depends upon how hard you throw it: Is its initial velocity above, equal to, or less than its escape velocity?
That's a quantitative question. Referring to quantitative mathematics as "bible thumping" isn't going to advance your understanding of physics.
Like I said, you've been spun a yarn. Come on Clinger, think it through. Think for yourself. Don't let Hawking do your thinking for you.
If I were content to let Hawking think for me, would I have found those two minor errors in his equations?
How many errors have you found,
Farsight? You have yet to identify a single technical error in any "bible" I've cited.