# Cosmological principle and quasar distribution

## Main Question or Discussion Point

Our cosmology model follows the cosmological principle according to wich we are not in a privileged place in the universe and there is homogeneity, but if you take a look at the distribution of quasars in the universe there seems to be a "quasar spherical void" roughly one billion lightyears in radius around us. There seems to be a big flaw in the homogeneity of the universe precisely with center in us.
This doesn't seem to agree very well with the Cosmological principle in wich our cosmology is based together with GR. Is this something to worry about? Or it doesn't really matter?

Jonathan Scott
Gold Member
Our cosmology model follows the cosmological principle according to wich we are not in a privileged place in the universe and there is homogeneity, but if you take a look at the distribution of quasars in the universe there seems to be a "quasar spherical void" roughly one billion lightyears in radius around us. There seems to be a big flaw in the homogeneity of the universe precisely with center in us.
This doesn't seem to agree very well with the Cosmological principle in wich our cosmology is based together with GR. Is this something to worry about? Or it doesn't really matter?
The standard explanation is that this relates to time, in that the conditions for forming new quasars effectively cease after the universe reaches a certain age.

The standard explanation is that this relates to time, in that the conditions for forming new quasars effectively cease after the universe reaches a certain age.
Really? it doesn't sound very convincing, rather looks like an ad hoc justification without any base.
In fact that kind of explanation makes the cosmological principle unfalsifiable, since any deviation from the principle can be blamed on "evolution" causes.

Last edited:
bcrowell
Staff Emeritus
Gold Member
Really? it doesn't sound very convincing, rather looks like an ad hoc justification without any base.
In fact that kind of explanation makes the cosmological principle unfalsifiable, since any deviation from the principle can be blamed on "evolution" causes.
Actually this kind of observation is an observational test that big-bang cosmology passes and that steady-state cosmology fails. If we observed that the universe looked the same at all redshifts, then it would falsify big-bang cosmology and support steady-state.

The cosmological principle is far from unfalsifiable. There is some good material on the topic here: http://en.wikipedia.org/wiki/Cosmological_principle

Staff Emeritus
2019 Award
We know a 100,000 year old universe looks different than a 13.7Gy old universe. Why should a 13.7 Gy universe look like a 10, 8 or 5 Gy old universe?

We know a 100,000 year old universe looks different than a 13.7Gy old universe. Why should a 13.7 Gy universe look like a 10, 8 or 5 Gy old universe?
The Cosmological principle? the Generalized Copernican principle? The very principles FRW cosmology is based on. It all comes down to homogeneity at large spatial distances. People seems to forget selectively that Gly is also a spatial distance measure, not just a time span.
The way you depict it with the universe looking very different at different radial distances from us is against the Cosmological principle among other things because it locates us at the center of that particular evolution and our cosmology is based on the opposite assumption:"no privileged observer".
Homogeneity means that every observer in the universe must observe approximately the same at large distances, don't you see something odd if every observer no matter how near or far from us must see the same evolution (the universe looking a certain way at different radial distances)? Either there is homogeneity or a different evolution for every observer located sufficiently far from other observers. I'm afraid you can't coherently have the two scenarios at the same time.

Jonathan Scott
Gold Member
The Cosmological principle? the Generalized Copernican principle? The very principles FRW cosmology is based on. It all comes down to homogeneity at large spatial distances. People seems to forget selectively that Gly is also a spatial distance measure, not just a time span.
The way you depict it with the universe looking very different at different radial distances from us is against the Cosmological principle among other things because it locates us at the center of that particular evolution and our cosmology is based on the opposite assumption:"no privileged observer".
Homogeneity means that every observer in the universe must observe approximately the same at large distances, don't you see something odd if every observer no matter how near or far from us must see the same evolution (the universe looking a certain way at different radial distances)? Either there is homogeneity or a different evolution for every observer located sufficiently far from other observers. I'm afraid you can't coherently have the two scenarios at the same time.
Yes you can. If quasars died out everywhere around the same time, then everyone existing at the present day will see the same effect of a spherical void, regardless of location.

Yes you can. If quasars died out everywhere around the same time, then everyone existing at the present day will see the same effect of a spherical void, regardless of location.
It is not so easy to define a "same time" and "present day" for distant observers in GR.
If you set the time all quasars happened to die out about 1GY ago from our location, how does that translate to a galaxy 13.2 Gly from us (like the last one discovered recently:see WP), supposing it is still existing at present day from our point of view, that is 13,2 Gy after the light we see left that galaxy. Would you say that if we were observing with our telescopes from that location the quasar would ceased to exist also at a radius of 1Gly? Or would we observe a different evolution?

Last edited:
Chalnoth
Quasars are a result of lots of gas falling into the supermassive blackholes at the centers of galaxies. Once the gas is used up, the quasars die out. So, there were lots of quasars when galaxies were first forming. But now they are far less common because there aren't as many galaxies with enough gas in their cores.

Quasars are a result of lots of gas falling into the supermassive blackholes at the centers of galaxies. Once the gas is used up, the quasars die out. So, there were lots of quasars when galaxies were first forming. But now they are far less common because there aren't as many galaxies with enough gas in their cores.
And I was saying that "now" is not a meaningful global concept in GR, it only has meaning as cosmic time for us as observers who choose a certain preferred frame of reference, but our "now" can not be compared with "now" at cosmological distances,(that is if you use GR as the theory to understand the universe and consider it a curved manifold). So our chronology doesn't have to be the same as the chronology as seen from a distant point from our POV.
So it is easy to see that spatial homogeneity is not a valid concept in a curved manifold for different points sufficiently distant between each other, because their timescales can't be meaningfully compared.
There is no way you can claim (if you know something about relativity) that all quasars died out approximately at the "same time" for all possible observers in a curved spacetime universe. But a cosmological principle that includes spatial homogeneity demands that from any point we should see approximately the same matter distribution (no privileged observer), you seem to imply that from any point in the universe the same matter distributions at different distances should be observed, and I'm saying that that would demand an absolute time, not just a congruence for our POV in our preferred frame.
If there is some specific point where I'm wrong in what I'm saying I would like it to be explained to me.

Last edited:
Jonathan Scott
Gold Member
And I was saying that "now" is not a meaningful global concept in GR, it only has meaning as cosmic time for us as observers who choose a certain preferred frame of reference, but our "now" can not be compared with "now" at cosmological distances,(that is if you use GR as the theory to understand the universe and consider it a curved manifold). So our chronology doesn't have to be the same as the chronology as seen from a distant point from our POV.
So it is easy to see that spatial homogeneity is not a valid concept in a curved manifold for different points sufficiently distant between each other, because their timescales can't be meaningfully compared.
There is no way you can claim (if you know something about relativity) that all quasars died out approximately at the "same time" for all possible observers in a curved spacetime universe. But a cosmological principle that includes spatial homogeneity demands that from any point we should see approximately the same matter distribution (no privileged observer), you seem to imply that from any point in the universe the same matter distributions at different distances should be observed, and I'm saying that that would demand an absolute time, not just a congruence for our POV in our preferred frame.
If there is some specific point where I'm wrong in what I'm saying I would like it to be explained to me.
On the cosmological scale there is an approximate rest frame everywhere, which is roughly the average frame of the local galaxies. If there were not, galaxy redshifts would not be any use as distance indicators. That means that there is effectively a global time to that level of accuracy.

Jonathan Scott
Gold Member
I must admit I agree that it seems a bit odd that quasars just happened to be doing fine since the start of the universe - or apparently even before(!), according to some results from very high redshift quasar metallicity values - but then all ran out everywhere quite recently, given how non-uniform their properties seem in other ways (with unexpectedly weak correlations between brightness, redshifts and spectral characteristics).

However, it doesn't violate the cosmological principle.

Staff Emeritus
2019 Award
The Cosmological Principle does not say that the universe looks the same at all times.

On the cosmological scale there is an approximate rest frame everywhere, which is roughly the average frame of the local galaxies. If there were not, galaxy redshifts would not be any use as distance indicators. That means that there is effectively a global time to that level of accuracy.
Yes, there is a rest frame, but it is claimed that is just a consequence of a coordinate choice, according to GR we should be able to use a different coordinate choice, with different cosmic time and we would still be able to use galaxy redshift as useful distance indicators from our location (remember redshift is a coordinate-independent measure).
So there is not really a global time that is the same for every distant location in the universe, we can assign a cosmic time for our particular observer position, but that doesn't guarantee you that from a distant position observers share the same "global time", that's impossible to determine in GR because in a curved manifold parallel transport with a connection permits connecting the geometries of nearby points only, it is a local procedure, unlike it is possible in flat spacetime, it doesn't work that way for points separated cosmological distances, where ambiguity is introduced thru path-dependence of the connection. But this connection is defined locally, in order to have a "global time" we would have to have a "global connection, no such thing in GR, it's not possible to define a global connection because in general parallel transport is path dependent, the only spaces with global connections are those with no curvature. If there were really a "global time" we could unambiguously define distant velocities in cosmology but as you might know we can't.
The Cosmological Principle does not say that the universe looks the same at all times.
You're right, it doesn't. From WP: "The cosmological principle is usually stated formally as 'Viewed on a sufficiently large scale, the properties of the Universe are the same for all observers.'... Here "observers" means any observer at any location in the universe...The two testable structural consequences of the cosmological principle are homogeneity and isotropy. Homogeneity means that the same observational evidence is available to observers at different locations in the universe ("the part of the Universe which we can see is a fair sample").
So it doesn't specifically say the universe looks the same at all times, but it does say it looks approximately the same from all locations, right? Now, in a curved spacetime locations separated by large distances don't share the same "global time" as explained above. So if you impose that they should observe fairly the same, you are imposing it for whatever times every location is at. So in an indirect way you are either giving a meaningless homogeneity concept for a curved manifold, or indirectly saying the universe looks the same at all times.

Chalnoth
And I was saying that "now" is not a meaningful global concept in GR, it only has meaning as cosmic time for us as observers who choose a certain preferred frame of reference, but our "now" can not be compared with "now" at cosmological distances,(that is if you use GR as the theory to understand the universe and consider it a curved manifold). So our chronology doesn't have to be the same as the chronology as seen from a distant point from our POV.
Just define the global time slicing as one in which the CMB temperature is isotropic and the same temperature everywhere. With that definition, what I wrote holds.

Just define the global time slicing as one in which the CMB temperature is isotropic and the same temperature everywhere. With that definition, what I wrote holds.
As i explained in my previous post that time slicing is a local one for each distant location in the universe that allows to have a cosmic time from the limit of the observable universe from each location but that is valid only for that location because in a curved manifold you can't have such global time, that is only possible in a Newtonian univere or in a Minkowski spacetime because both are flat.

Chalnoth
As i explained in my previous post that time slicing is a local one for each distant location in the universe that allows to have a cosmic time from the limit of the observable universe from each location but that is valid only for that location because in a curved manifold you can't have such global time, that is only possible in a Newtonian univere or in a Minkowski spacetime because both are flat.
No, it's quite global. It is arbitrary, in that there are many other potential choices, but that doesn't mean it isn't a global time slicing.

No, it's quite global. It is arbitrary, in that there are many other potential choices, but that doesn't mean it isn't a global time slicing.
It's global for us, wich is not the same as being globally valid for any observer in the universe, people keep mixing properties of flat spaces with those of curved manifolds. And in any case it can't be checked experimentally, so we must go by what the current theory (GR) and differential geometry say, and that is that in a curved manifold there is no way to unambiguously compare cosmic times from distant points in the universe.

Chalnoth
It's global for us, wich is not the same as being globally valid for any observer in the universe, people keep mixing properties of flat spaces with those of curved manifolds. And in any case it can't be checked experimentally, so we must go by what the current theory (GR) and differential geometry say, and that is that in a curved manifold there is no way to unambiguously compare cosmic times from distant points in the universe.
You're not getting me. It's an arbitrary choice, but a perfectly valid one. And in this case it is a choice that would make sense to any sitting on any planet on any galaxy in our universe, because it is directly related to the proper time since the CMB was emitted. It wouldn't be a terribly useful time slicing for an observer moving through our universe with relativistic speed, but then we don't care much about those observers for the purpose of this kind of discussion.

You're not getting me. It's an arbitrary choice, but a perfectly valid one. And in this case it is a choice that would make sense to any sitting on any planet on any galaxy in our universe,
This is perfectly ok, I guess you are missing my point too.

because it is directly related to the proper time since the CMB was emitted.
Proper time depends on the state of motion of the clocks that measure it and I've been saying all along that in GR this can't be ascertained without ambiguity, so you can't compare proper times of points of the universe separated by large distances.

So I'm still not finding arguments to make compatible the peculiar quasar distribution with the cosmological principle.

Chalnoth
Proper time depends on the state of motion of the clocks that measure it
Yes, but due to the fact that the expansion of the universe tends to dampen motions, this point is irrelevant, so that as far as time dilation is concerned you can effectively consider everything to be stationary with respect to the CMB.

Now, if you want to consider some hypothetical observer moving at relativistic velocities with respect to the CMB, that's fine. But it doesn't apply to any galaxy in the universe.

Edit: Just to put a few numbers on this, our motion with respect to the CMB is about $v/c = 0.001$. In the densest of clusters, you might get up to around $v/c = 0.01$, but that is a rarity. That would cause a time dilation of 0.005% with respect to the global time reference frame of the CMB. Certainly measurable if you were careful about it, but irrelevant for nearly all purposes.

Last edited:
The Standard Model depends on the CP (and GR) and although the CP isn't "unfalsifiable" it's also unprovable.

The quasar issue isn't the only problem for the CP. The quadrupole and octupole modes of the WMAP that seem to orient along the ecliptic plane is a problem. The Pioneer anomaly is a problem. Inflation being in tatters doesn't help.

I know of one paper that provides data regarding concentric rings at intervals radiating out... from here, home.

A bounded, finite Universe with a gravitational center would explain a lot of things, and it doesn't require an unprovable CP. But nobody is brave enough to touch it, because of the implications.

Chalnoth
The Standard Model depends on the CP (and GR) and although the CP isn't "unfalsifiable" it's also unprovable.
Er, what? It is only "unprovable" in the sense that all science is unprovable: proof is something that can never be done in science, since all of science relies upon inductive reasoning which is impossible to ever prove.

The quasar issue isn't the only problem for the CP. The quadrupole and octupole modes of the WMAP that seem to orient along the ecliptic plane is a problem. The Pioneer anomaly is a problem. Inflation being in tatters doesn't help.
The quadrupole and octupole modes of WMAP are well within the expected statistical bounds. See here: http://arxiv.org/abs/1001.4758
They show that all of the "anomalies" seen in the WMAP data are just down to not doing the statistics right.

The Pioneer anomaly is just due to the spacecraft itself and has nothing to do with fundamental physics. The prime suspect is the asymmetrical radiation from the Pioneer probe.

I know of one paper that provides data regarding concentric rings at intervals radiating out... from here, home.
Which paper would this be?

A bounded, finite Universe with a gravitational center would explain a lot of things, and it doesn't require an unprovable CP. But nobody is brave enough to touch it, because of the implications.
A bounded, finite universe with a gravitational center would actually explain precisely nothing about our observations. If anything, our observations were pointing in the other direction, with one of the potential proposed explanations for the accelerated expansion being that we live in an especially underdense region of the universe. However, this view has since been ruled out by observations.

Yes, but due to the fact that the expansion of the universe tends to dampen motions, this point is irrelevant, so that as far as time dilation is concerned you can effectively consider everything to be stationary with respect to the CMB.

Now, if you want to consider some hypothetical observer moving at relativistic velocities with respect to the CMB, that's fine. But it doesn't apply to any galaxy in the universe.

Edit: Just to put a few numbers on this, our motion with respect to the CMB is about $v/c = 0.001$. In the densest of clusters, you might get up to around $v/c = 0.01$, but that is a rarity. That would cause a time dilation of 0.005% with respect to the global time reference frame of the CMB. Certainly measurable if you were careful about it, but irrelevant for nearly all purposes.
This seems a bit confusing. You say that expansion "dampens motions", it's the first time I hear it put like that, I thought expansion is exacly the opposite, I mean the reason people in the past thought our universe was static was precisely for the reasons you are giving about the motion of galaxies.
Where you don't seem to realize your argument is contradictory in that if the CMB is a global time reference frame valid for all locations in the universe, and if you at the same time claim all galaxies have irrelevant motions wrt that global frame,because we "can effectively consider everything to be stationary with respect to the CMB", you are directly shooting down expansion.
Besides you can't assert that a frame is arbitrary and globally valid for all points in the universe at the same time. An arbitrary choice of frame is perfectly valid for any location, but if all locations share the CMB as a valid rest frame, then you are imposing an absolute frame in GR, I'm not sure this is possible. You can't have your cake and eat it too. Either the CMB is just a local arbitrary valid reference frame, or it's a global reference frame, in this case global means absolute, because if a reference frame can be shared by all points in a manifold so that as you are doing in your post you can assign proper velocities to galaxies located at any point in the manifold based on a shared referenced frame (CMB), you are basically describing a Minkowski spacetime.
So I don't know if you are really sure about your arguments or you are not taking very seriously this discussion.

Chalnoth